- Email: [email protected]
Stimulated prostaglandin E2 release from rat skin, in vitro
Life
ELSEVIER
STIMULATED
PII SOO24-3205(98)00176-3
PROSTAGLANDIN
sciem vol. at, No. 22, pp. 2045-1055,1998 ckpyeght 0 1958ElSnkr !3cimce Inc. Printedin the USA. All @Its relend 0024-~/!xl$19.00 t .m
E2 RELEASE FROM RAT SKIN, IN VITRO
S.K. Sauer IXI, D. Schafer #, M. Kress, P.W. Reeh Institut fur Physiologie und Experimentelle Pathophysiologie IXI, Universititstr. Institut fur Klinische Immunologie #, Universitatsstr. 29, Universitlt Erlangen-Ntimberg, D-91054 Erlangen
17;
(Received in final form March 4,1998)
Summary The excitatory effect of bradykinin (BK) and of low pH on nociceptors appears to partly depend on secondary release of prostaglandins from the surrounding tissue. Rat skin, in vitro, is introduced as a novel model to measure basal and stimulated release of PGE, and, in future, other substances relevant to nociception, such as neuropeptides. Flaps of hairy skin (n=57) from the rat saphenous region of the hindpaw were subcutaneously excised and fixed on acrylic rods, the corium side exposed. The preparations were equilibrated in carbogen gassed ,,synthetic interstitial fluid“ (SIF) for 30 minutes. The skin flaps were then immersed for 5 minutes each in 9 consecutive glass tubes, which were mounted in a shaking bath at 32°C. Each tube was filled with 5 ml of gassed SIF, the third tube contained inflammatory mediator(s) dissolved in SIF or solutions of low pH. After passage of the skin flap, the eluates were deep frozen (-70°C) and the PGE z content measured, off-line, using an enzyme immuno-assay. As stimulants, BK at 10m5M (n=9) and 1u6 M (n=4) and BK in equimolar combination with histamine (HA) and serotonin (5HT; lo5 M: n=8, 10e6 M: n=6, lo-’ M: n=6) dose-dependently increased PGEz release. Considering the total amount of PGE, secreted the combination of inflammatory mediators caused a significantly greater release of PGE, at lo5 and 10s6 M (~~0.01, Kruskal-Wallis test) than BK stimulation alone. Racemic flurbiprofen caused a profound depression of basal and stimulated release. Solutions of high proton concentration are known to stimulate and sensitize nociceptors. However, phosphate buffered SIF at pH 6.1 and 6.4 caused a substantial and significant decrease of the PGE2 release, probably due to low-pH block of phospholipases. Thus, algogenic potency of mediators does not necessarily match their pro-inflammatory action. Key Words: nociception, inflammation, bradykinh, histamine, serotonin, protons
Address for Correspondence: Dept. of Physiology I., Universitatsstr. Germany; phone +49-9131-856729, fax +49-9131-852497, e-mail erlangen.de
17, D-91054 Erlangen, [email protected]
Chdcally-evoked PGQ Release from Rat Skin
Vol. 62, No. 22, PBS
Introduction The algogenic effects of bradykinin (BK) seem to depend partly on a sensitizing action of prostaglandin E2 (PGIQ, that is released in many tissues in response to BK (1). Positive feedback is thought to result from this interaction, increasing the nociceptor exciting effect of BK. This theory is supported by suppressive effects of prostaglandin-synthesis inhibitors on responses of nociceptors to stimulation with BK (2, 3) and, further, by the effect of exogenous PGE, which can enhance BK-induced excitation in some preparations (4, 5, 6). However, dependence of BK responses on PGE, is not consistently found: In the neonatal rat tail-spinal of BK-induced reflex responses by non-steroidal cord preparation, the suppression antiinflammatory drugs (NSAIDs) could only be seen at low concentrations of BK (3). In the guinea pig trachea preparation BK responses of vagal afferents could not be suppressed by ibuprofen (7). Species differences could also be of importance, since effects which were obtained in knee joint afferents of the cat (6) could not be reproduced in ankle joint afferents of the rat (8). In behavioral experiments, NSAIDs could not reduce BK-induced writhing in mice (9) which is in contrast to BK-induced hyperalgesia in rats (10). In particular in the rat skin, in vitro, BK-induced nociceptor responses are not augmented by PGE, nor reduced by diclofenac (11). Blockade of the PGE, synthesis and substitution of exogenous PGE, did not exert influence on the nociceptor responses to BK in a combination of inflammatory mediators in normal and inflamed skin (12). In this preparation, like in many others, it is serotonin (5-HT) that facilitates the excitatory effect of BK which is of importance for the interaction of inflammatory mediators in inducing pain (11, 13). Low pH as in tissue acidosis is another relevant cofactor in inflammatory pain (14). NSAlDs have been shown to effectively suppress pain from experimental acidosis in human skin (15, 16). It is not known whether inhibition of prostaglandin synthesis is involved in this action or not. Again, low pH could stimulate PGE, release, as in endothelial cells (17), and by that, enhance the stimulatory effect of protons on nociceptors. The controversial findings raise the question whether rat skin, in vitro, is at all able to release PGE, in response to BK or low pH, and whether a possible effect could be facilitated by additional synergistic inflammatory mediators. Therefore, we developed a release model of isolated rat skin and measured PGE, liberation into the eluation fluid. PG synthesis has previously been measured in isolated cell preparations and bioptates from various tissues. Here, the skin was chosen since it is the organ on which by far most of the electrophysiological, pharmacological and behavioral pain studies were conducted but hardly any biochemical ones. Hence, we hope to fill a gap between the disciplines with correlative data. For the same reason, isolated skin preparations have recently been employed to study release of substance P and calcitonin-gene related peptide from cutaneous nerve endings (18, Averbeck and Reeh, submitted (1998); Kress, Guthmann, Averbeck and Reeh, submitted (1998)). Some of the present results have previously been communicated in abstract form (19). Methods For this novel release model we used skin flaps of the lower hindpaw of adult male Wistar rats. After sacrificing the animals in a container filled with pure CO, atmosphere, the hairy skin of both hindpaws was subcutaneously excised sparing the bigger vessels and the saphenous and sural nerves. The preparation started at knee level and included the toes. F rom 29 rats skin flaps (n=57) were obtained with a mean weight of 1.08 2 0.22 g (range 1.95 g- 0.66 g). After dissection, the flaps were wrapped around arylic glass rods (diam. 6 mm) with the corium side
Vol. 62, No. 22, 1998
Chemically-evoked PGQ Release from Rat Skin
2047
exposed and fixed with threads. The preparations were washed for 30 minutes on ice in 250 ml “synthetic interstitial fluid” at pH 7.4 (SIF containing (mM) 108 NaCI, 3.48 KCI, 3.5 MgS04 26 NaHCO,, 11.7 NaHzP04, 1.5 CaCl,, 9.6 Na-gluconate, 5.55 glucose, 7.6 sucrose). The cooling of the preparation during the initial washout was meant to suppress induction of PGE2 synthesis due to the surgical procedure. Subsequently, the skin flaps were placed in nine consecutive glass tubes filled with 5 ml SIF previously bubbled with carbogen (95% OZ, 5% CO,). These tubes were mounted in a shaking bath at 32 “C. After five minutes of incubation, the preparations were forwarded to the next tube, and the remaining eluates were deep frozen at -70°C. The third incubation step always contained the stimulating solution (see below). The amounts of PGE, released into the eluates were quantified in duplicate using a competitive enzyme immuno assay (EIA) which has been described previously (20). Assay procedures as well as intra-and interassay validation have recently been published (21). In brief, PGE z was coupled to bovine serum albumin (BSA), which was then used to coat the wells of microtiter plates at pH 9.6 for 18 hours (200 @/well). After this and every following incubation step the plates were washed using PBS-Tween 0.1 % (v/v). Unspecific binding sites in the wells were blocked with 1% BSA in PBS for 60 minutes. Standard curves were assessed using seven different concentrations of PGE, (1 to 500 pg/well) in duplicate. Samples and standards (100 pg/well) were incubated with a monoclonal mouse antibody against PGE, (mab E2Rl; 100 ul/weIl - a generous gift of K. Brune, Dept. Pharmacol., University Erlangen, Germany) at 4°C for 18 hours. Thereafter the biotinylated goat-anti-mouse antibody was added (200 @/well for 2 hours at 37°C). Following this, the EIA was incubated with the streptavidin-biotin complex coupled to horseradish peroxidase (HRP) for 1 hour at 37 “C (200 p,l/well). As a chromogenic substrate we used ABTS (2,2’-azino-bis [3 ethylbenzthiazoline-6-sulfate]) for 45 minutes at room temperature (200 @/well). The absorbance was determined at 405 nm (reference filter 490 nm). A recovery rate of 98% for PGE2 characterized the assay. Detection limit of the PGE z-EIA was 3 pg/well. The intra- and interassay coefficients of variation (cv) for PGE 2 were 5.6% and 8.1%, respectively. To determine the viability of the assay under low pH conditions, the acidic samples collected were measured using standards at low pH (6.1 and 6.4). The validation assays performed for low pH standards resulted in intra- and interassay coefftcients of variation (cv) of 8.9% and 9.7% for pH 6.1, 8.1% and 9.2% for pH 6.4, which were larger than in standard assays at neutral pH. Searching for a possible reason we found that the recovery rates for PGE z were systematically lower than normal in acidic samples that had been frozen and thawed, whereas the recovery of PGE, in freshly prepared acidic standard solutions was 98% as above. To account for this loss of PGE, in the thawed acidic samples and to prevent from underestimating the real PGE, content, we determined a mean correcting factor of 1.24 for solutions of pH 6.1 and of 1.05 for pH 6.4 with which the particular PGE2 values were multiplied before entering further data analysis. PGE, values (pg) are given per 1 ml of the eluation fluid SIF and referred to lg fresh-weight cutaneous tissue.
of
For chemical stimulation we used the following solutions, all of them diluted in SIF, aerated with carbogen (95% O2 and 5% COZ) and applied at 32 “C. . Bradykinin (BK-triacetate, and pH 7.4.
Sigma) was used at 10m5M (n=9) and 10m6M (n=4) concentration
2048
Chemically-evoked PC% Release from Rat Skin
Vol. 62, No, 22, 1998
An equimolar combination of inflammatory mediators (IM) which was composed of BK, histamine (HA) and serotonin (5HT) (all from Sigma) at pH 7.4 and applied at 10 -5 M (n=8), 10m6M (n=6) and 10m7M (n=6). BK (n=4) and IM (n=4), each at 1O-5 M, were used to stimulate the skin treated with flurbiprofen. Racemic flurbiprofen (a generous gift of G. Geisslinger, Dept. Pharmacol., University Erlangen, Germany) was used to block the endogenous prostaglandin synthesis; the rats received a single dose of 25 mg/kg i.p. one hour before dissection. Additionally, we added flurbiprofen at 10m6M concentration to the SIF solution in each test tube. Stimulation solutions of low pH were prepared using phosphate buffered SIF. The bicarbonate content of SIF was replaced by phosphate buffer in proportions to reach pH values of 6.4 (n=8) and 6.1 (n=8), which were fine adjusted using drops of HCl or NaOH. 1500
Pg 1250
n=4
\
-----
250
lmr 0
0
5
10 15 20 on ice
25
30
I
35
40
45
50
55
60
65
70
75 min
at 32°C
Fig. I Release of PGE2 from the rat isolated skin under resting conditions, i.e. without any chemical stimulation. Note the temperature effect upon transferal of the preparations into the warmer glass tubes after the pre-wash period, on ice. The amount of PGE 2 measured is referred to lg fresh weight of cutaneous tissue in 1 ml eluation fluid. Dafa analysis: Results concerning groups of identical experiments are given as means + s.e.m. The numbers of experiments (n) in the figures refer to skin flaps from different animals. For calculation of the total amount of stimulated PGE, release, values comprising samples three to six were pooled. Statistical non-parametric comparisons within one group of experiments were made using the Wilcoxon matched pairs test which compared PGE, values before and during stimulation. Comparisons between different experimental groups were performed using the nonparametric analysis of variance (Kruskal-Wallis test) and U-tests (Mann-Whitney). Values of p < 0.05 were considered significant.
VoL tX$ No. 22,19%
Chcmkally-cvoked
l’G% Release fromRat Skin
Results
The cutaneous release of PGE, was remarkably influenced by the temperature step of more than 20 “C after the initial 30 minutes pre-wash period (see methods). The temperature rise induced substantial increase and interindividual variability of the PGEz release which was addressed in a separate series of experiments (n=4; fig 1). During the initial pre-wash period at about lO”C, PGE, values varied between 100 and 300 pg and were increased by more than 4 times when the preparations were transferred into SIF at 32°C. After an initial overshoot in this group of skin flaps (n=4; peak values 890 2 300 pg), however, relatively stable intraindividual PGE2 levels were reached and maintained over the subsequent 40 minutes averaging to 456.08 f 34.38 pg (n=4 x 8 consecutive values pooled). In the main study, PGE, levels measured in the first two incubation steps were taken for baseline: average values of 421.5 f 46.8 pg (2 sem) in the first and 433.7 + 52.5 pg in the second sample (n=49) were found. These baseline values match the values found in the second phase of the temperature experiment. Nonetheless, we have drawn the consequence for future studies to perform the initial pre-wash and equilibration of the skin preparation at a temperature of 32°C. release induced by inflammatory mediators: The stimulation with BK at 10q5 M concentration significantly increased the PGE2 level by more than three times in the sample containing BK as compared to the preceeding control sample (fig. 2). BK at 10M6M yielded a smaller increase of PGE, baseline which did just not reach significance (probably due to n=4). At the high BK concentration complete and significant recovery was reached only in the second wash-out step after BK. Flurbiprofen (pre-) treatment (25 mg/kg i.p. and 10 -6 M in the eluation solution) almost abolished basal and PGE, release stimulated by BK 10M5 M.
PGE2
1500 PB BK lo’:M, (n=9) : ;z ;$+ (n=4)
1250
l
1000
+flurbiprofen,
(~4)
P 4 2
750 a: p=O.O4vs. bl. b: p=O.O6vs. bl. c: p=O.O05vs. BK 105M
c-4
g
500 250 0 I ;
Ib
1;
;o
;5
3b
3b
40
45 min
Fig. 2 BK (1V5 and 1N6 M) stimulation for 5 minutes (grey area) induced transiently enhanced PGE, release from the rat skin. Flurbiprofen, a cyclooxygenase blocker, was administered to the alive rats and contained in the incubation fluid (25 mg/kg i.p.; 1U6 M in SIF); it completely attenuated PGE, secretion.
2050
ChemicalIywked
PGQ Relesse from Rat Skin
Vol. 62,No. 22,1998
The combination of inflammatory mediators BK, HA and 5-HT (,,inflammatory soup “, IM) was applied at three different concentrations and caused a protracted release of PGE, (fig. 3). IM at lo9 M concentration significantly augmented the average release of PGE z by a factor of 3.7; IM at low6M concentration was also significantly (p=O.O2) effective with a factor of 2.6. Stimulation with IM at lo-’ M concentration evoked an increase of the PGE 2 release by a factor of two which just did not reach level of significance as compared to baseline (p= 0.07, n=6). Again, flurbiprofen treatment of the animals and preparations resulted in a complete depression of the basal and stimulated prostaglandin synthesis as tested with stimulation by IM at 10.’ M concentration. For comparison of the releasing efficacies of BK and of the combination of inflammatory mediators, we calculated the total amount of stimulated PGE2 release during and after stimulation until baseline was reached again (comprising samples 3 to 6; fig. 4). Bradykinin and the combination of inflammatory mediators increased the total PGE, release dose-dependently (p=O.O3 for BK and p=O.O4 for IM-stimulation, Kruskal-Wallis test). The pooling of the data revealed a “ceiling“ effect of the increasing IM concentration at 10e6 M. In contrast to stimulation with BK, the inflammatory mediators at 10e5 M concentration did not significantly cause further enhancement of the total PGE 2 release.
A l n l
IM 10.‘M, (n=8) IM 10-6M,(s6) IM 10.‘M, (n=6) IM 1n5M +flurbiprofen, (n=4)
a: p=O.O3vs. bl. b: p=O.O2vs. bl.
0
0
I
r
5
10
I
15
I
I
1
I
1
20
25
30
35
40
1 45
Fig. 3 The combination of inflammatory mediators (IM: equimolar BK, HA and 5-HT) evoked massive PGE, secretion from the rat skin, and the effects outlasted the period of chemical stimulation (grey area). Again flurbiprofen (25 mg/kg i.p.; 10 -6 M in SIF) completely blocked basal and stimulated PGE, release. The combination of inflammatory mediators was, however, significantly more effective than BK alone at the two equimolar concentrations (fig. 4). IM 1O-7 M was about equipotent to 10e5 M BK, i.e. addition of 5-HT and HA to BK caused a hundredfold increase in potency. Flurbiprofen treatment significantly suppressed total PGE, release after BK and IM (10m5 M) stimulation (p=O.OOSand p=O.O06, Kruskal-Wallis test).
Chemidyevokcd
Vol. 62, No. 22, 19!38
PGE$ Rekase from Rat Skin
2051
PGE2 release reduced by solutions of low pH: To verify whether another potent stimulus, that
excites and sensitizes nociceptors, is also effective to release prostaglandins, we used solutions of low pH. Solutions of pH 6.1 and 6.4, phosphate buffered SIF, produced profound and outlasting suppression of the PGE, release (fig. 5). Exposure to pH 6.4 caused a more transient but equally strong suppression as pH 6.1 with rapid recovery within five minutes. PGE, remained significantly below baseline for 20 min after exposure to pH 6.1. The relative supression was about the same and not significantly different comparing pH 6.1 and pH 6.4 application. Also, bicarbonte buffered SIF bubbled with pure CO 2 to create a pH of 6.1 caused a marked decrease of the mean PGE, release (p=O.O7,n=4; data not shown). 3500 Pl3
A
3000
T
2500
I
p=o.o4
I
2000 .-a
2 1500
p=o.o3 I
1
BK 1 o”M n=9
BK 10-6M n=4
BK 1O-‘M +fl n=4
IM 1 O-“M n=8
IM I OdM n=6
IM 1O-‘M n=6
IM ‘Z-Y n=4
Fig. 4 Summary showing total amounts of PGE2 released during and after chemical stimulations, pooling data from samples 3-6 (= lo-30 min). The combination of inflammatory mediators (IM) was significantly more effective than BK stimulation at comparable concentrations (p=O.OOOl,Kruskal-Wallis test). Discussion
The novel release model of the rat skin, in vitro, is suitable to study mediator release - here PGE 2 - in response to chemical stimulation. Stimulation with BK alone and, up to six-fold more effective, with an equimolar combination of inflammatory mediators (IM: bradykinin, BK, histamine, HA, serotonin, .5-HT) significantly increased PGE, release from the isolated skin. The treatment with a cyclooxygenase blocker, flurbiprofen, caused a complete suppression of basal and stimulated release. Surprisingly, solutions of low pH, which potently excite and sensitize nociceptors, induced a profound suppression of the prostaglandin release. In the rat hairy skin preparation different types of cells may be sources of PGE 2: keratinocytes, tibroblasts, melanocytes and Langerhans cells in the superficial epidermal layers and fibroblasts, endothelial cells, mast cells and macrophages in deeper layers - the corium. Since the corium side of the skin was exposed to chemical stimulation (see methods), this tissue may contribute the major part of the PGE2 released, although the lymphatic clefts of the corium allow for rapid fluid convection reaching the epidermal border. Also postganglionic efferents of the sympathetic
Chemically-evoked PC+
2052
Vol. 62, No. Z&l998
Release from Rat Skin
nervous system to the skin are suggested to be involved in the prostanoid chemical stimulation (22).
release after noxious
Arachidonic acid provides the main source for lipid mediators such as prostaglandins, leukotrienes, thromboxanes and platelet activating factor. The hydrolysis of membrane phospholipids to arachidonate by phospholipases is thought to be rate-limiting in the prostanoid cascade. Phospholipase A2 (PLA,) is represented by a multifarious group of enzymes, secretory extracellular forms and intracellular ones (23). 900 -) n
pH 6.1 pb, (n=8) pH 6.4 pb, (n=8)
0
200 a: p=O.Ol vs. bl. b: p=O.Ol vs. bl. 5
IO
15
20
25
30
35
40
45
Fig. 5 Exposure of the skin to solutions of low pH (grey area) depressed PGE, release significantly. Phosphate buffered SIF at pH 6.1 produced profound and outlasting suppression of the PGE, release; pH 6.4 was equally effective but more transient in action. Rise of the intracellular Ca++ concentration is a common principle for PLAZ activation and translocation to the plasma membrane. This is notable because many of the inflammatory mediators (BK, HA, 5-HT, protons) are reported to cause an increase of the intracellular Ca ++ concentration and, following this, are potential stimulators of PLA, (24, 25, 26, 27, 28). The successive step in prostaglandin synthesis is the conversion of arachidonate to PGG I and PGH,, which is catalyzed by the cyclooxygenases (COX I and II) (for review see (29)). Bradykinin is one of the most potent endogenous algogens; it excites nociceptive primary afferents and causes pain (for review see (30)). BK has been shown to release prostaglandins from different tissues on different routes of application (1, 31). In addition, BK sensitizes nociceptive afferents to physical stimuli, and this has been suggested to be mediated through release of prostaglandins (3, 2, 5). However, 10s8 M concentration of BK is sufficient to induce the sensitizing effect, whereas PGE2 needs to be applied at least in 10s5 M concentration to induce nociceptor sensitization to heat (32). In our experimental setup, BK dose-dependently evoked PGE, secretion at 10e6 M and 10m5M concentration. This fits well to BK concentrations effective to stimulate prostaglandin release and to excite nociceptors as reported from previous work (1, 31). The rapid accumulation of PGE, in five minutes induced by relatively short lasting stimulation with inflammatory mediators is in agreement with previous findings (29). For
Vol. 62, No. 22, 19!%
Chemically-evoked PGQ Release from Rat Skin
2053
cultured fibroblasts a half maximal release was achieved after two minutes of BK stimulation at tom6M concentration whereas interleukins and phorbol ester caused a more delayed but sustained PG release (31). Histamine-induced prostaglandin release has been shown in many different tissues and cell types (33, 34). This release is mediated by the Hl receptor in membranes of keratinocytes, myocytes and endothelial cells (35, 36). Serotonin plays an important role in inflammation and in pain mechanisms (37). So far, seven different receptors for 5-HT have been isolated (for review see (26)) which partly facilitate the activation of PLA, due to their second messenger pathways (DAG/IP,). A direct activation of the PLAZ has - to our knowledge - not been reported. The combination of inflammatory mediators, BK, HA and 5-HT, resulted in further enhancement of the PGEz release as compared to BK stimulation (see fig 4). In electrophysiological experiments, the combination of inflammatory mediators activated about 30% more of the primary nociceptive afferents in rat skin than BK alone, and the excitatory effect was much stronger (11, 13). The enhanced responsiveness to IM of the cutaneous nociceptors is due to mutually sensitizing and additive effects of the combined inflammatory mediators, to which the endogenous prostaglandins, triggered for example by BK, did not contribute: The mediator combination (BK, 5-HT, HA) had the same impact on nociceptors in the skin whether or not the cyclooxygenase was blocked (by flurbiprofen) and PGE, (10” M) was added (12). In agreement with this, the IM-induced nociceptor excitation significantly increased from 1O‘6 M to 10e5 M concentration (Reischl & Reeh., unpublished), while the induced PGE2 release in the present work stagnated. The stimulation of the rat skin with phosphate-buffered solutions of pH 6.1 and 6.4 significantly decreased PGE, release under basal levels. This is at first surprising since in inflamed tissues, pH values down to 5.4 have been measured (38, 39), and high PGE, levels have been found (29). But pH and prostaglandin values have never been measured simultaneously in one and the same study. The decrease of PGE, release induced by exposure to low pH fits well with enzymatic properties of the PLA,. Several isoforms of PI..& are reported to have pH optima for their enzymatic activity in the alkaline range, and this is true for secreted as well as cytosolic forms and whether or not the iso-enzyme is Cat+ dependent (23). Changes of the extracellular pH were effective to induce an increased formation of IP 3 and a Cat+ peak in fibroblasts, which effects were remarkably similar to those of BK (10 -’ M) stimulation. The authors suppose the same intracellular pathway and pool of Ca ++ to be mobilized (40). Also, in capsaicin-sensitive sensory neurons, a putative cellular model of nociception, low-pH stimulation evoked an influx of Cat+ which by itself may not be sufficient to activate PLA, (28). In contrast, in endothelial and smooth muscle cells low pH (pH 6.8) stimulation produced an indomethacin (O.lpM) sensitive release of PGE, although low pH did not change the intracellular Cat+ concentrations or induce Cat’ influx in these experiments (17). The authors assume that intracellular pH changes stimulated PG synthesis by activating a Ca ++independent PLA,. Thus, in the complex cutaneous tissue relative inhibition of the PLA2 by low pH seems to prevail. This may be overcome in case of inflammation when mediators strongly boost intracellular Cat+ and activate the PLA, and, thus, allow for coexistence of high PG and proton concentrations. Solutions of low pH potently excite and sensitize nociceptive primary afferents in the skin-nerve preparation (41). In a human acidosis pain model, topically applied acetylsalicylic and salicylic acid dose-dependently and competitively attenuated pain induced by low pH (15). Following
2054
Chemically-evoked PG% Release from Rat Skin
Vol. 62, No. 22, 1998
this. one would have expected that stimulation with solutions of low pH in the present skin preparation would lead to increased but not to decreased PGE, values. The cyclooxygenase activity is not disturbed by low pH, since the pH optimum lies in the acid range (42). Thus, the action of acetylsalicylic acid against pain from tissue acidosis should not be due to the drug’s COX blocking effect, but rather to direct antinociceptive effects of yet unknown mechanism. However, suppression of PGEz synthesis by acidosis has yet to be confirmed in human skin. In conclusion, rat skin is capable, like other tissues, to release PGE, upon stimulation with certain algogenic chemicals. This release is dose-dependent, differentiated and reflects the mutual interactions of inflammatory mediators. Thus, lack of sensitizing prostaglandin effects on rat cutaneous nociceptors is not due to a deficiency of the surrounding tissue to release PG. Instead, a deficiency of the rat skin nociceptors in reception of the PG signal must be assumed. Further experiments will be done to elucidate mechanisms of PGE z release, which may partly be mediated through an action of neuropeptides released upon algogenic chemical stimulation of primary afferents. Additionally, the suggested role of postganglionic sympathetic efferents in prostaglandin release deserves experimental attention (22) . Ack~rowledgmrcn~s: The Allergology, University excellent technical Forschungsgemeinschaft,
authors like to thank Prof. H.-W. Baenkler (Medical Clinic III., Dept. of Erlangen) for providing lab facilities and Mrs. Jana Schramm for assistance. This work was supported by the Deutsche SFB 353 - grant B12. References
I. 2. 3. 4. 5. 6. 7. 8. 9. 10. Il. 12. 13. 14. IS. 16. 17.
H. JUAN, Naunyn-Schmiedebergs Arch Pharmacol 300 77-85 (1977). K. MIZUMURA, J. SAT0 and T. KUMAZAWA, Pfliigers Archiv - Europ. J. Physiol. 408 565-572 (1987). A. RUEFF and A. DRAY, Neurosci. 54 527-53.5 (1993). G.J. BIRRELL, D.S. McQUEEN, A. IGGO and B.D. GRUBB, Neurosci. 54 537-544 (1993). V. NEUGEBAUER, H.-G. SCHAIBLE and R.F. SCHMIDT, Pfliigers Archiv - Europ. J. Physiol. 415 330-335 (1989). H.-G. SCHAIBLE and R.F. SCHMIDT, J. Physiol. - London 403 91-104 (1988). A.J. FOX, P.J. BARNES, L. URBAN and A. DRAY, J. Physiol. - London 469 21-35 ( 1993). B.D. GRUBB, G.J. BIRRELL, D.S. MCQUEEN and A. IGGO, Exp. Brain Res. 84 383392 (1991). T. WALTER, T.T. CHAU and B.M. WELCHMAN, Agents Actions 27 375-377 (1989). S.G. KHASAR, P.G. GREEN and J.D. LEVINE, Neurosci. Let. 153 215-218 (1993). E. LANG. A. NOVAK, P.W. REEH and H.Q. HANDWERKER, J. Neurophysiol. 63 887-901 (1990). P.W. REEH and S. BREHM, Sot. Neurosci. Abstr. 19 (1993). W. KESSLER, C. KIRCHHOFF, P.W. REEH and H.O. HANDWERKER, Exp. Brain Res. 91 467-476 (1992). P.W. REEH and K.H. STEEN. The Polymodal Nociceptor, T. Kumazawa, L. Kruger and K. Mizumura (eds), 143-1.51, Elsevier Science BV. Amsterdam (1996). K.H. STEEN, P.W. REEH and H.W. KREYSEL, Pain 64 71-82 (1996). K.H. STEEN, P.W. REEH and 1J.W. KREYSEL. Pain 62 339-347 (1995). P. HSU, M.L.C. ALBUQUERQUE and C.W. LEFFLER, Am. J. Physiol. 37 HS91603 ( 1995).
Vol. 62, No. 22, 1998
18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.
Chemically-evoked PC?% Release from Rat Skin
S. KILO, C. HARDING-ROSE, K.M. HARGREAVES and C.M FLORES, Pain 73 201207 (1997). S. BREHM, D. SCHAFER, H.-W. BAENKLER and P.W. REEH, Pfliigers Archiv Europ. J. Physiol. 249 133 (1995). K. BRUNE, M. REINKE, R. LANZ and B.A. PESKAR, FEBS Letters 186 46-50 (1985). D. SCHAFER, U. LINDENTHAL, M. WAGNER, P.L. BGLCSKEI and H.-W. BAENKLER, Thorax 51919-923 (1997). J.D. LEVINE, Y.O. TAIWO, S.D. COLLINS and J.K. TAM, Nature 323 158-160 (1986). K.B. GLASER, D. MOBILIO, J.Y. CHANG and N. SENKO, Trends Pharmacol. Sci. 14 92-98 (1993). J. AXELROD, R.M. BURCH and C.L. JELSEMA, Trends Neurosci. 11 117-123 (1988). G. REISER, F. BINMGLLER, P.N. STRONG and B. HAMPRECHT, Brain Res. 506 205-214 (1990). D. HOYER, D.E. CLARKE, P.R. FOZARD, G.R. MARTIN, E.J. MYLECHARANE, P.R. SAXENA and P.P.A. HUMPHREY, Pharmacol. Rev. 46 157-203 (1994). R. KURATA, I. TAKAYANAGI and T. HISAYAMA, Br J Pharmacol. 110 875-881 (1993). U. ZEILHOFER, P.W. REEH, D. SWANDULLA and M. KRESS, J. Neurophysiol. 76 2834-2840 (1996). A.L. WILLIS and D.L. SMITH, The Handbook of Immrtnopharmacology: Lipid Mediators, F.M. Cunningham (ed), 2-32, Academic Press, London, San Diego (1994). A. DRAY and M. PERKINS, Trends Neurosci. 16 99-104 (1993). R.M. BURCH and J. AXELROD, Proc. Natl. Acad. Sci. USA 83 6374-6378 (1986). K. MIZUMURA, M. MINAGAWA, Y. TSUJII and T. KUMAZAWA, Neurosci. Let. 161 117-119 (1993). M.K. CHURCH, Y. OKAYAMA and S. EL-LATI, Skin Pharmacol. 4 15-24 (1991). R.C. BENYON, C. ROBINSON and M.K. CHURCH, Br. J. Pharmacol. 97 898-904 (1989). H.A. BULL, J. COHEN and P.M. DOWD, J. Invest. Dermatol. 97 787-792 (1991). M.C. LINSSEN, W. ENGELS, P.J. LEMMENS, V.V. HEIJNEN, M. VAN BILSEN, RENEMANRS and G.J. VAN DER VUSSE, Am. J. Physiol. 264 H973-82 (199~). H.O. HANDWERKER, P.W. REEH and K.H. STEEN, Serotonin and Pain, J.-M. Besson (ed), 1-15, Elsevier Science Pb, Amsterdam (1990). C. Hi&LER, Klinische Wochenschrift 8 1569-1572 (1929). U. ISSBERNER, P.W. REEH, H.W. KREYSEL and K.H. STEEN, Abstr 81hWorld Congr Pain, 123 45, IASP press, Seattle (1996). J.B. SMITH, S.D. DWYER and L. SMITH, J. Biol. Chem. 264 8723-8728 (1989). K.H. STEEN, P.W. REEH, F. ANTON and H.O. I-IANDWERKER, J. Neurosci. 12 86-95 (1992). M. HEMLER and W.E. LANDS, J. Biol. Chem. 2515575-5579 (1976).
2055
ELSEVIER
STIMULATED
PII SOO24-3205(98)00176-3
PROSTAGLANDIN
sciem vol. at, No. 22, pp. 2045-1055,1998 ckpyeght 0 1958ElSnkr !3cimce Inc. Printedin the USA. All @Its relend 0024-~/!xl$19.00 t .m
E2 RELEASE FROM RAT SKIN, IN VITRO
S.K. Sauer IXI, D. Schafer #, M. Kress, P.W. Reeh Institut fur Physiologie und Experimentelle Pathophysiologie IXI, Universititstr. Institut fur Klinische Immunologie #, Universitatsstr. 29, Universitlt Erlangen-Ntimberg, D-91054 Erlangen
17;
(Received in final form March 4,1998)
Summary The excitatory effect of bradykinin (BK) and of low pH on nociceptors appears to partly depend on secondary release of prostaglandins from the surrounding tissue. Rat skin, in vitro, is introduced as a novel model to measure basal and stimulated release of PGE, and, in future, other substances relevant to nociception, such as neuropeptides. Flaps of hairy skin (n=57) from the rat saphenous region of the hindpaw were subcutaneously excised and fixed on acrylic rods, the corium side exposed. The preparations were equilibrated in carbogen gassed ,,synthetic interstitial fluid“ (SIF) for 30 minutes. The skin flaps were then immersed for 5 minutes each in 9 consecutive glass tubes, which were mounted in a shaking bath at 32°C. Each tube was filled with 5 ml of gassed SIF, the third tube contained inflammatory mediator(s) dissolved in SIF or solutions of low pH. After passage of the skin flap, the eluates were deep frozen (-70°C) and the PGE z content measured, off-line, using an enzyme immuno-assay. As stimulants, BK at 10m5M (n=9) and 1u6 M (n=4) and BK in equimolar combination with histamine (HA) and serotonin (5HT; lo5 M: n=8, 10e6 M: n=6, lo-’ M: n=6) dose-dependently increased PGEz release. Considering the total amount of PGE, secreted the combination of inflammatory mediators caused a significantly greater release of PGE, at lo5 and 10s6 M (~~0.01, Kruskal-Wallis test) than BK stimulation alone. Racemic flurbiprofen caused a profound depression of basal and stimulated release. Solutions of high proton concentration are known to stimulate and sensitize nociceptors. However, phosphate buffered SIF at pH 6.1 and 6.4 caused a substantial and significant decrease of the PGE2 release, probably due to low-pH block of phospholipases. Thus, algogenic potency of mediators does not necessarily match their pro-inflammatory action. Key Words: nociception, inflammation, bradykinh, histamine, serotonin, protons
Address for Correspondence: Dept. of Physiology I., Universitatsstr. Germany; phone +49-9131-856729, fax +49-9131-852497, e-mail erlangen.de
17, D-91054 Erlangen, [email protected]
Chdcally-evoked PGQ Release from Rat Skin
Vol. 62, No. 22, PBS
Introduction The algogenic effects of bradykinin (BK) seem to depend partly on a sensitizing action of prostaglandin E2 (PGIQ, that is released in many tissues in response to BK (1). Positive feedback is thought to result from this interaction, increasing the nociceptor exciting effect of BK. This theory is supported by suppressive effects of prostaglandin-synthesis inhibitors on responses of nociceptors to stimulation with BK (2, 3) and, further, by the effect of exogenous PGE, which can enhance BK-induced excitation in some preparations (4, 5, 6). However, dependence of BK responses on PGE, is not consistently found: In the neonatal rat tail-spinal of BK-induced reflex responses by non-steroidal cord preparation, the suppression antiinflammatory drugs (NSAIDs) could only be seen at low concentrations of BK (3). In the guinea pig trachea preparation BK responses of vagal afferents could not be suppressed by ibuprofen (7). Species differences could also be of importance, since effects which were obtained in knee joint afferents of the cat (6) could not be reproduced in ankle joint afferents of the rat (8). In behavioral experiments, NSAIDs could not reduce BK-induced writhing in mice (9) which is in contrast to BK-induced hyperalgesia in rats (10). In particular in the rat skin, in vitro, BK-induced nociceptor responses are not augmented by PGE, nor reduced by diclofenac (11). Blockade of the PGE, synthesis and substitution of exogenous PGE, did not exert influence on the nociceptor responses to BK in a combination of inflammatory mediators in normal and inflamed skin (12). In this preparation, like in many others, it is serotonin (5-HT) that facilitates the excitatory effect of BK which is of importance for the interaction of inflammatory mediators in inducing pain (11, 13). Low pH as in tissue acidosis is another relevant cofactor in inflammatory pain (14). NSAlDs have been shown to effectively suppress pain from experimental acidosis in human skin (15, 16). It is not known whether inhibition of prostaglandin synthesis is involved in this action or not. Again, low pH could stimulate PGE, release, as in endothelial cells (17), and by that, enhance the stimulatory effect of protons on nociceptors. The controversial findings raise the question whether rat skin, in vitro, is at all able to release PGE, in response to BK or low pH, and whether a possible effect could be facilitated by additional synergistic inflammatory mediators. Therefore, we developed a release model of isolated rat skin and measured PGE, liberation into the eluation fluid. PG synthesis has previously been measured in isolated cell preparations and bioptates from various tissues. Here, the skin was chosen since it is the organ on which by far most of the electrophysiological, pharmacological and behavioral pain studies were conducted but hardly any biochemical ones. Hence, we hope to fill a gap between the disciplines with correlative data. For the same reason, isolated skin preparations have recently been employed to study release of substance P and calcitonin-gene related peptide from cutaneous nerve endings (18, Averbeck and Reeh, submitted (1998); Kress, Guthmann, Averbeck and Reeh, submitted (1998)). Some of the present results have previously been communicated in abstract form (19). Methods For this novel release model we used skin flaps of the lower hindpaw of adult male Wistar rats. After sacrificing the animals in a container filled with pure CO, atmosphere, the hairy skin of both hindpaws was subcutaneously excised sparing the bigger vessels and the saphenous and sural nerves. The preparation started at knee level and included the toes. F rom 29 rats skin flaps (n=57) were obtained with a mean weight of 1.08 2 0.22 g (range 1.95 g- 0.66 g). After dissection, the flaps were wrapped around arylic glass rods (diam. 6 mm) with the corium side
Vol. 62, No. 22, 1998
Chemically-evoked PGQ Release from Rat Skin
2047
exposed and fixed with threads. The preparations were washed for 30 minutes on ice in 250 ml “synthetic interstitial fluid” at pH 7.4 (SIF containing (mM) 108 NaCI, 3.48 KCI, 3.5 MgS04 26 NaHCO,, 11.7 NaHzP04, 1.5 CaCl,, 9.6 Na-gluconate, 5.55 glucose, 7.6 sucrose). The cooling of the preparation during the initial washout was meant to suppress induction of PGE2 synthesis due to the surgical procedure. Subsequently, the skin flaps were placed in nine consecutive glass tubes filled with 5 ml SIF previously bubbled with carbogen (95% OZ, 5% CO,). These tubes were mounted in a shaking bath at 32 “C. After five minutes of incubation, the preparations were forwarded to the next tube, and the remaining eluates were deep frozen at -70°C. The third incubation step always contained the stimulating solution (see below). The amounts of PGE, released into the eluates were quantified in duplicate using a competitive enzyme immuno assay (EIA) which has been described previously (20). Assay procedures as well as intra-and interassay validation have recently been published (21). In brief, PGE z was coupled to bovine serum albumin (BSA), which was then used to coat the wells of microtiter plates at pH 9.6 for 18 hours (200 @/well). After this and every following incubation step the plates were washed using PBS-Tween 0.1 % (v/v). Unspecific binding sites in the wells were blocked with 1% BSA in PBS for 60 minutes. Standard curves were assessed using seven different concentrations of PGE, (1 to 500 pg/well) in duplicate. Samples and standards (100 pg/well) were incubated with a monoclonal mouse antibody against PGE, (mab E2Rl; 100 ul/weIl - a generous gift of K. Brune, Dept. Pharmacol., University Erlangen, Germany) at 4°C for 18 hours. Thereafter the biotinylated goat-anti-mouse antibody was added (200 @/well for 2 hours at 37°C). Following this, the EIA was incubated with the streptavidin-biotin complex coupled to horseradish peroxidase (HRP) for 1 hour at 37 “C (200 p,l/well). As a chromogenic substrate we used ABTS (2,2’-azino-bis [3 ethylbenzthiazoline-6-sulfate]) for 45 minutes at room temperature (200 @/well). The absorbance was determined at 405 nm (reference filter 490 nm). A recovery rate of 98% for PGE2 characterized the assay. Detection limit of the PGE z-EIA was 3 pg/well. The intra- and interassay coefficients of variation (cv) for PGE 2 were 5.6% and 8.1%, respectively. To determine the viability of the assay under low pH conditions, the acidic samples collected were measured using standards at low pH (6.1 and 6.4). The validation assays performed for low pH standards resulted in intra- and interassay coefftcients of variation (cv) of 8.9% and 9.7% for pH 6.1, 8.1% and 9.2% for pH 6.4, which were larger than in standard assays at neutral pH. Searching for a possible reason we found that the recovery rates for PGE z were systematically lower than normal in acidic samples that had been frozen and thawed, whereas the recovery of PGE, in freshly prepared acidic standard solutions was 98% as above. To account for this loss of PGE, in the thawed acidic samples and to prevent from underestimating the real PGE, content, we determined a mean correcting factor of 1.24 for solutions of pH 6.1 and of 1.05 for pH 6.4 with which the particular PGE2 values were multiplied before entering further data analysis. PGE, values (pg) are given per 1 ml of the eluation fluid SIF and referred to lg fresh-weight cutaneous tissue.
of
For chemical stimulation we used the following solutions, all of them diluted in SIF, aerated with carbogen (95% O2 and 5% COZ) and applied at 32 “C. . Bradykinin (BK-triacetate, and pH 7.4.
Sigma) was used at 10m5M (n=9) and 10m6M (n=4) concentration
2048
Chemically-evoked PC% Release from Rat Skin
Vol. 62, No, 22, 1998
An equimolar combination of inflammatory mediators (IM) which was composed of BK, histamine (HA) and serotonin (5HT) (all from Sigma) at pH 7.4 and applied at 10 -5 M (n=8), 10m6M (n=6) and 10m7M (n=6). BK (n=4) and IM (n=4), each at 1O-5 M, were used to stimulate the skin treated with flurbiprofen. Racemic flurbiprofen (a generous gift of G. Geisslinger, Dept. Pharmacol., University Erlangen, Germany) was used to block the endogenous prostaglandin synthesis; the rats received a single dose of 25 mg/kg i.p. one hour before dissection. Additionally, we added flurbiprofen at 10m6M concentration to the SIF solution in each test tube. Stimulation solutions of low pH were prepared using phosphate buffered SIF. The bicarbonate content of SIF was replaced by phosphate buffer in proportions to reach pH values of 6.4 (n=8) and 6.1 (n=8), which were fine adjusted using drops of HCl or NaOH. 1500
Pg 1250
n=4
\
-----
250
lmr 0
0
5
10 15 20 on ice
25
30
I
35
40
45
50
55
60
65
70
75 min
at 32°C
Fig. I Release of PGE2 from the rat isolated skin under resting conditions, i.e. without any chemical stimulation. Note the temperature effect upon transferal of the preparations into the warmer glass tubes after the pre-wash period, on ice. The amount of PGE 2 measured is referred to lg fresh weight of cutaneous tissue in 1 ml eluation fluid. Dafa analysis: Results concerning groups of identical experiments are given as means + s.e.m. The numbers of experiments (n) in the figures refer to skin flaps from different animals. For calculation of the total amount of stimulated PGE, release, values comprising samples three to six were pooled. Statistical non-parametric comparisons within one group of experiments were made using the Wilcoxon matched pairs test which compared PGE, values before and during stimulation. Comparisons between different experimental groups were performed using the nonparametric analysis of variance (Kruskal-Wallis test) and U-tests (Mann-Whitney). Values of p < 0.05 were considered significant.
VoL tX$ No. 22,19%
Chcmkally-cvoked
l’G% Release fromRat Skin
Results
The cutaneous release of PGE, was remarkably influenced by the temperature step of more than 20 “C after the initial 30 minutes pre-wash period (see methods). The temperature rise induced substantial increase and interindividual variability of the PGEz release which was addressed in a separate series of experiments (n=4; fig 1). During the initial pre-wash period at about lO”C, PGE, values varied between 100 and 300 pg and were increased by more than 4 times when the preparations were transferred into SIF at 32°C. After an initial overshoot in this group of skin flaps (n=4; peak values 890 2 300 pg), however, relatively stable intraindividual PGE2 levels were reached and maintained over the subsequent 40 minutes averaging to 456.08 f 34.38 pg (n=4 x 8 consecutive values pooled). In the main study, PGE, levels measured in the first two incubation steps were taken for baseline: average values of 421.5 f 46.8 pg (2 sem) in the first and 433.7 + 52.5 pg in the second sample (n=49) were found. These baseline values match the values found in the second phase of the temperature experiment. Nonetheless, we have drawn the consequence for future studies to perform the initial pre-wash and equilibration of the skin preparation at a temperature of 32°C. release induced by inflammatory mediators: The stimulation with BK at 10q5 M concentration significantly increased the PGE2 level by more than three times in the sample containing BK as compared to the preceeding control sample (fig. 2). BK at 10M6M yielded a smaller increase of PGE, baseline which did just not reach significance (probably due to n=4). At the high BK concentration complete and significant recovery was reached only in the second wash-out step after BK. Flurbiprofen (pre-) treatment (25 mg/kg i.p. and 10 -6 M in the eluation solution) almost abolished basal and PGE, release stimulated by BK 10M5 M.
PGE2
1500 PB BK lo’:M, (n=9) : ;z ;$+ (n=4)
1250
l
1000
+flurbiprofen,
(~4)
P 4 2
750 a: p=O.O4vs. bl. b: p=O.O6vs. bl. c: p=O.O05vs. BK 105M
c-4
g
500 250 0 I ;
Ib
1;
;o
;5
3b
3b
40
45 min
Fig. 2 BK (1V5 and 1N6 M) stimulation for 5 minutes (grey area) induced transiently enhanced PGE, release from the rat skin. Flurbiprofen, a cyclooxygenase blocker, was administered to the alive rats and contained in the incubation fluid (25 mg/kg i.p.; 1U6 M in SIF); it completely attenuated PGE, secretion.
2050
ChemicalIywked
PGQ Relesse from Rat Skin
Vol. 62,No. 22,1998
The combination of inflammatory mediators BK, HA and 5-HT (,,inflammatory soup “, IM) was applied at three different concentrations and caused a protracted release of PGE, (fig. 3). IM at lo9 M concentration significantly augmented the average release of PGE z by a factor of 3.7; IM at low6M concentration was also significantly (p=O.O2) effective with a factor of 2.6. Stimulation with IM at lo-’ M concentration evoked an increase of the PGE 2 release by a factor of two which just did not reach level of significance as compared to baseline (p= 0.07, n=6). Again, flurbiprofen treatment of the animals and preparations resulted in a complete depression of the basal and stimulated prostaglandin synthesis as tested with stimulation by IM at 10.’ M concentration. For comparison of the releasing efficacies of BK and of the combination of inflammatory mediators, we calculated the total amount of stimulated PGE2 release during and after stimulation until baseline was reached again (comprising samples 3 to 6; fig. 4). Bradykinin and the combination of inflammatory mediators increased the total PGE, release dose-dependently (p=O.O3 for BK and p=O.O4 for IM-stimulation, Kruskal-Wallis test). The pooling of the data revealed a “ceiling“ effect of the increasing IM concentration at 10e6 M. In contrast to stimulation with BK, the inflammatory mediators at 10e5 M concentration did not significantly cause further enhancement of the total PGE 2 release.
A l n l
IM 10.‘M, (n=8) IM 10-6M,(s6) IM 10.‘M, (n=6) IM 1n5M +flurbiprofen, (n=4)
a: p=O.O3vs. bl. b: p=O.O2vs. bl.
0
0
I
r
5
10
I
15
I
I
1
I
1
20
25
30
35
40
1 45
Fig. 3 The combination of inflammatory mediators (IM: equimolar BK, HA and 5-HT) evoked massive PGE, secretion from the rat skin, and the effects outlasted the period of chemical stimulation (grey area). Again flurbiprofen (25 mg/kg i.p.; 10 -6 M in SIF) completely blocked basal and stimulated PGE, release. The combination of inflammatory mediators was, however, significantly more effective than BK alone at the two equimolar concentrations (fig. 4). IM 1O-7 M was about equipotent to 10e5 M BK, i.e. addition of 5-HT and HA to BK caused a hundredfold increase in potency. Flurbiprofen treatment significantly suppressed total PGE, release after BK and IM (10m5 M) stimulation (p=O.OOSand p=O.O06, Kruskal-Wallis test).
Chemidyevokcd
Vol. 62, No. 22, 19!38
PGE$ Rekase from Rat Skin
2051
PGE2 release reduced by solutions of low pH: To verify whether another potent stimulus, that
excites and sensitizes nociceptors, is also effective to release prostaglandins, we used solutions of low pH. Solutions of pH 6.1 and 6.4, phosphate buffered SIF, produced profound and outlasting suppression of the PGE, release (fig. 5). Exposure to pH 6.4 caused a more transient but equally strong suppression as pH 6.1 with rapid recovery within five minutes. PGE, remained significantly below baseline for 20 min after exposure to pH 6.1. The relative supression was about the same and not significantly different comparing pH 6.1 and pH 6.4 application. Also, bicarbonte buffered SIF bubbled with pure CO 2 to create a pH of 6.1 caused a marked decrease of the mean PGE, release (p=O.O7,n=4; data not shown). 3500 Pl3
A
3000
T
2500
I
p=o.o4
I
2000 .-a
2 1500
p=o.o3 I
1
BK 1 o”M n=9
BK 10-6M n=4
BK 1O-‘M +fl n=4
IM 1 O-“M n=8
IM I OdM n=6
IM 1O-‘M n=6
IM ‘Z-Y n=4
Fig. 4 Summary showing total amounts of PGE2 released during and after chemical stimulations, pooling data from samples 3-6 (= lo-30 min). The combination of inflammatory mediators (IM) was significantly more effective than BK stimulation at comparable concentrations (p=O.OOOl,Kruskal-Wallis test). Discussion
The novel release model of the rat skin, in vitro, is suitable to study mediator release - here PGE 2 - in response to chemical stimulation. Stimulation with BK alone and, up to six-fold more effective, with an equimolar combination of inflammatory mediators (IM: bradykinin, BK, histamine, HA, serotonin, .5-HT) significantly increased PGE, release from the isolated skin. The treatment with a cyclooxygenase blocker, flurbiprofen, caused a complete suppression of basal and stimulated release. Surprisingly, solutions of low pH, which potently excite and sensitize nociceptors, induced a profound suppression of the prostaglandin release. In the rat hairy skin preparation different types of cells may be sources of PGE 2: keratinocytes, tibroblasts, melanocytes and Langerhans cells in the superficial epidermal layers and fibroblasts, endothelial cells, mast cells and macrophages in deeper layers - the corium. Since the corium side of the skin was exposed to chemical stimulation (see methods), this tissue may contribute the major part of the PGE2 released, although the lymphatic clefts of the corium allow for rapid fluid convection reaching the epidermal border. Also postganglionic efferents of the sympathetic
Chemically-evoked PC+
2052
Vol. 62, No. Z&l998
Release from Rat Skin
nervous system to the skin are suggested to be involved in the prostanoid chemical stimulation (22).
release after noxious
Arachidonic acid provides the main source for lipid mediators such as prostaglandins, leukotrienes, thromboxanes and platelet activating factor. The hydrolysis of membrane phospholipids to arachidonate by phospholipases is thought to be rate-limiting in the prostanoid cascade. Phospholipase A2 (PLA,) is represented by a multifarious group of enzymes, secretory extracellular forms and intracellular ones (23). 900 -) n
pH 6.1 pb, (n=8) pH 6.4 pb, (n=8)
0
200 a: p=O.Ol vs. bl. b: p=O.Ol vs. bl. 5
IO
15
20
25
30
35
40
45
Fig. 5 Exposure of the skin to solutions of low pH (grey area) depressed PGE, release significantly. Phosphate buffered SIF at pH 6.1 produced profound and outlasting suppression of the PGE, release; pH 6.4 was equally effective but more transient in action. Rise of the intracellular Ca++ concentration is a common principle for PLAZ activation and translocation to the plasma membrane. This is notable because many of the inflammatory mediators (BK, HA, 5-HT, protons) are reported to cause an increase of the intracellular Ca ++ concentration and, following this, are potential stimulators of PLA, (24, 25, 26, 27, 28). The successive step in prostaglandin synthesis is the conversion of arachidonate to PGG I and PGH,, which is catalyzed by the cyclooxygenases (COX I and II) (for review see (29)). Bradykinin is one of the most potent endogenous algogens; it excites nociceptive primary afferents and causes pain (for review see (30)). BK has been shown to release prostaglandins from different tissues on different routes of application (1, 31). In addition, BK sensitizes nociceptive afferents to physical stimuli, and this has been suggested to be mediated through release of prostaglandins (3, 2, 5). However, 10s8 M concentration of BK is sufficient to induce the sensitizing effect, whereas PGE2 needs to be applied at least in 10s5 M concentration to induce nociceptor sensitization to heat (32). In our experimental setup, BK dose-dependently evoked PGE, secretion at 10e6 M and 10m5M concentration. This fits well to BK concentrations effective to stimulate prostaglandin release and to excite nociceptors as reported from previous work (1, 31). The rapid accumulation of PGE, in five minutes induced by relatively short lasting stimulation with inflammatory mediators is in agreement with previous findings (29). For
Vol. 62, No. 22, 19!%
Chemically-evoked PGQ Release from Rat Skin
2053
cultured fibroblasts a half maximal release was achieved after two minutes of BK stimulation at tom6M concentration whereas interleukins and phorbol ester caused a more delayed but sustained PG release (31). Histamine-induced prostaglandin release has been shown in many different tissues and cell types (33, 34). This release is mediated by the Hl receptor in membranes of keratinocytes, myocytes and endothelial cells (35, 36). Serotonin plays an important role in inflammation and in pain mechanisms (37). So far, seven different receptors for 5-HT have been isolated (for review see (26)) which partly facilitate the activation of PLA, due to their second messenger pathways (DAG/IP,). A direct activation of the PLAZ has - to our knowledge - not been reported. The combination of inflammatory mediators, BK, HA and 5-HT, resulted in further enhancement of the PGEz release as compared to BK stimulation (see fig 4). In electrophysiological experiments, the combination of inflammatory mediators activated about 30% more of the primary nociceptive afferents in rat skin than BK alone, and the excitatory effect was much stronger (11, 13). The enhanced responsiveness to IM of the cutaneous nociceptors is due to mutually sensitizing and additive effects of the combined inflammatory mediators, to which the endogenous prostaglandins, triggered for example by BK, did not contribute: The mediator combination (BK, 5-HT, HA) had the same impact on nociceptors in the skin whether or not the cyclooxygenase was blocked (by flurbiprofen) and PGE, (10” M) was added (12). In agreement with this, the IM-induced nociceptor excitation significantly increased from 1O‘6 M to 10e5 M concentration (Reischl & Reeh., unpublished), while the induced PGE2 release in the present work stagnated. The stimulation of the rat skin with phosphate-buffered solutions of pH 6.1 and 6.4 significantly decreased PGE, release under basal levels. This is at first surprising since in inflamed tissues, pH values down to 5.4 have been measured (38, 39), and high PGE, levels have been found (29). But pH and prostaglandin values have never been measured simultaneously in one and the same study. The decrease of PGE, release induced by exposure to low pH fits well with enzymatic properties of the PLA,. Several isoforms of PI..& are reported to have pH optima for their enzymatic activity in the alkaline range, and this is true for secreted as well as cytosolic forms and whether or not the iso-enzyme is Cat+ dependent (23). Changes of the extracellular pH were effective to induce an increased formation of IP 3 and a Cat+ peak in fibroblasts, which effects were remarkably similar to those of BK (10 -’ M) stimulation. The authors suppose the same intracellular pathway and pool of Ca ++ to be mobilized (40). Also, in capsaicin-sensitive sensory neurons, a putative cellular model of nociception, low-pH stimulation evoked an influx of Cat+ which by itself may not be sufficient to activate PLA, (28). In contrast, in endothelial and smooth muscle cells low pH (pH 6.8) stimulation produced an indomethacin (O.lpM) sensitive release of PGE, although low pH did not change the intracellular Cat+ concentrations or induce Cat’ influx in these experiments (17). The authors assume that intracellular pH changes stimulated PG synthesis by activating a Ca ++independent PLA,. Thus, in the complex cutaneous tissue relative inhibition of the PLA2 by low pH seems to prevail. This may be overcome in case of inflammation when mediators strongly boost intracellular Cat+ and activate the PLA, and, thus, allow for coexistence of high PG and proton concentrations. Solutions of low pH potently excite and sensitize nociceptive primary afferents in the skin-nerve preparation (41). In a human acidosis pain model, topically applied acetylsalicylic and salicylic acid dose-dependently and competitively attenuated pain induced by low pH (15). Following
2054
Chemically-evoked PG% Release from Rat Skin
Vol. 62, No. 22, 1998
this. one would have expected that stimulation with solutions of low pH in the present skin preparation would lead to increased but not to decreased PGE, values. The cyclooxygenase activity is not disturbed by low pH, since the pH optimum lies in the acid range (42). Thus, the action of acetylsalicylic acid against pain from tissue acidosis should not be due to the drug’s COX blocking effect, but rather to direct antinociceptive effects of yet unknown mechanism. However, suppression of PGEz synthesis by acidosis has yet to be confirmed in human skin. In conclusion, rat skin is capable, like other tissues, to release PGE, upon stimulation with certain algogenic chemicals. This release is dose-dependent, differentiated and reflects the mutual interactions of inflammatory mediators. Thus, lack of sensitizing prostaglandin effects on rat cutaneous nociceptors is not due to a deficiency of the surrounding tissue to release PG. Instead, a deficiency of the rat skin nociceptors in reception of the PG signal must be assumed. Further experiments will be done to elucidate mechanisms of PGE z release, which may partly be mediated through an action of neuropeptides released upon algogenic chemical stimulation of primary afferents. Additionally, the suggested role of postganglionic sympathetic efferents in prostaglandin release deserves experimental attention (22) . Ack~rowledgmrcn~s: The Allergology, University excellent technical Forschungsgemeinschaft,
authors like to thank Prof. H.-W. Baenkler (Medical Clinic III., Dept. of Erlangen) for providing lab facilities and Mrs. Jana Schramm for assistance. This work was supported by the Deutsche SFB 353 - grant B12. References
I. 2. 3. 4. 5. 6. 7. 8. 9. 10. Il. 12. 13. 14. IS. 16. 17.
H. JUAN, Naunyn-Schmiedebergs Arch Pharmacol 300 77-85 (1977). K. MIZUMURA, J. SAT0 and T. KUMAZAWA, Pfliigers Archiv - Europ. J. Physiol. 408 565-572 (1987). A. RUEFF and A. DRAY, Neurosci. 54 527-53.5 (1993). G.J. BIRRELL, D.S. McQUEEN, A. IGGO and B.D. GRUBB, Neurosci. 54 537-544 (1993). V. NEUGEBAUER, H.-G. SCHAIBLE and R.F. SCHMIDT, Pfliigers Archiv - Europ. J. Physiol. 415 330-335 (1989). H.-G. SCHAIBLE and R.F. SCHMIDT, J. Physiol. - London 403 91-104 (1988). A.J. FOX, P.J. BARNES, L. URBAN and A. DRAY, J. Physiol. - London 469 21-35 ( 1993). B.D. GRUBB, G.J. BIRRELL, D.S. MCQUEEN and A. IGGO, Exp. Brain Res. 84 383392 (1991). T. WALTER, T.T. CHAU and B.M. WELCHMAN, Agents Actions 27 375-377 (1989). S.G. KHASAR, P.G. GREEN and J.D. LEVINE, Neurosci. Let. 153 215-218 (1993). E. LANG. A. NOVAK, P.W. REEH and H.Q. HANDWERKER, J. Neurophysiol. 63 887-901 (1990). P.W. REEH and S. BREHM, Sot. Neurosci. Abstr. 19 (1993). W. KESSLER, C. KIRCHHOFF, P.W. REEH and H.O. HANDWERKER, Exp. Brain Res. 91 467-476 (1992). P.W. REEH and K.H. STEEN. The Polymodal Nociceptor, T. Kumazawa, L. Kruger and K. Mizumura (eds), 143-1.51, Elsevier Science BV. Amsterdam (1996). K.H. STEEN, P.W. REEH and H.W. KREYSEL, Pain 64 71-82 (1996). K.H. STEEN, P.W. REEH and 1J.W. KREYSEL. Pain 62 339-347 (1995). P. HSU, M.L.C. ALBUQUERQUE and C.W. LEFFLER, Am. J. Physiol. 37 HS91603 ( 1995).
Vol. 62, No. 22, 1998
18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.
Chemically-evoked PC?% Release from Rat Skin
S. KILO, C. HARDING-ROSE, K.M. HARGREAVES and C.M FLORES, Pain 73 201207 (1997). S. BREHM, D. SCHAFER, H.-W. BAENKLER and P.W. REEH, Pfliigers Archiv Europ. J. Physiol. 249 133 (1995). K. BRUNE, M. REINKE, R. LANZ and B.A. PESKAR, FEBS Letters 186 46-50 (1985). D. SCHAFER, U. LINDENTHAL, M. WAGNER, P.L. BGLCSKEI and H.-W. BAENKLER, Thorax 51919-923 (1997). J.D. LEVINE, Y.O. TAIWO, S.D. COLLINS and J.K. TAM, Nature 323 158-160 (1986). K.B. GLASER, D. MOBILIO, J.Y. CHANG and N. SENKO, Trends Pharmacol. Sci. 14 92-98 (1993). J. AXELROD, R.M. BURCH and C.L. JELSEMA, Trends Neurosci. 11 117-123 (1988). G. REISER, F. BINMGLLER, P.N. STRONG and B. HAMPRECHT, Brain Res. 506 205-214 (1990). D. HOYER, D.E. CLARKE, P.R. FOZARD, G.R. MARTIN, E.J. MYLECHARANE, P.R. SAXENA and P.P.A. HUMPHREY, Pharmacol. Rev. 46 157-203 (1994). R. KURATA, I. TAKAYANAGI and T. HISAYAMA, Br J Pharmacol. 110 875-881 (1993). U. ZEILHOFER, P.W. REEH, D. SWANDULLA and M. KRESS, J. Neurophysiol. 76 2834-2840 (1996). A.L. WILLIS and D.L. SMITH, The Handbook of Immrtnopharmacology: Lipid Mediators, F.M. Cunningham (ed), 2-32, Academic Press, London, San Diego (1994). A. DRAY and M. PERKINS, Trends Neurosci. 16 99-104 (1993). R.M. BURCH and J. AXELROD, Proc. Natl. Acad. Sci. USA 83 6374-6378 (1986). K. MIZUMURA, M. MINAGAWA, Y. TSUJII and T. KUMAZAWA, Neurosci. Let. 161 117-119 (1993). M.K. CHURCH, Y. OKAYAMA and S. EL-LATI, Skin Pharmacol. 4 15-24 (1991). R.C. BENYON, C. ROBINSON and M.K. CHURCH, Br. J. Pharmacol. 97 898-904 (1989). H.A. BULL, J. COHEN and P.M. DOWD, J. Invest. Dermatol. 97 787-792 (1991). M.C. LINSSEN, W. ENGELS, P.J. LEMMENS, V.V. HEIJNEN, M. VAN BILSEN, RENEMANRS and G.J. VAN DER VUSSE, Am. J. Physiol. 264 H973-82 (199~). H.O. HANDWERKER, P.W. REEH and K.H. STEEN, Serotonin and Pain, J.-M. Besson (ed), 1-15, Elsevier Science Pb, Amsterdam (1990). C. Hi&LER, Klinische Wochenschrift 8 1569-1572 (1929). U. ISSBERNER, P.W. REEH, H.W. KREYSEL and K.H. STEEN, Abstr 81hWorld Congr Pain, 123 45, IASP press, Seattle (1996). J.B. SMITH, S.D. DWYER and L. SMITH, J. Biol. Chem. 264 8723-8728 (1989). K.H. STEEN, P.W. REEH, F. ANTON and H.O. I-IANDWERKER, J. Neurosci. 12 86-95 (1992). M. HEMLER and W.E. LANDS, J. Biol. Chem. 2515575-5579 (1976).
2055