The leukotriene B4 receptors BLT1 and BLT2 form an antagonistic sensitizing system in peripheral sensory neurons

Abstract

Sensitization of the heat-activated ion channel transient receptor potential vanilloid 1 (TRPV1) through lipids is a fundamental mechanism during inflammation-induced peripheral sensitization. Leukotriene B4 is a proinflammatory lipid mediator whose role in peripheral nociceptive sensitization is not well understood to date. Two major G-protein-coupled receptors for leukotriene B4 have been identified: the high-affinity receptor BLT1 and the low-affinity receptor BLT2. Transcriptional screening for the expression G-protein-coupled receptors in murine dorsal root ganglia showed that both receptors were among the highest expressed in dorsal root ganglia. Calcium imaging revealed a sensitization of TRPV1-mediated calcium increases in a relative narrow concentration range for leukotriene B4 (100–200 nm). Selective antagonists and neurons from knock-out mice demonstrated a BLT1-dependent sensitization of TRPV1-mediated calcium increases. Accordingly, leukotriene B4-induced thermal hyperalgesia was mediated through BLT1 and TRPV1 as shown using the respective knock-out mice. Importantly, higher leukotriene B4 concentrations (>0.5 μm) and BLT2 agonists abolished sensitization of the TRPV1-mediated calcium increases. Also, BLT2 activation inhibited protein kinase C- and protein kinase A-mediated sensitization processes through the phosphatase calcineurin. Consequently, a selective BLT2-receptor agonist increased thermal and mechanical withdrawal thresholds during zymosan-induced inflammation. In accordance with these data, immunohistochemical analysis showed that both leukotriene B4 receptors were expressed in peripheral sensory neurons. Thus, the data show that the two leukotriene B4 receptors have opposing roles in the sensitization of peripheral sensory neurons forming a self-restricting system.

Introduction

Leukotriene B4 (LTB4)³ is well known as a proinflammatory lipid mediator, which causes the recruitment of neutrophils to the site of inflammation and enhances pain sensation (nociception) through the release of pronociceptive mediators by the recruited immune cells (1). LTB4 is generated through lipoxygenase 5 and leukotriene A4-hydrolase (2, 3) and acts mainly through two G-protein-coupled receptors: BLT1 and BLT2 whereby BLT1 has a higher affinity for LTB4 (pK_d 9.2) than BLT2 (pK_d 7.2) (4–7). The nociceptive effects of LTB4 are attributed mainly to the recruitment of neutrophils (1), causing a long-lasting (several hours) hyperalgesia (8). In contrast, LTB4 reduces the proresolving lipid 18S-resolvin LTB4-induced NF-κB activation, as well as neutrophil infiltration by antagonizing BLT1, thereby limiting inflammation in mice (9, 10).

Ion channels such as the transient receptor potential channels transduce physical (e.g. heat or cold) and chemical (e.g. endogenous lipids) stimuli in the peripheral nociceptive system (11–13). One member of this family, the transient receptor potential vanilloid 1 (TRPV1) ion channel is a key element in peripheral nociceptor sensitization and modulation during inflammation. TRPV1 activity can be sensitized by G-protein-coupled receptors, such as bradykinin or prostaglandin E₂ receptors (13, 14), which induce TRPV1-phosphorylation by PKC or PKA, respectively (15, 16). Although activation of PKA and PKC leads to phosphorylation and sensitization of TRPV1, activation of the protein phosphatase 2B (calcineurin) causes dephosphorylation and desensitization of TRPV1 (17, 18). In regard to LTB4, BLT1 activation has been shown to increase nociception (19, 20), and it has been shown that at very high concentrations (EC₅₀ = 11.7 μm), LTB4 may act directly on TRPV1 as a weak agonist (20).

Although the proinflammatory and pronociceptive roles of BLT1 have been studied more intensively, the relevance of the BLT2 receptor in inflammation and especially in nociception is not fully understood. For BLT2, both pro- and anti-inflammatory effects have been described. For example, BLT2 is expressed in colon cryptic cells appears to protect against dextran sodium sulfate-induced colitis, possibly by enhancing barrier function in epithelial cells of the colon (21). Similarly, in the mouse model of ovalbumin-induced allergic airway disease, BLT2 has a protective role (22). In contrast, BLT2-deficient mice have a reduced incidence and severity of disease in autoantibody-induced inflammatory arthritis, including protection from bone and cartilage loss (23).

Here, we describe that BLT1 and BLT2 are coexpressed in peripheral sensory neurons and that low concentrations of LTB4 sensitize TRPV1-mediated intracellular calcium increases through BLT1-induced PKC activation, whereas higher LTB4 concentrations desensitize TRPV1-mediated calcium increases through BLT2-induced calcineurin activation. Fittingly, in vivo selective activation of BLT2 caused strong antinociceptive effects on thermal hyperalgesia and mechanical allodynia.

Results

Low concentrations of LTB4 sensitize TRPV1-mediated calcium increases through BLT1

G-protein-coupled receptors play an important role in the modulation of the peripheral nociceptive system. To investigate which receptors are present in dorsal root ganglia, we performed a semiquantitative mRNA screen for 546 known G-protein-coupled receptors with RNA of DRGs from adult mice. The LTB4 receptors BLT1 and BLT2 were among the most abundant receptors with the 14^th and 7^th highest mRNA copy numbers, respectively (Table 1, GEO accession number GSE75476).

Table 1.

List of the 30 highest expressed mRNAs of GPCRs in murine DRGs

	Gene	Receptor name	Target copies/ng cDNAa	Ratio (target/Gapdh)b
1	PTGER3	EP3 receptor	2658.4	25.8
2	ADRA2C	Adrenoceptor α2C	2603.7	22.1
3	GPR88	G-protein-coupled receptor 88	2000.8	25.9
4	ADRA2A	Adrenoceptor α2A	1828.4	15.4
5	GPR153	G-protein-coupled receptor 153	1790.7	16.2
6	ADRB1	β1-Adrenoceptor	1613.9	13.6
7	BLTR2	BLT2	1257.5	11.6
8	PTGER1	EP1 receptor	823.9	8.9
9	OXTR	Oxytocin receptor	812.6	7.3
10	TAS1R1	Taste receptor type 1 member 1	768.7	10.2
11	NPFFR1	Neuropeptide FF receptor 1	758.1	6.8
12	FZD8	Frizzled class receptor 8	758.1	10.8
13	FZD5	Frizzled class receptor 5	650.9	8.7
14	BLTR	BLT1	642.0	5.9
15	GPR62	G-protein-coupled receptor 62	615.8	5.8
16	ADRA1D	α1D-Adrenoceptor	590.7	5.4
17	CRHR2	Corticotropin-releasing hormone receptor 2	547.4	7.4
18	GPR26	G-protein-coupled receptor 26	539.8	5.2
19	GPR25	G-protein-coupled receptor 25	525.1	4.8
20	ADRB3	β3-adrenoceptor	510.7	4.3
21	GPR137B	G-protein-coupled receptor 137B	473.2	6.1
22	GPR54	Kisspeptin receptor	453.9	4.3
23	LGR6	Leucine-rich repeat-containing GPCR6	450.8	4.9
24	Vmnr2r54	Vomeronasal receptor	447.7	4.8
25	FZD7	Frizzled family receptor 7	435.4	5.8
26	HTR6	5-Hydroxytryptamine receptor 6	429.4	4.3
27	GPR135	G-protein-coupled receptor 135	429.4	5.5
28	DRD4	Dopamine receptor D4	414.8	3.5
29	P2Y4	Pyrimidinergic receptor P2Y4	403.5	4.4
30	DRD5	Dopamine receptor D5	392.4	3.3

^a Total RNA was isolated from DRGs of untreated adult mice and subjected to semiquantitative RT-PCR. The data are corrected for target copies/ng cDNA.

^b The relative expression of GPCR genes expressed as ratio between copy number of target/ng cDNA and copy number of Gapdh/ng cDNA.

Previously it has been demonstrated in electrophysiological experiments with isolated membrane patches of sensory neurons that LTB4 acts at concentrations >10 μm as a weak agonist for TRPV1 (20). Therefore, we investigated in this study the impact of BLT1 and BLT2 activation on sensitization the ion channel TRPV1 in peripheral sensory neurons. To examine sensitization of TRPV1-mediated responses, calcium imaging experiments were performed with cultured DRG neurons from adult mice. Here, dorsal root ganglia neurons were first stimulated with capsaicin (200 nm) alone, followed by an incubation with LTB4 prior to a second capsaicin stimulation. Sensitization of TRPV1-mediated calcium increases by LTB4 was calculated as the ratio between the second peak (cells treated with LTB4 and capsaicin) and the first peak (cells stimulated with capsaicin alone). The second capsaicin stimulations was induced 10 min after the first to avoid a decrease of the second capsaicin response caused by desensitization of TRPV1 (24). In accordance with a previous report (20), no elevation of intracellular calcium concentrations was observed during the incubation of LTB4 (Fig. 1A). However, preincubation with LTB4 significantly enhanced the capsaicin-stimulated calcium increases reaching a maximum with 100 nm LTB4 (Fig. 1, A and B). Surprisingly, the sensitizing effect disappeared when LTB4 concentrations were further increased (Fig. 1B). To investigate the potential involvement of BLT1 in this sensitization process, we first tested the effect of LTB4 on dorsal root ganglia cultures from BLT1-deficient mice and observed that neither 0.1 nor 1 μm LTB4 was able to sensitize TRPV1-mediated calcium increases in the absence of BLT1 (data not shown). Also, incubation with the BLT1 antagonist U75302 abolished the sensitizing effect of 100 nm LTB4 in dorsal root ganglia cultures from wild-type mice (Fig. 1, C and D). Because BLT1 is known to activate PKC signaling, we hypothesized that LTB4 may sensitize TRPV1 through PKC activation. Fittingly, the PKC inhibitor bisindolylmaleimide (BIM) (Fig. 1, C and D) eliminated the sensitizing effect of LTB4. To test whether or not the sensitizing effect of LTB4 is restricted to TRPV1-expressing neurons, we tested TRPM8-positive neurons, which are an individual TRPV1-negative subgroup of nociceptors (25). Therefore, dorsal root ganglia neurons were stimulated twice with menthol (100 μm) and with 100 nm LTB4 or its vehicle prior to the second menthol stimulus. These two menthol stimuli were followed by a capsaicin stimulus (250 nm, 20 s) to ensure that menthol responders are TRPV1-negative. Interestingly, LTB4 was not able to sensitize TRPM8-mediated calcium increases (data not shown), suggesting that the sensitizing effect is limited to TRPV1-expressing neurons.

Figure 1.

LTB4 sensitizes TRPV1-mediated calcium influx through BLT1. A–D, calcium imaging from DRG culture from C57BL/6N (A and B) or BLT1 knock-out mice (C and D). A, capsaicin (200 nm) stimulates intracellular calcium increases in DRG neurons (black line). Incubation with LTB4 before a second stimulation sensitizes intracellular calcium increases (gray line). Neurons were identified by responses to KCl (50 mm). Shown is a representative trace. B, same as A except that different LTB4 concentrations were used. Sensitization represents the ratio of the second peak by the first peak. The data are presented as means ± S.E. (n = 32–86; one-way ANOVA; Kruskal-Wallis test; Dunn's post hoc test). *, p < 0.05; ****, p < 0.0001. C and D, same as A and B except that DRGs were incubated with the BLT1 antagonist U75302 (1 μm) or the PKC inhibitor BIM (1 μm) before LTB4 treatment. The data are shown as means ± S.E. (n = 45–118). DRGs of three animals were evaluated. Statistics were derived using one-way ANOVA (Kruskal-Wallis test, Dunn's post hoc test). *, p < 0.05; ***, p < 0.001.

Next, we injected 20 ng of LTB4 in the hind paw of adult mice and determined the thermal withdrawal thresholds in these mice. This dose has previously been shown to be sufficient to evoke a maximal mechanical allodynia in mice (8, 26). Accordingly, we observed a long-lasting decrease of the paw withdrawal thresholds after LTB4 injection (Fig. 2A), which is known to depend on the recruitment of neutrophils (8). Indeed, we found increased neutrophil numbers in the paw 2 h after LTB4 injection (Fig. 2B), whereas the number of macrophages (F4–80⁺) was not altered. Importantly, 1 h after LTB4 injection, before neutrophil recruitment occurs (Fig. 2B), wild-type but not BLT1-deficient mice reacted to LTB4 injection with a decrease of thermal withdrawal thresholds (Fig. 2C). Likewise, genetic deletion of TRPV1 prevented LTB4-induced thermal hyperalgesia at this early time point (Fig. 2D). Taken together, our data show that LTB4 sensitizes TRPV1-mediated calcium increases in sensory neurons and subsequently thermal hyperalgesia through BLT1.

Figure 2.

BLT1 and TRPV1 mediate LTB4-induced thermal hyperalgesia. A, thermal withdrawal thresholds after injection of 10 μl of LTB4 (20 ng) or vehicle in one hind paw of wild-type mice. The data are shown as means ± S.E. (n = 6; two-way ANOVA; Bonferroni post correction). *, p < 0.05; **, p < 0.01. B, FACS determination of F4–80 (macrophages)- and Ly6G (neutrophils)-positive cells in the paw of untreated wild-type mice and 1 or 2 h after LTB4 injection. The data are shown as means ± S.E. (n = 4–6; one-way ANOVA; Bonferroni post hoc test). *, p < 0.05. C, thermal paw withdrawal latencies before and after injection of 10 μl of LTB4 (20 ng) in one hind paw of wild-type or BLT1 knock-out mice. The data are show as means ± S.E. (n = 7–8; two-way ANOVA; Bonferroni post correction). **, p < 0.01. D, thermal paw withdrawal latencies before and after injection of 10 μl LTB4 (20 ng) in one hind paw of wild-type or TRPV1 knock-out mice. The data are shown as means ± S.E. (n = 6; two-way ANOVA; Bonferroni post correction). ***, p < 0.001. PWL, paw withdrawal latency.

LTB4 desensitizes TRPV1-mediated calcium increases through BLT2

Because LTB4 sensitizes TRPV1-mediated calcium increases only in a very narrow concentration range and BLT1 has a higher affinity to LTB4 as compared with BLT2, we investigated whether or not activation of BLT2 is responsible for the decreased sensitization observed at higher LTB4 concentrations. The selective BLT2 agonist CAY10583 (6) alone had no effect on TRPV1-mediated calcium increases in DRG cultures but abolished the LTB4-induced sensitization of TRPV1-mediated calcium increases (Fig. 3A). Fittingly, blockage of BLT2 by the BLT2 antagonist LY255283 increased sensitization of the TRPV1-mediated calcium increases at high (1 μm) LTB4 concentrations (Fig. 3B). To investigate whether or not BLT2 activation inhibits sensitization produced by other mediators in addition to LTB4, we tested the effect of the BLT2 agonist on the sensitization induced by bradykinin and a prostaglandin E2 receptor 4 (EP4) agonist. Notably, bradykinin induces TRPV1 sensitization through the PLC/PKCϵ pathway, similar to the one used by BLT1 (16). Accordingly, the BLT2-agonist CAY10583 abolished bradykinin-induced TRPV1 sensitization (Fig. 3C). Next, we tested whether or not BLT2 activation also affects PKA-mediated TRPV1 sensitization. Therefore we used EP4 activation, which induces TRPV1 sensitization through the cAMP/PKA pathway (27). Incubation with the selective EP4 receptor agonist ONO-AE-329 (28) induced a robust sensitization of TRPV1-mediated calcium increases, which was abolished in presence of the BLT2 agonist (Fig. 3D).

Figure 3.

Activation of BLT2 causes desensitization of TRPV1-mediated calcium influx. Calcium imaging was performed on DRG cells as in Fig. 1 except that cells were incubated with agonists or antagonists 2 min before LTB4 application. A, LTB4-induced (100 nm) TRPV1 sensitization in absence and presence of the BLT-2 agonist CAY10583 (400 nm). The data are shown as means ± S.E. (n = 50–159; one-way ANOVA; Kruskal-Wallis test; Dunn's multiple comparisons test). ***, p < 0.001. B, BLT2 inhibition increases TRPV1 desensitization induced by high dose of LTB4 (1 μm). LTB4-induced (1 μm) TRPV1 desensitization in absence or presence of the BLT2 antagonist LY255283 (10 μm). The data are shown as means ± S.E. (n = 26–93; one-way ANOVA; Kruskal-Wallis test; Dunn's multiple comparisons test). ***, p < 0.001. C and D, BLT2 activation suppresses TRPV1 sensitization induced by different signaling pathways. Capsaicin-induced TRPV1 activation was sensitized using bradykinin (0.5 μm) (C) or the EP4 agonist ONO-AE-329 (500 nm) (D) in the absence or presence of the BLT2 agonist CAY10583 (400 nm). The data are shown as means ± S.E. (C, n = 47–87; D, n = 52–109; one-way ANOVA; Kruskal-Wallis test, Dunn's multiple comparisons test). *, p < 0.05; **, p < 0.01; ***, p < 0.001.

To investigate whether or not BLT2 desensitizes TRPV1-mediated calcium increases through the PKC pathway, we sensitized TRPV1 using 8-bromo-cAMP and found that PKC inhibition did not affect BLT2-induced desensitization of the TRPV1-mediated calcium increases (Fig. 4A). Notably, PKA- and PKC-mediated TRPV1 sensitization is based on TRPV1 phosphorylation (29), which can be reversed by protein phosphatase 2B (calcineurin) (17). Because of the unselective effect of BLT2 activation on PKA- and PKC-mediated TRPV1 sensitization, we hypothesized that the effect of BLT2 activation might be mediated by calcineurin. Therefore, we sensitized TRPV1-mediated calcium influx with a low LTB4 concentration (100 nm) in the absence or presence of the protein-phosphatase 2B inhibitors cyclosporin A and FK506 (tacrolimus) (30, 31). Both inhibitors were used at concentrations that had no effect on LTB4-induced sensitization of TRPV1-mediated calcium increases by itself. Both inhibitors completely abolished the BLT2-mediated desensitization (Fig. 4B), which strongly suggests that calcineurin mediates the BLT2-induced desensitization of the TRPV1-mediated calcium increase. Next, we tested whether or not BLT1 and BLT2 are expressed in sensory neurons in dorsal root ganglia. We found that BLT1 and BLT2 were expressed in CGRP- and IB4-positive cells, which represent peptidergic and non-peptidergic sensory neurons, respectively (Fig. 5, A and B).

Figure 4.

BLT2-mediated desensitization of TRPV1-mediated calcium influx is calcineurin dependent. A, BLT2 activation suppresses PKA-dependent TRPV1 sensitization by 8-bromo-cAMP independently of PKC. Capsaicin-induced TRPV1 activation was sensitized using 8-bromo-cAMP (20 μm) with or without preincubation with the PKC inhibitor BIM (1 μm) and/or the BLT2 agonist CAY10583 (400 nm). The data are shown as means ± S.E. (n = 24–73; one-way ANOVA; Kruskal-Wallis test; Dunn's multiple comparisons test). ***, p < 0.001 (comparison with vehicle); ###, p < 0.001 (comparison between conditions); #, p < 0.05 (comparison between conditions). B, capsaicin-induced TRPV1 activation was sensitized using LTB4 (100 nm) with or without preincubation with the BLT2 agonist CAY10583 (400 nm) and/or the calcineurin inhibitors cyclosporin A (CspA, 100 nm) and FK506 (10 nm). The data are shown as means ± S.E. (n = 53–123; one-way ANOVA; Kruskal-Wallis test; Dunn's multiple comparisons test). ***, p < 0.001 (comparison to vehicle); #, p < 0.05 (comparison between treatments).

Figure 5.

BLT1 and BLT2 are expressed in sensory neurons. A and B, representative images showing the colocalization of BLT1 and BLT2 with IB4 (A) and CGRP (B) in dorsal root ganglia. The white bar represents 30 μm. Arrows depict cells expressing LTB4 receptors.

BLT2 activation decreases zymosan-induced thermal hyperalgesia and mechanical allodynia

In the next step we tested whether or not BLT2 activation is able to decrease nociception. We found that intraplantar application of the BLT2 agonist CAY10583 significantly increased thermal (Fig. 6A) but not mechanical paw withdrawal latencies (Fig. 6B). In accordance with the proposed TRPV1-desensitizing mechanism, we observed no effect of the BLT2 agonist on thermal withdrawal latencies in TRPV1-deficient mice (data not shown).

Figure 6.

BLT2 activation decreases basal thermal hyperalgesia. A, thermal withdrawal thresholds after intraplantar injection of 10 μl of BLT2 agonist Cay10583 (5 μm) or vehicle (0.9% NaCl) in wild-type mice. The data are shown as means ± S.E. (n = 6; two-way ANOVA; post t test with Bonferroni correction). *, p < 0.05; **, p < 0.01; ***, p < 0.001. B, mechanical withdrawal thresholds after intraplantar injection of 10 μl of BLT2 agonist Cay10583 (5 μm) or vehicle (0.9% NaCl) in wild-type mice. The data are shown as means ± S.E. (n = 6; two-way ANOVA; post t test with Bonferroni correction). No significant differences were detected between treatments.

Next, the effect of the BLT2 agonist CAY10583 on nociceptive behavior during peripheral inflammation was tested. Zymosan injection decreases thermal and mechanical withdrawal thresholds within 2–4 h. Afterward paw withdrawal thresholds returned gradually over the following 48 h to baseline level (Fig. 7A). We administered CAY10583 intraperitoneally 4 and 24 h after the zymosan injection and found that 5 and 10 mg/kg of CAY10583 significantly reduced thermal (Fig. 7A) and mechanical paw withdrawal latencies (Fig. 7B). Notably, TRPV1-mediated mechanical pain during inflammation involves indirect mechanisms such as mechanical stimulation-induced release of mediators from epithelial cells that subsequently act on nociceptors also activating TRPV1 (14). Accordingly, TRPV1-deficient mice did not exhibit zymosan-induced thermal hyperalgesia and had a reduced mechanical allodynia (Fig. 7, C and D). In accordance with our data showing that the BLT2-agonist reduces sensitization of TRPV1-mediated calcium increases, injection of the BLT2-agonist had no significant effect on zymosan-induced mechanical allodynia in TRPV1 knock-out mice (data not shown). Finally, we determined the endogenous LTB4 concentrations in the paw at the different time points after zymosan injection via LC-MS/MS. We found significantly increased LTB4 level early (1 and 2 h) after zymosan injection, which went back to baseline level 24 h after zymosan injection (Fig. 7E). Thus, the slightly stronger antinociceptive effect of the BLT2 agonist may be based on the low level of endogenous LTB4 in the paw at this time point. Taken together, the data show that BLT2 activation is able to decrease with a high efficacy nociceptive behavior during inflammatory pain.

Figure 7.

BLT2 activation decreases zymosan-induced thermal hyperalgesia and mechanical allodynia. A, thermal thresholds after injection of 20 μl of zymosan (3 mg/ml) in one hind paw of wild-type mice. 100 μl of CAY10582 or vehicle was administered intraperitoneally at the indicated times (black arrows). The data are shows as means ± S.E. (n = 6; two-way ANOVA; Bonferroni post correction). *, p < 0.05 (comparison to vehicle); ****, p < 0.0001 (comparison to vehicle); ##, p < 0.01 (comparison between treatments). B, same as A except that mechanical paw withdrawal latencies were determined. The data are shown as means ± S.E. (n = 6; two-way ANOVA; Bonferroni post correction). *, p < 0.05; **, p < 0.01. C, thermal thresholds after injection of 20 μl of zymosan (3 mg/ml) in one hind paw of wild-type or TRPV1 knock-out mice. The data are shows as means ± S.E. (n = 6; two-way ANOVA; Bonferroni post correction). **, p < 0.01; ****, p < 0.0001. D, same as C except that mechanical thresholds were determined. Thresholds within the genotypes were compared with baseline levels using one-way ANOVA and Bonferroni post correction. *, p < 0.05; **, p < 0.001. Thresholds between the groups were compared using two-way ANOVA and Bonferroni post correction. ##, p < 0.001. E, LTB4 level in untreated or zymosan-injected paws of wild-type mice. The data are shown as means ± S.E. (n = 5; one-way ANOVA; Bonferroni post hoc test). **, p < 0.01; ***, p < 0.001. PWL, paw withdrawal latency.

Discussion

LTB4 is a common proinflammatory mediator that is released mainly by neutrophils at the site of inflammation. It is not only a powerful chemoattractant for neutrophils but has also reported to be able to sensitize nociceptors, thereby increasing nociception. In this regard it has been shown that C-polymodal nociceptors and C-mechanoheat nociceptors are sensitized by LTB4 (32). As consequence LTB4 lowers the average heat threshold from 45 °C by ∼10 °C and decreases mechanical thresholds by almost 80% (32). BLT2 activation increased thermal but not mechanical thresholds in naïve mice. In contrast BLT2 activation was able to increase both thermal and mechanical withdrawal latencies during zymosan-induced inflammation. The broader response during inflammation is not surprising, because sensitization of nociceptors during an inflammation and stimulation of non-neuronal cells allows certain afferents to acquire responsiveness to mechanical stimuli to which they are, under naïve conditions, insensitive (14, 33, 34).

Although most of the proinflammatory effects of LTB4 are based on BLT1 activation, the previously known pronociceptive effects of LTB4 were thought to be due to a weak agonism of TRPV1, which occurs at relative high concentrations (EC₅₀ = 11.7 μm) (20). Here, we found that 100-fold lower LTB4 concentrations (100 nm) sensitize TRPV1-mediated calcium increases through BLT1 and that, more surprisingly, higher LTB4 concentrations (0.4–1 μm) desensitize the TRPV1-mediated calcium influx through BLT2. Because BLT2-induced desensitization is mediated by the phosphatase calcineurin, BLT2 activation can inhibit not only BLT1-mediated sensitization but also other sensitization pathways (EP4 agonist or bradykinin) mediated by either PKA or PKC. This property makes BLT2 an interesting target for the development of novel analgesics, which aim to inhibit TRPV1-mediated pain sensations. Because of the prominent role of TRPV1 in pain perception, various direct TRPV1 inhibitors have been developed over the last years as potential analgesics. Unfortunately, so far TRPV1 inhibitors are tainted by their side effects of inducing hyperthermia (35). Because BLT2 activation is only reversing the sensitizing effects of pronociceptive inhibitors without altering basal TRPV1 activity, BLT2 agonists would circumvent this side effect.

BLT2 is highly expressed in the small intestine, colon, spleen, liver, and pancreas, as well as on T-cells, mast cells, peripheral sensory neurons, and keratinocytes (5, 6, 36, 37). Most of its functions in these tissues and cells are to date unknown or not fully understood, and BLT2 knock-outs appear as phenotypically normal. Interestingly, it was previously shown that BLT2 expression in CD4 T-cells has a protective role in allergic airway inflammation reducing the pathophysiology of asthma (22). Likewise, BLT2 is expressed in colon cryptic cells and appears to protect against dextran sodium sulfate-induced colitis, possibly by enhancing barrier function in epithelial cells of the colon (21). Thus, selective activation of BLT2 by agonists might produce, in addition to analgesia, additional beneficial effects in certain pathophysiological settings.

The fact that LTB4 concentration-dependently induces opposing effects through its different receptors is not an unusual concept in neurobiology. In this regard, it was shown that long-term potentiation is inhibited by dopamine receptor 1 receptors (49, 50) but reduced by D2 receptors (51, 52). Similarly, the neurotransmitter GABA has both inhibitory and stimulating effects on short-term memory. Low doses of GABA strongly inhibit reinforced memory seemingly activating GABA_Creceptors, whereas higher doses enhance weakly reinforced memory via GABA_A receptors (38). Finally, in midbrain dorsal periaqueductal gray neurons, anandamide can either facilitate or inhibit excitatory transmission through TRPV1 or cannabinoid receptor 1, respectively (39). Accordingly, anandamide induces both panicolytic and panicogenic effects in rats via its actions at cannabinoid receptor 1 and TRPV1, respectively (37).

In the nociceptive system, P2Y receptors (40) and serotonergic signaling (41) are examples for such antagonistic responses to the same mediator. So are P2Y₁, P2Y₁₂, P2Y₁₃, and P2Y₁₄ receptors expressed in neurons of mouse dorsal root ganglia (42) that can act either pronociceptively (P2Y₁) or antinociceptively (P2Y₁₂, P2Y₁₃, and P2Y₁₄) (40).

However, LTB4 receptors are expressed in the same sensory neurons and have opposing roles in the regulation of TRPV1-mediated calcium increases. In the above mentioned examples for self-restricting signaling pathways in neurons, they serve to limit the consequences of an excessive activation of the affected signal pathway. In case of anandamide it prevents overshooting anxiety and panicogenic behavior, whereas in case of dopamine and GABA receptors, it restricts long-term potentiation development and, subsequently, modulates short term memory. For P2Y receptors it is believed that under normal conditions their opposing effects are balanced and that ADP increases during an inflammation shift this balance toward the pronociceptive effects (40). Considering these findings, the antagonism of the two LTB4 receptors may serve to restrict the input of a single pronociceptive mediator in nociceptor sensitization. This effect would be especially important for mediators such as LTB4 that are synthesized temporarily in high amounts to allow neutrophil recruitment to inflammation sites. These peak LTB4 concentrations could produce through BLT1 an excessive sensitization of peripheral sensory neurons that is limited or even reversed by activation of BLT2.

Experimental procedures

Animals

BLT1 and TRPV1 knock-out mice on C57BL/6J background were supplied by The Jackson Laboratory. Wild-type mice were supplied by Janvier (Le Genest, France). All mice were compared with age- and sex-matched wild-type mice. The animals were cared for corresponding to International Association for the Study of Pain (IASP) guidelines. For all experiments the ethics guidelines for investigations in conscious animals were obeyed, and the procedures were approved by the local ethics committee.

Reagents

LTB4, CAY10583, U75302, and LY255283 were from Cayman Chemicals (Ann Arbor, MI). Capsaicin was from Sigma-Aldrich, and ONO-AE-329 was a kind gift from Prof. R.M. Nüsing (Hospital of the Goethe University, Frankfurt, Germany). BIM, 8-bromo-cAMP, and cyclosporin A were purchased from Biomol (Hamburg, Germany); zymosan A was from Sigma-Aldrich; and bradykinin was from Tocris (Bristol, UK). Anti-BLT1 and anti-BLT2 antibodies were from Hölzel Diagnostika (Köln, Germany), anti-TRPV1 antibody was from Alomone Labs, FITC-labeled IB4 was from Sigma-Aldrich, and anti-cGRP antibody was from Alpha Diagnostics International Inc.

GPCR expression analysis

Dorsal root ganglia were dissected from adult mice on both sides of the intervertebral foramina from the segments S1 up to C3 and frozen immediately. RNA isolation and transcription were performed by using the RNeasy kit (Qiagen) according to the manufacturer's instructions. A total of 10 μg of RNA was isolated from 10 mice, pooled, and used as template for semiquantitative RT-PCR of 546 murine GPCRs as described previously (43). Semiquantitative RT-PCR was done by using primers designed with the Roche's online tool (Table 2), and quantification and data analysis were performed by using the LightCycler 480 Probe Master System (Roche). 25 ng of cDNA/reaction were used to generate Ct values. 6 ng of genomic DNA was used with the same primer pair as standard to calculate gene copy number per nanogram RNA. The housekeeping gene (Gapdh) was used as a control for cDNA preparation and PCRs. To show the relative abundance of each GPCR, relative expression values were calculated by taking into account the copy number of Gapdh/ng and expressed as ratio between copy number of target gene and of Gapdh (Table 1).

Table 2.

List of primers for GPCRs named in Table I

Gene	Forward primer	Reverse primer
PTGER3	GAAGTCTTTCCTGCTGTGCAT	AGGAGCTGCCCCACTAAGTC
ADRA2C	aagcggtagagtacaacctgaag	agcggagatgagccacac
GPR88	GCCACCGCAAGTCCACTA	CGATGCCCAGGTAGCAGT
ADRA2A	ggtgacactgacgctggtt	gcacgttgccaaatactgtg
GPR153	CGGTTCTCCCACGATGAT	CCAGAGCTGCCAAGTCCT
ADRB1	catcatgggtgtgttcacg	ggaaagccttcaccacgtt
BLTR2	CGCCTGGGAATGAGACAC	GCCGCCAGCAGTAGAAAG
VIPR2	cagtgtgaactgaaaagaagatgg	agctgtgcagccggtagt
OXTR	CATCACCTTCCGCTTCTACG	CCACCTGCAAGTATTTGACCA
TAS1R1	gcttcctcatgctgggttc	tcccgaagaagctgtagagg
NPFFR1	TGCTGCTGCTCATCGACTAC	AGAAGGCCAGCCAGTGTG
FZD8	gtagcatgtacacgcccatc	gttcacacacagagcgacaag
FZD5	TACTGGGTGCTCATGCTCAA	ATCCAGACTCCCGACGTG
BLTR	ctgctggtgctgaacttgg	gaggaaaaagggagcagtga
GPR62	GGCGGTTCTCGTAGAGGTG	CTAAGTAGAAGGCATCTTGAAGGTC
ADRA1D	actgagtgcagctgtgttgc	gaaggcccagaagcctagaa
CRHR2	AACACGACCTTGGACCAGAT	GGGGCACGGTCTCTCTACTA
GPR26	gctgctggtgggcactat	caggaggcagagcagcac
GPR25	CGAACCCTGGAGTCCTAGC	GGCACAACTCCACTTGGTC
ADRB3	ctgctagcatcgagaccttgt	aagggttggtgacagctagg
GPR137B	GTGGCAGCCAACTCGTTC	AGACACACGGGGAAGCAGTA
GPR54	agtgcgcaccaaggtctc	gcttgaagcaccaggaaca
LGR6	TGTCAAGTCAGTCCTTCTGGTG	CAGGTAGAGCAGTGGGTTGAG
Vmn2r54	ttacataaacagggtgatgagagg	agcaggaaaaacagggtgag
FZD7	CGTCTTCAGCGTGCTCTACA	TCATAAAAGTAGCAGGCCAACA
HTR6	acgtcggacctgatggtg	catacagcgcgttcagca
GPR135	CCGTGCTCATCATGATTATCTTT	GCACCAGGAAGCAGTAGGG

Primary mouse dorsal root ganglia cultures

As previously described (24, 44) murine dorsal root ganglia were dissected on both sides of the intervertebral foramina from the segments S1 up to C3 and immediately stored in ice-cold Hanks' balanced salt solution containing Ca²⁺ and Mg²⁺. Afterward, the dorsal root ganglia were incubated with collagenase/dispase (500 units/ml collagenase (Biochrom AG, Berlin, Germany); 2.5 units/ml dispase (Roche) in neurobasal medium containing penicillin (100 units/ml), and streptomycin (100 g/ml), (all from Invitrogen) at 37 °C for 75 min. Dorsal root ganglia were washed two times with neurobasal medium containing 10% FCS and incubated with 0.05% trypsin/EDTA for 10 min. Washes were repeated twice, and the tissues were dissociated with a 1-ml Gilson pipette in 300 μl of neurobasal medium with 10% FCS, penicillin (100 units/ml), and streptomycin (100 μg/ml). 25 μl of the cell suspension was seeded on poly-l-lysine-coated coverslips and incubated for 2 h, before the addition of 1 ml of neurobasal medium containing l-glutamine (2 mm), B-27 (all from Invitrogen), and gentamicin (50 μg/ml) (BioWhittaker Europe, Verviers, Belgium).

Calcium-imaging experiments

Experiments were done 24 h after DRG cell preparation. The cells were loaded with Fura-2-AM for 45 min with medium containing 5 μm Fura-2-AM, 0.002% pluronic F-127 (both Biotinum). For determination of TRPV1 sensitization, the cells were stimulated twice with 200 nm capsaicin in Hanks' balanced salt solution (145 mm NaCl, 1.25 mm CaCl₂, 1 mm MgCl₂, 5 mm KCl, 10 mm d-glucose, and 10 mm HEPES, adjusted to pH 7.3, all from Sigma-Aldrich). For calcium imaging a Leica calcium-imaging setup, composed of a Leica DMI 4000 microscope equipped with a DFC360 FX (CCD) camera, Fura-2 filters, and an N-Plan 10/0.25 Ph1 objective, was used as previously described (24, 44). For recordings and calculation, the LAS AF software (Leica) was used. Sensitization of TRPV1-mediated calcium increases by LTB4 is expressed as the ratio of the second peak (cells treated with LTB4 and capsaicin) by the first peak (cells stimulated with capsaicin alone). Treatments with inhibitors, agonists, or antagonists were performed for 2 min before the second capsaicin stimulation, except for BIM, which was incubated for 45 min prior the experiment.

Behavioral tests

The experimenter was unaware of the treatments or the genotypes during all behavioral experiments. Mechanical withdrawal thresholds of the plantar side of a hind paw were determined using a plantar aesthesiometer (Dynamic Plantar Aesthesiometer, Ugo Basile). A steel rod (20-mm diameter) was pushed against the paw with ascending force until a strong and immediate withdrawal occurred. Thermal withdrawal thresholds were determined as described previously (45). The paw withdrawal latency was taken to be the mean of at least four consecutive trials with at least 20 s in between (24, 46). Peripheral inflammation was induced by subcutaneous injection of 20 μl of 3 mg/ml zymosan into the plantar side of one hind paw (47). For animal experiments (but not for other experiments), the BLT2 agonist CAY10583 was suspended in a solution of 1% methylcellulose and 1% soybean lecithin (both from Sigma-Aldrich) to avoid high concentrations of DMSO.

FACS analysis and sorting

The cells were isolated from zymosan-induced edemas and prepared for FACS analysis as previously described (48). FACS analysis was performed on a FACS Canto flow cytometer (Becton Dickinson). The cells were stained with PE-labeled anti-F4/80 (Biolegend) and FITC-labeled anti-Ly6G (eBiosciences) antibodies.

Liquid chromatography-tandem mass spectrometry

LC-MS/MS analysis of LTB4 from paws was performed as described previously (26). Paws were prepared at different time points after zymosan injection paws. Standard stock solutions with 2500 ng/ml of LTB4 and internal standard with 25 mg/ml of LTB4-d4 were prepared in methanol. Working standards were obtained by further dilution with a final concentration range of 0.1-250 ng/ml. Sample treatment was performed using liquid-liquid extraction. Therefore, homogenized tissue was extracted twice with 600 μl of ethyl acetate. The combined organic phases were removed at a temperature of 45 °C under a gentle stream of nitrogen. The residues were reconstituted with 50 μl of methanol/water/butylated hydroxytoluene (50:50:0.1, v/v/v), centrifuged for 2 min at 10,000 × g, and stored in glass vials.

The LC-MS/MS system consisted of a QTrap 5500 (AB Sciex, Darmstadt, Germany) equipped with a Turbo-V source operating in negative electrospray ionization mode, an Agilent 1200 binary HPLC pump and degasser (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (CTC Analytics, Zwingen, Switzerland). High-purity nitrogen for the mass spectrometer was produced by a NGM 22-LC-MS nitrogen generator (CMC Instruments, Eschborn, Germany). For the chromatographic separation a Gemini NX C18 column and precolumn was used (150 × 2-mm inner diameter, 5-μm particle size, and 110 Å pore size; Phenomenex, Aschaffenburg, Germany). A linear gradient was used at a flow rate of 0.5 ml/min with a total run time of 17.5 min. Mobile phase A consists of water:ammonia (100:0.05, v/v), and mobile phase B consists of acetonitrile:ammonia (100:0.05, v/v). The gradient changed from 85% A to 10% within 12 min. These conditions were held for 1 min. Then the mobile phase shifted back to 85% A within 0.5 min, and it was maintained for 4 min to re-equilibrate the column. 20 μl of the extracted samples were injected into the LC-MS/MS system. Quantification was performed with Analyst software version 1.5 (AB Sciex) using the internal standard method (isotope-dilution mass spectrometry). Ratios of analyte peak area and internal standard area (y axis) were plotted against concentration (x axis), and calibration curves were calculated by least-squares regression with 1/concentration² weighting.

Immunohistology

DRG sections were fixed for 10 min in 4% paraformaldehyde in PBS, permeabilized for 10 min with 0.1% Triton in PBS, blocked with 3% BSA in PBS for 1 h at room temperature, and incubated with antibodies against BLT1 or BLT2 (Bioss) and CGRP (Abcam) or FITC-labeled lectin IB4 (Sigma-Aldrich) followed by a AF488 or Cy5-labeled secondary antibody (Sigma-Aldrich).

Statistics

All statistics were calculated with GraphPad Prism 6 software (GraphPad Software, Inc.). Gaussian distribution was tested with D'Agostino-Pearson normality test. Differences between two independent groups were tested with Student's t test in case of Gaussian distribution. To compare more than two groups in repeated measurements, a one-way ANOVA was performed. For nonparametric testing of more than two groups, the Kruskal-Wallis test with the Dunn's multiple comparisons test was done. Results from behavior experiments were tested with a two-way ANOVA with Bonferroni post-test correction.

Author contributions

All authors were involved in planning experiments and discussion and interpretation of the data and critically revised and approved the manuscript. S. Z., A. L., and M. S. performed calcium imaging experiments; M. S., K. K., E.-M. T., J. S., H. J., and N. D. performed behavioral experiments; S. T. and S. O. did the GPCR expression screen; S. P. and J. S. performed FACS analysis; S. P. and K. S. did MELC analyses; C. A. and N. F. did LTB4 measurements; G. G. and K. S. designed the experiments; and S. Z., S. P., and K. S. wrote the manuscript.