Prostaglandin D 2 improves IL-31-induced alloknesis : itch-stimulation becomes pain-stimulation in mouse skin

We examined the effect of prostaglandin D2 (PGD2) on IL-31-induced alloknesis in mice. We investigated itch-associated scratching behavior (long-lasting scratching: LLS) as an indicator of itching. Topically applied PGD2 and dexamethasone each significantly suppressed IL-31 (acute or repeated administration)-induced LLS. However, neither PGD2 nor dexamethasone suppressed the IL-31 (repeated administration)-induced increase in IL-31 receptor A mRNA expression in dorsal root ganglia. Next, we investigated the scratching behavior induced by pruritogens (histamine, serotonin and compound 48/80) or algogens (acetylcholine, bradykinin and capsaicin) in IL-31-induced itchy-skin mice. In naïve mice, these irritants caused many counts of short-lasting scratching (SLS) and a few counts of LLS for 60 min. In contrast, IL-31-induced itchy-skin mice showed significant increases in LLS counts compared with naïve mice. Topical application of PGD2 significantly suppressed LLS counts induced by these pruritogens or algogens in IL-31-induced itchy-skin mice. The anti-pruritic efficacy of PGD2 may be related to its ability to reverse the hyperkinesis of primary afferents; thus, PGD2 improves IL-31-induced alloknesis. In a hot-plate test, topical application of PGs (PGD2, PGE2 and PGI2) caused a decrease in the nociceptive threshold in naïve mice, indicating that PGs enhance pain. These results suggest that PGs (especially PGD2) can change itch-stimulation into pain-stimulation in itchy-skin mice caused by pretreatment with IL-31. Therefore, it is possible that IL-31 and PGs may play a role in the regulation of a primary sensory nerves for the sensation of itch and pain.


Introduction
Recent studies on itch have been based on the human pruriceptive sense, and have found no discernable differences between the nociceptive stimuli examined.Since itching elicits a strong desire to scratch, the measurement of scratching is useful for evaluating itching [1] .In a previous investigation that measured spontaneous scratching in NC/Nga mice, an animal model of atopic dermatitis [2] , two kinds of scratching behavior were observed: long-lasting

RESEARCH ARTICLE
scratching (LLS, over 1.5 s) and short-lasting scratching (SLS, 0.3 -1.5 s).In that study, LLS was frequently seen in spontaneous skin-lesioned NC/Nga mice, but not in other strains of mice.In contrast, SLS was frequently seen in both skin-lesioned NC/Nga mice and other strains of mice.These results suggest that SLS is a form of social and/or hygiene behavior, while LLS is the true itching response in these mice.Therefore, we investigated LLS as an indicator of itching [3,4] .
Previously, we reported that topical application of prostaglandins (PGs), especially PGD2, significantly suppressed LLS counts in skin-lesioned NC/Nga mice, via a specific prostanoid DP1 receptor [5] .It is known that the itch sensation can be reduced by the painful sensations caused by scratching.The inhibition of itch by painful stimuli has been experimentally demonstrated by the use of various painful stimuli.We demonstrated that the scratching of mouse skin with a stainless steel wire brush (mechanical scratching) was associated with a significant elevation of cutaneous PG levels [6]   , and these PGs suppressed the itch-associated scratching behavior (LLS) observed in mite-infected mice [7] .On the other hand, it is well known that PGs are associated with inflammation, and their administration was found to enhance pain [8] .
Touch-or brush-evoked pruritus around an itching site has been termed "alloknesis" [9,10] , whereas pin prick-evoked itching sensations around an itching site have been termed "hyperknesis".Recent studies in patients with chronic itching have demonstrated that repetitive application of painful stimuli, such as electrical or noxious heat stimuli, or scratching stimuli distal to an itchy stimuli, may be perceived as an itch [11,12] .This may also explain why scratching aggravates itching and induces a vicious cycle of scratching-induced itching [13] .Clarification of the precise mechanisms of itchy skin could have major therapeutic implications, but a suitable animal model for studying this phenomenon does not yet exist.Pruritus is an important symptom of atopic dermatitis; however, the major pruritogen(s) have not yet been identified.Interleukin-31 (IL-31) is a possible mediator of itching, and induces both severe pruritus and dermatitis in mice [14] .Recently, we reported that intradermal injection of pruritogens and algogens in BALB/c mice that were cohoused with skin-lesioned NC/Nga mice or pretreated with IL-31 increased itch-associated scratching (LLS) counts, and suggested that this phenomenon might be similar to "alloknesis" or "hyperknesis" [15] .However, the sites of action of Il-31 have not been clarified.In this study, we investigated the relationship between IL-31, an alloknesis-inducer, and PGD2, an allodynia-inducer, to elucidate the regulatory mechanism of itch and pain.

Animals
Male 8-week-old BALB/c mice were purchased from SLC Japan (Shizuoka, Japan).The animals were housed under conditions of controlled temperature (23±3 ℃ ), humidity (50±20 %) and lighting (lights on from 7:00 am to 7:00 pm).All animals were given free access to food and tap water.All procedures for animal experiments were approved by the Committee for Animal Experimentation at the International University of Health and Welfare and were in accordance with the Guidelines for Proper Conduct of Animal Experiments (Science Council of Japan, 2006).

Measurement of scratching counts
Scratching was measured as described previously [17] .The number of scratches was detected automatically and evaluated objectively using MicroAct (Neuroscience, Tokyo, Japan) [18] .The analysis parameters for detecting waves were Threshold; 0.1 V, Event gap; 0.2 s, Minimum duration; 1.5 s, Maximum frequency; 20 Hz, and Minimum frequency; 2 Hz.

Effect of PGD2 on IL-31 (acute administration)-induced itch-associated scratching behavior in mice
A single intradermal injection of IL-31 (1g/site, i.d.) was carried out at 10:00 AM.This dosage is based on our previous report [15] .PGD2 (0.1 %) was dissolved in EtOH and applied topically at 1:00 PM to the rostral part of the back of mice at 3 h after IL-31 injection.The scratch count was measured as described above and compared between groups.

Quantitative real-time PCR
Total RNA was extracted from the dorsal skin of each mouse by Trizol (Invitrogen-Carlsbad, CA, USA) and digested using amplification-grade DNase I (Invitrogen), according to the manufacturer's instructions.cDNA was synthesized by the SuperScript III First-Strand Synthesis System (Invitrogen).Quantitative real-time PCR was performed with SYBR Green Master Mix, using an Applied Biosystems 7700 Sequence Detection System (Applied Biosystems, Foster City, CA, USA).The PCR primers for IL-31 were designed using PRIMER 3 software, and primers for -actin were purchased from TAKARA BIO (Otsu, Shiga, Japan).Primer sequences were as follows: IL-31RA ('5-CCA GAA GCT GCC ATG TCG AA-3' and 5'-TCT CCA ACT CGG TGT CCC AAC-3'), -actin (5'-TGA CAG GAT GCA GAA GGA GA-3' and 5'-GCT GGA AGG TGG ACA GTG AG-3').Relative expression levels were calculated by the relative standard curve method as outlined in the manufacturer's technical bulletin.A standard curve was generated using fluorescent data from four-fold serial dilutions of total RNA of the sample with the highest expression.This was then used to calculate the relative amounts of target mRNA in test samples.Quantities of all targets in test samples were normalized to the corresponding -actin RNA transcript in the skin samples.

Measurement of the nociceptive effect of prostaglandins
The hot-plate test was adapted from Eddy and Leimbach (1953).Mice were placed on a hot-plate maintained at 51 ± 0.5 ℃, and the latency to either paw-lick or an attempt to escape by jumping was recorded.PGD2, PGE2 and PGI2 dissolved in EtOH at 0.1% (w/v) were topically applied to the footpad of mice.Animals were tested before and 30, 60, 90, 120, 150, 180 min after the application of PGs.To prevent tissue damage, mice that showed no response within 60 sec were removed from the hot-plate and assigned a score of 60 sec.The percentage of nociception (nociceptive index) was calculated according to the formula: [(T1-T0)/(T2-T0)]x100, where T0 and T1 were the latencies observed before and after the application of PGs, respectively, and T2 was the cut-off time (60 sec).

Data analysis
Parameters are reported as the mean  S.E.Nociceptive effects are represented by the area under the curve that was calculated for each mouse by the trapezoidal method and expressed as a mean percentage for the vehicle-treated group.The parametric Student's t-test with the Bonferroni correction was used to analyze the data.P values of less than 0.05 were considered statistically significant.

Discussion
Itch-associated scratching behavior is an important symptom for the development of dermatitis in NC/Nga mice, but the major pruritogens have not yet been identified.We previously reported that LLS counts correlated with the dermatitis score in NC/Nga mice, and LLS was inhibited by dexamethasone or PGD2, but not by ketotifen [3,4] .In this study, PGD2 and dexamethasone significantly suppressed IL-31-induced LLS counts, while ketotifen had no effect.These effects are similar to those on spontaneous scratching behavior in NC/Nga mice, suggesting that IL-31 participates in the sensation of itching and promotes LLS in skin-lesioned NC/Nga mice.However, neither PGD2 nor dexamethasone suppressed the expression of DRG IL-31RA mRNA in mice.This result indicates that the suppressive effect of PGD2 on IL-31-induced LLS was not due to a decrease in the expression of DRG IL-31 receptor A. It has been reported that mechanical scratching of the skin significantly increases the cutaneous PGD2 level [6] and the cutaneous application of PGD2 significantly suppresses LLS in skin-lesioned NC/Nga mice [7] .Considering these reports, the present findings suggest that PGD2 may keep an itch in check through functional antagonism for alloknesis or hyperknesis caused by IL-31.Tissue injury results in the release of many chemical mediators, such as ATP, bradykinin, amines, protons, prostanoids, cytokines and peptide [19] .These inflammatory mediators also act to modify the response properties of primary afferent neurons to subsequent stimuli (peripheral sensitization).Alternatively, the responses to noxious stimuli may be enhanced (hyperalgesia) or normally innocuous stimuli may produce pain (allodynia).PGE2 and PGI2 have been shown to influence inflammation, and their administration was found to reproduce the major signs of inflammation including augmented pain [8,20] .They are synthesized by the constitutive enzyme cyclooxygenase-1 (COX-1) and its isoform enzyme COX-2, which is induced in peripheral tissue by inflammatory stimuli [21] .Interestingly, it has been reported that COX-1-deficient mice show reduced nociceptive activity [22] .On the other hand, the topical application of PGD2, PGE2 and PGI2 significantly suppressed LLS in skin-lesioned NC/Nga mice, and their inhibitory activities were in the order PGD2 >> PGI2 > PGE2.In this study, PGD2 significantly suppressed IL-31-induced LLS.In addition, PGD2, PGE2 and PGI2 increased nociceptive effects and their nociceptive activities were in the order PGD2 > PGI2 > PGE2, which is the same as the order of their anti-pruritic activities [5] .Our previous data showed that topical application of the selective COX-1 inhibitor CS-560, but not the selective COX-2 inhibitor NS-398, clearly increased LLS [23] .Furthermore, the cutaneous level of PGD2 in COX-1-deficient mice was significantly lower than that in wild type, while the cutaneous levels of PGE2, PGF and PGI2 in COX-1-deficient mice were almost the same as those in wild type [24] .These previous and present findings suggest that cutaneous PGD2 could be mainly produced by COX-1, and may play a critical role in the regulation of the sensation of both itch and pain.
Recently, we investigated LLS induced by several pruritogens and algogens i.e., histamine, serotonin, compound 48/80, acetylcholine, bradykinin and capsaicin, in mice with itchy skin.In NC/Nga mice with itchy skin caused by mite-infestation, LLS induced by pruritogens (histamine, serotonin and compound 48/80) or algogens (acetylcholine, bradykinin and capsaicin) was significantly increased compared to that in non-mite-infested mice [15] .NC/Nga mice with itchy skin caused by mite-infestation showed increased cutaneous IL-31 mRNA expression [25,26] .Based on these previous findings, in the present study we examined the effects of pruritogens and algogens on LLS in mice that had been systemically pretreated with IL-31.As a result, in mice with itchy skin caused by intravenous pretreatment with IL-31, not only pruritogens but also algogens significantly increased LLS compared to that in vehicle-pretreated mice.These previous and present findings suggest that IL-31 may cause alloknesis or hyperknesis, i.e., it changes non-selective irritant stimulation into itch-stimulation in mouse skin.Therefore, it is possible that pain and itch are transmitted on the same nerve fibers, and a sensation is perceived as pain or itch depending on the operation of IL-31.Furthermore, the present study also demonstrated that PGD2 decreased LLS counts induced by the cutaneous injection of pruritogens or algogens in itchy skin caused by IL-31.These results indicate that PGD2 improves IL-31-induced alloknesis or hyperknesis, i.e., it changes non-selective irritant stimulation into pain-stimulation in mouse skin.If we consider all of these data together, the sensation of pain and itch may be regulated by PGD2 (allodynia-inducer) and/or IL-31 (alloknesis-inducer), through their functional antagonism (Fig. 6).For example, when a mite irritates the skin, IL-31 is expressed in that region of the skin, and an itch sensation is produced in response to various kinds of cutaneous stimulation.On the other hand, PGD2 is produced in response to inflammation such as that in response to a burn, and this may be involved in the onset of pain in response to various kinds of cutaneous stimulation.

Figure 1 .
Figure 1.Effect of PGD2 on IL-31-induced itch-associated scratching behavior in mice.LLS; long-lasting scratching behavior (over 1.5 s) induced by IL-31 (1 g/site, i.d.) was measured for 24 h in BALB/c mice (A).A red arrow-shaped mark indicates the site of injection of vehicle or IL-31.A blue arrow-shaped mark indicates the site of topical application of EtOH or PGD2.Green line indicate vehicle + EtOH-treated mice, purple line indicate vehicle + PGD2-treated mice, blue line indicate IL-31 + EtOH-treated mice, and orange line indicate IL-31 + PGD2-treated mice.The lateral axis indicates the clock hour, and the shaded area represents the dark phase (7:00 pm to 7:00 am).Total LLS counts for 24 h after the injection of IL-31 (B).The green column indicates vehicle + EtOH-treated mice, the purple column indicates vehicle + PGD2-treated mice, the blue column indicates IL-31 + EtOH-treated mice, and the orange column indicates IL-31 + PGD2-treated mice.Each value represents the Mean±S.E.*P<0.05 and ***P<0.001compared with the respective values in each vehicle-treated group.

Figure 5 .
Figure 5.Effect of PGs on the pain threshold in response to thermal stimuli estimated by the hot-plate test in mice.Latency: the time until either paw-lick or jumping (A).Nociceptive index; the percentage of the nociceptive area from 30-120 min was calculated according to the latencies obtained before and after PGs application (B).A red arrow-shaped mark indicates the site of application of ethanol (vehicle) or PGD2.Blue line indicate ethanol (vehicle)-treated mice, orange line indicate prostaglandin D2 (PGD2)-treated mice, green line indicate prostaglandin E2 (PGE2)-treated mice, and purple line indicate prostaglandin I2 (PGI2)-treated mice.Each value represents the Mean±S.E.*P<0.05,**P<0.01 and ***P<0.001compared with the vehicle-applied group.

Figure 6 .
Figure 6.Schematic diagram of the putative roles of IL-31 and PGD2 in the regulation of the sensation of cutaneous itch and pain in mice.IL-31 can change non-selective-stimulation into itch-stimulation.In contrast, PGD2 can change non-selective-stimulation into pain-stimulation transmitted by the primary nerves of C-fibers and by second-order nerves and spinothalamic tract neurons in the spinal cord.This suggests that IL-31 and PGD2 regulate the perception of sense (pain or itch) through their mutual functional antagonism.