The Serotoninergic Receptors of Human Dendritic Cells: Identification and Coupling to Cytokine Release ¹

Serotonin (5-hydroxytryptamine (5-HT)) ⁴ is a well-characterized neurotransmitter and vasoactive amine involved in the regulation of a large number of physiological functions such as sleep, appetite, and behavior (1, 2). 5-HT has also immunomodulatory effects by regulating a wide variety of cell responses such as migration, phagocytosis, superoxide anion generation, and cytokine production (3, 4, 5, 6). 5-HT is released at inflammatory sites by IgE-activated mast cells and platelets (7), and recent findings point to a role of 5-HT in the pathophysiology of asthma (8). The wide variety of 5-HT-mediated functions is paralleled by the high pharmacological complexity of responses due to existence of different classes of serotoninergic receptors (5-HTR) (9). The 5-HTR₁ class consists of at least five subtypes named 5-HTR_1A, 5-HTR_1B, 5-HTR_1D, 5-HTR_1E, and 5-HTR_1F. 5-HTR_1A interacts with several G proteins eliciting different responses (10, 11, 12). 5-HTR_1B and 5-HTR_1D are coupled to formation of inositol phosphates through interaction with pertussis toxin-sensitive G_i/o and pertussis toxin-insensitive Gα₁₅ proteins (12, 13). The G protein-coupled 5-HTR₂ class includes three different subtypes: 5-HTR_2A, 5-HTR_2B, and 5-HTR_2C (14, 15, 16, 17). 5-HTR₃ receptors are ligand-gated cation channels triggering depolarization of the plasma membrane through activation of Na⁺ and K⁺ fluxes (18). The 5-HTR₄ receptor has two splice variants (5-HTR_4a and 5-HTR_4b) (19, 20). The heptahelical 5-HTR₅ subtype is less characterized (21). 5-HTR₆ and 5-HTR₇ are linked to G_s protein-mediated stimulation of adenylyl cyclase (22, 23, 24, 25, 26, 27, 28, 29).

Dendritic cells (DC) are APCs specialized in activating naive T lymphocytes to initiate primary immune responses (30, 31). DC originate from hemopoietic stem cells that migrate into target sites to capture Ags. During circulation through the body DC undergo maturation, a process that entails acquisition of high levels of surface MHC and costimulatory molecules, expression of different chemokines, and production of cytokines. DC migrate to secondary lymphoid organs where they play a crucial role in the development of Th1/Th2-modulated immune responses through release of cytokines and chemokines (32). DC also produce several proinflammatory cytokines such as TNF-α, IL-1β, IL-6, and IL-8 that profoundly affect the outcome of inflammatory reactions (33). In this study, we characterized the biological activity of 5-HT in DC, showing that 5-HT-mediated responses depend on the differentiation stage of DC. 5-HT induced Ca²⁺ mobilization from intracellular stores in immature, but not in LPS-matured, DC. On the contrary, 5-HT triggered, in mature DC, Ca²⁺ influx through the plasma membrane, cAMP increase, IL-1β and IL-8 release, while it reduced secretion of IL-12 and TNF-α.

5-HT, 5-methoxytryptamine (2-MHT), N-methyl-5-HT (2Me5HT), R-(-)-DOI-hydrochloride (DOI), ketanserin, recombinant human complement fragment 5a (C5a), pertussis toxin (PTX), and lysophosphatidylcholine were obtained from Sigma-Aldrich (Deisenhofen, Germany); 5-carboxamidotryptamine maleate (5-CT), BRL-54443, 8-hydroxy-DPAT-hydrobromide (8-HDPAT), anpirtoline hydrochloride (AnHCL), pimozide, RS-39604 hydrochloride, and SB-269970 hydrochloride were purchased from Tocris (Bristol, U.K.). Macrophage inflammatory protein-3β/chemokine ligand 19 (MIP-3β/CCL19) from PeproTech (London, U.K.).

Peripheral mononuclear cells were separated from buffy coats using a Ficoll gradient. After separation, the leukocyte-containing pellet was resuspended in 2 ml of PBS containing 0.15% EDTA and 0.5% BSA. Cells were separated with anti-CD14 mAb-coated MicroBeads using Macs single use separation columns from Miltenyi Biotec (Bergisch Gladbach, Germany). The CD14⁺ cells were cultured for 5 days in RPMI 1640 medium containing 10% FCS, 1% glutamine, 50 IU/ml penicillin, 50 μg/ml streptomycin, 1,000 U/ml IL-4, and 10,000 U/ml GM-CSF (Promocell, Heidelberg, Germany) at 37°C in a humidified atmosphere with 5% CO₂. These cells were CD14^neg, CD1a^pos, CD80^low, CD83^low, CD86^low, and >95% CD115^high and are also referred to as immature DC. Maturation of DC was induced by a 24 h incubation in the presence of 3 μg/ml LPS (LPS; Sigma-Aldrich). Mature DC were >95% CD80^high, CD86^highCD83^high, and CD115^low. mAbs and their respective isotype controls were from Coulter-Immunotech (Krefeld, Germany).

The mRNA was isolated with QIAshredder and RNeasy kits (Qiagen, Hilden, Germany). mRNA, Moloney murine leukemia virus reverse transcriptase and pd(N)₆ primers (Life Technologies, Gaithersburg, MD) were used to obtain cDNA. All oligonucleotides used as primers in PCR were designed to recognize sequences specific for each target cDNA. Primer sequences are as follows: 5-HTR_1A (411-bp product): sense: 5′-GCC GCG TGC GCT CAT CTC G-3′, antisense: 5′-GCG GCG CCA TCG TCA CCT T-3′; 5-HTR_1B (460-bp product): sense: 5′-CAG CGC CAA GGA CTA CAT TTA CCA-3′, antisense: 5′-GAA GAA GGG CGG CAG CGA GAT AGA-3′; 5-HTR_1E (461-bp product): sense: 5′-CAA GAG GGC CGC GCT GAT GAT-3′; antisense: 5′-CTG CCT TCC GTT CCC TGG TGG TGC TA-3′; 5-HTR_2A (359-bp product): sense: 5′-ACT CGC CGA TGA TAA CTT TGT CCT-3′, antisense: 5′-TGA CGG CCA TGA TGT TTG TGA T-3′; 5-HTR_2B (416-bp product): sense: 5′-GGC CCC TCC CAC TTG TTC T-3′, antisense: 5′-TAG GCG TTG AGG TGG CTT GTT-3′; 5-HTR_2C (449-bp product): sense: 5′-TGT GCC CCG TCT GGA TTT CTT TAG-3′, antisense: 5′-CTC TTC CTC GGC CGT ATT CCT CTT-3′; 5-HTR₃ (448/352 bp) sense: 5′-CCG GCG GCC CCT CTT CTA T-3′, antisense: 5′-GCA AAG TAG CCA GGC GAT TCT CT-3′; 5-HTR₄ (411 bp) sense: 5′-GGC CTT CTA CAT CCC ATT TCT CCT-3′, antisense: 5′-CTT CGG TAG CGC TCA TCA TCA CA-3′; 5-HTR₆ (342 bp) sense: 5′-CCG CCG GCC ATG CTG AAC G-3′, antisense: 5′-GCC CGA CGC CAC AAG GAC AAA AG-3′; 5-HTR₇ (436 bp) sense: 5′-GCG CTG GCC GAC CTC TC-3′, antisense: 5′-TCT TCC TGG CAG CCT TGT AAA TCT-3′; β₂-microglobulin (259 bp): sense: 5′-CCT TGA GGC TAT CCA GCG TA-3′, antisense: 5′-GTT CAC ACG GCA GGC ATA CT-3′.

Thirty PCR cycles were run at 94°C (denaturation, 1 min), 62°C (annealing, 1 min), and 72°C (extension, 1 min). The generated products were subjected to electrophoresis on a 2% agarose gel and visualized by ethidium bromide staining. Intensity of the different bands in PCR gels was quantified by measuring the OD with a OneDscan computer software package (Scanalytics, Fairfax, VA). The cDNA amplification was linear in an amplification range of 24–34 cycles. The identity of the PCR products was confirmed by sequencing after cloning using pCRII vectors. Controls run without reverse transcriptase yielded no PCR products.

Total RNA was extracted using the RNeasy kit according to the manufacturer’s protocol (Qiagen). Briefly, DNase I (Invitrogen, San Diego, CA) treatment, 1 μg of total RNA from each sample, was used as a template for the reverse transcription reaction. Fifty microliters of cDNA were synthesized using M-MLV reverse transcriptase and pd(N)₆ primers (Life Technologies). All samples were reverse transcribed under the same conditions (25°C for 10 min, 48°C for 30 min) and from the same reverse transcription master mix to minimize differences in reverse transcription efficiency. All oligonucleotide primers for real-time PCR were designed using Primer 3 software (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3www.cgi) and synthesized by Invitrogen.

For iCycler reaction, a master mix of the following compounds was prepared to the indicated end concentration: 10 μl of SYBR Green master mix (Bio-Rad, Hercules, CA), 6 μl of water, 1 μl of sense and 1 μl of antisense primers (500 nM). This master mix (18 μl) was filled in the iCycler strips and 2 μl of cDNA (0.625, 2.5, 10, or 40 ng reverse-transcribed total RNA) was added as PCR template. The following iCycler experimental run protocol was used: denaturation program (95°C for 9 min), amplification, and quantification program repeated 40 times (95°C for 30 s, 60°C for 30 s, 72°C for 30 s), melting curve program (60–95°C with a heating rate of 0.1°C per second), and finally a cooling step to 4°C. Emitted fluorescence for each reaction was measured during the extension phase. Real-time PCR efficiency (E) was calculated from the given slopes, with the iCycler software, as previously described (34).

The cycle threshold (CT), i.e., the cycle number at which the amount of the amplified gene reaches threshold fluorescence, was determined by using the iCycler software. The relative expression ratio (R) of the different target genes was calculated based on efficiency (E) and cycle threshold (CT), deviation of an unknown sample vs a control, and compared with the housekeeping gene GAPDH, as previously described (34, 35).

Ca²⁺ transients were measured in DC loaded with the Ca²⁺ indicator fura-2/AM (Calbiochem, La Jolla, CA) by using the digital fluorescence microscope unit Attofluor (Zeiss, Oberkochen, Germany). Briefly, DC were incubated with 2 μM fura-2/AM for 30 min at 37° C in a Ca²⁺- and Mg²⁺-free Hanks’ BSA solution. Cells were then washed twice and finally resuspended in the same buffer containing 1.5 mM CaCl₂ and MgCl₂. Traces were followed spectrofluorometrically and Ca²⁺ transients were determined by multiple cell acquisitions with the 340/380 wavelength excitation ratio at an emission wavelength of 505 nm. Curves shown are representatives of the whole cell population.

IL-8 was measured in DC supernatants by ELISA (BD PharMingen, San Diego, CA). IL-1β was determined by using ELISA kits from Amersham Pharmacia Biotech (Piscataway, NJ). IL-12 and TNF-α present in DC supernatants were measured by ELISA using matched pair mAbs from R&D Systems (Abingdon, U.K.). Samples were assayed in triplicate for each condition.

Intracellular cAMP levels were determined by an enzyme immunoassay (Amersham Pharmacia Biotech) according to the manufacturer’s instructions. Forskolin (Sigma-Aldrich) was used as a positive control. cAMP levels are expressed as the index representing the ratio between values obtained in stimulated cells and cells incubated in control medium.

Unless stated otherwise, data are expressed as mean ± SEM. ANOVA was used to compare experimental groups to control values. When the global test of differences was significant at the 5% level, pairwise tests of differences between groups were applied (Tukey’s comparison test). For PCR bands, statistical analysis was performed by the Dunnet comparison test (ANOVA).

Expression of mRNA for the different 5-HTR subtypes was analyzed by RT-PCR in immature and mature DC. Fig. 1,A shows that immature and LPS-matured DC expressed mRNA for 5-HTR_1B, 5-HTR_1E, 5-HTR_2A, and 5-HTR_2B receptors. The long splice variant of the 5-HTR₃ and 5-HTR₄ mRNA were found. Expression of the 5-HTR₇ mRNA was also detected (Fig. 1,B). We found no transcripts for 5-HTR_1A, 5-HTR_1D, 5-HTR_1F, 5-HTR_2C, 5-HTR₅, and 5-HTR₆ receptors in DC (data not shown). Extensive characterization of 5-HTR isotypes present in DC was performed by real-time PCR and relative quantification, at different time points during DC maturation (Fig. 1, C and D). Expression of the transcript of 5-HTR_1B, 5-HTR_1E, and 5-HTR_2B subtypes significantly decreased, while 5-HTR₄ and 5-HTR₇ transcripts increased after LPS addition. Expression levels of mRNA for the ligand-gated cation channel 5-HTR₃ and the heptahelical 5-HTR_2A subtype did not significantly change during maturation.

Functional expression of 5-HTR in DC was analyzed by measuring intracellular Ca²⁺ changes elicited by stimulation of 5-HTR. Stimulation of immature and mature DC with 5-HT induced a rapid and dose-dependent Ca²⁺ increase both in immature and mature DC although the extent of this response was higher in immature DC (Fig. 2). To study involvement of the different 5-HTR subtypes in Ca²⁺ transients in immature DC, cells were stimulated with different 5-HTR agonists. Fig. 3,A shows that 5-CT which is a preferential agonist at 5-HTR₁, 5-HTR₄, and 5-HTR₇ subtypes, as well as the selective 5-HT_1B agonist AnHCL (Fig. 3,B) induced a spiking Ca²⁺ rise followed by a slow declining phase. Experiments performed with the 5-HT_1E/F agonist BRL 54443 (Fig. 3,C) and the 5-HTR₂ agonist DOI (Fig. 3,D) show that these subtypes are also functional in immature DC. Moreover, we were able to show that incubation of immature DC with the selective 5-HTR₂ antagonist ketanserin (100 μM for 30 min) before stimulation with the 5-HT₂ agonist DOI completely abolished the Ca²⁺ increase induced by this compound, while it failed to block 5-HTR₁- and 5-HTR₃-mediated responses (data not shown). In contrast to immature DC, mature DC did not respond to any of the above mentioned agonists (Fig. 3 E). Unresponsiveness was not due to a generalized defect in the Ca²⁺ response as the chemokine MIP-3β/CCL19 was able to trigger a Ca²⁺ response in mature DC (data not shown).

5-HTR₃ is a ligand-gated cation channel triggering Ca²⁺ influx from the extracellular milieu and consequently plasma membrane depolarization. As shown in Fig. 4, the 5-HTR₃ agonist 2-methyl-5HT induced Ca²⁺ transients in immature as well as in mature DC. In contrast, the 5-HTR₄ agonist 2-MHT and the 5-HTR₇ agonist 8-HDPAT did not induce any Ca²⁺ response in immature and mature DC (data not shown). Besides Ca²⁺ influx through the plasma membrane, Ca²⁺ transients can be due to mobilization of the ion from the intracellular stores. To better discriminate between the two pathways, DC were stimulated with 5-HT in the absence of extracellular Ca²⁺. The Ca²⁺ chelator EGTA reduced 5-HT-induced Ca²⁺ transients by ∼20% in immature DC. In contrast, EGTA completely abolished the Ca²⁺ response in mature DC showing that it was entirely due to influx through the plasma membrane. Chelation of extracellular Ca²⁺ did not affect the 5-HTR₁- and 5-HTR₂-mediated intracellular Ca²⁺ mobilization induced by AnHCL, BRL 54443, DOI, and 5-CT. In contrast, the 5-HTR₃-mediated response in immature and mature DC was totally blocked by EGTA (Table I).

Mobilization of Ca²⁺ from intracellular stores by heptahelical receptors is often mediated via PTX-sensitive G_i/o proteins (16, 17, 18). To study participation of G_i/o-proteins in 5-HTR₁-mediated signaling, immature DC were preincubated with PTX and then stimulated with 5-CT, AnHCL, and BRL 54443 (Fig. 5). PTX almost completely abolished responses induced by these agonists. To exclude that lack of response of DC upon treatment with pertussis toxin was due to a cytotoxic effect of the molecule, cells were also stimulated with the 5-HTR₃ agonist 2-methyl-5HT, or the P2X receptor agonists αβ-meATP and BzATP. PTX did not affect the Ca²⁺ response triggered by these agonists, showing that its inhibitory effects were not due to cytotoxicity (data not shown).

5-HTR₄ and 5-HTR₇ couple to G_s proteins and stimulate adenylyl cyclase. To investigate functional coupling of 5-HTR₄ and 5-HTR₇ in DC, cAMP levels were analyzed after stimulation with 5-HT, the 5-HTR₄ agonist 2-MHT as well as the 5-HTR₇ agonist 8-HDPAT. 5-HT, 2-MHT, and 8-HDPAT did not elicit cAMP increase in immature DC, whereas they induced accumulation of this second messenger in a concentration-dependent manner in LPS-matured DC (Fig. 6).

Recent evidence suggests that 5-HT modulates the production of cytokines in T cells and monocytes (3, 4). Fig. 7,A shows that 5-HT added together with LPS, dose-dependently increased the production of IL-8 in mature DC, while it did not affect cytokine production in immature DC. Fig. 7,B shows that the 5-HTR₇ agonist 8-HDPAT, the 5-HTR₄ agonist 2-MHT, the 5-HTR₃ agonist 2-methyl-5HT and the 5-HTR_1,4,7 agonist 5-CT had a significant and concentration-dependent effect on IL-8 secretion. Half-maximal and maximal effects were seen at 10⁻⁵ and 10⁻⁴ M, respectively. In contrast, the selective 5-HTR₁ and 5-HTR₂ agonists had no effect on IL-8 secretion. Moreover, 5-HTR agonists stimulated a time-dependent release of IL-8 (Fig. 7,C). In addition, relative mRNA quantification by real-time PCR showed that 5-HT-induced IL-8 release was paralleled by enhanced IL-8 mRNA levels (Fig. 7 D). Maximal expression of the mRNA for IL-8 was detected after 8 h and slowly declined over the 24-h time course.

Moreover, 5-HT induced IL-1β secretion from mature, but not immature, DC (Fig. 8,A). Stimulation of the ionotropic and G_s protein-coupled 5-HTR₃, 5-HTR₄, and 5-HTR₇ subtypes triggered IL-1β secretion (Fig. 8,B). The 5HT-mediated effect on IL-1β release was time-dependent with significant increases starting from 8 h (Fig. 8 C). However, relative mRNA quantification by real-time PCR showed that expression of IL-1β mRNA was not changed by 5-HT stimulation in mature DC (data not shown).

To study involvement of different 5-HT receptor subtypes on IL-1β and IL-8 release, experiments with the selective 5-HT₄ receptor antagonist RS-39604 and the 5-HT₇ receptor antagonists pimozide or SB-269970 were also performed (Table II). Preincubation of LPS-maturing DC with the isotype-specific receptor antagonists (10⁻⁷ M) before stimulation with 2-MHT or 8-HDPAT (10⁻⁴ M) completely abolished 5-HT-mediated effects on IL-8 and IL-1β secretion, while it failed to block 5-HTR₃-mediated responses.

It is known that stimulation of G_s protein-coupled receptors inhibits IL-12 and TNF-α secretion in mature DC (36, 37). Therefore, we characterized the effect of 5-HT on these important cytokines. As shown in Fig. 9, 5-HT inhibited IL-12 and TNF-α production in mature DC in a dose-dependent manner, while it had no effects on immature DC. Using selective 5-HTR agonists we showed that 5-HT-mediated inhibition of IL-12 and TNF-α production was due to activation of the 5-HTR₄ and 5-HTR₇ subtypes (Table III).

5-HT is present in the periphery at high concentrations, in platelets, basophils, and mast cells (2) and it is released during platelet aggregation or IgE stimulation. There is accumulating evidence to support a regulatory function of 5-HT in the immune system (2, 3, 4, 5, 6). A role for 5-HT in the pathogenesis of bronchial asthma has also been recently proposed (8). Pharmacological and molecular studies revealed the existence of different 5-HTR subtypes classified either as ligand-gated cation channels or in the G protein-coupled receptor superfamily.

In this study, we show that DC expressed several functional 5-HTR subtypes. In addition, we also found that 5-HTR mRNA expression levels were modulated during DC maturation. Immature, compared with mature, DC expressed higher mRNA levels of the 5-HTR_1B, 5-HTR_1E, and 5-HTR_2B subtypes. Comparable mRNA levels of the two splice variants of 5-HTR₃ and 5-HTR_2A were found in immature and mature DC, whereas mRNA levels of the 5-HTR₄ and 5-HTR₇ subtypes were higher in mature DC.

To investigate functional expression of the different 5-HTR subtypes during DC maturation, we analyzed in more detail the intracellular signaling pathways activated by 5-HT. 5-HTR_1B, 5-HTR_1E, 5-HTR_2A, and 5-HTR_2B receptors couple to PTX-sensitive G_i/o as well as to PTX-insensitive G_q proteins. Stimulation of these receptors also activates phospholipase C which cleaves phosphoinositides into diacylglycerol and inositol 1,4,5-trisphosphate, inducing mobilization of Ca²⁺ from the intracellular stores. By monitoring agonist-dependent Ca²⁺ changes, we showed that 5-HTR_1B, 5-HTR_1E, 5-HTR_2A, and 5-HTR_2B were functional and coupled to G_i/o proteins only in immature DC. Moreover, we found that the cation channel 5-HTR₃ was functional both in immature and mature DC. These findings suggest that while in immature DC, 5-HT induced intracellular Ca²⁺ concentration changes via 5-HTR₁ and 5-HTR₂-mediated Ca²⁺ mobilization from the intracellular stores, besides the ligand-gated cation channel 5-HTR₃-mediated Ca²⁺ influx. In mature DC, the only active pathway seemed to be that mediated by 5-HTR₃. 5-HTR₄ and 5-HTR₇ couple via G_s to stimulate adenylyl cyclase (22, 27). Functional expression of these two receptors was demonstrated in mature DC. We showed that 5-HT induced an increase in cAMP concentration in these cells. The shift in 5-HT-induced G_i/o protein-dependent Ca²⁺ response to adenylyl cyclase-mediated cAMP formation during the maturation process was well in accordance with the increased mRNA expression levels of the 5-HTR₄ and 5-HTR₇ subtypes during DC maturation. However, to explain this functional shift one cannot exclude other mechanisms besides transcriptional down- and/or up-regulation of single 5-HTR subtypes. Retention of receptor molecules into submembraneous vesicles or posttranslational modifications of G protein subunits can also be hypothesized.

To get insight into the physiological significance of 5-HT in DC, cytokine secretion was analyzed. We found that stimulation of 5-HTR₃, 5-HTR₄ and 5-HTR₇ subtypes mediated the release of IL-1β and IL-8. However, mRNA analyses suggested that 5-HT modulated secretion of IL-1β and IL-8 by two different mechanisms. Enhanced IL-8 mRNA levels upon stimulation of DC with 5-HT would suggest a transcriptionally regulated effect. In contrast, unchanged mRNA levels of IL-1β in immature and mature DC indicate that 5-HT would affect a posttranscriptional regulatory step in IL-1β production. In this context, it might be of interest that 5-HT has been recently involved in the pathogeneses of asthma (8). Several studies have shown that allergen challenge causes, in humans as well as in animal models, an IL-8-mediated recruitment of neutrophils in the lung, and also an IL-1β-dependent alteration of airway smooth muscle responses (38, 39). Therefore, it can be hypothesized that in patients with acute severe asthma, IL-8 released by 5-HT-activated DC may cause neutrophil infiltration, and that secretion of IL-1β would then exacerbate the proasthmatic changes due to airway smooth muscle hyperresponsiveness. DC are critical effectors in both initiating and modulating immune responses because they capture, process, and transport Ags to secondary lymphoid organs, where they prime T cells (30, 31). Depending on the microenvironment, DC can regulate the outgrowth of T cell subsets. In the presence of IL-12 they induce Th1 cells, whereas with IL-4 there is induction of Th2 cell subsets (37, 40). In this context, it is of interest that previous publications reported that cAMP inhibits IL-12 production from DC, promotes Th2 cell differentiation, and suppresses Th1 priming (40, 41).

In summary, our study shows that 5-HT activates, in a maturation-dependent manner, different DC signaling pathways. These data further stress the immunomodulatory role of 5-HT at peripheral sites.