Expression and Function of Serotonin 2A and 2B Receptors in the Mammalian Respiratory Network

Neurons of the respiratory network in the lower brainstem express a variety of serotonin receptors (5-HTRs) that act primarily through adenylyl cyclase. However, there is one receptor family including 5-HT2A, 5-HT2B, and 5-HT2C receptors that are directed towards protein kinase C (PKC). In contrast to 5-HT2ARs, expression and function of 5-HT2BRs within the respiratory network are still unclear. 5-HT2BR utilizes a Gq-mediated signaling cascade involving calcium and leading to activation of phospholipase C and IP3/DAG pathways. Based on previous studies, this signal pathway appears to mediate excitatory actions on respiration. In the present study, we analyzed receptor expression in pontine and medullary regions of the respiratory network both at the transcriptional and translational level using quantitative RT-PCR and self-made as well as commercially available antibodies, respectively. In addition we measured effects of selective agonists and antagonists for 5-HT2ARs and 5-HT2BRs given intra-arterially on phrenic nerve discharges in juvenile rats using the perfused brainstem preparation. The drugs caused significant changes in discharge activity. Co-administration of both agonists revealed a dominance of the 5-HT2BR. Given the nature of the signaling pathways, we investigated whether intracellular calcium may explain effects observed in the respiratory network. Taken together, the results of this study suggest a significant role of both receptors in respiratory network modulation.

5-HT 2B Rs have been implicated in anxiety, schizophrenia, autism, migraine, and spreading depression [32]. In addition, 5-HT 2B R-dependent serotonin uptake influences the plasma serotonin level [33]. 5-HT 2B Rs are also important regulators of embryonic development; inactivation of the 5-HT 2B R gene leads to partial embryonic and early neonatal death in mice [34]. In the respiratory network, it has been shown that 5-HT 2B Rs enhance rhythmic motor discharge activity recorded in neonatal mice in vitro [35].
Until now, there were no detailed descriptions published of 5-HT 2B R distribution in the brainstem with expression data only available for the neocortex, cerebellum, dorsal hypothalamus, and medial amygdala [36].
In the present report, we provide a detailed account of the expression and distribution of 5-HT 2B Rs and 5-HT 2A Rs in the ponto-medullary respiratory network including respiratory motor population of the cervical spinal cord and brainstem. Using a monospecific anti-5-HT 2B R-antibody prepared in our laboratory and a commercially available 5-HT 2A R antibody, we show that 5-HT 2B Rs and 5-HT 2A Rs are expressed in neurons of the pre-Bötzinger complex (pre-BötC), an essential kernel of the respiratory network associated with the primary rhythmogenesis [37][38][39][40].
Furthermore, our study also demonstrates that both receptors affect discharge properties in the phrenic motor output.

Materials and Methods
The experimental procedures were performed in accordance with European Community and National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. The Ethics Committee of the Georg-August-University, Göttingen, Germany approved the study and assigned the approval ID T1108 to this work.

Antibody generation
The polyclonal antibodies against the rat 5-HT 2B R were generated by immunizing three New Zealand White rabbits (Charles River) with a 16mer peptide derived from the second intracellular loop of the rat 5-HT 2B R amino acid sequence (NH 2 -CAISLDRYIAIKKPIQ-COOH; NCBI-Accession No.: NP_058946). For immunization purposes, peptides were coupled to keyhole limpet hemocyanin (KLH) according to standard protocols. The rabbits were immunized with 300 mg of KLH-coupled peptide in Hunter's adjuvant (TiterMax Gold, Sigma) five times (28-daysintervall). The resulting antiserum was affinity-purified against the immunizing peptide.

Western Blot detection of 5-HT 2B R protein
Brain stem tissue isolated from both male Sprague-Dawley rats and male C57BL/6J mice were resuspended in 200 ml cell lysis buffer (50 mM Tris/HCl, pH 8.0, 150 mM NaCl, 2% w/v SDS, 1% NP40, 0.5% Na-Deoxycholate) supplemented with protease inhibitor cocktail (Sigma). Protein content was determined using a Lowry assay for high SDS concentration according to manufacturer's instructions (BioRad). 40 mg of total protein of each sample were boiled in 56 Laemmli buffer (250 mM Tris/HCl, pH 6.8; 10 mM EDTA, 10% w/v SDS, 5% v/v 2-mercaptoethanol, 50% v/v glycerol, 0.5% w/v bromophenol blue) for 5 min at 95uC and then separated using a precast SDS-PAGE (Novex). Proteins were transferred to a nitrocellulose membrane. The membrane was blocked with 4% w/v BSA/TBS/0.05%Tween (pH 7.4) for 45 min at RT. 5-HT 2B R protein was detected with a self-made monospecific polyclonal antibody (1:1,000 dilution) after incubation for 3 hours at RT. After extensive washing, appropriate secondary horseradish peroxidase (HRP)-conjugated antibodies (Dianova, Hamburg, Germany) were used at a dilution of 1:20,000 for 2 hours at RT. The visualization of the antigen-antibody reaction was performed with enhanced chemiluminescence (ECL) kit (BioRad, Germany).

Immunohistochemistry
(a) Preparation of brain tissue. Male juvenile Sprague-Dawley rats (P25-P32) were deeply anesthetized with isoflurane (1-Chloro-2,2,2-trifluoroethyl-difluoromethylether, Abbott, Wiesbaden, Germany) until they were unresponsive to painful stimuli (pressure applied to a forepaw). A thoracotomy was performed, and animals were transcardially perfused with 50 ml of 0.9% NaCl followed by 200 ml of 4% phosphate-buffered formaldehyde (10 ml/min). The brain was removed and post-fixed for 4 hours with the same fixative at 4uC, cryoprotected in 10% sucrose for 2 hours followed by 30% sucrose in 0.1 M phosphate buffer overnight at 4uC, and then frozen at 225uC. Series of 20-and 40-mm-thick brain sections were cut from cervical spinal cord to midbrain collicular level using a freezing microtome (Frigocut, Reichert-Jung, Germany).
(c) Immunofluorescence. Sections were permeabilized with 0.2% Triton X-100 for 30 min at RT and then washed two times with PBS (pH 7.4). Non-specific binding sites were blocked with PBS containing 5% BSA for 1 h at RT. Sections were incubated with primary antibodies (2-5 mg/ml) for 4 hours at RT. After washing, sections were incubated for 2 hours at RT in the dark with species-specific Cy2-or Cy5-conjugated secondary antibodies (Dianova, Germany; 2% BSA/PBS, antibody dilution 1:400). Neuronal immunofluorescence was analyzed with a confocal laserscanning microscope Meta-LSM 510 (Zeiss, Germany) using laser lines at 488 nm (Ar/Kr laser) and at 633 nm (He/Ne laser). Confocal images were processed by using overlays of two channels with the LSM 510 software provided by Zeiss. Digital images were taken at 2,04862,048 dpi and were imported into Adobe Photoshop CS4, were digitally adjusted if necessary for brightness and contrast and were assembled into plates. Subsequent imaging procedures (cell counting) were performed using ImageJ (http://rsb.info.nih.gov/ij/).

Molecular Biology
(a) Generation of expression constructs. Brain tissue from one male rat at P11 was explanted and used for total RNA isolation with the OLS RNA kit (OLS, Germany) according to manufacturer's instructions. The total RNA was used in one-step RT-PCR (Invitrogen) using primer pairs for the 5-HT 2A R gene [Htr2a, F (59-atggaaattctttgtgaagac-39)/R (59-tcacacacagctaaccttttc-39)] and 5-HT 2B R gene [Htr2b, F (59-atggcttcatcttataaaatgtc-39)/ R (59-ctatatgtagctgacttggtcttc-39)], respectively. The cycling program used for RT-PCR comprised of: initial reverse transcription at 55uC for 30 min followed by denaturation at 94uC for 2 min. 40 cycles of denaturation at 94uC for 15 sec, annealing at 57uC for 30 sec, and elongation at 68uC for 90 sec were concluded with a final elongation step at 68uC for 5 min. The resulting RT-PCR fragment was purified from the gel and cloned into pTarget expression vector (Promega). Sequencing validated the correct insert identity.
(b) Transfection of cell lines. Murine neuroblastoma cell line N1E-115 was obtained from ATCC and maintained at 37uC in humid atmosphere with 5% CO 2 and passaged every second day. For transfection, cells were seeded 24 hours prior to transfection at a density of 100,000 cells in 4-well-plates (Nunc) on acid-washed and poly-L-lysine coated 12 mm round glass cover slips. Cells were transfected with 2 mg DNA and 2 ml Lipofectamine (Invitrogen) in 500 ml OptiMEM (Invitrogen) per well and kept under normal culture conditions for 20 hours, afterwards fresh OptiMEM replaced the medium.
(c) Detection of endogenous Htr2b in cell lines by RT-PCR. Total RNA from 10 7 non-transfected cells or cells transfected with 6 mg of the plasmid encoding 5-HT 2B R was prepared using the OLS RNA kit. One mg of total RNA each was entered in the one-step-RT-PCR reaction, and the resulting PCR fragment was analyzed on an agarose gel. While the 59-sequence of murine and rat Htr2b is identical, the 39-sequences do differ. Therefore, for RT-PCR the rat forward primer was used, while the reverse primer for mouse was 59-ctatatgtagctgacctgctcttc-39.
(d) RT-PCR analyses of Htr2a and Htr2b of rat brain tissue. The total RNA of the cortex, hypoglossal nucleus, and pre-BötC dissected from corresponding 300-mm-thick slice preparations was isolated using GenElute TM mammalian total RNA kit (Sigma). First strand cDNA was synthesized from 1 mg total RNA using SuperScript TM first-strand synthesis system with random hexamers according to manufacturer's instructions (Invitrogen). Samples without reverse transcription (w/o RT) served as negative controls for the following PCR to exclude amplification of genomic DNA. For the PCR, specific forward and reverse primers were derived from different exons of the 5-HT 2A R and 5-HT 2B R cDNA to avoid amplification of genomic DNA. The cDNA sequences were obtained from the National Center for Biotechnology Information (NCBI; http://www.ncbi.nlm.nih. gov/). Specificity of selected primers was tested by partial sequencing of the amplification products for their identification by SeqLab company (Göttingen, Germany). The following primer pairs were used for amplification: The PCR reaction mixture for one sample was composed of 1-2 ml cDNA, 1 ml forward primer, 1 ml reverse primer, 1 ml dNTPs (100 mM dNTP mix), 1 ml DMSO, 5 ml NH 4 buffer (106), 2 ml MgCl 2 (50 mM solution), and 1 ml PANScript red DNA polymerase (PAN Biotech, Germany). The mixture was filled up to 50 ml with DEPC-treated water. The following program was used for the PCR reaction: initial denaturation at 94uC, 4 min/ 386[denaturation 94uC, annealing 1 min/55uC, extension 1 min/ 72uC, 2 min]/final elongation 72uC, termination 10 min/4uC hold. Actb (b-Actin) was used as an internal standard for all PCR reactions.
(e) Real-time RT-PCR. The relative quantification of Htr2a and Htr2b gene expression in specific rat tissues was done by real-time RT-PCR analysis. Spinal cord, inferior olive, pre-Bötzinger complex, and parabrachial complex were dissected from corresponding 300mm-thick cryostat sections (P32; n = 3 animals) under visual control. The total ribonucleic acid (RNA) of homogenized brain tissue was isolated using the TrizolH method according to manufacturer's instructions (GibcoBRL) and its concentration was determined using the NanoDrop ND-1000 spectrophotometer followed by its quality and integrity measurement by electrophoresis on RNA 6000 LabChipH kit (Agilent 2100 Bioanalyzer). The RNA was transcribed into the corresponding deoxyribonucleic acid (cDNA) using the iScript cDNA Synthesis Kit (BioRad). The following primer pairs were designed by using the Primer3 program (http://frodo.wi. Gel electrophoresis revealed a single polymerase chain reaction (PCR) product, and the melting curve analysis showed a single peak for all amplification products. The PCR products were sequenced and blasted to confirm the correct identity of each amplicon. Ten-fold serial dilutions generated from cDNA of each sample were used as a reference for the standard curve calculation to determine primer efficiency. Triplicates of all real-time PCR reactions were performed in a 25 ml mixture containing 1/20 volume of the sample cDNA preparation from 250 ng total RNA, 400 nM of each primers, and 16 Fast-SYBR Green Master Mix (Applied Biosystems, USA).
The PCR-reactions were performed as follows: initial activation at 95uC for 60 s, 42 cycles of (denaturation 95uC/10 s, annealing and extension 60uC/30 s), and a final gradual increase of 0.5uC in temperature from 60uC to 90uC.
All real-time quantifications were performed using the iCycler iQ system (BioRad) and were adjusted by using the method according to Pfaffl [41].

Calcium imaging of cells recombinantly expressing 5-HT 2A Rs or 5-HT 2B Rs
The perfused brainstem preparation is, due to its thickness and need for constant perfusion not suited for microscopic analysis. Therefore, we opted to do the calcium imaging in murine neuroblastoma N1E-115 cells, where endogenous expression of 5-HT 2 Rs is negligible, but are known to signal via the PLC-DAG pathway [42,43]. Another advantage of transfection is the control over which receptors (5-HT 2A R, 5-HT 2B R or both) are expressed in individual cells, avoiding the need for antagonists and simplifying analysis. 12-16 hours post transfection, cells were transferred to calcium-free imaging medium (130 mM NaCl, 3.5 mM KCl, 1.25 mM NaH 2 PO 4 , 24 mM NaHCO 3 , 1.2 mM MgSO 4 , 10 mM Glucose) and incubated with Fluo-4-AM (Invitrogen) at a final concentration of 5 mM for 30 min at 37uC. The Fluo-4-AM stock solution was prepared as 2 mM using 10% pluronic acid F-127 in DMSO (Sigma) and was diluted just before use. After incubation, cells were washed with calcium-free medium and Fluo-4-AM was allowed to hydrolyze for another 30 min in the presence of probenicid to avoid leeching of fluorescent probe from the cell.
For calcium imaging, Fluo-4-AM loaded cells were transferred to a recording chamber equipped with an inverted Olympus microscope (IX71) with appropriate filters (515 nm beam splitter and a 535/50 band-pass filter) and a triggered LED light source (PreciseExcite, CoolLED) with 465 nm excitation. Images were taken for 300 msec at 1 sec intervals. After recording baseline fluorescence, cells were stimulated with 1000 nM serotonin (Sigma) in calcium-free medium. This concentration was chosen based on a dose-response curve giving a linear response for serotonin stimulation between 500 and 1500 nM. For all experiments, a 106, 1.0 NA objective (Olympus) was used.
To compare calcium measurements between experiments, we calculated the apparent fluorescent intensity F/F 0 by dividing the fluorescent intensity (F) at every time point by the average fluorescence recorded before stimulation (F 0 ). Data were statistically analyzed with Student's t-test and presented as mean 6 standard deviation (s. d.).
Perfused brainstem preparation of rats (a) Perfused brainstem preparation of rats. The experiments on the perfused brainstem preparation [44] were performed on male Sprague-Dawley rats (P25-P32, 90-150 g) that were housed under a 12 h light/dark cycle, with food and water provided ad libitum.
Animals were deeply anesthetized with halothane until they were unresponsive to a forepaw pinch, decerebrated at the precollicular level and cerebellectomized, bisected below the diaphragm, and the skin was removed. The upper body was placed into a recording chamber and perfused retrogradely via the thoracic aorta with ACSF (containing in mM: MgSO 4 1.25; KH 2 PO 4 1.25; KCl 5; NaCl 125; CaCl 2 2.5; NaHCO 3 25; glucose 10, 1.25% Ficoll and aerated with carbogen (5% CO 2 /95% O 2 ; pH 7.35 at 30uC). The perfusate was collected, filtered twice and re-circulated. Norcuronium-bromide (0.5 mg/200 ml) was added for muscle relaxation. The perfusion pressure was set to 45 to 55 mm Hg.
(b) Phrenic nerve signal processing. A silver wire immersed in bath solution within a capillary suction electrode picked up phrenic nerve activity representing the respiratory motor output to the diaphragm and inspiratory rib cage muscles. Phrenic nerve signals were amplified 2,000-5,0006, filtered (lowpass, 7,000 Hz cutoff frequency; high pass, 8 Hz) and integrated (time constant, 100 ms). The processed signals were digitized by a PowerLab 8/30 microprocessor and stored using LabChart 7 software (ADInstruments, Australia).
(c) Analysis of phrenic nerve discharge properties. Discharges of a representative one-minute duration were measured in the absence of (control) and after intra-arterial perfusion with ACSF containing a 5-HT 2A or 5-HT 2B receptor agonist or antagonist. Measurements of drug effect were made at 5-minute intervals. The peak of the integrated discharge (mV) was used as an estimator of discharge intensity and normalized to the control, which was set to 100%. Discharge frequency (bursts per minute) was calculated from the integrated signals. Values (mean 6 standard error of the mean) for amplitude and frequency were calculated from consecutive discharges that occurred over one minute during control and when drug effects were maximal.
All statistical tests (paired t-test) for pharmacological experiments were performed using GraphPad Prism version 5.0d for MacOS X.

Production and characterization of monospecific anti-5-HT 2B R antibodies
The peptide for immunization was derived from the second intracellular loop of the rat 5-HT 2B R-sequence (NH 2 -CAISL-DRYIAIKKPIQ-COOH; fig. 1Aa). The specificity of the monospecific polyclonal anti-5-HT 2B R antibody was tested in three different test systems: Immunoblot analysis (n = 3) of both mouse and rat brainstem lysate revealed a specific band at 48 kDa, which is in accordance with the predicted relative molecular mass of the receptor (fig. 1Ab). Murine neuroblastoma cells recombinantly expressing rat 5-HT 2B R ( fig. 1B) were used for specificity testing of the antibody, while the neocortex and the hypoglossal nucleus (XII) was selected to test immunohistochemistry in tissue ( fig. 1C) based on previous positive results reported by Duxon [36].
The control cells faintly expressed the mouse 5-HT 2B R that is also recognized by the antibody because of sequence homology. After transfection with the rat receptor the antibody labeling revealed a strong fluorescent signal. Also, both brain regions selected showed strong 5-HT 2B R reactivity (fig. 1Bc).
The anti-5-HT 2B R antibody immunoreactivity on cells as well as on neurons of both regions was effectively blocked after preincubation of the primary antibody with a 50-fold molar excess of the peptide that was used for immunization indicating specificity. As a control, RT-PCR analysis confirmed 5-HT 2B R-specific mRNA expression in cells within both regions ( fig. 1Cc). To analyze receptor expression at the protein level within the ponto-medullary respiratory network we applied our self-made monospecific polyclonal anti-5-HT 2B R antibody in combination with a commercially available monoclonal anti-5-HT 2A R antibody (BD Bioscience, San Diego, USA). Both receptor subtypes were expressed in crucial parts of the respiratory network such as the pre-BötC and the pontine Kölliker-Fuse nucleus ( fig. 3, 4). A detailed analysis of the pre-BötC, the supposed kernel essential for the generation of the primary respiratory rhythm (Smith et al., 1991), showed a strong co-expression of 5-HT 2A R and 5-HT 2B R.
In the dorsolateral pons expression of 5-HT 2B R within the parabrachial complex (PB) and Kölliker-Fuse (KF) nucleus was weak, compared to the 5-HT 2A R ( fig. 4D-F). Contrary, the 5-HT 2A R showed dense expression in the KF and lateral crescent nucleus of the PB ( fig. 4A-C). Both nuclei are closely linked with respiratory control. In addition more modest expression was observed in the external lateral, central, and dorsal nuclei of the PB. Curiously, both 5-HT 2A R and 5-HT 2B R showed dense expression in the internal lateral nucleus (il) of the PB (fig. 4B, E), a subnucleus of the PB-complex that is still undefined in its physiological functions.

Analysis of systemic effects of 5-HT 2A Rs and 5-HT 2B Rs on respiratory activity
The effects of systemic application of specific antagonists for 5-HT 2A R and 5-HT 2B R (Altanserin hydrochloride and LY 272015, respectively) and specific agonists for 5-HT 2A R and 5-HT 2B R (TCB-2 and BW 723C86, respectively) on spontaneous breathing activity were investigated in male Sprague-Dawley rats using the perfused brainstem preparation [44]. We found that pharmacological manipulation of 5-HT 2 Rs can either change the amplitude or the frequency of the phrenic nerve activity (PNA).
Application To analyze the calcium signaling of both receptors, we expressed them recombinantly either alone or together in neuroblastoma cells. To avoid artifacts, relatively low amounts of DNA were transfected to avoid overexpression.
The recombinant approach allowed us to image single cells and correlate resulting signals to a defined complement of receptors, which would have not been possible in perfused brainstem preparations.
In neuroblastoma cells expressing 5-HT 2A Rs and 5-HT 2B Rs alone, either agonist evoked a release of cytosolic Ca 2+ from intracellular stores (see figure 7 and table 1) with a large initial calcium spike. While the reactions of individual cells varied slightly, the mean Ca 2+ increase of cells expressing 5-HT 2A R (peak F/F 0 of 2.5960.8) was similar to those expressing 5-HT 2B R (peak transfected N1E-115 cells (b). The anti-5-HT 2B R antibody-dependent staining indicated a strong labeling of N1E-115 cells that had been transiently transfected with the rat 5-HT 2B R (b). Non-transfected cells expressing the mouse 5-HT 2B R showed a weak neuronal immunofluorescent signal that corresponds with a weak PCR signal (amplicon size 1114 bp) for the mouse 5-HT 2B R-mRNA (Htr2b) (c). Samples without reverse transcription (w/o RT) served as negative controls. (C) (a, b) Immunohistochemistry. Both pyramidal neurons of the cortex and motoneurons of the hypoglossal nucleus (XII) revealed a strong 5-HT 2B R immunoreactivity (-IR) (a) that was effectively blocked after pre-incubation of the antibody with a 50-fold molar excess of the peptide CAISLDRYIAIKKPIQ (+ peptide) that was used for immunization (b). Insets in (a) show labeled neurons at a higher magnification. Immunolabeling was performed using the PAP-method with diaminobenzidine as chromogen. (c) RT-PCR analysis of the rat cortex and hypoglossal nucleus. The 5-HT 2B R-specific mRNA (Htr2b) was detectable in neurons within both the rat cortex and the hypoglossal nucleus (XII) (amplicon size 380 bp). doi:10.1371/journal.pone.0021395.g001 F/F 0 of 2.2361.1). In contrast, the calcium increase was significantly faster for 5-HT 2A R (time to peak 30.569.6) than for 5-HT 2B R (time to peak 57.768.4). The calcium level almost returned to baseline levels within ,35 seconds exhibiting no significant differences for both receptors, with a half-peak-width of 36.761.7 sec for 5-HT 2A R and 34.663.2 for 5-HT 2B R ( fig. 7B).
Co-application of 5-HT 2A R and 5-HT 2B R agonists had unexpected effects. The Ca 2+ signal (79.4468.9 sec) was significantly slower from onset to peak than the signals produced by either agonist alone (p,0.001), while the time of onset was similar to 5-HT 2A R alone. In addition, the fluorescence signal at its peak (F/F 0 , 1.4460.17) was notably smaller (p,0.001). While the duration of the calcium peak by co-stimulation of both receptors was nearly doubled (half-peak-width of 86.3265.0 sec; p,0.001), the amount of released calcium, measured as ''area under the curve'', was very similar no matter if the receptors are expressed alone or together.
As our real-time PCR analysis showed that 5-HT 2A R and 5-HT 2B R are expressed in different amounts, with 5-HT 2A R being in 5-to 10-fold excess, we also transfected N1E cells with DNA ratios of 5-HT 2A R to 5-HT 2B R of 5:1 and 1:5, respectively. Regardless of the DNA ratios, the presence of 5-HT 2B R always

Discussion
This study reveals the locations of 5-HT 2A R and 5-HT 2B R in regions of the ponto-medullary respiratory network, including sites where the receptors are co-expressed. We demonstrate that agonist activation of the receptors evokes dramatic changes in discharge activity recorded from the phrenic motor output, and that activation of each type of receptor has distinctive effects on phrenic nerve discharge intensity and duration. Through the use of selective receptor antagonists, we found that only the 5-HT 2A R constitutively modulates phrenic motor output. In neuroblastoma cells transfected with 5-HT 2A R and 5-HT 2B R, we discovered distinctly different calcium signal kinetics when each type of 5-HT receptor was activated. We also uncovered unexpected effects on signal amplitude and time course when 5-HT 2A R and 5-HT 2B R are coactivated.
In the paragraphs to follow, we discuss each of these aspects in turn, along with their physiological implications for respiratory motor output modulation.

Distribution of 5-HT 2B and 5-HT 2A receptors in regions of the ponto-medullary respiratory network
Within the medulla and pons, functionally defined respiratory regions provide input to cranial motoneurons controlling the airways, and to spinal motoneurons activating inspiratory and expiratory pump muscles. A variety of neurotransmitters and modulators involved in respiratory control have been identified in many of these respiratory related compartments (reviewed by Alheid and McCrimmon [48]). In all these regions, 5-HT 2A R and coupled protein kinase dependent signaling pathways have been identified functionally and anatomically [49]. Until now, however, the distribution of 5-HT 2B Rs had not been investigated, and nothing had been known about their functional importance to respiratory control. Our study shows that 5-HT 2A R and 5-HT 2B R are co-localized in the Kolliker-Fuse and Parabrachial regions of the Pons, and in the BötC and pre-BötC of the ventral medulla with an approximate 5-fold stronger expression of 5-HT 2A R in all regions.
Respiratory neurons in the PB and KF constitute the pontine respiratory group. A variety of respiratory neuronal types are found in this region [50,51], which receives axonal projections from the ventral respiratory column, and from the nucleus of the solitary tract (NTS). The NTS itself is a receiving station for pulmonary afferents from the lungs and upper airways. Based on 5-HT 2A R and 5-HT 2B R co-expression patterns, our study would predict contributions by both types of receptors to respiratory modulation within the KF-PB complex, with a modulatory role played by 5-HT 2B Rs.
Based on both in vitro and in vivo [52,53] studies, the pre-BötC was identified as a medullary region essential for respiratory rhythm generation. The region may play a prominent role in inspiratory phase control, although the pre-BötC contains populations of neurons that exhibit a variety of respiratory related discharge patterns . Previously, 5-HT 1A , 5-HT 2A , 5-HT 4 , and 5-HT 7 receptors have been identified in the pre-BötC by immuno-labeling [8,12,17]. The BötC houses a prominent population of expiratory neurons that provide widespread inhibitory projections within the ventral respiratory column (VRC), targeting both inspiratory and expiratory bulbospinal neurons as well as respiratory-related cranial motoneurons [54]. Some expiratory BötC neurons also send axon collaterals to the spinal cord, reaching at least as far as the phrenic motor nucleus [48,55]. Our present study of 5-HT 2A R and 5-HT 2B R coexpression in the pre-BötC and BötC suggests that, as in the pontine respiratory group, 5-HT 2A R modulation is predominant.
Differential effects of 5-HT 2A and 5-HT 2B receptor agonists and antagonists on phrenic nerve discharge properties Activation of 5-HT 2A Rs by the selective, CNS-permeable agonist TCB-2 [47,56] increased PNA discharge amplitude but not frequency, whereas the 5-HT 2A R antagonist Altanserin decreased both amplitude and frequency: in fact, discharge frequency decreased below control level. Although speculative, we suggest that 5-HT 2A R agonists target two functionally different populations of respiratory neurons: bulbospinal inspiratory neurons, leading to increased phrenic motor output, and propriobulbar inspiratory phase-regulating neurons that determine discharge frequency. Altanserin's capacity to reduce discharge frequency below control levels indicates that constitutive 5-HT 2A R activation is substantial in inspiratory phase regulating neurons.
The 5-HT 2B R agonist BW 723C68 increased only frequency, indicating that only neurons involved in discharge rate regulation were affected. However, the antagonist LY 272015 changed neither amplitude nor frequency. This suggests that in our in situ perfused brainstem preparation, 5-HT 2B Rs though present and activated by BW 723C68 were not constitutively activated.
When 5-HT 2A R and 5-HT 2B R agonists were given concurrently, a frequency increase that exceeded the singular effects of either agonist occurred. This is an expected outcome if frequencycontrolling neurons are preferred targets for 5-HT 2B R modulation in the ponto-medullary respiratory compartment.

5-HT 2B and 5-HT 2A receptors effects on Ca 2+ signaling
Because 5-HT 2A and 5-HT 2B receptors both utilize an intracellular signal pathway that leads to a buildup of cytoplasmic  Ca 2+ , we used the Ca 2+ signal as a measure of receptor activation by TCB-2 and by BW 723C68. Neuroblastoma cells, for reasons presented earlier, are advantageous for measuring the magnitude and kinetics of intracellular Ca 2+ fluctuations. Nonetheless, we acknowledge that there are limitations in relating Ca 2+ signals detected in neuroblastoma cells to discharge properties recorded from the phrenic motor output, and we interpret our results with this caveat in mind. Agonist activation of 5-HT 2A receptors produced a Ca 2+ signal that was rapid in onset, somewhat faster in time course to peak and larger in magnitude than the Ca 2+ transient produced by 5-HT 2B receptor activation. Another characteristic difference was an initial small dip in the signal due to 5-HT 2A receptors, whereas a small Ca 2+ transient preceded the predominant 5-HT 2B R dependent signal. Their respective signal profiles may reflect differences in dynamic interactions involving receptor activation, cell membrane Ca-channel openings as well as Ca 2+ release and uptake processes in intracellular organelles. A detailed interpretation of Ca 2+ signals recorded from the respiratory rhythmic rodent slice preparation can be found in published studies by Keller and coworkers [57,58]. Those studies illustrate Ca 2+ signals that are proportional to respiratory discharge intensity. Assuming that Ca 2+ signaling in neuroblastoma cells can be equated with signaling in cells of the brainstem respiratory network, we interpret our findings as follows. The faster and larger Ca 2+ signal may reflect more efficient 5-HT 2A R coupling to the PLC-DAG-PKC signal pathway, or to the spatial arrangement of receptor, Ca 2+ -channel, and organelles involved in Ca 2+ release and uptake. We can offer no ready explanation for the slowing and diminution of the Ca 2+ signal when 5-HT 2A and 5-HT 2B receptors co-expressed in neuroblastoma cells were coactivated. Perhaps different isoforms of DAG or PKC activated by 5-HT 2A R and 5-HT 2B R compete for phosphorylation sites on membrane calcium channels and storage sites and interact negatively.

Conclusion
Taking together the results from all sets of experiments (distribution, co-expression, phrenic nerve activity, and Ca 2+ signaling), we formulate the following working hypothesis: 5-HT 2A R is the dominant receptor governing respiratory control as evidenced by its stronger expression and constitutive activity. 5-HT 2B R, being present in all respiratory nuclei analyzed and coexpressed with 5-HT 2A R in many cells, although at a much reduced level, may act as a dose-dependent modulator. If more serotonin is released than needed to activate all 5-HT 2A Rs, 5-HT 2B Rs become activated. This would allow the system to regulate the respiratory rhythm by controlling serotonin release. Activation of spare 5-HT 2B R has the strongest effect on respiratory frequency. This regulation could, at least in part, be due to Ca 2+ signaling, as the presence of 5-HT 2B R changes the kinetics of the Ca 2+ signaling observed for both receptors alone, without altering the overall amount of mobilized calcium.