Neuropeptide Receptors Provide a Signalling Pathway for Trigeminal Modulation of Olfactory Transduction

European Journal of Neuroscience

European Journal of Neuroscience, Vol. 37, pp. 572–582, 2013 doi:10.1111/ejn.12066

NEUROSYSTEMS

Neuropeptide receptors provide a signalling pathway for trigeminal modulation of olfactory transduction

Philipp Daiber,1 Federica Genovese,1 Valentin A. Schriever,2 Thomas Hummel,2 Frank Mohrlen€ 1 and Stephan Frings1 1Department of Molecular Physiology, Centre for Organismal Studies, Heidelberg University, Heidelberg, Germany 2Interdisciplinary Centre for Smell & Taste, University Hospital Dresden, Dresden, Germany

Keywords: bimodal, CGRP, irritants, odorants

Abstract The mammalian olfactory epithelium contains olfactory receptor neurons and trigeminal sensory endings. The former mediate odor detection, the latter the detection of irritants. The two apparently parallel chemosensory systems are in reality interdependent in various well-documented ways. Psychophysical studies have shown that virtually all odorants can act as irritants, and that most irritants have an odor. Thus, the sensory perception of odorants and irritants is based on simultaneous input from the two systems. Moreover, functional interactions between the olfactory system and the trigeminal system exist on both peripheral and central levels. Here we examine the impact of trigeminal stimulation on the odor response of olfactory receptor neurons. Using an odorant with low trigeminal potency (phenylethyl alcohol) and a non-odorous irritant (CO2), we have explored this interaction in psychophysical experiments with human subjects and in electroolfactogram (EOG) recordings from rats. We have demonstrated that simultaneous activation of the trigeminal system attenuates the perception of odor intensity and distorts the EOG response. On the molecular level, we have identified a route for this cross-modal interaction. The neuropeptide calcitonin-gene related peptide (CGRP), which is released from trigeminal sensory fibres upon irritant stimulation, inhibits the odor response of olfactory receptor neurons. CGRP receptors expressed by these neurons mediate this neuromodulatory effect. This study demonstrates a site of trigeminal–olfactory interaction in the periphery. It reveals a pathway for trigeminal impact on olfactory signal processing that influences odor perception.

Introduction

The mammalian nose harbours a set of distinct chemosensory organs uli with low, if any, trigeminal potency (examples are vanillin, H2S that differ in their chemosensory receptors, their specific wiring in and phenylethyl alcohol), and the only known irritant with little or the brain and their role in the animal’s responses to environmental no detectable olfactory percept in humans is CO2 (Bensafi et al., cues. In rodents, five such organs can be distinguished: the main 2008). Thus it is highly likely that each sniff elicits a bimodal neu- olfactory epithelium, the trigeminal system, the vomeronasal organ, ronal response: activation of olfactory receptor neurons and, at the the septal organ and the Grueneberg ganglion (Ma, 2007, 2010; same time, activation of trigeminal sensory fibres. Importantly, the Munger et al., 2009; Dauner et al., 2012). The human nose, how- concomitant trigeminal activity detectably alters odorant perception ever, appears to operate with only two systems, the main olfactory (Kobal & Hummel, 1988; Cometto-Muniz & Hernandez, 1990; epithelium and the trigeminal system. Although these two systems Brand, 2006). Key features of olfactory performance, including sen- are generally held to be responsible for the separate sensory modali- sitivity and odor discrimination, can be modified by co-stimulation ties, one mediating olfaction and the other nociception, psychophysi- of trigeminal sensory fibres (Jacquot et al., 2004). Interaction cal and electrophysiological studies have clearly demonstrated that between the two sensory systems is thus a relevant factor in olfac- the intimate intertwining of the two systems precludes this distinc- tory performance. tion. In fact, practically every odorant co-stimulates the trigeminal At which sites and through which pathways trigeminal activity system and virtually all irritants co-stimulate the main olfactory epi- impacts on the olfactory system is currently not understood. There thelium (Cain, 1977; Doty et al., 1978; Cain & Murphy, 1980; Liv- is evidence that interaction occurs both in the olfactory epithelium ermore et al., 1992; Silver, 1992; Hummel & Livermore, 2002; (Kratskin et al., 2000) and in various brain regions including the Brand, 2006). Only a very few odorants are selective olfactory stim- olfactory bulb (Schaefer et al., 2002) and, possibly, in the medio- dorsal thalamic nucleus, where the two systems converge (Inokuchi et al., 1993; Brand, 2006). Earlier recordings from the frog olfactory epithelium revealed that odor-induced field potentials [electroolfacto- Correspondence: Professor Dr S. Frings, as above. grams (EOGs)] were altered when the ophthalmic branch of the tri- E-mail: [email protected] geminal nerve was stimulated antidromically, an effect that was Received 31 July 2012, revised 23 October 2012, accepted 24 October 2012 partially recapitulated by application of the trigeminal neuropeptide

© 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd Trigeminal modulation of olfactory receptor cells 573 substance P to the epithelium (Bouvet et al., 1987, 1988). This Table 1. Air-phase concentrations of odorants in the odor tubes observation suggests a neuromodulatory role for trigeminal fibres that appears to be mediated by neuropeptides and targeted at the Molarity (M) PEA conc. (ppm) IAA conc. (ppm) AIC conc. (ppm) olfactory receptor neurons. Release of trigeminal neuropeptides dur- À À À 0.0003 0.0228 9 10 3 2.54 9 10 3 2.37 9 10 3 ing a sniff may, therefore, alter the response characteristics of olfac- 0.001 0.0759 9 10À3 8.46 9 10À3 7.91 9 10À3 tory receptor neurons. 0.003 0.228 9 10À3 0.0254 0.0237 Here we have examined this hypothesis by investigating the sen- 0.01 0.760 9 10À3 0.0847 0.0791 À3 sory response to 2-phenylethyl alcohol (PEA), a floral odorant with 0.03 2.28 9 10 0.254 0.228 9 À3 particularly weak trigeminal potency. We have also studied its mod- 0.1 7.64 10 0.852 0.795 0.3 0.0232 2.59 2.41 ulation by the trigeminal agonist CO2 in human subjects. For com- 1 0.0815 9.09 8.33 parison, we measured PEA responses in rat olfactory epithelium. 3 0.288 32.1 28.0 Because of the pronounced CO2 sensitivity of the rat chemosensory 6 0.782 87.2 68.1 system we used the trigeminal neuropeptide calcitonin-gene related Undiluted 1.61 180 168 peptide (CGRP) to examine trigeminal–olfactory interactions. We Calculated concentrations (in ppm) of phenylethyl alcohol (PEA), isoamylac- explored the trigeminal innervation of the olfactory epithelium and etate (IAA) and allylisothiocyanate (AIC) in the gas phase of odor tubes. the expression of CGRP receptors in this tissue. Our studies provide Stock solutions, prepared at the indicated concentrations (in M) in mineral a concept for trigeminal modulation of the primary olfactory signal oil, were added to filter paper within the odor tube at room temperature. Val- and its consequence for odor perception. ues of ppm were calculated for 20 °C from the vapor pressure and the molar fraction of the two fluid components.

Materials and methods 5; CaCl2, 1; MgCl2, 1; HEPES, 10; glucose, 10; and pyruvate, 1; pH All experiments with human subjects followed the Declaration of adjusted to 7.4 with NaOH. The reference air flow was deodorised, Helsinki on biomedical research involving human subjects and was humidified and adjusted to 0.1 L/min. Recording electrodes were approved by the Ethics Committee from the University of Dresden pulled to a tip aperture of 20–25 lm from borosilicate glass capillaries Medical School (EK332092011, EK118072003). All participants (OD 1.5 mm, ID 0.87 mm) using a Flaming–Brown puller (Sutter provided written informed consent. Experiments with animal tissues Instruments, Novato, CA, USA) and filled with Ringer’s solution. The were performed in accordance with the Animal Protection Law and local surface potential of the olfactory epithelium was amplified (DP- the guidelines and permissions of Heidelberg University. 301; Warner Instruments), digitised (BNC 2120; National Instru- ments) and processed using the WINWCP software provided by Strath- clyde University, UK. Odorants were dissolved in mineral oil Psychophysical investigations (BioUltra, Sigma 697934; Sigma) at the concentrations listed in Forty healthy subjects, 29 female and 11 male, age range 20– Table 1 in M. Forty microlitres of each solution was placed inside an 32 years (mean Æ SD, 24.95 Æ 3.13 years) participated in this odor tube from which the odorant could be injected under computer study. All subjects were normosmic as ascertained using the 16-item control into the air stream using a pneumatic pico pump (PV830; odor identification test from the Sniffin’ Sticks test kit (mean score, WPI, Sarasota, FL, USA). For dose–response experiments, increasing 14.06 Æ 1.08; score range, 13–16; Hummel et al., 1997). Stimuli concentrations were applied at 2-min intervals. Odor concentrations in were presented monorhinally using a computer-controlled olfactome- the air space of the odor tubes were calculated from the respective ter (Olfactometer OM6b; Burghart, Wedel, Germany) with an air- vapour pressures according to Table 1. No attempt was made to esti- flow of 12 L/min. PEA was used for olfactory stimulation and CO2 mate odor concentrations at the sensory surface. for trigeminal stimulation. The concentrations for both stimuli ran- Submerged EOGs were recorded from similar preparations. A ged from 5 to 20% v/v. The stimulus duration for PEA was set to constant flow of Ringer’s solution (in mM: NaCl, 120; NaHCO3, 25; 200 ms. For CO2 stimulus, durations of 200, 1000, 2000 and KCl, 5; BES, 5; MgSO4, 1; CaCl2, 1; and glucose, 10; pH 7.4) was 3000 ms were used. The inter-stimulus interval ranged between 27 applied flowing over all olfactory turbinates in the caudal direction. and 33 s. The two stimuli were presented simultaneously. The study Recording electrodes were pulled from capillaries (GB150-10; Sci- was divided into sessions of 12–15 min each. The nostril side was ence Products) to an opening aperture of 1 lm and a resistance of changed after each session and the order of sessions was randomised 1MΩ. The pipettes were tip-filled with 1–2% agar and then back- between subjects. After each stimulus pair, subjects were asked to filled with pipette solution: (in mM) NaCl, 145; KCl, 5; HEPES, 10; rate the intensity of the odor (PEA) as well as the intensity of the MgCl2, 1; and CaCl2, 1; pH adjusted to 7.4 with NaOH. For stimu- irritation (CO2) separately on a visual analog scale from 0 to 100. lus application, a three-barrel capillary (3B120F-4; WPI) was pulled Data were analysed using the program package SPSS (version 17; with a vertical puller and subsequently bevelled on a Micro Grinder SPSS Inc., Chicago, Ill., USA). For statistical comparisons, t-tests (EG-44; Narishige) at an angle of 45°. The tip diameter of each for paired samples were used. Pearson statistics were employed for barrel was 1 lm. The capillary was connected to a Pico Pump correlational analyses. (PV 830; WPI), and stimuli were generated by applying a pressure pulse of 80–100 ms to one of the barrels. CGRP (1 lM) was either applied through one of the barrels of the stimulation capillary or, for Electroolfactography preincubation, included into the superfusing Ringer’s solution. Rats were anesthetised by exposure to an air–isoflurane mixture and killed by cervical dislocation. Skull and nasal bone were split to Cell sorting and PCR analysis expose the olfactory turbinates. For air-phase EOGs, the preparation was placed on a 2% agarose block and mounted in an interface cham- Rat olfactory turbinates were prepared, cut into small pieces and ber where the sensory surface was exposed to water-saturated air. The kept in ice-cold calcium-free solution: (in mM:) NaCl, 125; KCl, 5; agarose was dissolved in Ringer’s solution: (in mM) NaCl, 140; KCl, EGTA, 10; and HEPES, 10; pH adjusted to 7.4 with NaOH.

For dissociation of the olfactory epithelium, the tissue was first 8 min. Primer pairs are listed in Table 2. The resulting PCR prod- exposed to the same solution containing 0.1% (w/v) trypsin for ucts were verified by sequencing. 30 min at 35 °C. To remove DNA that leaked from damaged cells, the tissue pieces were incubated in (mM) 140 mM NaCl, 2 MgCl ,2 2 In situ hybridisation EGTA, 10 HEPES, 0.01% DNase, pH 7.4/NaOH for 10 min at 37 °C. Following trituration with a fire-polished Pasteur pipette, cells were DNA fragments of the CRLR gene (calcitonin receptor-like recep- centrifuged at 300 g for 5 min at 20 °C. The pellet was resuspended tor; GenBank Acc. No. NM_012717.1) and the CRCP gene (CGRP- in calcium-free solution containing 1 mM fluorescein-di (b-galactopyr- receptor component protein; NM_053670.3) were amplified using anoside) and stained in this solution for 2 min at 4 °C. After a second RT-PCR from rat olfactory epithelium, subcloned into pGEM-T vec- staining step with 1.5 lM propidium iodide (30 min, 4 °C), cells were tor (Promega) and sequenced. Primer pairs were rCRCP and rCRLR centrifuged (300 g,20°C), and the pellet was resuspended in cal- (Table 2). In vitro transcription was performed with the DIG RNA cium-free solution. The cell count was adjusted to 105–107 cells/mL labelling mix (Roche) according to the manufacturer’s instructions. using a haemocytometer. Fluorescence activated cell sorting (FACS) In situ hybridisation was performed as described in Schaeren-Wie- analysis was performed in a FACSAria Cell Sorter (BD Biosciences). mers & Gerfin-Moser (1993) with the following modifications: the Live olfactory sensory neurons were identified by propidium iodide probes were partially hydrolysed by alkaline treatment to 200– exclusion and fluorescein fluorescence. 300 bp. Cryosections (12–30 lm) from rat nasal cavity were col- RNA from the collected olfactory receptor neurons ( 4.5 9 105 lected onto Superfrost plus slides (Fisher Scientific, Houston, Texas) cells each) was extracted using the Dynabeads mRNA Direct and allowed to dry at room temperature for 15 min. Before acetyla- Kit (SKU 610-11; Invitrogen) or the Magnetic mRNA Isolation Kit tion, the sections were fixed in 4% paraformaldehyde, followed by (no. S1550S; New England Biolabs) according to the manufacturers’ digestion with proteinase K (50 lg/mL) for 5 min at 37 °C. Hybri- instructions. For RNA isolation from whole rat olfactory epithelium, disation was performed with ~300 ng/mL of each probe for 15 h at tissue was dissected from olfactory turbinates of adult rats. For 55 °C. Following high stringency washing at 60 °C, the probes human tissue samples, up to four biopsies were taken from each were detected with an anti-digoxigenin antibody (Roche) conjugated patient in the area of the lateral superior wall of the nasal cavity, to alkaline phosphatase and nitro-blue tetrazolium/5‐bromo‐4‐chloro‐ close to the radix of the medial turbinate and from the opposite dor- 3‐indolylphosphate as substrate. For controls, slides were hybridised sal septum (Leopold et al., 2000). Tissues were shock-frozen in to the sense probes. The staining was analysed using a Nikon liquid nitrogen and pulverised. Isolation of mRNA from rat and Eclipse 90i microscope equipped with a DS-Ri1 digital camera human olfactory epithelium was done with the innu-PREP RNA (Nikon AG, Dusseldorf,€ Germany). Mini Kit (Analytik Jena AG Lifescience). After DNase I treatment (RNase-free; Fermentas) cDNA was synthesised using 10 ng Immunohistochemistry mRNA, random hexamer primers and SuperscriptTM III Reverse Transcriptase (Invitrogen). PCR amplification was performed on Rats and olfactory marker protein (OMP)–green fluorescent protein 0.2 ng single-stranded cDNA with 2 U Taq DNA polymerase (GFP) mice were used for immunohistochemistry. In OMP–GFP (Axon). Cycling conditions were 94 °C for 3 min, 94 °C for 20 s, mice, the gene that encodes the OMP, a marker for mature olfactory 58 °C for 20 s and 72 °C for 30 s, for 32 cycles and then 72 °C for receptor neurons, is replaced by GFP (Potter et al., 2001). These mice were kindly provided by Dr Peter Mombaerts (MPI for Bio- Table 2. Primer sequences for PCR with rat and human cDNA physics, Frankfurt, Germany). Following isoflurane anesthesia and cervical dislocation, the nasal cavity of killed animals was perfused Product fi Target Primer length for 1 h with paraformaldehyde (PFA) xative (in mM:Na2HPO4, protein orientation Primer sequence (bp) 8.1; NaH2PO4, 1.9; and NaCl, 0.13; pH 7.4; with 4% paraformaldehyde) or with picric acid fixative (in mM:Na2HPO4, 121; NaH2PO4 rRAMP1 F GCCGGGACCCTGACTATGGTA 29; pH 7.4; with 4% PFA and 0.2% picric acid). The fixatives were R GATGCCCTCTGTGCGCTTGC 362 introduced into the nasal cavity under reduced pressure in a desicca- rCRCP F CGCCGGGCAGCAGAACTTGA tor. After three 20-min washing steps in PBS (in mM:Na2HPO4, R CTGCCCCATCTCGAGCCCTCT 576 8.1; NaH2PO4, 1.9; and NaCl, 0.13; pH 7.4) the preparation was – ° rCRLR F CAAACAGACTTGGGAGTCACTAGGAA dehydrated in 10 30% saccharose solution at 4 C overnight. Cryo- R CCACTGCCGTGAGGTGAATGATTGTC 505 sections were prepared at À20 °C (CM3050S; Leica Microsystems), and sections 8–20 lm thick were fixed to gelatine-coated glass rTRPA1 F AAGGGGCCTTGTTTCTTAGTG R TCATGGATAATGGGCATTGGA 498 slides with 4% PFA for 5 min. For antigen retrieval, sections were heated to 100 °C for 3 min in 10 mM sodium citrate, pH adjusted to rActin F GGTCATCACTATCGGCAATGAGC 6.0 with NaOH, then equilibrated with sodium borate buffer (120 g R GGACAGTGAGGCCAGGATAGAGC 300 H3BO3 per litre, pH adjusted to 8.0 with NaOH), and finally washed hRAMP1 F GGCCCATCACCTCTTCATGACC with PBS. Sections were dried and prepared for immunostaining by R AATGCCCTCAGTGCGCTTGCTC 391 equilibrating with 5% v/v Chemiblocker (no. 2170; Millipore) and hCRCP F AGTGAAGGATGCCAATTCTGCGCTTCTC 0.5% v/v Triton X-100 in PBS for 1 h. Primary antisera were R TTGGTCAACTTGTGGCTTTTCAATGCTG 234 applied in the same solution for 2 h. Sections were then washed hCRLR F TTGGACACGGATTGTCTATTGCATCA three times in PBS, dried and incubated with the secondary antibody for R CAGCACAAATTGGGCCATGGATAATG 462 90 min at room temperature. After three further washing steps, nuclei l hActin F CTGGGACGACATGGAGAAAA were stained with 0.3 M DAPI (no. D1306; Invitrogen), washed, dried R AAGGAAGGCTGGAAGAGTGC 564 and mounted with cover slips using Aqua PolyMount (no. 18606; Polysciences). The following primary antisera were used: rabbit anti- Prefix r, rat; prefix h, human. mCGRP (antigenic peptide sequence: SCNTATCVTHRLAGLLSRSGG

AB C D

Fig. 1. Psychophysical determination of trigeminal effects on PEA perception. (A) Ratings of the odor intensity of PEA at different concentrations. Mean ratings are displayed with SDs for each PEA concentration. Saturation was reached at 10%. *P < 0.05, **P < 0.01. (B) Ratings of irritation intensity at CO2 stimuli of different concentrations and stimulus durations. Mean ratings and SDs. (C) Ratings of odor intensity for a 20% PEA stimulus in the absence or presence of CO2 stimuli with different concentrations and durations. A signiﬁcant effect was detected at 20% CO2 at all durations. **P < 0.01; ***P < 0.001. (D) Nega- tive correlation between perceived odor intensity (PEA) and irritation intensity (CO2). The odor intensity was set to 20% PEA, while the CO2 concentration ranged from 5 to 20%. CO2 stimulus durations were 200, 1000, 2000 and 3000 ms. The solid line was obtained by linear regression and the correlation coefﬁcient is 0.98 (P < 0.001).

VVKDNFVPTNVGSEAF, polyclonal; PC205L; Calbiochem); this zel et al., 1999), the samples were frozen in liquid nitrogen and antiserum selectively stained peptidergic fibres, consistent with ground. Five micrograms of total protein was used for each earlier results (Finger et al., 1990). Rabbit anti-hCRLR (antigenic cAMP-ELISA. Protein concentration was determined by Bradford peptide sequence DGNWFRHPASNRTWTNYTQCNVNTHEKVK- assays. TALNLFYLTIIGHGLSIASLL, polyclonal (ab83697; Abcam and SAB 2100335; Sigma); these two antisera localised CGRP receptor protein to ORNs, in accordance with in situ hybridisation studies and RT-PCR Results analysis of FACS-purified ORNs. Rabbit anti-hCRCP (antigenic peptide Psychophysical investigations of trigeminal–olfactory sequence: TPCRHQSPEIVREFLTALKSHKLTKAEKLQLLNHRPVTA- interactions VEIQLMVEESEERLTEEQIEALLHTVTSILPAEPEAEQKK NT NSN - VAMDEEDPA, polyclonal; HPA007216; Sigma); the specificity To examine effects of trigeminal co-stimulation on odor perception, of this antiserum is supported by consistency with our in situ hybri- we used a pair of stimuli which are selective agonists for either the disation and RT-PCR results. Goat anti-hezrin (C-terminal peptide), olfactory system (PEA) or the trigeminal system (CO2). The stimuli polyclonal (sc-6409; Santa Cruz) raised against a 15- to 25-amino were presented to human subjects using an olfactometer at a flow rate acid (aa)-immunising antigenic peptide mapping between aa residues of 12 L/min. After each stimulation, subjects were asked to rate the 490 and 540 of human origin. We have confirmed the specificity of intensity of perception on a visual analogue scale ranging from 0 to this antibody for microvilli staining in nasal epithelia using confocal 100. The appropriate concentration ranges for the two stimuli were microscopy (Hengl et al., 2010; Dauner et al., 2012). Secondary tested separately. A plateau in odor intensity rating was achieved at antisera were donkey anti-rabbit, Alexa Fluor 488 and 568 (A21206 10% PEA delivered for 200 ms (Fig. 1A). No further significant and A10042; Molecular Probes) and donkey anti-goat, Alexa increase in odor intensity rating was observed at higher PEA concen- Fluor 488 and 568 (A11057 and A11055; Molecular Probes). trations. CO2-induced irritation was detectable at 10% CO2 delivered Control experiments without primary antisera did not show fluo- for 1–3 s but was more pronounced at 20% CO2 delivered for 1–3s rescence signals. For confocal microscopy, we used a Nikon (Fig. 1B). Based on these results, we applied PEA and CO2 at concen- TE2000-E with a C1Si spectral imaging confocal laser scanning trations of 5–20% in the subsequent interaction studies. When the two system and a Nikon Ti with an A1Rsi confocal laser scanning sys- stimuli were co-applied, the intensity ratings for PEA decreased with tem, both provided by the Nikon Imaging Center at Heidelberg increasing CO2 stimulus intensity (Fig. 1C). Intensity rating for 20% University. PEA was reduced when combined with 5–20% CO2 (Fig. 1C; P 0.01). Thus, trigeminal activation attenuates odor perception. This effect is illustrated in Fig. 1D, which depicts the negative correla- Determination of cAMP production tion between ratings of odor intensity and irritation intensity Quantification of cAMP-synthesis was performed using the cAMP (r = 0.98, P < 0.001). The stronger the perceived trigeminal stimula- Enzyme Immunoassay Kit (CA200; Sigma) with a solubilisation tion, the weaker the odor perception. solution containing (in mM) HEPES, 50; NaCl, 100; MgCl2,1; and CaCl , 1. The substrate was p-nitrophenyl phosphate (N7653; 2 Trigeminal potency in the EOG Sigma). Samples of olfactory epithelium from three rats were tested in triplicate for each experimental condition, with a total of To explore the effect of trigeminal activity on olfaction in the rat, six samples per animal and test compound. Each sample was pre- we employed EOG recordings from rat olfactory epithelium. We treated for 2 min with 300 lM 3-isobutyl-1-methylxanthine first asked whether odorants with different trigeminal potencies pro- (IBMX; I5879; Sigma) to inhibit cAMP degradation by phospho- duce different EOGs. Using an air-based odor-delivery system, we diesterases. Following a 5-min incubation with 10 lM CGRP, examined the electrical response to PEA, amyl acetate and mustard 10 lM forskolin or Henkel 100 odor mix (1 : 1000 dilution; Wet- oil (allylisothiocyanate; AIC), compounds with increasing trigeminal

ABCD

Fig. 2. Air EOG recordings at increasing trigeminal potency. (A–C) Electrical response of rat olfactory epithelium to 100-ms stimuli of the odorants PEA, IAA and AIC at increasing odorant concentrations. PEA displays a simple phasic–tonic response whereas IAA and AIC display secondary effects that distort the simple time-course. (D) Dose–response relations obtained from n = 5 to 7 animals per odorant. Means and SDs are displayed together with Hill-fits (for PEA and IAA) to establish the Hill coefficients (PEA, 1.1; IAA, 0.6). Concentrations refer to the gas phase in the odor tube before injection into the air stream. potencies. PEA induced biphasic responses, an initial transient fol- A lowed by a steady plateau (Fig. 2A). The resulting dose–response relation for peak values had a sigmoid saturating shape and a slope that was characterised by a Hill coefficient of 1.1 (Fig. 2D, black). Saturation in the EOG response corresponded well with the saturation of perceived odor intensity seen in Fig. 1A. In contrast, amyl acetate, a fruity odorant with intermediate trigeminal potency in both humans and rats (Doty et al., 1978; Silver & Moulton, 1982), displayed prolonged responses, biphasic EOG shapes at high concentrations (Fig. 2B) and no clear saturation of the dose–response curve (Fig. 2D, blue). When fitted with a Hill equation, a Hill coefficient of 0.6 indicated a diminished slope and, hence, the deviation from a simple binding paradigm. Finally, AIC, a pungent odorant with high trigeminal potency, generated multiphasic EOG recordings (Fig. 2C) B with non-saturating dose–response relations (Fig. 2D, red). These data indicate that an estimate of trigeminal potency in rat can be derived from EOG recordings. Odorants with highest trigeminal potencies show the strongest deviation from a binding process as described by the Hill equation. Thus, in accordance with psychometric data from humans (Doty et al., 1978) and with trigeminal nerve recordings from rat (Silver & Moulton, 1982), our observation of a simple dose–response relation in the EOG indicates that PEA specif- ically activates the olfactory system in the rat and has little effect on the trigeminal system. Fig. 3. Trigeminal fibres in the rat olfactory epithelium. (A) Coronal section of an olfactory turbinate showing a dense plexus of CGRP-positive fibres in Density of peptidergic trigeminal innervation of the olfactory the lamina propria (LP) and several single fibres (arrows) traversing the epithelium layer of olfactory receptor neurons (ORN) towards the apical surface which is formed by sustentacular cells (SC). Confocal Z-stack corresponding to To explore the proximity of olfactory receptor neurons to sites of 36 lm in depth. (B) Three individual trigeminal fibres reaching the very top neuropeptide release, we stained peptidergic fibres in cryosections of the olfactory epithelium. One (right) even seems to reach the tissue surface which is labelled with red ezrin immunofluorescence. from rat olfactory epithelium using a CGRP antiserum. Figure 3A shows an image that represents a 36-lm-thick coronal section (reconstructed from a confocal Z-stack) of an olfactory turbinate. A along the entire length of the fibres. The fibres ended near the apical dense plexus of CGRP-positive fibres can be seen in the subepitheli- surface, mostly right underneath the microvilli layer which was al lamina propria. Originating from this plexus, peptidergic fibres stained with an ezrin antiserum (Fig. 3B). Occasionally, fibres rise toward the epithelial surface in small groups, with a distance of appeared to reach the apical surface (Fig. 3B, right). However, this 100 lm between individual groups. Thus, each olfactory receptor was an infrequent observation and may be the result of a preparation neuron is separated by not more than 50 lm from the next trigemi- artifact. In accordance with previous data from frog and rat nal fibre. We estimated a density of 400–600 fibres per mm2 for the (Getchell et al., 1989; Zielinski et al., 1989; Finger et al., 1990), epithelium covering the turbinates. Trigeminal innervation was den- our data indicate a regularly spaced peptidergic innervation of the ser on the caudal part of the septum, which is also covered with olfactory epithelium with most olfactory receptor neurons < 50 lm olfactory epithelium. CGRP-containing varicosities could be seen away from a release site of neuropeptides.

Fig. 5. CGRP induced net cAMP synthesis in rat olfactory epithelium. cAMP levels were measured in tissue lysates by ELISA following a 5-min pre-incubation with 300 lM IBMX to inhibit phosphodiesterase. Test compounds (10 lM CGRP, 10 lM forskolin, or the odorant mixture Henkel 100, 1 : 1000) were applied for 2 min. Student’s t-test, ***P < 0.01; number of experiments: n = 16 for CGRP and n = 4 each for forskolin and odorants.

Fig. 4. Apically applied CGRP induced an EOG-like response. (A) Applica- AB tion of a CGRP pulse (1 lM, 100 ms) caused a transient negative potential in the submerged-EOG conﬁguration. For comparison, a strong odorant stimulus (Henkel 100; 1 : 1000, 100 ms) was applied during recording from the same spot on the olfactory epithelium. Both signals were suppressed by MDL-12, 330A (30 lM), an inhibitor of adenylyl cyclase type III. (B) Niﬂu- 2+ mic acid (NFA; 300 lM), a blocker of the Ca -activated chloride channel anoctamin2/TMEM16B, blocked the response to 1 lM CGRP.

Modulation of EOG response by the neuropeptide CGRP To examine whether CGRP acts directly on olfactory sensory neurons we performed submerged EOGs, in which the olfactory epithelium was covered with Ringer’s solution. Neuropeptides and pharmacological agents were dissolved in the Ringer’s solution and could be applied through a gravity-driven superfusion system to the tissue surface. Pulsed odor stimuli were added to a constant stream Fig. 6. Reduction of EOG amplitude by CGRP. (A) Two submerged-EOG of Ringer’s solution by a stimulation pipette with 1 lm tip aperture, recordings from the same spot on the rat olfactory epithelium. The control was 100 lM PEA (100 ms) and the test was 100 lM PEA + 1 lM CGRP connected to a pump that generated brief pressure pulses. The stim- (100 ms). The neuropeptide reduced the EOG amplitude by 60%. (B) Mean uli elicited transient negative potentials, similar in time course and reduction of EOG amplitude by 1 lM CGRP, averaged across 15 experi- amplitude to the ones recorded in the air-phase EOGs. When a pulse ments. The Mann–Whitney U-test indicates a highly significant reduction in EOG amplitude (***P < 0.01). of 1 lM CGRP was applied through a stimulation pipette, a small EOG-like signal was observed. Its amplitude amounted to 20% of the maximal odor-induced EOG recorded at the same site of the olfactory epithelium (Fig. 4A, solid lines). To find out whether this time in isolated olfactory epithelium using a cAMP-ELISA. In these excitatory CGRP effect involved components of the odor-transduc- binding assays, degradation of cAMP was suppressed by 300 lM tion cascade, we applied CGRP together with a compound that IBMX, an inhibitor of the phosphodiesterase PDE1C in olfactory interrupts this pathway. In the presence of 30 lM MDL-12, an inhib- receptor neurons (Yan et al., 1995). At 10 lM CGRP, cAMP itor of adenylyl cyclase type III, both odorant response and CGRP synthesis was increased to a similar extent as by a strong olfactory response were reduced by 90% (Fig. 4A, dotted lines). Downstream stimulation (Fig. 5). The maximum rate of cAMP synthesis was of cAMP synthesis, the olfactory transduction cascade contains an achieved with 10 lM forskolin, an agonist of adenylyl cyclase amplification step that is mediated by Ca2+-activated ClÀ channels. type III. These data suggest that CGRP targets olfactory sensory These channels (anoctamin 2; Pifferi et al., 2009; Stephan et al., neurons directly, and that CGRP is able to induce cAMP signalling. 2009; Billig et al., 2011) are sensitive to the channel blocker niflu- When 1 lM CGRP was applied together with an odor stimulus mic acid, as demonstrated in previous EOG studies (Nickell et al., (100 lM PEA, 80-ms pulse), the EOG amplitude was reduced while 2007) and in recordings from isolated olfactory receptor neurons the time course remained unchanged (Fig. 6A). On average, the (Boccaccio & Menini, 2007). The CGRP response in our recordings observed EOG amplitude was reduced to 55 Æ 41% of the PEA was largely suppressed by co-application of 300 lM niflumic acid value (Fig. 6B; Mann–Whitney test, P = 0.003, n = 15). The co- (Fig. 4B). These observations suggest that CGRP has the capacity application of PEA and CGRP in the same pulse has the disadvan- to weakly trigger the cAMP-dependent transduction process in olfac- tage that the two substances may have different kinetics of action on tory receptor neurons. To test CGRP-induced cAMP synthesis the olfactory receptor neurons. We therefore examined whether pre- directly, we measured cAMP accumulation during a 2-min exposure incubation with 1 lM CGRP would expose a slower effect on the

ABreceptors for CGRP. Receptors targeted to the cilia probably explain the sensitivity of the olfactory epithelium to apically applied CGRP, as observed in our experiments.

Discussion Activation of the nasal trigeminal system by odorants is well documented in rats (Silver, 1992) and even better in humans, where the examination of patients with isolated congenital anosmia (people born without an olfactory sense) enables the analysis of trigeminal chemoperception without olfactory interference (Frasnelli et al., 2007; Iannilli et al., 2007, 2011; Croy et al., 2012). One aspect of fi Fig. 7. this topic that has not been investigated in detail is the signi cance Prolonged exposure to CGRP suppressed the odor response. fi (A) Following an odor pulse (100 ms, 100 lM PEA) in the absence of CGRP of stimulus-induced neuropeptide release from trigeminal bres in (trace 0), 10 lM CGRP was added to the superfusion in the submerged EOG the olfactory epithelium. When stimulated with high concentrations fi con guration. Consecutive odor stimuli were repeated at 4-min intervals and of the non-odorous gas CO2, human subjects report irritation, a per- produced decreasing EOG amplitudes. (B) Results from three experiments τ = ception that is based on the activation of nasal trigeminal sensory show an exponential decline in EOG amplitudes ( 13 s) in the presence of fi fi CGRP, while odor responses without CGRP were fairly constant. bres. We nd that the intensity rating for the non-irritant olfactory stimulus PEA is reduced by pungent CO2 stimuli, a finding that is in line with the first systematic study of the impact of CO2 on olfac- PEA response. Figure 7A illustrates that this was indeed the case. tory perception (Cain & Murphy, 1980). This attenuating effect on odor perception may be caused by peripheral or central interactions PEA pulses applied at 4-min intervals in the presence of 1 lM CGRP in the superfusion solution elicited EOG responses that between the two sensory systems. Here we demonstrate that periph- declined with a time constant of 13 s (Fig. 7B). Thus, the trigeminal eral olfactory sensory neurons express CGRP receptors and respond peptide CGRP induced a fast, weakly excitatory, effect followed by to CGRP application. The receptors provide an input channel for a prolonged inhibitory effect on the odor response recorded in the modulation by this trigeminal neuropeptide. ‘ ’ ‘ ’ EOG. In this study, we used the pure odorant PEA and the pure irritant CO2 in order to selectively stimulate the olfactory and trigeminal systems. PEA belongs to the very few odors which do not cause CGRP receptors are expressed in olfactory sensory neurons irritation in humans and rats, even at elevated concentrations (Doty Our EOG data suggest that olfactory receptor neurons may express et al., 1978; Silver & Moulton, 1982). PEA has been shown to stim- CGRP receptors. To test this, we isolated rat olfactory receptor neu- ulate microvillus solitary chemosensory cells in the nasal anterior rons by FACS, using the high intrinsic b-galactosidase activity of respiratory epithelium at a concentration of 5 mM (Lin et al., 2008). these neurons (Liberles & Buck, 2006). The cDNA derived from This is interpreted as irritant stimulation because these chemosensory these cells contained messages encoding all three components of the cells are connected to the trigeminal system. In contrast, similar CGRP receptor complex (Fig. 8A): CRLR, CRCP (Evans et al., microvillus cells in the olfactory epithelium have no discernible tri- 2000) and receptor activity-modifying protein 1 (RAMP1; McLat- geminal association and are not considered to serve a chemosensory chie et al., 1998). In cDNA derived from entire olfactory epithe- function (Hansen & Finger, 2008). It is, therefore, not likely that lium, we found similar signals in preparations from rat and human PEA produces irritation in the olfactory system via local solitary tissue (Fig. 8A). A major receptor for irritants in the trigeminal sys- chemosensory cells. It is, however, conceivable that respiratory tri- tem is TRPA1 (Wang et al., 2008, 2010, 2011). We detected no geminal fibres form collaterals with olfactory fibres. In this case, a evidence of TRPA1 expression in olfactory receptor neurons PEA signal mediated by respiratory chemosensory cells may cross (Fig. 8A), a finding that excludes direct activation of these neurons the epithelial boundary and cause neuropeptide release in the olfac- by the irritant CO2 through TRPA1 channels (Jordt et al., 2004; tory epithelium. Other odorants, such as isoamyl acetate (IAA) with Wang et al., 2010). In situ hybridisation revealed expression of its fruity odor, display increasing trigeminal stimulation at higher CRLR and CRCP message in olfactory receptor neurons (Fig. 8B). concentrations. In our EOG experiments from rats, this difference is In accordance with these results, immunohistochemistry with a poly- reflected in distortions of time courses and dose–response relations clonal antiserum directed against the receptor protein CRLR showed of the two compounds. While the EOGs indicate a single response a specific staining of olfactory receptor neurons in all cell compart- process during PEA stimulation, the double-peaked EOG shapes ments ranging from the apical surface to axon bundles in the lamina with IAA and the reduced slope of the IAA dose–response curve propria (Fig. 9A). Similar immunosignals were observed with reveal that additional processes emerge as the IAA concentration CRCP (Fig. 9B). To obtain robust immunosignals, tissue sections increases. This effect is even more apparent in the EOGs obtained had to be treated with hot citrate buffer for antigen retrieval. This with AIC. Several distinct processes distort the EOG time course procedure often caused loss of cilia from the apical surface. Without and produce a non-saturating dose–response relation. Thus, the antigen retrieval, preservation of cilia was better but the sub-apical occurrence of multiple components in the EOG recording correlates immunosignals were weaker. In these experiments we were, how- qualitatively with the trigeminal potency of an odorant. For the pres- ever, able to observe intense ciliary staining against CRLR and ent study, this analysis allowed us to accept PEA as an odorant with- CRCP (Fig. 9C). Ciliary staining was confined to the olfactory epi- out manifest trigeminal potency in rat olfactory epithelium, a finding thelium and was absent from the cilia of the respiratory epithelium that is in accordance with an earlier study of multiunit responses (Fig. 9D). Similar results were obtained with two different CRLR recorded from the rat ethmoid nerve (Silver & Moulton, 1982). antisera (not shown). In accordance with the in situ hybridisation In its response to CO2 stimuli, the olfactory epithelium can results, these data show that olfactory receptor neurons expressed display two distinct modes of sensitivity. Several amphibia and

Fig. 8. Transcription of CGRP-receptor genes in olfactory receptor neurons. (A) RT-PCR results obtained with primers against the three components of CGRP receptors: calcitonin receptor-like receptor (CRLR), CGRP-receptor component protein (CRCP) and receptor activity-modifying protein 1 (RAMP1). All three gene products were detected in cDNA generated from isolated rat olfactory receptor neurons (rORN), from whole rat olfactory epithelium (rOE) and from whole human olfactory epithelium (hOE). The message for TRPA1, the trigeminal receptor for CO2 and other irritants, was not present in olfactory tissue. The positive control was cDNA from rat trigeminal ganglia (rTG). (B) In situ hybridisation with probes for CRCP and CRLR reveals expression of both messages in the olfactory epithelium. A strong signal is localised in the layer of olfactory receptor neurons (ORN). Weaker signals appear in sustentacular cells (SC) and in the subepithelial lamina propria (LP). The signal is absent from control sections treated with sense probes.

Fig. 9. Expression of CGRP-receptor proteins in rodent olfactory receptor neurons. (A) A coronal cryosection of rat olfactory turbinate displays specific expression of CRLR protein in the olfactory epithelium (OE). Somata, dendrites and axon bundles (AB) of olfactory receptor neurons are stained. Other cell types display weaker immunosignals. The section was treated for antigen retrieval. (B) A similar expression pattern is shown for CRCP in the rat olfactory epithelium. (C) Ciliary expression of CRLR protein (red) can be seen at the apical surface of the olfactory epithelium in an OMP–GFP mouse, processed without antigen retrieval. The sub-apical CRLR immunosignal is weaker under this condition. (D) Expression of CRCP in the OMP-GFP mouse. Expression stops at the transi- tion area between olfactory epithelium (OE) and respiratory epithelium (RE). mammals, including the rat, have the ability to smell low concentra- activated, again through acidification. Importantly, the protons that tions of CO2 (Coates, 2001). The CO2 sense is thought to be induce trigeminal activity are not those released in the olfactory involved in the regulation of respiration and in the control of protec- mucus or in the interstitial fluid, but those released within the axo- tive reflexes such as sneezing. Behavioural tests with rats yielded plasm of the trigeminal fibres. Studies of TRPA1-channel gating in detection thresholds of 1–3% CO2 (Youngentob et al., 1991; Ferris trigeminal ganglion neurons recently revealed that the channels are et al., 2007). This high-sensitivity mode of CO2 detection depends opened by intracellular acidification (Wang et al., 2010). As CO2 on the activity of carbonic anhydrase which catalyses the synthesis can readily diffuse across plasma membranes, the carbonic anhydr- of carbonic acid (Tarun et al., 2003; Kimoto et al., 2004; Ferris ase reaction inside the sensory endings can trigger a drop in intrafi- + À et al., 2007). The resulting acidification (H2CO3 ? H + HCO3 ) bre pH. The precise extent of this intracellular acidification has not induces activity in a small subset of olfactory receptor neurons which yet been measured, and the intrafibre concentration of carbonic an- are located in the most dorsal recesses of the olfactory epithelium (Hu hydrase is not known. However, considering the small accessible et al., 2007). These neurons have an exceptionally high CO2 sensitivity volume within the fibres, acidification would be expected to be more (detection threshold ~0.066% CO2) and convey their information to the pronounced within the fibres than in the surrounding fluid with its most caudal area of the main olfactory bulb, the necklace olfactory sys- much larger volume. In human subjects, Shusterman & Avila (2003) tem, a part of the rodent olfactory system that appears to be specialised measured the acidification of nasal mucosal pH with extracellular in CO2 detection (Luo, 2008). In humans, there is no necklace olfac- pH electrodes during CO2 stimuli similar to the ones used in the tory system and no high-sensitivity CO2 detection; CO2 has no odor present study (5 L/min, 3 s duration, 20% CO2). The extracellular for us. At higher CO2 concentrations, however, trigeminal fibres are pH decreased from basal levels of ~7.4 by only 0.05–0.1 pH units

© 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience, 37, 572–582 580 P. Daiber et al. during each CO2 pulse. These minute decrements in extracellular 2007), and we found an inhibitory effect of CGRP on odor pH reflect efficient pH buffering of the extracellular medium. The response. It is not likely that CGRP acts as an inhibitory odor- advantage of CO2 detection by intracellular acidification is obvious: ant through classical odorant receptors in the chemosensory cilia. larger pH changes can be triggered by CO2 inside the axoplasm. Inhibitory odorants use a different signalling pathway than excit- With respect to the extracellular medium, the trigeminal fibres atory odorants. The canonical excitatory pathway is mediated by appear not to act as pH electrodes but rather as as CO2 electrodes, cAMP (Kleene, 2008), while inhibitory odorants use phosphati- independent of volume and pH buffer capacity of the surrounding dylinositol-3,4,5-trisphosphate as second messenger (Spehr et al., fluid. In conclusion, using CO2 as ‘pure’ irritant in our psychophysi- 2002; Ukhanov et al., 2010). Our data show that olfactory cal experiments is justified by the lack of a necklace olfactory system receptor neurons produce cAMP in response to CGRP applica- in humans and by the mode of operation of trigeminal fibres as tion, as is generally seen with CGRP receptors in other tissues high-threshold CO2 detectors. (Walker et al., 2010). Thus, it appears that the olfactory CGRP For psychophysical studies, we used PEA and CO2 concentra- receptor couples to adenylyl cyclase type III, the only isoform tions sufficient to produce robust responses but low enough to pre- known to be active in olfactory sensory neurons (Pfeuffer et al., vent unnecessary discomfort. As application dynamics and stimulus 1989; Bakalyar & Reed, 1990). Our discovery of the triad length are both essential parameters in CO2 stimulation (Hummel CRLR, CRCP and RAMP1 in olfactory receptor neurons points et al., 2003; Wise et al., 2003), we also tested stimuli of different to a functional role for CGRP receptors in shaping the primary durations. Our data revealed an inverse relation between the ratings olfactory signal. The precise signalling pathway for the CGRP of odor intensity and irritation intensity. When PEA and CO2 were effect awaits detailed biochemical studies. applied simultaneously, the perceived odor intensity decreased as Taken together, our results demonstrate that stimulation of the perceived irritation intensity increased. This finding is in line the nasal trigeminal system exerts an inhibitory effect on the with the results from Cain & Murphy (1980) although it has to be perception of odor intensity, and that CGRP-mediated inhibi- pointed out that, in contrast to the previous study with the bimodal tion of olfactory receptor neurons may contribute to this n-amyl butyrate stimulus, we used a highly selective olfactory stim- cross-modal interaction. ulus to prevent trigeminal activation by the odorous stimulus itself. Our EOG experiments on rat olfactory epithelium were designed Acknowledgements to clarify whether peripheral processes contribute to the inhibitory trigeminal effect. We circumvented the issue of CO2-induced pH We thank Professor Trese Leinders-Zufall for introducing us to the sub- changes by directly applying the trigeminal neuropeptides CGRP to merged-EOG technique and Dr Johannes Reisert for the design of the air- phase EOG. We are indebted to Volker Gudziol for his help with the bioptic the apical surface of the olfactory epithelium. CGRP is present in material. This work was supported by grants from the CellNetworks Cluster this tissue, enclosed in varicosities of the trigeminal fibres that tra- of Excellence (P.D.) and by the DFG Priority Program 1392 ‘Integrative verse the neuroepithelium up to the apical border at a density of Analysis of Olfaction’ (S.F. and T.H). 400–600 fibres per mm2. Thus, the source of CGRP is a dense sys- fi ~ l tem of parallel bres with a maximum distance of 50 m between Abbreviations each receptor cell and the next trigeminal fibre. This distance may be a decisive factor in the speed of the trigeminal impact on olfac- AIC, allylisothiocyanate (mustard oil); CGRP, calcitonin-gene related tion. To our knowledge, neuropeptide diffusion rates and ranges peptide; CRCP, CGRP-receptor component protein; CRLR, calcitonin receptor-like receptor; EOG, electroolfactogram or electroolfactography; FACS, have not been determined for olfactory epithelium. They may be fluorescence activated cell sorting; GFP, green fluorescent protein; IAA, approached in future by diffusion modelling based on microdialysis isoamyl acetate; IBMX, 3-isobutyl-1-methylxanthine; OMP, olfactory marker research (Bungay et al., 2011). Meanwhile, our findings are in protein; PEA, 2-phenylethyl alcohol; RAMP1, receptor activity-modifying accordance with earlier studies which have demonstrated that rodent protein 1. olfactory epithelium is innervated by the anterior ethmoidal nerve, which branches off the ophthalmic division of the trigeminal system. The epithelium mainly contains peptidergic nociceptive nerve termi- References nals, with CGRP and substance P as neuropeptides (Silverman & Bakalyar, H.A. & Reed, R.R. (1990) Identification of a specialized Kruger, 1989; Finger et al., 1990; Finger & Bottger, 1993; Schaefer adenylyl cyclase that may mediate odorant detection. Science, 250, – et al., 2002). Our immunohistochemical data show that receptors for 1403 1406. Bensafi, M., Iannilli, E., Gerber, J. & Hummel, T. (2008) Neural coding of CGRP and the receptor-associated protein CRCP were detected all stimulus concentration in the human olfactory and intranasal trigeminal over the olfactory receptor neurons, from the apical sensory cilia to systems. Neuroscience, 154, 832–838. + the axons in the lamina propria. Simultaneous application of PEA Billig, G.M., Pal, B., Fidzinski, P. & Jentsch, T.J. (2011) Ca2 -activated 14 – and CGRP reduced the odor response, both upon single pulse-like Cl- currents are dispensable for olfaction. Nat. Neurosci., ,763 769. Boccaccio, A. & Menini, A. (2007) Temporal development of cyclic nucleo- stimulations and with repeated odor stimulation during continuous tide-gated and Ca2+-activated ClÀ currents in isolated mouse olfactory sen- CGRP superfusion. This finding is consistent with our psychophysi- sory neurons. J. Neurophysiol., 98, 153–160. cal data. It points to a direct channel of interaction between the two Bouvet, J.F., Delaleu, J.C. & Holley, A. (1987) Olfactory receptor cell sensory systems, provided by CGRP release from trigeminal fibres function is affected by trigeminal nerve activity. Neurosci. Lett., 77, – and by CGRP receptors on olfactory receptor neurons. While differ- 181 186. Bouvet, J.F., Delaleu, J.C. & Holley, A. (1988) The activity of olfactory ences in chemosensory function between humans and rats are obvi- receptor cells is affected by acetylcholine and substance P. Neurosci. Res., ous, our results disclose similar inhibitory trigeminal effects on the 5, 214–223. olfactory system in the two organisms. This finding corroborates the Brand, G. (2006) Olfactory/trigeminal interactions in nasal chemoreception. 30 – suitability of the rodent system as model for further research into Neurosci. Biobehav. Rev., , 908 917. Bungay, P.M., Sumbria, R.K. & Bickel, U. (2011) Unifying the mathematical chemosensory interactions. modeling of in vivo and in vitro microdialysis. J. Pharm. Biomed. Anal., It has previously been demonstrated that CO2 stimulation 55,54–63. causes CGRP release from trigeminal neurons (Vause et al., Cain, W.S. (1977) Bilateral interaction in olfaction. Nature, 268,50–52.

Cain, W.S. & Murphy, C.L. (1980) Interaction between chemoreceptive Kimoto, M., Iwai, S., Maeda, T., Yura, Y., Fernley, R.T. & Ogawa, Y. modalities of odour and irritation. Nature, 284, 255–257. (2004) Carbonic anhydrase VI in the mouse nasal gland. J. Histochem. Cy- Coates, E.L. (2001) Olfactory CO(2) chemoreceptors. Respir. Physiol., 129, tochem., 52, 1057–1062. 219–229. Kleene, S.J. (2008) The electrochemical basis of odor transduction in verte- Cometto-Muniz, J.E. & Hernandez, S.M. (1990) Odorous and pungent brate olfactory cilia. Chem. Senses, 33, 839–859. attributes of mixed and unmixed odorants. Percept. Psychophys., 47, Kobal, G. & Hummel, C. (1988) Cerebral chemosensory evoked 391–399. potentials elicited by chemical stimulation of the human olfactory Croy, I., Negoias, S., Novakova, L., Landis, B.N. & Hummel, T. (2012) and respiratory nasal mucosa. Electroencephalogr. Clin. Neurophysi- Learning about the functions of the olfactory system from people without ol., 71,241–250. a sense of smell. PLoS ONE, 7, e33365. Kratskin, I., Hummel, T., Hastings, L. & Doty, R. (2000) 3-Methylindole Dauner, K., Lissmann, J., Jeridi, S., Frings, S. & Mohrlen, F. (2012) Expres- alters both olfactory and trigeminal nasal mucosal potentials in rats. Neu- sion patterns of anoctamin 1 and anoctamin 2 chloride channels in the roReport, 11, 2195–2197. mammalian nose. Cell Tiss. Res., 347, 327–341. Leopold, D.A., Hummel, T., Schwob, J.E., Hong, S.C., Knecht, M. & Kobal, Doty, R.L., Brugger, W.E., Jurs, P.C., Orndorff, M.A., Snyder, P.J. & G. (2000) Anterior distribution of human olfactory epithelium. Laryngo- Lowry, L.D. (1978) Intranasal trigeminal stimulation from odorous vola- scope, 110, 417–421. tiles: psychometric responses from anosmic and normal humans. Physiol. Liberles, S.D. & Buck, L.B. (2006) A second class of chemosensory recep- Behav., 20, 175–185. tors in the olfactory epithelium. Nature, 442, 645–650. Evans, B.N., Rosenblatt, M.I., Mnayer, L.O., Oliver, K.R. & Dickerson, I.M. Lin, W., Ogura, T., Margolskee, R.F., Finger, T.E. & Restrepo, D. (2008) (2000) CGRP-RCP, a novel protein required for signal transduction at cal- TRPM5-expressing solitary chemosensory cells respond to odorous irri- citonin gene-related peptide and adrenomedullin receptors. J. Biol. Chem., tants. J. Neurophysiol., 99, 1451–1460. 275, 31438–31443. Livermore, A., Hummel, T. & Kobal, G. (1992) Chemosensory event-related Ferris, K.E., Clark, R.D. & Coates, E.L. (2007) Topical inhibition of nasal potentials in the investigation of interactions between the olfactory and the carbonic anhydrase affects the CO2 detection threshold in rats. Chem. somatosensory (trigeminal) systems. Electroencephalogr. Clin. Neurophysi- Senses, 32, 263–271. ol., 83, 201–210. Finger, T.E. & Bottger, B. (1993) Peripheral peptidergic fibers of the trigem- Luo, M. (2008) The necklace olfactory system in mammals. J. Neurogenet., inal nerve in the olfactory bulb of the rat. J. Comp. Neurol., 334, 22, 229–238. 117–124. Ma, M. (2007) Encoding olfactory signals via multiple chemosensory sys- Finger, T.E., St Jeor, V.L., Kinnamon, J.C. & Silver, W.L. (1990) tems. Crit. Rev. Biochem. Mol. Biol., 42, 463–480. Ultrastructure of substance P- and CGRP-immunoreactive nerve Ma, M. (2010) Multiple olfactory subsystems convey various sensory signals. fibers in the nasal epithelium of rodents. J. Comp. Neurol., 294, In Menini, A. (Ed.), Neurobiology of Olfaction. CRC Press, Boca Raton, 293–305. FL, pp. 225–240. Frasnelli, J., Schuster, B. & Hummel, T. (2007) Interactions between olfac- McLatchie, L.M., Fraser, N.J., Main, M.J., Wise, A., Brown, J., Thompson, tion and the trigeminal system: what can be learned from olfactory loss. N., Solari, R., Lee, M.G. & Foord, S.M. (1998) RAMPs regulate the trans- Cereb. Cortex, 17, 2268–2275. port and ligand specificity of the calcitonin-receptor-like receptor. Nature, Getchell, M.L., Bouvet, J.F., Finger, T.E., Holley, A. & Getchell, T.V. 393, 333–339. (1989) Peptidergic regulation of secretory activity in amphibian olfactory Munger, S.D., Leinders-Zufall, T. & Zufall, F. (2009) Subsystem organiza- mucosa: immunohistochemistry, neural stimulation, and pharmacology. tion of the mammalian sense of smell. Ann. Rev. Physiol., 71, 115– Cell Tiss. Res., 256, 381–389. 140. Hansen, A. & Finger, T.E. (2008) Is TrpM5 a reliable marker for chemosen- Nickell, W.T., Kleene, N.K. & Kleene, S.J. (2007) Mechanisms of neuronal sory cells? Multiple types of microvillous cells in the main olfactory epi- chloride accumulation in intact mouse olfactory epithelium. J. Physiol., thelium of mice. BMC Neurosci., 9, 115. 583, 1005–1020. Hengl, T., Kaneko, H., Dauner, K., Vocke, K., Frings, S. & Mohrlen, F. Pfeuffer, E., Mollner, S., Lancet, D. & Pfeuffer, T. (1989) Olfactory adenylyl (2010) Molecular components of signal amplification in olfactory sensory cyclase. Identification and purification of a novel enzyme form. J. Biol. cilia. Proc. Natl. Acad. Sci. USA, 107, 6052–6057. Chem., 264, 18803–18807. Hu, J., Zhong, C., Ding, C., Chi, Q., Walz, A., Mombaerts, P., Matsun- Pifferi, S., Dibattista, M. & Menini, A. (2009) TMEM16B induces chloride ami, H. & Luo, M. (2007) Detection of near-atmospheric concentra- currents activated by calcium in mammalian cells. Pflugers Arch., 458, tions of CO2 by an olfactory subsystem in the mouse. Science, 317, 1023–1038. 953–957. Potter, S.M., Zheng, C., Koos, D.S., Feinstein, P., Fraser, S.E. & Hummel, T. & Livermore, A. (2002) Intranasal chemosensory function of Mombaerts, P. (2001) Structure and emergence of specific olfactory glo- the trigeminal nerve and aspects of its relation to olfaction. Int. Arch. Oc- meruli in the mouse. J. Neurosci., 21, 9713–9723. cup. Environ. Health, 75, 305–313. Schaefer, M.L., Bottger, B., Silver, W.L. & Finger, T.E. (2002) Trigeminal Hummel, T., Sekinger, B., Wolf, S.R., Pauli, E. & Kobal, G. (1997) ‘Sniffin’ collaterals in the nasal epithelium and olfactory bulb: a potential route for sticks’: olfactory performance assessed by the combined testing of odor direct modulation of olfactory information by trigeminal stimuli. J. Comp. identification, odor discrimination and olfactory threshold. Chem. Senses, Neurol., 444, 221–226. 22,39–52. Schaeren-Wiemers, N. & Gerfin-Moser, A. (1993) A single protocol to detect Hummel, T., Mohammadian, P., Marchl, R., Kobal, G. & Lotsch, J. (2003) transcripts of various types and expression levels in neural tissue and cul- Pain in the trigeminal system: irritation of the nasal mucosa using short- tured cells: in situ hybridization using digoxigenin-labelled cRNA probes. and long-lasting stimuli. Int. J. Psychophysiol., 47, 147–158. Histochemistry, 100, 431–440. Iannilli, E., Gerber, J., Frasnelli, J. & Hummel, T. (2007) Intranasal trigemi- Shusterman, D. & Avila, P.C. (2003) Real-time monitoring of nasal mucosal nal function in subjects with and without an intact sense of smell. Brain pH during carbon dioxide stimulation: implications for stimulus dynamics. Res., 1139, 235–244. Chem. Senses, 28, 595–601. Iannilli, E., Bitter, T., Gudziol, H., Burmeister, H.P., Mentzel, H.J., Chopra, Silver, W.L. (1992) Neural and pharmacological basis for nasal irritation. A.P. & Hummel, T. (2011) Differences in anosmic and normosmic group Ann. NY Acad. Sci., 641, 152–163. in bimodal odorant perception: a functional- MRI study. Rhinology, 49, Silver, W.L. & Moulton, D.G. (1982) Chemosensitivity of rat nasal trigemi- 458–463. nal receptors. Physiol. Behav., 28, 927–931. Inokuchi, A., Kimmelman, C.P. & Snow, J.B. Jr. (1993) Convergence of Silverman, J.D. & Kruger, L. (1989) Calcitonin-gene-related-peptide-immu- olfactory and nasotrigeminal inputs and possible trigeminal contributions noreactive innervation of the rat head with emphasis on specialized sen- to olfactory responses in the rat thalamus. Eur. Arch. Otorhinolaryngol., sory structures. J. Comp. Neurol., 280, 303–330. 249, 473–477. Spehr, M., Wetzel, C.H., Hatt, H. & Ache, B.W. (2002) 3-phosphoinositides Jacquot, L., Monnin, J. & Brand, G. (2004) Influence of nasal trigeminal modulate cyclic nucleotide signaling in olfactory receptor neurons. Neuron, stimuli on olfactory sensitivity. C. R. Biol., 327, 305–311. 33, 731–739. Jordt, S.E., Bautista, D.M., Chuang, H.H., McKemy, D.D., Zygmunt, P.M., Stephan, A.B., Shum, E.Y., Hirsh, S., Cygnar, K.D., Reisert, J. & Zhao, H. Hogestatt, E.D., Meng, I.D. & Julius, D. (2004) Mustard oils and cannabi- (2009) ANO2 is the cilial calcium-activated chloride channel that noids excite sensory nerve fibres through the TRP channel ANKTM1. Nat- may mediate olfactory amplification. Proc. Natl. Acad. Sci. USA, 106, ure, 427, 260–265. 11776–11781.

Tarun, A.S., Bryant, B., Zhai, W., Solomon, C. & Shusterman, D. (2003) Wang, Y.Y., Chang, R.B., Allgood, S.D., Silver, W.L. & Liman, E.R. Gene expression for carbonic anhydrase isoenzymes in human nasal (2011) A TRPA1-dependent mechanism for the pungent sensation of weak mucosa. Chem. Senses, 28, 621–629. acids. J. Gen. Physiol., 137, 493–505. Ukhanov, K., Corey, E.A., Brunert, D., Klasen, K. & Ache, B.W. (2010) Wetzel, C.H., Oles, M., Wellerdieck, C., Kuczkowiak, M., Gisselmann, G. & Inhibitory odorant signaling in Mammalian olfactory receptor neurons. J. Hatt, H. (1999) Speciﬁcity and sensitivity of a human olfactory receptor Neurophysiol., 103, 1114–1122. functionally expressed in human embryonic kidney 293 cells and Xenopus Vause, C., Bowen, E., Spierings, E. & Durham, P. (2007) Effect of carbon Laevis oocytes. J. Neurosci., 19, 7426–7433. dioxide on calcitonin gene-related peptide secretion from trigeminal neu- Wise, P.M., Wysocki, C.J. & Radil, T. (2003) Time-intensity ratings of nasal rons. Headache, 47, 1385–1397. irritation from carbon dioxide. Chem. Senses, 28, 751–760. Walker, C.S., Conner, A.C., Poyner, D.R. & Hay, D.L. (2010) Regulation of Yan, C., Zhao, A.Z., Bentley, J.K., Loughney, K., Ferguson, K. & Beavo, J. signal transduction by calcitonin gene-related peptide receptors. Trends A. (1995) Molecular cloning and characterization of a calmodulin-depen- Pharmacol. Sci., 31, 476–483. dent phosphodiesterase enriched in olfactory sensory neurons. Proc. Natl. Wang, Y.Y., Chang, R.B., Waters, H.N., McKemy, D.D. & Liman, E.R. Acad. Sci. USA, 92, 9677–9681. (2008) The nociceptor ion channel TRPA1 is potentiated and Youngentob, S.L., Hornung, D.E. & Mozell, M.M. (1991) Determination of car- inactivated by permeating calcium ions. J. Biol. Chem., 283, 32691– bon dioxide detection thresholds in trained rats. Physiol. Behav., 49,21–26. 32703. Zielinski, B.S., Getchell, M.L. & Getchell, T.V. (1989) Ultrastructural evi- Wang, Y.Y., Chang, R.B. & Liman, E.R. (2010) TRPA1 is a component of dence for peptidergic innervation of the apical region of frog olfactory epi- the nociceptive response to CO2. J. Neurosci., 30, 12958–12963. thelium. Brain Res., 492, 361–365.