Abstract Capsaicin-sensitive sensory C-fibers derived from vagal ganglia innervate the visceral organs, and respond to inflammatory mediators and noxious stimuli. These neurons play an important role in maintenance of visceral homeostasis, and contribute to the symptoms of visceral inflammatory diseases. Vagal sensory neurons are located in two ganglia, the jugular ganglia (derived from the neural crest), and the nodose ganglia (from the epibranchial placodes). The functional difference, especially in response to immune mediators, between jugular and nodose neurons is not fully understood. In this study, we microscopically isolated murine nodose and jugular capsaicin-sensitive / Trpv1-expressing C-fiber neurons and performed transcriptome profiling using ultra-low input RNA sequencing. RNAseq detected genes with significantly differential expression in jugular and nodose neurons, which were mostly involved in neural functions. Transcriptional regulators, including Cited1, Hoxb5 and Prdm12 showed distinct expression patterns in the two C-fiber neuronal populations. Common and specific expression of immune receptor proteins was characterized in each neuronal type. The expression of immune receptors that have received little or no attention from vagal sensory biologists is highlighted including receptors for certain chemokines (CXCLs), interleukins (IL-4) and interferons (IFNα, IFNγ). Stimulation of immune receptors with their cognate ligands led to activation of the C-fibers in isolated functional assays. Introduction The vagal sensory nervous system plays an important role in maintenance of visceral homeostasis. A large subset of sensory nerves in the vagus nerves can also respond to inflammatory mediators and noxious stimuli [[44]1]. Afferent nerves in the skin that respond to noxious stimuli were coined by Sherrington as “nociceptors” where he concluded they “provide the skin a so to say sense of its own potential injury” [[45]2]. There is an advantage to the host if visceral organs also have a mechanism to provide a sense of its own potential injury, and this may explain why large numbers of visceral afferent nerves have nociceptive properties (are stimulated by inflammatory mediators and noxious stimuli). As with the somatosensory system, visceral nociceptors are typically capsaicin-sensitive slow-conducting C-fibers. Dysregulation of nociceptive function likely participates in the symptoms of visceral inflammatory diseases [[46]3–[47]9]. A better understanding of activating mechanisms of visceral nociceptors may therefore uncover novel targets for new therapeutic strategies for these disorders. In addition to dorsal root ganglia (DRG), a sizeable proportion of the visceral nociceptive innervation is derived from small capsaicin-sensitive (TRPV1-expressing) C-fiber neurons in vagal sensory ganglia. The majority of vagal C-fibers are derived from neurons situated in the nodose ganglia. Embryologically, these neurons arise from the epibranchial placodes. Another less appreciated component of the vagal C-fiber innervation of visceral tissues is the fibers derived from neurons situated in the jugular (supranodose) ganglia. These neurons arise embryonically not from the epibranchial placodes, but like the spinal afferent neurons in the dorsal root ganglion (DRG), from the neural crest [[48]10]. Whereas a proportion of the jugular neurons are strictly speaking non-vagal as they project their axons to the ear via the auricular nerve, and also to the oral pharynx via pharyngeal nerves, retrograde tracing studies in many mammals including, mice[[49]11, [50]12], rat[[51]13, [52]14], guinea pig [[53]15, [54]16], cats[[55]17, [56]18], dogs[[57]19] and monkeys[[58]20, [59]21] reveal numerous jugular neurons that project axons down the vagus nerves. The relative contribution from placodal vs. neural crest vagal afferent innervation likely depends on the visceral organ in question, with thoracic viscera generally receiving a much richer jugular afferent innervation than sub-diaphragmatic viscera [[60]22]. We and others have noted that the respiratory tract, esophagus, and heart are innervated by large numbers of nodose and jugular neurons, most of which have nociceptive properties [[61]22–[62]27]. The nodose and jugular C-fibers share certain C-fiber defining phenotypic properties, including high mechanical threshold and responsiveness to capsaicin. However, the nodose vs. jugular differentiation is important to consider because the receptor expression and activation profile of placodal (nodose) C-fibers differs considerably from neural crest derived (jugular) C-fibers innervating the same tissue [[63]11, [64]12, [65]28–[66]32]. Moreover, recent studies are revealing that vagal jugular afferent nerves terminate in distinct regions of the central nervous system from nodose afferent nerves, with the former terminating largely in the paratrigeminal nucleus and the later terminating in the nucleus of the solitary tract [[67]33, [68]34]. This may explain why activation of jugular and nodose nociceptors within a given tissue, e.g. respiratory tract, can lead to distinctly different reflexes [[69]35]. The nodose and jugular neurons in the mouse typically coexist within an elongated ganglion, which is often inappropriately referred to simply as the nodose ganglion [[70]23]. Although microarrays and RNAseq analyses have been carried out on mouse vagal sensory neurons, these studies likely reflect a mixture of both placodal-derived nodose neurons and neural crest-derived jugular neurons as well as capsaicin-sensitive and–insensitive neurons [[71]36, [72]37]. In this study we individually isolated capsaicin-sensitive nodose and capsaicin-sensitive jugular vagal sensory neurons using two independent methods, and performed ultra-low input RNA sequencing to characterize the global gene expression in these two types of neurons. To add new insights into this key question we specifically focused our analysis on the expression of receptors for inflammatory mediators. Materials and methods Isolation of neurons based on capsaicin response All experiments were approved by the Johns Hopkins Animal Use and Care Committee. The right and left vagal sensory ganglia were isolated from wildtype mice. We have previously reported that in mice the jugular and nodose neurons typically form a fused jugular/nodose ganglion (JNG) complex [[73]12] ([74]Fig 1A and 1B). The jugular neurons are situated in the rostral portion of the ganglia, whereas the caudal portion is composed nearly exclusively of nodose neurons ([75]Fig 1B). Therefore, from each JNG the most rostral (jugular) and the most caudal (nodose) portion were used. The ganglia were enzymatically dissociated and the responsiveness of individual neurons to the selective activator of nodose C-fibers, purinergic P2X agonist α,β-methylene-ATP (10 μM) and to the general C-fiber activator TRPV1 agonist capsaicin (1μM) was evaluated by Fura-2 AM intracellular calcium assay as previously described [[76]11, [77]12]. After determining the responsiveness to agonists, the neurons were collected under the microscope into a glass pipette (pulled to tip diameter 50–150 μm) by applying negative pressure (up to 5 neurons/ pipette). The pipette tip was then broken in a PCR tube containing RNAse inhibitor, immediately frozen on dry ice and stored at -80°C. A total of 31 α,β-methylene-ATP-unresponsive, capsaicin-responsive neurons (jugular C-fiber neurons) from the rostral portion (15 cells were pooled for Replicate1 and 16 cells were pooled for Replicate2) and 29 α,β-methylene-ATP-responsive, capsaicin-responsive neurons (nodose C-fiber neurons) from caudal portion (14 cells were pooled for Replicate1 and 15 cells were pooled for Replicate2) were collected. Care was taken to avoid adhering cells, debris and glial cells. Fig 1. Jugular and nodose C-fiber neurons were isolated based on capsaicin sensitivity and Trpv1 expression. [78]Fig 1 [79]Open in a new tab (A) Dissection of right jugular/nodose complex in the mouse. SCG denotes superior cervical ganglion (modified from [[80]12]. (B) The rostral part of the jugular/nodose ganglion (JNG) complex is formed by jugular (J) neurons while the caudal part is formed by nodose (N) neurons. The jugular (neural crest-derived) neurons are visualized by X-Gal staining while nodose (placodes-derived) neurons remain unstained in Wnt1Cre/R26R mouse [[81]12]. Neurons isolated from the rostral section of the ganglion (J) were considered jugular neurons, whereas those from the more caudal aspect (N) were considered nodose neurons. The upper-central part of the ganglion (X) often comprise a mixture of nodose and jugular neurons and was avoided. (C-D) The collection of capsaicin-responsive / Trpv1-positive C-fiber neurons. The jugular or nodose portions of JNG were enzymatically dissociated, capsaicin-sensitive C-fiber neurons were identified by one of two methods and collected. (C) Identification of C-fiber neurons by capsaicin-responsiveness by intracellular calcium assay in wild-type mice (capsaicin selection). (D) Identification of C-fiber neurons by tdTomato immunofluorescence in the Trpv1-tdTomato mice (Trpv1-tdTomato selection). Generation of Trpv1-Td-tomato reporter mice and isolation of neurons based on Trpv1 expression Trpv1-tdTomato mice was generated by mating Trpv1^Cre mice with tdTomato reporter strain Gt(ROSA)26Sortm14(CAG-tdTomato)Hze using standard approach. The ganglia were dissected and dissociated the same way as for capsaicin response-based isolation described above. Individual red fluorescent (Td-tomato-positive) neurons from the most caudal nodose portion and from the most rostral jugular portion were collected under fluorescent microscope by using rhodamine (red) filter set. Total of 60 nodose and 60 jugular Td-tomato-positive neurons (30 cells were pooled for each replicate) were collected from 2 mice ([82]Fig 1C). For comparison, 30 Td-tomato-negative nodose neurons were collected from the same mice (15 cells were pooled for each replicate). Ultra-low input RNA sequencing Cells from capsaicin response based isolation (capsaicin-responsive jugular and nodose neurons) were processed at the Johns Hopkins Deep Sequencing and Microarray Core. Cells from TRPV1 expression based isolation (Td-tomato-positive jugular and nodose neurons, and Td-tomato-negative nodose neurons) were processed at MedImmune Deep Sequencing and Microarray Core independently following the same protocol. Briefly, RT-PCR and cDNA synthesis was performed using SMARTer Ultra Low Input RNA Kit v3 following manufacturer’s protocol (Clontech). Cells were used directly as starting material without RNA extraction to avoid extra loss of RNA. First stand cDNA was synthesized using 3’ SMART CDS Primer II A and SMARTer IIA Oligonucleotide with SMARTScribe Reverse Transcriptase. dT priming was used to eliminate DNA contamination. The cDNA was then amplified by 15 cycles of LD PCR and purified using Agencourt AMPure XP kit. Amplified cDNA was validated using Agilent 2100 BioAnalyzer. Illumina paired-end DNA sequencing library was generated using Low Input Library Prep Kit v2 following manufacturer’s protocol (Clontech). The libraries were purified using AMPure beads and quality validated using Agilent 2100 BioAnalyzer. Sequencing was performed on Hiseq 2000. 50bp paired-end sequencing was performed with ~30 million reads generated per sample for capsaicin response based isolation, and ~250 million reads generated per sample for TRPV1 expression based isolation. RNA sequencing data analysis Quality of the RNAseq data, such as the overall sequencing score, over-represented reads, kmer presence, was evaluated using the FastQC package ([83]http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Sequencing reads were aligned to mouse reference genome mm10 using Tophat2 (v2.0.9) [[84]38] and Bowtie2 (2.1.0.0)[[85]39]. Default parameters were used, allowing maximum of 2 mismatches during alignment. Raw counts were generated using HTseq[[86]40]. Data was normalized and counts per million (CPM) values for each gene were generated using the Deseq2 package. Group comparison was performed in Deseq2 using generalized linear model assuming negative binomial distributions[[87]41]. False discovery rate (FDR) was generated with Benjamini and Hochberg correction. Genes with significant expression change was defined as fold change ≥ 2 and FDR ≤ 0.1. To improve the quality of analysis, additional filtering was applied to only keep the genes with CPM ≥ 10 in both replicates in jugular or nodose or both of the two populations. Functional annotation of the genes was performed in PANTHER classification system ([88]http://pantherdb.org/). Pathway analysis was performed using DAVID ([89]https://david.ncifcrf.gov/tools.jsp) against Gene Ontology biological process and molecular function database. Pathway enrichment FDR was generated from p values with Benjamini and Hochberg correction. Enriched pathways were ranked by FDR of enrichment. Heatmap and Principle component analysis were generated in R using log2 transformed counts per million (CPM) values. Other graphing and statistics were performed using R and Microsoft Office Excel. Raw and processed RNA sequencing data has been uploaded to Gene Expression Omnibus ([90]https://www.ncbi.nlm.nih.gov/geo/) with accession number [91]GSE102123. Intracellular Ca2+ measurements For measurement of intracellular Ca^2+ concentrations ([Ca^2+][i]), nodose neurons were isolated from Pirt-GCaMP3 heterozygote mice in which the genetically encoded Ca^2+ indicator GCaMP3 is expressed under the sensory neuron-specific Pirt promoter as previously described[[92]42]. Briefly, mice were killed by CO[2] inhalation and subsequent exsanguinations. Both sides of jugular/nodose ganglia were dissected and cleared of adhering connective tissues. The lower two thirds of ganglia (containing mostly nodose neurons) were cut out for subsequent enzymatic digestion using type 1A collagenase (2mg/ml) and dispase II (2mg/ml). The isolated nodose neurons were kept at 37 C in L-15 medium containing 10% of fetal bovine serum for use within 24 hours. GCaMP3 was excited at 488 nm, and the emission of its green fluorescence at 525 nm was captured by a Nikon DS camera (30–60 frames/min) under the control of imaging software NIS-Elements AR. The images were continuously captured before and during agonist application. The intensity of whole-cell fluorescence for each cell under study was measured as a function of time, normalized to the resting fluorescence (F/F[0]) and used as an index of the [Ca^2+][i]. The experiment was performed at room temperature. Bath solution contained (mM): NaCl 136, KCl 5.4, MgCl[2] 1, CaCl[2] 1.5, HEPES 10 and glucose 10 with pH adjusted to 7.35 with NaOH. The stock solution of mouse interferon-α A (PBL Assay Science, Piscataway, NJ), recombinant mouse interference-γ and recombinant human CXCL8 (R&D Systems, Minneapolis, MN) was prepared in PBS containing 0.1% BSA, and sphingosine-1 phosphate (Tocris, Bristol, UK) in ethanol. Extracellular recording The innervated isolated trachea-lung preparation was prepared as previously described [[93]43]. Briefly, the airways and lungs with their vagus nerve including jugular/nodose ganglia were placed in a dissecting dish containing Krebs bicarbonate buffer solution composed of (mM) 118 NaCl, 5.4 KCl, 1.0 NaH[2]PO[4], 1.2 MgSO[4], 1.9 CaCl[2], 25.0 NaHCO[3] and 11.1 dextrose, and equilibrated with 95% O[2] and 5% CO[2] (pH 7.2–7.4). The left jugular/nodose ganglion was gently pulled into the adjacent compartment of the chamber through a small hole and pinned. Both compartments were separately superfused with Krebs solution, which was warmed by a warming jacket (39–42°C) to keep airway tissues and ganglia at 37°C. Action potentials were recorded at the level of the cell body strategically positioned in the vagal sensory ganglion using a sharp extracellular glass electrode filled with 3M NaCl (impedance 2MΩ). The recorded action potentials were amplified (Microelectrode AC amplifier 1800; A-M Systems, Everett, WA, USA), filtered (0.3 kHz of low cut-off and 1 kHz of high cut-off), and monitored on an oscilloscope (TDS340; Tektronix, Beaverton, OR, USA) and a chart recorder (TA240; Gould, Valley View, OH, USA). The scaled output from the amplifier was captured and analyzed by a Macintosh computer using NerveOfIt software (Phocis, Baltimore, MD, USA). We have found that this technique does not record action potentials in “through fibers”. For example, positioning the electrode away from cell bodies on the vagus itself, fails to record action potential stimulated by receptive fields, or be electrical nerve stimulation. For measuring conduction velocity, an electrical stimulation (S44; Grass Instruments, Quincy,MA,USA) was applied on the core of the receptive field. The conduction velocity was calculated by dividing the distance along the nerve pathway by the time delay between the shock artifact and the action potential evoked by electrical stimulation. If a C-fiber (<1 m s^-1) was found, the recording was started. C-fibers were stimulated by 1 ml of vehicle, α,β-methylene ATP (10 μM), shingosine-1phosphate (S-1P) (0.1, 1 and 10 μM), or Thr-Phe-Leu-Leu-Arg-NH2 (TFLLR) (3 μM) injected into the lung through the trachea. Results Differential gene expression in Trpv1-tdTomato-positive jugular vs. nodose neurons We compared the transcriptome between jugular vs. nodose neurons in the capsaicin selection and Trpv1 selection dataset, respectively. Previously, we have studied the function and expression of several ion channels (P2X[2], P2X[3], 5-HT[3])[[94]11, [95]12, [96]28, [97]29], G-protein coupled receptors (PAR1, PAR2, adenosine A[1] receptor) [[98]31, [99]32], receptors for neurotrophins and neurotrophic factors (TrkA, TrkB and GFRα[3]) [[100]11, [101]12] and neuropeptides (PPT-A products) [[102]11, [103]12] that differentiate between nodose and jugular capsaicin-sensitive/TRPV1-positive C-fiber neurons using functional, immunohistochemical and/or RT-PCR assays. We used this knowledge to evaluate the validity of our approach. The predictions from previous studies were satisfied in every case in the transcriptome analysis of jugular and nodose neurons using either the capsaicin-sensitivity or Trpv1-tdTomato selection to delineate the C-fibers ([104]Table 1 and references therein).