Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 18 January 2019

Dorsal Root Ganglion Neuron Types and Their Functional Specialization

Abstract and Keywords

Primary sensory neurons of the dorsal root ganglion (DRG) respond and relay sensations that are felt, such as those for touch, pain, temperature, itch, and more. The ability to discriminate between the various types of stimuli is reflected by the existence of specialized DRG neurons tuned to respond to specific stimuli. Because of this, a comprehensive classification of DRG neurons is critical for determining exactly how somatosensation works and for providing insights into cell types involved during chronic pain. Here, we review the recent advances in unbiased classification of molecular types of DRG neurons in the perspective of known functions as well as predicted functions based on gene expression profiles. The data show that sensory neurons are organized in a basal structure of three cold-sensitive neuron types, five mechano-heat sensitive nociceptor types, four A-Low threshold mechanoreceptor types, five itch-mechano-heat–sensitive nociceptor types and a single C–low-threshold mechanoreceptor type with a strong relation between molecular neuron types and functional types. As a general feature, each neuron type displays a unique and predicable response profile; at the same time, most neuron types convey multiple modalities and intensities. Therefore, sensation is likely determined by the summation of ensembles of active primary afferent types. The new classification scheme will be instructive in determining the exact cellular and molecular mechanisms underlying somatosensation, facilitating the development of rational strategies to identify causes for chronic pain.

Keywords: DRG, sensory neurons, pain, itch, thermosensation, single-cell RNA sequencing, classification, neuron types, specialization


The somatosensory system processes information that organisms “feel,” such as joint position, muscle stretch, pain, pressure, temperature, and touch. In contrast to other senses, somatosensation uses organs and tissues as receptive fields. These receptive fields contain nerve endings arising from primary sensory neurons of the trigeminal ganglia in the head region and of the dorsal root ganglia (DRG) in the trunk. The system is composed of a diverse array of peripheral nerve endings specialized to detect various sensory modalities. These modalities include the detection of innocuous (nonharmful) stimuli, such as warmth, cooling, and light touch, as well as noxious (damage-causing) stimuli, such as touching a hot stove, holding your hand in ice water, or a dropping a brick on your foot. In addition to discrete sensations, the somatosensory system conveys information about the position of your body in space, allowing you to determine when your arm is stretched or folded. A vital function of the somatosensory system is to provide information about the occurrence or threat of injury. Hence, the sensation of pain by its aversive nature helps detect and protect against potentially damaging stimuli in the external environment. It is also a warning system for internal damage, as represented by deep pain and thermal sensation from organs and tissues.

The initial processing of somatosensory information from the periphery is mediated by these primary sensory neurons, which terminate on second-order sensory neurons in the central nervous system. Primary sensory neurons are pseudounipolar, with one process attached to the cell body that bifurcates into two branches. One branch, termed the distal process, projects to cutaneous or deep peripheral tissues, and the other, referred to as the proximal process, terminates in the dorsal horn of the spinal cord or sensory nuclei of the brain stem. Thus, the distal process, depending the location of the terminal, can be activated by internal or external stimuli. Activation results in depolarization and action potential propagation to the central nervous system. This makes primary sensory neurons unique in the nervous system because an action potential initiated in the terminal of the distal process can propagate to a presynaptic terminal in the spinal cord and higher central nervous system regions without the need for passing through the cell body (Amir & Devor, 2003). The distal process can end either in specialized end organs, which filter sensory stimuli, or as free terminals. Stimuli that result in an action potential being initiated at the distal nerve process can be chemical, thermal, or mechanical in nature.

The existence of distinct types of sensory neurons with different response profiles underlies the ability to discriminate between different types of sensations, such as warmth, cold, pain, itch, as well as mechanical senses like touch and proprioception. It is mostly the unmyelinated neurons with free nerve endings in the skin that initiate and transduce noxious stimuli, thermoception, and pruriception (itch sensation), although some myelinated Aδ and Aβ fibers are also involved in the transduction of noxious stimuli (Djouhri & Lawson, 2004). The neurons that respond preferentially to noxious stimuli are termed nociceptors and typically have high thresholds of activation. In addition, nociceptors can be regarded as either polymodal, responding to a variety of noxious stimuli (i.e., mechanical and intense heat), or unimodal, responding to only a single type of noxious stimulation. Nociception (i.e., the process by which we detect noxious stimuli) is a vital function because it provides information about the occurrence or threat of injury; however, some neuronal types are instead devoted to innocuous (nonpainful) stimuli, such as touch sensations and proprioceptive functions, the latter for control and awareness of body position and balance. These mechanosensitive neurons not only are myelinated low-threshold mechanoreceptive neurons (LTMRs), but also include some unmyelinated neurons, termed C–low-threshold mechanoreceptors (C-LTMRs) that become activated in response to light mechanical stimulation (Abraira & Ginty, 2013). With respect to the quality of a particular sensation, a topographical order in the ascending system is more or less maintained. For instance, when the activity of a sensory neuron is triggered by a specific stimulus, such as heat, this signal eventually passes to an area in the brain uniquely attributed to that area on the body, therefore allowing the processed stimulus to be felt at the correct location. Nonnociceptive stimuli typically have higher fidelity and acuity than those that are noxious. Furthermore, painful sensations involve not only perception, but also emotional and motivational components (McMahon & Koltzenburg, 2006).

Historically, sensory neurons have been categorized by their degree of myelination and associated conduction velocity. This classification gives rise to four main class types: heavily and moderately myelinated A fibers (Aα and Aβ, respectively); thinly myelinated Aδ fibers; and unmyelinated C fibers. The degree of myelination has a profound effect on the rate of action potential conduction, with heavily myelinated A fibers conducting at a rate of 70–120 m.s−1, while unmyelinated C fibers conduct far more slowly at a rate of 0.5–2 m.s−1. These neurophysiological differences in conduction velocity have also been associated with discrete functions. Heavily and moderately myelinated A fibers are largely associated with the detection of light touch as well as relaying proprioceptive information; however, numerous studies have also reported a significant number of nociceptors within the A-fiber populations (as determined by conduction velocity properties), with relative proportions ranging from 18% to 65%, depending on species (Djouhri & Lawson, 2004). Thinly myelinated Aδ fibers are associated with the detection of a variety of innocuous and noxious stimuli, while unmyelinated C fibers are mostly involved in detecting nociceptive stimuli; however, some C fibers are also responsible for relaying innocuous light touch stimulation. Importantly, given the heterogeneity of fiber responses to both innocuous and noxious stimuli between and within different fiber classes, fiber classification alone does not provide a sufficient means to predict neuronal function.

Sensory neurons can also be classified on the basis of their sensitivity to specific modality types and intensities. Over the decades, numerous electrophysiological studies have been undertaken to investigate the responses of sensory neurons to specific modality types and intensities across different species. These studies have shown that there is significant heterogeneity within the sensory neuron population, with some neurons only responding to a single innocuous or noxious modality, thus being unimodal in function, while others respond to a variety of modality types and intensities, thus being polymodal in function. Within the nociceptive population, the estimated proportion of unimodal and polymodal fibers varies considerably across the literature, with polymodal fibers making up anywhere between 35% and almost 100% of nociceptive C fibers, and unimodal fibers ranging from 10% to 15% for mechanical sensitivity and 10–25% for heat sensitivity (Dubin & Patapoutian, 2010). In addition, approximately 10–25% of nociceptors are unresponsive under basal conditions and are termed silent nociceptors.

Other, nonelectrophysiological approaches have also been used to investigate the sensitivities of sensory neurons in vivo. A recent in vivo imaging study investigating the modality profiles of all sensory neurons in response to plantar stimulation in mice showed that 19% of all responding neurons were responsive to both mechanical and noxious heat stimulation (MH); 11% were responsive only to noxious heat stimulation (H); 4% were responsive to mechanical, noxious heat, and cold stimulation (MHC); and 56% were responsive only to mechanical stimulation (Wang et al., 2018). Similar observations have also been reported by other in vivo imaging studies (Emery et al., 2016). Such differences in observed levels of polymodality between electrophysiological and imaging studies is puzzling and likely due to several factors, including the technique used to measure neuronal response, the stimulation area being investigated, the physiological condition of the animal, as well as the selection criteria used to define polymodality. What is clear, however, is that within the sensory neuron population, specifically nociceptors, there is a wide variety of responses to different stimuli, which are significantly altered following injury (Emery et al., 2016; Kim et al., 2016; Smith-Edwards, DeBerry, Saloman, Davis, & Woodbury, 2016; Yarmolinsky et al., 2016).

Because different types of neurons respond and relay different types of somatosensation, a comprehensive classification of DRG neurons is critical for establishing the roles of the different neurons in basal sensation and during chronic pain. However, most methods until today have been based on measuring a limited number of features and, because of this, may contain insufficient information to accurately identify cell types. This chapter summarizes what we know about the molecular aspects of DRG sensory neurons, with an emphasis on molecular types of sensory neurons.

Classification Based on Neurochemistry

Neurochemical features of primary sensory neurons have been extensively studied and correlated to functional properties over the past decades. These studies are focused on identifying one or a few marker genes, with these genes indicating neuronal types with different functions. The markers have been used to study the function of neurons expressing the marker gene and for identification of peripheral and central projection patterns. The chosen genes often encode cell surface molecules (such as receptors, ion channels, and cell adhesion molecules), calcium-binding proteins, and neuropeptides. This work has been instrumental for deciphering how primary sensory neurons are organized. Consequently, the literature on assignment of function based on individual markers is extensive. Although new unbiased and more comprehensive classification strategies of DRG neuron types using single-cell RNA sequencing is available (see the section on molecular classification), knowledge gained from already-known features needs to be incorporated into a new unbiased classification. For DRG sensory neurons, two main morphologically defined neuronal subtypes were initially identified: the large light and the small dark neurons. Large light neurons stain for neurofilament 200 (RT97, gene Nefh) and are myelinated with a conduction velocity of Aα/β-neuron category (>12 m.s−1), or thinly myelinated Aδ neurons (2–12 m.s−1); small dark neurons are neurofilament 200–negative unmyelinated nociceptive C neurons (<1.3 m.s−1) (Lawson & Waddell, 1991). The size distribution of the Aδ group is skewed, with a peak of small cells and tail of medium-size cells, suggesting the existence of both LTMR and nociceptor types of Aδ neurons.

The large light neurons of the A-fiber LTMRs types are involved in innocuous touch sensation and proprioception. Mechanical touch stimulation broadly induces two distinct current types in responding neurons: rapidly adapting (RA) currents, which decay quickly after activation, and slowly adapting (SA) currents, which decay more slowly. Proprioceptive neurons innervating the Golgi tendon organs and muscle spindles can be more precisely identified by their expression of TrkC (Ntrk3) and the calcium-binding protein parvalbumin (Pvalb) (Ernfors, Lee, Kucera, & Jaenisch, 1994). A large number of studies have employed both mouse genetics and direct neurochemical staining to mark neurons and their processes to resolve subtypes of touch-sensitive neurons, mostly based on TrkB, TrkC, Ret, and calcium-binding protein expression. These studies correlate cell soma expression with immunoreactivity in fibers of sensory end organs such as Meissner corpuscles, Pacinian corpuscles, Merkel cells, and various types of hair follicle innervation (longitudinal lanceolate and circumferential) or with deficits in the corresponding gene knockout mice (Montano, Perez-Pinera, Garcia-Suarez, Cobo, & Vega, 2010). LTMR neuron types have also been classified using transcription factors such as MafA, c-Maf, Runx3, and Shox2 in combination with TrkB, TrkC, and Ret expression. Based on developmental patterns of expression and the neurochemistry of peripheral fibers (Bourane et al., 2009; Carroll, Lewin, Koltzenburg, Toyka, & Thoenen, 1998; L. Li et al., 2011; Luo, Enomoto, Rice, Milbrandt, & Ginty, 2009; Wende et al., 2012), a number of predictions could be made to identify neuron types by these markers (Lallemend & Ernfors, 2012). The results showed that Shox2 drives the generation of TrkB+ LTMRs, while RA neurons arise closely together during development as identified by Ret expression. However, the identification of neuron types in these studies was partly confounded by developmental changes in expression for the studied receptors and transcription factors.

More recent studies have made use of conditional strategies to genetically mark highly TrkB-expressing neurons. This strategy in combination with physiology has shown that TrkBhigh neurons are of the Aδ-LTMR type, terminating as lanceolate endings in hair follicles and mediating direction selectivity (Rutlin et al., 2014). Based on functionally overexpressing the TrkC ligand or generating genetic mouse reporter mice for TrkC, both the SA Merkel fibers as well as SA circumferential and Aβ-field LTMRs fiber in hair follicles are represented by neurons coexpressing TrkC and Ret (Bai et al., 2015; McIlwrath, Lawson, Anderson, Albers, & Koerber, 2007). Aβ RA-LTMRs innervating Meissner and Pacinian corpuscles express TrkBlow and calbindin (Calb) (Wende et al., 2012). Thus, based on expression of a few markers, several types of LTMR neuron types can be identified.

Additionally, nociceptors can be subdivided based on a limited number of neurochemical features. A major division is between peptidergic and nonpeptidergic neurons. Peptidergic neurons are defined by containing neuropeptides such as substance P (SP, Tac1) and calcitonin gene–related peptide (CGRP, Calca) or somatostatin (Som, SSt) (Hokfelt et al., 1976; Hokfelt, Kellerth, Nilsson, & Pernow, 1975; Wiesenfeld-Hallin et al., 1984) or those that do not contain these neuropeptides, but instead contain the fluoride-resistant acid phosphatase (FRAP) and bind the plant lectin Griffonia simplicifolia I-B4 (IB4) (Nagy & Hunt, 1982; Silverman & Kruger, 1988a, 1988b). Although some overlap between these populations has been reported (Carr, Yamamoto, & Nagy, 1990; Dalsgaard et al., 1984), it was not realized that the overlapping population consists of very different neuronal types. The peptidergic neurons are C- or Aδ-fiber neurons on the basis of conduction velocity, while nonpeptidergic neurons are all of C-fiber type (Lawson, McCarthy, & Prabhakar, 1996; McCarthy & Lawson, 1990). The majority of peptidergic neurons express the tyrosine kinase receptor TrkA (Ntrk1), which is the receptor for nerve growth factor (NGF) (Kaplan, Martin-Zanca, & Parada, 1991), while nonpeptidergic neurons do not (Averill, McMahon, Clary, Reichardt, & Priestley, 1995). Some peptidergic neurons also express neuropeptide Y2 receptor (Npy2r) (Brumovsky et al., 2005). Nonpeptidergic neurons instead express the tyrosine kinase glial cell line–derived neurotrophic factor receptor Ret along with the coreceptors GFRα1 and GFRα2 (Gfra1, Gfra2) (Bennett et al., 1998; Kashiba, Uchida, & Senba, 2001). Subpopulations of nonpeptidergic neurons express the Mas-related G protein–coupled receptors, which define at least three subgroups of IB4-binding and Ret-expressing nonpeptidergic neurons: Mrgprd (Mas1-related G protein–coupled receptor D), MrgprA1/MrgprC11, and MrgprB4 (Bender et al., 2002; Dong, Han, Zylka, Simon, & Anderson, 2001). Embedded among unmyelinated C fibers are also some neuronal types that are not nociceptors, but respond to cooling, and the C-LTMRs. Cooling neurons express the cold-activated ion channel TrpM8. RNA transcripts for the channel are present in both Aδ and C fibers that are TrkA positive but do not contain the heat-activated Trp channel TrpV1, CGRP, or neurofilament or display IB4 binding (Kobayashi et al., 2005; Peier et al., 2002). The C-LTMRs not only respond to pinprick and skin indentation but also are particularly responsive to innocuous gentle stroking of the hairy skin, but not glabrous skin, or to heat or chemical stimulation. These last fibers appear to mediate pleasant touch (Olausson et al., 2008; A. B. Vallbo, Olausson, & Wessberg, 1999). C-LTMRs are a separate type of nonpeptidergic Ret+ neuron that express tyrosine hydroxylase (Th) and Vglut3 but have no IB4 binding (Brumovsky, Villar, & Hokfelt, 2006; Seal et al., 2009), consistent with the fact that only about two thirds of Ret+ neurons bind IB4 (thus, also expressed in some A-fiber LTMR and in nonpeptidergic neurons). Nociceptors can also be distinguished according to their differential expression of channels that confer sensitivity to heat (TrpV1, TrpM2, TrpM3); cold (TrpM8); acidic milieu (acid-sensitive ion channels, ASICs); and various chemical irritants (TrpA1, Mrgprd) (Cook, Christensen, Tewari, McMahon, & Hamilton, 2018; Julius, 2013; Julius & Basbaum, 2001; Vriens & Voets, 2018).

Molecular Classification of DRG Sensory Neurons

Recent classification of primary afferents has made use of bulk transcriptional profiles of different types of DRG neurons. This includes the use of fluorescent-activated cell sorting of neurons expressing specific genes, such as Ntrk3 (TrkC) or Trpv1 (TrpV1), or by parallel quantitative polymerase chain reaction (PCR) of candidate genes in individual DRG neurons from three distinct neuronal populations (neurons that express IB4, neurons that express NaV1.8 but are negative for IB4, and those that express both NaV1.8 and Pvalb) (Goswami et al., 2014; Lee, Friese, Mielich, Sigrist, & Arber, 2012). These results provide insight into the transcriptomes of these groups of cells but were neither recordings of the transcriptome within single cells nor based on an unbiased strategy and hence do not represent a complete classification of sensory neuron types.

Single-cell RNA sequencing contains up to 28,000 features of data, is fully quantitative, and if performed without any selection of neuron types, allows unbiased identification of molecular types of neurons. Thus, single-cell RNA sequencing is more likely to identify cell types correctly than more biased strategies, such as the use of morphological or neurophysiological parameters, which contains markedly fewer and less quantitative features that can be used to accurately identify neuron types. It is therefore likely that molecularly identified neuronal types will be guiding future research and that morphological, physiological, and other features will be the most informative if defined within the domains of molecular types.

With respect to molecular analysis of sensory neurons, the complexity of the library and sequencing depth increase the accuracy and potential utility of the data; however, there are several sources of noise in single-cell transcriptome experiments, which may confound the results. These include both biological fluctuations and technical noise, caused by temperature differences, differences in sequencing depth between cells, pipetting errors, PCR bias during amplification, and reverse transcription efficiency. Only a fraction of the messenger RNAs (mRNAs) present in the cell are successfully converted for sequencing, and because mRNA levels are proportional to cell size, small cells with low mRNA content will generate greater errors than larger cells. This probably affects studies of primary sensory neurons more than cells of many other tissues, as neuronal size varies greatly in the DRG. However, technical noise can be alleviated by increasing the number of analyzed cells (Shapiro, Biezuner, & Linnarsson, 2013). Primary sensory neurons are also different from most other neurons in that they are intimately associated with satellite cells covering the cell soma of each neuron. Large DRG neurons associate with more satellite cells than small DRG neurons, and the dissociation of DRG neurons for single-cell RNA sequencing does not always successfully remove all satellite cells. However, satellite cells are very small compared to the neurons, and as mRNA content is proportional to cell size, incomplete removal of satellite cells is unlikely to affect the data significantly unless the samples are very deeply sequenced. Satellite cell contributions can also be mitigated by single-cell RNA sequencing of the satellite cell population to computationally subtract satellite cell signature genes.

Studies using unbiased single-cell RNA sequencing of somatosensory DRG neurons have been performed, with one study encompassing 197 neurons that were very deeply sequenced detecting 10,950 ± 1,218 genes per neuron without in vivo validation, with 17 distinct neuronal types predicted (C. L. Li et al., 2016). Data from this study were more recently reanalyzed, predicting only nine neuronal types (C. Li, Wang, Chen, & Zhang, 2018). In 2015, Usoskin et al. analyzed 622 neurons detecting about 3,900 ± 1,880 genes per neuron, but included in vivo validation of the predicted neuron types (Usoskin et al., 2015). We based this chapter on the Usoskin et al. (2015) classification, as the identified types in Li et al. (2016) largely concurred with Usoskin et al.’s work after reanalysis. Likewise, more recent resequencing of 1,580 DRG neurons confirmed the clusters/cell types identified by Usoskin et al. in 2015 (Zeisel et al., 2018).

Thus, the Usoskin et al. (2015) classification is consistent with ontogeny of DRG neuron types, with established functions of predicted neuron types, and furthermore has been confirmed to exist in vivo in the DRG. This study identified 11 neuronal types: one unmyelinated peptidergic C-fiber neuron type, termed PEP1, and one lightly myelinated Aδ-nociceptor population, termed PEP2; five types of neurofilament heavy chain (Nefh) expressing A-fiber neurons, of which three were LTMRs (termed NF1, NF2, NF3); and two proprioceptive populations (termed NF4 and NF5). Three types of “nonpeptidergic neurons” were identified, termed NP1, NP2, and NP3. However, NP2 and NP3 neuronal types contain markers for both peptidergic neurons (i.e., neuropeptides) and for nonpeptidergic neurons (P2rx3 expression and predicted IB4 binding). Finally, one C-LTMR neuron type expresses Vglut3 (SLC17A8) and Th and was therefore referred to as the TH population.

The more recent resequencing (Zeisel et al., 2018) (Figure 1) identified the same neuronal types, but because more neurons were sequenced, variability within some of the neuronal types identified by Usoskin et al. (2015) was observed, thus predicting subclusters within PEP1, NP1, and NP2 populations. For simplicity, in this chapter the PEP1 subclusters are called PEP1.1, PEP1.2, PEP1.3, and PEP1.4; the subclusters in NP1 are called NP1.1 and NP1.2; and NP2 subclusters are called NP2.1 and NP2.2. In addition, Zeisel et al. (2018) identified three Trpm8-expressing subclusters (Trpm8.1, Trpm8.2, Trpm8.3) that were not discovered by Usoskin et al. (2015) due to the limited numbers of neurons sequenced. LTMR neurons are underrepresented in numbers in the work of Zeisel et al. (2018) as compared to that of Usoskin et al. (2015); as a consequence, NF2 and NF3 could not be separated. Also, proprioceptors that were identified as two distinct, but highly similar, clusters by Usoskin et al. (2015) were identified as a single cluster in the newer data of Zeisel et al. (2018). The new data had a shallower sequencing depth than Usoskin et al. (2015) used and therefore instead relied on larger sample sizes. Thus, the data on LTMRs are likely to be more reliable in the work of Usoskin et al. (2015) than in the resequencing (Zeisel et al., 2018).

Dorsal Root Ganglion Neuron Types and Their Functional SpecializationClick to view larger

Figure 1. Transcriptomic organization of distinct mouse DRG sensory neuron clusters. t-SNE representation, each dot is one cell.

The most comprehensive classification therefore merges the identified LTMRs in Usoskin et al. (2015) and the nociceptors and proprioceptors from Zeisel et al. (2018). This results in a total of 18 neuron types: 3 Trpm8-expressing clusters, 5 PEP clusters, 3 NF clusters representing A-fiber LTMRs, 1 NF cluster with proprioceptors, 1 TH-containing C-LTMR cluster, 2 nonpeptidergic (NP1) clusters, 2 Mrgpra3-containing clusters (NP2), and finally 1 Sst/Nts-containing cluster (NP3) (Figure 2). Gene expression can be browsed among the different neuron types or downloaded ( [data from Usoskin et al., 2015] and for Zeisel et al.’s [2018] data).

Dorsal Root Ganglion Neuron Types and Their Functional SpecializationClick to view larger

Figure 2. Hierarchical organization and gene expression profiles of distinct subsets of mouse DRG sensory neurons. The PSPEP, PSNF and PSNP nomenclature of neuron types on top refers to the classification presented in Zeisel et al., 2018 (available at The TrpM8, PEP1, PEP2, NF, TH, NP nomenclature below represents the relationship of these neuron types to the classification presented in Usoskin et al., 2015, and is used throughout this chapter.

Hierarchical Relationship of Neuronal Types

The hierarchical organization of identified neuronal types also represents an estimation of the similarities between them. Based on the work of Zeisel et al. (2018), three major branches are evident, the first including Trpm8 and PEP classes of neurons (Trpm8.1, Trpm8.2, Trpm8.3, PEP1.1, PEP1.2, PEP1.3, PEP1.4, PEP2); the second A-LTMRs and proprioceptors (NF1, NF2, NF3, NF4); and the last C-LTMR (TH), nonpeptidergic (NP1.1, NP1.2), Mrgpra3 (NP2.1, NP2.2), and Sst/Nts-containing cluster (NP3) (Figure 1). Among Trpm8 neurons, the three types are relatively similar, while PEP1.1–PEP1.4 separate from PEP2 neurons, consistent with previous results (Usoskin et al., 2015). Among A-LTMRs, NF1 is most dissimilar, while NF2, NF3, and NF4 are more like each other. Among the remaining neuron types, C-LTMRs are most distinct, while NP1.1 and NP1.2 are relatively similar, as are NP2.1 and NP2.2, but not NP3. Gene expression is characterized by the gene expressed uniquely in each of the neuron types, genes that share expression in highly related neurons, as well as genes expressed by related, but more distinct, neuron types in accordance with the hierarchical relationship (Figure 2).

Specialization of Molecular Neuron Types to Specific Functions

The different types of sensory neurons are organized in terms of peripheral innervation, functional properties encoded by the molecular profile of the neurons, and their central terminations in the spinal cord. This organization is stereotypical and reproducible between individuals and likely provides the basis for modality-specific sensation. The response of nerve endings to chemical activation by various spices, inflammatory compounds, and growth factors has been studied. Furthermore, as briefly mentioned, actual transducer molecules that are activated in highly selective and specific ways to certain stimuli, but no other types of stimuli, have been identified and include TrpV1, TrpM8, TrpA1, TrpM2, TrpM3, MrgprD, Mrgpra3, Piezo2, and more. Examination of their expression among neuronal types allows for functional prediction of the neuronal types. Although no comprehensive attempt has been made to functionally interrogate the molecular types of sensory neurons, several studies corroborated the existence of discrete neuron types engaged in particular types of sensation. However, interpreting these results in light of molecular neuronal types can be confounded by the use of Cre mouse driver lines in many studies because gene expression may change during development, while the classified neuronal types are based on young adults.

Overall, a principal structure of cold-sensitive neurons (Trpm8 containing), mechano-heat–sensitive nociceptors (PEP1 and PEP2 types), A-LTMRs, and itch-mechano-heat–sensitive nociceptors is evident (Figure 3), although the molecular profiles predict variability in response characteristics between neuronal types within these categories (Figure 2). A brief description follows of each neuronal type in relation to expression of transducers and markers known to relate to specific functions. It is evident from this account that there is a strong relation between molecular neuron types and functional types (Figure 3).

Dorsal Root Ganglion Neuron Types and Their Functional SpecializationClick to view larger

Figure 3. Taxonomy of somatosensory DRG neurons. Modality, fiber-type association and hierarchical relationship of clades of molecular types of mouse DRG sensory neurons. Proposed cutaneous or deep tissue of innervation is indicated.

Trpm8-containing neuronal types are grouped into three related neuron types (Trpm8.1, Trpm8.2, Trpm8.3); peptidergic into four (PEP1.1–PEP1.4); Aδ nociceptors (PEP2); Aδ LTMR (NF1); Aβ RA-LTMR (NF2); Aβ SA-LTMR (NF3); and proprioceptors (NF4). The remaining five neuron types include C-LTMRs (TH), nonpeptidergic (NP1.1, NP1.2), Mrgpra3 containing (NP2.1, NP2.2) or Sst/Nts containing (NP3). As evident in the discussion that follows, the hierarchical relationship based on molecular properties also reflects functional relationships.

Cold-Sensing Neurons (Trpm8.1, Trpm8.2, Trpm8.3 Neuron Types)

Cold-sensing neurons make up a small subpopulation of DRG neurons, with approximately 20% of cultured neurons responding to cold stimulation in vitro. As in vitro studies are restricted to investigating the effect of cooling on the cell soma, which is a nonphysiological sensing mechanism, it is currently unclear how this population represents the cold-sensing population of neurons in vivo (Hjerling-Leffler, Alqatari, Ernfors, & Koltzenburg, 2007; Reid & Flonta, 2001). In 2002, the putative cold sensor Trpm8 was cloned, paving the way to a vast array of studies investigating its role in innocuous cooling and noxious cold perception (McKemy, Neuhausser, & Julius, 2002). TrpM8 is activated at temperatures below 28°C and is thus a prime candidate for mediating the perception of cooling. Genetic knockout studies showed that Trpm8 deletion substantially reduced the sensitivity of mice to cold temperatures, thus highlighting the indispensable role of TrpM8 in the detection of environmental cold (Bautista et al., 2007; Colburn et al., 2007; Dhaka et al., 2007). Interestingly, Trpm8-/- mice still display aversive behavior toward noxious cold, albeit it to a reduced extent, suggesting that TrpM8 is not the only cold-sensing mechanism present in sensory neurons. Follow-up studies where the entire Trpm8-expressing population was ablated using diphtheria toxin led to a far greater loss in noxious cold sensitivity, with a complete loss of innocuous cold sensitivity in mice (Knowlton et al., 2013; Pogorzala, Mishra, & Hoon, 2013). These studies highlight the significant role of the Trpm8-expressing sensory neuron populations (Trpm8.1, Trpm8.2, Trpm8.3) in cold sensing, while indicating that other molecular mechanisms must participate in the detection of noxious cold.

Over the last 10–15 years, the search for novel putative cold sensors has continued, with numerous molecular candidates, including TrpA1, NaV1.8, and NaV1.9, being ascribed crucial roles in cold sensing in physiological or pathological conditions (Lolignier et al., 2016). Of these candidates, significant effort has been focused on the role of TrpA1 in cold sensing, with numerous studies investigating the effect of Trpa1 deletion on cold sensing in mice. Unfortunately, the outcomes of these studies were mixed, presenting both supporting and opposing results; therefore, there is currently no clear consensus on the role of TrpA1 in acute cold sensation (Bandell et al., 2004; Bautista et al., 2006; Jordt et al., 2004; Karashima et al., 2009; Kwan et al., 2006; Story et al., 2003; Zhou, Suzuki, Uchida, & Tominaga, 2013; Zurborg, Yurgionas, Jira, Caspani, & Heppenstall, 2007). More recently, an in vivo imaging study showed that less than 1% of Trpa1-expressing sensory neurons in the trigeminal ganglia displayed any activity to cold, adding further uncertainty to the role of this channel in normal cold sensing (Yarmolinsky et al., 2016). Instead, Trpm8 was shown to mark all cold-sensitive neurons, which can be functionally divided into three types, potentially providing a functional link to the three molecularly distinct Trpm8 populations (Trpm8.1, Trpm8.2, Trpm8.3) revealed using single-cell RNA sequencing (Yarmolinsky et al., 2016; Zeisel et al., 2018) and based on Nefh expression TrpM8.1 and TrpM8.2 are predicted as C-fibers while TrpM8.3 Ad-fibers.

Other potential mechanisms for cold sensing involve the activity of voltage-gated potassium channels, specifically KV1.1 and KV1.2, which are involved in attenuating neuronal hyperexcitability through the generation of a hyperpolarizing potassium brake current IKD. Interestingly, KV1.1 and KV1.2 are functionally enriched in neurons that do not respond to cold under normal physiological conditions (González et al., 2017). This hypothesizes that these channels are responsible for preventing neurons from firing in response to environmental cold. Indeed, pharmacological inhibition of these channels causes a profound increase in neuronal activity in response to cold, thus supporting their potential role in fine-tuning a neuron’s response to cold stimulation (González et al., 2017). It is important to note that at extreme temperatures (<0°C), the number of nociceptive neurons that respond increases dramatically, with 100% of neurons responding at temperatures down to −18°C (Simone & Kajander, 1997). The global activation of all nociceptors at extreme cold temperatures is likely to reflect cold-induced damage rather than the activation of specific cold-sensitive transducers and therefore highlights the potential difficulty in identifying molecular mechanisms responsible for transducing varying degrees of noxious cold stimulation.

Mechano-Noxious Heat Neurons (PEP1.1, PEP1.2, PEP1.3, PEP1.4, PEP2)

At the molecular level, mechano-heat neurons can be regarded as peptidergic (PEP), as substance P (Tac1) is one of the most commonly used neuropeptides to identify this neuronal population. However, it needs to be stressed that all nociceptors except NP1.1 and NP1.2 populations contain neuropeptides. Noxious heat rapidly activates heat-sensing nociceptors (Cesare & McNaughton, 1996), most of which belong to the unmyelinated C–and lightly myelinated Aδ-fiber types, with few Aβ fibers, and have a threshold of activation around 43°C (Nagy & Rang, 1999; Treede, Meyer, & Campbell, 1998). The TrpV1 ion channel is activated by heat and low pH and plays an important role in inflammation-induced heat hypersensitivity (Caterina et al., 1997, 2000; Davis et al., 2000; Tominaga et al., 1998); however, Trpv1 knockout mice display only minor deficits in acute noxious heat sensing (Caterina et al., 2000; Davis et al., 2000). Trpm3 knockout mice also exhibit reduced heat sensitivity (Vriens et al., 2011), but the combined elimination of both TrpV1 and TrpM3 heat-sensitive channels only produces mild deficits in basal heat responsiveness (Vandewauw et al., 2018). Nevertheless, eliminating Trpv1, Trpm3, and Trpa1 in triple-knockout mice prevents noxious heat sensation at both the cellular and behavioral levels (Vandewauw et al., 2018).

These results show that any one of these channels alone is sufficient to convey acute noxious heat sensation. Trpv1 is abundantly expressed in the peptidergic PEP1.1, PEP1.2, PEP1.3, PEP1.4 neurons and in NP3, indicating these to be critical for inflammatory hyperalgesia. Trpa1 is expressed in NP1.1, NP1.2, and a proportion of PEP1.2 and PEP1.4, while Trpm3 is low and only in some cells of NP1.1, NP1.2, and peptidergic PEP1.1, PEP1.2, PEP1.3, PEP1.4 neurons. Thus, there are various combinations of coexpression, with neurons among the PEP1.2 and PEP1.4 populations expressing all three channels, PEP1.1 and PEP1.2 populations expressing Trpv1 and Trpm3, and finally NP1.1 and NP1.2 populations expressing Trpm3 and Trpa1. Interestingly, the NP3 population of neurons only expresses Trpv1 among the noxious heat-sensitive Trp channels.

Some ion channels, such as Piezo1 and Piezo2, are activated by mechanical stimuli and are critically required for stretch-activated channel activity in mammalian cells. Piezo2, but not Piezo1, is expressed in DRG neurons, where it is critical for touch sensation (Coste et al., 2010; Ranade et al., 2014). Piezo2 is expressed in several of the proposed mechano-noxious heat neurons, but its role for noxious mechanical sensitivity is unclear because the Piezo2 knockout shows deficits only in light-touch sensation (Ranade et al., 2014). In the absence of an identified high-threshold mechanotransducer, uncertainty remains concerning the mechanosensitivity of the neurons classified here as mechano-noxious heat. However, excluding C-LTMRs, a significant proportion of the remaining C-fiber population is polymodal, responding to both mechanical and heat noxious stimuli (Smith & Lewin, 2009). All mechano-heat neurons as well as the NP3 class of neurons express Tac1 (encoding substance P), and noxious mechanical stimuli elicit substance P release in the spinal cord dorsal horn; essentially all of substance P–positive neurons respond to both noxious mechanical and heat stimuli, with the exception of a few that respond to only high-threshold mechanical stimuli but not heat (Duggan, Hendry, Morton, Hutchison, & Zhao, 1988; Lawson, Crepps, Bao, Brighton, & Perl, 1994; Lawson, Crepps, & Perl, 1997). Therefore, it is likely that the PEP1.1–PEP1.4 sensory neuron populations represent mechano-noxious heat neurons.

Unlike the unmyelinated C–nociceptors of the PEP1 class, PEP2-containing neurons are likely to be lightly myelinated Aδ nociceptors. TAC1 is largely absent from nociceptive Aδ units (Lawson et al., 1997). Instead, all Aδ mechano-heat nociceptors expressed Calca, although measured numbers were very low, creating some uncertainty (Lawson, Crepps, & Perl, 2002). PEP2 is the only nociceptive neuron type that contains Nefh at similar levels to those observed in the NF1 Aδ-fiber population. Based on this, we predict PEP2 to be lightly myelinated Aδ nociceptors and therefore to participate in acute pain sensation. TrpV1, but no other noxious heat Trp channels, is present in PEP2. Furthermore, TrpV1 seems not to be expressed in all PEP2 neurons, predicting only some of the PEP2 Aδ nociceptors to be heat responsive.

A-LTMRs and Proprioceptive Neurons (NF1, NF2, NF3, NF4)

The different molecular types of LTMRs and proprioceptors can be predicted based on expression of tyrosine kinase receptors and calcium-binding proteins. NF1 neurons are the Aδ LTMRs that are lightly myelinated and the most sensitive velocity detectors that rapidly adapt to sustained stimulation. Direction-selective hairy skin mechanoreceptors exist in multiple hair types (Horch, Tuckett, & Burgess, 1977) and are conveyed through activation of longitudinal lanceolate endings. Longitudinal lanceolate endings are terminal nerve branches extending along the vertical axis of the hair follicle (Zimmerman, Bai, & Ginty, 2014). Ntrk2 is expressed in these A-fiber LTMRs but not Ret (L. Li et al., 2011; Rutlin et al., 2014), and the NF1 class of LTMRs is the only one expressing high levels of Ntrk2 without Ret. In addition, NF1 neurons express intermediate levels of Nefh, indicating that they are thinly myelinated, consistent with an Aδ-fiber conduction velocity.

The NF2 neurons are RA mechanoreceptors terminating as Meissner corpuscles in the dermal papillae, as longitudinal lanceolate endings in hair follicles, and as Pacinian corpuscles in the periostea of bones and in joints. These neurons respond to movement of the skin, vibration, and the onset/offset of sustained indentation and are critical for the perception of texture and shape. The peripheral nerves of these Aβ RA LTMRs express Calb, Ret, and low levels of Ntrk2, but not Ntrk3 (Wende et al., 2012). The Aβ RA LTMRs have been shown to arise during development from an early Ret-expressing population of sensory precursors, and signaling downstream of Ret is critical for their development (Bourane et al., 2009; Luo et al., 2009). Thus, based on expression of Ret, Calb, and Ntrk2 at lower levels, NF2 neurons can be identified as RA LTMRs. This neuronal cluster therefore seems to serve the function of all types of Aβ RA LTMRs. Hence, in this case, filtering at the end organ shapes the differences in response profile between the different types of RA mechanoreceptors rather than molecular differences between sensory neurons.

The SA mechanoreceptors innervate in mouse Merkel cells. These are located in Merkel touch domes and as dermal Merkel cells, located around hair follicles. These nerves respond to skin movement and static indentation stretch important for the discriminative perception of texture, object shape, and pressure (Zimmerman et al., 2014). Neurotrophin 3 (NT3) is required to maintain innervation of Merkel cells (Airaksinen et al., 1996; McIlwrath et al., 2007). Thus, these neurons express the NT3 receptor Ntrk3. There is only one kind of LTMR neuron that expressed Ntrk3, and these neurons coexpress Ret (Bourane et al., 2009; McIlwrath et al., 2007). In addition, circumferential endings of hair follicles are also Ret and Ntrk3 expressing (Bai et al., 2015). The circumferential LTMR endings are field LTMRs that are sensitive to gentle stroking of the skin but not hair deflection (Bai et al., 2015; Horch et al., 1977). These results reveal that SA LTMR and field LTMR neurons likely belong to the NF3 neuron type.

Proprioceptive neurons are a unique neuronal type represented by the NF4 cluster of cells. Proprioceptive neurons in the DRG innervate the muscle spindles and Golgi tendon organs located in the muscle, acting to feed back information about muscle tension as well as joint position. Such information is essential for planning of movements and also for refining those already in progress. Proprioceptive neurons rely on NT3 for their survival during development and express Ntrk3 and Pvalb (Ernfors et al., 1994). The only neuron type expressing these genes is NF4; hence, this class of neurons represents proprioceptive neurons. Usoskin et al. (2015) identified two types of proprioceptive neurons, NF4 and NF5, and these two types are proposed to be merged into one type, because there were differences only in levels of expression without any specific marker genes uniquely distinguishing one from the other. However, it is possible that sequencing more neurons deeper may lead to the identification of molecular differences between the three known functional types of proprioceptive neurons (1a, II, and 1b neurons innervating the muscle spindles and Golgi tendon organs) that here are represented by one proprioceptive neuron type, NF4.

C–Low-Threshold Mechanoreceptors (TH)

Neurons expressing Th and Slc17a8 in adult are C-LTMRs. The affective aspect of pleasant touch elicited by low mechanical forces such as soft brush stroking is signaled in humans by C-LTMRs (Johansson, Trulsson, Olsson, & Westberg, 1988; Loken, Wessberg, Morrison, McGlone, & Olausson, 2009; A. Vallbo, Olausson, Wessberg, & Norrsell, 1993; A. B. Vallbo et al., 1999). The peripheral nerves of these neurons do not appear in glabrous skin (Bessou, Burgess, Perl, & Taylor, 1971; A. Vallbo et al., 1993). In rodent skin, C-LTMRs terminate as lanceolate endings in hair follicles with electrophysiological properties similar to human (L. Li et al., 2011; Loken et al., 2009; A. Vallbo et al., 1993). C-LTMRs are the only neuronal type expressing Th and SLC17A8 (Vglut3) (L. Li et al., 2011; Seal et al., 2009); hence, this is a unique neuron type dedicated to convey the affective aspect of touch.

Itch-Mechano-Heat Neurons (NP1, NP2, NP3 Neurons)

The NP1.1 and NP1.2 neurons are the only subset containing high levels of MrgprD. MrgprD is expressed in a subset of sensory neurons (Zylka, Dong, Southwell, & Anderson, 2003), and these neurons terminate in the most superficial layer of the epidermis (Zylka, Rice, & Anderson, 2005). In vivo recordings of MrgprD-expressing neurons reveal them to be noxious mechanosensitive neurons, and about half of these also respond to heat (Liu et al., 2012). Mrgprd-expressing neurons are required in ablation experiments for a normal response to light punctate mechanical stimuli evoked by von Frey filaments, but heat sensation remains unchanged (Cavanaugh et al., 2009). β-Alanine, a natural compound, directly binds and activates Mrgprd receptors (Shinohara et al., 2004). β-Alanine induces itch, and this effect is dependent on Mrgprd, while these neurons do not respond to the histamine itch compounds (Shinohara et al., 2004). Some of the NP1.1 and NP1.2 neurons express Trpm3 or Trpa1, and very few express Trpv1, consistent with that some are heat responsive but lack responsiveness to the TrpV1 agonist capsaicin (Dussor, Zylka, Anderson, & McCleskey, 2008). In this study, these neurons were also shown to lack response to the TrpA1 agonist cinnamaldehyde, which is inconsistent with expression of TrpA1 in some of the neurons (Dussor et al., 2008). Thus, NP1 neurons are itch-mechano-heat neurons for which their role for “nonhistamine” itch induced by β-alanine is critical for this neuron type, but they also participate in noxious mechanical sensation (von Frey stimuli). Although some of these neurons are heat responsive, their role for eliciting integrated behavioral heat-induced pain responses is negligible.

NP2.1 and NP2.2 are the only neuronal types expressing Mrgpra3. Chloroquine induces itch and is a ligand for MRGPRA3. Mrgpra3-expressing itch neurons innervate the skin, but not any other tissue examined, explaining why itch is felt in the skin but not in deeper tissues, such as muscles or organs. These neurons are nevertheless polymodal nociceptive neurons with slow conduction velocity (0.5 m/s) and also respond to noxious heat as well as noxious mechanical stimuli (Han et al., 2013; Liu et al., 2009), which is unexpected considering that very few of the cells express Trpv1, and essentially none expresses Trpm3 or Trpa1. However, Mrgpra3 is abundantly expressed in most or all of the NP2 class of neurons. Ablation of MrgprA3-positive neurons specifically affects histamine-dependent and histamine-independent itch but not acute noxious heat, cold, or mechanical pain (Han et al., 2013). Ablation experiments have shown that these neurons play a negligible role in pain behavior yet are critical for itch-related behavior, including both histamine-dependent and -independent itch, indicating that these neurons play a broad role in itch sensation, with only minor contributions to heat and mechanical nociception.

NP3 neurons are similar to NP2 neurons as they seem to be broad itch neurons; in addition, they seem to convey nonnoxious warmth sensation. NP3 neurons express a number of unique markers present only in these neurons, including somatostatin (Sst), neurotensin (Nts), natriuretic peptide type B (NPPB), interleukin 31 receptor A (IL-31ra) and its coreceptor oncostatin M receptor (Osmr), the cysteinyl leukotriene receptor 2 (Cysltr2), 5-hydroxytryptamine receptor 1F (Htr1f), as well as Npy2r. Although not unique to NP3 neurons, they also express TrpV1, predicting that they respond to noxious heat. Deletion of NPPB or Sst, both expressed in the NP3 neurons and which act as neurotransmitters in spinal cord, removes nearly all behavioral responses to a range of itch-inducing agents (Huang et al., 2018; Mishra & Hoon, 2013). Sst also differentiates itch and pain signaling, as deletion of Sst increases sensitivity to noxious heat pain responses (Huang et al., 2018). IL-31ra, Osmr, and Cysltr2 are also implicated in chronic allergic itch (Dillon et al., 2004; Taylor-Clark, Nassenstein, & Undem, 2008; Usoskin et al., 2015). Ablation of Sst-expressing neurons causes a loss at the cellular level of both histamine and nonhistamine agents; however, scratching behavior by interleukin 31 and Htr1f is affected, but not by histamine (Stantcheva et al., 2016), suggesting that NP2 neurons are more critical for histamine itch than NP3 neurons, while NP3 neurons are the only ones available for responses to IL-31- and Htr1f-induced itch. Thus, while NP1 neurons mediate a very specific type of itch (β-alanine induced), NP2 and NP3 neurons have broader itch functions, as well as selective itch functions for particular compounds. Thus, as already mentioned, NP2 neurons are the only neuron type mediating chloroquine itch, while NP3 neurons are the only type mediating interleukin 31 and HTR1F itch. NP3 neurons are also the only DRG neuron type that expresses Trpm2. Trpm2-deficient mice show a marked deficit in sensing nonnoxious warm temperatures (Tan & McNaughton, 2016), and NP3 neurons are therefore predicted to be the only neuron type conveying ambient warm sensitivity through TrpM2, in addition to noxious thermal (Morton, Hutchison, Hendry, & Duggan, 1989), consistent with the presence also of TrpV1 in the NP3 class of neurons. Myelinated afferents expressing the neuropeptide Y receptor NPY2R transmit suprathreshold pinprick-evoked intense mechanical pain (Arcourt et al., 2017). NP3 neurons are the only neuron type expressing Npy2r, and these are C nociceptors consistent with previous studies showing its presence in C nociceptors (Brumovsky et al., 2005), while those studied by Arcourt et al. (2017) were myelinated A-fiber neurons. It is possible that the Npy2r-Cre transgenic mouse line used in the study of Arcourt et al. (2017) does not fully recapitulate adult endogenous expression.

Sensory Neurons Innervating Cutaneous Versus Deep Tissues

Many neurons expressing Tac1, Calca, and Ntrk1 also innervate deep tissues in addition to innervating the skin. Furthermore, some Mrgprd (NP1) and cold-sensitive (TrpM8) neurons also innervate deep tissues (Hockley et al., 2018). Thus, based on this, it is possible that some of the identified neuron types are dedicated to deep tissue innervation. Neurons projecting to deep tissues (muscle, bladder, colon, etc.) are enriched in markers, including Asic3, Prokr2, Chrna7, Serpinb1b, Smr2, and Fam19a1 (Hockley et al., 2018; Yang et al., 2013). These markers are present in PEP1.2 and PEP2. Cdhr1 and Ptgir, which also are enriched in deeply projecting neurons, are in addition to PEP1.2 and PEP2 present in PEP1.4 and NP2.2. Finally, Ntm is present in some neurons projecting to deep tissues (Hockley et al., 2018). Ntm is among nociceptors exclusively expressed in TrpM8.3 neurons. Therefore, based on the hierarchical relationship, many neuron types may exist in two related versions, one that innervates cutaneous and another deep tissues. Thus, assuming that markers mentioned are present more or less exclusively in deeply innervating neurons and absent in cutaneous ones, the following relationship emerges: The related cold-sensitive Trpm8.1 and TrpM8.2 neuron types are cutaneous, while TrpM3 is deep; the related PEP1.1 and PEP1.2 represent one cutaneous (PEP1.1) and one deep (PEP1.2) neuron type; the related PEP1.3 and PEP1.4 represent one cutaneous (PEP1.3) and one deep (PEP1.4) neuron type; the related NP2.1 and NP2.2 represent one cutaneous (NP2.1) and one deep (NP2.2) neuron type. It cannot be excluded that the related neuron types NP1.1 and NP1.2 also represent cutaneous and deep tissue–innervating versions of these neuron types, although evidence for this is missing. It is possible that some neuron types could innervate both cutaneous and deep tissues, such as PEP2 and perhaps NP3.


A general feature of primary somatosensory neurons emerges based on molecular identification of neuronal types. Each neuron class is largely responsible for conveying multiple modalities and intensities, while also having unique and predictable response profiles. The variety of available neurotransmitters, including neuropeptides, may participate in the refinement of stimulus encoding in the second-order neurons of the spinal cord. Furthermore, the response to a single stimulus modality is likely to be dependent on numerous molecular transducers, potentially across several neuronal types, which act in concert to initiate a finely tuned modality-specific response. Therefore, as a general feature, sensation is likely determined by summation of the ensembles of primary afferent types activated. The very recent identification of subgroups of neuronal types (Zeisel et al., 2018) predicts that further functional specializations are yet to be discovered. In addition, primary sensory neurons defined by their physiological responses to heat, cold, and noxious mechanical stimuli are only partly assigned to molecular types, and a systematic assignment of molecular types to function is warranted.

The unbiased classification of primary sensory neurons at the single-cell level allows for the functional prediction of response profiles to various types of stimuli, such as thermal, chemical, and mechanical. This provides the basis from which cell physiology and neurotransmission properties, and the contribution of identified genes involved in these characteristics of the neurons, can be validated in rational experiments. As it identifies unique markers for each neuron type, direct functional manipulation of cell types can also be achieved to establish the morphology of peripheral terminals and the connectivity patterns in the spinal cord, which it is hoped will help to resolve how summation of sensation through the engagement of various ensembles of primary sensory neuron types is decoded.


Abraira, V. E., & Ginty, D. D. (2013). The sensory neurons of touch. Neuron, 79(4), 618–639. doi:10.1016/j.neuron.2013.07.051Find this resource:

Airaksinen, M. S., Koltzenburg, M., Lewin, G. R., Masu, Y., Helbig, C., Wolf, E., … Meyer, M. (1996). Specific subtypes of cutaneous mechanoreceptors require neurotrophin-3 following peripheral target innervation. Neuron, 16(2), 287–295. Retrieved from this resource:

Amir, R., & Devor, M. (2003). Electrical excitability of the soma of sensory neurons is required for spike invasion of the soma, but not for through-conduction. Biophysical Journal, 84(4), 2181–2191. doi:10.1016/s0006-3495(03)75024-3Find this resource:

Arcourt, A., Gorham, L., Dhandapani, R., Prato, V., Taberner, F. J., Wende, H., … Lechner, S. G. (2017). Touch receptor-derived sensory information alleviates acute pain signaling and fine-tunes nociceptive reflex coordination. Neuron, 93(1), 179–193. doi:10.1016/j.neuron.2016.11.027Find this resource:

Averill, S., McMahon, S. B., Clary, D. O., Reichardt, L. F., & Priestley, J. V. (1995). Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Journal of NeuroscienceEuropean Journal of Neuroscience, 7(7), 1484–1494. Retrieved from this resource:

Bai, L., Lehnert, B. P., Liu, J. W., Neubarth, N. L., Dickendesher, T. L., Nwe, P. H., … Ginty, D. D. (2015). Genetic identification of an expansive mechanoreceptor sensitive to skin stroking. Cell, 163(7), 1783–1795. doi:10.1016/j.cell.2015.11.060Find this resource:

Bandell, M., Story, G. M., Hwang, S. W., Viswanath, V., Eid, S. R., Petrus, M. J., … Patapoutian, A. (2004). Noxious cold ion channel TRPA1 is activated by pungent compounds and bradykinin. Neuron, 41(6), 849–857. Retrieved from this resource:

Bautista, D. M., Jordt, S. E., Nikai, T., Tsuruda, P. R., Read, A. J., Poblete, J., … Julius, D. (2006). TRPA1 mediates the inflammatory actions of environmental irritants and proalgesic agents. Cell, 124(6), 1269–1282. doi:10.1016/j.cell.2006.02.023Find this resource:

Bautista, D. M., Siemens, J., Glazer, J. M., Tsuruda, P. R., Basbaum, A. I., Stucky, C. L., … Julius, D. (2007). The menthol receptor TRPM8 is the principal detector of environmental cold. Nature, 448(7150), 204–208. doi:10.1038/nature05910Find this resource:

Bender, E., Buist, A., Jurzak, M., Langlois, X., Baggerman, G., Verhasselt, P., … Luyten, W. (2002). Characterization of an orphan G protein-coupled receptor localized in the dorsal root ganglia reveals adenine as a signaling molecule. Proceedings of the National Academy Sciences of the United States of America, 99(13), 8573–8578. doi:10.1073/pnas.122016499Find this resource:

Bennett, D. L., Michael, G. J., Ramachandran, N., Munson, J. B., Averill, S., Yan, Q., … Priestley, J. V. (1998). A distinct subgroup of small DRG cells express GDNF receptor components and GDNF is protective for these neurons after nerve injury. Journal of Neuroscience, 18(8), 3059–3072. Retrieved from this resource:

Bessou, P., Burgess, P. R., Perl, E. R., & Taylor, C. B. (1971). Dynamic properties of mechanoreceptors with unmyelinated (C) fibers. Journal of Neurophysiology, 34(1), 116–131. doi:10.1152/jn.1971.34.1.116Find this resource:

Bourane, S., Garces, A., Venteo, S., Pattyn, A., Hubert, T., Fichard, A., … Carroll, P. (2009). Low-threshold mechanoreceptor subtypes selectively express MafA and are specified by Ret signaling. Neuron, 64(6), 857–870. doi:10.1016/j.neuron.2009.12.004Find this resource:

Brumovsky, P., Stanic, D., Shuster, S., Herzog, H., Villar, M., & Hokfelt, T. (2005). Neuropeptide Y2 receptor protein is present in peptidergic and nonpeptidergic primary sensory neurons of the mouse. Journal of Comparative Neurology, 489(3), 328–348. doi:10.1002/cne.20639Find this resource:

Brumovsky, P., Villar, M. J., & Hokfelt, T. (2006). Tyrosine hydroxylase is expressed in a subpopulation of small dorsal root ganglion neurons in the adult mouse. Experimental Neurology, 200(1), 153–165. doi:10.1016/j.expneurol.2006.01.023Find this resource:

Carr, P. A., Yamamoto, T., & Nagy, J. I. (1990). Calcitonin gene-related peptide in primary afferent neurons of rat: Co-existence with fluoride-resistant acid phosphatase and depletion by neonatal capsaicin. Neuroscience, 36(3), 751–760. Retrieved from this resource:

Carroll, P., Lewin, G. R., Koltzenburg, M., Toyka, K. V., & Thoenen, H. (1998). A role for BDNF in mechanosensation. Nature Neuroscience, 1(1), 42–46. doi:10.1038/242Find this resource:

Caterina, M. J., Leffler, A., Malmberg, A. B., Martin, W. J., Trafton, J., Petersen-Zeitz, K. R., … Julius, D. (2000). Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science, 288(5464), 306–313. Retrieved from this resource:

Caterina, M. J., Schumacher, M. A., Tominaga, M., Rosen, T. A., Levine, J. D., & Julius, D. (1997). The capsaicin receptor: A heat-activated ion channel in the pain pathway. Nature, 389(6653), 816–824. doi:10.1038/39807Find this resource:

Cavanaugh, D. J., Lee, H., Lo, L., Shields, S. D., Zylka, M. J., Basbaum, A. I., & Anderson, D. J. (2009). Distinct subsets of unmyelinated primary sensory fibers mediate behavioral responses to noxious thermal and mechanical stimuli. Proceedings of the National Academy Sciences of the United States of America, 106(22), 9075–9080. doi:10.1073/pnas.0901507106Find this resource:

Cesare, P., & McNaughton, P. (1996). A novel heat-activated current in nociceptive neurons and its sensitization by bradykinin. Proceedings of the National Academy Sciences of the United States of America, 93(26), 15435–15439. Retrieved from this resource:

Colburn, R. W., Lubin, M. L., Stone, D. J., Jr., Wang, Y., Lawrence, D., D’Andrea, M. R., … Qin, N. (2007). Attenuated cold sensitivity in TRPM8 null mice. Neuron, 54(3), 379–386. Retrieved from this resource:

Cook, A. D., Christensen, A. D., Tewari, D., McMahon, S. B., & Hamilton, J. A. (2018). Immune cytokines and their receptors in inflammatory pain. Trends in Immunology, 39(3), 240–255. doi:10.1016/ this resource:

Coste, B., Mathur, J., Schmidt, M., Earley, T. J., Ranade, S., Petrus, M. J., … Patapoutian, A. (2010). Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels. Science, 330(6000), 55–60. doi:10.1126/science.1193270Find this resource:

Dalsgaard, C. J., Ygge, J., Vincent, S. R., Ohrling, M., Dockray, G. J., & Elde, R. (1984). Peripheral projections and neuropeptide coexistence in a subpopulation of fluoride-resistant acid phosphatase reactive spinal primary sensory neurons. Neuroscience Letters, 51(1), 139–144. Retrieved from this resource:

Davis, J. B., Gray, J., Gunthorpe, M. J., Hatcher, J. P., Davey, P. T., Overend, P., … Sheardown, S. A. (2000). Vanilloid receptor-1 is essential for inflammatory thermal hyperalgesia. Nature, 405(6783), 183–187. doi:10.1038/35012076Find this resource:

Dhaka, A., Murray, A. N., Mathur, J., Earley, T. J., Petrus, M. J., & Patapoutian, A. (2007). TRPM8 is required for cold sensation in mice. Neuron, 54(3), 371–378. doi:10.1016/j.neuron.2007.02.024Find this resource:

Dillon, S. R., Sprecher, C., Hammond, A., Bilsborough, J., Rosenfeld-Franklin, M., Presnell, S. R., … Gross, J. A. (2004). Interleukin 31, a cytokine produced by activated T cells, induces dermatitis in mice. Nature Immunology, 5(7), 752–760. doi:10.1038/ni1084Find this resource:

Djouhri, L., & Lawson, S. N. (2004). Abeta-fiber nociceptive primary afferent neurons: A review of incidence and properties in relation to other afferent A-fiber neurons in mammals. Brain Research, Brain Research Reviews, 46(2), 131–145. doi:10.1016/j.brainresrev.2004.07.015Find this resource:

Dong, X., Han, S., Zylka, M. J., Simon, M. I., & Anderson, D. J. (2001). A diverse family of GPCRs expressed in specific subsets of nociceptive sensory neurons. Cell, 106(5), 619–632. Retrieved from this resource:

Dubin, A. E., & Patapoutian, A. (2010). Nociceptors: The sensors of the pain pathway. Journal of Clinical Investigation, 120(11), 3760–3772. doi:10.1172/JCI42843Find this resource:

Duggan, A. W., Hendry, I. A., Morton, C. R., Hutchison, W. D., & Zhao, Z. Q. (1988). Cutaneous stimuli releasing immunoreactive substance P in the dorsal horn of the cat. Brain Research, 451(1–2), 261–273. Retrieved from this resource:

Dussor, G., Zylka, M. J., Anderson, D. J., & McCleskey, E. W. (2008). Cutaneous sensory neurons expressing the Mrgprd receptor sense extracellular ATP and are putative nociceptors. Journal of Neurophysiology, 99(4), 1581–1589. doi:10.1152/jn.01396.2007Find this resource:

Emery, E. C., Luiz, A. P., Sikandar, S., Magnúsdóttir, R., Dong, X., & Wood, J. N. (2016). In vivo characterization of distinct modality-specific subsets of somatosensory neurons using GCaMP. Science Advances, 2(11), e1600990. doi:10.1126/sciadv.1600990Find this resource:

Ernfors, P., Lee, K. F., Kucera, J., & Jaenisch, R. (1994). Lack of neurotrophin-3 leads to deficiencies in the peripheral nervous system and loss of limb proprioceptive afferents. Cell, 77(4), 503–512. Retrieved from this resource:

González, A., Ugarte, G., Restrepo, C., Herrera, G., Piña, R., Gómez-Sánchez, J. A., … Madrid, R. (2017). Role of the excitability brake potassium current I. Journal of Neuroscience, 37(12), 3109–3126. doi:10.1523/jneurosci.3553–16.2017Find this resource:

Goswami, S. C., Mishra, S. K., Maric, D., Kaszas, K., Gonnella, G. L., Clokie, S. J., … Iadarola, M. J. (2014). Molecular signatures of mouse TRPV1-lineage neurons revealed by RNA-Seq transcriptome analysis. Journal of Pain, 15(12), 1338–1359. doi:10.1016/j.jpain.2014.09.010Find this resource:

Han, L., Ma, C., Liu, Q., Weng, H. J., Cui, Y., Tang, Z., … Dong, X. (2013). A subpopulation of nociceptors specifically linked to itch. Nature Neuroscience, 16(2), 174–182. doi:10.1038/nn.3289Find this resource:

Hjerling-Leffler, J., Alqatari, M., Ernfors, P., & Koltzenburg, M. (2007). Emergence of functional sensory subtypes as defined by transient receptor potential channel expression. Journal of Neuroscience, 27(10), 2435–2443. doi:10.1523/JNEUROSCI.5614–06.2007Find this resource:

Hockley, J. R. F., Taylor, T. S., Callejo, G., Wilbrey, A. L., Gutteridge, A., Bach, K., … Smith, E. S. J. (2018). Single-cell RNAseq reveals seven classes of colonic sensory neuron. Gut. doi:10.1136/gutjnl-2017-315631Find this resource:

Hokfelt, T., Elde, R., Johansson, O., Luft, R., Nilsson, G., & Arimura, A. (1976). Immunohistochemical evidence for separate populations of somatostatin-containing and substance P-containing primary afferent neurons in the rat. Neuroscience, 1(2), 131–136. Retrieved from this resource:

Hokfelt, T., Kellerth, J. O., Nilsson, G., & Pernow, B. (1975). Substance p: Localization in the central nervous system and in some primary sensory neurons. Science, 190(4217), 889–890. Retrieved from this resource:

Horch, K. W., Tuckett, R. P., & Burgess, P. R. (1977). A key to the classification of cutaneous mechanoreceptors. Journal of Investigative Dermatology, 69(1), 75–82. Retrieved from this resource:

Huang, J., Polgar, E., Solinski, H. J., Mishra, S. K., Tseng, P. Y., Iwagaki, N., … Hoon, M. A. (2018). Circuit dissection of the role of somatostatin in itch and pain. Nature Neuroscience, 21, 707–716. doi:10.1038/s41593-018-0119-zFind this resource:

Johansson, R. S., Trulsson, M., Olsson, K. A., & Westberg, K. G. (1988). Mechanoreceptor activity from the human face and oral mucosa. Experimental Brain Research, 72(1), 204–208. Retrieved from this resource:

Jordt, S. E., Bautista, D. M., Chuang, H. H., McKemy, D. D., Zygmunt, P. M., Hogestatt, E. D., … Julius, D. (2004). Mustard oils and cannabinoids excite sensory nerve fibres through the TRP channel ANKTM1. Nature, 427(6971), 260–265. doi:10.1038/nature02282Find this resource:

Julius, D. (2013). TRP channels and pain. Annual Review of Cell and Developmental Biology, 29, 355–384. doi:10.1146/annurev-cellbio-101011-155833Find this resource:

Julius, D., & Basbaum, A. I. (2001). Molecular mechanisms of nociception. Nature, 413(6852), 203–210. doi:10.1038/35093019Find this resource:

Kaplan, D. R., Martin-Zanca, D., & Parada, L. F. (1991). Tyrosine phosphorylation and tyrosine kinase activity of the trk proto-oncogene product induced by NGF. Nature, 350(6314), 158–160. doi:10.1038/350158a0Find this resource:

Karashima, Y., Talavera, K., Everaerts, W., Janssens, A., Kwan, K. Y., Vennekens, R., … Voets, T. (2009). TRPA1 acts as a cold sensor in vitro and in vivo. Proceedings of the National Academy Sciences of the United States of America, 106(4), 1273–1278. doi:10.1073/pnas.0808487106Find this resource:

Kashiba, H., Uchida, Y., & Senba, E. (2001). Difference in binding by isolectin B4 to trkA and c-ret mRNA-expressing neurons in rat sensory ganglia. Brain Research Molecular Brain Research, 95(1–2), 18–26. Retrieved from this resource:

Kim, Y. S., Anderson, M., Park, K., Zheng, Q., Agarwal, A., Gong, C., … Dong, X. (2016). Coupled activation of primary sensory neurons contributes to chronic pain. Neuron, 91(5), 1085–1096. doi:10.1016/j.neuron.2016.07.044Find this resource:

Knowlton, W. M., Palkar, R., Lippoldt, E. K., McCoy, D. D., Baluch, F., Chen, J., & McKemy, D. D. (2013). A sensory-labeled line for cold: TRPM8-expressing sensory neurons define the cellular basis for cold, cold pain, and cooling-mediated analgesia. Journal of Neuroscience, 33(7), 2837–2848. doi:10.1523/jneurosci.1943–12.2013Find this resource:

Kobayashi, K., Fukuoka, T., Obata, K., Yamanaka, H., Dai, Y., Tokunaga, A., & Noguchi, K. (2005). Distinct expression of TRPM8, TRPA1, and TRPV1 mRNAs in rat primary afferent neurons with A delta/C-fibers and colocalization with trk receptors. Journal of Comparative Neurology, 493(4), 596–606. doi:10.1002/cne.20794Find this resource:

Kwan, K. Y., Allchorne, A. J., Vollrath, M. A., Christensen, A. P., Zhang, D. S., Woolf, C. J., & Corey, D. P. (2006). TRPA1 contributes to cold, mechanical, and chemical nociception but is not essential for hair-cell transduction. Neuron, 50(2), 277–289. doi:10.1016/j.neuron.2006.03.042Find this resource:

Lallemend, F., & Ernfors, P. (2012). Molecular interactions underlying the specification of sensory neurons. Trends in Neurosciences, 35(6), 373–381. doi:10.1016/j.tins.2012.03.006Find this resource:

Lawson, S. N., Crepps, B. A., Bao, J., Brighton, B. W., & Perl, E. R. (1994). Substance P-like immunoreactivity (Sp-Li) in dorsal-root ganglia (DRGs) of anesthetized guinea-pigs is related to sensory receptor-type in a-fiber and C-fiber neurons. Journal of Physiology–London, 476P, P39–P39. Retrieved from this resource:

Lawson, S. N., Crepps, B. A., & Perl, E. R. (1997). Relationship of substance P to afferent characteristics of dorsal root ganglion neurones in guinea-pig. Journal of Physiology–London, 505(1), 177–191. doi:10.1111/j.1469–7793.1997.00177.xFind this resource:

Lawson, S. N., Crepps, B., & Perl, E. R. (2002). Calcitonin gene-related peptide immunoreactivity and afferent receptive properties of dorsal root ganglion neurones in guinea-pigs. Journal of Physiology, 540(Pt. 3), 989–1002. Retrieved from this resource:

Lawson, S. N., McCarthy, P. W., & Prabhakar, E. (1996). Electrophysiological properties of neurones with CGRP-like immunoreactivity in rat dorsal root ganglia. Journal of Comparative Neurology, 365(3), 355–366. doi:10.1002/(SICI)1096-9861(19960212)365:3<355::AID-CNE2>3.0.CO;2-3Find this resource:

Lawson, S. N., & Waddell, P. J. (1991). Soma neurofilament immunoreactivity is related to cell size and fibre conduction velocity in rat primary sensory neurons. Journal of Physiology, 435, 41–63. Retrieved from this resource:

Lee, J., Friese, A., Mielich, M., Sigrist, M., & Arber, S. (2012). Scaling proprioceptor gene transcription by retrograde NT3 signaling. PLoS One, 7(9), e45551. doi:10.1371/journal.pone.0045551Find this resource:

Li, C., Wang, S., Chen, Y., & Zhang, X. (2018). Somatosensory neuron typing with high-coverage single-cell RNA sequencing and functional analysis. Neuroscience Bulletin, 34(1), 200–207. doi:10.1007/s12264-017-0147-9Find this resource:

Li, C. L., Li, K. C., Wu, D., Chen, Y., Luo, H., Zhao, J. R., … Zhang, X. (2016). Somatosensory neuron types identified by high-coverage single-cell RNA-sequencing and functional heterogeneity. Cell Research, 26(1), 83–102. doi:10.1038/cr.2015.149Find this resource:

Li, L., Rutlin, M., Abraira, V. E., Cassidy, C., Kus, L., Gong, S., … Ginty, D. D. (2011). The functional organization of cutaneous low-threshold mechanosensory neurons. Cell, 147(7), 1615–1627. doi:10.1016/j.cell.2011.11.027Find this resource:

Liu, Q., Sikand, P., Ma, C., Tang, Z., Han, L., Li, Z., … Dong, X. (2012). Mechanisms of itch evoked by beta-alanine. Journal of Neuroscience, 32(42), 14532–14537. doi:10.1523/JNEUROSCI.3509–12.2012Find this resource:

Liu, Q., Tang, Z., Surdenikova, L., Kim, S., Patel, K. N., Kim, A., … Dong, X. (2009). Sensory neuron-specific GPCR Mrgprs are itch receptors mediating chloroquine-induced pruritus. Cell, 139(7), 1353–1365. doi:10.1016/j.cell.2009.11.034Find this resource:

Loken, L. S., Wessberg, J., Morrison, I., McGlone, F., & Olausson, H. (2009). Coding of pleasant touch by unmyelinated afferents in humans. Nature Neuroscience, 12(5), 547–548. doi:10.1038/nn.2312Find this resource:

Lolignier, S., Gkika, D., Andersson, D., Leipold, E., Vetter, I., Viana, F., … Busserolles, J. (2016). New insight in cold pain: Role of ion channels, modulation, and clinical perspectives. Journal of Neuroscience, 36(45), 11435–11439. doi:10.1523/jneurosci.2327–16.2016Find this resource:

Luo, W., Enomoto, H., Rice, F. L., Milbrandt, J., & Ginty, D. D. (2009). Molecular identification of rapidly adapting mechanoreceptors and their developmental dependence on ret signaling. Neuron, 64(6), 841–856. doi:10.1016/j.neuron.2009.11.003Find this resource:

McCarthy, P. W., & Lawson, S. N. (1990). Cell type and conduction velocity of rat primary sensory neurons with calcitonin gene-related peptide-like immunoreactivity. Neuroscience, 34(3), 623–632. Retrieved from this resource:

McIlwrath, S. L., Lawson, J. J., Anderson, C. E., Albers, K. M., & Koerber, H. R. (2007). Overexpression of neurotrophin-3 enhances the mechanical response properties of slowly adapting type 1 afferents and myelinated nociceptors. European Journal of Neuroscience, 26(7), 1801–1812. doi:10.1111/j.1460–9568.2007.05821.xFind this resource:

McKemy, D. D., Neuhausser, W. M., & Julius, D. (2002). Identification of a cold receptor reveals a general role for TRP channels in thermosensation. Nature, 416(6876), 52–58. doi:10.1038/nature719Find this resource:

McMahon, S. B., & Koltzenburg, N. (2006). Wall and Melzack’s textbook of pain (5th ed.). London, England: Churchill Livingston.Find this resource:

Mishra, S. K., & Hoon, M. A. (2013). The cells and circuitry for itch responses in mice. Science, 340(6135), 968–971. doi:10.1126/science.1233765Find this resource:

Montano, J. A., Perez-Pinera, P., Garcia-Suarez, O., Cobo, J., & Vega, J. A. (2010). Development and neuronal dependence of cutaneous sensory nerve formations: Lessons from neurotrophins. Microscopy Research and Technique, 73(5), 513–529. doi:10.1002/jemt.20790Find this resource:

Morton, C. R., Hutchison, W. D., Hendry, I. A., & Duggan, A. W. (1989). Somatostatin: Evidence for a role in thermal nociception. Brain Research, 488(1–2), 89–96. Retrieved from this resource:

Nagy, I., & Rang, H. (1999). Noxious heat activates all capsaicin-sensitive and also a sub-population of capsaicin-insensitive dorsal root ganglion neurons. Neuroscience, 88(4), 995–997. Retrieved from this resource:

Nagy, J. I., & Hunt, S. P. (1982). Fluoride-resistant acid phosphatase-containing neurones in dorsal root ganglia are separate from those containing substance P or somatostatin. Neuroscience, 7(1), 89–97. Retrieved from this resource:

Olausson, H., Cole, J., Rylander, K., McGlone, F., Lamarre, Y., Wallin, B. G., … Vallbo, A. (2008). Functional role of unmyelinated tactile afferents in human hairy skin: Sympathetic response and perceptual localization. Experimental Brain Research, 184(1), 135–140. doi:10.1007/s00221-007-1175-xFind this resource:

Peier, A. M., Moqrich, A., Hergarden, A. C., Reeve, A. J., Andersson, D. A., Story, G. M., … Patapoutian, A. (2002). A TRP channel that senses cold stimuli and menthol. Cell, 108(5), 705–715. doi:10.1016/S0092-8674(02)00652-9Find this resource:

Pogorzala, L. A., Mishra, S. K., & Hoon, M. A. (2013). The cellular code for mammalian thermosensation. Journal of Neuroscience, 33(13), 5533–5541. doi:10.1523/jneurosci.5788–12.2013Find this resource:

Ranade, S. S., Woo, S. H., Dubin, A. E., Moshourab, R. A., Wetzel, C., Petrus, M., … Patapoutian, A. (2014). Piezo2 is the major transducer of mechanical forces for touch sensation in mice. Nature, 516(7529), 121-U330. doi:10.1038/nature13980Find this resource:

Reid, G., & Flonta, M. L. (2001). Physiology. Cold current in thermoreceptive neurons. Nature, 413(6855), 480. doi:10.1038/35097164Find this resource:

Rutlin, M., Ho, C. Y., Abraira, V. E., Cassidy, C., Bai, L., Woodbury, C. J., & Ginty, D. D. (2014). The cellular and molecular basis of direction selectivity of Adelta-LTMRs. Cell, 159(7), 1640–1651. doi:10.1016/j.cell.2014.11.038Find this resource:

Seal, R. P., Wang, X., Guan, Y., Raja, S. N., Woodbury, C. J., Basbaum, A. I., & Edwards, R. H. (2009). Injury-induced mechanical hypersensitivity requires C-low threshold mechanoreceptors. Nature, 462(7273), 651–655. doi:10.1038/nature08505Find this resource:

Shapiro, E., Biezuner, T., & Linnarsson, S. (2013). Single-cell sequencing-based technologies will revolutionize whole-organism science. Nature Reviews Genetics, 14(9), 618–630. doi:10.1038/nrg3542Find this resource:

Shinohara, T., Harada, M., Ogi, K., Maruyama, M., Fujii, R., Tanaka, H., … Hinuma, S. (2004). Identification of a G protein-coupled receptor specifically responsive to beta-alanine. Journal of Biological Chemistry, 279(22), 23559–23564. doi:10.1074/jbc.M314240200Find this resource:

Silverman, J. D., & Kruger, L. (1988a). Acid phosphatase as a selective marker for a class of small sensory ganglion cells in several mammals: Spinal cord distribution, histochemical properties, and relation to fluoride-resistant acid phosphatase (FRAP) of rodents. Somatosensory Research, 5(3), 219–246. Retrieved from this resource:

Silverman, J. D., & Kruger, L. (1988b). Lectin and neuropeptide labeling of separate populations of dorsal root ganglion neurons and associated “nociceptor” thin axons in rat testis and cornea whole-mount preparations. Somatosensory Research, 5(3), 259–267. Retrieved from this resource:

Simone, D. A., & Kajander, K. C. (1997). Responses of cutaneous A-fiber nociceptors to noxious cold. Journal of Neurophysiology, 77(4), 2049–2060. doi:10.1152/jn.1997.77.4.2049Find this resource:

Smith, E. S. J., & Lewin, G. R. (2009). Nociceptors: A phylogenetic view. Journal of Comparative Physiology A: Neuroethology Sensory Neural and Behavioral Physiology, 195(12), 1089–1106. doi:10.1007/s00359-009-0482-zFind this resource:

Smith-Edwards, K. M., DeBerry, J. J., Saloman, J. L., Davis, B. M., & Woodbury, C. J. (2016). Profound alteration in cutaneous primary afferent activity produced by inflammatory mediators. Elife, 5, e20527. doi:10.7554/eLife.20527Find this resource:

Stantcheva, K. K., Iovino, L., Dhandapani, R., Martinez, C., Castaldi, L., Nocchi, L., … Heppenstall, P. A. (2016). A subpopulation of itch-sensing neurons marked by Ret and somatostatin expression. EMBO Reports, 17(4), 585–600. doi:10.15252/embr.201540983Find this resource:

Story, G. M., Peier, A. M., Reeve, A. J., Eid, S. R., Mosbacher, J., Hricik, T. R., … Patapoutian, A. (2003). ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell, 112(6), 819–829. Retrieved from this resource:

Tan, C. H., & McNaughton, P. A. (2016). The TRPM2 ion channel is required for sensitivity to warmth. Nature, 536(7617), 460–463. doi:10.1038/nature19074Find this resource:

Taylor-Clark, T. E., Nassenstein, C., & Undem, B. J. (2008). Leukotriene D4 increases the excitability of capsaicin-sensitive nasal sensory nerves to electrical and chemical stimuli. British Journal of Pharmacology, 154(6), 1359–1368. doi:10.1038/bjp.2008.196Find this resource:

Tominaga, M., Caterina, M. J., Malmberg, A. B., Rosen, T. A., Gilbert, H., Skinner, K., … Julius, D. (1998). The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron, 21(3), 531–543. doi:10.1016/S0896-6273(00)80564-4Find this resource:

Treede, R. D., Meyer, R. A., & Campbell, J. N. (1998). Myelinated mechanically insensitive afferents from monkey hairy skin: Heat-response properties. Journal of Neurophysiology, 80(3), 1082–1093. doi:10.1152/jn.1998.80.3.1082Find this resource:

Usoskin, D., Furlan, A., Islam, S., Abdo, H., Lonnerberg, P., Lou, D., … Ernfors, P. (2015). Unbiased classification of sensory neuron types by large-scale single-cell RNA sequencing. Nature Neuroscience, 18(1), 145–153. doi:10.1038/nn.3881Find this resource:

Vallbo, A., Olausson, H., Wessberg, J., & Norrsell, U. (1993). A system of unmyelinated afferents for innocuous mechanoreception in the human skin. Brain Research, 628(1–2), 301–304. Retrieved from this resource:

Vallbo, A. B., Olausson, H., & Wessberg, J. (1999). Unmyelinated afferents constitute a second system coding tactile stimuli of the human hairy skin. Journal of Neurophysiology, 81(6), 2753–2763. doi:10.1152/jn.1999.81.6.2753Find this resource:

Vandewauw, I., De Clercq, K., Mulier, M., Held, K., Pinto, S., Van Ranst, N., … Voets, T. (2018). A TRP channel trio mediates acute noxious heat sensing. Nature, 555(7698), 662–666. doi:10.1038/nature26137Find this resource:

Vriens, J., Owsianik, G., Hofmann, T., Philipp, S. E., Stab, J., Chen, X. D., … Voets, T. (2011). TRPM3 is a nociceptor channel involved in the detection of noxious heat. Neuron, 70(3), 482–494. doi:10.1016/j.neuron.2011.02.051Find this resource:

Vriens, J., & Voets, T. (2018). Sensing the heat with TRPM3. Pflügers Archiv: European Journal of Physiology, 470(5), 799–807. doi:10.1007/s00424-017-2100-1Find this resource:

Wang, F., Belanger, E., Cote, S. L., Desrosiers, P., Prescott, S. A., Cote, D. C., & De Koninck, Y. (2018). Sensory afferents use different coding strategies for heat and cold. Cell Reports, 23(7), 2001–2013. doi:10.1016/j.celrep.2018.04.065Find this resource:

Wende, H., Lechner, S. G., Cheret, C., Bourane, S., Kolanczyk, M. E., Pattyn, A., … Birchmeier, C. (2012). The transcription factor c-Maf controls touch receptor development and function. Science, 335(6074), 1373–1376. doi:10.1126/science.1214314Find this resource:

Wiesenfeld-Hallin, Z., Hokfelt, T., Lundberg, J. M., Forssmann, W. G., Reinecke, M., Tschopp, F. A., & Fischer, J. A. (1984). Immunoreactive calcitonin gene-related peptide and substance P coexist in sensory neurons to the spinal cord and interact in spinal behavioral responses of the rat. Neuroscience Letters, 52(1–2), 199–204. Retrieved from this resource:

Yang, F. C., Tan, T., Huang, T., Christianson, J., Samad, O. A., Liu, Y., … Ma, Q. (2013). Genetic control of the segregation of pain-related sensory neurons innervating the cutaneous versus deep tissues. Cell Reports, 5(5), 1353–1364. doi:10.1016/j.celrep.2013.11.005Find this resource:

Yarmolinsky, D. A., Peng, Y., Pogorzala, L. A., Rutlin, M., Hoon, M. A., & Zuker, C. S. (2016). Coding and plasticity in the mammalian thermosensory system. Neuron, 92(5), 1079–1092. doi:10.1016/j.neuron.2016.10.021Find this resource:

Zeisel, A., Hochgerner, H., Lonnerberg, P., Johnsson, A., Memic, F., van der Zwan, J., … Linnarsson, S. (2018). Molecular architecture of the mouse nervous system. Cell, 174(4), 999–1014. doi:10.1016/j.cell.2018.06.021Find this resource:

Zhou, Y., Suzuki, Y., Uchida, K., & Tominaga, M. (2013). Identification of a splice variant of mouse TRPA1 that regulates TRPA1 activity. Nature Communications, 4, 2399. doi:10.1038/ncomms3399Find this resource:

Zimmerman, A., Bai, L., & Ginty, D. D. (2014). The gentle touch receptors of mammalian skin. Science, 346(6212), 950–954. doi:10.1126/science.1254229Find this resource:

Zurborg, S., Yurgionas, B., Jira, J. A., Caspani, O., & Heppenstall, P. A. (2007). Direct activation of the ion channel TRPA1 by Ca2+. Nature Neuroscience, 10(3), 277–279. doi:10.1038/nn1843Find this resource:

Zylka, M. J., Dong, X., Southwell, A. L., & Anderson, D. J. (2003). Atypical expansion in mice of the sensory neuron-specific Mrg G protein-coupled receptor family. Proceedings of the National Academy Sciences of the United States of America, 100(17), 10043–10048. doi:10.1073/pnas.1732949100Find this resource:

Zylka, M. J., Rice, F. L., & Anderson, D. J. (2005). Topographically distinct epidermal nociceptive circuits revealed by axonal tracers targeted to Mrgprd. Neuron, 45(1), 17–25. doi:10.1016/j.neuron.2004.12.015Find this resource: