In the literature cited, there are occasional disagreements about which specific cells show expression in reporter fusion transgenics. We have preserved identifications found in a paper in this table, even for cases in which there was a likely mis-identification. (We have attempted to identify such instances in the footnotes.) In some cases, the situation is less clear, and could reflect real expression differences arising from reporter constructs with different regions of sequence, integration sites, etc. Each Summary List of neurons represents our best judgment based on the evidence available. Where conflicts arise, we have typically deferred to studies that are 1) more recent, 2) comprehensive, and/or 3) using constructs more likely to represent the true expression pattern (i.e., fosmid or insertions into native loci).

Curtis M. Loer^a & James B. Rand^b

^aDept. of Biology, University of San Diego, San Diego, CA; ^bOklahoma Center for Neuroscience, University of Oklahoma Health Sciences Center, Okalahoma City, OK

By Class:
ADF, AIA, AIM, AIN, AIY,  ALN, AS, ASJ, AVA, AVB, AVD, AVE, AVG, AWB, CA(m), CEM(m), DA, DB,  DVA, DVE(m), DX1(m), DX2(m), HOB(m), HSN(h),  I1, I3, IL2,  M1, M2, M4, M5, MC,  PCB(m), PCC(m), PDA, PDB, PDC(m), PGA(m), PLN, PVC, PVN, PVP, PVV(m), PVX(m), PVY(m), PVZ(m), R1A(m), R2A(m), R3A(m), R4A(m), R6A(m), RIB, RIF, RIH, RIR, RIV, RMD, RMF, RMH, SAA, SAB, SDQ, SIA, SIB, SMB, SMD, SPC(m), SPV(m), URA, URB, URX, VA, VB, VC(h)

By Class (with numbers and notes):
ADF(2), AIA(2), AIM(2)⁴, AIN(2), AIY(2), ALN(2), AS1-11, ASJ(2), AVA(2), AVB(2), AVD(2), AVE(2), AVG⁵, AWB(2), CA1-9(m)⁴, CEM(4, m), DA1-9, DB1-7,  DVA⁶, DVE(m), DX1(m)⁷, DX2(m)⁷, HOB(m), HSN(2, h), I1(2)⁵, I3⁵, IL2(6), M1⁵, M2(2)⁵, M4, M5, MC(2)⁵, PCB(2, m), PCC(2, m), PDA, PDB, PDC(m)⁸, PGA(m)⁸, PLN(2), PVC(2), PVN(2)⁵, PVP(2), PVV(m), PVX(m), PVY(m), PVZ(m), R1A(2, m), R2A(2, m), R3A(2, m), R4A(2, m), R6A(2, m), RIB(2)⁸, RIF(2), RIH, RIR, RIV(2), RMD(6), RMF(2), RMH(2), SAA(4)⁵, SAB(3), SDQ(2), SIA(4), SIB(4), SMB(4), SMD(4), SPC(2, m), SPV(2, m), URA(4), URB(2), URX(2), VA1-12, VB1-11, VC1-6(h) 

Pharynx: I1(2)⁵, I3⁵, M1⁵, M2(2)⁵, M4, M5, MC(2)⁵

Ventral nerve cord & body: AS2-10, CAn(1-9)(m)⁴, DA2-7, DB3-7, HSN(2, h), SDQ(2), VA2-11, VB3-11, VC1-6(h)   

Tail: ALN(2), AS11, DA8-9, DVA⁶, DX1(m)⁷, DX2(m)⁷, HOB(m), PCB(2, m), PCC(2, m), PDA, PDB, PDC(m)⁸, PGA(m)⁸, PLN(2), PVC(2), PVN(2)⁵, PVP(2), PVV(m), PVX(m), PVY(m), PVZ(m), R1A(2, m), R2A(2, m), R3A(2, m), R4A(2, m), R6A(2, m), SPC(2, m), SPV(2, m), VA12

Note: Somas found in the retrovesicular ganglion (RVG) are listed in the head; those in the preanal ganglion (PAG) are listed in the tail.

[Note: Monoamine catabolism pathways are only partially characterized in C. elegans. N-acetylation and N-succinylation are likely significant destinations for monoamines including dopamine (Artyukhin et al., 2013). Oxidation via AMX-2/MAO is likely, similar to the well-characterized pathways in other animals (best known from vertebrates, especially mammals). There are no clear orthologs of catechol-o-methyl transferase (COMT); however, worm comt genes encode proteins with a methyltransferase associated domain also found in mammalian COMT.²⁴ Several different monoamine inactivating modifications are found among invertebrate phyla, including N-acetylation, γ-glutamylation, β-alanylation, sulfation, etc. (see review by Sloley, 2004), although it is not always clear whether these are associated with neurons, or that the modifications are catabolic in nature.]

Aldehyde Dehydrogenase (ALDH)

ALDH

Succinic Semialdehyde Dehydrogenase (SSADH)

ALDH

SUMMARY - Dopaminergic Neurons (adult): Hermaphrodite (n = 8), Male (n = 14+2)¹³

By Class:
ADE, CEP, PDE, R5A(m), R7A(m), R9A(m), SPSo(m)

By Class (with numbers and notes):
ADE(2), CEP(4), PDE(2), R5A(2, m), R7A(2, m), R9A(2, m), SPSo(2, m)¹³

[Note: Monoamine catabolism pathways are only partially characterized in C. elegans. Succinylation appears to be a significant pathway for TA and OA (Artyukhin et al., 2013). TA and OA may also be catabolized via MAO, likely AMX-2. See Dopamine catabolism for further notes on possible monoamine metabolism, and lists of genes that may be involved.]

By Class (with numbers and notes):
HOA, possibly R8A, R8B, RIM, uv1

By Body Region:
Head: RIM(2)²⁹
Body: uv1(4, h) (non-neuronal)
Tail (male only): HOA, possibly R8A(2), R8B(2)³⁰

By Class: RIC

By Body Region:
Head: RIC(2)
Body: gonadal sheath (non-neuronal)

By Class:
ADA, ADL, AFD, AIB, AIM, AIZ,  ALM, ASE, ASG, ASH, ASK, AQR, AUA, AVM, AWC, BAG, CP0, CP5-7, DVC, FLP,  HOA(m), I2, I5, IL1, LUA, M3, MI, OLL, OLQ,  PCA(m), PHA, PHB, PHC, PLM, PVD, PVQ, PQR, PVR, PVV(m), R2B(m), R5A(m), R6B(m), R9A(m), RIA, RIG, RIM, URY

By Class (with numbers and notes):
ADA(2), ADL(2), AFD(2), AIB(2), AIM(2)⁵, AIZ(2),  ALM(2), ASE(2), ASG(2)³⁸, ASH(2), ASK(2), AQR, AUA(2), AVM, AWC(2), BAG(2), CP0, CP5-7, DVC, FLP(2),  HOA(m), I2(2), I5⁴³, IL1(6), LUA(2), M3(2), MI, OLL(2), OLQ(4),  PCA(2, m) PHA(2), PHB(2)43, PHC(2), PLM(2), PVD(2), PVQ(2), PQR(2), PVR, PVV(m), R2B(2, m), R5A(2, m), R6B(2, m), R9A(2, m), RIA(2), RIG(2), RIM(2), URY(4)

By Body Region:
Head: ADA(2), ADL(2), AFD(2), AIB(2), AIM(2)⁵, AIZ(2), ASE(2), ASG(2)³⁸, ASH(2), ASK(2), AQR, AUA(2), AWC(2), BAG(2), FLP(2), IL1(6), OLL(2), OLQ(4), RIA(2), RIG(2), RIM(2), URY(4)     Pharynx: M3(2), MI, I2(2), I5⁴³    Ventral nerve cord & body: ALM(2), AVM, CP0, CP5 - CP7 (m)    Tail: DVC, HOA(m), LUA(2), PCA(2, m), PHA(2), PHB(2)⁴³, PHC(2), PLM(2), PVD(2), PVQ(2), PQR(2), PVR, PVV(m), R2B (2, m), R5A(2, m), R6B (2, m), R9A (2, m)

Sex-shared neurons: ASI, AVF, AVH, AVJ, AVK, AWA, BDU, CAN²⁰, PVM, PVT, PVW [only in hermaphrodites], RID, RIP, RMG, I4, and I6
Male-specific neurons: MCM, DVE, DVF, DX3/4, R4B, R5B, R7B, and SPD

¹ The genes unc-17 and cha-1 are coexpressed in an operon, controlled by a single promoter upstream of unc-17 (Alfonso et al., 1994a).
² Because of the genomic structure of the unc-17-cha-1 locus, it is likely that the genes are co-expressed in the same neurons. However, expression of cha-1 was confirmed only for the unc-17 -expressing neurons identified in the Duerr et al., 2008 study.
³ Most reporter transgenics to date have been made with a few kbp of upstream genomic sequence fused to the reporter (transcriptional, or translational near the N-terminus), and therefore may lack important regulatory elements (distant upstream, within coding, and downstream). Larger fosmid-based reporters typically have 35-40 kbp of genomic sequence, both upstream and downstream, with the reporter fused to the full length coding sequence at the C-terminus, or separated from the gene of interest by an SL2-splice site creating an artificial operon (e.g., see Dolphin & Hope, 2006; Sarov et al., 2006; Tursun et al., 2009). Therefore, transgenics with these fosmid (or BAC) constructs may be more likely to reproduce the expression pattern of the endogenous locus. Issues with both small fusion reporter constructs and fosmid reporters that are extrachromosomal or inserted into ectopic genomic locations may also progressively be resolved with CRISPR-generated knock-in reporters of various types inserted as single copies into the endogenous locus.
⁴ AIM neurons are dual-transmitter glutamatergic/serotonergic in hermaphrodites (Serrano-Saiz et al., 2013); in males, AIM neurons are glutamatergic/serotonergic until L3, then switch to a cholinergic/serotonergic dual-transmitter phenotype (Pereira et al., 2015).
⁵ These cells express unc-17 but not cho-1 (Pereira et al., 2015).
⁶ DVA was originally (mis)identified as DVC by Duerr et al., 2008.
⁷ Pereira et al., 2015 originally identified 2 male-specific tail neurons as DVE & DVF; Serrano-Saiz et al., 2017 revised this identification to be DX1/2.
⁸ Expression of cho-1 is strong in RIB, but unc-17 expression is very weak (Pereira et al., 2015).
⁹ VC4-5 express unc-17 but not cho-1 (Pereira et al., 2015); VC4-5 also release serotonin (Duerr et al., 1999).
¹⁰ Uptake was Na⁺-dependent, and blocked by 1 μm HC3 (criterion for high-affinity uptake) (Okuda et al., 2000).
¹¹ Staining with acetylthiocholine (Method: Karnovsky & Roots, 1964).
¹²Although the genes ace-4 and ace-3 are coexpressed in a bi-cistronic operon (controlled by a single promoter upstream of ace-4), the level of ace-4 mRNA is exceedingly low, and there is no detectable ACE-4 protein or enzymatic activity present in worms (Combes et al., 2000). Also, there is no known ace-4-orthologous gene in nematodes outside of genus Caenorhabditis; only an ace-3 gene is present.
¹³ Spicule socket cells SPSo ('non-neuronal' support cells) comprise 2 syncitial cells with 4 nuclei each that likely release DA to promote sperm release, acting much like neurons (LeBouef et al., 2014). Interestingly, SPSo cells do not appear to express DAT-1, and SPSo-mediated behavior is not rescued by exogenous DA (LeBouef et al., 2014).
¹⁴ AADC is also known as 5HTP Decarboxylase or Dopa Decarboxylase (DDC); in animals, the enzyme generally has broad substrate specificity, catalyzing both serotonin and dopamine synthesis (reviewed by Zhu & Jorio, 1995).
¹⁵ BH4 is also required for the function of other aromatic amino acid hydroxylases such as phenylalanine hydroxylase (pah-1), and the lipid metabolic enzyme alkylglycerol monooxygenase (agmo-1); therefore, expression of BH4 synthesis and regeneration genes is not neuron-specific. They are highly expressed in the hypodermis, and also in the intestine (Loer et al., 2015). BH4 is also required for nitric oxide synthase (NOS) function; however, NOS has not been identified in C. elegans (Gusarov et al. 2013).
¹⁶ Reporters for these BH4 synthesis and regeneration genes to date have shown limited expression in identified dopaminergic and serotonergic neurons (Zhang et al., 2014; Loer et al., 2015), despite mutant phenotypes indicating function in those cells promoting serotonin and dopamine synthesis (Loer et al., 2015).
¹⁷ QDPR is also known as Dihydropteridine reductase (DHPR).
¹⁸ Vesicular monoamine transporter (VMAT) is used in common by all monoaminergic neurons to transport the neurotransmitter into synaptic vesicles for release (Eiden & Weihe, 2011). In C. elegans, this includes neurons using dopamine, tyramine, octopamine, serotonin and perhaps one or more as yet unidentified monoamines ²⁰.
¹⁹ VMAT colocalizes with synaptic vesicles (Duerr et al., 1999).
²⁰Expression of a cat-1/VMAT fosmid (but no other monoaminergic markers) may indicate CAN uses an unidentified monoamine (Serrano-Saiz et al., 2017).
²¹ VMAT's relative affinity for monoamine substrates: dopamine ~ tyramine > serotonin > norepinephrine ~ octopamine > histamine (Duerr et al., 1999).
²² DAT expressed in human cells mediates Na⁺ and Cl^--dependent uptake of DA better than norepinephrine or other transmitters. Transport by DAT was blocked by tricyclics (especially imipramine) and other monoamine transport inhibitors (Jayanthi et al., 1998).
²³ DAT mediates Na⁺ and Cl^--dependent uptake of DA in both heterologous (tsA-201) and cultured native C. elegans cells, and shows electrogenic activity (Carvelli et al., 2004).
²⁴ Among the named C. elegans comt genes, four encode proteins orthologous to mammalian COMT-domain containing protein 1 (COMT-D1), and are most similar to bacterial and plant known and putative O-methyltransferases, including Caffeic Acid O-Methyltransferase (also abbreviated COMT) and Caffeoyl CoA O-Methyltransferase (CCoAOMT, Ferrer et al., 2005). It is plausible that one or more of the worm COMT proteins could act on a catechol-containing substrate based on similarity to CCoAMTs - i.e., the substrate caffeoyl CoA has a catechol structure that is methylated on the 3-hydroxyl by CCoAMT. The worm COMT proteins, however, have very limited or no significant sequence homology to mammalian COMTs except among S-adenosylmethionine binding residues (SAM aka AdoMet, the methyl donor) found in all AdoMet-dependent methyltransferases (e.g., see Martin & McMillan, 2002). Currently, there is no evidence suggesting comt gene function in neurotransmitter metabolism; no informative mutant phenotypes (by RNAi) and no expression patterns have been reported.
²⁵ Among the amx gene encoded proteins, the AMX-2 predicted protein is most similar to a mammalian monoamine oxidase (MAO-A), and has been shown to metabolize serotonin (Schmid et al., 2015). AMX-1 may be a histone demethylase (orthologous to lysine-specific histone demethylases). The protein encoded by amx-3 (not listed) is most similar to polyamine and spermine oxidases; no expression pattern has been reported to date.
²⁶ There are 13 identified alh genes that encode proteins orthologous or highly similar to mammalian aldehyde dehydrogenases (ALDHs). Those shown here have relevant reporter expression patterns. To date, there are no expression patterns reported for alh-2, alh-3, alh-4, alh-5, alh-11, and alh-12; alh-8 is expressed in muscle mitochondria (Meissner et al., 2011); alh-9 is expressed only in hypodermis, during embryogenesis (Mounsey et al., 2002); alh-13 is expressed in the adult intestine (McKay et al., 2003). See also GABA catabolism regarding alh-7 / SSADH. Aldehyde dehydrogenases convert a wide array of both endogenous and exogenous aldehydes to carboxylic acids in amino acids, biogenic amines, carbohydrates, lipids, etc. In humans, the same ALDH (ALDH1A1) that converts DOPAL to DOPAC in DA neurons elsewhere converts retinal to retinoic acid, and oxidizes the ethanol metabolite acetaldehyde (Marchitti et al., 2008). ALDH1A1 is also capable of mediating GAD-independent GABA synthesis in mammalian brain via an alternative pathway from putrescine (Seiler & Al-Therib, 1974; Kim et al., 2015).
²⁷ Found in Hope Expression Pattern DB (Reference: Hope et al., 1996).
²⁸ Method of  P. D. Evans, 1978.
²⁹ The uterine uv1 cells appear to be neuroendocrine. They express neurosecretory proteins and neuropeptides, contain neuroscretory vesicles, and likely release tyramine to inhibit egg laying (Alkema et al., 2005 and references therein).
³⁰ Neither R8A or R8B express a cat-1/VMAT fosmid, making it unclear whether they actually use tyramine as a neurotransmitter (Serrano-Saiz et al., 2017).
³¹ FIF for serotonin in C. elegans is weak and extremely labile (Horvitz et al., 1982), which likely explains why it has been seen only in the NSM neurons.
³² GAIF method of de la Torre, 1980; comparable to FIF, but apparently more robust in C. elegans.
³³ 5HT-IR in VC4, 5 can be weak and unreliable (Duerr et al., 1999), and/or varies with antiserum and staining protocol. Serrano-Saiz et al., 2017 did not observe 5HT-IR or cat-1/VMAT fosmid reporter expression in VC4, 5.
³⁴ 5HT-IR in CA neurons (CA1-4) is rare (Loer & Kenyon, 1993), but is not reported by others (e.g., Serrano-Saiz et al., 2017). CA neurons also do not express other serotonergic marker genes. Strong 5HT-IR in likely CA homologs is observed in other nematodes, including other Caenorhabditis species (Loer & Rivard, 2007).
³⁵ Serrano-Saiz et al., 2017 suggest that AIM, CEM, I5 and URX are '5HT clearance' neurons - they saw no expression of tph-1 or cat-1/VMAT in the cells. Unlike the other cells, which are weakly 5HT immunoreactive, AIM stains strongly and reliably with serotonin antibodies.
³⁶ The male-specific, unpaired 'right preanal ganglion' (RPAG) neuron was first identified as either PDC or PGA (Loer & Kenyon, 1993); it is now known to be PGA (Serrano-Saiz et al., 2017).
³⁷ 5HT-IR and cat-1/VMAT fosmid expression in PVW is sexually dimorphic, found only in the adult male (Serrano-Saiz et al., 2017).
³⁸ ASG is glutamatergic, but 5HT-IR and tph-1 reporter expression appear after exposure to hypoxic conditions (Pocock & Hobert, 2010).
³⁹ Expression of tph-1 and bas-1 in AIMs and RIH is rare with most reporters; these cells may be '5HT-absorbing' via MOD-5/SERT (and serotonin-releasing) rather than 5HT-synthesizing neurons (Kullyev et al., 2010; Jafari et al., 2011). No expression was observed in the RIH and AIM (and VC4, VC5) neurons with tph-1 fosmid reporter (Serrano-Saiz et al., 2017).
⁴⁰ Reported CAT-1/VMAT antibody staining in AIM (Duerr et al., 1999) may actually be in the tyraminergic RIM, based on a cat-1 fosmid reporter that otherwise fully replicates previous antibody staining (Serrano-Saiz et al., 2017).
⁴¹ MOD-5 expressed in mammalian cells mediates Na⁺-dependent uptake of 5HT but not DA, histamine, norepinephrine, GABA, Glu, or Gly. Uptake is blocked by SSRIs, tricyclics, and non-specific monoamine transport inhibitors (Ranganathan et al., 2001).
⁴² AA-NAT activity may be catabolic for serotonin and/or synthetic for melatonin. Although there are no clear orthologs of hydroxyindole-o-methyltransferase (HIOMT) in C. elegans, synthesis of melatonin, and a reporter expression pattern for a very weak homolog (Y74C9A.3/homt-1) has been reported (Tanaka et al., 2007; Migliori et al., 2012).
⁴³ I5 and PHB are glutamatergic (Serrano-Saiz et al., 2013), and may also be serotonergic (Sawin et al., 2000).
⁴⁴ Cells / tissues likely mediate GABA clearance and/or recycling; they do not express unc-47 (VGAT) (Gendrel et al., 2016).
⁴⁵ Anti-GABA staining in AVA, AVB and AVJ is weak; the cells express no other GABA-related marker genes (Gendrel et al., 2016).
⁴⁶ UNC-46 is a BAD-LAMP-related protein required for trafficking UNC-47/VGAT to synaptic vesicles at synapses, specifically in GABAergic neurons in worms. Rescuing UNC-46-GFP fusion proteins co-localize with synaptic varicosities (Schuske et al., 2007).
⁴⁷ Expression is weak and inconsistent.
⁴⁸ There are discrepancies between the snf-11 expression patterns reported by Jiang et al. (2005) and those reported by Mullen et al. (2006); the Jiang et al observations are likely due to an error in promoter construction (see Mullen et al., 2006).
⁴⁹ Male-specific expression of snf-11 in VD12 reported in Gendrel et al., 2016 is instead expression in the male-specific CP9 (Serrano-Saiz et al., 2017).
⁵⁰ SNF-11 expressed in Xenopus oocytes mediates Na⁺- and Cl ^--dependent high affinity uptake of GABA; uptake is blocked by GAT inhibitors nipecotic acid
and SKF89976A (Mullen et al., 2006).
⁵¹ Ubiquitous expression of gta-1 (single GABA-T orthologous gene) is consistent with a role beyond GABA metabolism (Gendrel et al., 2016).
⁵² Lee et al., 1999 reported eat-4 expression in NSM, AVJ, and IL2, and Ohnishi et al., 2011 reported eat-4 expression in the AVA, AVE, SIB, RMD and ASJ neurons; however, subsequent analysis with the same transgenic reporter constructs did not replicate those cell IDs (Serrano-Saiz et al., 2013).
⁵³ Reporter expression localizes to vesicular structures in the soma, indicating VGLU-2 may not be involved in synaptic vesicle function; however, vglu-2 mutants have AIN-associated olfactory behavior defects indicating function. In C. elegans alone (not other sequenced Caenorhabditis species), there is a degenerate duplicate gene, vglu-3, for which reporters show no expression (Serrano-Saiz et al., 2019).
⁵⁴ There is no glt-2 (the sequence originally named glt-2 was found to be a splice variant of glt-1). Although all glt genes are listed here, it seems unlikely that all are involved in neuronal glutamate use.
⁵⁵ glt-4 may be a primarily presynaptic neuronal GluT.