CHAPTER TWENTY-FIVE

Terminal Selectors of Neuronal

Identity
Oliver Hobert1
Department of Biological Sciences, Howard Hughes Medical Institute, Columbia University, New York, USA
1
Corresponding author: e-mail address: or38@columbia.edu

Contents
1. Historical Background
2. Coregulation of Terminal Effector Genes by Terminal Selectors
3. Terminal Selectors Initiate and Maintain the Terminally Differentiated State
4. Terminal Selectors Act in Reiteratively Used Combinations
5. Terminal Selectors Control Regulatory Subroutines
6. Parallel Regulatory Routines
7. Diversification of Neuronal Identity via Selective Repressibility
8. Generality of the Terminal Selector Concept
9. Regulation of Terminal Selector Genes
10. Conclusions
Acknowledgments
References

455
459
463
464
466
467
469
469
471
471
471
472

Abstract
The analysis of the developmental programs that define many different neuron types in
the nematode Caenorhabditis elegans has revealed common themes in how distinct
terminal differentiation programs are controlled. Rather than being controlled in a
piece-meal manner, terminal identity features of a mature neuron are often coregulated
by so-called terminal selector transcription factors. Here, I summarize the terminal selector concept and emphasize core features of this concept in the C. elegans system such as
coregulation of terminal effector batteries, combinatorial control mechanisms, and the
coupling of initiation and maintenance of neuronal identity.

1. HISTORICAL BACKGROUND
Nervous system development has traditionally been studied with a
top-down focus, defining inductive events early in embryonic patterning
and elucidating signaling events and transcriptional regulatory cascades that
Current Topics in Developmental Biology, Volume 116
ISSN 0070-2153
http://dx.doi.org/10.1016/bs.ctdb.2015.12.007

2016 Elsevier Inc.

All rights reserved.

455

456

Oliver Hobert

progressively restrict developmental fates as embryogenesis proceeds

(Edlund & Jessell, 1999; Hemmati-Brivanlou & Melton, 1997). This
top-down approach is well illustrated by Waddingtons landscape model
in which a ball travels down specific valleys, making several choices along
the way (Waddington, 1957). An alternative perspective, the bottomup angle, views the problem of neuronal development from the standpoint
of the end product of neuronal development, asking how the terminal features of specific neuron types in a mature nervous system are genetically
programmed. The bottom-up perspective has been extensively taken in
the Caenorhabditis elegans field, not by design but as the result of the initial
efforts of Sydney Brenner to identify C. elegans mutants that affect specific
animal behaviors. Brenners first behavioral mutants displayed various types
of uncoordinated (unc) locomotory behavior (Brenner, 1974). The ensuing
establishment of various assays that measure the response of animals to specific sensory cues then lead to the identification of viable mutants that selectively affected very specific behaviors such as chemotaxis behavior (Lewis &
Hodgkin, 1977), thermotaxis behavior (Hedgecock & Russell, 1975),
mechanosensory behavior (Chalfie & Sulston, 1981), or odortaxis behavior
(Sengupta, Colbert, & Bargmann, 1994). Strikingly, the molecular characterization of these behavioral mutants showed that many of them affect transcription factors that turned out to act at the terminal step of differentiation
of postmitotic neurons (listed in Table 1; Baran et al., 1999; Finney,
Ruvkun, & Horvitz, 1988; Hobert et al., 1997; Jin et al., 1994; Prasad
et al., 1998; Satterlee et al., 2001; Sengupta et al., 1994; Uchida et al.,
2003; Way & Chalfie, 1988).
The phenotypic analysis of these mutants profited from some of the key
defining features of the C. elegans system: (a) an exceptionally well-described
cell lineage history (Sulston, 1983), (b) a most detailed anatomical description of all individual neuron types (White, Southgate, Thomson, & Brenner,
1986), (c) the knowledge of functional features of many individual neuron
types based on laser ablation (e.g., Bargmann & Horvitz, 1991; Mori &
Ohshima, 1995), and (d) the availability of a vast collection of molecular
markers in the form of gfp reporter transgenes (Chalfie, Tu, Euskirchen,
Ward, & Prasher, 1994). These molecular markers, built by the community
over the past 20 years (curated at www.wormbase.org), monitor the
neuron-type specific expression of hundreds of genes that define terminal
properties of a neuron type (e.g., genes encoding neurotransmitter synthesis,
receptors, ion channel, etc.). Together, these anatomical, functional, and
molecular features assign precise and unique phenotypic spaces to

457

Terminal Selectors of Neuronal Identity

Table 1 Classic Behavioral Mutants that Originally Defined Terminal Selectors

Transcription
Behavioral Phenotype
Gene Factor Type
Leading to Identification Neuron Class Affected

unc-3

EBF/COE

unc-30 Pitx-type
homeodomain

Uncoordinated
locomotiona

Cholinergic ventral cord

motorneuronsb

Uncoordinated
locomotiona

GABAergic ventral cord

motorneuronsc
Cholinergic PVP
interneuronsd

unc-42 Homeodomain

Uncoordinated
locomotiona

Glutamatergic ASH sensory

neuronse
Glutamatergic AIB
interneuronsf

unc-86 Brn3-type POU Mechanosensory defectsg Mechanosensory ALM/

homeodomain
AVM/PLM/PVM neuronsh
Serotonergic HSN motor
neuronsi
Serotonergic NSM neuronsj
Inner labial IL2 sensorymotor neuronsj
URA motor neuronsj
URB interneuronsj
Glutamatergic URX
sensory neuronsk
Glutamatergic AQR/PQR
sensory neuronsk
Glutamatergic PVR
interneuronk
Mechanosensory defectsg Mechanosensory ALM/
AVM/PLM/PVM neuronsl

mec-3

Lhx1/5 LIM
homeodomain

che-1

C2H2 Zn finger Chemotaxis defectsm

ttx-1

Otx
homeodomain

Thermotaxis defectso

Gustatory ASE sensory

neuronsn
Thermosensory AFD
neuronp
Continued

458

Oliver Hobert

Table 1 Classic Behavioral Mutants that Originally Defined Terminal Selectorscont'd

Transcription
Behavioral Phenotype
Gene Factor Type
Leading to Identification Neuron Class Affected

ttx-3

Lhx2/7 LIM
homeodomain

Thermotaxis defectso

Cholinergic AIY
interneuronq
Cholinergic AIA
interneuronj
Serotonergic NSM neuronj
Glutamatergic ASK sensory
neuronk

odr-7

C4 Zn finger

Odortaxis defectsr

Olfactory AWA neurons

Brenner (1974).
Kratsios, Stolfi, Levine, and Hobert (2011), Kratsios et al. (2015), and Prasad et al. (1998). unc-3 locomotory defects match the effect expected from ablating of A- and B-type motor neurons (Chalfie et al.,
1985).
c
Jin, Hoskins, and Horvitz (1994). unc-30 locomotory defects match the defects observed upon losing
signaling by GABAergic D-type motor neurons (McIntire, Jorgensen, & Horvitz, 1993).
d
Pereira et al. (2015).
e
Baran, Aronoff, and Garriga (1999) and Serrano-Saiz et al. (2013). unc-42 mutants phenocopy the sensory defects observed in ASH-ablated animals (Baran et al., 1999).
f
E. Serrano-Saiz and O. Hobert (unpublished data).
g
Chalfie and Sulston (1981).
h
Duggan, Ma, and Chalfie (1998). unc-86 mutants phenocopy ablation of mechanosensory neurons
(Chalfie et al., 1985).
i
Desai, Garriga, McIntire, and Horvitz (1988). unc-86 mutants phenocopy loss of HSN neurons.
j
Zhang et al. (2014).
k
Serrano-Saiz et al. (2013).
l
Duggan et al. (1998), Way and Chalfie (1988), and Zhang et al. (2002). mec-3 mutants phenocopy ablation of mechanosensory neurons (Chalfie et al., 1985).
m
Lewis and Hodgkin (1977).
n
Etchberger et al. (2007) and Uchida, Nakano, Koga, and Ohshima (2003). che-1 mutants phenocopy the
ablation of the ASE neurons (Bargmann & Horvitz, 1991).
o
Hedgecock and Russell (1975).
p
Satterlee et al. (2001). ttx-1 mutants phenocopy ablation of the AFD neurons (Mori & Ohshima, 1995).
q
Altun-Gultekin et al. (2001). ttx-3 mutants phenocopy ablation of the AFD neurons (Mori & Ohshima,
1995; Tsalik & Hobert, 2003).
r
Sengupta et al. (1994).
s
odr-7 mutants phenocopy ablation of the AWA neurons (Bargmann, Hartwieg, & Horvitz, 1993).
Many additional terminal selectors were identified through forward and reverse genetic approaches and
are not listed here. See an upcoming review (O. Hobert, WIREs Dev. Biol., in preparation) for a more
comprehensive summary. As illustrated in Fig. 2, one terminal selector can control distinct neuronal differentiation programs through its pairing with distinct cofactors.
b

individual neuron types which then permitted a detailed assessment of how

the regulatory factors defined by the classic behavioral mutants (listed in
Table 1) affect neuronal differentiation. Strikingly, all of the transcription
factors listed in Table 1 were found to have very broad effects on the adoption of the correct terminal identity of very specific neuron types. They

Terminal Selectors of Neuronal Identity

459

affect the functional properties of the respective neuron types in the sense
that the behavioral phenotypes of these mutants match the phenotypic consequences of laser ablating these neuron types and they broadly affect
manyif not allknown molecular markers normally expressed specifically
in the respective neuron; however, they do not affect the generation of these
neurons. Given the neuron identity-defining properties, I proposed to call
these transcription factors terminal selectors (Hobert, 2008, 2011). This
terminology is an extension of the selector gene concept, originally proposed by Garcia-Bellido (1975). Classic selector genes are genes that
define the identity of specific domains of a developing organism and they
act transiently during specific phases of development. In contrast, terminal
selectors control the terminally differentiated state of a neuron, i.e., they initiate and maintain the stable identity of nondividing differentiated cells.
I will provide here a concise summary of the properties of terminal selectors as revealed primarily by studies in C. elegans, and I will organize these
properties around specific themes that are also highlighted in Fig. 1. The
description of these properties is based on the few original examples that
led to the original terminal selector definition (Hobert, 2008, 2011) but also
covers a plethora of recent examples that have allowed further generalization
of this concept. The themes that I will discuss are
Identity-defining terminal effector genes are coregulated.
Terminal selectors initiate and maintain the terminally differentiated state.
Terminal selectors act in reiteratively used combinations.
Terminal selectors control regulatory subroutines.
Parallel regulatory routines define specific aspects of neuronal identity.
Diversification of neuronal identity is achieved via selective gene
repression.
Generality of the terminal selector concept.

2. COREGULATION OF TERMINAL EFFECTOR GENES

BY TERMINAL SELECTORS
Every fully differentiated neuron type is uniquely defined by the
expression of a unique combination of nuts and bolts, terminal effector
genes that code for proteins that define the functional features of a mature
neuron type, such as neurotransmitter receptors, neurotransmitter synthesizing enzymes, neuropeptides, ion channels, sensory receptors, and synaptic
recognition molecules (see Table 2 for an exemplary list of terminal effector
genes). As illustrated in Fig. 1, the identification of cis-regulatory elements

Hourglass
topology

Multiple inputs

Continuous
expression
(autoregulation)

Operate in
combinations

Terminal
selectors

Terminal selector
binding site

Parallel
regulatory routines
TF b

DAF-19

TF c

Direct
coregulation
TF

Terminal effector genes: X

(Neurotransmitter synthesis &
transport, ion channels,
receptors, synaptic connectivity
neuropeptides, etc.)

X2

X3

X4

X5

TF

X6

X7

In neuron type #1

A2

A3

Pansensory
identity

Neuron-type
specific identity
Unique identity

Terminal
selectors

A1

B1

B2

B3

C1

C2

C3

Subroutines

Terminal
selectors

In neuron type #2

Terminal
selectors

Morphological
features

Shared with other neuron types

Modular
organization
of effectors

In neuron type #3

X1

Figure 1 See legend on opposite page.

Panneuronal
identity

Terminal Selectors of Neuronal Identity

461

required for the coexpression of terminal effector genes in specific neuron

types, and the analysis of the expression of many terminal effector genes in
many different terminal selector mutant backgrounds demonstrated a simple
common theme: neuron-type-specific terminal effector batteries are organized into regulons, i.e., they are all coregulated, and this coregulation is
mediated by shared cis-regulatory motifs and their cognate transcription factors in the form of specific terminal selector transcription factors (or combination thereof; as discussed below) (Cinar et al., 2005; Eastman et al.,
1999; Etchberger et al., 2007; Flames & Hobert, 2009; Gordon &
Hobert, 2015; Kratsios et al., 2011; Wenick & Hobert, 2004; Zhang
et al., 2002).
The broadness of the effect that terminal selectors exert on neuronal
identity is illustrated by the way that these effects have been measured in
many different neurons types: the repertoire of gfp reporter transgenes that
the C. elegans community has built over the past two decades is large enough
that for the majority of the 118 anatomically defined neuron types anything
between 3 and several dozen molecular markers have been described.
These reporters most often mark biochemically unlinked molecular features
of a neuron and provide a random snapshot of the identity of a neuron. In
most cases examined, terminal selectors affect either all or most of the
markers examined. In several instances many dozen markers were tested
(Etchberger et al., 2007; Kratsios et al., 2011; Wenick & Hobert, 2004),
in others just a handful of markers were tested (e.g., Serrano-Saiz et al.,

Figure 1 Key features of gene expression programs in mature neuron types. (A) Terminal selector regulons. Red squares highlight key concepts. Sustained expression of
terminal selectors is often, but not always ensured by autoregulation. Small rectangles
illustrate binding sites for transcription factors and indicate direct regulation of effector
genes by terminal selectors. Terminal selectors shown on the left control neuron-type
specific identity features that assign a unique identity to a neuron. See Table 2 for an
exemplary list of effector genes. Parallel regulatory routines such as those that regulate
pansensory features (via DAF-19) are controlled by factors that could also considered to
be terminal selectors; but they do not assign unique identities. Morphology regulators
control generic aspect of a neuron's morphology, such as placement of axons/dendrites
into specific fascicles or axo/dendritic polarity. TF, transcription factor. (B) Modular organization of terminal effector genes. X1 is a representative of any effector gene expressed
in more than one neuron type (e.g., the vesicular transporter of glutamate, which is
expressed in all glutamatergic neurons, but not in GABA or cholinergic neurons). This
schematic also indicates the key principle of combinatorial reusage of the same
terminal selector in different neuron types, as schematized in more detail in Fig. 2.

462

Oliver Hobert

Table 2 Examples of Effector Genes of Terminal Selectors

Categories
Examples

Neurotransmitter pathway GAD, TPH, TH, etc.

synthesis
Vesicular neurotransmitter VGLUT, VGAT, VMAT, VAChT
transporters
Neurotransmitter
receptors

ACh-gated, Glu-gated, GABA-gated ion channels,

metabotropic ACh-, Glu-receptors, biogenic amine
receptors

Ion channels

DEG/EnaC, TRP, TWK, Ca2+, K+ channels

Gap junctions

Innexins

Sensory receptors

Olfactory GPCRs, receptor-type guanylyl cyclases,

mechanosensory MEC channels

Neuropeptides

FMRFamides, NLPs

Neuropeptide receptors

FMRFamide-, tachykinin-, somatostatin-like

receptors (and others)

Synaptic organizers

oig-1, madd-4, neuroligin

Signaling

TGF (dbl-1, unc-129), Slit/slt-1, G-proteins

Others

Tetraspanins, IgSF, etc.

The table is meant to illustrate the range of biochemical activities that are under terminal selector control.
The examples shown were derived from the following studies: Cinar, Keles, and Jin (2005), Eastman,
Horvitz, and Jin (1999), Etchberger et al. (2007), Flames and Hobert (2009), Gordon and Hobert
(2015), Howell, White, and Hobert (2015), Kratsios et al. (2015, 2011), Serrano-Saiz et al. (2013),
Wenick and Hobert (2004), and Zhang et al. (2002).

2013). Ideally, one would want to be able to profile neurons in wild-type

and terminal selector-deficient neurons to assess the full extent of the effects.
While this has not been done yet, the random sampling of the markers
examined, combined with the broad defects observed in terminal selector
mutants, make the assessment of effects of terminal selector on neuronal
identity only an extrapolation, albeit a reasonable one.
Coregulation of a multitude of identity features is not self-evident.
A priori, one could easily envision that individual identity features could
be independently regulated via distinct transcriptional regulatory factors.
Transcriptome profiling studies of individual neuron types (in invertebrates
or vertebrates) usually reveal dozens of transcription factors expressed in a
mature neuron type (e.g., Etchberger et al., 2007), hence the job of initiating

Terminal Selectors of Neuronal Identity

463

and maintaining the expression of terminal gene batteries could, in theory,

be distributed over a number of distinct factors. However, as studies on the
transcription factors shown in Table 1 (as well as ensuing analysis of additional terminal selectors) revealed, the expression of many, perhaps most terminal markers for neuronal identity is affected in terminal selector mutants,
indicating that coregulation is achieved via a limited number of regulatory
factors.
Even in a nervous system as comparatively simple as C. elegans, only very
few terminal effector genes are exclusively expressed in a single neuron type
(e.g., olfactory receptor proteins). The much more prevalent scenario is that
genes, like neurotransmitter receptor subunits, neurotransmitter synthesizing enzymes, neuropeptides, or various types of ion channels are expressed in
several distinct neuron types. In each specific neuron type, these effector
genes are under control of neuron-type specific terminal selectors. This
effectively means that such effector genes harbor a modular array of response
elements each activated by the different terminal selectors present in each
neuron class where it is expressed (Fig. 1B). This quite intuitive concept
has been experimentally validated in a number of cases, including the loci
encoding the vesicular glutamate transporter (Serrano-Saiz et al., 2013) or
the vesicular acetylcholine or choline transporter (Kratsios et al., 2011;
Wenick & Hobert, 2004).

3. TERMINAL SELECTORS INITIATE AND MAINTAIN

THE TERMINALLY DIFFERENTIATED STATE
A key defining feature of terminal selectors is that they are not only
required to initiate terminal differentiation programs but are continuously
expressed throughout the life of the neuron (and animal) to maintain the differentiated state of a neuron (Doitsidou et al., 2013; Etchberger, Flowers,
Poole, Bashllari, & Hobert, 2009; Flames & Hobert, 2009; Kratsios et al.,
2011). This has been amply illustrated through temporally controlled, late
removal of terminal selectors not just in C. elegans but also in mice
(Deneris & Hobert, 2014; Holmberg & Perlmann, 2012; Laguna et al.,
2015). The continuous expression of terminal selectors is often ensured
through autoregulatory feedback mechanisms (Baran et al., 1999;
Etchberger et al., 2007; Hobert et al., 1997; Sagasti, Hobert, Troemel,
Ruvkun, & Bargmann, 1999; Way & Chalfie, 1989).
This continuous role sets terminal selectors clearly apart from earlier and
transiently acting transcription factors. For example, the C. elegans Zic gene

464

Oliver Hobert

ortholog REF-2 is required for the terminal differentiation of the AIY

interneurons, but it is only transiently expressed in the progenitor of the
AIY interneurons, where it acts to initiate the expression of the two cooperatively acting terminal selectors of AIY identity, ttx-3 and ceh-10 (Bertrand
& Hobert, 2009). If the only role of REF-2 is to turn a terminal selector,
REF-2 should not be considered a terminal selector because it is not
required to maintain the differentiated state and it cannot be considered
to be the terminal link of a regulatory cascade to a terminal effector gene
(such as a neurotransmitter pathway gene). In principle, one could envision a
scenario in which expression of a terminal effector is initiated by a transiently
acting factor binding to some cis-regulatory element in the effector gene
locus and its expression is then maintained via a separate cis-regulatory element that is controlled by a terminal selector. However, such a scenario has
not yet been documented.

4. TERMINAL SELECTORS ACT IN REITERATIVELY USED

COMBINATIONS
Terminal selectors are usually not single proteins that operate in complete isolation. Terminal selectors rather function in the context of combinations of cooperating factors (Fig. 2; Wenick & Hobert, 2004; Xue, Tu, &
Chalfie, 1993; Zhang et al., 2014). In one recently published example, four
terminal selectors cooperate in a given neuron type (Doitsidou et al., 2013).
In an unpublished example, seven terminal selectors cooperate on the level
of the cis-regulatory elements of the downstream effector genes (N. Flames
and O. Hobert, unpublished data). In many cases, no partners for terminal
selectors have been identified, but this should not be taken as evidence that
terminal selectors can act completely alone.
The biochemical basis for cooperation of terminal selectors appears to be
distinct in different cellular contexts. In some cases, terminal selectors exhibit
cooperative DNA binding (Wenick & Hobert, 2004; Xue et al., 1993). For
example, the UNC-86 and MEC-3 homeodomain proteins bind cooperatively to cis-regulatory elements required for the expression of effector genes
that define touch neuron identity (Duggan et al., 1998; Xue et al., 1993). In
other cases, terminal selectors may bind DNA in a noncooperative manner,
but exert cooperative effects possibly through the cooperative recruitment
of cofactors (Doitsidou et al., 2013; Gordon & Hobert, 2015). In yet other
cases, the type and extent of cooperation is unclear. For example, the serotonergic NSM neurons fail to terminally differentiate in double mutant

465

Terminal Selectors of Neuronal Identity

Terminal selector
CEH-43
(Dlx)

TTX-1
(Otx)

CEH-10
(Chx)

TTX-3
(Lhx)

AST-1
(ETS)

CEH-14
(Lhx)

UNC-86
(Pou)
AIY neuron
identity
HSN neuron
identity
NSM neuron
identity
PVR neuron
identity
AFD neuron
identity
Dopa neuron
identity

Figure 2 Combinatorial activity of terminal selectors. Selected examples are shown, some
not showing the complete set of factors (e.g., dopaminergic neurons are controlled by the
combined action of AST-1, CEH-43 plus a Pbx-type homeodomain factor). Note that all
except one factor (AST-1) are homeodomain transcription factors.

animals lacking the POU homeobox gene unc-86 and the LIM homeobox
gene ttx-3 (Zhang et al., 2014). However, in ttx-3 and unc-86 single mutants,
some identity features are either not affected at all or are already affected as
strongly as in the double mutant. There may be a spectrum of distinct types
of cooperation between terminal selectors, depending on cis-regulatory
architecture of individual effector genes.
There are a number of key implications of the combinatorial activity of
terminal selectors:
(1) It explains how one and the same terminal selector can control very
distinct fates. For example, UNC-86 is required to define the terminal
identity of vastly different neuron types (Table 1) and in several cases
we understand how. UNC-86 works cooperatively with MEC-3 to
control touch neuron effector genes via a well-defined UNC-86/
MEC-3 binding site (Xue et al., 1993). In contrast, in the BDU neurons,
UNC-86 binds to different sites and collaborates with another factor, the
Zn finger transcription factor PAG-3 (Gordon & Hobert, 2015).
(2) Conversely, the combinatorial logic also explains why one factor can
operate as a terminal selector in one cell type, while not having any
discernible effect on the same set of target genes in another cell type.

466

Oliver Hobert

For example, the ttx-3 LIM homeobox genes, expressed in the three
cholinergic interneuron types AIY, AIA, and AIN, controls the cholinergic properties of the AIY and AIA interneurons (as well as all other
known properties of these neurons), but does not control the cholinergic identity of the AIN interneurons (Zhang et al., 2014).
(3) Combinatorial activities provide a solution to the specificity problem
of transcription factors. Single transcription factors usually interact with
only about 412 nucleotides. Because such small sites are found too
commonly in the genome, the activation of very specific target genes
is likely mediated by the much more restrictive co-occurrence of combinations of terminal selector binding sites.
(4) Combinatorial function can also provide a solution to the sufficiency
problem. In many cases tested, terminal selectors found to be required
to control a specific differentiation program are sufficient to induce this
differentiation program in only some cellular contexts (e.g., Tursun,
Patel, Kratsios, & Hobert, 2011; Wenick & Hobert, 2004). A broader
effect may be achieved only upon misexpression of the complete set of
combinatorially acting terminal selectors.
The examination of more and more C. elegans neuron types has uncovered
another striking feature of transcription factor combinations. Even though
the C. elegans genome has a repertoire of almost 1000 transcription factors,
most terminal selectors are homeodomain transcription factors, which constitute only 10% of all C. elegans transcription factors. In those cases where
non-homeodomain terminal selectors have been identified (e.g., the ETS
domain factor AST-1), the non-homeodomain factor cooperates with a
homeodomain transcription factor (Doitsidou et al., 2013; Serrano-Saiz et
al., 2013). In cases where no homeodomain factor has been identified, it
is conceivable that a homeodomain factor still awaits identification. The preponderance of homeodomain transcription factors may be a reflection of
their very ancient use as drivers of neuronal identity.

5. TERMINAL SELECTORS CONTROL REGULATORY

SUBROUTINES
A number of transcription factors have been identified that control
only some, and perhaps in total very few genes in a terminally differentiated
neuron. One example is the ceh-23 homeobox gene which affects expression
of a GPCR gene, but not other identity aspects of the cholinergic AIY interneuron (Altun-Gultekin et al., 2001); similarly, the ceh-14 homeobox gene
affects expression of neuropeptides in the BDU neurons, but does not affect

Terminal Selectors of Neuronal Identity

467

other BDU identity features (Gordon & Hobert, 2015). Notably, in all these
cases, these transcriptional subroutines are under direct control of a terminal
selector (schematized in Fig. 1A). That is, the ceh-23 gene, as well as its target, the GPCR gene sra-11, are under direct control of the terminal selectors
of AIY identity, ttx-3 and ceh-10, which control scores of other terminal
identity features of the AIY interneuron (Altun-Gultekin et al., 2001;
Wenick & Hobert, 2004). Similarly, ceh-14 and its neuropeptide targets
are controlled by the terminal selectors of BDU identity, unc-86 and
pag-3 (Gordon & Hobert, 2015). Hence, in systems biology parlance, terminal selector frequently acts in feedforward loops (Alon, 2006) to control
effector genes (Fig. 1A). It is easily conceivable that any terminal selector
may regulate scores of regulatory subroutines. For example, the terminal
selector che-1 directly controls not only scores of nuts- and bolts-type terminal identity features (e.g., chemoreceptors, neuropeptides, and channels) via
well-defined CHE-1 binding sites but also controls the expression of dozens
of transcription factors in the ASE gustatory neurons; each of these transcription factors may control specific regulatory subroutines (Etchberger et al.,
2007). Indeed, some of these transcription factors ensure the left/right asymmetric expression of chemoreceptors in the left or right ASE neurons, yet
these factors do not control other identity features of the ASE neurons
(Chang, Johnston, & Hobert, 2003; Hobert, 2014).
Should those subroutine regulators also be considered as terminal selectors of these subroutines? The most striking difference is that in all known
cases, loss of these subroutine regulators still leaves the affected neuron with a
recognizable identity. For example, while loss of ceh-14 leads to loss of neuropeptide genes in BDU, the neuron is still recognizable as a BDU neuron in
terms of intact expression of many other marker genes and overall morphology. In contrast, loss of pag-3, the terminal selector of BDU results in loss of
all known molecular markers and alterations in neuron morphology. Even in
cases where some terminal identity features are unaffected in a terminal
selector mutant, these features are not sufficient to assign the neuron a specific identity. Therefore, it seems reasonable to propose that the terminal
selector term should be restricted to regulatory factors whose loss results
in a neuron not expressing a recognizable identity (or having switched to
an alternative identity; Gordon & Hobert, 2015).

6. PARALLEL REGULATORY ROUTINES

Terminal selectors control the expression of many if not all terminal
effector genes that are expressed in a neuron-type-specific manner and which, in

468

Oliver Hobert

a combinatorial manner define the unique identity of a neuron type. Such

effector genes include biosynthetic enzymes and transporters for specific
neurotransmitter systems, neurotransmitter receptors, ion channel, synaptic
adhesion molecules, neuropeptide, sensory receptors, and others (Table 2).
However, neuron identity-determining terminal selectors do not control the
expression of all genes in a terminally differentiated neuron type. Extensive
phenotypic analysis has shown that features that are shared by many or all
neuron types are not affected by removal of terminal selectors that control
neuron-type-specific features. The most striking feature that are commonly
unaffected by terminal selector are panneuronal genes (e.g., genes expressed in
all neurons, such as those encoding for synaptic vesicle proteins) or pansensory
genes (genes involved in building the ciliated structures of sensory neurons)
(Altun-Gultekin et al., 2001; Etchberger et al., 2007; Flames & Hobert,
2009; Kratsios et al., 2011). This argues that regulatory routines exist that
act in parallel to terminal selectors of neuron-type-specific identity
(Fig. 1A). While regulators of panneuronal gene expression are presently
ill defined (Stefanakis, Carrera, & Hobert, 2015), the parallel regulatory routine that controls pansensory identity features has been well documented.
The RFX-type transcription factor daf-19 controls the coordinated expression of genes involved in building cilia (Burghoorn et al., 2012; Swoboda,
Adler, & Thomas, 2000). daf-19 can therefore be considered to be a terminal
selector of ciliated identity.
While terminal selectors have profound effects on synaptic connectivity
of a neuron (Howell et al., 2015; Kratsios et al., 2015; Zhang et al., 2014;
Pereira et al., 2015), some overt morphological features remain relatively
intact in terminal selector mutants. For example, in unc-3 or unc-30 terminal
selector mutants, ventral nerve cord motor neurons are miswired, motor
axons still project in the ventral and dorsal nerve cord. Similarly, in che-1
terminal selector mutants, dendrites of the otherwise misspecified ASE
chemosensory neurons are still present and its cell body is correctly placed
(Uchida et al., 2003). Similarly, in unc-86 terminal selector mutants, the
HSN motor neuron fails to adopt terminal identity features such as its neurotransmitter identity, but still migrates normally (Desai et al., 1988). Hence,
there must be factors that act in parallel to terminal selectors to determine
some basic morphological features. There are few examples of such regulators. In the case of the HSN neurons, the transiently acting Zn finger transcription factor egl-43 controls migration but not other features HSN
identity (Baum, Guenther, Frank, Pham, & Garriga, 1999). A recent study
in flies has shown that transcription factors called morphology TF (mTFs)
define axonal/dendritic patterning features of motor neurons, but not other

Terminal Selectors of Neuronal Identity

469

terminal features, such as neurotransmitter identity (Enriquez et al., 2015).

Transiently acting mTFs are selectors of terminal morphological features of a
neuron, but their transient function in regulating effector genes that are
likely just transiently expressed (e.g., genes involved in axon pathfinding
or cell migration) clearly sets them apart from terminal selectors that are constitutively present in a neuron to initiate and maintain constitutively
expressed terminal identity features of a neuron.
Taken together, gene expression programs in terminally differentiated
neurons may be controlled by parallel acting regulons that define distinct
identity aspects; some of these identity aspects define the unique identity of a
neuron, while others define more generic aspects of a neuron identity, such
as pansensory features or some overt morphological features (Fig. 1A).

7. DIVERSIFICATION OF NEURONAL IDENTITY VIA

SELECTIVE REPRESSIBILITY
Not every single neuron type appears to have its own, unique complement of terminal selector(s). Rather, distinct but related neuronal subtypes can be diversified by modulating the activity of a terminal selector
on specific subsets of targets. This has been illustrated in two different contexts: the left and right ASE neurons are two neuronal subtypes that are very
similar to one another but differ in their complement of chemoreceptor
expression. Their identity is not controlled by distinct terminal selectors,
but is controlled by the same terminal selector, che-1 (Etchberger et al.,
2007). However, the ability of che-1 to activate some target genes (that contain CHE-1 binding sites) is prevented by the presence of repressor proteins
in one of the two ASE subtypes (Etchberger et al., 2009; Hobert, 2014). This
selective repression results in the ASEL- and ASER-expressing distinct subsets of genes, while simultaneously also expressing a vast amount of similar
genes (under control of che-1). Similarly, the terminal selector unc-3 controls
the expression of multiple genes that are expressed differentially in distinct
motor neuron subtypes. The ability of UNC-3 to activate some genes in
some subtypes is restrained by subtype-specific repressor proteins
(Kratsios et al., 2011; Winnier et al., 1999; our unpublished data).

8. GENERALITY OF THE TERMINAL SELECTOR CONCEPT

At the time of writing the original essay on the terminal selector concept, only a handful of examples of such selector genes has been described
(Hobert, 2008). By now, transcriptional regulators that control terminal

470

Oliver Hobert

URBL

ADFL

AWAL
ASEL
ASHL

AUAL
AIAL

OLQVL
SIBVL

URAVL
IL2VL

IL1VL

VL

EP

SMDDL

RMDDL
R

DL

SAA
RIH

EV

SMBVL

AINL

RIML

SIBDL AWCL

RMDL

AVJL

AVDL
RIBL

BAGL

URYVL

RIVL

AVHL

ASGL

AWBL

AIBL

AVEL
IL1L

ASIL

ASKL

URXL
R
SMDVL IAL
SAAVL
AFDL
AVAL

RMDVL

RIPL RMEL

Head

ADLL

PD

CE

AV
BL

OLLL
IL2L

AL
LA

RID

RMED
IL1DL OLQDL
DL
URADL URYDL

IL2

ASJL

RICL

ADEL

AIZL

ADAL

SIAVL

BD

RMGL

FLPL

RIS

SIADL

SM

AIML

AIYL
DB2

AVL

AVKL

VR
SAB
SABVL RIFL

RIFR AVG
RIGL

SABD
DD1

RIGR

DA1

DB1

Mid-body
Glu

Colored:
Identity regulator
known

ALML
CANL

ACh

HSNL

GABA
Aminergic

DA3

DD2

DB4

DA4

DB5

DD3

DA5

DD4

DB6

DA6

DD5

DB7

DA7

Unknown NT

White:
Identity regulator
unknown

Tail

Glu, ACh,
aminergic,
or unknown

DVA
PHBL
PVQL
PVPR
PVT

DD6
PVPL

PHAL

ALNL
LUAL

DVC

PVCL

PQR
PLML

DA9

DA8

Figure 3 An overview of identity regulator in an L1 stage animal. Colored neurons are

those in which terminal selector-type transcription factors have been identified that affect
multiple identity features, including (in those cases known) the neurotransmitter identity
of the neuron (color-coded). The only currently ambiguous cases included in this figure
are the command interneurons, in which one factor (unc-3) affects neurotransmitter
identity, but not other identity features, while another factor (unc-42) controls other
identity features, but not neurotransmitter identity (Pereira et al., 2015). Pharyngeal
neurons and postembryonically generated neurons are not shown. References: Altun,
Chen, Wang, and Hall (2009), Cassata et al. (2000), Doitsidou et al. (2013), Etchberger
et al. (2007), Gonzalez-Barrios et al. (2015), Gordon and Hobert (2015), Guillermin,
Castelletto, and Hallem (2011), Jin et al. (1994), Hutter (2003), Kim, Kim, and Sengupta
(2010), Kratsios et al. (2015, 2011), Lanjuin, VanHoven, Bargmann, Thompson, and
Sengupta (2003), Nokes et al. (2009), Pereira et al. (2015), Satterlee et al. (2001),
Sengupta et al. (1994), Serrano-Saiz et al. (2013), Sze, Zhang, Li, and Ruvkun (2002),
Van Buskirk and Sternberg (2010), Zhang et al. (2014), and Zhang et al. (2002).

identity features of more than two-thirds of the 118 neuron classes have been
identified (Fig. 3). Obviously, there is no other nervous system that is even
close in the extent to which neuronal differentiation programs have been
described throughout the nervous system. A concise summary of all these
will be provided in an upcoming comprehensive review that is currently
in preparation. Not all the regulators have been described in the same depth

Terminal Selectors of Neuronal Identity

471

as others. In contrast to the original, almost all or nothing examples in

which a mutation in a terminal selector affects the expression of dozens
of terminal markers, two 50/50 examples have now been described in
which some terminal markers are affected, while others are not (Pereira
et al., 2015; Zhang et al., 2014). It is conceivable that in these apparently
exceptional cases, the respective neuron may harbor several terminal selectors that can partially compensate for the loss of each other, as proposed for
the case of the NSM neurons (Zhang et al., 2014).

9. REGULATION OF TERMINAL SELECTOR GENES

As exemplified by the case of the terminal selector ttx-3, the induction
of terminal selector gene expression is complex. Regulatory control regions
of ttx-3 integrate a number of distinct autonomous and nonautonomous
inputs (Bertrand & Hobert, 2009; Murgan et al., 2015). Terminal selectors
could therefore perhaps be viewed as critical junctions in a regulatory hourglass topology, sampling a diversity of upstream inputs and then branching
out again to control a multitude of targets (Fig. 1A). The regulation of more
terminal selectors needs to be studied to assess how broadly applicable this
concept is.

10. CONCLUSIONS
The purpose of this chapter has been to provide a description of the
general features of gene regulatory programs that operate in terminally differentiated neuron types throughout the C. elegans nervous system. I have
focused here entirely on the C. elegans system. I will leave it to researchers
of other model systems to assess the extent of similarity between gene
expression programs in C. elegans in their favorite system and I hope that
some of the findings made in C. elegans may provide experimental testable
hypotheses for researchers in other systems.

ACKNOWLEDGMENTS
I thank former and current lab members whose work has been instrumental in building the
terminal selector concept. I thank lab members as well as Nuria Flames, Evan Deneris,
Richard Mann, Doug Allan and Stefan Thor for comments on the manuscript. Work in
my laboratory is funded by the National Institutes of Health and the Howard Hughes
Medical Institute.

472

Oliver Hobert

REFERENCES
Alon, U. (2006). An introduction to systems biology: Design principles of biological circuits. 1st ed.
London, UK: Chapman & Hall/CRC.
Altun, Z. F., Chen, B., Wang, Z. W., & Hall, D. H. (2009). High resolution map of
Caenorhabditis elegans gap junction proteins. Developmental Dynamics, 238, 19361950.
Altun-Gultekin, Z., Andachi, Y., Tsalik, E. L., Pilgrim, D., Kohara, Y., & Hobert, O.
(2001). A regulatory cascade of three homeobox genes, ceh-10, ttx-3 and ceh-23, controls cell fate specification of a defined interneuron class in C. elegans. Development, 128,
19511969.
Baran, R., Aronoff, R., & Garriga, G. (1999). The C. elegans homeodomain gene unc-42
regulates chemosensory and glutamate receptor expression. Development, 126,
22412251.
Bargmann, C. I., Hartwieg, E., & Horvitz, H. R. (1993). Odorant-selective genes and neurons mediate olfaction in C. elegans. Cell, 74, 515527.
Bargmann, C. I., & Horvitz, H. R. (1991). Chemosensory neurons with overlapping functions direct chemotaxis to multiple chemicals in C. elegans. Neuron, 7, 729742.
Baum, P. D., Guenther, C., Frank, C. A., Pham, B. V., & Garriga, G. (1999). The
Caenorhabditis elegans gene ham-2 links Hox patterning to migration of the HSN motor
neuron. Genes & Development, 13, 472483.
Bertrand, V., & Hobert, O. (2009). Linking asymmetric cell division to the terminal differentiation program of postmitotic neurons in C. elegans. Developmental Cell, 16, 563575.
Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics, 77, 7194.
Burghoorn, J., Piasecki, B. P., Crona, F., Phirke, P., Jeppsson, K. E., & Swoboda, P. (2012).
The in vivo dissection of direct RFX-target gene promoters in C. elegans reveals a novel
cis-regulatory element, the C-box. Developmental Biology, 368, 415426.
Cassata, G., Kagoshima, H., Andachi, Y., Kohara, Y., Durrenberger, M. B., Hall, D. H.,
et al. (2000). The LIM homeobox gene ceh-14 confers thermosensory function to
the AFD neurons in Caenorhabditis elegans. Neuron, 25, 587597.
Chalfie, M., & Sulston, J. (1981). Developmental genetics of the mechanosensory neurons of
Caenorhabditis elegans. Developmental Biology, 82, 358370.
Chalfie, M., Sulston, J. E., White, J. G., Southgate, E., Thomson, J. N., & Brenner, S. (1985).
The neural circuit for touch sensitivity in Caenorhabditis elegans. The Journal of Neuroscience, 5, 956964.
Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., & Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science, 263, 802805.
Chang, S., Johnston, R. J., Jr., & Hobert, O. (2003). A transcriptional regulatory cascade that
controls left/right asymmetry in chemosensory neurons of C. elegans. Genes & Development, 17, 21232137.
Cinar, H., Keles, S., & Jin, Y. (2005). Expression profiling of GABAergic motor neurons in
Caenorhabditis elegans. Current Biology, 15, 340346.
Deneris, E. S., & Hobert, O. (2014). Maintenance of postmitotic neuronal cell identity.
Nature Neuroscience, 17, 899907.
Desai, C., Garriga, G., McIntire, S. L., & Horvitz, H. R. (1988). A genetic pathway for the
development of the Caenorhabditis elegans HSN motor neurons. Nature, 336, 638646.
Doitsidou, M., Flames, N., Topalidou, I., Abe, N., Felton, T., Remesal, L., et al. (2013). A
combinatorial regulatory signature controls terminal differentiation of the dopaminergic
nervous system in C. elegans. Genes & Development, 27, 13911405.
Duggan, A., Ma, C., & Chalfie, M. (1998). Regulation of touch receptor differentiation by
the Caenorhabditis elegans mec-3 and unc-86 genes. Development, 125, 41074119.
Eastman, C., Horvitz, H. R., & Jin, Y. (1999). Coordinated transcriptional regulation of the
unc-25 glutamic acid decarboxylase and the unc-47 GABA vesicular transporter by the

Terminal Selectors of Neuronal Identity

473

Caenorhabditis elegans UNC-30 homeodomain protein. The Journal of Neuroscience, 19,

62256234.
Edlund, T., & Jessell, T. M. (1999). Progression from extrinsic to intrinsic signaling in cell fate
specification: A view from the nervous system. Cell, 96, 211224.
Enriquez, J., Venkatasubramanian, L., Baek, M., Peterson, M., Aghayeva, U., & Mann, R. S.
(2015). Specification of individual adult motor neuron morphologies by combinatorial
transcription factor codes. Neuron, 86, 955970.
Etchberger, J. F., Flowers, E. B., Poole, R. J., Bashllari, E., & Hobert, O. (2009). Cisregulatory mechanisms of left/right asymmetric neuron-subtype specification in
C. elegans. Development, 136, 147160.
Etchberger, J. F., Lorch, A., Sleumer, M. C., Zapf, R., Jones, S. J., Marra, M. A.,
et al. (2007). The molecular signature and cis-regulatory architecture of a C. elegans gustatory neuron. Genes & Development, 21, 16531674.
Finney, M., Ruvkun, G., & Horvitz, H. R. (1988). The C. elegans cell lineage and differentiation gene unc-86 encodes a protein with a homeodomain and extended similarity to
transcription factors. Cell, 55, 757769.
Flames, N., & Hobert, O. (2009). Gene regulatory logic of dopamine neuron differentiation.
Nature, 458, 885889.
Garcia-Bellido, A. (1975). Genetic control of wing disc development in Drosophila. Ciba
Foundation Symposium, 29, 161182.
Gonzalez-Barrios, M., Fierro-Gonzalez, J. C., Krpelanova, E., Mora-Lorca, J. A.,
Pedrajas, J. R., Penate, X., et al. (2015). Cis- and trans-regulatory mechanisms of gene
expression in the ASJ sensory neuron of Caenorhabditis elegans. Genetics, 200, 123134.
Gordon, P. M., & Hobert, O. (2015). A competition mechanism for a homeotic neuron
identity transformation in C. elegans. Developmental Cell, 34, 206219.
Guillermin, M. L., Castelletto, M. L., & Hallem, E. A. (2011). Differentiation of carbon
dioxide-sensing neurons in Caenorhabditis elegans requires the ETS-5 transcription factor. Genetics, 189, 13271339.
Hedgecock, E. M., & Russell, R. L. (1975). Normal and mutant thermotaxis in the nematode
Caenorhabditis elegans. Proceedings of the National Academy of Sciences of the United States of
America, 72, 40614065.
Hemmati-Brivanlou, A., & Melton, D. (1997). Vertebrate neural induction. Annual Review of
Neuroscience, 20, 4360.
Hobert, O. (2008). Regulatory logic of neuronal diversity: Terminal selector genes and selector motifs. Proceedings of the National Academy of Sciences of the United States of America, 105,
2006720071.
Hobert, O. (2011). Regulation of terminal differentiation programs in the nervous system.
Annual Review of Cell and Developmental Biology, 27, 681696.
Hobert, O. (2014). Development of left/right asymmetry in the Caenorhabditis elegans nervous system: From zygote to postmitotic neuron. Genesis, 52, 528543.
Hobert, O., Mori, I., Yamashita, Y., Honda, H., Ohshima, Y., Liu, Y., et al. (1997). Regulation of interneuron function in the C. elegans thermoregulatory pathway by the ttx-3
LIM homeobox gene. Neuron, 19, 345357.
Holmberg, J., & Perlmann, T. (2012). Maintaining differentiated cellular identity. Nature
Reviews. Genetics, 13, 429439.
Howell, K., White, J. G., & Hobert, O. (2015). Spatiotemporal control of a novel synaptic
organizer molecule. Nature, 523, 8387.
Hutter, H. (2003). Extracellular cues and pioneers act together to guide axons in the ventral
cord of C. elegans. Development, 130, 53075318.
Jin, Y., Hoskins, R., & Horvitz, H. R. (1994). Control of type-D GABAergic neuron differentiation by C. elegans UNC-30 homeodomain protein. Nature, 372, 780783.

474

Oliver Hobert

Kim, K., Kim, R., & Sengupta, P. (2010). The HMX/NKX homeodomain protein MLS-2
specifies the identity of the AWC sensory neuron type via regulation of the ceh-36 Otx
gene in C. elegans. Development, 137, 963974.
Kratsios, P., Pinan-Lucarre, B., Kerk, S. Y., Weinreb, A., Bessereau, J. L., & Hobert, O.
(2015). Transcriptional coordination of synaptogenesis and neurotransmitter signaling.
Current Biology, 25, 12821295.
Kratsios, P., Stolfi, A., Levine, M., & Hobert, O. (2011). Coordinated regulation of cholinergic motor neuron traits through a conserved terminal selector gene. Nature Neuroscience,
15, 205214.
Laguna, A., Schintu, N., Nobre, A., Alvarsson, A., Volakakis, N., Jacobsen, J. K.,
et al. (2015). Dopaminergic control of autophagic-lysosomal function implicates Lmx1b
in Parkinsons disease. Nature Neuroscience, 18, 826835.
Lanjuin, A., VanHoven, M. K., Bargmann, C. I., Thompson, J. K., & Sengupta, P. (2003).
Otx/otd homeobox genes specify distinct sensory neuron identities in C. elegans. Developmental Cell, 5, 621633.
Lewis, J. A., & Hodgkin, J. A. (1977). Specific neuroanatomical changes in chemosensory
mutants of the nematode Caenorhabditis elegans. The Journal of Comparative Neurology,
172, 489510.
McIntire, S. L., Jorgensen, E., & Horvitz, H. R. (1993). Genes required for GABA function
in Caenorhabditis elegans. Nature, 364, 334337.
Mori, I., & Ohshima, Y. (1995). Neural regulation of thermotaxis in Caenorhabditis elegans.
Nature, 376, 344348.
Murgan, S., Kari, W., Rothbacher, U., Iche-Torres, M., Melenec, P., Hobert, O.,
et al. (2015). Atypical transcriptional activation by TCF via a Zic transcription factor
in C. elegans neuronal precursors. Developmental Cell, 33, 737745.
Nokes, E. B., Van Der Linden, A. M., Winslow, C., Mukhopadhyay, S., Ma, K., &
Sengupta, P. (2009). Cis-regulatory mechanisms of gene expression in an olfactory neuron type in Caenorhabditis elegans. Developmental Dynamics, 238, 30803092.
Pereira, L., Kratsios, P., Serrano-Saiz, E., Sheftel, H., Mayo, A., Hall, D. H., et al. (2015). A
cellular and regulatory map of the cholinergic nervous system of C. elegans. Elife, 4.
pii: e12432. doi: 10.7554/eLife.12432.
Prasad, B. C., Ye, B., Zackhary, R., Schrader, K., Seydoux, G., & Reed, R. R. (1998). unc3, a gene required for axonal guidance in Caenorhabditis elegans, encodes a member of
the O/E family of transcription factors. Development, 125, 15611568.
Sagasti, A., Hobert, O., Troemel, E. R., Ruvkun, G., & Bargmann, C. I. (1999). Alternative
olfactory neuron fates are specified by the LIM homeobox gene lim-4. Genes & Development, 13, 17941806.
Satterlee, J. S., Sasakura, H., Kuhara, A., Berkeley, M., Mori, I., & Sengupta, P. (2001). Specification of thermosensory neuron fate in C. elegans requires ttx-1, a homolog of otd/
Otx. Neuron, 31, 943956.
Sengupta, P., Colbert, H. A., & Bargmann, C. I. (1994). The C. elegans gene odr-7 encodes
an olfactory-specific member of the nuclear receptor superfamily. Cell, 79, 971980.
Serrano-Saiz, E., Poole, R. J., Felton, T., Zhang, F., de la Cruz, E. D., & Hobert, O. (2013).
Modular control of glutamatergic neuronal identity in C. elegans by distinct
homeodomain proteins. Cell, 155, 659673.
Stefanakis, N., Carrera, I., & Hobert, O. (2015). Regulatory logic of pan-neuronal gene
expression in C. elegans. Neuron, 87, 733750.
Sulston, J. E. (1983). Neuronal cell lineages in the nematode Caenorhabditis elegans. Cold
Spring Harbor Symposia on Quantitative Biology, 48, 443452.
Swoboda, P., Adler, H. T., & Thomas, J. H. (2000). The RFX-type transcription factor
DAF-19 regulates sensory neuron cilium formation in C. elegans. Molecular Cell, 5,
411421.

Terminal Selectors of Neuronal Identity

475

Sze, J. Y., Zhang, S., Li, J., & Ruvkun, G. (2002). The C. elegans POU-domain transcription
factor UNC-86 regulates the tph-1 tryptophan hydroxylase gene and neurite outgrowth
in specific serotonergic neurons. Development, 129, 39013911.
Tsalik, E. L., & Hobert, O. (2003). Functional mapping of neurons that control locomotory
behavior in Caenorhabditis elegans. Journal of Neurobiology, 56, 178197.
Tursun, B., Patel, T., Kratsios, P., & Hobert, O. (2011). Direct conversion of C. elegans
germ cells into specific neuron types. Science, 331, 304308.
Uchida, O., Nakano, H., Koga, M., & Ohshima, Y. (2003). The C. elegans che-1 gene
encodes a zinc finger transcription factor required for specification of the ASE
chemosensory neurons. Development, 130, 12151224.
Van Buskirk, C., & Sternberg, P. W. (2010). Paired and LIM class homeodomain proteins
coordinate differentiation of the C. elegans ALA neuron. Development, 137, 20652074.
Waddington, C. H. (1957). The strategy of the genes. London, UK: Allen and Unwin.
Way, J. C., & Chalfie, M. (1988). mec-3, a homeobox-containing gene that specifies differentiation of the touch receptor neurons in C. elegans. Cell, 54, 516.
Way, J. C., & Chalfie, M. (1989). The mec-3 gene of Caenorhabditis elegans requires its own
product for maintained expression and is expressed in three neuronal cell types. Genes &
Development, 3, 18231833.
Wenick, A. S., & Hobert, O. (2004). Genomic cis-regulatory architecture and trans-acting
regulators of a single interneuron-specific gene battery in C. elegans. Developmental Cell,
6, 757770.
White, J. G., Southgate, E., Thomson, J. N., & Brenner, S. (1986). The structure of the nervous system of the nematode Caenorhabditis elegans. Philosophical Transactions of the Royal
Society of London. Series B, Biological Sciences, 314, 1340.
Winnier, A. R., Meir, J. Y., Ross, J. M., Tavernarakis, N., Driscoll, M., Ishihara, T.,
et al. (1999). UNC-4/UNC-37-dependent repression of motor neuron-specific genes
controls synaptic choice in Caenorhabditis elegans. Genes & Development, 13,
27742786.
Xue, D., Tu, Y., & Chalfie, M. (1993). Cooperative interactions between the Caenorhabditis
elegans homeoproteins UNC-86 and MEC-3. Science, 261, 13241328.
Zhang, F., Bhattacharya, A., Nelson, J. C., Abe, N., Gordon, P., Lloret-Fernandez, C.,
et al. (2014). The LIM and POU homeobox genes ttx-3 and unc-86 act as terminal selectors in distinct cholinergic and serotonergic neuron types. Development, 141, 422435.
Zhang, Y., Ma, C., Delohery, T., Nasipak, B., Foat, B. C., Bounoutas, A., et al. (2002).
Identification of genes expressed in C. elegans touch receptor neurons. Nature, 418,
331335.