2 Combinatorial Chemistry & High Throughput Screening, 2009, 12, 2-23

1386-2073/09 $55.00+.00 © 2009 Bentham Science Publishers Ltd.
Evolution of the Human Ion Channel Set
Timothy J. Jegla*,1, Christian M. Zmasek2, Serge Batalov3 and Surendra K. Nayak4
1The Scripps Research Institute, La Jolla, CA 92037, USA
2Burnham Institute, San Diego, CA 92037, USA
3Genomics Institute of the Novartis Research Foundation, San Diego, CA 92121, USA
4Exelexis, San Diego, CA 92121, USA
Abstract: Ion channels are intimately involved in virtually every physiological process of consequence in humans. Their
importance is underscored by the identification of numerous “channelopathies”, human diseases caused by ion channel
mutations. Ion Channels have consequently been viewed as fertile ground for drug discovery and, indeed, they represent
one of the largest target classes for current medicines. The future prospects of ion channels as a target class are tied to the
functional characterization of the human ion channel set on a genomic scale. The focus of this review is to describe the
molecular diversity and conservation of human ion channels. The human genome contains at least 232 genes that encode
the pore-forming subunits of plasma membrane ion channels. Comparative genome analysis shows that most human ion
channel gene families have their origins in the earliest metazoans but the human genes are largely derived from duplica-
tions that took place in the vertebrate lineage. The mouse and human ion channel gene sets are virtually identical, but dif-
fer significantly from fish channel sets. Genome comparisons highlight a number of highly conserved channel families
that do not yet have specifically defined functional roles in vivo. These channel families are likely to have non-redundant
functions in metazoans and represent some of the best new opportunities for channel target prospecting. Furthermore, ge-
nome-wide patterns of sequence conservation can now be used to refine strategies for the identification of gene-specific
channel probes.
Keywords: Ion channels, evolution, Kv, TRP, Nematostella, Monosiga, Kv, K2P.
INTRODUCTION
Characterizing the functional and molecular diversity of
human ion channel genes has been of considerable interest to
the scientific community because of the pervasive role chan-
nels play in human physiology. Numerous causal associa-
tions have been established between ion channel mutations
and human disease. A growing list of “channelopathies”
links ion channel dysfunction to a long list of major human
health problems including epilepsy, migraine, cardiac ar-
rhythmia, kidney failure, and blindness. Ion channels are
theoretically attractive drug targets because they are often
the end effectors of signaling cascades and processes. Chan-
nels have indeed become one of the principle target classes
exploited by the pharmaceutical industry, and the success of
ion channel-based drugs continues to peak interest in the
identification of new channel targets.
The set of human ion channel genes has now been largely
defined through two decades of gene hunting and the human
genome project. The search for new channel drug targets is
therefore moving beyond gene discovery to comparative
evaluation of the finite set of channel genes in experiments
aimed at determining disease relevance. Nevertheless, there
is still a lot to be gained from approaching such work with a
genome-level understanding of ion channel diversity and
evolution. The purpose of this review is to explore the


*Address correspondence to this author at the Department of Cell Biology,
ICND-210C, The Scripps Research Institute, 10550 North Torrey Pines
Road, La Jolla, CA 92037, USA; Tel: (858)-784-2506;
Fax: (858)-784-7149; E-mail: tjegla@scripps.edu
molecular diversity and origins of the human ion channel set
with an eye towards issues relevant to drug discovery. A
thorough examination of the human channel set suggests that
many opportunities for channel-based drug discovery remain
to be exploited. For instance, a surprisingly large number of
human ion channels still lack chemical probes of sufficient
quality to drive target validation research. Furthermore, ge-
nome sequences from organisms representing the major
kingdoms of cellular life, including a phylogenetically di-
verse set of metazoans, now gives us a much clearer picture
of how the human ion channel genes evolved. Such informa-
tion is valuable for understanding how to use model species
and reveals a substantial number of highly conserved chan-
nel types that remain functionally uncharacterized in vivo.
Genome-wide analysis of sequence conservation patterns can
inform target site selection and identify potential selectivity
problems for ion channel drug discovery efforts.
The scope of this review will be limited to genes that
encode the integral pore-forming subunits of plasma mem-
brane ion channel genes. Some gene families, such as the
ABC transporters, where only a few members have ion
channel activity, are also excluded. Many channels consist of
both pore-forming and regulatory subunits, and most chan-
nels are undoubtedly part of large macromolecular signal
transduction complexes. However, the state of our knowl-
edge regarding these complexes is still primitive, and thus it
is not yet possible to perform a comprehensive genome-wide
survey of channel diversity beyond the level of the pore-
forming subunits. Thus discussion of regulatory subunits will
be limited.

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 3
CHANNEL NAMING CONVENTIONS
Channel names have been a source of constant confusion
over the years. Multiple reports of novel mammalian channel
genes would often appear in the literature virtually simulta-
neously under distinct names. Recent efforts to standardize
names of ion channel genes have come from two different
groups: 1) The International Union of Basic and Clinical
Pharmacology (IUPHAR, http://www.iuphar-db.org/index_
ic.jsp), which represents the attempt of ion channel research-
ers to “put their house in order”, and 2) The HUGO Gene
Nomenclature Committee (http://www.genenames.org/),
which has the goal of assigning a unique name to each hu-
man gene. The IUPHAR names are preferred within the
channel field and will be used here when available, but the
HUGO name list is currently more comprehensive and is
widely used among genome biologists. Both conventions
effectively give channel genes a unique identifier, although
the IUPHAR names are generally more descriptive of chan-
nel function (KV for K
+ channel, voltage-gated, for instance).
It should be noted that each naming system has inconsisten-
cies with molecular phylogeny which can lead to misinter-
pretations of evolutionary relationships if not understood.
HUGO names are well separated by gene family, but often
do not differentiate between distinct gene subfamilies. As an
example, the HUGO designations for the 8 human members
of the Ether-ago-go (EAG) K+ channel gene family
(KCNH1-8) do not differentiate between three distinct gene
subfamilies Eag (KCNH1, KCNH5), Erg (KCNH2, KCNH6,
KCNH7) and Elk (KCNH3, KCNH4, KCNH8). The
IUPHAR names focus on the functional role over phyloge-
netic origin. Thus in cases where distinct gene families have
functional similarity, the names can gloss over phylogenetic
relationships. For instance, the Kv1 through Kv12 designa-
tions for voltage-gated potassium channels wrap genes from
three structurally distinct families into the same naming con-
vention. Thus it is not apparent from the Kv10-12 designa-
tion that EAG family potassium channels are actually more
closely related to cyclic nucleotide-gated cation channels and
hyperpolarization-gated cation channels [1] than they are to
other voltage-gated K+channels. Nevertheless, these naming
systems have been extremely helpful for efforts to classify
mammalian ion channel genes, and the few imperfections are
easily learned. A comprehensive table of IUPHAR and
HUGO channel names for voltage-gated cation channels can
be found is presented in Yu and Catterall (2004) [1].
WHAT IS AN ION CHANNEL?
There is no one signature sequence that identifies ion
channels, as there is for the vast majority of eukaryotic ser-
ine/threonine protein kinases [2, 3]. For instance, voltage-
gated Cl- channels and K+ channels share no significant se-
quence homology and have distinct structural basis for ion
permeation [4, 5]. Thus as a protein class, ion channels may
have evolved de novo several times in response to distinct
physiological needs. Cartoon representations of the major
structural classes of ion channels identified in mammalian
genomes are shown in Fig. (1). Channel pores differ widely
in their transmembrane (TM) topology and subunit
stoichiometry, ranging from 6-36 TM domains and 2-5
subunits. This variation in structure underlies a vast diversity
in channel pores and gating mechanisms. The most basic
uniting feature of channel proteins is simply the ability to
form an ion-permeable pathway through biological mem-
branes. Channels differ in definition from ion transporters in
that the movement of ions through the pore occurs passively.
The energy for ion movement is simply provided by existing
transmembrane electrochemical gradients for the permeant
ions. Channels are not able to move ions against these gradi-
ents. Transporters, on the other hand, can move ions in op-
position to electrochemical gradients because they couple ion
transport to energy derived from alternate sources, such as
ATP hydrolysis. Alternatively, many transporters reduce
energy requirements for transporting ions against gradients
by co-transporting a second ion in an energetically favorable
direction. Transporters can therefore establish concentration
gradients of ions that are exploited by ion channels and rep-
resent an intriguing therapeutic target class in their own
right. However, channels are better suited for rapid signaling
because they typically support higher conduction rates. The
functional separation between transporters and channels in
reality may be more of a continuum than a sharp division.
The CLC Cl- channel family blurs the line between trans-
porters and channels [6]; the crystallized bacterial CLC-ec1
and mammalian CLCs expressed in intracellular organelles
appear to be H+/Cl- co-transporters [5, 7-11], while other
mammalian family members are indeed plasma membrane
ion channels.
BASIC ION CHANNEL GATING MECHANISMS
Conduction through ion channels is passive, but gating of
the pore in most cases requires energetic input. The vast ma-
jority of channels have evolved to derive the energy required
for gating from one or more of three sources: transmembrane
voltage, mechanical force and ligand binding. Voltage-gated
channels tap the energy stored in transmembrane voltage
gradients to drive the conformational changes required for
opening and closing the pore. Voltage-gated CLC family Cl-
channels and voltage-gated cation channels have distinct
structures and evolutionary origins, and both are found in all
major divisions of cellular life. CLC family Cl- channels
have a double-barreled pore [5, 12] (Fig. 1). Each subunit
contains its own pore and the channels function as homodi-
mers [5, 13]. They do not have an intrinsic voltage sensor;
instead, voltage-sensitivity derives from occupancy of a
chloride binding site in the pore [14]. Voltage-gated cation
channels are far more numerous and functionally diverse
than Cl- channels in all metazoan genomes and form by far
the largest group of channels that can be traced to a single
evolutionary origin. They share a similar structural motif
consisting of a four transmembrane voltage sensor domain
(VSD) coupled to a 2 TM domain ion pore motif (Fig. 1).
Voltage changes cause movement of a positively charged
helix (S4) in the VSD which in turn imparts a mechanical
force on the pore domain to open or close the channel [15].
The mechanism and magnitude of voltage sensor movements
during gating is currently a topic of much debate. Crystal
structures of the voltage sensor so far reveal only the open
state [15-17]. The issue is of significant relevance for drug
discovery because the voltage sensor domain, present in ap-
proximately one half of all human ion channels, is a promis-
ing target site for drug binding [18]. Understanding the na-
ture and variability of voltage sensor movements among dis-
tinct channel subtypes will help to structurally define prom-
ising binding pockets for chemical modulators.

4 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
The 6 TM domain voltage-gated cation channel subunit
motif may have its origins in the fusion of two highly adapt-
able domains. The 2 TM pore domain exists in isolation in
inward rectifier K+ channels (which also have ancient pro-
karyotic origins). Similarly, the identification of a VSD cou-
pled to an enzyme instead of a channel pore [19] and a VSD
functioning independently as voltage-gated proton channel
[20, 21] suggests that the voltage sensor may have evolved
independently of the pore domain. The two domains could
have been fused during a later event. However, these particu-
lar isolated VSD proteins may not have the same phyloge-
netic spread as canonical voltage-gated ion channels. They
have so far been identified only in deuterostomes. Therefore,
it is also possible that proteins with isolated voltage sensors
could instead represent functional derivatives of voltage-
gated ion channels.
Mechanisms of ligand-gating in ion channels are far
more diverse, as is the structure of ligand gated ion channels.
Indeed, a common theme in the functional diversification of
voltage-gated cation channels has been the addition of spe-
cialized ligand-gating domains. An excellent example is the
large-conductance Ca2+-activated K+ channel (BK) which
contains a typical voltage-gated cation channel motif ap-
pended to an RCK-containing C-terminal domain involved
in divalent gating [22-25]. The mammalian Slo channel
(KCa1) contains an additional “bowl” domain that is also
involved in Ca2+-dependent gating [24, 26]. Depolarization
and Ca2+ work together in an allosteric manner to gate the





















Fig. (1). Structural cartoons of metazoan ion channel subunits and pores. Transmembrane topology of ion channel subunits is depicted with
rectangles representing transmembrane (TM) domains. Cartoons are oriented with extracellular domains at the top and intracellular domains
at the bottom. Schematic drawings of channel pores are shown from above; individual peptide subunits each have a different color. (A) KIR
inward rectifier K+ channels have a 2TM topology with an intervening membrane-embedded selectivity filter. Functional channels are
tetramers. K2P channels contain tandemly linked KIR-like motifs for a total of 4 TMs and 2 filters; they form pores as dimers. Metazoan
ionotropic glutamate receptors contain an inverted KIR-like motif with an appended TM domain (grey) and function as tetramers. ORAI
channels have 4 TMS, share and no homology with KIR channels, but function as tetramers. (B) Voltage-gated cation channels have a 4 TM
voltage sensor domain (VSD) appended to the basic KIR pore motif (P). Most voltage-gated cation channels function as tetramers; the VSD
of individual subunits is adjacent to the neighboring pore motif. TPC Ca2+ channels contain two tandemly linked voltage-gated channel mo-
tifs and probably have a dimeric conduction pathway. Voltage-gated Na+ and Ca2+ channels and the Na+ leak channel NALCN contain four
voltage-gated channel homology domains and form pores as monomers. (C) P2X receptors and ASIC/DEG family sodium channels contain
2 TM domains and function as trimers. They surprisingly share no significant sequence homology with each other. (D) C-loop neurotransmit-
ter receptors have extracellular termini, 4 TMs and form pores as pentamers. The thick gray line depicts the extracellular disulfide bridge
from which they derive their name. (E) CLC family chloride channel subunits have a complicated transmembrane topology; the subunit car-
toon has been simplified to show the antiparallel structural repeat in CLC subunits. Functional channels are dimers with a separate pore in
each subunit.

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 5
pore [27]. The result is a channel that will only open at
physiological voltages if depolarization is coupled to a rise
in intracellular Ca2+. A more extreme example of ligand-
gating in the voltage-gated channel superfamily is presented
by cyclic nucleotide-gated (CNG) cation channels: voltage
sensing has essentially been lost (despite the retention of a
VSD) and gating is driven almost exclusively by binding of
cyclic nucleotides to a cytoplasmic cyclic nucleotide-binding
domain (CNBD) [28]. These channels serve as the basis of
mammalian vision and smell; they generate sensory poten-
tials in response to changes in cyclic nucleotide levels driven
by activation of opsins or olfactory receptors. Hyperpolariza-
tion-gated cation channels (HCN) and Ether-a-go-go family
K+ channels share the basic body plan of CNG channels, but
voltage remains the dominant gating mechanism [28, 29].
Ligand-gated ionotropic neurotransmitter receptors,
comprising the ionotropic glutamate receptors and the C-
loop neurotransmitter receptors, differ markedly from volt-
age-gated cation channels and from each other in pore struc-
ture. Mammalian glutamate receptor subunits have 3 trans-
membrane domains and function as tetramers, while the C-
loop receptors have four transmembrane domains and func-
tion as pentamers (Fig. 1). Interestingly, both classes of neu-
rotransmitter receptor have their origins in prokaryotes [30-
32], suggesting that they initially evolved in response to fill a
metabolic role, perhaps coupling ion flux to detection of nu-
trients. Glutamate receptors may have evolved from K+
channels; a prokaryotic glutamate receptor carries the K+-
selective pore motif but has been inverted in the membrane
[30]. The third transmembrane domain has been appended to
metazoan glutamate receptors, but they still share a
tetrameric structure with their putative K+ channel ancestors.
Other ligand gated ion channels include P2X (ATP) recep-
tors and amiloride-sensitive sodium channels, typified by the
H+-activated ASIC channels and nematode degenerins
(DEG). While these channel types share no recognizable
amino acid homology, both are believed to have two trans-
membrane domains and function as trimers [33, 34]. This
similarity implies a common evolutionary origin, but further
ties have not yet been found. The structural diversity of
ligand-gated ion channels suggests that ligand-gating arose
de novo many times in the evolutionary history of ion chan-
nels.
Mechanosensory ion channels form the basis of hearing
and touch in mammals. Structurally distinct mechanosensory
channels have been identified in bacteria (typified by MscL)
[35] and metazoans (the ASIC/DEG family [36-39] and cer-
tain TRP channels [40], reviewed in [41]), but the molecular
identities of specific mammalian touch receptors, including
the hearing transduction channels, remains poorly under-
stood. Thus the true molecular diversity of mechanosensory
ion channels probably has not yet been fully established. It is
not clear whether all mammalian mechanosensory channels
will turn out to be known members of the ASIC/DEG and
TRP gene families, or whether some may belong to as yet
unidentified channel families. Regardless, touch sensitivity,
like ligand gating appears to have arisen multiple times in
ion channel evolution.
A fourth ion channel gating mechanism that has recently
been characterized is temperature-dependent gating. TRP
channels from the TRPV, TRPM and TRPA families have
been implicated in temperature-sensitive gating and given
the moniker ThermoTRPS (reviewed in [42]). The mecha-
nism of temperature gating and structures involved have not
been definitively characterized; both direct and allosteric
effects on voltage-dependent gating have been proposed [43,
44]. All thermosensitive ion channels identified to date are
TRP channels, but thermosensitivity cannot be predicted
from phylogeny since ThermoTRPs are found on three sepa-
rate branches of the TRP family tree. Each branch also con-
tains channels that do not appear thermosensitive, complicat-
ing the identification of candidate thermosensor sequence
motifs.
A RETROSPECTIVE HISTORY OF ION CHANNEL
GENE DISCOVERY
It is worth examining the history of ion channel gene
discovery because understanding how ion channel genes
were identified in the past shows us how mammalian ion
channel genes might be found in the future. Identification of
channel sequences in genome databases is entirely dependent
upon previous characterization of channel coding sequences
in the wet lab. The well for homology-based identification of
human ion channel genes with current channel sequences
appears to have run dry, yet there still may be a few uniden-
tified ion channels hiding in the genome sequence. However,
the picture of ion channel diversity presented here is likely to
be substantially complete, based on what we know about the
human genome sequence.
The history of ion channel gene discovery can essentially
be divided into three eras: 1) “Brute-force” cloning in the
absence of sequence information, 2) Cloning by sequence
homology and 3) genome database mining, or cloning by
sequence homology in silico. The brute-force era yielded the
initial ion channel sequences in the 1980s and early 1990s
that served to enable future eras of homology-based gene
identification. Methods such as biochemical purification of
peptide fragments (with high affinity ligands), expression
cloning and genetics were at one time the standard tech-
niques used to identify channel genes. Shosaku Numa pio-
neered ion channel cloning through application of biochemi-
cal approaches. Among his achievements, he led efforts to
identify the first nicotinic acetylcholine receptors [45, 46],
the first voltage-gated calcium channels [47], and the first
voltage-gated sodium channels [48], essentially launching
ion channel biology into the molecular age. Another notable
success of the early era of channel cloning was the identifi-
cation of the first K+ channel through gene walking to the
Drosophila Shaker locus [49-51]. Shaker had previously
been shown to have abnormal action potentials and lack a
voltage-gated K+ current in flight muscle [52, 53]. Brute
force approaches took a back seat to homology-based clon-
ing during the 1990s in terms of the numbers of genes identi-
fied, but continued to reveal novel classes of ion channels. In
fact, the most recently discovered channel family, ORAI (or
CRACM), was first identified in Drosophila through a func-
tional genomic screen for store operated calcium channels
[54]. A mammalian ortholog of this channel underlies the
CRAC current [55-57] which plays a key role in lymphocyte
proliferation. The CRAC channel has been one of the most
sought channel targets; searches over the last decade had
centered on various TRP family candidates. Unfortunately,
ORAI1 expression is near ubiquitous, suggesting that spe-

6 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
cific targeting of lymphocyte CRAC channels may be chal-
lenging. All told, the original members of 30 of the 45 func-
tionally distinct ion channel gene families that we know to
be widely conserved among metazoans were identified using
“brute force” approaches. However, only about 15% of the
mammalian ion channel genes have been identified using
such techniques, since model organisms were often used for
initial studies.
The observation that channels often come in multigene
families launched the second era of cloning in which re-
search groups raced to find additional gene family members
using the founder sequences as low stringency hybridization
probes or guides for the design of degenerate PCR cloning
strategies. For instance, the Shaker gene served as a hybridi-
zation probe for the identification of three novel voltage-
gated K+ channels in Drosophila [58] and numerous mam-
malian homologs. Each new sequence refined cloning strate-
gies by highlighting regions of conservation. This was par-
ticularly important to the development of effective degener-
ate PCR strategies, which require only two short regions (4-7
amino acid residues) of high conservation for cloning of
novel sequences. Homology cloning was most effective for
identifying novel members of previously identified gene
families; nevertheless 9 new metazoan channel families were
first identified based on sequence similarity to other gene
families. Almost half (112) of the mammalian ion channel
gene set was first identified by wet lab homology cloning.
The final major era of gene discovery has been enabled
by genome sequencing. Sequences can now be “homology
cloned” in silico from vast databases of expressed sequence
tags (ESTs) and genome sequence using known channel se-
quences and a variety of sequence comparison algorithms,
most notably variations of BLAST [59]. Sequence compari-
son algorithms have at times proved more sensitive then wet
lab approaches for identifying distant members of channel
gene families. However, we now know that most metazoan
ion channel families had already been identified in the wet
lab. Thus only 6 new metazoan ion channel families (CaV3,
KCa2, KCa4, K2P, TPC and VGCNL1, or NALCN) were first
identified using in silico searches [60-66]. Nevertheless, al-
most one third (84) of the mammalian ion channel genes
were identified with database searches. In silico cloning has
been one of the most fruitful methods for identifying human
channel orthologs.
Most researchers in the channel field (especially those in
the pharmaceutical industry), expected to find hundreds of
additional channel genes as the human genome unfolded.
Thus the relatively small number of ion human channel
genes may at first seem somewhat surprising. During the
early days of genome sequencing it was generally assumed
that the scientific community had so far identified only a
small fraction of human genes; estimates of the number of
human genes ranged as high as 50,000 to 120,000 [67, 68].
The expectation was built primarily on two assumptions: 1)
humans are much more complex than simple invertebrates
such as the worm and fly, and this complexity must be re-
flected in numbers of genes, and 2) early analysis of ESTs
showed very high numbers of unique clusters. More cautious
estimates still predicted 30,000 to 40,000 human genes [69].
However, the number of unique EST clusters steadily
dropped as new ESTs were sequenced and mapped to the
genome, and a large number of ESTs appear to be spurious
[70]. Through comparative analysis of mammalian genomes,
Clamp et al. [70] now suggest a tight estimate of ~20,000
human protein coding genes. Even though channel numbers
are lower than original expectations, pore-forming channel
subunits still represent more than 1% of the human genome.
Only 168 human-specific genes (relative to mouse and dog)
were identified in the Clamp study. In this context the obser-
vation that few ion channel genes differ between mouse and
human is not surprising.
Despite the fact that the human genome has not yielded
the vast numbers of ion channel genes, genomics is having a
tremendous impact on ion channel drug discovery. The se-
cret is that genomes have been more valuable for structural
characterization and target validation than for identification
of new human channels. This is a key point because vali-
dated drug targets are in much shorter supply than poten-
tially interesting sequences. Genome databases importantly
enabled the identification of a wide variety of ion channels in
Prokaryotes (reviewed in [71]). We now know that most of
the major structural classes of ion channels, including those
thought to be specifically tailored to nervous system func-
tion, had their origins in prokaryotes. Prokaryotic channels
have been more amenable to crystallization and served to
bring ion channels into the age of structural biology [4, 5, 22,
32, 72, 73]. Structures and structural models of mammalian
ion channels will have an enormous impact on the future of
ion channel drug discovery because of the information they
will provide on gating mechanisms and potential drug bind-
ing sites. Genome sequences from numerous model organ-
isms now give us a high resolution picture of how mammal-
ian ion channel genes and gene families evolved. This in-
formation can be used to inform choice of model organism
and provides an under-exploited resource for structure func-
tion studies. Comparison of gene sequence and function
across distant species aids in the rapid identification of func-
tionally important conserved residues. For instance compre-
hensive comparison of KV family sequences revealed that a
phenyalanine residue in the S2 TM domain is the most con-
served residue in KV channels outside the pore [16]. This
residue is proposed to form a hydrophobic plug that sepa-
rates intracellular and extracellular aqueous crevices in the
voltage sensor [16]. Combining comprehensive sequence
comparisons with pharmacology will undoubtedly help iden-
tify key residues in drug binding pockets. Furthermore, de-
tailed genomic maps (including the human HapMap [74]) are
revolutionizing the identification of disease genes and thus
provide one of the best current sources of target validation
[75]. The number of identified human channelopathies, has
been increasing rapidly in recent years and is likely to do so
for quite some time. Personal genomic information promises
safer, more effective targeting of drugs, provided that devel-
opment costs are not prohibitive.
PROSPECTS FOR IDENTIFICATION OF ADDI-
TIONAL ION CHANNEL GENES
It is almost certain that all human ion channel genes in
identified channel families have now been found because the
current human genome sequence is virtually complete (at
least with respect to gene-bearing regions). However, there is
still a possibility that new channel families will continue to
be identified. The history of ion channel gene discovery

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 7
shows that “brute force” is usually needed to identify new
structures. ORAI channels share virtually no homology with
previously known calcium channels; subunits appear to have
4 transmembrane domains and may function as tetramers
[76]. ORAI was not identified as a channel with sequence
searches despite the fact that its sequence sat in public data-
bases for years. Identification required a non-biased func-
tional approach. There is every reason to expect that other
new channel families might be found using traditional “brute
force” methods. The sequence feature that we know novel
channel genes will contain are TM domains. There are still
~200 predicted human TM proteins with little to no homol-
ogy to genes of known function in the REFSEQ database
[77]. Any additional ion channel genes are likely to come
from this pool of genes, but their identification will have to
await functional characterization. A direct strategy of func-
tionally expressing these proteins could be one of the fastest
ways to proceed.
CHANNELS ARE OLD, METAZOAN CHANNELS
ARE UNIQUE
One of the great surprises of the genome sequencing era
has been the finding that many of the gene families that we
specifically associate with locomotion, multicellular com-
munication or the metazoan nervous system actually evolved
in prokaryotes. The superfamily of voltage-gated cation
channels, which includes the majority of mammalian ion
channels, not only first appeared in prokaryotes, but diversi-
fied into several structural classes that are still found in
metazoans. Inward rectifiers, basic voltage-gated K+ chan-
nels, Ca2+-activated K+ channels, cyclic nucleotide–gated
channels and a tetrameric (single homology domain) volt-
age-gated Na+ channel are all present in various prokaryotic
species [71]. Interestingly, the cyclic-nucleotide gated chan-
nel is K+-selective and lacks the characteristic C-linker gat-
ing domain that connects the channel pore to the CNBD in
the metazoan channel family [72]. Thus the mechanism of
activation by cyclic nucleotides may have evolved substan-
tially in the metazoan lineage. Homologs of metazoan Cl-
channels are also present and have provided the first Cl-
channel structures [5]. Furthermore, a prokaryotic K+-
selective glutamate receptor and a prokaryotic acid-sensitive
C-loop receptor point to prokaryotic origins of the major
ionotropic neurotransmitter receptors [30-32]. So far, of the
major independently evolved structural classes of metazoan
ion channels, only amiloride-sensitive sodium channels, P2X
receptors, ORAI channels have not been identified in pro-
karyotes. Prokaryotic channels, however, do not closely re-
semble their metazoan cousins and systematically lack do-
mains that specify metazoan gene families. For instance
KvAP is clearly not a Shaker-like channel; it has no T1 do-
main and shares a similarly low level of sequence conserva-
tion with all metazoan Kv family channels.
There has been substantial diversification of ancient pro-
karyotic channel structures within single -celled eukaryotes
to produce new channel classes. Psuedotetrameric Na+ and
Ca2+ channels with four homology domains first appear in
single cell eukaryotes (there is one gene in yeast), and were
likely produced by 2 rounds of intragenic duplication of an
ancestral Na+ channel inherited from prokaryotes. A single
TRP-like channel is present in yeast [78], but it does not

clearly group with any metazoan TRP channel family. Choan-
oflagelletes, which are the closest known single cell ancestors
of metazoans [79], show the beginnings of diversification of
channel sequences into the recognizable metazoan gene fami-
lies. Examination of the draft genome of the choanoflagellete
Monosiga brevicollis [79] shows that it has a recognizable
NaV family sodium channel, a CaV1 (L-type) family calcium
channel and a homolog of the vertebrate two-domain Ca2+
channel TPC1. Two-pore Ca2+channels are also found in
plants [80], although it is not yet clear if they originate from
the same duplication event. The apparent ancient origin of
TPC1 raises the question of whether it is a relic of the first
intragenic duplication of the ancestral monomeric Na+ (or
Ca2+) channel. Other characteristically metazoan channels that
are found in Monosiga include all three subfamilies of CLC
channels found in metazoans, a cyclic nucleotide gated cation
channel with the C-linker motif that segragates with the meta-
zoan CNG
subfamily, a P2X receptor, 2 TRPA channels, 2
TRPM channels, 2 TRPML channels, a KCa2 channel (Na
+-
activated K+ channel), a KIR channel, 2 K2P channels and a KV
channel. A single two-pore K+ channel has been identified in
yeast [62], but it represents the fusion of a 6TM channel with a
2 TM channel, and thus probably does not share a common
origin with metazoan K2P channels. In contrast, both choan-
oflagellete K2P channels and plant two-pore K
+ channels [80]
share the 2TM x2 motif of metazoan K2P channels. The
Monosiga Kv channel is specifically similar to Shaker-like Kv
channels (Kv1-4), but still lacks the characteristic N-terminal
tetramerization domain. It is interesting that Monosiga has a
KCa2 channel (Na
+-activated) but apparently no KCa1 channel
(Ca2+-activated). Thus KCa1 may have evolved from KCa2 in
metazoans. This implies that the high-sensitivity Ca2+-gating
of metazoan KCa1 channels has a new origin and could be
mechanistically distinct from gating RCK-dependent gating of
the ancestral prokaryotic MthK channel by millimolar Ca2+.
The presence of only 15 recognizably metazoan channel fami-
lies in choanoflagelletes implies that the majority of channel
gene families conserved in higher metazoans are metazoan-
specific.
CONSERVATION OF CHANNEL FAMILIES (BUT
NOT GENES) THROUGHOUT METAZOAN EVOLU-
TION
We conducted a detailed analysis of the channel genes in
the human genome and the genomes of a diverse set of
metazoans to explore the molecular diversity and evolution-
ary origins of the human ion channel set. A summary of the
ion channel genes found in each genome is given in Table 1.
The human, mouse, tunicate, fly and nematode channel sets
have been previously reported, and the numbers of genes we
find are in good agreement [1, 81-84]. Almost all the genes
in these latter species are well annotated in public genome
databases. In addition, Yu and Catterall [1] provide compre-
hensive tables of sequence information for voltage-gated
cation channels from human, nematode and fly. Full identifi-
cation of channel genes in the genomes of pufferfish [85],
chicken [86], mosquito [87] and sea anemone [88] required
BLAST searches due to incomplete annotation of the ge-
nome drafts. BLAST searches of all genomes with multiple
members of all channel families were conducted to ensure
that all channel genes had been identified. A phylogeny of

8 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
Table 1. Summary of the Number of Genes Encoding Plasma Membrane Ion Channels in Phylogenetically Diverse Metazoan
Species

Species Hs Mm Gg Fr Ci Dm Ag Ce Nv
Total Channel Genes 235 231 211 311 112 132 147 227 219
Voltage-Gated Cation Channel Superfamily
K+ Channels
KV1-6,8,9 27 27 24 42 6 5 6 11 44
KV7 5 5 5 8 2 1 2 3 1
KV10-12 8 8 6 8 6 3 3 2 5
KIR 15 15 14 23 4 3 6 3 4
K2P 15 15 12 20 5 11 10 44 29
KCa1,4 4 4 4 5 2 2 2 2 2
KCa2 4 4 3 6 1 1 1 4 1
Non-Selective K Channels
HCN 4 4 3 7 3 1 3 0 2
CNG 6 6 6 10 4 4 4 6 1
TRP Channels
TRPA 1 1 1 1 6 4 8 2 6
TRPN 0 0 0 0 1 1 1 1 2
TRPML 3 3 3 4 1 1 1 1 2
TRPP 3 3 3 2 1 1 0 1 1
TRPC 6 7 6 8 9 3 3 3 2
TRPM 8 8 7 8 2 1 1 4 3
TRPV 6 6 5 4 2 2 2 5 2
4 Domain Channels
CaV1-3 10 10 10 17 3 3 4 3 8
NALCN 1 1 1 1 1 1 1 2 1
NaV 10 10 9 9 4 2 2 0 5
Ca2P (TPC) 2 2 2 2 2 0 0 0 1
Misc. VG Cation Channels
CATSPER 4 4 1 0 3 0 0 0 0
Hv 









Other Plasma Membrane Ion Channel Families
Glutamate Receptors 18 18 18 32 9 29 32 13 16
C-Loop Receptors
CHRN/HTR3 21 18 17 39 14 10 14 53
GABR/GLR 24 24 24 29 8 12 11 32
48
ZAC 1 0 0 0 0 0 0 0 0
Other Cation Channels


ORAI 3 

4 1 1 1 1 1
ASC 9 8 7 5 6 27 26 25 25
P2XR 7 7 7 7 0 0 0 0 1
Chloride Channels (CLC) 9 9 9 9 5 3 3 6 6
Gene families are given in the left hand column and grouped by structural similarity. The “Non-selective K+ Channel” families HCN and CNG belong to the EAG superfamily that
includes KV10-12 K
+ channels, but are not K+-selective themselves. Species key: Hs, Homo sapiens; Mm, Mus musculus; Gg, Gallus gallus (chicken); Fr, Fugu rubripes (puffer
fish); Ci, Ciona intestinalis (tunicate); Dm, Drosophila melanogaster; Ag, Anopheles gambiae (mosquito); Ce, Caenorhabditis elegans (nematode); Nv, Nematostella vectensis (sea
anemone). Unexpectedly large expansions of individual gene families in various invertebrates are indicated with italics and underlining. Channel genes from human, mouse, Droso-
phila and nematode are well annotated in genome databases and could largely be identified with text queries. Extensive TBLASTN searches of genomic sequence with multiple
query sequences failed to reveal additional gene family members. A similar TBLASTN strategy was used to identify all channel genes encoded in the pufferfish, tunicate, mosquito
and sea anemone genomes. Full coding regions were manually assembled when necessary from the genome drafts based on homology and predicted intron/exon boundaries. The
nucleotide sequence of each identified gene was then queried using BLASTN against the appropriate genome draft to make sure it represented a truly unique gene sequence. Identi-
fied channel protein sequences were queried against mammalian refseq set using TBLASTN to double check family classification. A high degree of accuracy for this methodology in
assembling channel sets is suggested by the fact that gene predictions for Ciona are virtually identical to those published by Okamura et al. [84], despite the fact that they were inten-
tionally assembled independently.

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 9
metazoans depicting the relationships of the species examined
here is shown in Fig. (2). The phylogeny is based on analysis
of the evolutionary conservation of single copy genes [85] and
represents the consensus view of metazoan evolution. The
species examined here represent the two major clades of mod-
ern bilateral metazoans, protostomes (nematode, fly and mos-
quito) and deuterostomes (Ciona and the vertebrates), and the
ancient, radially symmetric Cnidaria (Nematostella). Cnidari-
ans, which include jellyfish, hydra, sea anemones, coral and
comb jellies split from Bilateria very early, probably soon
after the evolution of the first neural networks. It is immedi-
ately apparent that the vast majority of mammalian channel
families are present in all metazoan genomes, including
Nematostella. Thus most metazoan channel families evolved
in basal metazoans in a time spanning from the advent of
multi-cellularity through the establishment of the nervous sys-
tem. An evolutionary tree of the voltage-gated cation channel
superfamily predicted from genome comparisons is presented
in Fig. (3). There is a slow accumulation of channel subunit
structures in prokaryotes and single-cell eukaryotes followed
by a rapid diversification in basal metazoans. The high con-
servation of channel families throughout metazoan evolution
points to a functionally non-redundant core set of ion channel
genes. The early emergence of this set suggests that most
originally evolved in response to specific needs in neuronal
signaling. There has been almost no innovation in channel
subunit structure since the split of Cnidaria and Bilateria, de-
spite considerable evolution of the anatomical complexity of
physiological systems. 45 classes of ion channel widely con-
served in Bilateral metazoans are listed in Table 2. 42 of these
are found in Nematostella.
The human channel set (as defined in this review) con-
sists of 235 genes. It has often been suggested that K+ chan-
nel genes are the most diverse and numerous mammalian ion
channel genes, and the human genome does indeed have 78
genes for K+ channel pore-forming subunits. All known
mammalian K+ channels derive from the voltage-gated
cation channel superfamily lineage and they are by far the
most abundant members of this superfamily. However, when
other channel families are considered as well, Na+-selective,
Ca2+-selective and non-selective cation channels together
outnumber K+ channels in all genomes (124 to 78 in hu-
mans). The higher diversity of these channels probably has
its origins in selective pressure to couple cellular excitation
to a broad array of stimuli. Cl- channel diversity is much
more limited with only 33 human genes in two major fami-
lies (CLC and the GABA-A/Glycine receptors of the C-loop
family). Since only 5/9 members of the mammalian CLC
family may form plasma-membrane Cl- channels, the num-
ber may in fact only be 28 genes. Cl- channel diversity ap-
pears to be low in all metazoan genomes. One can speculate
that the relatively shallow transmembrane concentration gra-
dients for Cl- ions found in biological systems reduce the
usefulness of Cl- signaling for many applications.
Comparison of gene numbers between species shows that
channel gene numbers also do not correlate with anatomic or
nervous system complexity. The numbers of channels in
mammals and simple metazoans such as nematodes and sea
anemones is surprisingly similar, and the number of channels
ranged widely from a low of 112 in tunicate to 311 in fish
(both deuterostomes). Most gene families in invertebrates are


















Fig. (2). Evolutionary tree of metazoans depicting the phylogenetic relationships of metazoan species included in genome-wide analyses of
ion channels. Choanoflagelletes represent the most recent single-celled ancestor of the Metazoa. Molecular analysis indicates that cnidarians
split from Bilateria very early in metazoan evolution. The last common ancestor of cnidarians and bilateral metazoans may have been one of
the first metazoans with a true nervous system. Chordates and ecdysozoans represent the two major clades of bilateral metaozans in our
analysis: the deuterostomes and the protosostomes (respectively). Tunicates represent chordate ancestors of the vertebrates.

10 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.


































Fig. (3). Evolutionary relationships and origins of the conserved Metazoan families of the voltage-gated cation channel superfamily. Back-
ground colors indicate the organisms in which evolutionary events occurred (labels at right margin); branch lengths do not reflect time for
display purposes. Branches are labeled by gene family at the bottom and branch color indicates K+ selectivity (red), Na+ and/or Ca2+ perme-
able tetrameric channels (blue), dimeric channels (X2 homology domains) (grey) and monomeric channels (X4 homology domains) (black).
Note that Na+/Ca2+-selectivity appears to have arisen 4 times during the evolution of the gene family, while K+-selectivity arose only once.
Separate origins for Na+ permeability in CNG and HCN channels are supported by the observation that HCN channels do not share a deletion
of pore residues that eliminates K+-selectivity in all CNG channels. The inclusion of ionotropic glutamate receptors (left group) in the tree is
based on the theory that they originated from an inversion of the KIR motif in prokaryotes. The ? splitting the Delta glutamate receptor line
indicates that in is not clear which glutamate receptor family gave rise this newer family. Major structural innovations are depicted with
subunit cartoons and arrows. The pore domain is shown in red and the VSD domain is shown in green. The original prokaryotic KIR motif is
faithfully preserved in only a single branch (star). Intragenic duplications of motifs are indicated with brackets and numbers. Other appended
domains include the RCK domains of KCa1,4,5 channels (only one of two domains is depicted), the calmodulin-binding domain of KCa2
channels (CAM), a cyclic nucleotide-binding domain (CNG), the PAS domain of EAG family channels, the tetramerization domain of
Shaker-like KV channels (T1) and ankryin repeats (of variable number) in TRP channel families. The TRPN family and a second invertebrate
TRPV family are not included in the tree because they have been lost in mammals.

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 11

Table 2. Metazoan Conserved Ion Channel Families. 45 Families of Ion Channels Genes are Widely Conserved in Metazoans; 43
are Present in the Humans. Proposed Conserved Functions of Each Family are Listed

Channel Family Human Genes Conserved Function Earliest Appearance
TRPA 1 TRPA1 Sensory TRP Choanoflagellete
TRPN 0 Sensory TRP Sea Anemone
TRPML 3 TRPML1-3 Endosomal TRPs Choanoflagellete
TRPP 3 TRPP1-3 Mechano TRP Sea Anemone
TRPC 6 TRPC1, 3-7 Gq-activated TRP Sea Anemone
TRPM 8 TRPM1-8 various Choanoflagellete
TRPV 6 TRPV1-6 thermoTRP Sea Anemone
TRPVinv 0 Sensory TRP Choanoflagellete
TPC 2 TPC1,2 Ca2+ channel? Choanoflagellete
ORAI 3 ORAI1-3 store-operated Ca2+ channel Sea Anemone
CaV1 4 CaV1.1-1.4 L-type Ca2+ channel Choanoflagellete
CaV2 3 CaV2.1-2.3 P/Q/N/R-type Ca2+ channel Sea Anemone
CaV3 3 CaV3.1-3.3 T-type Ca2+ channel Sea Anemone
NALCN 1 NALCN Na+ leak channel Sea Anemone
NaV1 10 NaV1.1-1.9, NaX Voltage-gated Na+ channels Choanoflagellete
KV1 8 KV1.1-1.8 KV, fast activation Sea Anemone
KV2 2 KV2.1,2.2 classic delayed-rectifier Sea Anemone
KV3 4 KV3.1-3.4 high threshold delayed-rectifier Sea Anemone
KV4 4 KV4.1-4.3 classic neuronal A-current Sea Anemone
KV7 (1) 1 KV7.1 Cardiac IKS Sea Anemone
KV7 (2) 4 KV7.2-7.5 Neuronal M-current Sea Anemone
KV10 4 KV10.1,10.2 sub-threshold K+ current Sea Anemone
KV11 4 KV11.1-11.3 Cardiac IKR Sea Anemone
KV12 4 KV12.1-12.3 sub-threshold K+ current Sea Anemone
HCN 4 HCN1-4 Ih Sea Anemone
CNG
4 CNG
1-4 cAMP-dependent cation channel Choanoflagellete
CNG 2 CNGb1,CNGb3 cAMP-dependent cation channel Nematode
KCa2 4 KCa2.1-2.3, KCa3.1 SK,IK channels Sea Anemone
KCa1 2 KCa1.1, KCa5.1 BK channels Sea Anemone
KCa4 2 KCa4.1, KCa4.2 Na+-activated K+ channels Choanoflagellete
KIR 15 KIR1.1, 2.1-.4,3.1-.4,4.1-.2,5.1,6.1-.2,7.1 Inward rectifier K+ channels Choanoflagellete
K2P 15 K2P1-7, K2P9-10, K2P12-13, K2P15-18 K+ leak channels Choanoflagellete
GLURA 4 GRIA1-4 AMPA receptors Sea Anemone
GLURK 5 GRIA1-4 Kainate receptors Sea Anemone
GLURNR1 1 GRIN1 NMDA receptor Sea Anemone
GLURNR2 6 GRIN2A-D, GRIN3A,B NMDA receptor Sea Anemone
P2X 7 P2X1-7 Ionotropic ATP receptors Choanoflagellete
ASIC/DEG 9 ACCN1-5, SCNN1A,B,D,G mechanosensitive sodium channels? Sea Anemone
GABA-A
10 GABRA1-6,G1-3,E GABA-A receptors Sea Anemone
GABA-A 4 GABRB1-3,Q GABA-A receptors Sea Anemone
NAchR 13 CHRNA1-6,B1-4,D,E,G typical nicotinic receptors Sea Anemone
NAchR
7-like 1 CHRNA7 Ca2+-selective nicotinic receptor Sea Anemone
CLC
4 CLC1,2, CLCK1,2 Chloride channels Choanoflagellete
CLC 3 CLC3-5 organelle Cl-/H+ transporters Choanoflagellete
CLC 2 CLC6,7 organelle Cl-/H+ transporters Choanoflagellete
The ancestral non-redundant functional properties of channel families listed in bold italics are poorly understood. The earliest appearance of the family in the genomes examined here
is also listed. Metazoan-like KIR, K2P, and CLC channels are found in other eukaryotes, suggesting earlier origins. HUGO nomenclature is used starting with Glutamate receptors
since IUPHAR names for these and subsequent channels have not yet been released.

12 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
significantly smaller than in vertebrates; many are repre-
sented by a single gene. However, nematodes, sea anemone
and insects (to a lesser extent) have large, phyla-specific
expansions in some gene families. For example, nematodes
have 85 C-loop receptors and 44 K2P channels, while Nema-
tostella has 44 Shaker-like Kv1-4 channels. Despite the fact
that all metazoans share a common set of ion channel types,
many of the extant genes must have separate evolutionary
histories in each major phyla.
THE HUMAN/MAMMALIAN CHANNEL SET
Comparative analysis of the human and mouse genomes
using either sequence homology or synteny shows a virtually
identical set of channel genes. The human channel set is
therefore essentially equivalent to the “mammalian” channel
set. The two terms will be used interchangeably here when
discussing channel evolution. The short list of channel genes
differing between human and mouse are shown in Table 3.
Four C-loop receptor genes and one ASIC/DEG family
epithelial sodium channel are unique to humans. The
HTR3C-E receptor cluster is present in most mammalian
genomes (including opossum), suggesting that it has recently
been lost in rodents [89]. These genes appear to be primarily
expressed in the GI track and their function in vivo is un-
known [90]. The other genes are not widespread and proba-
bly derive from recent duplications. The most physiologi-
cally significant difference is likely the loss of TRPC2 func-
tion in humans. A residual TRPC2 psuedogene is present in
the human genome. TRPC2 is the transduction channel for
pheromone responses in the rodent vomeronasal organ [91].
The absence of TRPC2 (coupled with the loss of most
pheromone receptors) indicates that vomeronasal-based
pheromone signaling does not occur in humans. The loss of
TRPC2 function appears to have occurred in the common
ancestor of apes and old world monkeys [92].
ORIGINS OF HUMAN ION CHANNEL GENES
Most of the gene duplications that produced the mammal-
ian ion channel set can be readily identified by building
phylogenetic trees for each gene family from diverse species.
Comparison of tunicate and human channels reveals system-
atic duplications in the vertebrate lineage [84]. A Kv7
(KCNQ) K+ channel family tree is shown in Fig. (4) as an
example. Two ancestral branches, or clades, contain the
mammalian neuronal m-current channels Kv7.2-7.5 and the
mammalian Kv7.1 channel (KVLQT1), respectively. M-
channels are important for regulating basal neuronal excitabil-
ity and have been linked to a number of neurological diseases
[93]. Kv7.1 has been linked to cardiac arrhythmia
and deafness [93]. A third highly-divergent branch of the Kv7
family is worm-specific and does not represent an ancestral
clade. Ancestral bilaterians likely had two Kv7 family genes,
one of which was duplicated twice in early vertebrate evolu-
tion to produce Kv7.2-7.5. The vertebrate Kv7.1, Kv7.2 and
Kv7.5 genes were duplicated again in fish producing sets of
Fugu co-orthologues for each mammalian gene. An additional
duplication in worms gave rise to the novel Kv7 gene. The
Kv7 tree shows the trends in metazoan ion channel evolution
that hold true for most families. The vast majority of mammal-
ian channel genes were produced by duplications of ancestral
deuterostome genes (represented by tunicates) early in the
vertebrate lineage. Most of these genes are therefore shared
with chicken and puffer fish. The duplication events producing
132 mammalian voltage-gated cation channels and 18 gluta-
mate receptors are summarized graphically in Fig. (5). Only
the final duplications that produced the extant genes are
shown: 88/101 occur in the common vertebrate lineage. Only
6 duplications are mammalian specific, and 7 occurred earlier
in metazoan evolution. This pervasive duplication is consistent
with the 2R hypothesis that two rounds of large-scale, perhaps
genome level, duplication occurred in the common vertebrate
lineage [94-98]. The hypothesis of large-scale duplication
includes an additional duplication during the evolution of
bony fish. We find that the pufferfish Fugu has by far the
highest number of channel genes and that almost half of the
channel genes shared by vertebrates have been duplicated
again in Fugu. The result of the fish-specific gene duplications
is that less than half of the mammalian genes have true 1:1
orthologs in fish. Thus there is the potential for substantial,
systematic divergence in expression pattern and function be-
tween mammal and fish channel genes. The 2R hypothesis
also predicts that few of the channel genes duplicated in the
vertebrate or fish lineages will be physically linked in the ge-
nome. This is indeed the case. Systematic physical dispersion
of the new vertebrate genes through chromosome rearrange-
ments seems unlikely since syntenic blocks are very large in
vertebrates and many can be traced all the way back to Nema-
tostella [88].
IMPLICATIONS OF GENE DUPLICATIONS FOR
USE OF MODEL ORGANISMS
It is not surprising that genome level analysis suggests
that rodents are excellent model organisms for ion channel
studies. While there are certainly documented cases of the
functional divergence of individual human and rodent ion
channel genes, the vast majority of channel genes appear
substantially identical in both sequence and expression


Table 3. Differences Between Human and Mouse Ion Channel Sets

Unique Human Gene Family Basic Function Why Not in Mouse?
ZAC C-loop Zinc-activated cation conductance Recent duplication
SCNN1D ASIC/DEG Epithelial sodium channel Recent duplication
HTR3C-E C-loop Ionotropic 5-HT receptors Loss in rodents
Unique Mouse Gene Family Function Why Not in Human?
TRPC2 TRPC Pheromone transduction channel Pseudogene

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 13

pattern. Genome analysis suggests that no other model spe-
cies can compare. One of the major surprises of genome
analysis is that zebrafish, which shares the large-scale dupli-
cations found in Fugu, could be a poor model for the func-
tion of many specific mammalian genes. The widespread
presence of co-orthologs of individual mammalian channels
in fish will have allowed an opportunity for substantial di-
vergence in function, especially with respect to expression
pattern. Therefore, when extrapolating results from zebrafish
studies to functions of mammalian genes, it would be wise to
closely examine the relationship of the gene in question to
mammalian orthologs. Nevertheless, even in cases where

zebrafish has undergone duplication and divergence, it cer-
tainly remains an excellent model when probing questions
more generally related to gene family or subfamily function.
Serious examination of the function of individual human ion
channels is simply not possible with invertebrate models
given the almost complete absence of 1:1 ortholog pairs.
KCa1, NALCN and the NR1 subunit of the NMDA receptor
are the only human ion channels that share 1:1 orthologous
relationships with their Drosophila counterparts. However,
invertebrates such as Drosophila and nematode may often be
the best models for questions directed towards understanding
conserved functions of the ancestral metazoan channel




























Fig. (4). Metazoan Kv7 (M-channel) family tree. Phylogenetic tree of metazoan Kv7 family constructed using the minimum evolution algo-
rithm implemented in MEGA4 [145]. Branch lengths show degree of divergence and numbers indicate bootstrap support (100 repetitions) for
key branch points. Species are labeled by prefix and color: human, mouse and rat, red, h,m,r; pufferfish, dark red, f; Ciona (tunicate), purple,
c; Drosophila and Anopheles (mosquito), green, d,a; nematodes C. elegans and C. briggsae, blue, ce,cb; and the sea anemone Nematostella,
light blue, nv. Gene duplication events were inferred using Forester [146] and are color coded by phylogenetic origin: red, basal vertebrates;
dark red outline, bony fish; blue, nematode; ancestral bilaterians, red with blue outline. Note that the most divergent branch is not ancestral
but a product of a nematode-specific duplication. Qualitatively identical results were obtained using neighbor joining or maximum parsi-
mony algorithms to build the tree.

14 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.







































Fig. (5). Phylogeny of the mammalian members of the voltage-gated ion channel superfamily predicted from detailed phylogenetic analysis
of individual gene families including all members from human, mouse, rat, pufferfish, chicken, tunicate, fly, mosquito and nematodes. Back-
ground colors indicate organisms in which duplication events occurred (labels at top). Question marks indicate uncertain gene family origins.
Key gene families are labeled at the right margin and bracketed with black bars. Duplication events were inferred using Forester [146]; only
the duplication events that produced extant mammalian channel genes are shown (red circles). Note that almost all occurred in vertebrate
evolution. We found that prediction of mammalian gene duplications was more robust when building phylogenies from individual channel
families rather than the whole superfamily at once because we could use much more sequence information from each gene.

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 15
families. For example, PLC-dependent activation of TRPC
channels, which is widespread in vertebrates, was first dis-
covered in Drosophila [99]. Many of channel families have
undergone little duplication outside the vertebrate line and
are thus represented by a single gene in Drosophila and
nematode. Loss-of-function studies are markedly simpler for
these families in invertebrate model systems. Genetic redun-
dancy may sometimes be a restricting criterion for similar
genetic studies in mice. Invertebrate models are thus likely to
remain relevant and informative for functional characteriza-
tion of ion channels until the conserved functions of meta-
zoan channel families are fully understood.
EVOLUTIONARY CONSERVATION SUGGESTS
NOVEL TARGETS
One of the basic assumptions of genetics is that sequence
conservation implies functional conservation and that gene
preservation over time implies non-redundant function.
Therefore, the 45 ancestral metazoan ion channel clades
identified in genome comparisons probably represent ion
channels classes that are fundamentally important to meta-
zoan physiology. We can predict a lot about the basic bio-
physical properties of channel genes simply from examina-
tion of sequence, and most conserved channel types have
been expressed in heterologous systems. Surprisingly, how-
ever, the non-redundant conserved functions of many of
these genes have not yet been identified. Understanding
these functional roles is absolutely critical for serious as-
sessment of target potential. The value of such work is per-
haps best illustrated by the recent functional characterization
of NALCN as the principle channel mediating sodium leak
in mammalian neurons [100]. The channel has a profound
effect on neuronal excitation and thus could be a very inter-
esting target for suppression of pathophysiological hyperex-
citability in diseases such as epilepsy. High conservation of
NALCN-like gene(s) in metazoan genomes had been known
for a number of years, and Drosophila ortholog had recently
been shown to control output of circadian pacemaker neu-
rons [101, 102]. The first clues to functional importance for
the NALCN family also came from Drosophila with the ob-
servation that hypomorphic alleles produce anesthetic resis-
tance and anatomical abnormalities [102, 103]. NALCN is
represented by a single gene in all the metazoan genomes
except nematode. Both nematode paralogs have recently
been linked to endocytosis defects, although the mechanism
is not clear [104]. Nematodes conspicuously lack NaV family
sodium channel genes, but appear to have sodium-dependent
action potentials [105, 106]. It will be interesting to see if
NALCN paralogs are involved. Another potentially highly
interesting conserved cation channel is the putative two-pore
Ca2+ channel TPC1. Its origin predates metazoans and simi-
lar genes are even found in plants. Metazoan TPC channels
have not yet been functionally expressed, but the plant chan-
nels do appear to be Ca2+-permeable [80]. All that is cur-
rently known about the mammalian TPC1 is that it is ex-
pressed in kidney (inner medullary collecting ducts), brain,
heart, lung, liver and spleen [60, 107]. Even less is known
about a second deuterostome paralog present in genome da-
tabases.
Other prominent ancestral channel clades that are “under-
characterized” with respect to conserved function include the
EAG family (Kv10-12) and the various TRP channel sub-
families. Kv11.1 (hERG) has a well-known role in the repo-
larization of cardiac action potentials [108, 109], but all three
EAG subfamilies appear to have evolved at a time in early
metazoans indicative of a fundamental role in neurons. In-
deed, all three families contain genes that are predominantly
expressed in the nervous system. However, their function in
neurons has largely remained a mystery. All encode K+
channels that are active in the sub-threshold voltage range
and thus have the potential to strongly influence neuronal
excitability. Kv12, or Elk, channels have the most hyperpo-
larized activation ranges of any mammalian KV channels.
Sub-threshold K+ channels are a therapeutically interesting
channel class because they hold the promise of silencing
hyperexcitability. However, these channels are structurally
diverse and many, like the EAG family members, remain
poorly characterized in neurons. Kv7 channels, the neuronal
m-channels, are a clear exception. Pathways that modulate
their activity (and neuronal excitability) are well docu-
mented: classic slow muscarinic EPSPs are mediated by Kv7
channel inhibition [110, 111]. PIP2 hydrolysis appears to be
the principle mechanism of inhibition [112]. Mutations in
Kv7.2 and Kv7.3 cause inherited epilepsy (BFNC) and Kv7
agonists have entered clinical trials as anti-epileptics (re-
viewed in [113]). No similar fundamental signaling role has
yet been identified for EAG channel subfamilies in the nerv-
ous system, despite superficial biophysical similarities, al-
though it has been suggested that Kv11 channels may con-
tribute to M-like currents in some neurons [114]. Interest-
ingly, KV10.1 (Eag1) has recently emerged as a target for
various cancers [115, 116].
TRP channels are among the hottest putative targets be-
cause of their involvement in sensory transduction and noci-
ception. ThermoTRPs that are activated by noxious tempera-
tures and compounds, principally TRPV1 and TRPA1, ap-
pear to have a central role in nociception [117-119], re-
viewed in [120]. We know a lot about the biophysics and cell
biology of a number of the 27 mammalian TRP channels
from extensive studies of these genes carried out over the
last decade. TRP channels display a broad range of cation
selectivities, current rectification, sensitivity to signaling
pathways, and activation by sensory stimuli such as touch,
temperature and chemicals. The confounding issue for un-
derstanding conserved functional roles of TRP channels is
that none of the described properties segregate by
phylogeny. For instance, thermosensation, mechanosensation
and activation by intracellular Ca2+ have all been attributed
to multiple TRP families, but only a subset of genes within
those families. It is hard to understand why 7 TRP channel
families would be conserved throughout metazoan evolution
with such substantial overlap in function. Some of the confu-
sion undoubtedly comes from a lack of understanding of
which stimuli act directly on TRP channels and which act on
upstream signaling pathways [120]. For instance, it is possi-
ble that some of the diverse stimuli that activate TRPA1
could work indirectly through increases in intracellular Ca2+
[121]. TRP channels are modulated by common signaling
pathways which probably complicates functional analysis by
heterologous overexpression. It is also entirely possible that
TRP channel conservation is more about localization and
links to cellular machinery than biophysics. Interestingly,
ankryin domains are found in TRPV, TRPC, TRPA and
TRPN channels and vary in number in a gene family specific

16 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
manner [41]. The roles of these domains have not been ex-
tensively characterized, but one can speculate that they me-
diate subfamily-specific protein-protein interactions. One
conserved function of TRPC channels appears to be activa-
tion by metabolites of DAG hydrolysis downstream of Gq-
coupled receptors [122]. Both Drosophila and mammalian
TRPC channels can be activated by Gq-coupled opsins. The
mechanism serves as the basis for Drosophila vision [122]
and may possibly be used for melanopsin-based circadian
phototransduction in mammals [123]. More family-specific
TRP functions will undoubtedly be discovered as research on
these intriguing channels continues and will provide a better
picture of which TRP channels have the best target potential.
UNTAPPED POTENTIAL FOR CHANNEL-BASED
INSECTICIDES
Development of new drugs targeting human channels
dominates efforts to identify therapeutically relevant ion
channel modulators because of the high potential for profit.
However, one could arguably have a greater impact on hu-
man health through the targeting of crop pests, insect disease
vectors and parasites. Ion channels are among the best tar-
gets for insecticides because block can quickly incapacitate
or kill by disrupting activity in the nervous system or mus-
cles [124]. Channel block is indeed the predominant strategy
used by venomous animals to capture and kill prey, as evi-
denced by the prevalence of channel toxins found in their
venoms. Resistance to current insecticides is becoming
widespread [125], and thus there is a need for a new genera-
tion of channel-targeted insecticides. Furthermore, com-
monly used pyrethroid insecticides have more toxic activity
against human sodium channels than originally thought
[126], suggesting that alternate targets might be preferred.
Comparisons of mammalian and insect channel sets reveal
significant separate gene duplication and should help in se-
lection of promising new targets, especially when combined
with the wealth of information on Drosophila channel mu-
tants that severely impair nervous system function. The sur-
prising level of sequence divergence between insect species
holds out the hope of pest-specific insecticides [83, 87]. Key
issues for the development of lead compounds necessarily
differ from those of drug discovery. For instance, cost of
synthesis and effects of environmental exposure can drive
insecticide development.
SEQUENCE CONSERVATION AND DRUG TARGET-
ING
One of the major challenges for successful exploitation
of new ion channel targets is validation. Huge gaps remain in
our understanding of how the observed molecular diversity
of ion channels corresponds to the physiological diversity of
ion channels in vivo. Mouse knockout models have been an
important source of information on channel function, and
will continue to be an important part of efforts to identify
therapeutically relevant channels. Introduction of disease
mutations into mouse orthologs has also provided experi-
mental models for the study of human channelopathies [127,
128]. Specific chemical probes are an important complement
to mouse genetic models for target validation. Use of a
chemical probe can allow for the assessment of function in
the absence of developmental changes or functional compen-
sation which can sometimes confound the use of mouse
knockouts. It is also far simpler to approach gain-of-function
validation with chemical probes. This is important since
many potential ion channel-based therapies rely on the use of
agonists. For instance, K+ channel openers are believed to
have great potential for the treatment of epilepsy. Use of
chemical probes on animal models provides a direct test of
the therapeutic potential of drugs. Since the best mammalian
model species for target validation is not always a mouse,
chemical probes are sometimes the only alternative for key
target validation experiments. A pervasive lack of highly-
specific chemical probes for most mammalian ion channels,
and even some prominent channel families (such as Kv12,
NALCN, K2P), presents a high hurdle for efficient validation
of channel targets. Complex pharmacological criteria can
often be used to identify channel families or subunits in na-
tive cells [129], but such an approach is inherently difficult
to apply across a broad range of target validation paradigms.
The vast majority of mammalian ion channels simply do not
have commercially available gene-specific small molecule
modulators.
The most specific channel probes are often peptide toxins
from the venom of spiders, scorpions, snakes and cone
snails. For instance, iberiotoxin (from a scorpion) is a high
affinity blocker of a single mammalian ion channel, the
large-conductance calcium-activated K+ encoded by KCa1
(mSlo) [130]. Venom peptides have provided numerous in-
sights into ion channel function, but have two potential
shortcomings for target validation in vivo (or as human
therapeutics). Most obviously, peptides are difficult to ad-
minister systemically. Ziconotide, a synthetic cone snail pep-
tide targeting N-type calcium channels has proven effective
for chronic pain, but its use is limited since it must be in-
jected intrathecally [131]. Second, most mammalian ion
channels simply do not have specific peptide toxin ligands.
Peptides with novel specificities continue to be found, but
there is no reason to assume that ligands for all channel types
will be found. If blocking a specific channel does not quickly
aid in the incapacitation of prey, then there is no reason to
expect to find venoms that specifically target the channel.
Another interesting strategy to achieve specific channel
block is to raise antibodies that recognize extracellularly
accessible regions of the channel pore such as the turret [4]
in voltage-gated cation channels [132]. One advantage of
targeting the turret is that it is a highly variable region. This
strategy could lead to the identification of specific blockers
for some additional channels, but antibodies suffer from the
same problems as peptide toxins when it comes to applica-
tion in vivo. Moreover, some channels have little extracellu-
lar exposure, so the technique may not be systematically
applicable.
A comprehensive effort to identify new generations of
small molecule chemical probes through HTS is clearly
needed to keep ion channel target validation efforts rolling.
The availability of sequences from all mammalian ion chan-
nels importantly provides the opportunity to systematically
assess sequence conservation on a whole genome scale. Such
information should greatly help efforts to produce selective
channel probes by highlighting the most promising protein
regions to target. As examples, the sequence conservation
patterns of mammalian K+ channels and TRP channels are
shown in Figs. (6-9). Conservation of key structural features
between members of gene subfamilies is typically around

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 17
60-70%, but varies significantly from family to family. K2P
channels and TRP channels generally are the least conserved
and KV channels are generally the most conserved. Not sur-
prisingly, the pore helices and selectivity filter are consis-
tently among the most highly conserved regions of these
channels. Voltage-sensor domain (VSD) conservation is
more variable (but generally lower), and most channel genes
display gene-specific sequence between structural domains
or at the protein extremities. While the value of extremities
for binding specificity is obvious, the likelihood of affecting
channel function by targeting these regions is less clear and
may have to be determined on a case by case basis.
K2P channels will probably pose few problems to the
generation of channel-specific probes. Mammalian K2P se-
quences can be divided into 3 evolutionarily distinct lineages
based on sequence comparison, intron/exon structure and
chromosomal location (Fig. 6). The K2P3 lineage (which
includes pH-sensitive TASK channels) is most highly con-
served, but still shows significant inter-gene variation in pore
helices. The other K2P lineages are so divergent that it is dif-
ficult to reconstruct meaningful phylogenies based on se-
quence homology alone. There is little sequence conserva-
tion between K2P lineages and no worrisome homologies to
other K+ channels. One of the main technical advantages for
targeting K2P channels is that they have large and highly
variable turret sequences. KIR channels are similarly diver-
gent from other K+ channels families, but share more homol-
ogy to each other. KIR3 (GIRK) channels are particularly
highly conserved and may present a challenge for the genera-
tion of gene-specific probes.
KV channels provide more difficulties for the generation of
specific probes because they are more highly conserved and
are more numerous (Fig. 7). Attaining highly gene-specific
probes for KV1-4 channels and KCa2 (SK) channels may re-
quire heroic efforts. Fortunately, venom toxins have proved a
good source of specific blockers for these channels and the
target validation status of most genes in these families is quite
advanced. BK-related KCa1,4,5 are highly divergent and are
not likely to display significant pharmacological cross-
reactivity. KV7 channels already have family-selective probes
such as XE991 and linopirdine, but gene-specific probes are
not yet widely available. The subfamilies of the EAG and TRP
superfamilies are for the most part quite divergent when com-
pared to shaker-like KV channels (Figs. 8,9). Pharmacological





















Fig. (6). Amino acid sequence conservation patterns in mammalian KIR and K2P channels. The predicted phylogeny is reconstructed from a
combination of minimum evolution analysis (constructed as outlined in Fig. (5)), gene duplications and (K2P only) cladistic analsysis of of
intron position and chromosomal location. Branch lengths do not reflect molecular divergence. Colors on the subunit cartoons show typical
amino acid identity shared between genes on the indicated branches. The K2P2 branch includes (in order) K2P16, K2P17, K2P5, K2P4, K2P
K2P10 and K2P2; the K2P1 branch includes K2P1, K2P7, K2P6 and K2P18; the K2P3 branch includes K2P12, K2P13, K2P3, K2P15 and K2P9. There
are distinct KIR branches for the KIR2 and KIR3 subfamilies, all other mammalian KIR channel genes group loosely and share little more se-
quence identity than general Pan-KIR background conservation. Pan-K2P levels of conservation are also shown. Cartoons with arrows indicate
the level of cross-family sequence identity shared between the K2P and KIR families.

18 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
divergence of the KV 10-12 EAG families is suggesting by the
contrast of the (unfortunately) rich pharmacology of HERG
(KV 10.1) [133] with the nondescript pharmacology of KV 12
channels [134]. Pharmacological diversity of TRP channels is
apparent from the fact that differential patterns of activation
by sensory compounds are a defining feature of sensory TRP
channels [120].
Emerging pictures of channel structures suggest that one
way to conquer selectivity issues for chemical probes may be
to exploit diverse binding pockets. The pore is generally the
most highly conserved feature of an ion channel, and thus it
may in some cases be difficult to achieve selectivity by tar-
geting the pore directly. Nevertheless, some highly-specific
venom peptides interact with poorly-conserved extracellular
turret sequences to achieve specific block of the neighboring
pore [135]. This approach may be difficult to mimic with
small molecules which are likely to be much smaller than
peptide toxins. The voltage sensor is an intriguing site of
interaction for development of chemical probes in the volt-
age-gated cation channels because it is generally less well
conserved than the pore domain and appears to have a deep
external aqueous cleft (reviewed in [136]) that could serve as
a binding pocket. However, the most conserved feature of
the voltage-sensor domain is the charge pattern, suggesting
that highly polar or charged probes may not be selective.
Hanatoxin, isolated from the Chilean rose tarantula, is the
best characterized of a growing list of tarantula toxins that
specifically block KV channels by binding to the voltage sen-
sor [137]. None of the toxins isolated so far is exclusively
selective for a single channel, but their ability to discriminate





























Fig. (7). Amino acid conservation patterns in mammalian KV1-9 and related KCa channels. Construction of predicted phylogeny is the same
as for Fig. (6); a scale is given for color coding. Conservation shared between the Shaker-like KV1-6,8,9 channels and their nearest neighbor
KV7 is shown in (A), between KCa2 and its nearest neighbors (KV1-9) in (B), and between BK family channels (KCa1,4,5) and their nearest
neighbors (KV1-9, KCa2) in (C).
T1
T1
T1
T1
T1
T1
T1
T1
CAM
CAM
RCK
Bowl
RCK
Bowl
RCK
Bowl
KV9
KV6
KV2
KV1
KV4
KV3
KV7
KCa2
KCa4
KCa1,5
Pan-KV1-6,8.9
Pan-KCa1,4,5
0 20 40 60 80 100
Percent Amino Acid Identity
A
B
C

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 19
between closely related KV channels suggests that targeting
the voltage sensor has the potential to provide a high degree
of probe specificity. One intriguing aspect of targeting the
voltage sensor is that it may be possible to achieve more
subtle modulation of channel function than through block of
the pore. For instance, Hanatoxin does not fully block KV
2.1, but instead shifts the voltage-activation curve +40mV
[137]. Channel function is substantially reduced at hyperpo-
larized voltages, but not entirely eliminated at depolarized
voltages. It may be possible to affect sub-maximal block
without careful titration of compound doses when targeting
the voltage sensor.
Other channel family-specific gating domains should also
provide a means for generating family-specific chemical
probes. Furthermore, it may also be possible to achieve sub-
tle modulation with this targeting strategy. For instance, cy-
clic nucleotides produce modest shifts in the voltage-
dependence of HCN channels [138]; thus modulators that
activate or silence the CNBD could be used to shift the acti-
vation threshold of HCN channels to reduce or enhance neu-
ronal excitability without completely removing HCN chan-
nel function. CNBDs are also found in CNG channels and
KV 10-12, but there is substantial divergence between these
channel families (Fig. 8). The KV 10-12 CNBD domains lack
residues predicted to coordinate cyclic nucleotides [139], and
thus may have novel ligand specificity. One note of caution
regarding the targeting of family specific regulatory domains
is that while the strategy avoids cross-family selectivity is-
sues, such domains are highly conserved within subfamilies.
They usually show the same level of conservation as the core
transmembrane regions (Figs. 6-9). Therefore appended gat-
ing domains provide an alternate targeting possibility but not
always a theoretical advantage for gene-specific probes.
Probes that target the interaction between ion channel
pore subunits and regulatory subunits have the potential for
specificity and a broad range of modulatory effects. Young
et al. (1998) proved such a strategy can work by using a
yeast complementation strategy to identify compounds that
disrupt interaction of the N-type Ca2+ channel with a beta
subunit to effect block of channel function [140]. Interaction
with regulatory subunits often differs between closely related
gene family members, providing an opportunity to obtain
specificity. For instance, neuronal KV 7 family channels are
highly conserved (Fig. 7), but differ in their ability to func-
tionally interact with mink/Mirp subunits, which affect their
trafficking and voltage-dependence [141]. Moreover, chan-
nel genes such as KCa1, are widely expressed but interact
with tissue-specific beta-subunits. The brain-specific beta-
subunit KCNMB4 confers resistance to iberiotoxin [142],
but might also provide a means for specifically targeting
neuronal BK channels. The somato-dendritic delayed recti-
fier K+ current in most neurons can be attributed to KV2.1,
but this channel subunit is known to form heteromeric chan-
nels in vitro with “silent” KV channels from the KV 5, KV 6,
KV 8 and KV 9 subfamilies [143]. These silent subunits are
true pore-forming subunits, but none form functional ho-
momeric channels. They appear to be the KV family equiva-
lent of CNG
subunits or the NR2 subunits of NMDA recep-




















Fig. (8). Amino acid conservation in the EAG gene superfamily. See Fig. (6) for details on phylogeny prediction and the scale for an expla-
nation of color coding. Note the relatively low level of conservation shared between the three principle families CNG, HCN and KV10-12.
Amino acid identity shared with other KV channels is below 20% across most of the channel structure (not shown).
PAS CNBD
PAS CNBD
PAS CNBD
PAS CNBD
CNBD
CNBD
CNBD
0 20 40 60 80 100
Percent Amino Acid Identity
CNGα
HCN
KV11
KV10
KV12
Pan-KV10-12
Pan-CNG/EAG Superfamily

20 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
tors. Silent KV channels have divergent sequence and are
expressed in restricted subsets of neurons. Thus they could
provide a means for targeting of delayed rectifiers in specific
neuronal populations if heteromerization indeed happens in
vivo. For instance, the delayed rectifier of mammalian rods
and cones is likely to be a heteromer of KV 2.1 and KV 8.2.
Both genes are expressed in photoreceptors and mutations in
KV 8.2 cause "cone dystrophy with supernormal rod elec-
troretinogram" [144], which is characterized by abnormal
transmission of photoreceptor responses to second order neu-
rons.
SUMMARY
Examination of the human ion channel set from an evolu-
tionary perspective provides important insights into issues
relevant to ion channel drug discovery. Human ion channel
gene families appear to have evolved very early in metazoan
evolution, but the duplications that gave rise to human ion
channel occurred predominantly in the common vertebrate
lineage. Large gaps remain in our knowledge of the non-
redundant physiological functions of conserved metazoan
channel families. Filling these gaps will provide an opportu-
nity to assess the therapeutic potential of functionally in-
triguing channels. One of the major barriers to functional
annotation of human ion channel genes remains a lack of
highly specific chemical probes. Strategies for development
of such probes can be developed based on genome-level
comparison channel sequence homology coupled with appli-
cation of the new high-throughput approaches to ion channel
analysis discussed in this special issue.





























Fig. (9). Amino acid conservation patterns in mammalian TRP channels. See Fig. (6) for details on phylogeny prediction; a color scale is
given for conservation level. (1) Conservation shared between TRPM channels and other TRP families; (2) Conservation shared between
TRPP and TRPML channels; (3) Conservation shared between TRPC, TRPA and TRPV channels. Pan-TRP conservation is not shown be-
cause it is difficult to produce meaningful amino acid alignments between distantly related TRP channels. The labeled TRPM1 branch in-
cludes TRPM1, TRPM3, TRPM6 and TRPM7. The TRPM2 branch includes TRPM2, TRPM4, TRPM5 and TRPM8.
0 20 40 60 80 100
Percent Amino Acid Identity
TRPV1-4
TRP
A
TRP
A
TRP
A
TRP
A
TRP
A
TRP
A
TRP
A
TRP
TRP
TRP
TRP
TRPV5,6
Pan-TRPV
TRPC3,6,7
TRPC1,4,5
TRPML
TRPP
TRPM2
TRPM1
Pan-TRPC
1
2
3
Pan-TRPM

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 21
REFERENCES
[1] Yu, F. H.; Catterall, W. A. Sci. STKE., 2004, (253), re15.
[2] Kostich, M.; English, J.; Madison, V.; Gheyas, F.; Wang, L.; Qiu,
P.; Greene, J.; Laz, T.M. Genome Biol., 2002, 3, (9), RE-
SEARCH0043.
[3] Manning, G.; Whyte, D.B.; Martinez, R.; Hunter, T.; Sudarsanam,
S. Science, 2002, 298, 1912-34.
[4] Doyle, D.A.; Morais Cabral, J.; Pfuetzner, R.A.; Kuo, A.; Gulbis,
J.M.; Cohen, S.L.; Chait, B. T.; MacKinnon, R. Science, 1998, 280,
69-77.
[5] Dutzler, R.; Campbell, E. B.; Cadene, M.; Chait, B. T.; MacKin-
non, R. Nature, 2002, 415(6869), 287-94.
[6] Miller, C. Nature, 2006, 440, 484-9.
[7] Accardi, A.; Miller, C. Nature, 2004, 427, 803-7.
[8] Accardi, A.; Walden, M.; Nguitragool, W.; Jayaram, H.; Williams,
C.; Miller, C. J. Gen. Physiol., 2005, 126, 563-70.
[9] Picollo, A.; Pusch, M. Nature, 2005, 436, 420-3.
[10] Pusch, M.; Zifarelli, G.; Murgia, A. R.; Picollo, A.; Babini, E. Exp.
Physiol., 2006, 91, 149-52.
[11] Scheel, O.; Zdebik, A. A.; Lourdel, S.; Jentsch, T. J. Nature, 2005,
436, 424-7.
[12] Middleton, R. E.; Pheasant, D. J.; Miller, C. Biochemistry, 1994,
33, 13189-98.
[13] Middleton, R. E.; Pheasant, D. J.; Miller, C. Nature, 1996, 383,
337-40.
[14] Pusch, M.; Ludewig, U.; Rehfeldt, A.; Jentsch, T. J. Nature, 1995,
373, 527-31.
[15] Long, S. B.; Campbell, E. B.; Mackinnon, R. Science, 2005, 309,
903-8.
[16] Long, S. B.; Tao, X.; Campbell, E. B.; MacKinnon, R. Nature,
2007, 450, 376-82.
[17] Long, S. B.; Campbell, E. B.; Mackinnon, R. Science, 2005, 309,
897-903.
[18] Alabi, A. A.; Bahamonde, M. I.; Jung, H. J.; Kim, J. I.; Swartz, K.
J. Nature, 2007, 450, 370-5.
[19] Murata, Y.; Iwasaki, H.; Sasaki, M.; Inaba, K.; Okamura, Y. Na-
ture, 2005, 435, 1239-43.
[20] Ramsey, I. S.; Moran, M. M.; Chong, J. A.; Clapham, D. E. Nature,
2006, 440, 1213-6.
[21] Sasaki, M.; Takagi, M.; Okamura, Y. Science, 2006, 312, 589-92.
[22] Jiang, Y.; Lee, A.; Chen, J.; Cadene, M.; Chait, B. T.; MacKinnon,
R. Nature, 2002, 417, 515-22.
[23] Jiang, Y.; Pico, A.; Cadene, M.; Chait, B. T.; MacKinnon, R. Neu-
ron, 2001, 29, 593-601.
[24] Xia, X. M.; Zeng, X.; Lingle, C. J. Nature, 2002, 418, 880-4.
[25] Ye, S.; Li, Y.; Chen, L.; Jiang, Y. Cell, 2006, 126, 1161-73.
[26] Schreiber, M.; Salkoff, L. Biophys. J., 1997, 73, 1355-63.
[27] Cox, D. H.; Cui, J.; Aldrich, R. W. J. Gen. Physiol., 1997, 110,
257-81.
[28] Craven, K. B.; Zagotta, W. N. Annu. Rev. Physiol., 2006, 68, 375-
401.
[29] Ganetzky, B.; Robertson, G. A.; Wilson, G. F.; Trudeau, M. C.;
Titus, S. A. Ann. N. Y. Acad. Sci., 1999, 868, 356-69.
[30] Chen, G. Q.; Cui, C.; Mayer, M. L.; Gouaux, E. Nature, 1999, 402,
817-21.
[31] Tasneem, A.; Iyer, L. M.; Jakobsson, E.; Aravind, L. Genome Biol.,
2005, 6, R4.
[32] Bocquet, N.; Prado de Carvalho, L.; Cartaud, J.; Neyton, J.; Le
Poupon, C.; Taly, A.; Grutter, T.; Changeux, J. P.; Corringer, P. J.
Nature, 2007, 445, 116-9.
[33] Barrera, N. P.; Ormond, S. J.; Henderson, R. M.; Murrell-Lagnado,
R. D.; Edwardson, J. M. J. Biol. Chem., 2005, 280,10759-65.
[34] Jasti, J.; Furukawa, H.; Gonzales, E. B.; Gouaux, E. Nature, 2007,
449, 316-23.
[35] Sukharev, S. I.; Blount, P.; Martinac, B.; Blattner, F. R.; Kung, C.
Nature, 1994, 368, 265-8.
[36] Price, M. P.; Lewin, G. R.; McIlwrath, S. L.; Cheng, C.; Xie, J.;
Heppenstall, P. A.; Stucky, C. L.; Mannsfeldt, A. G.; Brennan, T.
J.; Drummond, H. A.; Qiao, J.; Benson, C. J.; Tarr, D. E.; Hrstka,
R. F.; Yang, B.; Williamson, R. A.; Welsh, M. J. Nature, 2000,
407, 1007-11.
[37] Liu, J.; Schrank, B.; Waterston, R. H. Science, 1996, 273, 361-4.
[38] Huang, M.; Chalfie, M. Nature, 1994, 367, 467-70.
[39] Drummond, H. A.; Price, M. P.; Welsh, M. J.; Abboud, F. M. Neu-
ron, 1998, 21, 1435-41.
[40] Walker, R. G.; Willingham, A. T.; Zuker, C. S. Science, 2000, 287,
2229-34.
[41] Christensen, A. P.; Corey, D. P. Nat. Rev. Neurosci., 2007, 8, 510-
21.
[42] Bandell, M.; Macpherson, L. J.; Patapoutian, A. Curr. Opin. Neu-
robiol., 2007, 17, 490-7.
[43] Voets, T.; Droogmans, G.; Wissenbach, U.; Janssens, A.; Flock-
erzi, V.; Nilius, B. Nature, 2004, 430, 748-54.
[44] Brauchi, S.; Orta, G.; Salazar, M.; Rosenmann, E.; Latorre, R. J.
Neurosci., 2006, 26, 4835-40.
[45] Noda, M.; Takahashi, H.; Tanabe, T.; Toyosato, M.; Kikyotani, S.;
Hirose, T.; Asai, M.; Takashima, H.; Inayama, S.; Miyata, T.;
Numa, S. Nature, 1983, 301, 251-5.
[46] Noda, M.; Takahashi, H.; Tanabe, T.; Toyosato, M.; Furutani, Y.;
Hirose, T.; Asai, M.; Inayama, S.; Miyata, T.; Numa, S. Nature,
1982, 299, 793-7.
[47] Tanabe, T.; Takeshima, H.; Mikami, A.; Flockerzi, V.; Takahashi,
H.; Kangawa, K.; Kojima, M.; Matsuo, H.; Hirose, T.; Numa, S.
Nature, 1987, 328, 313-8.
[48] Noda, M.; Ikeda, T.; Suzuki, H.; Takeshima, H.; Takahashi, T.;
Kuno, M.; Numa, S. Nature 1986, 322, 826-8.
[49] Kamb, A.; Iverson, L. E.; Tanouye, M. A. Cell, 1987, 50, 405-13.
[50] Tempel, B. L.; Papazian, D. M.; Schwarz, T. L.; Jan, Y. N.; Jan, L.
Y. Science, 1987, 237, 770-5.
[51] Papazian, D. M.; Schwarz, T. L.; Tempel, B. L.; Jan, Y. N.; Jan, L.
Y. Science, 1987, 237, 749-53.
[52] Salkoff, L.; Wyman, R. Nature, 1981, 293, 228-30.
[53] Tanouye, M. A.; Ferrus, A.; Fujita, S. C. Proc. Natl. Acad. Sci.
USA, 1981, 78, 6548-6552.
[54] Vig, M.; Peinelt, C.; Beck, A.; Koomoa, D. L.; Rabah, D.; Koblan-
Huberson, M.; Kraft, S.; Turner, H.; Fleig, A.; Penner, R.; Kinet, J.
P. Science, 2006, 312, 1220-3.
[55] Prakriya, M.; Feske, S.; Gwack, Y.; Srikanth, S.; Rao, A.; Hogan,
P. G. Nature, 2006, 443, 230-3.
[56] Yeromin, A. V.; Zhang, S. L.; Jiang, W.; Yu, Y.; Safrina, O.; Caha-
lan, M. D. Nature, 2006, 443, 226-9.
[57] Vig, M.; Beck, A.; Billingsley, J. M.; Lis, A.; Parvez, S.; Peinelt,
C.; Koomoa, D. L.; Soboloff, J.; Gill, D. L.; Fleig, A.; Kinet, J. P.;
Penner, R. Curr. Biol., 2006, 16, 2073-9.
[58] Butler, A.; Wei, A. G.; Baker, K.; Salkoff, L. Science, 1989, 243,
943-7.
[59] Altschul, S. F.; Gish, W.; Miller, W.; Myers, E. W.; Lipman, D. J.
J. Mol. Biol., 1990, 215, 403-10.
[60] Ishibashi, K.; Suzuki, M.; Imai, M. Biochem. Biophys. Res. Com-
mun., 2000, 270, 370-6.
[61] Joiner, W. J.; Tang, M. D.; Wang, L. Y.; Dworetzky, S. I.; Bois-
sard, C. G.; Gan, L.; Gribkoff, V. K.; Kaczmarek, L. K. Nat. Neu-
rosci., 1998, 1, 462-9.
[62] Ketchum, K. A.; Joiner, W. J.; Sellers, A. J.; Kaczmarek, L. K.;
Goldstein, S. A. Nature, 1995, 376, 690-5.
[63] Kohler, M.; Hirschberg, B.; Bond, C. T.; Kinzie, J. M.; Marrion, N.
V.; Maylie, J.; Adelman, J. P. Science, 1996, 273, 1709-14.
[64] Lee, J. H.; Cribbs, L. L.; Perez-Reyes, E. FEBS Lett., 1999, 445,
231-6.
[65] Perez-Reyes, E.; Cribbs, L. L.; Daud, A.; Lacerda, A. E.; Barclay,
J.; Williamson, M. P.; Fox, M.; Rees, M.; Lee, J. H. Nature, 1998,
391, 896-900.
[66] Wei, A.; Jegla, T.; Salkoff, L. Neuropharmacology, 1996, 35, 805-
29.
[67] Fields, C.; Adams, M. D.; White, O.; Venter, J. C. Nat. Genet.,
1994, 7, 345-6.
[68] Liang, F.; Holt, I.; Pertea, G.; Karamycheva, S.; Salzberg, S. L.;
Quackenbush, J. Nat. Genet., 2000, 25, 239-40.
[69] Ewing, B.; Green, P. Nat. Genet., 2000, 25, 232-4.
[70] Clamp, M.; Fry, B.; Kamal, M.; Xie, X.; Cuff, J.; Lin, M. F.; Kel-
lis, M.; Lindblad-Toh, K.; Lander, E. S. Proc. Natl. Acad. Sci.
USA., 2007, 104, 19428-33.
[71] Kuo, M. M.; Haynes, W. J.; Loukin, S. H.; Kung, C.; Saimi, Y.
FEMS Microbiol. Rev., 2005, 29, 961-85.
[72] Chiu, P. L.; Pagel, M. D.; Evans, J.; Chou, H. T.; Zeng, X.; Gipson,
B.; Stahlberg, H.; Nimigean, C. M. Structure, 2007, 15, 1053-64.
[73] Jiang, Y.; Lee, A.; Chen, J.; Ruta, V.; Cadene, M.; Chait, B. T.;
MacKinnon, R. Nature, 2003, 423, 33-41.
[74] International HapMap Consortium. Nature, 2005, 437, 1299-320.
[75] Brinkman, R. R.; Dube, M. P.; Rouleau, G. A.; Orr, A. C.; Sa-
muels, M. E. Nat. Rev. Genet., 2006, 7, 249-60.

22 Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 Jegla et al.
[76] Mignen, O.; Thompson, J. L.; Shuttleworth, T. J. J. Physiol., 2008,
586, 419-25.
[77] Pruitt, K. D.; Tatusova, T.; Maglott, D. R. Nucleic Acids Res.,
2007, 35, (Database issue), D61-5.
[78] Palmer, C. P.; Zhou, X. L.; Lin, J.; Loukin, S. H.; Kung, C.; Saimi,
Y. Proc. Natl. Acad. Sci. USA, 2001, 98, 7801-5.
[79] King, N.; Westbrook, M.J.; Young, S.L.; Kuo, A.; Abedin, M.;
Chapman, J.; Fairclough, S.; Hellsten, U.; Isogai, Y.; Letunic, I.;
Marr, M.; Pincus, D.; Putnam, N.; Rokas, A.; Wright, K.J.; Zuzow,
R.; Dirks, W.; Good, M.; Goodstein, D.; Lemons, D.; Li, W.;
Lyons, J.B.; Morris, A.; Nichols, S.; Richter, D.J.; Salamov, A.;
Sequencing, J.G.; Bork, P.; Lim, W.A.; Manning, G.; Miller, W.T.;
McGinnis, W.; Shapiro, H.; Tjian, R.; Grigoriev, IV.; Rokhsar, D.
Nature, 2008, 451, 783-788.
[80] Very, A. A.; Sentenac, H. Trends Plant Sci., 2002, 7, 168-75.
[81] Bargmann, C. I. Science, 1998, 282, 2028-33.
[82] Littleton, J. T.; Ganetzky, B. Neuron, 2000, 26, 35-43.
[83] McCormack, T. J. Genome Biol., 2003, 4, R58.
[84] Okamura, Y.; Nishino, A.; Murata, Y.; Nakajo, K.; Iwasaki, H.;
Ohtsuka, Y.; Tanaka-Kunishima, M.; Takahashi, N.; Hara, Y.; Yo-
shida, T.; Nishida, M.; Okado, H.; Watari, H.; Meinertzhagen, I.
A.; Satoh, N.; Takahashi, K.; Satou, Y.; Okada, Y.; Mori, Y.
Physiol. Genomics., 2005, 22, 269-82.
[85] Aparicio, S.; Chapman, J.; Stupka, E.; Putnam, N.; Chia, J. M.;
Dehal, P.; Christoffels, A.; Rash, S.; Hoon, S.; Smit, A.; Gelpke,
M. D.; Roach, J.; Oh, T.; Ho, I. Y.; Wong, M.; Detter, C.; Verhoef,
F.; Predki, P.; Tay, A.; Lucas, S.; Richardson, P.; Smith, S. F.;
Clark, M. S.; Edwards, Y. J.; Doggett, N.; Zharkikh, A.; Tavtigian,
S. V.; Pruss, D.; Barnstead, M.; Evans, C.; Baden, H.; Powell, J.;
Glusman, G.; Rowen, L.; Hood, L.; Tan, Y. H.; Elgar, G.; Haw-
kins, T.; Venkatesh, B.; Rokhsar, D.; Brenner, S. Science, 2002,
297, 1301-10.
[86] International Chicken Genome Sequencing Consortium. Nature,
2004, 432, 695-716.
[87] Holt, R. A.; Subramanian, G. M.; Halpern, A.; Sutton, G. G.; Char-
lab, R.; Nusskern, D. R.; Wincker, P.; Clark, A. G.; Ribeiro, J. M.;
Wides, R.; Salzberg, S. L.; Loftus, B.; Yandell, M.; Majoros, W.
H.; Rusch, D. B.; Lai, Z.; Kraft, C. L.; Abril, J. F.; Anthouard, V.;
Arensburger, P.; Atkinson, P. W.; Baden, H.; de Berardinis, V.;
Baldwin, D.; Benes, V.; Biedler, J.; Blass, C.; Bolanos, R.; Boscus,
D.; Barnstead, M.; Cai, S.; Center, A.; Chaturverdi, K.; Christo-
phides, G. K.; Chrystal, M. A.; Clamp, M.; Cravchik, A.; Curwen,
V.; Dana, A.; Delcher, A.; Dew, I.; Evans, C. A.; Flanigan, M.;
Grundschober-Freimoser, A.; Friedli, L.; Gu, Z.; Guan, P.; Guigo,
R.; Hillenmeyer, M. E.; Hladun, S. L.; Hogan, J. R.; Hong, Y. S.;
Hoover, J.; Jaillon, O.; Ke, Z.; Kodira, C.; Kokoza, E.; Koutsos, A.;
Letunic, I.; Levitsky, A.; Liang, Y.; Lin, J. J.; Lobo, N. F.; Lopez,
J. R.; Malek, J. A.; McIntosh, T. C.; Meister, S.; Miller, J.;
Mobarry, C.; Mongin, E.; Murphy, S. D.; O'Brochta, D. A.;
Pfannkoch, C.; Qi, R.; Regier, M. A.; Remington, K.; Shao, H.;
Sharakhova, M. V.; Sitter, C. D.; Shetty, J.; Smith, T. J.; Strong,
R.; Sun, J.; Thomasova, D.; Ton, L. Q.; Topalis, P.; Tu, Z.; Unger,
M. F.; Walenz, B.; Wang, A.; Wang, J.; Wang, M.; Wang, X.;
Woodford, K. J.; Wortman, J. R.; Wu, M.; Yao, A.; Zdobnov, E.
M.; Zhang, H.; Zhao, Q.; Zhao, S.; Zhu, S. C.; Zhimulev, I.;
Coluzzi, M.; della Torre, A.; Roth, C. W.; Louis, C.; Kalush, F.;
Mural, R. J.; Myers, E. W.; Adams, M. D.; Smith, H. O.; Broder,
S.; Gardner, M. J.; Fraser, C. M.; Birney, E.; Bork, P.; Brey, P. T.;
Venter, J. C.; Weissenbach, J.; Kafatos, F. C.; Collins, F. H.;
Hoffman, S. L. Science, 2002, 298, 129-49.
[88] Putnam, N. H.; Srivastava, M.; Hellsten, U.; Dirks, B.; Chapman,
J.; Salamov, A.; Terry, A.; Shapiro, H.; Lindquist, E.; Kapitonov,
V. V.; Jurka, J.; Genikhovich, G.; Grigoriev, I. V.; Lucas, S. M.;
Steele, R. E.; Finnerty, J. R.; Technau, U.; Martindale, M. Q.;
Rokhsar, D. S. Science, 2007, 317, 86-94.
[89] Niesler, B.; Frank, B.; Kapeller, J.; Rappold, G. A. Gene, 2003,
310, 101-11.
[90] Niesler, B.; Walstab, J.; Combrink, S.; Moller, D.; Kapeller, J.;
Rietdorf, J.; Bonisch, H.; Gothert, M.; Rappold, G.; Bruss, M. Mol.
Pharmacol., 2007, 72, 8-17.
[91] Stowers, L.; Holy, T. E.; Meister, M.; Dulac, C.; Koentges, G.
Science, 2002, 295, 1493-500.
[92] Liman, E. R.; Innan, H. Proc. Natl. Acad. Sci. USA, 2003, 100,
3328-32.
[93] Robbins, J. Pharmacol. Ther., 2001, 90, 1-19.
[94] Amores, A.; Force, A.; Yan, Y. L.; Joly, L.; Amemiya, C.; Fritz,
A.; Ho, R. K.; Langeland, J.; Prince, V.; Wang, Y. L.; Westerfield,
M.; Ekker, M.; Postlethwait, J. H. Science, 1998, 282, 1711-4.
[95] Meyer, A.; Schartl, M. Curr. Opin. Cell Biol., 1999, 11, 699-704.
[96] Ohno, S. Curr. Opin. Genet. Dev., 1993, 3, 911-4.
[97] Postlethwait, J. H.; Yan, Y. L.; Gates, M. A.; Horne, S.; Amores,
A.; Brownlie, A.; Donovan, A.; Egan, E. S.; Force, A.; Gong, Z.;
Goutel, C.; Fritz, A.; Kelsh, R.; Knapik, E.; Liao, E.; Paw, B.; Ran-
som, D.; Singer, A.; Thomson, M.; Abduljabbar, T. S.; Yelick, P.;
Beier, D.; Joly, J. S.; Larhammar, D.; Rosa, F.; Westerfield, M.;
Zon, L.I.; Johnson, S.L.; Talbot, W.S. Nat. Genet., 1998, 18, 345-9.
[98] Sidow, A. Curr. Opin. Genet. Dev., 1996, 6, 715-22.
[99] Minke, B. Cell. Mol. Neurobiol., 2001, 21, 629-43.
[100] Lu, B.; Su, Y.; Das, S.; Liu, J.; Xia, J.; Ren, D. Cell, 2007, 129,
371-83.
[101] Lear, B. C.; Lin, J. M.; Keath, J. R.; McGill, J. J.; Raman, I. M.;
Allada, R. Neuron, 2005, 48, 965-76.
[102] Nash, H. A.; Scott, R. L.; Lear, B. C.; Allada, R. Curr. Biol., 2002,
12, 2152-8.
[103] Leibovitch, B. A.; Campbell, D. B.; Krishnan, K. S.; Nash, H. A. J.
Neurogenet., 1995, 10, 1-13.
[104] Jospin, M.; Watanabe, S.; Joshi, D.; Young, S.; Hamming, K.;
Thacker, C.; Snutch, T. P.; Jorgensen, E. M.; Schuske, K. Curr.
Biol., 2007, 17, 1595-600.
[105] Franks, C. J.; Pemberton, D.; Vinogradova, I.; Cook, A.; Walker,
R. J.; Holden-Dye, L. J. Neurophysiol., 2002, 87, 954-61.
[106] Vinogradova, I.; Cook, A.; Holden-Dye, L. Invert Neurosci., 2006,
6, 57-68.
[107] Hirosawa, M.; Nagase, T.; Ishikawa, K.; Kikuno, R.; Nomura, N.;
Ohara, O. DNA Res., 1999, 6, 329-36.
[108] Sanguinetti, M. C.; Jiang, C.; Curran, M. E.; Keating, M. T. Cell,
1995, 81, 299-307.
[109] Curran, M. E.; Splawski, I.; Timothy, K. W.; Vincent, G. M.;
Green, E. D.; Keating, M. T. Cell, 1995, 80, 795-803.
[110] Marrion, N. V. Annu. Rev. Physiol., 1997, 59, 483-504.
[111] Wang, H. S.; Pan, Z.; Shi, W.; Brown, B. S.; Wymore, R. S.;
Cohen, I. S.; Dixon, J. E.; McKinnon, D. Science, 1998, 282, 1890-
3.
[112] Suh, B. C.; Inoue, T.; Meyer, T.; Hille, B. Science, 2006, 314,
1454-7.
[113] Surti, T. S.; Jan, L. Y. Curr. Opin. Investig. Drugs., 2005, 6, 704-
11.
[114] Selyanko, A. A.; Delmas, P.; Hadley, J. K.; Tatulian, L.; Wood, I.
C.; Mistry, M.; London, B.; Brown, D. A. J. Neurosci., 2002, 22,
RC212.
[115] Gomez-Varela, D.; Zwick-Wallasch, E.; Knotgen, H.; Sanchez, A.;
Hettmann, T.; Ossipov, D.; Weseloh, R.; Contreras-Jurado, C.;
Rothe, M.; Stuhmer, W.; Pardo, L. A. Cancer Res., 2007, 67, 7343-
9.
[116] Pardo, L. A.; Contreras-Jurado, C.; Zientkowska, M.; Alves, F.;
Stuhmer, W. J. Membr. Biol., 2005, 205, 115-24.
[117] Bautista, D. M.; Jordt, S. E.; Nikai, T.; Tsuruda, P. R.; Read, A. J.;
Poblete, J.; Yamoah, E. N.; Basbaum, A. I.; Julius, D. Cell, 2006,
124, 1269-82.
[118] Kwan, K. Y.; Allchorne, A. J.; Vollrath, M. A.; Christensen, A. P.;
Zhang, D. S.; Woolf, C. J.; Corey, D. P. Neuron, 2006, 50, 277-89.
[119] Petrus, M.; Peier, A. M.; Bandell, M.; Hwang, S. W.; Huynh, T.;
Olney, N.; Jegla, T.; Patapoutian, A. Mol. Pain., 2007, 3, 40.
[120] Dhaka, A.; Viswanath, V.; Patapoutian, A. Annu. Rev. Neurosci.,
2006, 29, 135-61.
[121] Zurborg, S.; Yurgionas, B.; Jira, J. A.; Caspani, O.; Heppenstall, P.
A. Nat. Neurosci., 2007, 10, 277-9.
[122] Zuker, C. S. Proc. Natl. Acad. Sci. USA, 1996, 93, 571-6.
[123] Panda, S.; Nayak, S. K.; Campo, B.; Walker, J. R.; Hogenesch, J.
B.; Jegla, T. Science, 2005, 307, 600-4.
[124] Bloomquist, J. R. Annu. Rev. Entomol., 1996, 41, 163-90.
[125] Dong, K. Invert. Neurosci., 2007, 7, 17-30.
[126] Ray, D. E.; Fry, J. R. Pharmacol. Ther., 2006, 111, 174-93.
[127] Tamargo, J.; Caballero, R.; Nunez, L.; Gomez, R.; Vaquero, M.;
Delpon, E. Biosci., 2007, 12, 22-38.
[128] Berul, C. I. Physiol. Genomics., 2003, 13, 207-16.
[129] Bayliss, D. A.; Sirois, J. E.; Talley, E. M. Mol. Interv., 2003, 3,
205-19.
[130] Inoue, R.; Mori, Y. Curr. Drug Targets Cardiovasc. Haematol.
Disord., 2003, 3, 59-72.
[131] Wermeling, D. P. Pharmacotherapy., 2005, 25, 1084-94.

Evolution of the Human Ion Channel Set Combinatorial Chemistry & High Throughput Screening, 2009, Vol. 12, No. 1 23
[132] Benham, C. D. Nat. Biotechnol., 2005, 23, 1234-5.
[133] Vandenberg, J. I.; Walker, B. D.; Campbell, T. J. Trends Pharma-
col. Sci., 2001, 22, 240-6.
[134] Becchetti, A.; De Fusco, M.; Crociani, O.; Cherubini, A.; Restano-
Cassulini, R.; Lecchi, M.; Masi, A.; Arcangeli, A.; Casari, G.;
Wanke, E. Eur. J. Neurosci., 2002, 16, 415-28.
[135] Garcia, M. L.; Hanner, M.; Knaus, H. G.; Slaughter, R.; Kaczo-
rowski, G. J. Methods Enzymol., 1999, 294, 624-39.
[136] Tombola, F.; Pathak, M. M.; Isacoff, E. Y. Annu. Rev. Cell. Dev.
Biol., 2006, 22, 23-52.
[137] Swartz, K. J. Toxicon, 2007, 49, 213-30.
[138] Wainger, B. J.; DeGennaro, M.; Santoro, B.; Siegelbaum, S. A.;
Tibbs, G. R. Nature, 2001, 411, 805-10.
[139] Zagotta, W. N.; Olivier, N. B.; Black, K. D.; Young, E. C.; Olson,
R.; Gouaux, E. Nature, 2003, 425, 200-5.
[140] Young, K.; Lin, S.; Sun, L.; Lee, E.; Modi, M.; Hellings, S.; Hus-
bands, M.; Ozenberger, B.; Franco, R. Nat. Biotechnol., 1998, 16,
946-50.
[141] McCrossan, Z. A.; Abbott, G. W. Neuropharmacology, 2004, 47,
787-821.
[142] Meera, P.; Wallner, M.; Toro, L. Proc. Natl. Acad. Sci. USA, 2000,
97, 5562-7.
[143] Ottschytsch, N.; Raes, A.; Van Hoorick, D.; Snyders, D. J. Proc.
Natl. Acad. Sci. USA, 2002, 99, 7986-91.
[144] Wu, H.; Cowing, J. A.; Michaelides, M.; Wilkie, S. E.; Jeffery, G.;
Jenkins, S. A.; Mester, V.; Bird, A. C.; Robson, A. G.; Holder, G.
E.; Moore, A. T.; Hunt, D. M.; Webster, A. R. Am. J. Hum. Genet.,
2006, 79, 574-9.
[145] Tamura, K.; Dudley, J.; Nei, M.; Kumar, S. Mol. Biol. Evol., 2007,
24, 1596-9.
[146] Zmasek, C. M.; Eddy, S. R. Bioinformatics, 2001, 17, 821-8.


Received: February 15, 2008 Revised: February 20, 2008 Accepted: February 20, 2008