Download The development and evolutionary origin of barbels in

University of Colorado, Boulder
CU Scholar
Ecology & Evolutionary Biology Graduate Theses &
Dissertations
Ecology & Evolutionary Biology
Spring 1-1-2011
The development and evolutionary origin of
barbels in the channel catfish Ictalurus punctatus
(Siluriformes: Ictaluridae)
Michael Brent Hawkins
University of Colorado at Boulder, [email protected]
Follow this and additional works at: http://scholar.colorado.edu/ebio_gradetds
Part of the Ecology and Evolutionary Biology Commons
Recommended Citation
Hawkins, Michael Brent, "The development and evolutionary origin of barbels in the channel catfish Ictalurus punctatus (Siluriformes:
Ictaluridae)" (2011). Ecology & Evolutionary Biology Graduate Theses & Dissertations. Paper 13.
This Thesis is brought to you for free and open access by Ecology & Evolutionary Biology at CU Scholar. It has been accepted for inclusion in Ecology
& Evolutionary Biology Graduate Theses & Dissertations by an authorized administrator of CU Scholar. For more information, please contact
[email protected].
THE DEVELOPMENT AND EVOLUTIONARY ORIGIN OF BARBELS IN THE CHANNEL
CATFISH ICTALURUS PUNCTATUS (SILURIFORMES: ICTALURIDAE)
by
MICHAEL BRENT HAWKINS
B.A., University of Colorado, 2008
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirement for the degree of
Master’s of Arts
Department of Ecology and Evolutionary Biology
2011
This thesis entitled:
The development and evolutionary origin of barbels in the channel catfish Ictalurus punctatus
(Siluriformes: Ictaluridae)
written by Michael Brent Hawkins
has been approved for the Department of Ecology and Evolutionary Biology
David W. Stock, committee co-chair
Alexander Cruz, committee co-chair
Date
The final copy of this thesis has been examined by the signatories, and we
Find that both the content and the form meet acceptable presentation standards
Of scholarly work in the above mentioned discipline.
Hawkins, Michael Brent (M.A., Ecology and Evolutionary Biology)
The development and evolutionary origin of barbels in the channel catfish Ictalurus punctatus
(Siluriformes: Ictaluridae)
Thesis co-directed by Professor Alexander Cruz and Associate Professor David Stock
Understanding the origin of morphological novelties is an important goal of evolutionary
developmental biology. In pursuit of this goal, we have examined the developmental genetic
mechanisms that underlie growth and patterning in a largely overlooked group of morphological
novelties: the barbels of fishes. Barbels are appendages that project from the head region in a
large and disparate assortment of fish taxa, ranging from hagfishes to gobies. They often bear
sensory organs and can be supported by a rod of connective tissue, muscle, cartilage, or bone.
Considering the scattered distribution of barbels among fishes, along with the variability of
barbel position and composition, it is likely that barbels have originated independently in
multiple groups. We investigated the roles of genes known to be involved in the development of
other appendages in the developing barbels of the channel catfish, Ictalurus punctatus
(Siluriformes: Ictaluridae). Similar to other appendages, the barbels of I. punctatus express
members of the Bone morphogenetic protein (Bmp), Distal-less (Dlx), Fibroblast growth factor
(Fgf), Hedgehog (Hh), Tumor necrosis factor (Tnf), and Tnf receptor families. Other genes with
roles in appendage development were absent from barbels, however, including members of the
Dachshund (Dach) and Hox families. Treatment with pharmacological inhibitors of Hh signaling
revealed that this pathway is necessary for barbel outgrowth. I conclude that while the barbels of
catfishes arose via deployment of a general vertebrate outgrowth mechanism (an Fgf/Hh
feedback loop), additional features of the gene regulatory network underlying their development
overlap, but are distinct from, those of other appendages.
iii
This thesis is dedicated to my parents, Becky and Gari Hawkins.
ACKNOWLEDGEMENTS
I owe much to David Stock for his instruction and guidance on laboratory protocols,
experimental design, and data interpretation. Alexander Cruz and Daniel Medeiros provided
critical insights, discussion, and support over the course of the research. Thanks are also due to
Marcus Cohen. Mysti Martin, Leif Neitzel, and Paige Swanson assisted in channel catfish
husbandry.
This research was supported by a Raney Fund award from the American Society of
Ichthyologists and Herpetologists, a Grants-In-Aid of Research award from Sigma Xi, a GrantsIn-Aid of Research award from the Society for Integrative and Comparative Biology, a Beverly
Sears Graduate Student Grant award from the University of Colorado Graduate School, and an
EBIO Department Research Grant award from the Ecology and Evolutionary Biology Graduate
Research Funds and the Graduate School University Fellowships Funds.
v
CONTENTS
CHAPTER
I.
INTRODUCTION……………....................................................................1
II.
MATERIALS AND METHODS……………...……….………….............6
Animals..……………………………………………….............…....…6
Gene cloning and sequence analysis.…………………………..............6
in situ hybridization.…………………………………………………...8
Pharmacological inhibition of hedgehog signaling …………………...8
Immunohistochemistry and skeletal staining...…………………...……9
III.
RESULTS...………………………………………..……………………..10
Formation of barbels during embryonic stages …….………………...10
Expression of bmp4, fgf8a and shha in developing barbels...………..11
Expression of Dlx gene family members in the developing barbels....15
Expression of eda and its receptor in the developing barbels………..18
Other vertebrate appendage development
genes not expressed in the barbel..……………………………….19
Hedgehog signaling is required for barbel outgrowth………………..20
IV.
DISCUSSION.……………………………………………………..…….23
Maxillary barbels, paired appendages, and branchial rays..……….…23
Divergent Dlx gene expression patterns
between barbels and pharyngeal arches..……………………....…27
Divergent gene expression among maxillary and lower jaw barbels...28
The origin and evolution of barbels in fishes…...……………………30
BIBLIOGRAPHY....………………………………………………….…………....…34
vi
FIGURES
Figure
1.
The phylogenetic distribution of barbels across fishes at the ordinal level..3
2.
Barbels of the channel catfish I. punctatus.………………..………………5
3.
in situ hybridizations demonstrating expression of bmp4, fgf8a, and
shha during I. punctatus barbel development.………………….….....13
4.
Dlx gene expression in the developing barbels shown by
in situ hybridization..…………………………………………….…...17
5.
Barbel expression of ectodysplasin and its receptor demonstrated by
in situ hybridization..…………………………………………….…...19
6.
in situ hybridizations demonstrating the absence of hoxa13b, hoxd11a,
and dach1 from the developing barbels..……………………………..20
7.
Treatment with pharmacological inhibitors of hedgehog signaling
block barbel outgrowth..……………………………………………...22
8.
Schematic diagram comparing gene expression between
paired appendages and early and late maxillary barbels.......................24
vii
CHAPTER I
INTRODUCTION
The origin of morphological novelties, defined as structures lacking homologs within or
between organisms (Müller and Wagner 1991), has long been of interest to evolutionary
biologists (Moczek 2008). A major contribution of evolutionary developmental biology to the
question of how novelty arises has been the finding that novel morphological structures
frequently arise by the co-option of pre-existing gene regulatory networks for their patterning
rather than the de novo assembly of such networks (Shubin et al. 2009). Furthermore, in some
cases, the same gene regulatory networks have been co-opted in the origin of functionally similar
but independently evolved (analogous) structures, as in the appendages of vertebrates and
arthropods (Panganiban et al. 1997; Pueyo and Couso 2005). Such comparisons of appendages in
distantly related species have emphasized the similarities in the genetic control of their
development. However, vertebrates possess a diversity of appendage types, such as paired
appendages (limbs and fins), unpaired median fins, genitalia, epithelial appendages (hair,
feathers, teeth, and scales), and branchial rays (Kardong 2005), which exhibit developmental
differences as well as similarities (Chuong 1998; Freitas et al. 2006; Gillis et al. 2009; Lin et al.
2009; Zeller 2010). A functionally important and phylogenetically widespread group of
vertebrate appendages that have yet to be investigated extensively at the developmental genetic
level are the barbels of fishes (LeClair and Topczewski 2010).
Barbels (also known as cirri, feelers, tendrils, or tentacles) are appendages that project from
the head region of certain fishes, such as the “whiskers” from which catfishes receive their
popular name. Despite frequently being sensory in nature and playing a role in food detection,
1
barbels display extensive variety in their position, composition, developmental stage of
appearance, and function among species (Ryder 1883; Sato 1937; Fuiman 1984; Fox 1999).
They may exist as either unpaired structures on the midline or paired, bilaterally symmetric
appendages, may be located on the top, bottom, and sides of the snout, head, and throat, and may
range in number from one to dozens. While their composition is highly variable, they generally
consist of an epidermis surrounding a dermal core (Sato 1937). The epidermis is frequently
populated by taste buds, but olfactory and mechanosensory organs are also present in some
species (Herrick 1903). The dermal layer often contains blood vessels and nerves, and may
contain a supporting rod of connective tissue, cartilage, bone, or striated muscle (Sato 1937;
LeClair and Topczewski 2010). While most barbels are unbranched, secondary through
quaternary branching is found in some species. The point in ontogeny during which barbels
develop also varies from embryonic through adult stages. Uses of barbels beyond food detection
include luring prey and obtaining oxygen in hypoxic environments (Winemiller 1989; Helfman
et al. 2009).
Of the 62 orders of fishes listed by Nelson (2006), my survey of the literature suggests that
27 contain at least one species with barbels (Fig. 1). While once proposed to be homologous
structures (Balfour 1880), barbels are most commonly thought to have arisen independently in
many of these lineages, based on extensive differences in their position and composition, and
their scattered distribution in the fish phylogeny (Ryder 1890; Fox 1999). In many orders,
barbels are novel morphological structures by the definition of Müller and Wagner (1991), but
exceptions occur, such as the hyoid barbels of goatfishes, which are derived from branchiostegal
rays (Starks 1904; McCormick 1993; Kim et al. 2004).
2
Fig. 1. The phylogenetic distribution of barbels across fishes at the ordinal level. Orders
containing one or more fishes with barbels are highlighted in red. Representative species shown
at right demonstrate the various positions barbels can occupy. Tree topology adapted from
Nelson (2006). Illustrations taken from Dean 1895.
3
To begin to explore the developmental genetic mechanisms responsible for barbel evolution,
we chose to examine barbel development in the channel catfish Ictalurus punctatus
(Siluriformes: Ictaluridae). This species was selected because its barbels appear early in
embryonic development (Northcutt 2005), its embryos are available from commercial sources,
and published staging criteria for ictalurid catfishes are available (Armstrong and Child 1962).
Like all members of its family, the channel catfish bears four pairs of bilaterally symmetric
barbels: the nasal, maxillary, mandibular, and mental pairs (Fig. 2, A-B; Grizzle and Rogers
1976; Nelson 2005). Each of these barbels consists of an epidermis surrounding a dermal core.
The epidermis contains many taste buds, which are concentrated along the anterior edge of the
barbel and are innervated by rami of the facial nerve (Northcutt 2005). Contained in the dermal
layer are blood vessels, connective tissue, pigment cells, nerves, and a supportive rod of
“elastic/cell-rich” cartilage that runs the proximal/distal length of each barbel (Joyce and
Chapman 1978; Benjamin 1990). This rod of cartilage is located posterior to the central axis of
the barbel, and along with the taste bud-dense anterior edge of the epidermis, provides each
barbel with distinct anterior/posterior polarity (Fig. 2C).
A potential explanation for the seeming ease with which barbels have originated in disparate
fish lineages is the co-option of pre-existing developmental genetic mechanisms, such as those
used in the development of other vertebrate appendages. To test this hypothesis, we investigated
in the developing barbels of I. punctatus the expression of a subset of the genes known to control
the development of paired appendages, unpaired median fins, epithelial appendages, genitalia,
and branchial rays. Specifically, we cloned and examined by in situ hybridization the expression
of bmp4, dach1, dlx1a, dlx2a, dlx5a, dlx6a, eda, edar, fgf8a, hoxa13b, hoxd11a, and shha. We
also used pharmacological inhibitors of hedgehog signaling to test the role of this pathway in
4
barbel development. Our results suggest that barbels have arisen not by the co-option of an
elaborated pre-existing vertebrate appendage developmental genetic mechanism, but by
deploying an FGF8/SHH positive feedback loop, a general outgrowth program used in the
development of other vertebrate appendages.
Fig. 2. Barbels of the channel catfish I. punctatus. (A) Schematic diagram of the four barbel pairs
in an I. punctatus adult as seen in rostral view. (B) Head of I. punctatus juvenile cleared and
double stained with Alcian Blue (cartilage) and Alizarin Red (mineralized tissue) following
immunohistochemistry with anti-calretinin antibody to visualize taste buds. (C) Magnified
maxillary barbel of fish in (B) demonstrating morphological differentiation along the anteriorposterior axis, anterior to left, with anterior taste buds, medial nerves, and posterior cartilage rod.
(D) Schematic representation of barbel development during embryonic stages in I. punctatus. C,
cartilage rod; mdb, mandibular barbel; meb, mental barbel; mxb, maxillary barbel; N, nerves;
nsb, nasal barbel; TB, taste bud.
5
CHAPTER II
MATERIALS AND METHODS
Animals
Embryos of I. punctatus were purchased from Osage Catfisheries (Osage Beach, MO).
Newly fertilized eggs were shipped in chilled water and received in early epiboly stages. Upon
arrival, the large mass of adhesive eggs was separated by hand into fragments approximately 3
cm in diameter, each of which was placed in its own container inside a recirculating incubator.
Water in the containers (0.0003% Instant Ocean, 0.48 mM sodium bicarbonate, 0.1 mM HCl in
distilled water) was maintained at 26°C. Embryos were staged according to the I. nebulosus
staging table of Armstrong and Child (1962), which has been adapted for I. punctatus as well
(Northcutt 2005). Certain stages of I. punctatus development resembled a combination of stages
described for I. nebulosus, and these stages are designated here with two numbers (for example,
Stage 34/35). Due to imprecise fertilization timing and time spent in chilled water during
shipment, it is difficult to ascribe ages to each particular stage. However, the maxillary,
mandibular, and mental barbels are visible in Stage 40 embryos at around 4 days postfertilization (dpf). Embryos were fixed overnight in 4% paraformaldehyde in phosphate-buffered
saline (PBS) pH 7.4 at 4°C, rinsed in PBST (PBS + 0.1% Tween 20), and stored in 100%
methanol at -20°C.
Gene cloning and sequence analysis
Total cellular RNA was extracted from 14 dpf I. punctatus larvae using an RNAwiz kit
(Ambion). Random hexamer primers and Superscript II reverse transcriptase (Invitrogen) were
6
used to produce cDNA from this RNA. Genes known to be involved in the development of
vertebrate appendages were PCR-amplified from cDNA using the following degenerate primer
pairs designed using amino acid alignments between zebrafish and other species (added
restriction sites are underlined):
bmp4: GCCGGGATCCATGATHCCNGGNAAYMGNATG,
GCCGGAATTCCRCANCCYTCNACNACCAT
dach1: GCCGGGATCCCCNCARAAYAAYGARTGYAA,
GCCGGAATTCGCNCKRTTYTTYTGYTCNAC
dlx1a: GCCGGGATCCTTYATGGARTTYGGNCCNCC,
GCCGGAATTCATNARYTGNGGYTGYTGCAT
dlx2a: GCCGGGATCCATGACNGGNGTNTTYGAYAG,
GCCGGAATTCAADATNGTNCCNGCRCTNAC
dlx5a: GCCGGGATCCATGCAYCAYCCNWSNCARGA,
GCCGGAATTCTGNGGNSWRTTRCANGCCAT
dlx6a: GCCGGGATCCATGATGACNATGACNACNATG,
GCCGGAATTCTGRTGNGGNGANGARTACCA
eda: GCCGGGATCCCARGGNCARGARACNACNAT,
GCCGGAATTCCCNARRAANGTNGTRTGRTT
edar: GCCGGGATCCAGYGCNGARTAYTCNAGYTG,
GCCGGAATTCTCYTCRTCRCTRTCDATRCT
fgf8a: GCCGGGATCCTTYGCNTTYTGYTAYTAYGC,
GCCGGAATTCGGRTARTTDATRAARTCRAA
hoxa13b: GCCGGGATCCAARGARTTYGCNTTYTAYCA,
GCCGGAATTCTGRAACCADATNGTNACYTG
hoxd11a: GCCGGGATCCCCNCARGGNTTYGAYCARTT,
GCCGGAATTCTGRAACCADATYTTNACYTG
shha: GCCGGGATCCGCNTAYAARCARTTYATHCC,
GCCGGAATTCGARTACCARTGNAYNCCNTC
7
The resulting PCR products were cloned into pCR4-TOPO (Invitrogen) and subjected to
automated sequencing. Each sequence position was determined from at least three independent
clones, together representing both strands. To confirm the identity of clones, sequences were
translated and used in a BLAST search on GenBank (http://www.ncbi.nlm.nih.gov/genbank/).
Clones were ascribed to specific paralogs based on the top BLAST hits.
in situ hybridization
The whole mount in situ hybridization protocol used for I. punctatus specimens was adapted
from that used for zebrafish by Jackman et al. (2004). Proteinase K pre-treatment was carried out
at a concentration of 2.5 μg/mL for 30 minutes at room temperature. Gene expression was
examined at Stages 29/30, 31, 32/33, 34/35, 36, 37, 38, 39, and 40. Digoxigenin-labeled
antisense riboprobes were prepared using the cloned PCR fragments described above as
templates. Following in situ hybridization, larvae were cleared in 100% glycerol for whole
mount observation or dehydrated through a graded ethanol series and embedded in glycol
methacrylate (JB-4, Polysciences) for serial sectioning at 4 m using glass knives. Images of
whole mount specimens were captured with a Zeiss Axiocam digital camera mounted on a Zeiss
SV11 stereomicroscope. Dissected barbels and sections were visualized with Nomarski
differential interference contrast (DIC) optics on a Zeiss Axiovert 135 inverted compound
microscope and images captured with a Zeiss Axiocam digital camera.
Pharmacological inhibition of hedgehog signaling
Two pharmacological inhibitors of the hedgehog signaling pathway, cyclopamine (Chen et
al. 2002a) and SANT-1 (Chen et al. 2002b), were applied to I. punctatus embryos at various
8
stages of development. Cyclopamine (LC Labs) and SANT-1 (Tocris Bioscience) were dissolved
in 100% ethanol and DMSO, respectively, to obtain 10 mM stock solutions. Treatments were
carried out at 26°C in Nunclon 24-well plates, with one individual in 1 mL of medium per well.
Cyclopamine experiments were conducted with treatment concentrations of 50 and 100 μM in
30% Danieau’s medium + 1% ethanol, and SANT-1 experiments were conducted with treatment
concentrations of 10 and 20 μM in 30% Danieau’s medium + 1% DMSO. Vehicle-only control
groups for cyclopamine and SANT-1 treatments were raised in 1% ethanol and 1% DMSO,
respectively. Chorions were left intact for treatments beginning prior to hatching. Individuals
were kept in the same medium from the onset of treatment until being sacrificed at the end of the
experiment.
Immunohistochemistry and skeletal staining
Taste buds were visualized in 14 dpf I. punctatus larvae using anti-calretinin primary
antibody (Swiss Antibodies) following the protocol described by Northcutt (2005). After
developing the antibody stain, specimens were stained for bone and cartilage following the acidfree double stain method described by Walker and Kimmel (2007). While the
immunohistochemistry protocol alone stains taste buds brown, the skeletal staining protocol
causes this brown color to turn blue. Skeletal staining alone does not label taste buds.
9
CHAPTER III
RESULTS
Formation of barbels during embryonic stages
During embryonic stages following somitogenesis, the barbels of I. punctatus appear in the
order of the maxillary pair at Stage 31, followed by the mandibular pair at Stage 37, and finally
the mental pair at Stage 38 (Fig. 2D; Northcutt 2005). This pattern of appearance is consistent
with other catfishes (Ryder 1883). Maxillary barbels are first visible at Stage 31 as lateral
protrusions of the head that are located immediately posterior to the eye and immediately
anterior to the midbrain-hindbrain boundary. As the maxillary and mandibular processes of the
first pharyngeal arch become apparent at Stage 32, the maxillary barbel buds are positioned on
the former processes (Fig. 3B). Throughout embryonic development the maxillary barbels
migrate anteriorly with the mouth but retain their position at the margin of the upper jaw. Given
their location, it is likely that maxillary barbel mesenchyme is composed of first arch neural crest
cells.
The lower jaw barbels (mandibular and mental barbels) first appear several stages after the
maxillary barbels. At Stage 37, the mandibular barbels are first apparent as lateral buds of the
lower jaw. While referred to as mandibular barbels, close examination revealed that these barbels
are not derived from the mandibular pharyngeal arch but are outgrowths of the hyoid arch.
Mental barbels appear at Stage 38 and are also outgrowths of the hyoid arch, first appearing as
small buds that project ventrally from the lower jaw. The nasal barbels only appear later in larval
development, and were not examined in the present study.
10
In the maxillary, mandibular, and mental barbel pairs, outgrowth at first precedes cartilage
formation but later in development these processes are concurrent. Cartilage is visible in the
maxillary barbel by Stage 40, well after the barbel has grown past the lower jaw, and Alcian
Green stains cartilages in the maxillary and mandibular barbel pairs by Stage 42.
Expression of bmp4, fgf8a and shha in developing barbels
FGF8 and SHH play important roles in the development of several vertebrate appendage
types including the paired appendages, unpaired median fins, branchial rays, and genitalia. In the
developing paired appendages, the expression of FGF8 in the distal apical ectodermal ridge
(AER) induces expression of SHH in the zone of polarization activity (ZPA) located in the
posterior appendage bud mesenchyme (Lewandoski et al. 2000). Together, FGF8 and SHH form
a positive feedback loop that is necessary for proper outgrowth throughout development of the
fin/limb (Laufer et al. 1994; Niswander et al. 1994). SHH also acts as a morphogen and is
involved in the anterior/posterior patterning of fins/limbs (Riddle et al. 1993). A similar
mechanism of outgrowth has also been proposed for the development of the unpaired median
fins (Freitas et al. 2006; Abe et al. 2007), chondrichthyian branchial rays (Gillis et al. 2009), and
the genital tubercle of the mouse (Cohn 2004). The latter two structures differ somewhat from
limbs and fins in the details of the mechanism, however. Developing branchial rays express SHH
in a distal thickened epithelial ridge, rather than in a ZPA-type signaling center in the
mesenchyme (Gillis et al. 2009). In genital tubercle development, the FGF8 signaling center (the
urethral epithelium) is derived from endoderm rather than ectoderm, and no AER is present (Liu
et al. 2009). BMP4 also plays an important role in the development of the paired appendages and
the genital tubercle, where BMP signaling inhibits FGF signaling in the absence of BMP
11
inhibitors (Zuniga et al. 1999; Perriton et al. 2002). In the limb bud, BMP signaling plays roles in
outgrowth and proximal-distal patterning, and BMP4 is expressed in the AER and the anterior,
posterior, and distal bud mesenchyme (Pueyo and Couso 2005).
To investigate the involvement of BMP4, FGF8, and SHH in I. punctatus barbel
development, we analyzed the expression of bmp4, fgf8a, and shha via in situ hybridization (Fig.
3). At Stage 29/30, prior to any morphological indication of the maxillary barbel, bmp4 and fgf8a
are expressed in epithelium of first branchial arch where the maxillary barbel bud will first
appear. shha is expressed in the midline of the embryo at this stage, but is not in the region of the
future maxillary barbel. Expression patterns similar to that observed at Stage 29/30 are also
observed in Stages 31 and 32/33, with bmp4 and fgf8a expressed in the maxillary barbel bud
epithelium and shha expression absent from this barbel (Fig. 3, A-C).
Expression of shha is first observed in the maxillary barbel bud at Stage 34/35 (Fig. 3, F, I).
At this stage, both fgf8a and shha are expressed in the maxillary barbel bud epithelium in a
domain that spans the anterior/posterior length of the bud (Fig. 3, E-F), but expression is only
found in the ventral-most two-thirds of the bud (Fig. 3, H-I). bmp4 is also expressed at Stage
34/35 in the maxillary barbel bud epithelium in a domain across the anterior/posterior length of
the bud (Fig. 3D), but this bmp4 expression domain is limited to the dorsal two-thirds of the bud
epithelium (Fig. 3G). Patterns of bmp4, fgf8a, and shha expression similar to those described for
Stage 34/35 are observed at Stage 36 as well (data not shown).
Between Stages 36 and 37, bmp4 expression switches from an epithelial domain to the
anterior mesenchyme of the maxillary barbel, while fgf8a and shha maintain patterns of
epithelial expression similar to that observed at Stages 34/35 and 36 (Fig 3, J-L). However, the
intensity of staining for fgf8a expression in the maxillary barbel diminishes substantially from
12
13
Fig. 3. in situ hybridizations demonstrating expression of bmp4, fgf8a, and shha during I. punctatus barbel development. Dorsal
views with anterior to the left in (A-F, J-L); lateral views with anterior to the left in (M, P-R); Rostral views with dorsal to left in
(N-O). At Stage 32/33 (A-C), bmp4 and fgf8a are expressed in maxillary barbel epithelium (A-B, arrowheads) and shha is absent
from this tissue (C, arrowhead). bmp4 and fgf8a expression patterns at Stage 34/35 are similar to that in Stage 32/33 (D, E), but
shha is now expressed in maxillary barbel epithelium at Stage 34/34 (F, arrowhead). (G-I) Transverse sections of maxillary barbels
demonstrating epithelial expression of bmp4, fgf8a, and shha at Stage 34/35; section plane indicated by red line in (D). bmp4 is
expressed in the dorsal two-thirds of the epithelium (G, arrow) while fgf8a is expressed in the ventral two-thirds of the epithelium
(H, arrow), and shha is expressed in the distal tip of the barbel (I, arrow). At Stage 37 (J-L), bmp4 expression switches to the
maxillary barbel mesenchyme (J) while fgf8a and shha continue to be expressed in the epithelium (K-L). bmp4 continues to be
expressed in the anterior maxillary barbel mesenchyme (M, arrow) in Stage 38, while fgf8a expression is not detected in any of the
barbel pairs (N), and shha is expressed in the anterior maxillary barbel epithelium and mandibular barbel epithelium (O). At Stage
40 (P-R), bmp4 continues to be expressed in the anterior maxillary barbel mesenchyme (P, arrow) and is detected in the mandibular
barbel mesenchyme (P, arrowhead), while fgf8a expression is not detected in the maxillary (Q, arrow) or mandibular (Q,
arrowhead) barbels, and shha is expressed in the anterior epithelium of the maxillary (R, arrow) and mandibular barbels (R,
arrowhead). e, eye; mhb, midbrain-hindbrain boundary.
Stage 36 to Stage 37. From Stage 38 onward, fgf8a is not expressed in any of the barbel pairs
(Fig 3, N, Q).
At Stage 38 the maxillary barbel shha expression domain is confined to the anterior
epithelium (with the exception of the distal tip, where shha is expressed throughout the
epithelium), and this pattern continues throughout the remainder of maxillary barbel outgrowth
(Fig. 3 O, R), as does the expression of bmp4 in the anterior mesenchyme (Fig. 3, M, P).
Expression of bmp4 and shha are first observed in the mandibular barbel buds at Stage 38, with
shha expressed throughout the bud epithelium (Fig. 3O) and bmp4 expressed in the bud
mesenchyme (Fig. 3P).
The shha expression domain in the mandibular barbels is confined to the anterior epithelium
beginning at Stage 39 (Fig. 3R). shha expression is also first seen in the mental barbel bud
epithelium at this stage. It is not until Stage 40 that bmp4 is expressed in the mesenchyme of the
mental barbel buds. In contrast to the maxillary barbel buds, there is no stage of epithelial bmp4
expression of in the developing mandibular and mental barbel buds. While attempts to amplify
shhb from I. punctatus were unsuccessful, expression analysis in the cuckoo catfish, Synodontis
multipunctatus, suggests that this paralog is not expressed in the developing barbels (unpublished
data).
These results demonstrate that barbels are similar to paired appendages, unpaired median
fins, branchial rays, and genitalia in that each of these appendage types express FGF8 and SHH
during outgrowth. However, extensive differences exist in the expression domains of these genes
between barbels and other appendages, including expression in different tissue layers, expression
position, and duration of expression.
14
Expression of Dlx gene family members in the developing barbels
Members of the Dlx gene family are expressed during development in a variety of vertebrate
appendage types, and are also involved in the dorsoventral patterning of the pharyngeal arches
(Depew et al. 2002; Shigetani et al. 2005). In the paired appendages, Dlx genes are expressed in
the AER epithelium of the fin/limb buds (Thomas et al. 2000; Panganiban and Rubenstein 2002).
Dlx genes are also expressed in the developing unpaired median fins (Akimenko et al. 1994;
Ellies et al. 1997; Freitas et al. 2006; Abe et al. 2007). Both epithelial and mesenchymal Dlx
expression is observed in the teeth of fishes and mammals (Thomas et al. 1995; Zhao et al. 2000;
Jackman et al. 2004; Stock et al. 2006). Dlx5 and Dlx6 are expressed in the developing genital
tubercle of mice (Robledo et al. 2002). In mouse, nested expression of Dlx genes acts to pattern
the dorsoventral identity of the pharyngeal arch skeleton: Dlx1/2 are expressed throughout the
dorsal-ventral axis of the pharyngeal arches and specify dorsal identity, Dlx5/6 are expressed in a
ventrally-restricted domain and specify a ventral identity, and Dlx3/4 expression is restricted
more ventrally still (Depew et al. 2002). Similarly, studies in zebrafish have revealed that
dlx1a/2a are expressed throughout the dorsoventral axis of the arches, dlx5a/6a are expressed in
a ventrally restricted domain nested within the dlx1a/2a domain, and dlx3b/4a/4b are expressed
in a ventrally restricted domain nested within the dlx5a/6a domain (Talbot et al. 2010).
The number of Dlx homologs in I. punctatus is unknown, but the closely related zebrafish
possesses eight Dlx gene family members (Stock et al. 1996). Six of these eight Dlx genes form
three bi-gene clusters (dlx1a/dlx2a, dlx3b/dlx4b, and dlx5a/dlx6a), and the two genes within each
cluster are closely co-expressed (Ellies et al. 1997; Qiu et al. 1997). We successfully cloned and
analyzed in I. punctatus the expression of four genes comprising two of the three bi-gene
clusters: dlx1a, dlx2a, dlx5a, and dlx6a.
15
If the expression patterns of Dlx genes in the barbels followed that of the pharyngeal arches,
dlx1a/2a would be expressed in the maxillary and lower jaw barbels, and dlx5a/6a would be
expressed in the lower jaw barbels only. Instead, we found dlx1a, dlx2a, and dlx5a to be
expressed in the developing maxillary barbels (Fig. 4, A-C), while lower jaw barbels express
only dlx2a and dlx5a (Fig. 4, I-J). In contrast to these genes, dlx6a is not expressed in any
barbels at any stage examined (Fig. 4, D).
dlx2a is the first Dlx gene detected in the barbels, and is expressed in the maxillary barbel
bud epithelium at Stage 31, followed by dlx1a in the mesenchyme at Stage 32/33, and dlx5a in
the maxillary barbel epithelium and mesenchyme at Stage 34/35 (Fig. 4, A-C; data not shown).
The tissue layer in which these genes are expressed was verified by sectioning (Fig. 4, E-G).
Once initiated, expression of dlx1a and dlx5a continues in the maxillary barbel mesenchyme
throughout outgrowth of the structure (Fig. 4, H, J). dlx2a, on the other hand, follows a pattern of
expression similar to bmp4 in that dlx2a expression shifts from the epithelium to the
mesenchyme between Stages 36 and 37. From Stage 37 onward, dlx2a is expressed in the
maxillary barbel mesenchyme (Fig. 4I).
The first Dlx gene to be expressed in the mandibular barbel mesenchyme is dlx5a at Stage 38
and is followed by dlx2a at Stage 39 (Fig. 4, I-J). Expression of dlx2a and dlx5a in the mental
barbel bud mesenchyme begins at Stage 40, and expression of dlx2a and dlx5a in the lower jaw
barbel mesenchyme continues throughout barbel outgrowth. dlx1a, however, is not expressed in
the lower jaw barbels in any of the stages examined (Fig. 4H). These results demonstrate that the
developing barbels of I. punctatus deviate from Dlx expression patterns in the pharyngeal arches
as well as expected expression patterns based on Dlx bi-gene cluster co-expression.
16
Fig. 4. Dlx gene expression in the developing barbels shown by in situ hybridization. Dorsal
views with anterior to the left in (A-D). Rostral views with dorsal to left in (H-J). At Stage 34/35
(A-D), dlx1a, dlx2a, and dlx5a are expressed in the maxillary barbels (A-C, arrows), while dlx6a
expression is not detected in the barbels (D, arrow) but is present in the gill cover (D,
arrowhead). (E-F) transverse sections of Stage 34/35 specimens to determine tissue layers of Dlx
gene expression. dlx1a is expressed in the maxillary barbel mesenchyme (E, arrow), while dlx2a
is expressed in the dorsal epithelium of the maxillary barbel (F, arrow), and dlx5a is expressed in
the mesenchyme (G, arrow) and a small portion of the dorsal maxillary barbel epithelium (G,
arrowhead). (H-J) Dlx genes are differentially expressed in the maxillary and lower jaw barbels.
At Stage 38, dlx1a is expressed in the maxillary barbel mesenchyme but is absent from the lower
jaw barbels (H, arrow), while dlx2a and dlx5a are expressed in the mesenchyme of the maxillary
and mandibular barbels (I-J, arrows).
17
Expression of eda and its receptor in the developing barbels
The ectodysplasin signaling pathway includes the extracellular signaling ligand eda, its
receptor edar, and the receptor adaptor protein edaradd, which together are required for the
development of epithelial appendages (teeth, scales, hair, and feathers; Mikkola and Thesleff
2003). Zebrafish with mutations in eda or edar fail to develop teeth, scales, or fin rays (Harris et
al. 2008).
To determine if the ectodysplasin signaling pathway has a role in barbel development, we
examined the expression of eda and edar. Whole mount in situ hybridization for eda produced
ambiguous results, as most specimens displayed low levels of staining throughout the body that
may have been artifactual. Low levels of eda expression in the posterior maxillary barbel bud
mesenchyme were observed at Stage 34/35 (Fig. 5, A, C), but it is not clear that this staining
represents actual localization of the transcript.
Analysis of edar expression proved more successful. At Stages 34/35 and 36, edar is
expressed in the ventral maxillary barbel bud epithelium (Fig. 5, B, C) in a domain similar to
fgf8a. edar continues to be expressed in the maxillary barbel epithelium at Stage 37, but in a
domain that is restricted to the distal portion of the barbel. Expression of edar in the mandibular
barbel bud epithelium is also observed at this stage, and epithelial expression can be observed in
the mental barbel bud by Stage 39 (data not shown).
The expression of eda in the barbel mesenchyme and edar in the barbel epithelium
correspond to the tissue layers of eda and edar expression in developing cichlid teeth and
chicken feathers (Houghton et al. 2005; Fraser et al. 2009). edar is also expressed in the mouse
limb bud AER and developing fin rays of zebrafish paired appendages (Pispa et al. 2003; Harris
et al. 2008).
18
Fig. 5. Barbel expression of ectodysplasin and its receptor demonstrated by in situ hybridization.
Dorsal views with anterior to the left in (A-B). At Stage 34/35, eda appears to be expressed at
low levels in the barbel mesenchyme (A, C, arrows), but this result is questionable due to similar
levels of staining throughout specimens. edar is expressed in the ventral maxillary barbel
epithelium at Stage 34/35 (B, D, arrows).
Other vertebrate appendage development genes not expressed in the barbel
Dach proteins are transcription factors involved in patterning the proximal/distal axis of the
paired appendages (Kida et al. 2004). 5’ members of the HoxA and HoxD complexes play roles
in both the proximal/distal and anterior/posterior patterning of paired appendages (Zakany and
Duboule 2007). dach1, hoxa13b, and hoxd11a were not expressed in the developing barbels of I.
punctatus (Fig. 6, D-F). Detection of dach1, hoxa13b, and hoxd11a expression in the pectoral
fins and dach1 expression in the gill filaments demonstrates that these negative results are
unlikely to be an artifact of the in situ hybridization protocol (Fig. 6, A-C). These results show
that certain developmental genetic mechanisms used to pattern the proximal/distal and
anterior/posterior axes of the paired appendages are not deployed in the developing barbels.
19
Fig. 6. in situ hybridizations demonstrating the absence of hoxa13b, hoxd11a, and dach1 from
the developing barbels. Rostral views with dorsal at top in (D-E). Lateral views with anterior to
the left in (F). At Stage 38, hoxa13b and hoxd11a are expressed in the pectoral fin mesenchyme
(A, B, arrows) while dach1 is expressed in the developing gill filaments (C, arrow). Expression
of none of these genes is detected in the barbels during any of the stages examined, including
Stage 38 (A-C, arrows indicate maxillary barbel).
Hedgehog signaling is required for barbel outgrowth
To test the role of hedgehog signaling in barbel development we exposed I. punctatus to
cyclopamine and SANT-1, two pharmacological inhibitors of the hedgehog pathway (Chen et al.
2002a; Chen et al. 2002b). Treatment with 50 μM cyclopamine beginning at Stage 25 blocked
20
maxillary barbel initiation in all treated individuals (n = 12; Fig. 7A), while control embryos
raised in vehicle alone (1% EtOH) exhibited normal barbel outgrowth (n = 12; Fig. 7B) (p = 7.4
x 10-7, 2-tailed Fisher’s exact test). Embryos were sacrificed once control specimens reached
Stage 36. Identical results were obtained after treatment with 100 μM cyclopamine, 10 μM
SANT-1, and 20 μM SANT-1.
It is not clear if the negative effect of Hh signaling inhibition on maxillary barbel outgrowth
is a direct effect on the barbels themselves or an indirect effect due to the failure to develop
derivatives of the maxillary processes of the mandibular arch. Zebrafish embryos mutant for
shha or treated with hedgehog signaling inhibitors at stages similar to Stage 25 in I. punctatus
fail to form condensations that give rise to the upper jaw (Wada et al. 2005; Eberhart et al. 2007;
Schwend and Ahlgren 2009).
We therefore tested the requirement of hedgehog signaling for barbel outgrowth at later
stages by treating embryos with 50 μM cyclopamine beginning at Stages 35, 37, and 39, and
sacrificing specimens once controls reached Stage 42. Embryos beginning treatment at Stage 35
possessed only diminutive maxillary barbels (n = 19, data not shown), while controls exhibited
well-developed maxillary, mandibular, and mental barbels. Similarly, embryos treated beginning
at Stage 37 had smaller maxillary and mandibular barbels, with the mental pair entirely lacking
(n = 13, data not shown). Embryos beginning treatment at Stage 39 possessed maxillary,
mandibular, and mental barbels that were all of reduced size (n = 22, data not shown) compared
to controls. These results demonstrate the requirement for hedgehog signaling in the initial
growth and continued outgrowth of the maxillary and lower jaw barbels. Such a requirement is
similar between barbels and other vertebrate appendage types (Gillis et al. 2009).
21
Fig. 7. Treatment with pharmacological inhibitors of hedgehog signaling block barbel outgrowth.
Lateral views with anterior to the left in (A-B). Embryos treated with 50 μM cyclopamine
beginning at Stage 25 fail to develop barbels (B, arrowhead) while control embryos display
normal barbel outgrowth at Stage 36 (A, arrow).
22
CHAPTER IV
DISCUSSION
As shown by the results above, the barbels of I. punctatus express genes involved in the
development and patterning of other vertebrate appendage types. Of the 12 genes examined,
eight are expressed in the developing maxillary barbels, and only six of these eight genes are
expressed in the mandibular and mental barbels. The vertebrate appendage development genes
that are expressed in the barbels exhibit both similarities and differences in their expression
domains between the developing barbels and other vertebrate appendage types. The differences
suggest either that barbels did not evolve by the co-option of a complete appendage development
mechanism or that divergence in gene expression occurred following such co-option. As an
example of incomplete co-option, it is possible that barbels arose through the deployment of a
general vertebrate outgrowth program – an FGF8/SHH positive feedback loop.
Maxillary barbels, paired appendages, and branchial rays
Based on gene expression data presented here, barbel development proceeds in a manner that
is more similar to the development of the paired appendages and branchial rays than that of other
appendage types (Fig. 8). FGF8 semi-orthologs are expressed in the ectoderm prior to and during
early outgrowth of the barbels, paired appendages, and branchial rays (Mahmood et al. 1995;
Gillis et al. 2009); however, expression of fgf8a in the maxillary barbels is not confined to an
AER-like pseudostratified ridge of epithelium as it is for the latter two appendage types. Instead,
fgf8a is expressed in the basal portion of a multi-layered epithelium that caps the maxillary
23
Fig. 8. Schematic diagram comparing gene expression between paired appendages and early and
late maxillary barbels. Colored lines to the left of each black and white outline indicate epithelial
expression, while those to the right indicate mesenchymal expression. Gene expression in a
generalized paired appendage bud in (A), early maxillary barbel bud in (B), and a late stage
maxillary barbel in (C). Epi., epithelium; Mes., mesenchyme.
24
barbel bud mesenchyme like an umbrella. An additional discrepancy is that expression of fgf8a
in the maxillary barbels disappears prior to the complete outgrowth of the appendages, while in
the paired appendages and branchial rays FGF8 is expressed throughout outgrowth (Crossley et
al. 1996; Gillis et al. 2009). Barbels, paired appendages, and branchial rays also express SHH
semi-orthologs during development, although this expression is ectodermal in the case of barbels
and branchial rays, but is in lateral plate mesoderm-derived mesenchyme of the developing
paired appendages (Riddle et al. 1993; Gillis et al. 2009). Polarization of SHH semi-ortholog
expression along the anterior/posterior axis is observed in barbels, paired appendages, and
branchial rays; although SHH is anteriorly restricted in barbels and posteriorly restricted in
paired appendages and branchial rays (Riddle et al. 1993; Gillis et al. 2009). While the data
presented here demonstrate the expression of fgf8a and shha in the maxillary barbels, and the
requirement of hedgehog signaling for barbel outgrowth, future studies are needed to assess the
functional interactions of fgf8a and shha in the maxillary barbel bud. The existence of a FGF8SHH positive feedback loop is a hallmark of paired appendage and branchial ray development
(Crossley et al. 1996; Gillis et al. 2009), and investigation into the presence of such a loop in the
maxillary barbel is an important next step in understanding the evolution of barbels and the
developmental genetic mechanisms of barbel outgrowth. I. punctatus is limited in terms of the
experimental tools available to manipulate gene function, but cyclopamine and SU-5402, a
pharmacological inhibitor of FGF signaling (Mohammadi et al. 1997), could be used for such a
study.
Another interesting similarity between the maxillary barbels and paired appendages is the
expression of BMP4 during development. In the early limb bud, BMP4 is expressed in the
anterior, posterior, and distal mesenchyme, as well as the AER (Pueyo and Couso 2005). BMP
25
signaling would inhibit FGF signaling from the AER if it were not for the presence of SHHinduced inhibitors of BMP signaling (Zuniga et al. 1999). In the maxillary barbel bud, epithelial
fgf8a expression disappears shortly after bmp4 expression switches from the epithelium to the
anterior mesenchyme. It would be interesting to functionally investigate this correlated gene
expression in I. punctatus using the BMP signaling inhibitor dorsomorphin (Yu et al. 2008).
A final similar feature between maxillary barbels and paired appendages is the expression of
edar in the epithelium of the developing appendages. edar is expressed in the ventral epithelium
of the maxillary barbel (Fig. 5D) and the AER of the mouse limb bud (Tucker et al. 2000; Pispa
et al. 2003). However, the role of ectodysplasin signaling in paired appendage development is
not well understood (Harris et al. 2008), and it is not clear if the shared edar expression in
developing barbels and paired appendages is functionally significant.
Major points of divergence among barbel and paired appendage development are the
differences in Dlx gene expression patterns between these two appendage types, and the absence
of Dach and 5’ HoxA/HoxD complex genes in the developing barbels. All vertebrate Dlx genes
are expressed in the AER of the paired appendages (Panganiban and Rubenstein 2002), while
Dlx genes are expressed in the epithelium (early stage dlx2a and dlx5a) and the mesenchyme
(dlx1a, late stage dlx2a, dlx5a) of the barbels. In paired appendages, Dach1 and the 5’
HoxA/HoxD complex genes are involved in proximal distal patterning, and the 5’ HoxA/HoxD
complex genes have additional roles in patterning the anterior/posterior axis (Kida et al. 2004;
Zakany and Duboule 2007). The absence of such patterning genes might be expected, given the
lack of extensive morphological differentiation along these axes in barbels, and indicates major
differences between the developmental genetic mechanisms that pattern paired appendages and
those that pattern barbels. Thus, while barbel development shares certain features with the
26
development of other vertebrate appendage types, there are also substantial differences. Taken
altogether, these data suggest that I. punctatus barbels may develop by deploying a general
vertebrate outgrowth program, an FGF8/SHH positive feedback loop, that is found in the
developing paired appendages, unpaired median fins, genitalia, and branchial rays.
The deployment of a general vertebrate outgrowth mechanism to produce a morphological
novelty has been proposed to explain the evolution of genitalia (Liu et al. 2009). According to
this hypothesis, external genitalia have evolved not by co-opting the full limb development
program, but by deploying an AER-like outgrowth genetic cassette in the urethral epithelium.
Evidence for the ability of such a program to drive the outgrowth of head appendages is provided
by FGF8/SHH overexpression experiments in chick embryos and neural crest cultures
(Abzhanov and Tabin 2004). In experiments in vivo, groups of cells co-expressing endogenous
or ectopic FGF8 and SHH produced numerous cartilaginous protrusions on the head, and groups
of neural crest cells in culture which co-expressed FGF8 and SHH were able to form
chondrogenic nodules. The ability to produce outgrowths whenever FGF8 and SHH are both
expressed in the same tissue might be a general feature of the vertebrate head, and would provide
a mechanism to explain the ease with which barbels appear to arise in evolution.
Divergent Dlx gene expression patterns between barbels and pharyngeal arches
Similar to other vertebrate appendages and the pharyngeal arches, barbels were found to
express members of the Dlx gene family. However, Dlx gene expression patterns in the barbels
diverge from that observed in pharyngeal arches and expectations based on Dlx bi-gene cluster
co-expression in the pharyngeal arches and non-barbel vertebrate appendages (Ellies et al. 1997;
Qiu et al. 1997; Depew et al. 2002; Shigetani et al. 2005; Talbot et al. 2010). Based on their
27
positioning along the dorsoventral axis, dlx1a and dlx2a would be expected to be expressed in
the maxillary, mandibular, and mental barbel pairs, if the developing barbels followed the same
Dlx patterning mechanism as the pharyngeal arches (Shigetani et al. 2005). This expectation is
congruent with Dlx bi-gene cluster co-expression, wherein dlx1a/2a are expected to be closely
co-expressed (Ellies et al. 1997; Qiu et al. 1997). However, the developing barbels deviate from
both of these expectations, as dlx1a is expressed in the maxillary barbel mesenchyme but entirely
absent from the lower jaw barbels, while dlx2a is expressed at first in the maxillary barbel
epithelium, then switches to the maxillary barbel mesenchyme, and is expressed in the
mesenchyme of the lower jaw barbels. Barbel dlx5a and dlx6a expression patterns also deviate
from expectations based on Dlx bi-gene cluster co-expression. dlx5a is expressed in the
epithelium and mesenchyme of the maxillary barbels and the mesenchyme of the lower jaw
barbels, while dlx6a is not expressed in any of the barbel pairs. These findings reveal that the
regulation of Dlx genes, which is conserved between mouse and zebrafish, has diverged in I.
punctatus from the time Siluriformes and Cypriniformes split approximately 250 million years
ago (Peng et al. 2006). It will be of interest to examine Dlx gene expression in other catfishes to
investigate whether this change in Dlx gene regulation to specific to I. punctatus or if it is found
across siluriform fishes. If the latter situation is supported, then changes in Dlx gene regulation
might have played an important role in the origin of barbels in catfishes.
Divergent gene expression between maxillary and lower jaw barbels
While six of the genes examined are expressed in the developing maxillary, mandibular, and
mental barbel pairs, two additional genes are only expressed in the maxillary barbels during
development. dlx1a and fgf8a are both expressed in the developing maxillary barbels, but are not
28
expressed in either pair of developing lower jaw barbels. Due to gene duplication events in the
evolution of teleosts (Amores et al. 1998), both Dlx and Fgf gene families display a high degree
of paralogy, and these paralogs may play functionally redundant roles in barbel development.
Therefore, it is possible that another FGF ligand plays the same role in lower jaw barbels that
fgf8a does in the maxillary barbels. Similarly, it is also possible that one of the Dlx family genes
not examined in this study is expressed in the developing lower jaw barbels. However, we cannot
rule out that the lower jaw barbels do not use FGF signaling during development.
Genes that are expressed in both the maxillary and lower jaw barbels exhibit different
expression domains in the different barbel pairs. bmp4 and dlx2a are expressed in the maxillary
barbel epithelium up through Stage 36, and expression of both genes transitions to the
mesenchyme by Stage 37. In contrast, these genes are expressed only in the mesenchyme of the
mandibular and mental barbels. Thus, gene expression patterns in the developing lower jaw
barbels resemble those observed in the maxillary barbels at Stage 38, but have no expression
patterns equivalent to the maxillary barbels at Stage 37 or earlier. The unique expression patterns
found in the maxillary barbel are therefore not required in the development of the lower jaw
barbels.
A potential explanation for the differences in gene expression between the developing
maxillary and lower jaw barbels are the differences in developmental timing and position
between the maxillary, mandibular, and mental barbels. The maxillary barbels are derived from
the mandibular arch, while the lower jaw barbels grow from the hyoid arch. These two arches
have divergent transcriptional profiles, including the presence of Hox complex genes in the
hyoid that are absent from the Hox-default mandibular arch (Shigetani et al. 2005). Such
29
differences in the transcriptional background in which the maxillary and lower jaw barbels
develop may account for the observed differences in gene expression between the barbel pairs.
The origin and evolution of barbels in fishes
Maxillary barbels are invariably present across catfishes, but species may possess from one
to three additional barbel pairs (Burgess 1993; Diogo 2005). The plesiomorphic condition for
catfishes is to possess only the maxillary barbel pair, such as in Diplomystes, and the presence of
nasal or lower jaw barbels in catfishes are derived features (Diogo 2005). While nasal barbels
have been gained and lost numerous times, the presence and number of lower jaw barbels
follows a clear phylogenetic pattern. In the lineage leading to ictalurid catfishes, lower jaw
barbels first appear in the node leading to the non-diplomystid/non-loricarioid clade, where
species transition from having no lower jaw barbels to having both mandibular and mental barbel
pairs (Diogo 2005). In this sense, the order of appearance of barbel pairs during I. punctatus
development roughly recapitulates the order of appearance of barbels throughout catfish
evolutionary history. From this condition of having maxillary, mandibular, and mental barbel
pairs, certain lineages have independently lost one or both pairs of lower jaw barbels (Diogo
2005).
While the specific genetic events and selective pressures leading to the origin of maxillary
barbels in the siluriform common ancestor are not known, a plausible scenario can be proposed.
The presence of barbels in a species frequently correlates with feeding strategies in which
gustatory cues play a critical role in food detection, such as in benthic habitats where vision is
impeded by dark or turbid conditions (Helfman et al. 2009). In such circumstances, additional
sensory cells could provide a selective advantage. External taste buds are present on the lips and
30
heads of many fishes (Herrick 1903; Hansen et al. 2002), and an increase in the surface area of
this epithelium could increase the number of taste buds. The barbels of catfishes may have at
first only been present as smaller, less patterned protrusions of the upper jaw that expanded the
taste bud-bearing epithelium, and were only subsequently modified to become the elaborated
structures observed in modern catfishes. This hypothesis could be tested by examining barbel
gene expression in a catfish group exhibiting the plesiomorphic barbel condition. It may be the
case that the developing barbels of these catfishes display only a subset of the patterning genes
identified in I. punctatus barbels. Contrary to this hypothesis is the observation that Diplomystes
catfishes exhibit the highest density of taste buds along the anterior edge of their maxillary
barbels (Arratia and Huaquin 1995), suggesting that some anterior/posterior patterning
mechanism is already in place in these catfishes.
Given the structural and developmental similarities between the maxillary barbels and the
lower jaw barbels, one potential explanation for the origin of the lower jaw barbels is that they
evolved by co-opting portions of the maxillary barbel developmental program. Indeed, as shown
above, the maxillary and lower jaw barbels share similar gene expression patterns. Interestingly,
several loricarioid lineages have evolved a single pair of lower jaw barbels independently from
the non-diplomystid/non-loricarioid clade, indicating that lower jaw barbels are not historically
homologous across all catfishes (Diogo 2005). One such genus, Corydoras, is available in the pet
trade and has been spawned in captivity by hobbyists. It would be interesting to examine the
development of lower jaw barbels in species of this genus to see if they have evolved by coopting similar or different portions of the maxillary barbel development program as the lower
jaw barbels of the non-diplomystid/non-loricarioid catfishes.
31
One of the first hypotheses concerning the evolutionary relationships between the barbels of
fishes was proposed by Francis Balfour (1880). After describing developing sturgeon barbels and
gar adhesive organs (the latter of which he believed to be barbel rudiments), Balfour concluded,
“…the barbels of fishes must be phylogenetically derived from the papillae of a suctorial disc
adjoining the mouth (p. 89).” In a summary at the end of the chapter, Balfour went on to surmise
that such a disc was “a very primitive vertebrate organ, which has disappeared in the adult state
of almost all the Vertebrata (p. 98),” but still persists in certain bony fishes. Therefore, according
to Balfour’s hypothesis, the barbels of fishes are historically homologous structures inherited
from the vertebrate common ancestor, which have been subsequently lost repeatedly in nonbarbelled species.
John Adams Ryder (1890) rejected Balfour’s barbel homology hypothesis, favoring instead a
hypothesis that barbels are independently derived on the basis that the extensive differences in
position, composition, and developmental timing indicate “a want of community of descent” of
these structures. This is in agreement with the modern consensus, which is based on the scattered
phylogenetic distribution of barbeled species along with Ryder’s arguments. Thus, the barbels of
fishes from different orders are likely to be homoplastic structures produced through convergent
evolution. Given the many independent origins of barbels, we might not expect to observe
barbeled fishes from different orders recruiting the same genes to pattern their barbels during
development.
Based on our current understanding of barbel development, however, we cannot distinguish
between Ryder’s independent origin hypothesis and the reversal hypothesis put forth by H. B.
Pollard (1894; 1895). Similar to Balfour, Pollard proposed that the vertebrate common ancestor
possessed barbels. However, rather than believing that barbels were passed down in an unbroken
32
chain to contemporary barbelled vertebrates, Pollard thought that modern vertebrates with
barbels represent evolutionary reversals to the ancestral condition (Pollard 1895). If this were to
be the case, we would expect barbeled fishes from different orders to use the same ancestral
barbel developmental genetic mechanism. Such a result would be difficult to distinguish from the
parallel evolution of barbels. Only by examining barbel development in non-siluriform fishes can
we begin to investigate these ideas.
We have presented here the first study of gene expression during barbel development in
fishes. In the words of John Ryder, “The most interesting feature of the developmental evolution
of the young catfishes is the early appearance of the barbels (1883).” The more interesting work
is certainly yet to come, when barbel development can be compared and contrasted between
fishes from different orders that display differences in barbel position, composition, and
developmental timing. Loaches (order Cypriniformes), and sturgeons and paddlefish (order
Acipenseriformes), develop barbels early in development and can be spawned in the aquarium or
aquaculture facilities, making them tractable non-catfishes in which to study barbel development
(Bemis and Grande 1992; Snyder 2002; Fujimoto et al. 2006). Recent studies have advanced our
understanding of zebrafish “barbology” (LeClair and Topczewski 2010). Zebrafish barbels do
not appear until juvenile stages, making developmental studies difficult, but certain insights may
be obtained by examining adult barbel morphology in different mutant and transgenic lines. Such
investigations into the evolution and development of fish barbels holds promise to increase our
understanding of the developmental genetic mechanisms underlying morphological novelty and
convergence.
33
BIBLIOGRAPHY
Abe, G., Ide, H., and Tamura, K. 2007. Function of FGF signaling in the developmental
process of the median fin fold in zebrafish. Developmental Biology 304: 355-366.
Abzhanov, A., and Tabin, C. J. 2004. Shh and Fgf8 act synergistically to drive cartilage
outgrowth during cranial development. Developmental Biology 273: 134-148.
Akimenko, M. A., Ekker, M., Wegner, J., Lin, W., and Westerfield, M. 1994. Combinatorial
expression of three zebrafish genes related to distal-less: part of a homeobox gene code
for the head. Journal of Neuroscience. 14: 3475-3486.
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., and Postlethwait, J. H. 1998.
Zebrafish hox clusters and vertebrate genome duplication. Science 282: 1711-1714.
Armstrong, P. B., and Child, J. S. 1962. Stages in the development of Ictalurus nebulosus.
Syracuse University Press, Syracuse.
Arratia, G., and Huaquin, L. 1995. Morphology of the lateral line system and of the skin of
diplomystid and certain primitive loricarioid catfishes and systematic and ecological
considerations. Zoologisches Forschungsinst und Museum Alexander, Bohn.
Balfour, F. M. 1880. A Treatise on Comparative Embryology, Vol. 2. 1st Ed. MacMillan and Co.,
London.
Bemis, W. E., and Grande, L. 1992. Early development of the Actinopterygian head. I. External
development and staging of the paddlefish Polyodon spathula. Journal of Morphology
213: 47-83.
Benjamin, M. 1990. The cranial cartilages of teleosts and their classification. Journal of Anatomy
196: 153-172.
Burgess, W. E. 1993. An Atlas of Freshwater and Marine Catfishes: A Preliminary Survey of the
Siluriformes. TFH Publications, Neptune City.
Chen, J. K., Taipale, J., Young, K. E., Maiti, T., and Beachy, P. A. 2002a. Small molecule
modulation of Smoothened activity. Proceedings of the National Academy of Science
USA 99: 14071-14076.
Chen, J. K., Taipale, J., Cooper, M. K., and Beachy, P. A. 2002b. Inhibition of Hedgehog
signaling by direct binding of cyclopamine to Smoothened. Genes and Development 16:
2743-2748.
34
Chuong, C.-M. 1998. Morphogenesis of epithelial appendages: variations on top of a common
theme and implications in regeneration. In C.-M. Chuong (ed.). Molecular Basis of
Epithelial Appendage Morphogenesis. R. G. Landes, Austin. pp. 3-14.
Cohn, M. J. 2004. Developmental genetics of the external genitalia. Advances in Experimental
Medicine and Biology 545: 149-157.
Crossley, P. H., Minowada, G., MacArthur, C. A., and Martin, G. R. 1996. Roles for FGF8 in the
induction, initiation, and maintenance of chick limb development. Cell 84: 127-136.
Dean, B. 1895. Fishes, living and fossil: an outline of their forms and probable relationships. 1st
Ed. MacMillan and Co., New York.
Depew, M. J., Lufkin, T., and Rubenstein, J. L. 2002. Specification of jaw subdivisions by Dlx
genes. Science 298: 381-385.
Diogo, R. 2005. Morphological evolution, adaptations, homoplasies, constraints and
evolutionary trends: catfishes as a case study on general phylogeny and macroevolution.
Science Publishers, Enfield.
Eberhart, J. K., Swartz, M. E., Crump, J. G., and Kimmel, C. B. 2006. Early hedgehog signaling
from neural to oral epithelium organizes anterior craniofacial development. Development
133: 1069-1077.
Ellies, D. L., Stock, D. W., Hatch, G., Giroux, G., Weiss, K. M., and Ekker, M. 1997.
Relationship between the genomic organization and the overlapping embryonic
expression patterns of the zebrafish dlx genes. Genomics 45: 580-590.
Fox, H. 1999. Barbels and barbel-like tentacular structures in sub-mammalian vertebrates: a
review. Hydrobiologia 403: 153-193.
Fraser, G. J., Hulsey, C. D., Bloomquist, R. F., Uyesugi, K., Manley, N. R., and Streelman, J. T.
2009. An ancient gene network is co-opted for teeth on old and new jaws. PLoS Biology
7: 233-247.
Freitas, R. G., Zhang, G., and Cohn, M. J. 2006. Evidence that mechanisms of fin development
evolved in the midline of early vertebrates. Nature 442: 1033-1037.
Fuiman, L. A. 1984. Ostariophysi: development and relationships. In H. G. Moser (ed.).
Ontogeny and Systematics of Fishes. American Society for Ichthyology and Herpetology
Special Publication 1: 126–137.
Fujimoto, T., Kataoka, T., Sakao, S., Saito, T., Yamaha, E., and Arai, K. 2006. Developmental
stages and germ cell lineage of the loach (Misgurnus anguillicaudatus). Zoological
Science 23: 977-989.
35
Gillis, J. A., Dahn, R. D., and Shubin, N. H. 2009. Shared developmental mechanisms pattern the
vertebrate gill arch and paired fin skeletons. Proceedings of the National Academy of
Science USA 106: 5720-5724.
Grizzle, J. M., and Rogers, W. A. 1976. Anatomy and Histology of the Channel Catfish. Auburn
University Press, Alabama.
Hansen, A., Reutter, K., and Zeiske, E. 2002. Taste bud development in the zebrafish, Danio
rerio. Developmental Dynamics 223: 483-496.
Harris, M. P., Rohner, N., Schwarz, H., Perathoner, S., Konstantinidis, P., and Nusslein-Volhard,
C. 2008. Zebrafish eda and edar Mutants reveal conserved and ancestral roles of
ectodysplasin signaling in vertebrates. PLoS Genetics 4: e1000206.
Helfman, G. S., Collette, B. B., Facey, D. E., and Bowen, B. W. 2009. The Diversity of Fishes:
Biology, Evolution, and Ecology. 2nd Ed. Wiley-Blackwell, West Sussex.
Herrick, C. J. 1903. The organ and sense of taste in fishes. Bulletin of the U.S. Laboratory Fish
Commission, Washington 1902, 22: 237–272.
Houghton, L., Lindon, C., and Morgan, B. A. 2005. The ectodysplasin pathway in feather tract
development. Development 132: 863-872.
Jackman, W. R., Draper, B. W., and Stock, D. W. Fgf signaling is required for zebrafish tooth
development. Developmental Biology 274: 139-157.
Joyce, E. C., and Chapman, G. B. 1978. Fine structure of the nasal barbel of the channel catfish,
Ictalurus punctatus. Journal of Morphology 158: 109-153.
Kardong, K. V. 2008. Vertebrates: Comparative Anatomy, Function, Evolution. 5th Ed. McGraw
Hill Higher Education, Columbus.
Kida, Y. Y., Maeda, Y., Shiraishi, T., Suzuki, T., and Ogura, T. 2004. Chick Dach1 interacts
with the Smad complex and Sin3a to control AER formation and limb development along
the proximodistal axis. Development 131: 4179-4187.
Kim, B., Yabe, M., and Nakaya, K. 2001. Barbels and related muscles in Mullidae (Perciformes)
and Polymixiidae (Polymixiiformes). Ichthyological Research 48: 409-413.
Laufer, E., Nelson, C. E., Johnson, R. L., Morgan, B. A., and Tabin, C. 1993. Sonic hedgehog
and Fgf-4 act through a signaling cascade and feedback loop to integrate growth and
patterning of the developing limb bud. Cell 79: 993-1003.
LeClair, E. E., and Topczewski, J. 2010. Development and regeneration of the zebrafish
maxillary barbel: a novel study system for vertebrate tissue growth and repair. PLoS ONE
5: e8737.
36
Lewandoski, M., Sun, X., and Martin, G. R. 2000. Fgf8 signaling from the AER is essential for
normal limb development. Nature Genetics 26: 460-463.
Lin, C., Yin, Y., Veith, G.M., Fisher, A. V., Long, F., and Ma, L. 2009. Temporal and spatial
dissection of Shh signaling in genital tubercle development. Development 136: 39593967.
Mahmood, R., Bresnick, J., Hornbruch, A., Mahony, C., Morton, N., Colquhon, K., Martin, P.,
Lumsden, A., Dickson, C., and Mason, I. 1995. A role for FGF-8 in the initiation and
maintenance of vertebrate limb bud outgrowth. Current Biology 5: 797-806.
McCormick, M. I. 1993. Development and changes at settlement in the barbel structure of the
reef fish, Upeneus tragula (Mullidae). Environmental Biology of Fishes 37: 269-282.
Mikkola, M. L., and Thesleff, I. 2003. Ectodysplasin signaling in development. Cytokine Growth
Factor Reviews 14: 211-224.
Moczek, A. P. 2008. On the origins of novelty in development and evolution. Bioessays 30: 432447.
Mohammadi, M., McMahon, G., Sun, L., Tang, C., Hirth, P., Yeh, B. K., Hubbard, S. R., and
Schlessinger, J. 1997. Structures of the tyrosine kinase domain of fibroblast growth factor
receptor in complex with inhibitors. Science 276: 955-960.
Müller, G. B., and Wagner, G. P. 1991. Novelty in evolution: restructuring the concept. Annual
Reviews in Ecology & Systematics. 22: 229-256.
Nelson, J. S. 2006. Fishes of the World. 4th Ed. John Wiley & Sons, Hoboken.
Niswander, L., Jeffrey, S., Martin, G. R., and Tickle, C. 1994. A positive feedback loop
coordinates growth and patterning in the vertebrate limb. Nature 371: 609-612.
Northcutt, R. G. 2005. Taste bud development in the channel catfish. Journal of Comparative
Neurology 482: 1-16.
Panganiban, G., Irvine, S. M., Lowe, C., Roehl, H., Corley, L. S., Sherbon, B., Greneir, J. K.,
Kimble, J., Walker, M., Wray, G. A., Swalla, B. J., Martindale, M. Q., and Carroll, S. B.
1997. The origin and evolution of animal appendages. Proceedings of the National
Academy of Science USA 94: 5162-5166.
Panganiban, G., and Rubenstein, J. L. R. 2002. Developmental functions of the Distal-less/Dlx
homeobox genes. Development 129: 4371-4386.
37
Peng, Z. G., He, S., Wang, J., and Diogo, R. 2006. Mitochondrial molecular clocks and the
origin of the major Otocephalan clades (Pisces : Teleostei): A new insight. Gene 370:
113-124.
Perriton, C. L., Powles, N., Chiang, C., Maconochie, M. K., and Cohn, M. J. 2002. Sonic
hedgehog signaling from the urethral epithelium controls external genital development.
Developmental Biology 247: 26-46.
Pispa, J., Mikkola, M. L., Mustonen, T., and Thesleff, I. 2003. Ectodysplasin, Edar and
TNFRSF19 are expressed in complementary and overlapping patterns during mouse
embryogenesis. Gene Expression Patterns 3: 675-679.
Pollard, H. B. 1894. The “cirrhostomial” origin of the head of vertebrates. Anatomische Anzeiger
9: 349–359.
Pollard, H. B. 1895. The oral cirri of silurids and the origin of the head in vertebrates. Zool. Jb.
Anat. Ontog. Tiere 8: 379–424.
Pueyo, J. I., and Couso, J. P. 2005. Parallels between the proximal-distal development of
vertebrate and arthropod appendages: homology without an ancestor? Current Opinion in
Genetics & Development 15: 439-446.
Riddle, R. D., Johnson, R. L., Laufer, E., and Tabin, C. 1993. Sonic hedgehog mediates the
polarizing activity of the ZPA. Cell 75: 1401-1416.
Robledo, R. F., Rajan, L., Li, X., and Lufkin, T. 2002. The D1x5 and D1x6 homeobox genes are
essential for craniofacial, axial, and appendicular skeletal development. Genes and
Development 16: 1089-1101.
Ryder, J. A. 1883. Preliminary notice of the development and breeding habits of the Potomac
catfish, Amiurus albidus (Lesueur) Gill. Bulletin of the United States Fish Commission
15: 225-230.
Ryder, J. A. 1890. The sturgeon and sturgeon industries of the eastern coast of the United States,
with an account of experiments bearing upon sturgeon culture. Extracted from the
Bulletin of the United States Fish Commission, Vol. 8. Government Printing Office,
Washington. pp. 231-281.
Sato, M. 1937. Histological observations on the barbels of fishes. Science Reports of the Tohoku
Imperial University, Sendai, Japan (Fourth Series, Biology) 12: 265–276.
Schwend, T., and Ahlgren, S. C. 2009. Zebrafish con/disp1 reveals multiple spatiotemporal
requirements for Hedgehog-signaling in craniofacial development. BMC Developmental
Biology 9: 59.
38
Shubin, N. H., Tabin, C., and Carrol, S. 2009. Deep homology and the origins of evolutionary
novelty. Nature 457: 818-823.
Starks, E. C. 1904. The osteology of some Berycoid fishes. National Museum Proceedings 27:
601-619.
Shigetani, Y., Sugahara, F., and Kuratani, S. 2005. A new evolutionary scenario for the
vertebrate jaw. BioEssays 27: 331-338.
Snyder, D. E. 2002. Pallid and shovelnose sturgeon larvae – morphological description
and identification. Journal of Applied Ichthyology. 18: 240–265.
Stock, D. W., Ellies, D. L., Zhao, Z., Ekker, M., Ruddle, F. H., Weiss, K. M. 1996. The
evolution of the vertebrate Dlx gene family. Proceedings of the National Academy of
Science USA 93: 10858-10863.
Stock, D. W., Jackman, W. R., and Trapani, J. 2006. Developmental genetic mechanisms of
evolutionary tooth loss in cypriniform fishes. Development 133: 3127-3137.
Talbot, J. C., Johnson, S. L., and Kimmel, C. B. 2010. hand2 and Dlx genes specify dorsal,
intermediate and ventral domains within zebrafish pharyngeal arches. Development 137:
2507-2017.
Thomas, B. L., Porteus, M. H., Rubenstein, J. L., and Sharpe, P. T. 1995. The spatial localization
of Dlx-2 during tooth development. Connective Tissue Research 32: 27-34.
Thomas, B. L., Tucker, A. S., Qui, M., Ferguson, C. A., Hardcastle, Z., Rubenstein, J. L., and
Sharpe, P. T. 1997. Role of Dlx-1 and Dlx-2 genes in patterning of the murine dentition.
Development 124: 4811-4818.
Thomas, B. L., Liu, J. K., Rubenstein, J. L. R., and Sharpe, P.T. 2000. Independent regulation of
Dlx2 expression in the epithelium and mesenchyme of the first branchial arch.
Development 127: 217– 224.
Tucker, A. S., Headon, D. J., Schneider, P., Ferguson, B. M., Overbeek, P., Tschopp, J., and
Sharpe, P. T. 2000. Edar/eda interactions regulate enamel knot formation in tooth
morphogenesis. Development 127: 4691-4700.
Wada, N., Javidan, Y., Nelson, S., Carney, T. J., Kelsh, R. N., and Schilling, T. F. 2005.
Hedgehog signaling is required for cranial neural crest morphogenesis and
chondrogenesis at the midline in the zebrafish skull. Development 132: 3977-3988.
Walker, M. B., and Kimmel, C. B. 2006. A two-color acid-free cartilage and bone stain for
zebrafish larvae. Biotechnic & Histochemistry 82: 23-28.
39
Winemiller, K. O. 1989. Development of dermal lip protuberances for aquatic surface respiration
in South American characid fishes. Copeia 1989: 382-390.
Yu, B. U., Hong, C. C., Sachidanandan, C., Babitt, J. L., Deng, D. Y., Hoyng, S. A., Lin, H. Y.,
Bloch, K. D., and Peterson, R. T. 2008. Dorsomorphin inhibits BMP signals required for
embryogenesis and iron metabolism. Nature Chemical Biology 4: 33-41.
Zakany, J., and Duboule, D. 2007. The role of Hox genes during vertebrate limb development.
Current Opinion in Genetics & Development 17: 359-366.
Zeller, R. 2010. The temporal dynamics of vertebrate limb development, teratogenesis and
evolution. Current Opinion in Genetics & Development 20: 384-390.
Zhao, Z., Stock, D. W., Buchanan, A., Weiss, K. 2000. Expression of Dlx genes during the
development of the murine dentition. Development, Genes, and Evolution. 210: 270-275.
Zuniga, A., Haramis, A. P., McMahon, A. P., and Zeller, R. 1999. Signal relay by BMP
antagonism controls the SHH/FGF4 feedback loop in vertebrate limb buds. Nature 401:
598-602.
40