Abnormal Nuclear Pore Formation Triggers Apoptosis in the

advertisement
GASTROENTEROLOGY 2009;136:902–911
Abnormal Nuclear Pore Formation Triggers Apoptosis in the Intestinal
Epithelium of elys-Deficient Zebrafish
TANYA A. DE JONG–CURTAIN,* ADAM C. PARSLOW,* ANDREW J. TROTTER,* NATHAN E. HALL,*
HEATHER VERKADE,* TANIA TABONE,* ELIZABETH L. CHRISTIE,* MEREDITH O. CROWHURST,* JUDITH E. LAYTON,‡
IAIN T. SHEPHERD,§ SUSAN J. NIXON,储 ROBERT G. PARTON,储 LEONARD I. ZON,¶ DIDIER Y. R. STAINIER,#
GRAHAM J. LIESCHKE,‡ and JOAN K. HEATH*
*Ludwig Institute for Cancer Research, Royal Melbourne Hospital, Parkville, Victoria, Australia; ‡Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria,
Australia; §Department of Biology, Emory University, Atlanta, Georgia; 储Institute for Molecular Bioscience and Centre for Microscopy and Microanalysis, University of
Queensland, Brisbane, Queensland, Australia; ¶Stem Cell Program and Division of Hematology and Oncology, Children’s Hospital, Dana–Farber Cancer Institute,
Howard Hughes Medical Institute and Harvard Medical School, Boston, Massachusetts; #Department of Biochemistry and Biophysics, University of California,
San Francisco, California
BASIC–
ALIMENTARY TRACT
Background & Aims: Zebrafish mutants generated by
ethylnitrosourea-mutagenesis provide a powerful tool
for dissecting the genetic regulation of developmental
processes, including organogenesis. One zebrafish mutant, “flotte lotte” (flo), displays striking defects in intestinal, liver, pancreas, and eye formation at 78 hours
postfertilization (hpf). In this study, we sought to identify the underlying mutated gene in flo and link the
genetic lesion to its phenotype. Methods: Positional
cloning was employed to map the flo mutation. Subcellular characterization of flo embryos was achieved using
histology, immunocytochemistry, bromodeoxyuridine
incorporation analysis, and confocal and electron microscopy. Results: The molecular lesion in flo is a nonsense mutation in the elys (embryonic large molecule derived from yolk sac) gene, which encodes a severely
truncated protein lacking the Elys C-terminal AT-hook
DNA binding domain. Recently, the human ELYS protein
has been shown to play a critical, and hitherto unsuspected, role in nuclear pore assembly. Although elys messenger RNA (mRNA) is expressed broadly during early
zebrafish development, widespread early defects in flo are
circumvented by the persistence of maternally expressed
elys mRNA until 24 hpf. From 72 hpf, elys mRNA expression is restricted to proliferating tissues, including the
intestinal epithelium, pancreas, liver, and eye. Cells in
these tissues display disrupted nuclear pore formation;
ultimately, intestinal epithelial cells undergo apoptosis.
Conclusions: Our results demonstrate that Elys regulates digestive organ formation.
Z
ebrafish provide a powerful genetic model for the identification of molecular mechanisms that establish and
maintain the intestinal epithelium. The zebrafish intestine
has been shown to share many characteristics with its mammalian counterpart,1– 4 and a number of ethylnitrosoureamutagenized zebrafish lines exhibiting abnormalities in intestinal development have been reported.2,5–9 Some of the
underlying mutated genes have recently been identified using positional cloning.6 –9
flotte lotteti262c (flo) was identified on the basis of its
abnormal intestinal morphology at 96 hours postfertilization (hpf).5 An independent focused genetic screen10
carried out on the transgenic Tg(gutGFP)s854 background,
in which all endoderm-derived organs express green fluorescent protein (GFP),11 yielded several additional intestinal mutants, including s871, which exhibits an identical
phenotype to floti262c.
In this study, we provide a comprehensive characterization of the flo phenotype, revealing abnormalities in intestinal, liver, and pancreatic cell organization, growth, and
survival and use positional cloning to identify elys (embryonic large molecule derived from yolk sac) as the mutated
gene in flo. Elys (also known as Mel-2812,13) encodes an
AT-hook DNA binding protein, first discovered in mice,14
which was recently shown to play an indispensable role in
nuclear pore assembly.12,13,15–18 Nuclear pore complexes
(NPCs) are large, dynamic protein assemblies that form
9-nm diameter pores in the nuclear envelope through which
ions and small molecules diffuse and larger molecules, such
as proteins, RNAs, and ribonucleoprotein (RNP) particles,
are actively transported. NPCs comprise multiple copies of
approximately 30 individual components called nucleoporins (Nups), which are disassembled during every round of
cell division.19 The behavior of ELYS and components of
the Nup107–160 complex is particularly interesting. At the
onset of mitosis, these proteins disassociate from the nuclear envelope and localize to the mitotic spindle and attach
to the kinetochores of sister chromatids.12,13,15–17 At the end
of mitosis, they participate in a tightly regulated process to
rebuild the nuclear envelope by providing a scaffold for the
Abbreviations used in this paper: dpf, days postfertilization; Elys,
Embryonic large molecule derived from yolk sac; hpf, hours postfertilization; NPCs, nuclear pore complexes.
© 2009 by the AGA Institute
0016-5085/09/$36.00
doi:10.1053/j.gastro.2008.11.012
recruitment and assembly of other proteins and endoplasmic reticulum-derived membrane lipid components.19,20
The association of ELYS with chromatin appears to provide
a seeding point for the assembly of nucleoporins at the end
of mitosis, thereby compartmentalizing the chromosomes
within the cell.12,13,15–18
Here, we show that nuclear pore formation in the intestinal epithelium, liver, and photoreceptor cell layer of the
eye of flo embryos is severely disrupted, whereas nuclear
pores in most other tissues appear normal. In flo intestinal
epithelial cells, the consequence of aberrant nuclear pore
formation is apoptosis, consistent with an essential role for
Elys during vertebrate organogenesis.
Materials and Methods
Zebrafish Strains, Embryo Collection, and
Confocal and Electron Microscopy
Zebrafish embryos were obtained from natural
spawnings of either wild-type, floti262c (kind gift of the Max–
Planck-Institute of Developmental Biology, Tübingen, Germany), flos871, floti262c;Tg(gutGFP)s854, and floti262c;Tg(nkx2.2a:
mEGFP)1 fish and staged by morphologic criteria.21 All
procedures were approved by the Ludwig Institute for
Cancer Research Animal Ethics Committee. Imaging of
live embryos was carried out after anesthetizing embryos
with 200 mg/L benzocaine (Sigma–Aldrich, St. Louis,
MO) in embryo medium (for further details of microscopy and imaging, see supplemental material online at
www.gastrojournal.org). All images are of the floti262c allele unless otherwise stated and were imported into
CorelDRAW11 (Corel Corporation, Ottawa, Ontario,
Canada). Image manipulation was limited to levels, hue,
and saturation adjustments.
Histology, Immunocytochemistry, and Whole
Mount In Situ Hybridization
Histology and immunocytochemistry were performed
as described.1,22 Mucins and other carbohydrates secreted by
intestinal goblet cells were stained using alcian blue-periodic
acid-Schiff reagent.1 Primary antibodies were mouse anti-BrdU
(BD Biosciences Pharmingen), anti-Hu monoclonal antibody
(mAb) 16A11 (Invitrogen, Carlsbad, CA), and MAb414 (Covance, Princeton, NJ), which recognizes multiple nucleoporins
containing hydrophobic phenylalanine-glycine (FG) repeats,
including components of the Nup107–160 complex. Secondary antibodies were ZyMax horseradish peroxidase-conjugated
goat anti-mouse IgG (Zymed Laboratories, San Francisco, CA)
or Alexa Fluor 488/568 anti-mouse IgGs (Invitrogen). For
whole mount in situ hybridization, embryos were processed as
described22,23 (see supplemental material online at www.
gastrojournal.org).
Genetic Mapping and Positional Cloning of flo
For genetic mapping, floti262c heterozygotes on the
Tübingen (Tü)/TL background were crossed onto the polymorphic WIK strain.24 Mutant embryos were identified by
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
903
eye and intestinal defects that were visible with the stereomicroscope from 78 hpf. Full details of the steps taken
to clone the mutated gene in flo are provided in supplemental material online (see supplemental material online at
www.gastrojournal.org).
Detection of Wild-Type and Mutant elys
RNA Molecules Using Allele-Specific
Reverse-Transcription Polymerase Chain
Reaction
To distinguish between the expression of wild-type
and mutant elys messenger RNA (mRNA) molecules during
the development of flo embryos, total RNA was extracted
from pools of either wild-type or flo embryos at multiple
time points up to 96 hpf. Because wild-type and flo embryos
are only distinguishable by morphology from 78 hpf, genotyping of individual embryos prior to this stage was required prior to pooling. Embryos were individually arrayed
in a 96-well microtiter plate and homogenized in 125 ␮L
Trizol (Invitrogen). After addition of chloroform (25 ␮L),
two phases were separated according to the manufacturer’s
instructions. Genotyping was carried out on DNA extracted
from the lower phase. The corresponding upper phases of
embryos identified as either homozygous wild-type (⫹/⫹)
or homozygous flo were pooled for RNA extraction. The
presence or absence of wild-type and mutant elys mRNA
molecules in wild-type and flo embryos was determined
by allele-specific reverse-transcription polymerase chain
reaction (RT-PCR) using a temperature switch, four
primer system25 (see supplemental material online at
www.gastrojournal.org).
Detection of Cells in the S-Phase of the
Cell Cycle
To identify cells in the S-phase of the cell cycle,
bromodeoxyuridine (BrdU) incorporation by live embryos was analyzed as described.1
Results
flo Embryos Exhibit Gross Defects in
Intestinal, Liver, Pancreas, and Eye
Development
Using differential interference contrast (DIC) microscopy, several gross defects in the development of flo
embryos are evident (Figure 1A and B). The elaborate folding of the wild-type intestinal epithelium seen at 6 and 7
days postfertilization (dpf) is impaired in flo, and the mutant embryos also exhibit smaller eyes (microphthalmia), a
smaller, misshapen head, kyphosis, a rarely inflated swim
bladder, and impaired yolk resorption. The flo phenotype is
completely penetrant, and the animals die from 7 to 8 dpf.
Heterozygous flo carriers are phenotypically indistinguishable from their ⫹/⫹ siblings.
Using histology, we observed that wild-type intestinal
epithelial cells exhibit a regular, columnar morphology at
96 hpf (Figure 1C and E), whereas these cells in flo are
BASIC–
ALIMENTARY TRACT
March 2009
904
DE JONG–CURTAIN ET AL
GASTROENTEROLOGY Vol. 136, No. 3
BASIC–
ALIMENTARY TRACT
Figure 1. Intestinal abnormalities and microphthalmia in flo embryos. DIC images of wild-type
and flo embryos at 6 –7 dpf. (A
and B) The black arrows indicate,
from left to right, the 3 broad regions of the intestine: the intestinal
bulb and midintestine and posterior intestine. (A) The intestinal epithelium in wild-type embryos is
extensively folded (white arrows)
and is thinner and unfolded in flo
(white arrows). (B) In flo, yolk resorption is incomplete, and the
swim bladder rarely inflates. Microphthalmia is evident, and the
head is slightly smaller and misshapen. (C and D, left panels)
Sagittal histologic sections of wildtype and flo embryos at 96 hpf
stained with alcian blue periodic
acid-Schiff reagent showing the
entire intestinal tract. (C and D,
right panels) Transverse sections
of wild-type and flo eyes stained
with methylene blue/azureII show
disrupted cell layers in flo. White
arrow indicates optic nerve. (E and
F) The wild-type intestinal bulb epithelium is elaborating folds (arrow)
at 96 hpf but is thin and unfolded
in flo. Fewer goblet cells (turquoise
staining, arrows) are present in
the pharynx and midintestine (F)
compared with wild-type (E). Detached cells in the intestinal lumen
in flo (F, arrows) are not observed
in wild-type (E). es, esophagus;
gcl, ganglion cell layer; ipl, inner
plexiform layer; inl, inner nuclear
layer; ib, intestinal bulb; le, lens;
opl, outer plexiform layer; pcl,
photoreceptor cell layer; p, pharynx; rpe, retinal pigmented epithelium; sb, swim bladder; y, yolk.
Scale bars, C and D, 10 ␮m; E
and F, 20 ␮m (shown in F, last
panel).
disorganized, less polarized, frequently detached from the
monolayer (Figure 1D and F), and express the active form of
Caspase 3,1 indicating that they are in the execution phase
of apoptosis. Histology also reveals disruption to the 6 cell
layers of the flo retina at 84 hpf: the retinal pigmented
epithelial layer is expanded, and the photoreceptor cell layer
and the outerplexiform layer are largely absent (Figure 1C,
right, and Figure D, right).
At the ultrastructural level, wild-type intestinal epithelial
cells at 96 hpf exhibit tight junctions and a highly elaborated apical brush border, whereas these cells in flo embryos
have poorly developed apicobasal polarity and sparse apical
membrane microvilli, and the intestinal lumen is largely
occluded with cellular debris (see supplemental Figure 1A–E
online at www.gastrojournal.org). To determine the size and
morphology of the other endoderm-derived digestive organs, the flo allele was introduced onto the Tg(gutGFP)s854
background.11 Whereas the liver and pancreas anlagen in flo
and wild-type embryos are morphologically similar at 72
hpf, these organs are severely diminished in size in flo embryos by 98 hpf (see supplemental Figure 2A–D online at
www.gastrojournal.org). Additional defects in flo embryos
include impaired development of ceratobranchial cartilages
3–5 at 72 hpf, revealed using alcian blue staining, and the
absence of rag1-positive thymic lymphocytes at 96 hpf (data
not shown).
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
905
Figure 2. Abnormal cell polarization, nuclear morphology, and differentiation in flo intestinal epithelium. (A) Transverse sections through the intestinal
bulb of flo;Tg(nkx2.2a:mEGFP) embryos at 96 hpf, stained with rhodamine-phalloidin (red) to visualize F-actin and Hoechst 33342 (blue) to visualize
DNA. Compared with wild-type, cellular organization of the intestinal epithelium in flo is disrupted. White arrow in wild-type indicates an Nkx2.2amGFP-positive enteroendocrine cell; in flo the GFP fluorescence is associated with cellular debris in the lumen. (A, right panels) Higher magnification
images show nuclei (white arrows) are misshapen and of more variable size in flo compared with wild-type. (B and C) Fewer nkx2.2a-mEGFPenteroendocrine cells are present in the intestine of flo;Tg(nkx2.2a:mEGFP) embryos (dashed line) compared with wild-type. nkx2.2a-mEGFPpositive cells associated with pronephric ducts (white arrowheads) are unaffected in flo. (D) Lateral and ventral views of the vagal region of 55-hpf
embryos showing similar distribution of phox2b-expressing cells in flo and wild-type as they migrate along the intestine. Ventral views of dissected
intestines from 120-hpf wild-type (E, left panel) and flo (E, right panel) stained with anti-Hu antibody, showing a reduced and abnormal distribution
of differentiated enteric neurons (ENS) in the middle and posterior intestine of flo embryos at 120 hpf (white arrows). The numbers of ENS in a
10-somite midintestinal segment are as follows: wild-type ⫽ 139 ⫾ 3 (mean ⫾ SD); flo ⫽ 93 ⫾ 5 (n ⫽ 4; Student t test, P ⬍ .001). In a 10-somite
segment in the posterior intestine, the numbers are as follows: wild-type ⫽ 74 ⫾ 8; flo ⫽ 39 ⫾ 7 (n ⫽ 4; Student t test, P ⬍ .01).
Loss of All Three Cell Lineages in the
Intestinal Epithelium of flo Embryos
The flo mutation reduced the abundance of all three
cell lineages in the zebrafish intestinal epithelium. In wild-type
embryos, the most abundant cells are the absorptive enterocytes followed by the mucin-secreting goblet cells and the
relatively sparse hormone-secreting enteroendocrine cells. Goblet cells, which populate the pharynx, esophagus, and midintestine of wild-type embryos at 96 hpf, are virtually absent
from the pharynx and midintestine in flo embryos (Figure 1E
and F). Next, the distribution of enteroendocrine cells in flo
embryos was evaluated by introducing the flo allele onto the
Tg(nkx2.2a:mEGFP) background.1 Whereas EGFP-positive en-
teroendocrine cells are scattered throughout wild-type intestinal epithelium at 96 hpf (Figure 2A and B), they are found
rarely in flo intestinal epithelium (Figure 2A and C). Instead,
enhanced GFP (EGFP) fluorescence is largely associated with
apoptotic cells and cellular debris in the intestinal lumen (Figure 2A). To estimate the impact of the flo mutation on the
absorptive enterocyte cells, we counted the number of Hoescht
33342-positive nuclei in EGFP-negative cells in comparable
sections of wild-type and flo embryos through the intestinal
bulb region (Figure 2A). There are approximately 35% fewer
absorptive enterocytes in flo embryos at 78 hpf compared with
wild-type, and, by 96 hpf, the reduction is more severe (65%).
Interestingly, nuclei in the absorptive enterocytes of flo embryos are aberrantly shaped and positioned (Figure 2A).
BASIC–
ALIMENTARY TRACT
March 2009
906
DE JONG–CURTAIN ET AL
GASTROENTEROLOGY Vol. 136, No. 3
BASIC–
ALIMENTARY TRACT
Figure 3. Positional cloning reveals that elys is the mutated gene in flo. (A) Integrated genetic/physical map of chromosome 17 encompassing the flo
locus. Recombinants from 5200 meioses narrowed the genetic interval containing flo to a region spanned by 3 BACs (blue, green, and orange lines)
and 9 genes (red). Fine mapping of the last 2 flanking recombinants defined a 50-kb region (grey) containing a single gene, elys. The nucleotide
sequences of elys cDNAs from floti262 and flos871 embryos contain a C¡T transversion (B) and a T¡A transversion (C), respectively; both create a
premature stop codon. (D) Schematic representation of zebrafish Elys indicating 3 nuclear export sequences (NES), 2 WD40 protein binding domains
(WD), an AT-hook DNA binding domain, and 2 nuclear localization signals (NLS). Arrows indicate premature stop codons encoded by the flos871 and
floti262c alleles at codons 461 and 1319, respectively.
Role of flo in Enteric Neuron Differentiation
and Spreading
The broad impact of the flo mutation on intestinal
development prompted us to examine the behavior of neural crest-derived enteric neurons (ENS). ENS migration
from the vagal region to the intestine was followed using
the pan crest-specific marker crestin (36 hpf) and the ENS
precursor-specific marker phox2b (55 hpf). No differences in
the initial migration of ENS precursors are observed in flo
embryos at 36 hpf (data not shown) or along the length of
the intestine at 55 hpf (Figure 2D). From 96 to 120 hpf,
wild-type Hu-positive, differentiated ENS are spread along
the entire length and circumference of the intestine (Figure
2E), but, in flo, these cells are fewer, particularly in the
posterior intestine, and are restricted to the lateral margins
of the midintestine where they initially migrated in columns
(Figure 2E).3 However, ENS cells are not lost by apoptosis,
as shown by the absence of TUNEL staining at several time
points (data not shown). ENS migration has recently been
shown to depend on Hedgehog signaling from intestinal
endoderm22; therefore, this process may be impaired in flo
embryos because of the abnormal development of this tissue.
flo Encodes Elys, a Protein With an AT-Hook
DNA-Binding Motif
Positional cloning was used to identify the mutant
gene responsible for abnormal digestive organ development
in flo. We mapped the flo locus to a 50-kilobase (kb) interval
on chromosome 17 essentially encompassing a single gene,
elys (Figure 3A; see also supplementary material online at
www.gastrojournal.org). Zebrafish elys spans 28,464 base
pairs (bp) and comprises 36 exons, including a 5= noncoding exon. flo carries a codon 1319 (C¡T) base change in
exon 29 of elys that encodes a premature stop codon (Figure
3B). Individual embryo genotyping demonstrated invariable
linkage of the nonsense mutation to the flo phenotype.
That mutant elys is responsible for the flo phenotype was
confirmed by noncomplementation with an independent flo
allele (flos871) (see supplemental Figure 3A–D online at
www.gastrojournal.org) that was identified in a focused
genetic screen for mutants with defects in endoderm organ
formation.10 flos871 also harbors a nonsense mutation in elys:
a (T¡A) base change in codon 461 (Figure 3C).
Zebrafish elys encodes a protein of 2527 amino acids
(Figure 3D and supplemental Figure 4 online at www.
gastrojournal.org). Three nuclear export signals and two
WD-40 domains originally described in mouse Elys14 are
conserved. Two bipartite nuclear localization signals26 reside in the C-terminal region, and an AT-hook chromatin/
DNA binding sequence27 occurs at position 1994. Both
nonsense mutations are predicted to delete the DNA binding domain and nuclear localization signals. elys is conserved from worms to mammals (see supplemental Figure 5
online at www.gastrojournal.org).
March 2009
Analysis of elys mRNA Expression During
Zebrafish Development
907
the heterozygous mother during oogenesis, we conducted allele-specific RT-PCR. This approach (Figure
4A) allowed us to distinguish between wild-type elys
mRNA expressed maternally and mutant elys mRNA
expressed zygotically in genotyped ⫹/⫹ and flo embryos at multiple time points over the first 96 hpf.
⫹/⫹ embryos express only wild-type mRNA molecules
BASIC–
ALIMENTARY TRACT
The relatively late onset of the flo phenotype (⬎72
hpf) is surprising given the indispensable role of ELYS in
nuclear pore formation. To assess whether the early development of flo embryos is supported by maternal expression
of wild-type elys mRNA molecules deposited into the egg by
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
Figure 4. elys mRNA is expressed maternally and its expression pattern becomes more restricted by 72 hpf. (A) Schematic diagram illustrating the
positions of primers used to analyze elys mRNA expression by allele-specific RT-PCR. Asterisk indicates the position of the C¡T mutation in exon 29 that
provides the basis for allele-specific primer annealing. (B) Wild-type elys is expressed continuously in wild-type embryos as expected. Maternal expression
of wild-type elys mRNA in flo embryos is comparable with that of wild-type embryos until 8 hpf; thereafter, maternal elys expression progressively diminishes
and is barely detectable at 24 hpf. (C) Zygotic expression of mutant elys RNA is induced in flo from 12 hpf. M, DNA markers; DW, distilled water. (D and E)
Whole mount in situ hybridization (WISH) was used to determine the elys mRNA expression pattern from 24 to 96 hpf. (D) elys mRNA is broadly expressed
over the first 52 hpf. (D, left panel) elys expression in the retina (arrow), pharyngeal region (bracket), midbrain, and hindbrain (arrowhead) at 24 hpf. (D, second
panel) elys expression in the forebrain (arrowhead), midbrain-hindbrain boundary (arrow), liver (bracket) at 48 hpf. (D, third panel) elys expression in the fin
buds (arrowhead) and intestine (bracket) at 52 hpf. (D, fourth panel) elys expression in the retina (arrow), pharyngeal region (arrowhead), and liver (bracket).
(E) The elys mRNA expression pattern becomes more restricted at 72–96 hpf. (E, first panel) elys expression in the pharyngeal region (arrowhead), liver, and
anterior intestine (bracket), midbrain-hindbrain boundary (MHB) (arrow), and retina at 72 hpf. (E, second panel) elys expression in the liver and intestinal bulb
(bracket) and pharyngeal region (arrowhead) at 96 hpf. (E, third panel) Transverse section of a WISH embryo counterstained with nuclear fast red showing
elys mRNA in the cytoplasm of intestinal epithelial cells, liver, and pancreas at 96 hpf. (E, fourth panel) Barely detectable expression of elys in comparable
regions of a flo embryo at 96 hpf. Scale bar, D and E, 500 ␮m (except E, third panel).
908
DE JONG–CURTAIN ET AL
(Figure 4B and C) at all time points investigated. Comparable levels of wild-type elys mRNA are also expressed in flo embryos up to 8 hpf (Figure 4B). From 12
hpf, maternal expression of wild-type elys mRNA in flo
embryos progressively wanes until, at 48 hpf, no wildtype elys mRNA remains (Figure 4B). At the same time,
the expression of zygotic mutant elys mRNA molecules
in flo is induced and persists thereafter (Figure 4C).
To identify the specific tissues expressing elys mRNA, we
performed whole mount in situ hybridization analysis. At
24 hpf, elys is widely expressed in the anterior portion of the
embryo including the retina, pharyngeal region, midbrain,
and hindbrain (Figure 4D). At 48 to 52 hpf, elys expression
is present in a small area of the forebrain, the midbrainhindbrain boundary, the pharyngeal region, developing
liver, anterior intestine, and fin buds (Figure 4D). By 72 hpf,
elys expression is restricted to the pharyngeal region, anterior intestine, liver, fin buds, midbrain-hindbrain boundary,
GASTROENTEROLOGY Vol. 136, No. 3
and retina (Figure 4E). At 96 hpf, transverse sections of
whole mount preparations reveal strong expression of elys
mRNA in liver, pancreas, and intestinal epithelial cells (Figure 4E). In contrast, the presence of elys mRNA is barely
discernible in flo embryos at 96 hpf (Figure 4E), possibly
because of nonsense-mediated RNA decay. The discrepancy
here with the data in Figure 4B, in which mutant elys mRNA
is detectable at 96 hpf, may reflect the greater sensitivity of
the RT-PCR method compared with whole mount in situ
hybridization.
The Formation of Nuclear Pores Is Disrupted
in the Intestinal Epithelium of flo Embryos
Recent studies have established that Elys plays an
indispensable role in nuclear pore formation and mitosis.15–17 We therefore analyzed the distribution of NPCs in
various tissues of flo and wild-type embryos using monoclo-
BASIC–
ALIMENTARY TRACT
Figure 5. Aberrant distribution of nuclear pore complexes in the intestinal epithelium and liver of flo embryos at 78 hpf. (A–D) Thick transverse
sections of wild-type and flo embryos stained with rhodamine-phalloidin to detect F-actin (red), Hoechst 33342 to detect DNA (blue), and the
monoclonal antibody MAb414 to detect nuclear pore complex (NPC) proteins (green). Sections (200 ␮m) in the region of the intestinal bulb reveal a
punctate rim of fluorescence around the nuclei of cells (arrow) in wild-type intestinal epithelium (A, A’ [boxed area in A]) and liver (C, C’) but severe
disruption of this pattern in the intestinal epithelium (B, B’) and liver (D, D’) of flo embryos with staining largely associated with large cytoplasmic
aggregates (arrowheads) and cellular debris. (E) Transmission electron microscope (TEM) reveals abundant nuclear pores (arrowheads) in ultrathin
sections of intestinal epithelial cells in the region of the intestinal bulb of wild-type embryos but none in the corresponding tissue in flo. Meanwhile,
nuclear pores are evident in a periderm cell in flo (E, right panel). Scale bars, A and B, 50 ␮m; C and D, 25 ␮m; A’–D’, 10 ␮m; E (left panel), 2 ␮m;
E (center and right), 0.5 ␮m. N, nucleus; ib, intestinal bulb; L, liver; e, esophagus; sb, swim bladder.
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
909
Figure 6. Aberrant distribution
of nuclear pore complexes in
discrete regions of the brain and
eye of flo embryos at 78 hpf.
Transverse sections of wild-type
and flo embryos were analyzed
as described in the legend to
Figure 5. (A, A’) NPC localization
in the wild-type midbrain produces a punctate rim of fluorescence around the nuclei of cells.
A wild-type distribution is also
seen in ventral areas of the flo
midbrain (B, lower inset B’), but,
in neuroproliferative zones (dorsal midbrain and midline between the 2 hemispheres), cytoplasmic aggregates are found
(B, upper inset B’’, arrowheads).
(C and D) Similarly in the eye,
some areas of wild-type NPC localization are seen in flo embryos
(D, left inset D’), but, in the photoreceptor cell layer (D, right inset D’’), cytoplasmic aggregates
predominate. Scale bars, A–D,
50 ␮m; A’–D’’, 10 ␮m.
nal antibody MAb414. We found that the localization of
NPCs in intestinal epithelial cells of flo embryos is severely
disrupted at 78 hpf. Whereas immunofluorescent staining
of NPCs produces a punctate rim of fluorescence around
the circumference of the nucleus in wild-type intestinal
epithelial cells (Figure 5A and A’), the NPCs in intestinal
epithelial cells of flo embryos appear as large aggregates in
the cytoplasm (Figure 5B and B’). Similar aggregates have
been observed in human cell lines depleted of ELYS using
RNAi.5 When analyzed using transmission electron microscopy, the aggregates correspond to annulate lamellae comprising stacks of cytoplasmic pores.15 In flo embryos, detached epithelial cells present in the intestinal lumen
contain profuse cytoplasmic aggregates (Figure 5B’). In rare
cells, where NPCs remain associated with the nuclear rim,
the punctate staining is sparser than in wild-type (Figure
5B’). At 78 hpf, the localization of NPCs is also disrupted in
the flo liver in which cytoplasmic aggregates are abundant
(compare Figure 5C and C’ with Figure 5D and D’).
The ultrastructural appearance of nuclear pores in intestinal epithelial cells of wild-type and flo embryos at 82 hpf
was examined using transmission electron microscopy.
Whereas nuclear pores are readily identifiable in wild-type
intestinal epithelial cells (Figure 5E; approximately 5.1 per
nuclear profile), they are absent in flo intestinal epithelial
cells (Figure 5E). In contrast, other cells in flo embryos, such
as the peridermal cell shown in Figure 5E, display abundant,
apparently normal, nuclear pores (approximately 3.9 per
nuclear profile, compared with 4.4 in wild-type cells).
Outside the digestive system, the localization of NPCs
is relatively unperturbed. However, cytoplasmic aggregates are found in the most dorsally positioned cells in
the midbrain (compare Figure 6A and A’ with Figure 6B,
B’, and B’’) and in the photoreceptor cell layer of the
retinal epithelium (compare Figure 6C and C’ with Figure
D and D’’).
Aberrant Development of flo Embryos Is
Restricted to Proliferating Tissues at 72 hpf
Our observation that the flo phenotype is restricted
to specific organs appears to mirror the tissue-specific elys
gene expression pattern from 72 hpf. This pattern is also
reflected in the cell cycle activity of cells in these organs at
72 hpf. In wild-type and flo embryos, we observed a high
frequency of BrdU-positive cells in the intestinal epithelium
(Figure 7A–C), pharyngeal arches (Figure 7A and B), liver
(Figure 7A), dorsal midbrain, cerebellum, dorsal hindbrain
(Figure 7B and C), retinal epithelium (Figure 7B and C), and
pancreas (Figure 7B). Strikingly, the tissues demonstrating
the highest rates of BrdU uptake are the ones most severely
disrupted in flo embryos. In contrast, unaffected tissues in
flo, such as the somites, heart, and some areas of the brain,
are quiescent.
BASIC–
ALIMENTARY TRACT
March 2009
910
DE JONG–CURTAIN ET AL
GASTROENTEROLOGY Vol. 136, No. 3
Figure 7. Localized proliferative
activity in wild-type and flo embryos at 72 hpf. Sagittal sections
of wild-type and flo zebrafish
embryos showing BrdU-positive
nuclei (brown) in cells captured in
the S-phase of the cell cycle. (A
and B) A high proportion of proliferative cells is observed in the
entire intestinal epithelium (A, arrows; B, fourth panel), dorsal
midbrain (tectum), cerebellum,
dorsal hindbrain (B, first panel),
retinal epithelium (B, second
panel), pharyngeal arches (A, left
arrow; B, third panel), liver (A),
and pancreas (B, fourth panel).
(C) The same tissues are proliferative in flo at 72 hpf. c, cerebellum; hb, hindbrain; ib, intestinal
bulb; L, liver; P, pancreas; pa,
pharyngeal arches; sb, swim
bladder; t, tectum; y, yolk.
BASIC–
ALIMENTARY TRACT
Discussion
Here, we have shown, in the context of a vertebrate
organism, that elys plays an indispensable role during development and specifically during the growth and differentiation of the zebrafish digestive organs. That Elys plays a role
in nuclear pore formation was first suggested when antibodies to components of the Nup107–160 complex
were found to coimmunoprecipitate Xenopus and human
ELYS proteins.15 Further studies demonstrated that the
association of ELYS with chromatin acts as a seeding
point for nucleoporins, one of the earliest events in
nuclear pore assembly.12,13,15–18 In ELYS-depleted HeLa
cells, NPCs fail to reassemble at the end of mitosis,
causing cell cycle arrest and death.15,17 This, together
with the fact that Elys-deficient mouse embryos die
around the time of implantation,28 indicates that Elys is
an essential cellular protein. However, widespread developmental abnormalities are not observed in flo embryos
during early development. This is most likely due to
maternally deposited wild-type elys mRNA molecules,
which persist in flo embryos up to 24 hpf (Figure 4B).
Maternally encoded wild-type Elys protein is likely to
sustain flo embryos for an additional period thereafter.
Eventually, maternal elys expression diminishes to a level
that can no longer support normal development of flo
embryos. At this point, the affected organs are those that
express the highest levels of elys mRNA in wild-type
embryos at 72 hpf, including the digestive organs. Interestingly, BrdU incorporation analysis demonstrates that, at
72 hpf, many of the cells in these same organs are in
S-phase of the cell cycle (Figure 7). Thus, the tissuespecific requirement for elys expression in these organs
may be related to the need to reassemble nuclear pores at
the end of every round of mitosis.
An additional possibility is that Elys plays an active role
in determining tissue-specific gene expression during cell
differentiation. This suggestion is supported by strong evidence emerging from studies in yeast, flies, and mammals
that NPC composition and localization varies among differentiated cell types and during different stages of the cell
cycle and development. For example, mouse embryonic
stem cells deficient for Nup133, a component of the 107–
160 Nup complex that interacts directly with Elys,15 are able
to assemble nuclear pores but lack the capacity to differentiate along a neural pathway.29 Similarly, targeted disruption of Nup50 causes complex neural tube and central
nervous system abnormalities and growth retardation,30
whereas Nup96⫹/⫺ mice have impaired innate and adaptive
immunity because of a compromised ability to properly
express interferon-regulated proteins.31 Whereas studies so
far of Elys function have pointed to a global role in initiating NPC assembly at the end of mitosis, our study is the
first to demonstrate a tissue-specific requirement for elys
during vertebrate development. This raises the question of
whether Elys-containing nuclear pores play a specific role in
coordinating the genetic events that govern the growth and
differentiation of the intestinal epithelium and other digestive organs. Comparison of the tissue-specific expression
profiles of flo embryos and their wild-type siblings may be a
fruitful avenue to address this concept.
Supplementary Data
Note: To access the supplementary material
accompanying this article, visit the online version of
Gastroenterology at www.gastrojournal.org, and at doi:
10.1053/j.gastro.2008.11.012.
References
1. Ng AN, de Jong-Curtain TA, Mawdsley DJ, et al. Formation of the
digestive system in zebrafish: III. Intestinal epithelium morphogenesis. Dev Biol 2005;286:114 –135.
2. Pack M, Solnica-Krezel L, Malicki J, et al. Mutations affecting
development of zebrafish digestive organs. Development 1996;
123:321–328.
3. Wallace KN, Akhter S, Smith EM, et al. Intestinal growth and
differentiation in zebrafish. Mech Dev 2005;122:157–173.
4. Crosnier C, Vargesson N, Gschmeissner S, et al. Delta-Notch
signalling controls commitment to a secretory fate in the zebrafish intestine. Development 2005;132:1093–1104.
5. Chen JN, Haffter P, Odenthal J, et al. Mutations affecting the
cardiovascular system and other internal organs in zebrafish.
Development 1996;123:293–302.
6. Mayer AN, Fishman MC. Nil per os encodes a conserved RNA
recognition motif protein required for morphogenesis and cytodifferentiation of digestive organs in zebrafish. Development 2003;130:
3917–3928.
7. Chen J, Ruan H, Ng SM, et al. Loss of function of def selectively
up-regulates ⌬113p53 expression to arrest expansion growth of
digestive organs in zebrafish. Genes Dev 2005;19:2900 –2911.
8. Wallace KN, Dolan AC, Seiler C, et al. Mutation of smooth muscle
myosin causes epithelial invasion and cystic expansion of the
zebrafish intestine. Dev Cell 2005;8:717–726.
9. Yee NS, Gong W, Huang Y, et al. Mutation of RNA Pol III subunit
rpc2/polr3b leads to deficiency of subunit Rpc11 and disrupts
zebrafish digestive development. PLoS Biol 2007;5:e312.
10. Ober EA, Verkade H, Field HA, et al. Mesodermal Wnt2b signalling
positively regulates liver specification. Nature 2006;442:688 – 691.
11. Field HA, Ober EA, Roeser T, et al. Formation of the digestive system
in zebrafish. I. Liver morphogenesis. Dev Biol 2003;253:279 –290.
12. Galy V, Askjaer P, Franz C, et al. MEL-28, a novel nuclear-envelope
and kinetochore protein essential for zygotic nuclear-envelope assembly in C. elegans. Curr Biol 2006;16:1748 –1756.
13. Fernandez AG, Piano F. MEL-28 is downstream of the Ran cycle
and is required for nuclear-envelope function and chromatin
maintenance. Curr Biol 2006;16:1757–1763.
14. Kimura N, Takizawa M, Okita K, et al. Identification of a novel
transcription factor, ELYS, expressed predominantly in mouse
foetal haematopoietic tissues. Genes Cells 2002;7:435– 446.
15. Rasala BA, Orjalo AV, Shen Z, et al. ELYS is a dual nucleoporin/
kinetochore protein required for nuclear pore assembly and proper
cell division. Proc Natl Acad Sci U S A 2006;103:17801–17806.
16. Franz C, Walczak R, Yavuz S, et al. MEL-28/ELYS is required for the
recruitment of nucleoporins to chromatin and postmitotic nuclear
pore complex assembly. EMBO Rep 2007;8:165–172.
17. Gillespie PJ, Khoudoli GA, Stewart G, et al. ELYS/MEL-28 chromatin association coordinates nuclear pore complex assembly
and replication licensing. Curr Biol 2007;17:1657–1662.
18. Rasala BA, Ramos C, Harel A, et al. Capture of AT-rich chromatin
by ELYS recruits POM121 and NDC1 to initiate nuclear pore
assembly. Mol Biol Cell 2008;19:3982–3996.
19. Hetzer MW, Walther TC, Mattaj IW. Pushing the envelope: structure, function, and dynamics of the nuclear periphery. Annu Rev
Cell Dev Biol 2005;21:347–380.
20. Anderson DJ, Hetzer MW. Nuclear envelope formation by chromatin-mediated reorganization of the endoplasmic reticulum. Nat
Cell Biol 2007;9:1160 –1166.
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
911
21. Kimmel CB, Ballard WW, Kimmel SR, et al. Stages of embryonic
development of the zebrafish. Dev Dyn 1995;203:253–310.
22. Reichenbach B, Delalande JM, Kolmogorova E, et al. Endodermderived Sonic hedgehog and mesoderm Hand2 expression are
required for enteric nervous system development in zebrafish.
Dev Biol 2008;318:52– 64.
23. Thisse C, Thisse B, Schilling TF, et al. Structure of the zebrafish
snail1 gene and its expression in wild-type, spadetail and no tail
mutant embryos. Development 1993;119:1203–1215.
24. Rauch GJ, Granato M, Haffter P. A polymorphic zebrafish line for
genetic mapping using SSLPs on high percentage agarose gels.
Trends Tech Tips Online TO1208, 1997.
25. Tabone T, Hayden MJ. Method of amplifying nucleic acid. US
patent 60/973928. 2007 September.
26. Nakai K, Kanehisa K. A knowledge base for predicting protein localization sites in eukaryotic cells. Genomics 1992;14:897–911.
27. Aravind L, Landsman D. AT-hook motifs identified in a wide
variety of DNA-binding proteins. Nucleic Acids Res 1998;26:
4413– 4421.
28. Okita K, Kiyonari H, Nobuhisa I, et al. Targeted disruption of the
mouse ELYS gene results in embryonic death at peri-implantation
development. Genes Cells 2004;9:1083–1091.
29. Lupu F, Alves A, Anderson K, et al. Nuclear pore composition
regulates neural stem/progenitor cell differentiation in the
mouse embryo. Dev Cell 2008;14:831– 842.
30. Smitherman M, Lee K, Swanger J, et al. Characterization and
targeted disruption of murine Nup50, a p27(Kip1)-interacting
component of the nuclear pore complex. Mol Cell Biol 2000;20:
5631–5642.
31. Faria AM, Levay A, Wang Y, et al. The nucleoporin Nup96 is
required for proper expression of interferon-regulated proteins
and functions. Immunity 2006;24:295–304.
Received April 28, 2008. Accepted November 3, 2008.
Reprint requests
Address requests for reprints to: Joan K. Heath, Associate
Professor, Joint Head, Colon Molecular and Cell Biology Laboratory,
Ludwig Institute for Cancer Research, Post Office Royal Melbourne
Hospital, Parkville, Victoria 3050, Australia. e-mail: joan.heath@
ludwig.edu.au; fax: (613) 9341 3104.
Acknowledgements
The authors thank Annie Ng, Elke Ober, Holly Field, and Michel
Bagnat for the flos871 mutant; Val Feakes (histology); Stephen
Wilson (eye histology); Michel Bagnat and Helen Foote (genome
scans); Bastian Ackermann, Andrew Badrock, Rossana Chung, Stephen
Cody, Rachel Hancock, Sebastian Markmiller, Cam Tu Nguyen, Dora
McPhee, Elsbeth Richardson, and Alt Zantema (technical expertise);
Janna Taylor and Pierre Smith (graphics); and Matthias Ernst and Tony
Burgess (comments on the manuscript).
N.E.H’s current address is Department of Pharmacology, Monash
University, Victoria 3800, Australia.
H.V.’s current address is School of Biological Sciences, Monash
University, Victoria 3800, Australia.
Conflicts of interest
The authors disclose no conflicts.
Funding
The authors disclose the following: Supported by NHMRC Australia
project grants 280916 and 433614 (to J.K.H.); NHMRC Dora Lush
(Biomedical) Scholarship (to T.A.D.J–C.); NHMRC Howard Florey
Centenary Fellowship (to H.V.); Australian Research Council grant
DP0346823 (to G.J.L.); and NIH grants DK 067285 (to I.T.S.),
K058181, and DK060322 (to D.Y.R.S.).
BASIC–
ALIMENTARY TRACT
March 2009
911.e1
DE JONG–CURTAIN ET AL
Supplementary Material
Materials and Methods
Microscopy and Imaging
Fluorescent images of anesthetized floti262c;Tg(nkx2.2a:
mEGFP) embryos were obtained with a Nikon (Tokyo,
Japan) D-Eclipse C1 laser scanning confocal unit attached to a Nikon Eclipse TE2000-E inverted fluorescence microscope. DIC/confocal images of floti262c;Tg(gutGFP)s854 embryos (supplementary Figure 2) were taken
with a Bio-Rad MRC-1000/1024 dual boot laser-scanning
confocal system (Bio-Rad, Hercules, CA) equipped with a
100-mW Argon ion multiline laser attached to an inverted Nikon Eclipse TE-300 microscope.1 For transverse
sections, embryos were fixed in 2% paraformaldehyde
overnight at 4°C, embedded in 4% low melting temperature agarose (Cambrex BioScience, East Rutherford, NJ),
and sectioned at 200 ␮m intervals using a Leica (Solms,
Germany) VT1000S vibrating microtome. Floating sections were stained with 1:150 rhodamine-phalloidin (Biotium, Hayward, CA) and 5 ␮g/mL Hoechst33342 (Sigma–Aldrich, St. Louis, MO) in phosphate-buffered
saline (PBS) containing 0.1% Tween-20 and 0.1% DMSO
overnight at 4°C. For transmission electron microscopy
(TEM), embryos were fixed in 2.5% glutaraldehyde (Electron Microscopy Sciences, Hatfield, PA) in PBS for 2
hours at room temperature, rinsed in PBS, and embedded
in epon according to standard procedures. Transverse
sections were obtained approximately halfway through
the yolk.
Fish Lines, Genetic Mapping, and Positional
Cloning of flotte lotte
For genetic mapping, floti262c heterozygotes on the
Tübingen (Tü)/TL background were crossed onto the
polymorphic WIK strain.2 Mutant embryos were identified by eye and intestinal defects that were visible with
the stereomicroscope from 78 hours postfertilization
(hpf). Segregant analysis was performed on 2 pools of
DNA isolated from 20 homozygous flo mutant larvae and
20 wild-type siblings derived from the same mapping
cross. Amplification of simple sequence length polymorphisms (SSLPs) was achieved using polymerase chain
reaction (PCR), and the amplicons were resolved on 2%
1:1 agarose/MetaPhor agarose (BioWhittaker Molecular
Applications) gels. Flanking SSLP markers were identified, and an expressed sequence tag (EST) within the
region was radiolabelled and used as a probe to screen the
BAC Chori-211 library immobilized on filters (RZPD,
German Resource Center for Genome Research, Berlin,
Germany). Positive BAC clones were identified, and, in
conjunction with the zebrafish genome sequencing
project database (www.sanger.ac.uk/Projects/D_rerio), a
BAC contig spanning the mutant locus was assembled.
Sequencing of the genome encompassed by the BAC
contig, identified single nucleotide polymorphisms
GASTROENTEROLOGY Vol. 136, No. 3
(SNPs) and restriction fragment length polymorphisms
(RFLPs) that were used to narrow the interval to a region
encompassed by three sequenced, overlapping BACs.
Transcripts were predicted on the basis of homology
(for example using GenScan: http://genes/mit.edu/
GENSCAN.html) and by identifying ESTs. SNPs identified in this region resolved the interval to a single gene,
elys (embryonic large molecule derived from yolk sac), also
known as ahctf1 (GenBank accession No, AL954168).
Complementary DNA (cDNA) corresponding to this
gene was amplified by reverse-transcription (RT)-PCR
from homozygous flo embryos and subjected to nucleotide sequencing, which revealed a point mutation that
was invariably linked to the flo phenotype.
Domain and Motif Determination
The predicted protein sequence of zebrafish Elys
was analyzed for known domains and motifs. The AThook motif comprises conserved GRP residues surrounded by a number of basic residues.3 Nuclear localization signals (NLSs) were identified using the bipartite
definition used in PsortII.4
Sequence Alignment and Phylogenetic Tree
An Elys protein sequence alignment was performed using the clustalx program with default parameters. The phylogenetic analysis was carried out on the
most conserved region of the alignment (zebrafish amino
acids 1 to 1349) and was calculated in Clustalx using the
neighbor-joining method with default parameters.5 One
thousand bootstrap trials were performed to determine
the strengths of branch points.
Genotyping
A novel Tsp45I restriction enzyme site created by
the mutation in the floti262c allele provides an RFLP for
genotyping. Primers used to amplify a 482-base pair (bp)
fragment in exon 29 containing the flo mutation were as
follows: forward: 5=TGACATGCATGCCCTCTCTG and
reverse: 5=TAGCTGCTCCTCGCTTACGT.
Allele-Specific RT-PCR
Total RNA (1 ␮g) extracted from pooled embryos
was reverse transcribed using Superscript III (Invitrogen)
and subjected to allele-specific (AS) RT-PCR using a 4
primer system.6 One pair of primers (locus-specific primers) was designed with a high annealing temperature to
amplify a 821-bp fragment of elys (traversing exons 28
and 29) in the region encompassing the flo mutation
during the first phase of PCR (15 cycles at 92°C/30
seconds; 58°C/30 seconds; 72°C/60 seconds): forward
(F1) primer, 5=GCTCCCAAAGGCTCTGTTCA; reverse
(R1) primer, 5=GGTTCACCAACTCCCCCATT. A second
pair of primers (AS primers) was designed with a lower
annealing temperature to amplify a 147-bp fragment
within this region in the second phase of PCR (5 cycles
March 2009
at 92°C/10 seconds; 45°C/30 seconds followed by 15
cycles at 92°C/10 seconds; 53°C/30 seconds). Two sets
of AS primers were designed: one to amplify specifically
the wild-type elys sequence, the other to specifically
amplify the mutant elys sequence: WT forward (AS⫹/⫹)
primer, 5=CTGTTCCCACTCTCGTGC; flo forward (ASflo)
primer, 5=GCCTTCCCACTCTCGTGT, common reverse (R2) primer, 5=GCCCACACTATTGTTTGCTT.
During the first phase of PCR, a high annealing temperature (58°C) permits only the locus-specific primers (F1 and
R1) to anneal to the cDNA and enrich the region containing
the SNP of interest. In the second phase of PCR, the annealing temperature is decreased to 45°C to allow the AS
primers, which contain a 5=tag (underlined) that is noncomplementary to the cDNA template, to bind to the region containing the allele of interest. Once the AS primers
are incorporated into the DNA template, the 5=tag increases
the melting temperature of the amplicon, and the annealing
temperature is increased to 53°C. In these reactions, allele
specificity is conferred by the 3=nucleotide of the AS forward primer (alternative 3=nucleotide shown in italics),
which only anneals in the presence of that allele. The presence of ⫹/⫹ or mutant elys cDNAs is detected by the
presence of the product (147 bp) of the corresponding AS
forward and reverse primer pair. When absent, a PCR fragment corresponding to the product of the locus-specific
forward primer and the common AS reverse primer is observed (306 bp). In samples expressing both alleles, both
products are amplified.
Generation of elys Riboprobes for Whole
Mount In Situ Hydridization
To generate digoxigenin (DIG)-labeled elys riboprobes, 2 cDNA templates corresponding to regions in
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
911.e2
exons 9 and 29 were amplified by RT-PCR using T7 flanked
reverse primers and transcribed with T7 polymerase. Primers used for the 300-bp exon 29 PCR product were as
follows: forward: TGACATGCATGCCCTCTCTG and reverse: taatacgactcactatagggCAGGAGCTGTATGTTCCTCA
(T7 sequence in italics). Primers used for the 225-bp exon
9 PCR product were as follows: forward: AACGGGTGAATCCTTACGGA and reverse: taatacgactcactatagggCAAAGTTGTATGTGGTGGGA. Hybridized riboprobes were
detected using an anti-DIG antibody conjugated to alkaline
phosphatase according to the manufacturer’s instructions
(Roche Diagnostics). The two elys probes gave the same
pattern.
References
1. Cody SH, Xiang SD, Layton MJ, et al. A simple method allowing DIC
imaging in conjunction with confocal microscopy. J Microsc 2005;
217:265–274.
2. Rauch GJ, Granato M, Haffter P. A polymorphic zebrafish line for
genetic mapping using SSLPs on high percentage agarose gels.
Trends Tech Tips Online TO1208, 1997.
3. Aravind L, Landsman D. AT-hook motifs identified in a wide
variety of DNA-binding proteins. Nucleic Acids Res 1998;26:
4413– 4421.
4. Nakai K, Kanehisa K. A knowledge base for predicting protein
localization sites in eukaryotic cells. Genomics 1992;14:897–
911.
5. Thompson JD, Gibson TJ, Plewniak F, et al. The ClustalX windows
interface:flexible strategies for multiple sequence alignment aided
by quality analysis tools. Nucleic Acids Research 1997;25:4876 –
4882.
6. Tabone T, Hayden MJ. Method of amplifying nucleic acid. US
Patent 60 973 928. 2007.
7. Ober EA, Verkade H, Field HA, et al. Mesodermal Wnt2b signalling
positively regulates liver specification. Nature 2006;442:688 –
691.
911.e3
DE JONG–CURTAIN ET AL
GASTROENTEROLOGY Vol. 136, No. 3
Supplementary Figure 1. Ultrastructural characteristics of intestinal epithelium in wild type and flo embryos. (A–C) Wild-type intestinal epithelial
cells at 96 hpf demonstrate well-developed apicobasal polarity as evidenced by basally positioned nuclei (A), tight junctions (B, arrows), and the
elaboration of microvilli projecting from the apical surface into the intestinal lumen (A and B). The nuclei also exhibit conspicuous nuclear pores
(C, arrowheads). (D and E) The intestinal epithelium in flo is highly disorganized, with little evidence of apicobasal polarity. Occasional microvilli are
found on the apical surfaces of cells (E), and the intestinal lumen is largely occluded with cellular debris. Although nuclear pores are not detected on
the nuclear membranes of intestinal epithelial cells of flo embryos at 96 hpf, cells in other tissues, including the periderm (F) do exhibit normal nuclear
pores (arrowheads). Scale bars, A, B, E, 2 ␮m; D, 5 ␮m; C and F, 0.5 ␮m.
March 2009
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
911.e4
Supplementary Figure 2. The liver and pancreas are reduced in size in flo embryos. (A–D) The size and morphology of the digestive organs was
analyzed in wild-type and flo embryos crossed onto the Tg(gutGFP)s854 background.7 The normal pattern of GFP fluorescence in wild-type embryos
is seen in panels A and C, which reveal the morphology of the liver and pancreas, in left and right lateral views, respectively, at 98 hpf. At the same
time, the liver and pancreas anlagen in flo embryos are barely discernible (B and D). There was also less GFP fluorescence associated with the
intestine in flo (B and D) and detached, GFP-positive cells were present in the lumen (B, arrows).
911.e5
DE JONG–CURTAIN ET AL
Supplementary Figure 3. The impaired development of the digestive
organs and the eye in flo embryos is not complemented by crossing
onto the s871 background. (A–D) Gross morphologic appearance of
wildtype, floti262c, s871, and floti262c ⫻ s871 embryos at 98 hpf. Crossing
heterozygous floti262c carriers with carriers harboring an independent
allele s871, which was identified in a screen for mutants with abnormalities in endoderm organ formation,6 results in the generation of embryos
exhibiting the flo phenotype (D) in one quarter of the offspring. Particularly conspicuous features of the flo phenotype are the smaller eye,
which is reduced in diameter by approximately 25% (brackets), and the
lack of folding of the intestinal bulb epithelium. These data indicate that
both alleles correspond to the same genetic locus, ie, s871 is a new
allele of flo. e, eye; ib, intestinal bulb; sb, swim bladder; y, yolk.
GASTROENTEROLOGY Vol. 136, No. 3
March 2009
ELYS AND ZEBRAFISH INTESTINAL DEVELOPMENT
911.e6
Supplementary Figure 4. Alignment of the zebrafish Elys protein with human ELYS and mouse Elys. The zebrafish (Danio rerio) Elys protein
comprises 2527 amino acids, compared with 2188 in human (Homo sapiens) and 2243 in mouse (Mus musculus). It shares 38% identity (54%
similarity) with its mammalian orthologs, with most identity in the N-terminal 1–1240 residues. Known functional domains are shaded in grey,
including 3 nuclear export sequences (NES), 2 WD-40 protein-protein interaction domains (WD), an AT-hook DNA binding domain, and 2 bipartite
nuclear localization signals (NLS). Also shaded in grey are the positions of 2 premature stop codons (STOP) in the flos871 and floti262c alleles, at codons
461 and 1319, respectively. Sequences used: zebrafish (Danio rerio) (D.r.): our prediction from the ORF on BAC zK3I24; Homo sapiens (H.s.);
XM_001126456; Mus musculus (M.m.); NM_026375.
911.e7
DE JONG–CURTAIN ET AL
Supplementary Figure 5. Phylogenetic tree of the Elys family of proteins. The unrooted phylogenetic tree is based on a clustalx alignment
(using a conserved region of the zebrafish sequence 1–1349) of the
following sequences: hs (Homo sapiens) NP_056261.3; pt (Pan troglodytes) XP_514319.2; mm (Macaca mulatta) XP_001087893.1; mouse
(Mus musculus) NP_080651.1; rn (Rattus norvegicus) XP_341162.3; ec
(Equus caballus) XP_001490369.1; bt (Bos taurus) XP_580880.2; cf
(Canis familiaris) XP_537228.2; oa (Ornithorhynchus anatinus)
XP_001514988.1; md (Monodelphis domestic) XP_001377258.1; gg
(Gallus gallus) XP_419532.1; xl (Xenopus laevis) AAH86281.1; tn (Tetraodon nigroviridis) CAG11033.1; dr (Danio rerio) predicted from BAC
zK3I24; nv (Nasonia vitripennis) XP_001605706.1; am (Apis mellifera)
XP_001122084.1; ag (Anopheles gambiae str.) EAA07743.2; dm (Drosophila melanogaster) NP_608340.2; cb(b) (Caenorhabditis briggsae)
XP_001673587.1; cb(a) (Caenorhabditis briggsae) XP_001665636.1;
ce (Caenorhabditis elegans) NP_497987.2; One thousand bootstrap
trials were calculated (results displayed at the branch points) with the
neighbor-joining method used to determine the strengths of the phylogenetic relationships.
GASTROENTEROLOGY Vol. 136, No. 3
Download