Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Oct 22.
Published in final edited form as: Gastroenterology. 2008 Nov 6;136(3):902–911. doi: 10.1053/j.gastro.2008.11.012

Abnormal Nuclear Pore Formation Triggers Apoptosis in the Intestinal Epithelium of elys-Deficient Zebrafish

Tanya A de Jong-Curtain 1, Adam C Parslow 1, Andrew J Trotter 1, Nathan E Hall 1,7, Heather Verkade 1,8, Tania Tabone 1, Elizabeth L Christie 1, Meredith O Crowhurst 1, Judith E Layton 2, Iain T Shepherd 3, Susan J Nixon 4, Robert G Parton 4, Leonard I Zon 5, Didier Y R Stainier 6, Graham J Lieschke 2, Joan K Heath 1,*
PMCID: PMC3804769  NIHMSID: NIHMS506140  PMID: 19073184

Abstract

Background & Aims

Zebrafish mutants generated by ethylnitrosourea (ENU)-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 78hpf. 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. Sub-cellular characterization of flo embryos was achieved using histology, immunocytochemistry, bromodeoxyuridine incorporation analysis, 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, ELYS has been shown to play a critical, and hitherto unsuspected, role in nuclear pore assembly. Though elys mRNA is expressed broadly during early zebrafish development, widespread early defects in flo are circumvented by the persistence of maternally-expressed elys mRNA until 24hpf. From 72hpf, 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.

Conclusion

Our results demonstrate that Elys regulates digestive organ formation.

Introduction

Zebrafish 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 counterpart1–4 and a number of ethylnitrosourea (ENU)-mutagenized zebrafish lines exhibiting abnormalities in intestinal development have been reported2, 5–9. Some of the underlying mutated genes have recently been identified using positional cloning6–9.

flotte lotteti262c (flo), was identified on the basis of its abnormal intestinal morphology at 96 hours post-fertilization (hpf)5. An independent focused genetic screen10, carried out on the transgenic Tg(gutGFP)s854 background, in which all endoderm-derived organs express GFP11, 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 mouse14, which was recently shown to play an indispensable role in nuclear pore assembly12, 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 division19. 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 chromatids12, 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 recruitment and assembly of other proteins and endoplasmic reticulum-derived membrane lipid components19, 20. The association of ELYS with chromatin appears to provide a seeding point for the assembly of Nups at the end of mitosis, thereby compartmentalizing the chromosomes within the cell12, 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, while 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, 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 morphological criteria21. All procedures were approved by the Ludwig Institute for Cancer Research Animal Ethics Committee. Imaging of live embryos was carried out after anaesthetizing embryos with 200mg/L benzocaine (Sigma-Aldrich) 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. Image manipulation was limited to levels, hue, and saturation adjustments.

Histology, immunocytochemistry and wholemount in situ hybridization

Histology and immunocytochemistry were performed as described1, 22. Mucins and other carbohydrates secreted by intestinal goblet cells were stained using alcian blue-periodic acid Schiff reagent1. Primary antibodies were mouse anti-BrdU (BD Biosciences Pharmingen), anti-Hu mAb 16A11 (Molecular Probes) and MAb414 (Covance), which recognizes multiple Nups containing hydrophobic phenylalanine-glycine (FG) repeats, including components of the Nup107–160 complex. Secondary antibodies were ZyMax HRP-conjugated goat anti-mouse IgG (Zymed Laboratories) or Alexa Fluor 488/568 anti-mouse IgGs (Molecular Probes). For wholemount in situ hybridization (WISH), 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 strain24. Mutant embryos were identified by eye and intestinal defects that were visible with the stereomicroscope from 78hpf. Full details of the steps taken to clone the mutated gene in flo are provided in Supplemental Material online at www.gastrojournal.org.

Detection of wildtype and mutant elys RNA molecules using allele-specific (AS) RT-PCR

To distinguish between the expression of wildtype and mutant elys mRNA molecules during the development of flo embryos, total RNA was extracted from pools of either wildtype or flo embryos at multiple time-points up to 96hpf. Since wildtype and flo embryos are only distinguishable by morphology from 78hpf, 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 125µl Trizol (Invitrogen). After addition of chloroform (25µl), 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 wildtype (+/+) or homozygous flo were pooled for RNA extraction. The presence or absence of wildtype and mutant elys mRNA molecules in wildtype and flo embryos was determined by allele-specific (AS) 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 described1.

Results

flo embryos exhibit gross defects in intestinal, liver, pancreas and eye development

Using DIC microscopy, several gross defects in the development of flo embryos are evident (Figure 1A and 1B). The elaborate folding of the wildtype intestinal epithelium seen at 6 and 7 days post-fertilization (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 7–8dpf. Heterozygous flo carriers are phenotypically indistinguishable from their +/+ siblings.

Figure 1. Intestinal abnormalities and microphthalmia in flo embryos.

Figure 1

Open in a new tab

DIC images of wildtype and flo embryos at 6–7dpf. (A, B) The black arrows indicate, from L→R, the three broad regions of the intestine: the intestinal bulb, mid-intestine and posterior intestine. (A) The intestinal epithelium in wildtype 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, D left panels) Sagittal histological sections of wild-type and flo embryos at 96hpf stained with alcian blue periodic acid Schiff reagent showing the entire intestinal tract. (C, D right panels) Transverse sections of wildtype and flo eyes stained with methylene blue/azureII show disrupted cell layers in flo. White arrow indicates optic nerve. (E, F) The wildtype intestinal bulb epithelium is elaborating folds (arrow) at 96hpf, but is thin and unfolded in flo. Fewer goblet cells (turquoise staining, arrows) are present in the pharynx and mid-intestine (F) compared to wildtype (E). Detached cells in the intestinal lumen in flo (F, arrows) are not observed in wildtype (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, D 10µm; E, F 20µm (shown in F, last panel).

Using histology we observed that wildtype intestinal epithelial cells exhibit a regular, columnar morphology at 96hpf (Figure 1C and 1E), whereas these cells in flo are disorganized, less polarized, are frequently detached from the monolayer (Figure 1D and 1F) and express the active form of Caspase 31, indicating that they are in the execution phase of apoptosis. Histology also reveals disruption to the six cell layers of the flo retina at 84hpf: the retinal pigmented epithelial layer is expanded and the photoreceptor cell layer and the outerplexiform layer are largely absent (Figure 1C, right and Figure 1D, right).

At the ultrastructural level, wildtype intestinal epithelial cells at 96hpf exhibit tight junctions and a highly elaborated apical brush border while these cells in flo have poorly developed apicobasal polarity, sparse apical membrane microvilli and the intestinal lumen is largely occluded with cellular debris (see Supplemental Figure 1A–1E 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 background11. Whereas the liver and pancreas anlagen in flo and wildtype embryos are morphologically similar at 72hpf, these organs are severely diminished in size in flo by 98hpf (see Supplemental Figure 2A–D online at www.gastrojournal.org).

Additional defects of flo include impaired development of ceratobranchial cartilages 3–5 at 72hpf, revealed using alcian blue staining, and the absence of rag1-positive thymic lymphocytes at 96hpf (data not shown).

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 wildtype 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 mid-intestine of wildtype embryos at 96hpf are virtually absent from the pharynx and mid-intestine in flo (Figure 1E and 1F). Next, the distribution of enteroendocrine cells in flo was evaluated by introducing the flo allele onto the Tg(nkx2.2a:mEGFP) background1. While EGFP-positive enteroendocrine cells are scattered throughout wildtype intestinal epithelium at 96hpf (Figure 2A and 2B), they are found rarely in flo (Figure 2A and 2C). Instead, 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 Hoescht33342-positive nuclei in EGFP-negative cells in comparable sections of wildtype and flo embryos through the intestinal bulb region (Figure 2A). Interestingly, nuclei in the absorptive enterocytes of flo embryos are aberrantly shaped and positioned (Figure 2A). There are approximately 35% fewer absorptive enterocytes in flo embryos at 78hpf compared to wildtype and by 96hpf, the reduction is more severe (65%).

Figure 2. Abnormal cell polarization, nuclear morphology and differentiation in flo intestinal epithelium.

Figure 2

Open in a new tab

(A) Transverse sections through the intestinal bulb of flo;Tg(nkx2.2a:mEGFP) embryos at 96hpf, stained with rhodamine-phalloidin (red) to visualize F-actin and Hoechst 33342 (blue) to visualize DNA. Compared to wildtype, cellular organization of the intestinal epithelium in flo is disrupted. White arrow in wildtype indicates an Nkx2.2a-mGFP-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 to wildtype. (B, C) Fewer nkx2.2a-mEGFP-enteroendocrine cells are present in the intestine of flo;Tg(nkx2.2a:mEGFP) embryos (dashed line) compared to wildtype. nkx2.2a-mEGFP positive cells associated with pronephric ducts (white arrowheads) are unaffected in flo. (D) Lateral and ventral views of the vagal region of 55hpf embryos showing similar distribution of phox2b-expressing cells in flo and wildtype as they migrate along the intestine. Ventral views of dissected intestines from 120hpf wildtype (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 mid and posterior intestine of flo embryos at 120hpf (white arrows). The numbers of ENS in a 10-somite mid-intestinal segment are: wildtype = 139 ± 3 (mean ± S.D); flo = 93 ± 5 (n = 4; Student’s t-test P<0.001). In a 10-somite segment in the posterior intestine, the numbers are: wildtype = 74 ± 8; flo = 39 ± 7 (n = 4; Student’s t-test P<0.01).

Role of flo in enteric neuron differentiation and spreading

The broad impact of flo 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 (36hpf) and the ENS precursor-specific marker, phox2b (55hpf). No differences in the initial migration of ENS precursors are observed in flo at 36hpf (data not shown) or along the length of the intestine at 55hpf (Figure 2D). From 96–120hpf, wildtype 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 mid-intestine 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; this process may be impaired in flo due to 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 50kb interval on chromosome 17 essentially encompassing a single gene, elys (Figure 3A; see also Supplementary Material online at www.gastrojournal.org). Zebrafish elys spans 28,464bp and comprises 36 exons, including a 5’ non-coding exon. flo carries a codon 1319 (C→T) base change in exon 29 of elys that encodes a premature stop codon (Fig 3B). Individual embryo genotyping demonstrated invariable linkage of the nonsense mutation to the flo phenotype.

Figure 3. Positional cloning reveals that elys is the mutated gene in flo.

Figure 3

Open in a new tab

(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 nine genes (red). Fine mapping of the last two flanking recombinants defined a 50kb region (grey) containing a single gene, elys. The nucleotide sequences of elys cDNAs from floti262 and flos871embryos 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 three nuclear export sequences (NES), two WD40 protein binding domains (WD), an AT-hook DNA binding domain and two nuclear localization signals (NLS). Arrows indicate premature stop codons encoded by the flos871 and floti262c alleles at codons 461 and 1319, respectively.

That mutant elys is responsible for the flo phenotype was confirmed by non-complementation 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 formation10. 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).

Analysis of elys mRNA expression during zebrafish development

The relatively late onset of the flo phenotype (>72hpf) 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 wildtype elys mRNA molecules deposited into the egg by the heterozygous mother during oogenesis, we conducted allele-specific (AS) RT-PCR. This approach (Figure 4A) allowed us to distinguish between wildtype elys mRNA expressed maternally and mutant elys mRNA expressed zygotically in genotyped +/+ and flo embryos at multiple time-points over the first 96hpf. +/+ embryos express only wildtype mRNA molecules (Figure 4B and 4C) at all time-points investigated. Comparable levels of wildtype elys mRNA are also expressed in flo embryos up to 8hpf (Figure 4B). From 12hpf, maternal expression of wildtype elys mRNA in flo embryos progressively wanes until at 48hpf 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).

Figure 4. elys mRNA is expressed maternally and its expression pattern becomes more restricted by 72hpf.

Figure 4

Open in a new tab

(A) Schematic diagram illustrating the positions of primers used to analyze elys mRNA expression by allele-specific (AS) RT-PCR. Asterisk indicates the position of the C→T mutation in exon 29 that provides the basis for AS primer annealing. (B) Wildtype elys is expressed continuously in wildtype embryos as expected. Maternal expression of wildtype elys mRNA in flo embryos is comparable to that of wildtype embryos until 8hpf; thereafter maternal elys expression progressively diminishes and is barely detectable at 24hpf. (C) Zygotic expression of mutant elys RNA is induced in flo from 12hpf. M=DNA markers, DW=distilled water. (D, E) WISH was used to determine the elys mRNA expression pattern from 24–96hpf. (D) elys mRNA is broadly expressed over the first 52hpf. (D, left panel) elys expression in the retina (arrow), pharyngeal region (bracket), midbrain and hindbrain (arrowhead) at 24hpf. (D, second panel) elys expression in the forebrain (arrowhead), midbrain-hindbrain boundary (arrow), liver (bracket) at 48hpf. (D, third panel) elys expression in the fin buds (arrowhead) and intestine (bracket) at 52hpf. (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 72–96hpf. (E, first panel) elys expression in the pharyngeal region (arrowhead), liver and anterior intestine (bracket), MHB (arrow) and retina at 72hpf. (E, second panel) elys expression in the liver and intestinal bulb (bracket) and pharyngeal region (arrowhead) at 96hpf. (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 96hpf. (E, fourth panel) Barely detectable expression of elys in comparable regions of a flo embryo at 96hpf. Scale bar: D, E = 500µm (except E, third panel).

To identify the specific tissues expressing elys mRNA, we performed WISH analysis. At 24hpf, elys is widely expressed in the anterior portion of the embryo including the retina, pharyngeal region, midbrain and hindbrain (Figure 4D). At 48–52hpf elys expression is present in a small area of the forebrain, the midbrain-hindbrain boundary (MHB), the pharyngeal region, developing liver, anterior intestine and fin buds (Figure 4D). By 72hpf, elys expression is restricted to the pharyngeal region, anterior intestine, liver, fin buds, MHB and retina (Figure 4E). At 96hpf, transverse sections of wholemount 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 96hpf (Figure 4E), possibly due to nonsense-mediated RNA decay. The discrepancy here with the data in Figure 4B, where mutant elys mRNA is detectable at 96hpf, may reflect the greater sensitivity of the RT-PCR method compared to WISH.

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 mitosis15–17. We therefore analyzed the distribution of NPCs in various tissues of flo and wildtype embryos using monoclonal antibody, MAb414. We found that the localization of NPCs in intestinal epithelial cells of flo embryos is severely disrupted at 78hpf. Whereas immunofluorescent staining of NPCs produces a punctate rim of fluorescence around the circumference of the nucleus in wildtype intestinal epithelial cells (Figure 5A and 5A’), the NPCs in intestinal epithelial cells of flo appear as large aggregates in the cytoplasm (Figure 5B and 5B’). Similar aggregates have been observed in human cell lines depleted of ELYS using RNAi5. When analyzed using TEM, the aggregates correspond to annulate lamellae comprising stacks of cytoplasmic pores15. In flo, 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 wildtype (Figure 5B’). At 78hpf, the localization of NPCs is also disrupted in the flo liver where cytoplasmic aggregates are abundant (compare Figure 5C and 5C’ with Figure 5D and 5D’).

Figure 5. Aberrant distribution of nuclear pore complexes in the intestinal epithelium and liver of flo embryos at 78hpf.

Figure 5

Open in a new tab

(A–D) Thick transverse sections of wildtype 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 µm) in the region of the intestinal bulb reveal a punctate rim of fluorescence around the nuclei of cells (arrow) in wildtype 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) TEM reveals abundant nuclear pores (arrowheads) in ultrathin sections of intestinal epithelial cells in the region of the intestinal bulb of wildtype 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, B = 50 µm; C, D = 25 µm; A’–D’ = 10 µm; E (left panel) = 2 µm; E (centre and right) = 0.5 µm. (N) nucleus; (ib) intestinal bulb; (L) liver; (e) esophagus; (sb) swim bladder.

The ultrastructural appearance of nuclear pores in intestinal epithelial cells of wildtype and flo embryos at 82hpf was examined using TEM. Whereas nuclear pores are readily identifiable in wildtype 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 to 4.4 in wildtype 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 6A’ with Figure 6B, 6B’ and 6B’’) and in the photoreceptor cell layer of the retinal epithelium (compare Figure 6C and C’ with Figure 6D and 6D’’).

Figure 6. Aberrant distribution of nuclear pore complexes in discrete regions of the brain and eye of flo embryos at 78hpf.

Figure 6

Open in a new tab

Transverse sections of wildtype and flo embryos were analyzed as described in the legend to Figure 6. (A, A’) NPC localization in the wildtype midbrain produces a punctate rim of fluorescence around the nuclei of cells. A wildtype 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 two hemispheres), cytoplasmic aggregates are found (B, upper inset B’’, arrowheads). (C, D) Similarly in the eye, some areas of wildtype 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 µm, A’–D’’ = 10 µm.

Aberrant development of flo embryos is restricted to proliferating tissues at 72hpf

Our observation that the flo phenotype is restricted to specific organs appears to mirror the tissue-specific elys gene expression pattern from 72hpf. This pattern is also reflected in the cell cycle activity of cells in these organs at 72hpf. In wildtype and flo embryos, we observed a high frequency of BrdU-positive cells in the intestinal epithelium (Figure 7A–7C), pharyngeal arches (Figure 7A and 7B), liver (Figure 7A), dorsal midbrain, cerebellum, dorsal hindbrain (Figure 7B and 7C), retinal epithelium (Figure 7B and 7C) and pancreas (Figure 7B). Strikingly, the tissues demonstrating the highest rates of BrdU uptake are the ones most severely disrupted in flo. In contrast, unaffected tissues in flo, such as the somites, heart and some areas of the brain are quiescent.

Figure 7. Localized proliferative activity in wildtype and flo embryos at 72hpf.

Figure 7

Open in a new tab

Sagittal sections of wildtype and flo zebrafish embryos showing BrdU-positive nuclei (brown) in cells captured in the S-phase of the cell cycle. (A–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), pancreas (B, fourth panel). (C) The same tissues are proliferative in flo at 72hpf. (c) cerebellum; (hb) hindbrain; (ib) intestinal bulb, (L) liver; (P) pancreas; (pa) pharyngeal arches; (sb) swim bladder; (t) tectum; (y) yolk.

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 co-immunoprecipitate Xenopus and human ELYS proteins15. Further studies demonstrated that the association of ELYS with chromatin acts as a seeding point for Nups, one of the earliest events in nuclear pore assembly12, 13, 15–18. In ELYS-depleted HeLa cells, NPCs fail to reassemble at the end of mitosis, causing cell cycle arrest and death15, 17. This, together with the fact that Elys-deficient mouse embryos die around the time of implantation28, 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 wildtype elys mRNA molecules, which persist in flo embryos up to 24hpf (Figure 4B). Maternally encoded wildtype 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 wildtype embryos at 72hpf, including the digestive organs. Interestingly, BrdU incorporation analysis demonstrates that at 72hpf many of the cells in these same organs are in S-phase of the cell cycle (Figure 7). Thus the tissue-specific 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 Elys15, are able to assemble nuclear pores but lack the capacity to differentiate along a neural pathway29. Similarly, targeted disruption of Nup50 causes complex neural tube and CNS abnormalities and growth retardation30, while Nup96+/− mice have impaired innate and adaptive immunity due to a compromised ability to properly express interferon–regulated proteins31. While 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 wildtype siblings may be a fruitful avenue to address this concept.

Supplementary Material

01
02
03
04
05
06

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, Alt Zantema (technical expertise), Janna Taylor and Pierre Smith (graphics) and Matthias Ernst and Tony Burgess (comments on the manuscript).

Grant Support: This work was supported by NHMRC Australia project grants 280916 and 433614 (JKH), NHMRC Dora Lush (Biomedical) Scholarship (TADJ-C), NHMRC Howard Florey Centenary Fellowship (HV), Australian Research Council grant DP0346823 (GJL) and NIH grants DK 067285 (ITS), K058181 and DK060322 (DYRS).

Abbreviations

BrdU

bromodeoxyuridine

ENU

ethylnitrosourea

Elys

Embryonic large molecule derived from yolk sac

dpf

days post-fertilization

hpf

hours post-fertilization

TEM

transmission electron microscope

WISH

whole mount in situ hybridization

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

The authors declare no conflicts of interest related to this work.

References

  • 1.Ng AN, de Jong-Curtain TA, Mawdsley DJ, White SJ, Shin J, Appel B, Dong PD, Stainier DY, Heath JK. Formation of the digestive system in zebrafish: III. Intestinal epithelium morphogenesis. Dev Biol. 2005;286:114–135. doi: 10.1016/j.ydbio.2005.07.013. [DOI] [PubMed] [Google Scholar]
  • 2.Pack M, Solnica-Krezel L, Malicki J, Neuhauss SC, Schier AF, Stemple DL, Driever W, Fishman MC. Mutations affecting development of zebrafish digestive organs. Development. 1996;123:321–328. doi: 10.1242/dev.123.1.321. [DOI] [PubMed] [Google Scholar]
  • 3.Wallace KN, Akhter S, Smith EM, Lorent K, Pack M. Intestinal growth and differentiation in zebrafish. Mech Dev. 2005;122:157–173. doi: 10.1016/j.mod.2004.10.009. [DOI] [PubMed] [Google Scholar]
  • 4.Crosnier C, Vargesson N, Gschmeissner S, Ariza-McNaughton L, Morrison A, Lewis J. Delta-Notch signalling controls commitment to a secretory fate in the zebrafish intestine. Development. 2005;132:1093–1104. doi: 10.1242/dev.01644. [DOI] [PubMed] [Google Scholar]
  • 5.Chen JN, Haffter P, Odenthal J, Vogelsang E, Brand M, van Eeden FJ, Furutani-Seiki M, Granato M, Hammerschmidt M, Heisenberg CP. Mutations affecting the cardiovascular system and other internal organs in zebrafish. Development. 1996;123:293–302. doi: 10.1242/dev.123.1.293. [DOI] [PubMed] [Google Scholar]
  • 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. doi: 10.1242/dev.00600. [DOI] [PubMed] [Google Scholar]
  • 7.Chen J, Ruan H, Ng SM, Gao C, Soo HM, Wu W, Zhang Z, Wen Z, Lane DP, Peng J. Loss of function of def selectively up-regulates Delta113p53 expression to arrest expansion growth of digestive organs in zebrafish. Genes Dev. 2005;19:2900–2911. doi: 10.1101/gad.1366405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wallace KN, Dolan AC, Seiler C, Smith EM, Yusuff S, Chaille-Arnold L, Judson B, Sierk R, Yengo C, Sweeney HL, Pack M. Mutation of smooth muscle myosin causes epithelial invasion and cystic expansion of the zebrafish intestine. Dev Cell. 2005;8:717–726. doi: 10.1016/j.devcel.2005.02.015. [DOI] [PubMed] [Google Scholar]
  • 9.Yee NS, Gong W, Huang Y, Lorent K, Dolan AC, Maraia RJ, Pack M. Mutation of RNA Pol III subunit rpc2/polr3b Leads to Deficiency of Subunit Rpc11 and disrupts zebrafish digestive development. PLoS Biol. 2007;5:e312. doi: 10.1371/journal.pbio.0050312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Ober EA, Verkade H, Field HA, Stainier DY. Mesodermal Wnt2b signalling positively regulates liver specification. Nature. 2006;442:688–691. doi: 10.1038/nature04888. [DOI] [PubMed] [Google Scholar]
  • 11.Field HA, Ober EA, Roeser T, Stainier DY. Formation of the digestive system in zebrafish. I. Liver morphogenesis. Dev Biol. 2003;253:279–290. doi: 10.1016/s0012-1606(02)00017-9. [DOI] [PubMed] [Google Scholar]
  • 12.Galy V, Askjaer P, Franz C, Lopez-Iglesias C, Mattaj IW. MEL-28, a novel nuclear-envelope and kinetochore protein essential for zygotic nuclear-envelope assembly in C. elegans. Curr Biol. 2006;16:1748–1756. doi: 10.1016/j.cub.2006.06.067. [DOI] [PubMed] [Google Scholar]
  • 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. doi: 10.1016/j.cub.2006.07.071. [DOI] [PubMed] [Google Scholar]
  • 14.Kimura N, Takizawa M, Okita K, Natori O, Igarashi K, Ueno M, Nakashima K, Nobuhisa I, Taga T. Identification of a novel transcription factor, ELYS, expressed predominantly in mouse foetal haematopoietic tissues. Genes Cells. 2002;7:435–446. doi: 10.1046/j.1365-2443.2002.00529.x. [DOI] [PubMed] [Google Scholar]
  • 15.Rasala BA, Orjalo AV, Shen Z, Briggs S, Forbes DJ. 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. doi: 10.1073/pnas.0608484103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Franz C, Walczak R, Yavuz S, Santarella R, Gentzel M, Askjaer P, Galy V, Hetzer M, Mattaj IW, Antonin W. MEL-28/ELYS is required for the recruitment of nucleoporins to chromatin and postmitotic nuclear pore complex assembly. EMBO Rep. 2007;8:165–172. doi: 10.1038/sj.embor.7400889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Gillespie PJ, Khoudoli GA, Stewart G, Swedlow JR, Blow JJ. ELYS/MEL-28 chromatin association coordinates nuclear pore complex assembly and replication licensing. Curr Biol. 2007;17:1657–1662. doi: 10.1016/j.cub.2007.08.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Rasala BA, Ramos C, Harel A, Forbes DJ. Capture of AT-rich chromatin by ELYS recruits POM121 and NDC1 to initiate nuclear pore assembly. Mol Biol Cell. 2008;19:3982–2996. doi: 10.1091/mbc.E08-01-0012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 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. doi: 10.1146/annurev.cellbio.21.090704.151152. [DOI] [PubMed] [Google Scholar]
  • 20.Anderson DJ, Hetzer MW. Nuclear envelope formation by chromatin-mediated reorganization of the endoplasmic reticulum. Nat Cell Biol. 2007;9:1160–1166. doi: 10.1038/ncb1636. [DOI] [PubMed] [Google Scholar]
  • 21.Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF. Stages of embryonic development of the zebrafish. Dev Dyn. 1995;203:253–310. doi: 10.1002/aja.1002030302. [DOI] [PubMed] [Google Scholar]
  • 22.Reichenbach B, Delalande JM, Kolmogorova E, Prier A, Nguyen T, Smith CM, Holzschuh J, Shepherd IT. Endoderm-derived Sonic hedgehog and mesoderm Hand2 expression are required for enteric nervous system development in zebrafish. Dev Biol. 2008;318:52–64. doi: 10.1016/j.ydbio.2008.02.061. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Thisse C, Thisse B, Schilling TF, Postlethwait JH. Structure of the zebrafish snail1 gene and its expression in wild-type, spadetail and no tail mutant embryos. Development. 1993;119:1203–1215. doi: 10.1242/dev.119.4.1203. [DOI] [PubMed] [Google Scholar]
  • 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 [Google Scholar]
  • 25.Tabone T, Hayden MJ. Method of amplifying nucleic acid. 60/973,928. U.S. Pat. 2007
  • 26.Nakai K, Kanehisa K. A knowledge base for predicting protein localization sites in eukaryotic cells. Genomics. 1992;14:897–911. doi: 10.1016/S0888-7543(05)80111-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Aravind L, Landsman D. AT-hook motifs identified in a wide variety of DNA-binding proteins. Nucleic Acids Res. 1998;26:4413–4421. doi: 10.1093/nar/26.19.4413. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Okita K, Kiyonari H, Nobuhisa I, Kimura N, Aizawa S, Taga T. Targeted disruption of the mouse ELYS gene results in embryonic death at peri-implantation development. Genes Cells. 2004;9:1083–1091. doi: 10.1111/j.1365-2443.2004.00791.x. [DOI] [PubMed] [Google Scholar]
  • 29.Lupu F, Alves A, Anderson K, Doye V, Lacy E. Nuclear pore composition regulates neural stem/progenitor cell differentiation in the mouse embryo. Dev Cell. 2008;14:831–842. doi: 10.1016/j.devcel.2008.03.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Smitherman M, Lee K, Swanger J, Kapur R, Clurman BE. Characterization and targeted disruption of murine Nup50, a p27(Kip1)-interacting component of the nuclear pore complex. Mol Cell Biol. 2000;20:5631–5642. doi: 10.1128/mcb.20.15.5631-5642.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Faria AM, Levay A, Wang Y, Kamphorst AO, Rosa ML, Nussenzveig DR, Balkan W, Chook YM, Levy DE, Fontoura BM. The nucleoporin Nup96 is required for proper expression of interferon-regulated proteins and functions. Immunity. 2006;24:295–304. doi: 10.1016/j.immuni.2006.01.014. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS506140-supplement-01.doc (53KB, doc)
02
NIHMS506140-supplement-02.tif (16.9MB, tif)
03
NIHMS506140-supplement-03.tif (37MB, tif)
04
NIHMS506140-supplement-04.tif (42.6MB, tif)
05
NIHMS506140-supplement-05.tif (5.4MB, tif)
06
NIHMS506140-supplement-06.tif (3.4MB, tif)

RESOURCES