SEARCH

SEARCH BY CITATION

Keywords:

  • Shroom;
  • morphogenesis;
  • epithelial sheets

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Thickened epithelial sheets are found in a wide variety of organ systems and the mechanisms governing their morphogenesis remain poorly defined. We show here, through expression patterns and functional studies, that Shroom family proteins are broadly involved in generating thickened epithelial sheets. Through in situ hybridization, we report the temporal and spatial expression patterns of the four Shroom family members during early Xenopus development, from oocytes to tadpole stage embryos. Further, we show that Shroom1 and 2 mRNAs are maternally expressed, while Shroom3 and Shroom4 are zygotic transcripts. In addition, maternal Shroom1 and 2 mRNAs localize in the animal hemisphere of the Xenopus egg and early blastula. During later development, all four Shroom family proteins are broadly expressed in developing epithelial organs, and the epithelial cells that express Shrooms are elongated. Moreover, we show that ectopic expression of Shroom2, like Shroom3, is able to increase cell height and that loss of Shroom2 function results in a failure of cell elongation in the neural epithelium. Together, these data suggest that Shroom family proteins play an important role in the morphogenesis of several different epithelial tissues during development. Developmental Dynamics 238:1480–1491, 2009. © 2009 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Cell shape change is a critical event for proper morphogenesis during development. This process occurs by the regulation and reorganization of the actin and microtubule cytoskeletons. Neurulation, the morphogenetic process that gives rise to the hollow central nervous systems, is one developmental process that relies on such cytoskeletal regulation. During this process, the neuroepithelial cells, which are columnar at the beginning of neurulation, undergo apical constriction and apicobasal elongation, processes governed by actin filaments and microtubules, respectively (Burnside,1973). Despite many studies, it remains largely unknown which molecule or signal governs these cell-shape changes. However, recent studies have shown that Shroom3 drives both apical constriction and apicobasal cell elongation via organization of an actomyosin network at the apical surface and microtubule assembly along the apicobasal axis in neuroepithelial cells during neural tube closure (Hildebrand and Soriano,1999; Haigo et al.,2003; Hildebrand,2005; Lee et al.,2007).

Shroom3, an actin-binding protein, was first identified as a critical molecule for neural tube closure nearly a decade ago. Shroom3 mutation in mice caused severe neural tube closure defects, resulting in a phenotype resembling the shape of a mushroom, hence the gene was originally called Shroom and was later renamed Shroom3 (Hildebrand and Soriano,1999; Hagens et al.,2006a). It has also been shown that Shroom3 is required for proper neural tube closure in Xenopus (Haigo et al.,2003). Protein sequence analysis reveals that Shroom3 shares two highly conserved domains with two other proteins, previously known as apical protein Xenopus (APX) (Staub et al.,1992) and apical protein Xenopus-Like (APXL) (Schiaffino et al.,1995). These domains have been termed ASDs (Apx/Shroom domains). The centrally located domain is ASD1 and the C-terminal domain is ASD2 (Hildebrand and Soriano,1999). Recently, all related proteins that have an ASD domain have been grouped into one single family and renamed as the Shroom family proteins (Hagens et al.,2006a). There are four proteins in the Shroom family: APX, APXL, Shroom, and Kiaa1202. APX is now renamed as Shroom1, APXL is now Shroom2, Shroom is now Shroom3, and Kiaa1202 is now Shroom4 (Hagens et al.,2006a). These proteins will be referred to here by their Shroom family names.

Shroom1 was initially identified as a protein participating in amiloride-sensitive sodium channel activity in the apical surface of Xenopus epithelial cells (Staub et al.,1992). Additionally, in Xenopus renal epithelial cells, Shroom1 also has been shown to be associated with epithelial sodium channel (EnaC) and α-spectrin, a cytoskeletal molecule known to shape the plasma membranes of cells (Zuckerman et al.,1999). Shroom2 was first reported as a human homologue of APX, but has now been characterized as a distinct protein (Schiaffino et al.,1995; Hagens et al.,2006a). Shroom2 binds to the actin cytoskeleton and controls its organization (Dietz et al.,2006). In addition, it has been shown that Shroom2 controls the localization of melanosomes in the retinal pigment epithelium (RPE) and the distribution of γ-tubulin to the apical surface of naive epithelial cells in Xenopus (Fairbank et al.,2006). As the ectopic expression of Shroom1 or 3 also alters the distribution of γ-tubulin in epithelial cells, these Shroom family proteins are thought to govern the assembly of microtubules (Lee et al.,2007). Shroom4 is associated with human X-linked mental retardation (Hagens et al.,2006b). Unlike the other three Shroom proteins, which have both ASD1 and ASD2 domains, Shroom4 only contains an ASD2 domain. Shroom 4, like Shroom2 and 3, has been shown to interact directly with actin filaments (Yoder and Hildebrand,2007). Together these reports suggest that all of the Shroom family proteins are involved in the control of cellular architecture by utilizing the cytoskeletal elements, such as actin filaments, microtubules, and α-spectrin during developmental morphogenesis. However, the overall functions of these proteins and the specific functions of the ASD1 and ASD2 domains are still largely unclear.

To better understand the functions of Shroom family proteins during development, we observed the expression patterns of Shroom family mRNAs via in situ hybridization in Xenopus. Shroom1 and 2 are maternally expressed and their mRNAs are localized in the animal hemisphere. During early Xenopus development, all Shroom family genes, except Shroom4, are mostly expressed in epithelial tissues, including neural tissue and several placodes undergoing cell thickening. Although Shroom4 is expressed in mesodermal tissues, the cells expressing it are also elongated. Moreover, similar to Shroom3 (Lee et al.,2007), ectopic expression of Shroom2 causes cell heightening and this protein is required for cell elongation in the Xenopus neural plate. Together these data suggest that the Shroom family proteins play an important role in the morphogenesis of several different epithelial tissues during development.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Shroom1 and 2 mRNAs Are Maternally Expressed and Enriched in the Animal Hemisphere of the Oocyte and Early Blastula

To investigate when Shroom genes are expressed during embryonic development, we analyzed their mRNA levels from egg until mid-gastrula by reverse transcriptase PCR (RT-PCR). Both Shroom1 and Shroom2 mRNAs are detected in unfertilized eggs and their levels are stable during early cleavage stages. On the other hand, Shroom3 and 4 mRNAs are not detected during early development. Instead, Shroom3 mRNAs are first detected during the mid-blastula transition (MBT, around stage 9) and Shroom4 mRNAs are detected post MBT (Fig. 1).

thumbnail image

Figure 1. RT-PCR for Shroom family genes during early Xenopus development. Reverse transcriptase PCR (RT-PCR) was performed with specific primers of Shroom family genes in unfertilized egg, mid-blastrula (st. 8 and 9), early gastrula (st.10), and mid-gastrula (st. 11.5). ODC (ornithine decarboxylase) is used as a control. The weak bands for Shroom3 and 4 in egg and stage 8 are the result of genomic DNA contamination (data not shown).

Download figure to PowerPoint

Based on these expression patterns of Shroom1 and 2, we examined the spatial distribution of Shroom1 and 2 mRNAs in unfertilized eggs and early embryos by in situ hybridization. Surprisingly, these mRNAs appear to be localized asymmetrically along the animal-vegetal axis. mRNAs for both Shroom1 and 2 accumulated primarily in the animal hemisphere of both unfertilized eggs and cleavage stage embryos (Fig. 2A). The localization of Shroom1 and 2 mRNAs at the animal pole persisted into later cleavage stages, but the mRNA then appeared to be localized in or near the nuclei of the cells. At stage 6.5, Shroom1 and 2 mRNAs are also detected in the nucleus of some vegetal cells, although these levels are very low compared to levels seen in the animal hemisphere (Fig. 2A).

thumbnail image

Figure 2. mRNA localization of Shroom1 and 2 during oogenesis and early development. A: In situ hybridization against Shroom1 and 2 in unfertilized egg, 1-cell, 2-cell stage, and stage 6.5. B: In situ hybridization against Shroom1 and 2 during oogenesis. C: In situ hybridization against AN2, Vg1, and Histone H3. Animal to the top in all panels.

Download figure to PowerPoint

Several mRNAs such as AN1, AN2, AN3, and PABP have been previously reported to localize to the animal hemisphere in Xenopus (Rebagliati et al.,1985; Weeks and Melton,1987; Schroeder and Yost,1996). We, therefore, observed the distribution of Shroom1 and 2 mRNAs during oogenesis in comparison to these known, asymmetrically-localized mRNAs. Shroom1 and Shroom2 are expressed around the germinal vesicle or oocyte nucleus from the start of oogenesis, stage I (data not shown) (Dumont,1972). These mRNAs remain near the germinal vesicle during oogenesis even after it has moved towards the animal pole (Fig. 2B). The localization of AN2 mRNA, a transcript localized in the animal pole (Rebagliati et al.,1985; Weeks and Melton,1987), Vg1 mRNA, a transcript localized in the vegetal pole (Rebagliati et al.,1985; Melton,1987), and histone H3 and α-tubulin, transcripts distributed uniformly in both animal and vegetal hemispheres (Perry and Capco,1988), were examined. As expected, AN2 mRNA distribution is very similar to those of Shroom1 and 2 (Fig. 2C, top row). Vg1 mRNA localizes to the vegetal pole, as already reported (Rebagliati et al.,1985; Melton,1987) (Fig. 2C, middle row). However, though α-tubulin and histone H3 mRNA have been reported to localize in both animal and vegetal hemisphere by RNase protection assays, we unexpectedly found their localizations to be similar to that of Shroom1, Shroom2, and AN2 in our in situ hybridization data (Fig. 2C, bottom row and data not shown). Thus, while our in situ hybridizations suggest that Shroom1 and Shroom2 mRNA may be enriched animally, there remains some uncertainty. However, given that over-expression of Shroom2 or Shroom3 at blastula stages results in dramatic accumulation of pigment (Haigo et al.,2003; Fairbank et al.,2006; Lee et al.,2007), it is tempting to speculate that there is a link between animal enrichment of the Shroom family mRNAs and animal pigmentation in the Xenopus embryo.

Expression Pattern of Shroom Family Genes During Development

The temporal and spatial expressions of Shroom family genes during Xenopus development were analyzed by whole-mount in situ hybridization with several different stages of embryos from mid-gastrula to late tail bud stages.

Shroom1.

At stage 10.5, Shroom1 transcripts are detected throughout the animal 2/3 of the embryo, which excludes the entire blastopore and a part of the involuting marginal zone. This region corresponds to the ectodermal and mesodermal layers of the embryo but does not include the endoderm (Fig. 3A). During neurulation, the transcript is restricted to the neural plate region (Fig. 3B). The neural plate of Xenopus laevis consists of both a superficial layer and a deep layer (Fig. 4A) (Schroeder,1970). Shroom1 mRNA is mostly expressed in the deep layer of neuroepithelial cells (Figs. 3D, 4B). In addition, it is detected in the notochord and paraxial mesoderm of the somites from transverse sections of the neurula (Fig. 3D). After neural tube closure, Shroom1 is strongly expressed in the cement gland, and it is more strongly expressed in the dorsal portion than the ventral portion (Fig. 3E,G,H). Shroom1 transcripts are detected in the otic placode from early tadpole stage (Fig. 3F, I–K, yellow arrowhead) and in the pronephric tubes and duct in tadpole stage (Fig. 3I–K, white arrow).

thumbnail image

Figure 3. Expression pattern of Shroom1 in Xenopus. A: Lateral view of stage 10.5, animal to the top. B–D: Stage 17. B: Dorsal view, anterior to the left. C: Lateral view, anterior to the left, dorsal to the top. D: Transverse cross-section view. So, somite; Noto, notochord. Dorsal to the top. The sectioning level is indicated by yellow dotted line in C. E–G: Stage 22. E: Lateral view, anterior to the left, dorsal to the top. CG, cement gland. F: Dorsal view of anterior region, anterior to the left. OV, otic vesicle. G: Cement gland, dorsal to the top. H,I: Stage 25. H: Cement gland, dorsal to the top. I: Lateral view, anterior to the left, dorsal to the top. PN, prenephric tubes and duct. J,K: Stage 30. J: Lateral view, anterior to the left, dorsal to the top. K: High-magnification view of anterior region. Yellow arrowheads indicate the otic vesicle (OV) and white arrows indicate the prenephric tubes and duct (PN).

Download figure to PowerPoint

thumbnail image

Figure 4. Shroom expression in neural plate. A: Schematic depicting cross-section view of Xenopus neurula. B: Shroom1 expresses in the deep layer of neuroepithelial cells. C: Shroom2 expresses in the deep layer. D: Shroom3 expresses in the superficial layer. White dashed lines indicate the border between superficial layer and deep layer. White solid lines indicate the border between neural plate and mesodermal tissue.

Download figure to PowerPoint

Shroom2.

Similar to Shroom1, Shroom2 is expressed mostly in ectoderm and mesoderm but not in the endoderm during gastrulation (Fig. 5A). During neurulation, expression is restricted to the neural plate, especially the anterior neural plate, brain, and presumptive eye regions (Fig. 5B–D). A transverse section of neurula shows that Shroom2 is mostly expressed in the deep layer of neuroepithelial cells (Figs. 5E, 4C). As development proceeds, the expression increases in the forebrain, midbrain, and eyecup (Fig. 5F–I), whereas expression in the hindbrain and spinal cord region decreases (Fig. 5F,G). The transcript is first detected in the otic placode (Fig. 5G, yellow arrowhead) and kidney at the early tadpole stage, stage 21 (Fig. 5G, white arrow). Unexpectedly, it is also expressed in the notochord at stage 30 (Fig. 5M, red arrowhead). Later, it is expressed in the retinal pigment epithelium (RPE) (Fairbank et al.,2006).

thumbnail image

Figure 5. Expression pattern of Shroom2 in Xenopus. A: Stage 10.5, lateral view, animal to the top. B–E: Stage 17. B: Dorsal view, anterior to the left. C: Lateral view, anterior to the left, dorsal to the top. D: Anterior view, dorsal to the top. E: Transverse cross-section view, dorsal to the top. The sectioning level is indicated by yellow dotted line in C. F–I: Stage 21. F: Lateral view, anterior to the left, dorsal to the top. G: Dorsal view, anterior to the left. OV, otic vesicle; PN, prenephric tubes and duct. H: Ventral view, anterior to the left. I: Anterior view, dorsal to the top. J–L: Stage 25. J: Dorsal view, anterior to the left. K: Lateral view of anterior region, dorsal to the top. L: Anterior view, dorsal to the top. M: Stage 30, lateral view, anterior to the left, dorsal to the top. Noto, notochord. Yellow arrowheads indicate the otic vesicle (OV), white arrows indicate the prenephric tubes and duct (PN) and red arrowhead indicates the notochord (Noto).

Download figure to PowerPoint

Shroom3.

Shroom3 transcripts are detected broadly in mesoderm and ectoderm during gastrulation, but the expression is much stronger in the dorsal area than on the ventral (Fig. 6A,B). During neurulation, mRNA is expressed in the neural plate (Fig. 6C,D), but only in the superficial layer (Figs. 4D, 6E) where it controls apical constriction (Lee et al.,2007). This restriction to the superficial layer contrasts the patterns of Shroom1 and 2. In addition, Shroom3 is expressed in the anterior region of the neural plate, more specifically the placodal primordium (Fig. 6C,D) and neural crest of the trunk region (Fig. 6E). Immediately after neural tube closure, Shroom3 is expressed in several different placodes. It is initially expressed in the otic placode (Fig. 6F,H,K), and is later expressed in the lateral line placodes (Fig. 6F,H). Then finally expression is seen in the olfactory placodes (Fig. 6F,G,L). As previously reported, Shroom3 is strongly expressed in the cement gland in a pattern similar to Shroom1 (Fig. 6F,G,L) (Lee et al.,2007). Interestingly, Shroom3 transcripts are very strongly detected in the proctodeum at tadpole stages (Fig. 6F,G,J). Similar to Shroom2, Shroom3 is strongly expressed in the forebrain and midbrain (Fig. 6M). However, it is also expressed in the posterior neural tube until late tail bud stage (Fig. 6K). In addition, Shroom3 is expressed in the heart (Fig. 6K,L,N) and kidney (Fig. 6K) at late tail bud stages.

thumbnail image

Figure 6. Expression pattern of Shroom3 in Xenopus. A,B: Stage 10.5. A: Lateral view, animal to the top. B: Sagittal-section view, animal to the top. C–E: Stage 17. C: Dorsal view, anterior to the left. D: Anterior view, dorsal to the top. E: Transverse cross-section. NP, neural plate; NC, neural crest. Dorsal to the top. The sectioning level is indicated by yellow dotted line in C. F,G: Stage 23. F: Lateral view, dorsal to the top. OV, otic vesicle; LP, lateral line placode; Ol, olfactory placode; CG, cement gland; Pr, Proctodeum. G: Ventral view, anterior to the left. H–J: Stage 25. H: Dorsal view, anterior to the left. I: Anterior view, dorsal to the top. J: Posterior view, dorsal to the top. K,L Stage 30. K: Lateral view, anterior to the left, dorsal to the top. L–N: Head views of K. L: Lateral view, anterior to the left, dorsal to the top. M: Dorsal view, anterior to the top. PN, prenephric tubes and duct. Yellow star indicates the brain. N: Ventral view, anterior to the top. He, heart.

Download figure to PowerPoint

Shroom4.

At mid-gastrula, Shroom4 transcripts are detected in both the ectoderm and mesoderm (Fig. 7A,B). During neurulation, it is expressed in the cells of the presumptive somites (Fig. 7C,E,F), ventral mesodermal tissue (Fig. 7G) of the trunk region (Hausen and Riebessell,1991) and also in the future brain (Fig. 7D). At the early tadpole stage, transcripts of Shroom4 are detected in the brain regions including the eyecup, the boundary cells of the cement gland (Van Evercooren and Picard,1978), and the somites (Fig. 7H–K). At stage 30, Shroom4 is strongly expressed in the pronephric duct and tubules, otic vesicle, somite, and the region between the olfactory placode and cement gland, which will become the mouth (Fig. 7L). An interesting thing to note is that intense Shroom4 expression is seen in the most posterior of the somitic region, where the most recently formed somites reside (Fig. 7H,J,M, red arrowheads).

thumbnail image

Figure 7. Expression pattern of Shroom4 in Xenopus. A,B: Stage 10.5. A: Dorsal view, animal to the top. B: Sagittal-section view, animal to the top. C–G: Stage 17. C: Dorsal view, anterior to left. D: Anterior view, dorsal to top. E: Transverse cross-section view, dorsal to the top. The sectioning level is indicated by yellow dotted line in C. F,G: Dorsal part.(F) and ventral part (G) of E, dorsal to the top. H–K: Stage 21. H: Lateral view, anterior to the left, dorsal to the top. Som, somite. I: Anterior view, dorsal to the top. J: Dorsal view, anterior to the left. K: Lateral view of most anterior region, dorsal to the top. Black arrowhead indicates the boundary cells of the cement gland. L,M: Stage 30. L: Lateral view, anterior to the left, dorsal to the top. OV, otic vesicle; PN, prenephric tubes and duct. M: Most posterior region of L. Red arrowhead indicates the most posterior of somatic region.

Download figure to PowerPoint

Cell Shape in Shroom Protein-Expressing Regions

It has been reported that all Shroom family proteins interact with actin and that actin polymerization is required for their function (Prat et al.,1996; Hildebrand and Soriano,1999; Dietz et al.,2006; Yoder and Hildebrand,2007). Moreover, we have shown that ectopic expression of Shroom1, 2, or 3 causes accumulation of γ-tubulin, a microtubule nucleating protein, at the apical surface of epithelial cells (Fairbank et al.,2006; Lee et al.,2007). We have also demonstrated that Shroom3 induces apicobasal cell elongation in neuroepithelial cells and epithelial cells ectopically expressing it via microtubule assembly (Lee et al.,2007). Based on these data, it has been suggested that Shroom proteins function in the regulated control of cytoskeletal molecules and thereby aid in cell morphogenesis.

Furthermore, as shown in Figures 3–7 and Table1, Shroom family genes are expressed in many thickened epithelial sheets. To test whether Shroom family proteins are associated with certain cell shapes, we observed the cell shapes in Shroom gene–expressing tissues at a later stage. From sectioned embryos (Fig. 8A), we observed the cell shapes in the lateral line placode (expressing Shroom3)(Fig. 8B), proctodeum (expressing Shroom3) (Fig. 8C), cement gland (expressing Shroom1 and 3) (Fig. 8D), and otic placode (expressing Shroom 1–4) (Fig. 8E). In the otic placode and lateral line placode, the cells undergo apical constriction and apico-basal elongation as compared to neighboring cells. These observations, in concert with our previous findings in the neuroepithelial cells (Lee et al.,2007), suggest that Shroom3 is required for apical constriction and apico-basal elongation not only in neuroepithelial cells, but also in several other tissues during development. Moreover, just after the onset of Shroom1 and Shroom3 in the cement gland, these cells undergo significant elongation along their apico-basal axis (Fig. 8D). As in the cement gland, cells are very elongated in the proctodeum (Fig. 8C). Next, to test whether these tissues are exactly matched with Shroom-expressing cells, we applied fluorescence in situ hybridization against Shroom3 (Jekely and Arendt,2007; Trinh le et al.,2007). As shown in Figure 8F and G, only Shroom3-expressing cells are elongated and are sharply distinguishable from neighboring cells in morphology. These data support the idea that Shroom family proteins are generally involved in cell shape changes, including apico-basal elongation sometimes with apical constriction.

Table 1. Gene Expression Pattern of Shroom Family Members During Xenopus Developmenta
Stage tissueMaternal expressionGastrulaNeurulabTadpole stage
ProteinEctoMesoEndoCement glandOtic placodeKidneyEye-cupBrainUnique expression
  • a

    Ecto, ectodermal tissue; Meso, mesodermal tissue; Endo, endodermal tissue; SN, superficial neural plate; DN, deep neural plate; No, notodchord; So, somite, NC, neural crest; RPE, retinal pigment epithelium.

  • b

    Cross-section view of neurula.

  • c

    Forebrain and midbrain.

Shroom1+++equation image+++ 
Shroom2+++equation image++++cNotochord RPE
Shroom3++equation image++++cLateral line placodeOlfactory placodeProctodeumNeural tubeHeart
Shroom4++equation image++++Boundary cells of the cement glandPresumptive mouthSomite
thumbnail image

Figure 8. Cell shapes in Shroom-expressing tissues. A: Schematic depicting the sections of embryo imaged for B through G. B: Lateral line placode. C: Proctodeum. D: Cement gland (phalloidin). E: Otic vesicle. green is α-tubulin, red is propidium iodide in B, C, E. Green is phalloidin in D. F,G: Fluorescence images of in situ hybridization against Shroom3. DIC, Differential interference contrast image; α-tubulin, image stained with α-tubulin antibody; NBT/BCIP, NBT/BCIT precipitate following in situ hybridization by Shroom3 probe. It was detected by confocal microscopy. F: Otic vesicle. G: Lateral line placode. Scale bar = 50 μm.

Download figure to PowerPoint

Shroom Proteins Control Apicobasal Cell Elongation in Epithelia

Next, we tested whether other Shroom proteins could induce cell heightening, as Shroom3 does, by using gain and loss of function assays. For the gain of function assay, we ectopically expressed Shroom2 in epithelial cells of the early blastula to test if it was able to induce cell elongation. While the surface area of Shroom2-expressing cells is unchanged compared to that of non-Shroom2-expressing cells (Fairbank et al.,2006), Shroom2-expressing epithelial cells are apico-basally elongated (Fig. 9B, pink arrow), compared to non-Shroom2-expressing cells (Fig. 9B, yellow arrow). The height of Shroom2-overexpressing cells increases approximately 50% over that of neighboring cells (Fig. 9C). We also ectopically expressed Shroom2 in epidermal cells at the neurula stage. Although these epidermal cells are normally very flat and elongated along the lateral axis (Fig. 9D), Shroom2-expressing cells are apicobasally elongated (Fig. 9E, pink arrow). Another remarkable finding is that microtubule bundles are assembled parallel to the apicobasal axis in Shroom2-expressing cells (Fig. 9E). Shroom2-expressing cells are almost twice as tall as neighboring epidermal cells (Fig. 9G). Taken together with previous data that Shroom2, like Shroom3, controls γ-tubulin distribution, we suggest that Shroom2 regulates the alignment of microtubules to cause cell-shape changes and contributes to cell elongation during morphogenesis (Lee et al.,2007). To confirm that Shroom3 causes cell heightening, we measured the cell height in ectopically Shroom3-expressing cells of the neurula epidermis. Shroom3-expressing cells are two and a half times taller than normal neighboring epidermal cells (Fig. 9F,H). Because expression of Shroom proteins does not reorganize the cytoskeleton of mesenchamyal cells (Haigo et al.,2003), our data suggest that Shroom proteins broadly govern apicobasal cell elongation specifically in epithelia.

thumbnail image

Figure 9. Shroom proteins control cell heightening. A–C: Ectopic Shroom2 induces cell heightening in naive epithelial cells. A: Control. B: Shroom2-expressing cell. Green, γ-tubulin; red, myc-Shroom2). C: Graph of apicobasal cell height of control and Shroom2-expressing cells (mean ± s.e.m.; ctl, n = 16; Shroom2, n = 15). D–H: Ectopic Shroom2 expression induces cell heightening in epidermal cells. D: Control. E: Shroom2-expressing epidermis. F: Shroom3-expressing epidermis. Green, α-tubulin; red, myc-tagged Shroom2 or 3. G: Graph of apicobasal cell height in Shroom2-expressing epidermal cells (mean ± s.e.m.; ctl, n = 25;Shroom2, n = 25). H: Graph of apicobasal cell height in Shroom3-expressing epidermal cells (mean ± s.e.m : ctl, n = 39;Shroom3, n = 40). Scale bar = 20 μm. Yellow arrow bars are showing the way to measure the cell height for control cells (neighboring cells) and pink arrow bars are showing that for Shroom-expressing cells. I,L: Loss of function assay for Shroom2 in neural plate. Transverse cross-section view of neural plate. Green, α-tubulin; red, propidium iodide. I: One side-Shroom2 mismatched morpholino (MM)-injected embryo. J: One side-Shroom2 morpholino (MO)-injected embryo. The high-magnification views of control side cells (K) and Shroom2 MO-injected side cells (L). Scale bar = 50 μm.

Download figure to PowerPoint

As shown in Figure 5E, Shroom2 is expressed in the deep layer cells of the neural plate, which are elongated and in which microtubule bundles with apico-basal polarity are observed (Fig. 9I and K). To examine whether Shroom2 is required for cell elongation, we knocked down Shroom2 by expressing a splice-blocking morpholino (MO), an antisense oligonucleotide, in neuroepithelial cells (Fairbank et al.,2006). Similar to the Shroom3 MO, Shroom2 MO also causes severe neural tube defect of injected embryos (82.5%, n = 62, two experiments). However, while Shroom3 MO only disrupts the superficial layer and does not affect cell shape of the deep layer (Lee et al.,2007), knock-down of Shroom2 causes the disruption of cell shape in both the superficial layer and deep layer (Fig. 9J and L). These data suggest that Shroom2 is functionally involved in cell morphogenesis during development.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Here, we report the temporal and spatial expression patterns of Shroom family genes during Xenopus development. With the previous data that Shroom3 controls cell heightening in neural tissue, we hypothesized that other Shroom family proteins may also be involved in cell-shape changes. There are several lines of evidence that support this idea. First, all of the Shroom family proteins are able to directly bind actin filaments or be associated with actin filaments for their function (Prat et al.,1996; Hildebrand and Soriano,1999; Dietz et al.,2006; Yoder and Hildebrand,2007). Second, all of the Shroom family proteins are able to change the distribution of γ-tubulin, which is a key molecule in the regulation of microtubule assembly (Meads and Schroer,1995; Gunawardane et al.,2000). Similar to previously reported data that Shroom1, 2, and 3 can change the distribution of γ-tubulin in naive epithelial cells of Xenopus (Fairbank et al.,2006; Lee et al.,2007), we observed that Shroom4 can also change γ-tubulin distribution in those epithelial cells (data not shown). This observation suggests that Shroom family proteins may be broadly associated with γ-tubulin function. Finally, there are similar cell morphologies in the regions expressing these genes. For example, the neural plate, the cement gland, the proctodeum, the pronephric tubes and duct, and several placodes including otic, olfactory, lens, and lateral line placodes have a thickened cell shape. These findings imply that Shroom family proteins may play a role in cell elongation.

As shown previously with Shroom3, we show that Shroom2 is involved in cell heightening. During Xenopus neurulation, neuroepithelial cells in both the superficial and deep layers are elongated along their apico-basal axis (Schroeder,1970). Since it has been shown that Shroom3 controls cell elongation in only the superficial cell layer (Lee et al.,2007), it has remained a question as to which molecules control this process in the deep layer. By a loss-of-function assay, we show that Shroom2 is necessary for cell shape change in the deep layer of neuroepithelial cells (Fig. 9J, L). Furthermore, overexpression of Shroom2 induces cell heightening in two different epithelial cells (Fig. 9B, E). Together with data that Shroom1 and 2 are expressed in the deep layer of neuroepithelium, we suggest that they together may contribute to cell elongation in the deep layer.

Shroom4 is a distinct from other Shroom proteins, because it is mostly expressed in mesodermal tissue including the somites while the other Shroom family proteins are mostly expressed in epithelial cells. However, regions expressing Shroom4 have very similar cell shape to regions expressing the other Shroom proteins. Cells expressing this family of proteins tend to be very elongated and contain microtubule bundles arrayed along the apico-basal axis. Interestingly, Shroom4 is also expressed in the boundary of forming somites (Fig. 7H,M, red arrowheads). This region undergoes many events including dynamic cell-shape change during somitogenesis (Afonin et al.,2006). Based on these findings, we suggest that Shroom4 also may be involved in control of shape.

In this study, we report the expression patterns of Shroom family proteins during oogenesis and early development. Shroom1 and 2 mRNA are maternally expressed and they localize in the animal hemisphere in oocytes, unfertilized eggs, and early embryos. Moreover, cells in Shroom family protein–expressing regions in later embryos have a similar elongated cell shape. Added to previous data that Shroom3 controls cell shape, we show that Shroom2 is also involved in cell-shape changes by functional assays, both loss and gain of function assays. The molecular mechanisms underlying the elongation process are not known yet. However, we suggest that all Shroom proteins are required for cell-shape change in several different tissues during development.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Reverse Transcriptase PCR (RT-PCR)

RT-PCR was performed with the following primers and amplifying conditions: for Shroom1, 5′-TCTGGAGAAAGTGGTGAGCCTG-3′(forward) 5′-TCATTTGTAGCGGGTGGACG-3′(reverse), annealing temp.: 58°C and 28 cycles; for Shroom2, 5′-TCCTACTCCCGATTTTGTGC-3′(forward), 5′-CTGCTCCTGCATGTCTTTCA-3′ (reverse), annealing temp.: 58°C and 28 cycles; for Shroom3, 5′-TTATTGATTGAGCAACGGGAGC- 3′(forward), 5′-TGGAGGGGCATTGACACATTC-3′(reverse), annealing temp.: 58°C and 30 cycles; for Shroom4, 5′- CTCCTGCCCTGCTATGATGT-3′(for- ward), 5′-GCCTTTGAACCACCAACTTC-3′(reverse), annealing temp.: 58°C and 30 cycles; for ODC, 5′-GGCAAGGAATCACCCGAATG-3′(forward), 5′- GGCAACATAGTATCTCCCAGGCTC- 3′(reverse) annealing temp.: 58°C and 27 cycles.

Oocyte and Embryo Preparation

To obtain oocytes, ovarian tissue was isolated from a Xenopus laevis female and washed with calcium-free OR2 buffer (82.5 mM NaCl, 2.5 mM Kcl, 1 mM MgCl2, 1 mM NaH2PO4, 5 mM Hepes, 3.8 mM NaOH, pH 7.8). Oocytes were isolated by treatment of the ovarian tissue with type IA collagenase (1.5 mg/ml, Sigma) and were grouped based on features developed by Dumont (1972). Each stage of oocytes were fixed with 3.7 % formaldehyde in OR2 buffer.

Mouse Shroom2 mRNA or Xenopus Shroom3 mRNA was synthesized with mMESSAGE mMACHINE kit (Ambion). The capped mRNAs were injected into one dorsal cell of 4-cell stage embryo. For mosaic expression of Shroom, plasmid DNA containing Shroom2 (CS2-mAPXL) or Shroom3 (CS107-xShrm3) was injected into one ventral cell of 4-cell stage embryo. Antisense morpholino oligonucleotide or mismatched morpholino for Shroom2 (Fairbank et al.,2006) was injected one dorsal cell at 4-cell stage. Embryos were fixed in MEMFA (0.1 M MOPS, 2 mM EGTA, 1 mM MgSO4, 3.7% formaldehyde, pH 7.4) at proper developmental stages and were sectioned by using a Vibratome 1000 system (Davidson and Wallingford,2005).

In Situ Hybridization and Immunohistochemistry

In situ hybridization was performed as described previously (Sive et al.,2000). For probes of Shroom1, 2 and 4, NIBB xenopus cDNA clones were used as templates (Shroom1 :XL095f19, Shroom2 :XL031d13, Shroom4 :XL198e22), while a full-length xShroom3 construct cloned previously (Lee et al.,2007) was used for Shroom3 template.

Fixed embryos were immunostained as described before (Lee et al.,2007). Monoclonal anti-α-tubulin antibody (1:300 dilution, DM1A, Sigma) and rabbit polyclonal anti-γ-tubulin antibodies (1:300 dilution, Abcam) were used for primary antibodies and Alex Fluor-488 goat anti-mouse or anti-rabbit IgG was used for secondary antibody (Invitrogen, 1:300 dilution). Propidium iodide was used to 20 μg/ml (Sigma) to stain the nuclei. The images were obtained by using Zeiss LSM5 Pascal confocal microscope. Cell heights were measured by using LSM5 pascal software.

Fluorescence in situ hybridization were performed as described in recent reports (Jekely and Arendt,2007; Trinh le et al.,2007). First of all, in situ hybridization was performed as mentioned above. After the chromogenic reaction with BM purple (Roche), embryos were fixed in MEMFA and bleached in 1% H2O2, 5% formamide and 0.5 × SSC. The embryos were immunostained with α-tubulin as mentioned above. NBT/BCIT precipitates were detected by using a 633-nm laser and 650-nm long-pass filter of LSM5 laser confocal microscope.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Esther Kieserman for a critical reading of this manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES