- Email: [email protected]
progenitor cells of the intestinal epithelium
Gene Expression Patterns 6 (2005) 45–56 www.elsevier.com/locate/modgep
Rbm19 is a nucleolar protein expressed in crypt/progenitor cells of the intestinal epithelium James A. Lorenzena,d, Benedetta B. Bonaccia,d, Rachel E. Palmerb, Clive Wellsc, Jian Zhanga,d, Daniel A. Haberb, Allan M. Goldsteine, Alan N. Mayera,c,d,* a
Gastroenterology Section, Department of Pediatrics, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA b Cancer Center, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA 02129, USA c Department of Cell Biology, Neurobiology and Anatomy, Medical College of Wisconsin, Milwaukee, WI 53226, USA d Children’s Research Institute, Milwaukee, WI 53226, USA e Department of Pediatric Surgery, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA 02129, USA Received 19 April 2005; received in revised form 10 May 2005; accepted 11 May 2005 Available online 18 July 2005
Abstract Intestinal development and homeostasis rely on the coordination of proliferation and differentiation of the epithelium. To better understand this process, we are studying Rbm19, a gene expressed in the gut epithelium that is essential for intestinal morphogenesis and differentiation in the zebrafish (Development 130, 3917). Here we analyzed the expression of Rbm19 in several biological contexts that feature proliferation/differentiation cell fate decisions. In the undifferentiated embryonic gut tube, Rbm19 is expressed throughout the epithelium, but then becomes localized to the crypts of Lieberku¨hn of the adult intestine. Consistent with its expression in adult crypt/progenitor cells, expression is widespread in human colorectal carcinomas and dividing Caco-2 cells. Its expression in Caco-2 cells recapitulates the in vivo pattern, declining when the cells undergo confluence-induced arrest and differentiation. Rbm19 protein localizes to the nucleolus during interphase and to the perichromosomal sheath during mitosis, in accordance with the pattern described for other nucleolar proteins implicated in ribosome biogenesis. Interestingly, the loss of nucleolar rbm19, nucleolin/C23, and nucleophosmin/B23 in confluent Caco-2 cells did not signify loss of nucleoli as detected by electron microscopy. Taken together, these data point to the nucleolus as a possible locus for regulating the proliferation/differentiation cell fate decision in the intestinal epithelium. q 2005 Elsevier B.V. All rights reserved. Keywords: Intestine; Rbm19; Development; Nucleolus; Progenitor; Stem cell; Differentiation; Morphogenesis; Organogenesis; Caco-2 cells; Intestinal crypt; RNA-binding proteins; Nucleophosmin; Nucleolin; C23; B23; Preribosomal RNA processing; Ribosome biogenesis; Colorectal adenocarcinoma; Colon cancer; Mitosis
The vertebrate intestine offers a venue to study a broad range of basic biological processes: organ morphogenesis, growth, differentiation, homeostasis, and neoplasia. During development of the mammalian intestine, the epithelium changes from multilayer to single layer and the cells begin to express intestine-specific genes. A subset of cells remain undifferentiated and segregate to the basal portion of newly formed villi to form a progenitor compartment. Thereafter, * Corresponding author. Address: Department of Pediatric, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA. Tel.: C1 414 456 5894; fax: C1 414 456 6632. E-mail address: [email protected] (A.N. Mayer).
1567-133X/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.modgep.2005.05.001
progenitor cells continue to divide and differentiate as they migrate up the villus to replenish the epithelium on average every 3–5 days (Montgomery et al., 1999). The molecular control of the cell fate decisions that collectively establish and maintain the architecture of the intestine are being revealed through the manipulation of known candidate pathways (Sancho et al., 2004) and forward genetic screens (Amsterdam et al., 2004; Farber et al., 2001; Mayer and Fishman, 2003; Pack et al., 1996). Previously, we undertook a forward genetic approach to studying intestinal development in zebrafish and isolated the nil per os (npo) gene (Mayer and Fishman, 2003). The phenotype in the npo mutant is arrest of intestinal development at the primitive gut tube stage, with failure of subsequent growth and differentiation. Gene expression peaks in the gut
46
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
immediately prior to villus morphogenesis, consistent with an essential function in this step. npo was found to encode a unique protein with six RNA recognition motif (RRM) domains (5 domains in fungi and plants) that is highly conserved throughout all eukaryotes (Bjork et al., 2002; Mayer and Fishman, 2003). Studies of the npo ortholog in yeast, in the dipteran C. tentans, and in C. elegans, all concluded that the gene product is essential for the production of the 18S ribosomal RNA during ribosome biogenesis (Bjork et al., 2002; Jin et al., 2002; Saijou et al., 2004). We will hereafter refer to this gene by its HUGOapproved name, Rbm19. In this work we sought to determine how Rbm19 might function in the intestine of higher vertebrates. As an initial step, we examined its expression during development, in adult intestine, in a colon carcinoma cell line and in neoplastic tissue. In each of these contexts, we found Rbm19 to be differentially expressed in epithelial cells that exhibit a ‘crypt-progenitor’ phenotype, suggesting that it may be part of a pathway regulating proliferation/differentiation cell fate decisions.
1. Results and discussion 1.1. Rbm19 expression in developing and adult intestinal epithelium In the zebrafish, rbm19 expression is highly dynamic (Mayer and Fishman, 2003). Prior to intestinal morphogenesis, the gene is expressed throughout the primitive gut tube. Then the gut undergoes an anterior–posterior wave of intestinal differentiation (Andre et al., 2000), and rbm19 expression declines with the appearance of differentiated intestinal cells. A rare subset of cells continues to express rbm19, suggesting the possibility that these cells represent a remnant of undifferentiated endoderm fated to become epithelial progenitor cells. In mammals, epithelial morphogenesis of the intestine is well described, and the appearance of villi corresponds to the formation of a stem cell compartment in the intervillus epithelium (Klein, 1989). To detect Rbm19 expression in developing mouse intestine, we performed in situ hybridization (ISH), focusing on the stages during which the intestinal epithelium matures from the endoderm (Fig. 1). In whole mount specimens, we found the staining to be diffuse at E13.5, but increasingly restricted at E14.5. Expression then declines precipitously by E15.5. Rbm19 expression in E14.5 localizes to islands with a stereotypical spatial pattern, such that sharp bands of higher expression form in the rostral part of the island, and a gradient tapers off caudally (Fig. 1, panel B). These are highlighted in magnified views of the pylorus-duodenum junction (Fig. 1, panel B1) and the duodenal-jejunal border (Fig. 1, panel B2). At E15.5, the only expression detectable on whole-mount is in a segment of duodenum between the pylorus and the ampulla of Vater (Fig. 1D). Since, the temporal pattern in the mouse differed from that seen in
Fig. 1. Whole mount in situ hybridization for Rbm19 expression. Mouse (A–C) and chick (D–E). In the mouse Rbm19 is expressed in a dynamic, heterogeneous expression pattern, with more diffuse expression at E13.5, and progressive restriction of expression by E14.5. Magnified views of areas with increased expression show a distinct banding pattern (B1, B2). By E15.5 expression is undetectable except in the mesenteric aspect of the duodenum between the pylorus and ampulla of Vater (C). An overall similar pattern is seen in the developing chick intestine, with diffuse expression on day 10, then localization to the duodenum by day 13 (D and E). Histological section of chick intestine on E13 shows staining of the apical layer of the epithelium, in a pattern similar to the mouse intestine on E13.5 (Fig E, inset). Abbreviations: L, lumen, e, epithelium, m, mesenchyme.
zebrafish, we also performed in situ hybridization on chick intestine during the period of villus morphogenesis (Coulombre and Coulombre, 1958). In the chick, we found a similar developmental variation of Rbm19 compared with the mouse. At E10, Rbm19 expression is widely distributed across the intestine (Fig. 1D), and then at E13, expression is restricted to the duodenum (Fig. 1E). This pattern resembles the mouse intestine in the progression of diffuse to focal gene expression. We concluded that, while there may be interspecies differences in the exact developmental sequence, there is a general conservation of diffuse expression in the undifferentiated endoderm with progressive restriction of expression as the endoderm differentiates. Microscopic examination of embryonic gut stained for Rbm19 mRNA demonstrates that expression is mostly in the endoderm (Fig. 2). In accordance with the pattern seen on whole mount, we found a dramatic reduction in the extent of Rbm19 expression with the onset of villus morphogenesis. At E12.5 expression is distributed in multiple foci throughout the endoderm, then at E13.5 the signal localizes
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
47
Fig. 2. Histological analysis of Rbm19 expression in developing mouse intestine. In E12.5 embryos Rbm19 expression in the endoderm is mostly diffuse, but begins to show some spatial restriction by E13.5 (A and B, arrowheads). As villus morphogenesis becomes evident on E14.5, expression becomes increasingly restricted to the basal portions of the epithelium (C). A magnified area from an adjacent section to that shown in panel C (boxed) using a fluorescent substrate shows Rbm19 localized to discrete cytoplasmic foci in a subset of cells in the basal portion of the epithelium (arrowhead, panel D). After villus morphogenesis, Rbm19 is detected only in the intervillus region (arrowheads, panel E). Abbreviations: e, epithelium; m, mesenchyme; L, lumen Scale bars: A–C and E–F, 15 mm; D, 5 mm.
mostly to a cell layer just basal to the lumen. At E14.5, the epithelium is in the midpoint of a rearrangement from multito single-layered, and this is accompanied by a restriction of Rbm19 expression to the prospective villus base. As nascent villi form and the epithelium adopts a simple columnar morphology, expression becomes restricted to the epithelial cells in the intervillous regions (Fig. 2E). We noted that Rbm19-specific labeling appeared more punctate than expected for a diffuse cytoplasmic signal, so we performed the detection reaction with a fluorescent alkaline phosphate
substrate (see ‘Methods’). This revealed subcellular localization of the Rbm19 mRNA within apical cytoplasmic granules (Fig. 2D). The pattern is seen at all stages in which Rbm19 is detected, and is similar to that seen in zebrafish (Mayer and Fishman, 2003), suggesting conserved posttranscriptional control. The significance of this particular pattern of localization may relate to the subsequent localization of RBM19 protein to the nucleus. Other nuclear proteins such as c-myc are encoded by messages that are localized to the perinuclear cytoplasm where, after
48
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Fig. 3. Rbm19 expression in adult mouse small intestine. Whole mount in situ hybridization shows labeling along the base of the mucosa (A). Histological sections reveal Rbm19 mRNA in the crypts of Lieberku¨hn (B), with strict localization of expression to a subset of cells at the bases of the crypts (B, inset). In situ hybridization to Tcf-4 on an adjacent section shows a similar expression pattern (C). Rbm19 sense-strand control probe (D). Polyclonal antibodies made to recombinant human rbm19 was used for western blotting of a whole-cell Caco-2 extract fractionated by SDS-PAGE. We noted a band of about 100 kD, the size predicted for rbm19 (E). This antibody was used along with the nucleolar marker anti-B23 to label adult mouse duodenum. rbm19 are expressed in a virtually identical pattern selectively in the crypts of Lieberku¨hn (F–I). Scale bars: B–D and F–I, 20 mm; B and D insets, 10 mm.
translation, they are imported into the nucleus (Hesketh et al., 1994). Message localization may thus facilitate subsequent protein localization, though it is not always required for the latter (Jansen, 2001). Based on the progressive restriction of Rbm19 expression in the embryo, we predicted that expression in the adult intestine would be either absent or limited to a small subset of cells. Whole mount in situ hybridization of adult duodenum shows Rbm19 expression localized to
the bases of the villi (Fig. 3A) and sections of stained intestine show expression restricted to the crypts of Lieberku¨hn (Fig. 3B). On adjacent sections we compared Rbm19 expression to that of Tcf-4 (Fig. 3C), which is required for establishment of the stem cell compartment (Korinek et al., 1998). The expression patterns are very similar, consistent with their crypt-specific localization. As seen in the embryo, Rbm19 message localizes to discrete foci in the apical cytoplasm of a small subset of crypt cells
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
49
Fig. 4. Rbm19 expression in normal and neoplastic human colon. Specimens were from normal colon (A–C), adenomatous polyp (D–F) and colorectal adenocarcinoma (G–J). Sections stained with H&E (A, D and G) provide a view of the tissue architecture. In the normal colon, rbm19 is found mostly at the bases of the crypts, with occasionally increased signal at the lumenal surface. Of note are rare cells with particularly high levels of rbm19 indicated by arrowheads (B). In the benign polyp, epithelial hyperplasia is evident, but most of the cells do not express rbm19 at the higher level. The proportion of such cells is two–three-fold higher (E). In the carcinoma, high-expressing cells abound. Two different parts of the tumor are shown in panels H and I. In panel H, crypt-like structures embedded in a mesenchymal stroma predominate, and a majority of the epithelial cells lining these aberrant crypts express rbm19 at a level comparable to the rare ‘high-expressing’ cells seen in normal colon and benign polyps, indicated by the arrowheads. In Panel I, the tissue is less organized, suggesting a more aggressive phenotype correlates with rbm19 expression. Scale bars: A–H, J, 40 mm; I, 20 mm.
(Fig. 3B, inset). This suggests that Rbm19 message may be sequestered into a ribonucleoprotein complex and thus might be subject to translational control. Another possibility is that Rbm19 message may be segregated asymmetrically in analogy to Musashi-1, another RNA-binding protein expressed in the crypt (Potten et al., 2003) and known to specify a stem cell phenotype in other contexts (Okano et al., 2002). To monitor expression of rbm19 protein, we generated affinity-purified rabbit polyclonal antibodies to a fragment of the human protein. The antibody reacted with a single band on western blot of Caco-2 cell extracts (Fig. 3E). Immunofluorescence staining of adult mouse duodenum reveals selective labeling of the crypts, consistent with the in situ results (Fig. 3F–I). The protein is localized to the nucleus in discrete foci, suggestive of nucleoli. Indeed, double immunostaining with the nucleolar marker B23
reveals virtually complete co-localization at the sub-cellular level and also along the crypt-villus axis. Furthermore, this dynamic pattern is recapitulated in the Caco-2 cell culture model (see below). In summary, we find Rbm19 expression to be selective for undifferentiated crypt/progenitor cells in the intestinal epithelium. Its developmental expression may offer a means to follow the progressive restriction of cells to a crypt/progenitor cell fate during intestinal development. In adult intestine, Rbm19 joins the small but growing number of genes selectively expressed in the crypt (Sancho et al., 2004). 1.2. Rbm19 expression in human intestinal neoplasia Having demonstrated that RBM19 expression is spatially restricted to the crypts in the adult mouse intestine, we hypothesized that in human neoplastic tissue rbm19
50
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Fig. 5. Subcellular localization of rbm19. In subconfluent Caco-2 cells rbm19 is localized to discrete foci in the nucleus and, to a lesser extent, to smaller foci in the perinuclear cytoplasm (A–D). Panel A (inset) shows a 3-D surface-rendered view of rbm19 in structures that form rings and multiple lobes. Double immunofluorescence staining for rbm19 and B23 shows complete colocalization in subconfluent Caco-2 cells in the nucleus (E–G). Scale bars: 5 mm. Panel A inset: 6.5 mm across.
expression would be more widespread, since colonic neoplasia takes on many of the phenotypic characteristics of normal crypt cells (Sancho et al., 2004). Thus, we performed anti-rbm19 immunostaining of three separate specimens each of normal human colon, adenomatous polyp and colorectal carcinoma (Fig. 4). In the normal colonic mucosa, there are rare cells that express Rbm19 at high levels, and these are located in and around the crypt (Fig. 4B). In adenomatous polyps, the epithelium is hyperplastic but the cells remain polarized and organize along the basement membrane. In examining 100 high-power fields for each tissue type (about 35 fields per specimen), we detected highexpressing cells at a three-fold higher frequency compared with normal colon (4 vs. 1.3 cells/hpf) (Fig. 4E). In the carcinomas, the architecture of the tissue reflects unregulated crypt proliferation with no clear crypt-villus axis. Almost all of the epithelial cells lining these aberrant crypts express rbm19 at a level comparable to the rare highexpressing cells seen in normal colon (Fig. 4H). In less
organized areas of tumor, rbm19-expressing cells are quite abundant, and many of these cells have acquired a spindle-shaped morphology, indicating loss of polarity and increased cell motility (Fig. 4I). These data illustrate a positive correlation between degree of malignancy and rbm19 expression. 1.3. Rbm19 expression in Caco-2 cells Caco-2 cells are derived from a colorectal carcinoma, and thus would be expected to express rbm19. We tested this by immunofluorescence in actively dividing (subconfluent) cells, and found robust expression localizing to nuclear foci (Fig. 5). We noted that the signals form distinct ring structures, which would be consistent with localization in the granular compartment and dense fibrillar component of nucleoli (Dimario, 2004). Optical sectioning, deconvolution and three-dimensional rendering of the image (Fig. 5A, inset) revealed multi-lobed and hollow structures consistent
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
51
Fig. 6. Localization of rbm19 in Caco-2 cells during mitosis. At prophase, rbm19 staining becomes diffuse. At prometaphase, perichromosomal localization is first detected, and cytoplasmic signal persists. At metaphase, perichromosomal localization is clearly evident, with relatively less cytoplasmic staining. At anaphase, rbm19 expression colocalizes with the segregating chromosomes. At telophase, rbm19 continues to colocalize with the chromosomes, but the cytoplasmic foci are more prominent and localized to opposite sides of the spindle poles in the daughter cells (arrowheads), likely representing prenucleolar bodies. As cytokinesis occurs, (accompanied by chromosome decondensation and reformation of the nuclear membrane) rbm19 is found in a well-delineated structure in the center of the nucleus. Within this structure, rbm19 concentrates around the periphery of round structures about 1 mm in diameter, likely representing dense fibrillar component surrounding the fibrillar centers. Scale bars, 5 mm.
with this localization. Furthermore, rbm19 colocalizes with nucleophosmin/B23, the latter a well-established marker for these nucleolar compartments (Fig. 5E–G). Thus, the data are consistent with rbm19 participating in processing of preribosomal RNA (Fromont-Racine et al., 2003).
During mitosis, the nucleolus disassembles and the components relocate to different compartments of the cell. B23 and other non-ribosomal components of the nucleolus exhibit a distinctive spatio-temporal localization pattern with regard to the mitotic apparatus (Dundr et al., 1997).
52
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Fig. 7. Dynamic expression of rbm19 in newly confluent Caco-2 cells. Fully confluent Caco-2 cells are distinguished by enhanced localization of actin to the cell–cell interface, indicating adherens junction formation (A). Confluent cells exhibit abrupt loss of nuclear rbm19 and appearance of a diffuse signal (enclosed by dashed line) (B–D). Occasional confluent cells contained perinuclear rbm19 (D, inset). RBM19 expression as a function of Caco-2 cell differentiation declined both at the RNA and protein levels (E and F). Real-time RT-PCR was used to quantify RBM19 message, showing a nine-fold reduction of levels in concert with the induction of the differentiation marker sucrase–isomaltase; loading control (36b4) remains constant (E). Units on y-axis are for comparison of expression of a given gene at different time points, and cannot be used to compare absolute expression levels between genes. Western blot analysis of Caco-2 whole-cell extracts shows decline in rbm19 protein levels with the onset of confluence (F).
Given its co-localization with B23 during interphase, we were interested to know if rbm19 conformed to the pattern described for B23 during mitosis as well. Hence, we analyzed the localization of rbm19 in Caco-2 cells at the different stages of mitosis (Fig. 6). In prophase, rbm19 is no longer detected in nucleoli, consistent with nucleolar disassembly and the attendant dissociation of protein components. By late prophase, rbm19 can be seen associating with the chromosome periphery, where it remains throughout metaphase. During anaphase and telophase rbm19 continues to localize with the chromosomes, but in telophase rbm19 also localizes to prenucleolar
bodies (PNBs). As the chromosomes decondense during cytokinesis, rbm19 becomes restricted to incipient nucleoli, with increased localization in ring structures that correspond to the dense fibrillar component. In summary, this pattern conforms to the mitotic paradigm described for B23 and other nucleolar proteins involved in preribosomal RNA processing ((Dimario, 2004) and references therein). Caco-2 cells possess a unique property among colorectal carcinoma cell lines, in that upon reaching confluence, they undergo cell cycle arrest and differentiate into cells resembling small intestinal enterocytes (Rousset, 1986). Since, RBM19 expression is not expressed strongly in
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
53
Fig. 8. Concerted expression of rbm19 and nucleolar markers. Double labeling for rbm19 and either B23 (A–D) or C23 (E–H) in newly confluent Caco-2 cells. Absence of B23 or C23 in confluent cells from the nucleus precedes loss of rbm19, since we did not detect any B23-positive/rbm19-negative cells or C23positive/rbm19-negative cells. Electron micrograph of subconfluent and confluent (!14 days) Caco-2 cells demonstrates that morphologically normal nucleoli are still present in both. Morphological signs of differentiation are evident in the confluent cells with the detection of microvilli. Abbreviations: mv, microvilli; gc, granular compartment; dfc, dense fibrillar component; fc, fibrillar center. Scale bars: A–H, 30 mm. I and J, 3 mm; I (inset) 0.3 mm; J (inset) 1 mm.
differentiated cells of the mouse small intestinal villi nor in the glands of human colon, we hypothesized that expression would decline in Caco-2 cells upon confluence-induced differentiation. We found that shortly after the cells become confluent, rbm19 protein is no longer detected in the nucleus (Fig. 7). Rbm19 declines initially in the center of the colonies, followed by the colony periphery as cells from neighboring colonies encounter each other. As nuclear staining declines, a diffuse signal appears in its place, which is distinct from the granular appearance of the rbm19 signal in subconfluent cells entering mitosis (compare Fig. 7, panel B vs. Fig. 6, prophase). In a small number of the confluent cells, we noted perinuclear foci of rbm19, consistent with egress from the nucleus within a discrete particle (Fig. 7, panel D (arrowhead)). To quantify RBM19 expression as a function of cell phenotype we performed real-time PCR of Caco-2 RNA (Fig. 7, E and F). We noted a 10-fold decline in RBM19, while SUCRASE–ISOMALTASE expression increased (Fig. 7E). Rbm19 protein levels declined after the cells reached confluence, concordant with mRNA (Fig. 7F). Neither RBM19 RNA nor protein levels ever reach undetectable levels by these assays, presumably corresponding to a persistent diffuse signal detected by immunofluorescence. We considered that the decline of nuclear rbm19 could be a consequence of cell-cycle arrest,
and tested this possibility by imposing serum-starvation on preconfluent cells or treating them with the cytostatic drug rapamycin (10–20 nM). We found no differences in rbm19 immunofluorescence after 48 h of either treatment (not shown). To determine if the loss of nuclear rbm19 is unique to this protein, we performed double immunostaining for nucleophosmin and rbm19 in newly confluent Caco-2 cells. We found that both B23 and C23 proteins are also downregulated as the cells become confluent (Fig. 8, panels A– H)). Interestingly, we noted a significant number of confluent cells negative for B23 but still positive for rbm19, but never the reverse (Fig. 8 panel A vs. B). We noted a similar result for nucleolin and rbm19 (Fig. 8 panel E vs. F). Thus, expression of rbm19 may be regulated in concert with B23 and C23. These results raised the question of whether confluent Caco-2 cells still contain nucleoli. We tested this by performing electron microscopy on cells maintained in the confluent state for 14 days. In both subconfluent and confluent cells, we readily identified nucleoli (Fig. 8, panels I and J). In the confluent cells the nucleoli were appreciably smaller and less numerous, but they appeared morphologically normal, with discernible fibrillar center, dense fibrillar component, and granular compartment. Thus, the loss of B23, C23 and rbm19
54
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
immunostaining in confluent Caco-2 cells is not a result of the disassembly of the nucleolus, to the extent that we can discern by electron microscopy.
The probe for the chick homolog of Rbm19 was made from sequence in GenBank Accession no. AJ454823. 2.3. Antibodies
1.4. Conclusions Rbm19 represents one of over 30 genes identified from 2 independent genetic screens in zebrafish that link ribosome biogenesis to intestinal development (Amsterdam et al., 2004; Mayer and Fishman, 2003). Ribosome biogenesis has been known for decades to be essential for cell growth, but even twenty years ago very little was known regarding its regulation in eukaryotic cells (Nomura et al., 1984). Recently, the site of ribosome biogenesis, the nucleolus, has emerged as a center of control for the regulation of cell growth and proliferation (Sherr, 2004). In particular, preribosomal RNA processing has been identified as a target for key mediators of cell proliferation such as c-Myc (Schlosser et al., 2003) and the tumor suppressor p19Arf (Sugimoto et al., 2003). Organogenesis is therefore likely to be regulated, at least in part, through control of ribosome biogenesis. While this hypothesis has yet to be tested explicitly, the data presented here suggest Rbm19 to be a component of this regulatory mechanism.
2. Experimental procedures 2.1. Animals, cells and human tissue Mouse wild-type strains were either CD-1 from Charles River (Wilmington, MA) or C57BL/6 from Jackson (Bar Harbor, ME). Caco-2 cells were obtained from ATCC and maintained in MEM (Eagle’s) (Invitrogen) with 20% fetal calf serum. Human samples of colon and tumor tissue were anonymous specimens obtained from the Massachusetts General Hospital Tumor Bank. Chick embryos were obtained from fertilized white leghorn chicken eggs (Spafas, CT) incubated at 37 8C with 50% humidified air and staged by day of incubation (E). All animal procedures were performed in accordance with protocols approved by the MGH and MCW Animal Care and Use Committees. 2.2. In situ probes For detection of Rbm19 message in mouse tissues, we made digoxigenin-labeled probes (sense or antisense) using a full-length mouse Rbm19 cDNA clone as the template (IMAGE #3673396). The TCF-4 probe template was generated by RT-PCR from P19 mouse teratocarcinoma cell line RNA using the following primers: SP6 (sense)CACTAGATTTAGGTGACACTATAGAAGCTTCATTAAACAAGAGACC and T7 (antisense) CACTAGTAATACGACTCACTATAGGGCACTAGCTGACGTGAA TACC.
Rabbit polyclonal antibody was elicited to a His6-tagged fragment of human rbm19 (amino acids 50–511). Antiserum was affinity-purified using the antigen linked to sepharose-4B. Anti-B23 was from Zymed (now Invitrogen) (South San Francisco, CA), anti-C23 was from Santa Cruz Biotechnology, anti-a-tubulin was from Sigma (St Louis, MO). Secondary antibodies were from Jackson ImmunoResearch (West Grove, PA) or Molecular Probes (now Invitrogen). 2.4. In situ hybridization Embryonic gut was dissected from the embryo, fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) at 4 8C overnight and then stored in methanol at K20 8C. The procedure for in situ hybridization of whole gut was as described previously for whole embryos (Nagy et al., 2003). We found the proteinase K step to be particularly critical, with 20 mg/ml incubated at 25 8C for 30 min to give the best results. For section in situs, tissues were embedded in OCT and sectioned at 10 mm thickness using a Leica cryostat onto Superfrost plus gold slides (Fisher), dried at 50 8C for 20 min and then stored at K80 8C. Slides were washed in PBS 5 min !2, then treated with proteinase K 10 mg/ml for 20 min, washed in 2 mg/ml glycine for 5 min, PBS for 5 min !2, and post-fixed in 4% paraformaldehyde in PBS for 20 min. After washing in PBS for 5 min !2, samples were acetylated by treatment with 0.1 M triethanolamine hydrochloride (TEA) pH 8.0 for 2 min, then 0.1 M TEA with 0.3% acetic anhydride (added drop wise) for 10 min. The slides were washed in PBS 5 min, water 2 min, the subjected to prehybridization for 2 h at 70 8C in the following solution: 50% formamide, 10% dextran sulfate, 10 mM Tris HCl pH 7.6, 10 mM EDTA, 0.6 M NaCl, 1! Denhardt’s solution, 0.25% sodium dodecyl sulfate (SDS) and 0.2 mg/ml torula yeast RNA. For the hybridization reaction, this solution was replaced with digoxigeninlabeled probe at a concentration of 1 ng/ml and incubated overnight at 70 8C. The next day, the slides were washed in the following solutions at 60 8C since we found that higher temperatures led to loss of sample from the slides: 6X SSC/ 50% formamide for 10 min, 2X SSC for 10 min, then 0.2X SSC 30 min !2. Slides were then cooled to 37 8C and treated with 0.1 mg/ml RNase A for 30 min, then washed in 0.1X SSC for 30 min, then PBS with 0.1% Tween (PBT) for 5 min. Samples were blocked in 10% sheep serum/1% bovine serum albumin (BSA) in PBT at 4 8C for 30 min, then treated with sheep anti-digoxigenin F(ab) fragmentlinked alkaline phosphatase (Roche) 1:1000 for 30 min. After several washes in PBT, samples were incubated in 0.1 M sodium chloride/0.1 M Tris HCl pH 9.5/50 mM
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
MgCl2/0.1% Tween 20 (NTMT) followed by BM purple (neat). Staining was allowed to proceed until the negative control (sense probe) began to develop signal. For the Rbm19 in situs, this required about 2–3 days of staining with daily replacement of BM purple solution. In selected cases the HNPP fluorescent detection kit (Roche) was used to visualize Rbm19 expression. 2.5. Immunohistochemistry Tissue was fixed in 4% PFA in PBS overnight at 4 8C, then embedded in OCT and sectioned at 10 mm thickness. Sections were fixed to the slides and then washed with PBT several times, permiablilized with 1% Triton X-100, then digested with 10 mg/ml proteinase K for 20 min for antigen retrieval. After washing with PBT for 5 min !2, tissue sections were blocked with 10% goat serum/1% BSA/PBT for 30 min, then treated with 1:100 antibody in blocking solution for 1 h. After several washes in PBT, the slides were treated with biotinylated goat anti-rabbit (1:2000) in block, washed in PBT, then developed using the Vectastain ABC elite kit (Vector Labs, Burlingame, CA) and diaminobenzidine (Sigma). The specimens were dehydrated in ethanol and citrisolve, mounted with Permount and photographed in both brightfield and Nomarski views, which were captured and merged digitally in the program Openlab. 2.6. Immunofluorescence Fresh tissue was embedded in OCT and sectioned at 10 mm thickness. Cells for immunofluorescence were plated on laminincoated wells (2 mg/cm2, from Sigma) in Tissue-Tek chamber slides. Cells or tissue were fixed by treatment with 3.2% paraformaldehyde for 10 min, 150 mM glycine for 5 min !2, 0.2% Triton X-100 for 2 min, PBS 5 min !3, blocking with 5% sheep or goat serum in PBS for 1 h at room temperature, then primary antibody (1:500) in 5% sheep serum for 1 h at room temperature. Secondary antibodies were diluted 1:500 in 5% sheep or goat serum, and cells were counterstained with DAPI. Rhodamine phalloidin was from Molecular Probes (now Invitrogen). Images were captured using a Zeiss Axioplan motorized microscope to generate images across the z-axis using Openlab, followed by deconvolution with the program Volocity (Improvision, Lexington, MA).
55
TCTGGATGCCAAAAGGA;36B4 forward GCGACCTGGAAGTCCAACTA, reverse GTGAGGTCCTCCTT GGTGAA. Products were quantified using SYBR green dye using an ABI PRISMw 7000 Sequence Detection System using the following cycle program: 94 8C !3 min, (94 8C !30 s, 58 8C !30 s, 72 8C !1 min) !40. Products were verified by post-amplification melting curve analysis and agarose gel electrophoresis. Expression levels are expressed in arbitrary units for comparison of trends. 2.8. Western blot analysis Caco-2 cells were grown in T-25 culture flasks, harvested by scraping into 1 ml ice cold PBS, pelleted, then lysed in 25 mM HEPES pH 7.4/150 mM NaCl/10% glycerol/0.5% TX-100/10 mM sodium fluoride/1 mM sodium vanadate/ protease inhibitor cocktail (PMSFCbenzamidineCaprotininCleupeptin). 25 mg of protein was loaded per lane. After electrophoresis and transfer to nitrocellulose, blots were probed with 1:2000 anti-rbm19 followed by 1:30,000 peroxidase conjugated goat anti-rabbit. 2.9. Electron microscopy Cells were washed twice with growth media without serum and then fixed in situ on ice with 2.5% glutaraldehyde in 0.1M cacodylate buffer pH 7.4 for 1 h. After fixation the cells were post-fixed with 1% osmium tetroxide for 1 h on ice, rinsed in distilled water and en-block stained with 1% aqueous uranyl acetate for 3 h on ice. Cells were then scraped, pelleted, dehydrated in graded methanol and embedded in Epon 812 epoxy resin. Ultrathin sections, 60 nm, were cut, counterstained with Reynolds lead citrate and examined by transmission electron microscopy (Hitachi H 600 TEM).
Acknowledgements We thank the MGH Cancer Center Pathology Core for technical assistance, the MGH Tumor Bank for samples of human intestinal tissue, and Drs Stephen Duncan, Colin Rudolph, and members of the Mayer lab for critical review of the manuscript. This work was supported by grants from the NIH (R03DK067176-01) and the Children’s Research Institute.
2.7. Real-time PCR RNA from Caco-2 cells was isolated with Trizol, random hexamer-primed reverse transcribed (Superscript II, Invitrogen) and then amplified with the following primers: human Rbm19 forward GGCAAGCCTCGAACCAAA, reverse AGCATCCTGCCCTGGAATA;sucrase isomaltase forward GCTGTGTATGGAGAACGGGT, reverse GAA
References Amsterdam, A., Nissen, R.M., Sun, Z., Swindell, E.C., Farrington, S., Hopkins, N., 2004. Identification of 315 genes essential for early zebrafish development. Proc. Natl Acad. Sci. USA 101, 12792–12797.
56
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Andre, M., Ando, S., Ballagny, C., Durliat, M., Poupard, G., Briancon, C., et al., 2000. Intestinal fatty acid binding protein gene expression reveals the cephalocaudal patterning during zebrafish gut morphogenesis. Int. J. Dev. Biol. 44, 249–252. Bjork, P., Bauren, G., Jin, S., Tong, Y.G., Burglin, T.R., Hellman, U., et al., 2002. A novel conserved RNA-binding domain protein, RBD-1, is essential for ribosome biogenesis. Mol. Biol. Cell. 13, 3683–3695. Coulombre, A.J., Coulombre, J.L., 1958. Intestinal development. I. Morphogenesis of the villi and musculature. J. Embryol. Exp. Morphol. 6, 403–411. Dimario, P.J., 2004. Cell and molecular biology of nucleolar assembly and disassembly. Int. Rev. Cytol. 239, 99–178. Dundr, M., Meier, U.T., Lewis, N., Rekosh, D., Hammarskjold, M.L., Olson, M.O., 1997. A class of nonribosomal nucleolar components is located in chromosome periphery and in nucleolus-derived foci during anaphase and telophase. Chromosoma 105, 407–417. Farber, S.A., Pack, M., Ho, S.Y., Johnson, I.D., Wagner, D.S., Dosch, R., et al., 2001. Genetic analysis of digestive physiology using fluorescent phospholipid reporters. Science 292, 1385–1388. Fromont-Racine, M., Senger, B., Saveanu, C., Fasiolo, F., 2003. Ribosome assembly in eukaryotes. Gene 313, 17–42. Hesketh, J., Campbell, G., Piechaczyk, M., Blanchard, J.M., 1994. Targeting of c-myc and beta-globin coding sequences to cytoskeletalbound polysomes by c-myc 3 0 untranslated region. Biochem. J. 298 (Pt 1), 143–148. Jansen, R.P., 2001. mRNA localization: message on the move. Nat. Rev. Mol. Cell Biol. 2, 247–256. Jin, S.B., Zhao, J., Bjork, P., Schmekel, K., Ljungdahl, P.O., Wieslander, L., 2002. Mrd1p is required for processing of pre-rRNA and for maintenance of steady-state levels of 40 S ribosomal subunits in yeast. J. Biol. Chem. 277, 18431–18439. Klein, R.M., 1989. Small intestinal cell proliferation during development. In: Lebenthal, E. (Ed.), Human Gastrointestinal Development. Raven Press, New York, NY, pp. 367–392. Korinek, V., Barker, N., Moerer, P., van Donselaar, E., Huls, G., Peters, P.J., et al., 1998. Depletion of epithelial stem-cell compartments in the small intestine of mice lacking Tcf-4. Nat. Genet. 19, 379–383.
Mayer, A.N., Fishman, M.C., 2003. Nil per os encodes a conserved RNA recognition motif protein required for morphogenesis and cytodifferentiation of digestive organs in zebrafish. Development 130, 3917–3928. Montgomery, R.K., Mulberg, A.E., Grand, R.J., 1999. Development of the human gastrointestinal tract: twenty years of progress. Gastroenterology 116, 702–731. Nagy, A., Gertsenstein, M., Vintersten, K., Behringer, R., 2003. Manipulating the Mouse Embryo. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Nomura, M., Gourse, R., Baughman, G., 1984. Regulation of the synthesis of ribosomes and ribosomal components. Annu. Rev. Biochem. 53, 75–117. Okano, H., Imai, T., Okabe, M., 2002. Musashi: a translational regulator of cell fate. J. Cell Sci. 115, 1355–1359. Pack, M., Solnica-Krezel, L., Malicki, J., Neuhauss, S.C.F., Schier, A.F., Stemple, D.L., et al., 1996. Mutations affecting development of zebrafish digestive organs. Development 123, 321–328. Potten, C.S., Booth, C., Tudor, G.L., Booth, D., Brady, G., Hurley, P., et al., 2003. Identification of a putative intestinal stem cell and early lineage marker; musashi-1. Differentiation 71, 28–41. Rousset, M., 1986. The human colon carcinoma cell lines HT-29 and Caco2: two in vitro models for the study of intestinal differentiation. Biochimie 68, 1035–1040. Saijou, E., Fujiwara, T., Suzaki, T., Inoue, K., Sakamoto, H., 2004. RBD-1, a nucleolar RNA-binding protein, is essential for Caenorhabditis elegans early development through 18S ribosomal RNA processing. Nucleic Acids Res 32, 1028–1036. Sancho, E., Batlle, E., Clevers, H., 2004. Signaling pathways in intestinal development and cancer. Annu. Rev. Cell Dev. Biol. 20, 695–723. Schlosser, I., Holzel, M., Murnseer, M., Burtscher, H., Weidle, U.H., Eick, D., 2003. A role for c-Myc in the regulation of ribosomal RNA processing. Nucleic Acids Res. 31, 6148–6156. Sherr, C.J., 2004. Principles of tumor suppression. Cell 116, 235–246. Sugimoto, M., Kuo, M.L., Roussel, M.F., Sherr, C.J., 2003. Nucleolar Arf tumor suppressor inhibits ribosomal RNA processing. Mol. Cell 11, 415–424.
Rbm19 is a nucleolar protein expressed in crypt/progenitor cells of the intestinal epithelium James A. Lorenzena,d, Benedetta B. Bonaccia,d, Rachel E. Palmerb, Clive Wellsc, Jian Zhanga,d, Daniel A. Haberb, Allan M. Goldsteine, Alan N. Mayera,c,d,* a
Gastroenterology Section, Department of Pediatrics, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA b Cancer Center, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA 02129, USA c Department of Cell Biology, Neurobiology and Anatomy, Medical College of Wisconsin, Milwaukee, WI 53226, USA d Children’s Research Institute, Milwaukee, WI 53226, USA e Department of Pediatric Surgery, Massachusetts General Hospital and Harvard Medical School, Charlestown, MA 02129, USA Received 19 April 2005; received in revised form 10 May 2005; accepted 11 May 2005 Available online 18 July 2005
Abstract Intestinal development and homeostasis rely on the coordination of proliferation and differentiation of the epithelium. To better understand this process, we are studying Rbm19, a gene expressed in the gut epithelium that is essential for intestinal morphogenesis and differentiation in the zebrafish (Development 130, 3917). Here we analyzed the expression of Rbm19 in several biological contexts that feature proliferation/differentiation cell fate decisions. In the undifferentiated embryonic gut tube, Rbm19 is expressed throughout the epithelium, but then becomes localized to the crypts of Lieberku¨hn of the adult intestine. Consistent with its expression in adult crypt/progenitor cells, expression is widespread in human colorectal carcinomas and dividing Caco-2 cells. Its expression in Caco-2 cells recapitulates the in vivo pattern, declining when the cells undergo confluence-induced arrest and differentiation. Rbm19 protein localizes to the nucleolus during interphase and to the perichromosomal sheath during mitosis, in accordance with the pattern described for other nucleolar proteins implicated in ribosome biogenesis. Interestingly, the loss of nucleolar rbm19, nucleolin/C23, and nucleophosmin/B23 in confluent Caco-2 cells did not signify loss of nucleoli as detected by electron microscopy. Taken together, these data point to the nucleolus as a possible locus for regulating the proliferation/differentiation cell fate decision in the intestinal epithelium. q 2005 Elsevier B.V. All rights reserved. Keywords: Intestine; Rbm19; Development; Nucleolus; Progenitor; Stem cell; Differentiation; Morphogenesis; Organogenesis; Caco-2 cells; Intestinal crypt; RNA-binding proteins; Nucleophosmin; Nucleolin; C23; B23; Preribosomal RNA processing; Ribosome biogenesis; Colorectal adenocarcinoma; Colon cancer; Mitosis
The vertebrate intestine offers a venue to study a broad range of basic biological processes: organ morphogenesis, growth, differentiation, homeostasis, and neoplasia. During development of the mammalian intestine, the epithelium changes from multilayer to single layer and the cells begin to express intestine-specific genes. A subset of cells remain undifferentiated and segregate to the basal portion of newly formed villi to form a progenitor compartment. Thereafter, * Corresponding author. Address: Department of Pediatric, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA. Tel.: C1 414 456 5894; fax: C1 414 456 6632. E-mail address: [email protected] (A.N. Mayer).
1567-133X/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.modgep.2005.05.001
progenitor cells continue to divide and differentiate as they migrate up the villus to replenish the epithelium on average every 3–5 days (Montgomery et al., 1999). The molecular control of the cell fate decisions that collectively establish and maintain the architecture of the intestine are being revealed through the manipulation of known candidate pathways (Sancho et al., 2004) and forward genetic screens (Amsterdam et al., 2004; Farber et al., 2001; Mayer and Fishman, 2003; Pack et al., 1996). Previously, we undertook a forward genetic approach to studying intestinal development in zebrafish and isolated the nil per os (npo) gene (Mayer and Fishman, 2003). The phenotype in the npo mutant is arrest of intestinal development at the primitive gut tube stage, with failure of subsequent growth and differentiation. Gene expression peaks in the gut
46
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
immediately prior to villus morphogenesis, consistent with an essential function in this step. npo was found to encode a unique protein with six RNA recognition motif (RRM) domains (5 domains in fungi and plants) that is highly conserved throughout all eukaryotes (Bjork et al., 2002; Mayer and Fishman, 2003). Studies of the npo ortholog in yeast, in the dipteran C. tentans, and in C. elegans, all concluded that the gene product is essential for the production of the 18S ribosomal RNA during ribosome biogenesis (Bjork et al., 2002; Jin et al., 2002; Saijou et al., 2004). We will hereafter refer to this gene by its HUGOapproved name, Rbm19. In this work we sought to determine how Rbm19 might function in the intestine of higher vertebrates. As an initial step, we examined its expression during development, in adult intestine, in a colon carcinoma cell line and in neoplastic tissue. In each of these contexts, we found Rbm19 to be differentially expressed in epithelial cells that exhibit a ‘crypt-progenitor’ phenotype, suggesting that it may be part of a pathway regulating proliferation/differentiation cell fate decisions.
1. Results and discussion 1.1. Rbm19 expression in developing and adult intestinal epithelium In the zebrafish, rbm19 expression is highly dynamic (Mayer and Fishman, 2003). Prior to intestinal morphogenesis, the gene is expressed throughout the primitive gut tube. Then the gut undergoes an anterior–posterior wave of intestinal differentiation (Andre et al., 2000), and rbm19 expression declines with the appearance of differentiated intestinal cells. A rare subset of cells continues to express rbm19, suggesting the possibility that these cells represent a remnant of undifferentiated endoderm fated to become epithelial progenitor cells. In mammals, epithelial morphogenesis of the intestine is well described, and the appearance of villi corresponds to the formation of a stem cell compartment in the intervillus epithelium (Klein, 1989). To detect Rbm19 expression in developing mouse intestine, we performed in situ hybridization (ISH), focusing on the stages during which the intestinal epithelium matures from the endoderm (Fig. 1). In whole mount specimens, we found the staining to be diffuse at E13.5, but increasingly restricted at E14.5. Expression then declines precipitously by E15.5. Rbm19 expression in E14.5 localizes to islands with a stereotypical spatial pattern, such that sharp bands of higher expression form in the rostral part of the island, and a gradient tapers off caudally (Fig. 1, panel B). These are highlighted in magnified views of the pylorus-duodenum junction (Fig. 1, panel B1) and the duodenal-jejunal border (Fig. 1, panel B2). At E15.5, the only expression detectable on whole-mount is in a segment of duodenum between the pylorus and the ampulla of Vater (Fig. 1D). Since, the temporal pattern in the mouse differed from that seen in
Fig. 1. Whole mount in situ hybridization for Rbm19 expression. Mouse (A–C) and chick (D–E). In the mouse Rbm19 is expressed in a dynamic, heterogeneous expression pattern, with more diffuse expression at E13.5, and progressive restriction of expression by E14.5. Magnified views of areas with increased expression show a distinct banding pattern (B1, B2). By E15.5 expression is undetectable except in the mesenteric aspect of the duodenum between the pylorus and ampulla of Vater (C). An overall similar pattern is seen in the developing chick intestine, with diffuse expression on day 10, then localization to the duodenum by day 13 (D and E). Histological section of chick intestine on E13 shows staining of the apical layer of the epithelium, in a pattern similar to the mouse intestine on E13.5 (Fig E, inset). Abbreviations: L, lumen, e, epithelium, m, mesenchyme.
zebrafish, we also performed in situ hybridization on chick intestine during the period of villus morphogenesis (Coulombre and Coulombre, 1958). In the chick, we found a similar developmental variation of Rbm19 compared with the mouse. At E10, Rbm19 expression is widely distributed across the intestine (Fig. 1D), and then at E13, expression is restricted to the duodenum (Fig. 1E). This pattern resembles the mouse intestine in the progression of diffuse to focal gene expression. We concluded that, while there may be interspecies differences in the exact developmental sequence, there is a general conservation of diffuse expression in the undifferentiated endoderm with progressive restriction of expression as the endoderm differentiates. Microscopic examination of embryonic gut stained for Rbm19 mRNA demonstrates that expression is mostly in the endoderm (Fig. 2). In accordance with the pattern seen on whole mount, we found a dramatic reduction in the extent of Rbm19 expression with the onset of villus morphogenesis. At E12.5 expression is distributed in multiple foci throughout the endoderm, then at E13.5 the signal localizes
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
47
Fig. 2. Histological analysis of Rbm19 expression in developing mouse intestine. In E12.5 embryos Rbm19 expression in the endoderm is mostly diffuse, but begins to show some spatial restriction by E13.5 (A and B, arrowheads). As villus morphogenesis becomes evident on E14.5, expression becomes increasingly restricted to the basal portions of the epithelium (C). A magnified area from an adjacent section to that shown in panel C (boxed) using a fluorescent substrate shows Rbm19 localized to discrete cytoplasmic foci in a subset of cells in the basal portion of the epithelium (arrowhead, panel D). After villus morphogenesis, Rbm19 is detected only in the intervillus region (arrowheads, panel E). Abbreviations: e, epithelium; m, mesenchyme; L, lumen Scale bars: A–C and E–F, 15 mm; D, 5 mm.
mostly to a cell layer just basal to the lumen. At E14.5, the epithelium is in the midpoint of a rearrangement from multito single-layered, and this is accompanied by a restriction of Rbm19 expression to the prospective villus base. As nascent villi form and the epithelium adopts a simple columnar morphology, expression becomes restricted to the epithelial cells in the intervillous regions (Fig. 2E). We noted that Rbm19-specific labeling appeared more punctate than expected for a diffuse cytoplasmic signal, so we performed the detection reaction with a fluorescent alkaline phosphate
substrate (see ‘Methods’). This revealed subcellular localization of the Rbm19 mRNA within apical cytoplasmic granules (Fig. 2D). The pattern is seen at all stages in which Rbm19 is detected, and is similar to that seen in zebrafish (Mayer and Fishman, 2003), suggesting conserved posttranscriptional control. The significance of this particular pattern of localization may relate to the subsequent localization of RBM19 protein to the nucleus. Other nuclear proteins such as c-myc are encoded by messages that are localized to the perinuclear cytoplasm where, after
48
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Fig. 3. Rbm19 expression in adult mouse small intestine. Whole mount in situ hybridization shows labeling along the base of the mucosa (A). Histological sections reveal Rbm19 mRNA in the crypts of Lieberku¨hn (B), with strict localization of expression to a subset of cells at the bases of the crypts (B, inset). In situ hybridization to Tcf-4 on an adjacent section shows a similar expression pattern (C). Rbm19 sense-strand control probe (D). Polyclonal antibodies made to recombinant human rbm19 was used for western blotting of a whole-cell Caco-2 extract fractionated by SDS-PAGE. We noted a band of about 100 kD, the size predicted for rbm19 (E). This antibody was used along with the nucleolar marker anti-B23 to label adult mouse duodenum. rbm19 are expressed in a virtually identical pattern selectively in the crypts of Lieberku¨hn (F–I). Scale bars: B–D and F–I, 20 mm; B and D insets, 10 mm.
translation, they are imported into the nucleus (Hesketh et al., 1994). Message localization may thus facilitate subsequent protein localization, though it is not always required for the latter (Jansen, 2001). Based on the progressive restriction of Rbm19 expression in the embryo, we predicted that expression in the adult intestine would be either absent or limited to a small subset of cells. Whole mount in situ hybridization of adult duodenum shows Rbm19 expression localized to
the bases of the villi (Fig. 3A) and sections of stained intestine show expression restricted to the crypts of Lieberku¨hn (Fig. 3B). On adjacent sections we compared Rbm19 expression to that of Tcf-4 (Fig. 3C), which is required for establishment of the stem cell compartment (Korinek et al., 1998). The expression patterns are very similar, consistent with their crypt-specific localization. As seen in the embryo, Rbm19 message localizes to discrete foci in the apical cytoplasm of a small subset of crypt cells
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
49
Fig. 4. Rbm19 expression in normal and neoplastic human colon. Specimens were from normal colon (A–C), adenomatous polyp (D–F) and colorectal adenocarcinoma (G–J). Sections stained with H&E (A, D and G) provide a view of the tissue architecture. In the normal colon, rbm19 is found mostly at the bases of the crypts, with occasionally increased signal at the lumenal surface. Of note are rare cells with particularly high levels of rbm19 indicated by arrowheads (B). In the benign polyp, epithelial hyperplasia is evident, but most of the cells do not express rbm19 at the higher level. The proportion of such cells is two–three-fold higher (E). In the carcinoma, high-expressing cells abound. Two different parts of the tumor are shown in panels H and I. In panel H, crypt-like structures embedded in a mesenchymal stroma predominate, and a majority of the epithelial cells lining these aberrant crypts express rbm19 at a level comparable to the rare ‘high-expressing’ cells seen in normal colon and benign polyps, indicated by the arrowheads. In Panel I, the tissue is less organized, suggesting a more aggressive phenotype correlates with rbm19 expression. Scale bars: A–H, J, 40 mm; I, 20 mm.
(Fig. 3B, inset). This suggests that Rbm19 message may be sequestered into a ribonucleoprotein complex and thus might be subject to translational control. Another possibility is that Rbm19 message may be segregated asymmetrically in analogy to Musashi-1, another RNA-binding protein expressed in the crypt (Potten et al., 2003) and known to specify a stem cell phenotype in other contexts (Okano et al., 2002). To monitor expression of rbm19 protein, we generated affinity-purified rabbit polyclonal antibodies to a fragment of the human protein. The antibody reacted with a single band on western blot of Caco-2 cell extracts (Fig. 3E). Immunofluorescence staining of adult mouse duodenum reveals selective labeling of the crypts, consistent with the in situ results (Fig. 3F–I). The protein is localized to the nucleus in discrete foci, suggestive of nucleoli. Indeed, double immunostaining with the nucleolar marker B23
reveals virtually complete co-localization at the sub-cellular level and also along the crypt-villus axis. Furthermore, this dynamic pattern is recapitulated in the Caco-2 cell culture model (see below). In summary, we find Rbm19 expression to be selective for undifferentiated crypt/progenitor cells in the intestinal epithelium. Its developmental expression may offer a means to follow the progressive restriction of cells to a crypt/progenitor cell fate during intestinal development. In adult intestine, Rbm19 joins the small but growing number of genes selectively expressed in the crypt (Sancho et al., 2004). 1.2. Rbm19 expression in human intestinal neoplasia Having demonstrated that RBM19 expression is spatially restricted to the crypts in the adult mouse intestine, we hypothesized that in human neoplastic tissue rbm19
50
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Fig. 5. Subcellular localization of rbm19. In subconfluent Caco-2 cells rbm19 is localized to discrete foci in the nucleus and, to a lesser extent, to smaller foci in the perinuclear cytoplasm (A–D). Panel A (inset) shows a 3-D surface-rendered view of rbm19 in structures that form rings and multiple lobes. Double immunofluorescence staining for rbm19 and B23 shows complete colocalization in subconfluent Caco-2 cells in the nucleus (E–G). Scale bars: 5 mm. Panel A inset: 6.5 mm across.
expression would be more widespread, since colonic neoplasia takes on many of the phenotypic characteristics of normal crypt cells (Sancho et al., 2004). Thus, we performed anti-rbm19 immunostaining of three separate specimens each of normal human colon, adenomatous polyp and colorectal carcinoma (Fig. 4). In the normal colonic mucosa, there are rare cells that express Rbm19 at high levels, and these are located in and around the crypt (Fig. 4B). In adenomatous polyps, the epithelium is hyperplastic but the cells remain polarized and organize along the basement membrane. In examining 100 high-power fields for each tissue type (about 35 fields per specimen), we detected highexpressing cells at a three-fold higher frequency compared with normal colon (4 vs. 1.3 cells/hpf) (Fig. 4E). In the carcinomas, the architecture of the tissue reflects unregulated crypt proliferation with no clear crypt-villus axis. Almost all of the epithelial cells lining these aberrant crypts express rbm19 at a level comparable to the rare highexpressing cells seen in normal colon (Fig. 4H). In less
organized areas of tumor, rbm19-expressing cells are quite abundant, and many of these cells have acquired a spindle-shaped morphology, indicating loss of polarity and increased cell motility (Fig. 4I). These data illustrate a positive correlation between degree of malignancy and rbm19 expression. 1.3. Rbm19 expression in Caco-2 cells Caco-2 cells are derived from a colorectal carcinoma, and thus would be expected to express rbm19. We tested this by immunofluorescence in actively dividing (subconfluent) cells, and found robust expression localizing to nuclear foci (Fig. 5). We noted that the signals form distinct ring structures, which would be consistent with localization in the granular compartment and dense fibrillar component of nucleoli (Dimario, 2004). Optical sectioning, deconvolution and three-dimensional rendering of the image (Fig. 5A, inset) revealed multi-lobed and hollow structures consistent
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
51
Fig. 6. Localization of rbm19 in Caco-2 cells during mitosis. At prophase, rbm19 staining becomes diffuse. At prometaphase, perichromosomal localization is first detected, and cytoplasmic signal persists. At metaphase, perichromosomal localization is clearly evident, with relatively less cytoplasmic staining. At anaphase, rbm19 expression colocalizes with the segregating chromosomes. At telophase, rbm19 continues to colocalize with the chromosomes, but the cytoplasmic foci are more prominent and localized to opposite sides of the spindle poles in the daughter cells (arrowheads), likely representing prenucleolar bodies. As cytokinesis occurs, (accompanied by chromosome decondensation and reformation of the nuclear membrane) rbm19 is found in a well-delineated structure in the center of the nucleus. Within this structure, rbm19 concentrates around the periphery of round structures about 1 mm in diameter, likely representing dense fibrillar component surrounding the fibrillar centers. Scale bars, 5 mm.
with this localization. Furthermore, rbm19 colocalizes with nucleophosmin/B23, the latter a well-established marker for these nucleolar compartments (Fig. 5E–G). Thus, the data are consistent with rbm19 participating in processing of preribosomal RNA (Fromont-Racine et al., 2003).
During mitosis, the nucleolus disassembles and the components relocate to different compartments of the cell. B23 and other non-ribosomal components of the nucleolus exhibit a distinctive spatio-temporal localization pattern with regard to the mitotic apparatus (Dundr et al., 1997).
52
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Fig. 7. Dynamic expression of rbm19 in newly confluent Caco-2 cells. Fully confluent Caco-2 cells are distinguished by enhanced localization of actin to the cell–cell interface, indicating adherens junction formation (A). Confluent cells exhibit abrupt loss of nuclear rbm19 and appearance of a diffuse signal (enclosed by dashed line) (B–D). Occasional confluent cells contained perinuclear rbm19 (D, inset). RBM19 expression as a function of Caco-2 cell differentiation declined both at the RNA and protein levels (E and F). Real-time RT-PCR was used to quantify RBM19 message, showing a nine-fold reduction of levels in concert with the induction of the differentiation marker sucrase–isomaltase; loading control (36b4) remains constant (E). Units on y-axis are for comparison of expression of a given gene at different time points, and cannot be used to compare absolute expression levels between genes. Western blot analysis of Caco-2 whole-cell extracts shows decline in rbm19 protein levels with the onset of confluence (F).
Given its co-localization with B23 during interphase, we were interested to know if rbm19 conformed to the pattern described for B23 during mitosis as well. Hence, we analyzed the localization of rbm19 in Caco-2 cells at the different stages of mitosis (Fig. 6). In prophase, rbm19 is no longer detected in nucleoli, consistent with nucleolar disassembly and the attendant dissociation of protein components. By late prophase, rbm19 can be seen associating with the chromosome periphery, where it remains throughout metaphase. During anaphase and telophase rbm19 continues to localize with the chromosomes, but in telophase rbm19 also localizes to prenucleolar
bodies (PNBs). As the chromosomes decondense during cytokinesis, rbm19 becomes restricted to incipient nucleoli, with increased localization in ring structures that correspond to the dense fibrillar component. In summary, this pattern conforms to the mitotic paradigm described for B23 and other nucleolar proteins involved in preribosomal RNA processing ((Dimario, 2004) and references therein). Caco-2 cells possess a unique property among colorectal carcinoma cell lines, in that upon reaching confluence, they undergo cell cycle arrest and differentiate into cells resembling small intestinal enterocytes (Rousset, 1986). Since, RBM19 expression is not expressed strongly in
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
53
Fig. 8. Concerted expression of rbm19 and nucleolar markers. Double labeling for rbm19 and either B23 (A–D) or C23 (E–H) in newly confluent Caco-2 cells. Absence of B23 or C23 in confluent cells from the nucleus precedes loss of rbm19, since we did not detect any B23-positive/rbm19-negative cells or C23positive/rbm19-negative cells. Electron micrograph of subconfluent and confluent (!14 days) Caco-2 cells demonstrates that morphologically normal nucleoli are still present in both. Morphological signs of differentiation are evident in the confluent cells with the detection of microvilli. Abbreviations: mv, microvilli; gc, granular compartment; dfc, dense fibrillar component; fc, fibrillar center. Scale bars: A–H, 30 mm. I and J, 3 mm; I (inset) 0.3 mm; J (inset) 1 mm.
differentiated cells of the mouse small intestinal villi nor in the glands of human colon, we hypothesized that expression would decline in Caco-2 cells upon confluence-induced differentiation. We found that shortly after the cells become confluent, rbm19 protein is no longer detected in the nucleus (Fig. 7). Rbm19 declines initially in the center of the colonies, followed by the colony periphery as cells from neighboring colonies encounter each other. As nuclear staining declines, a diffuse signal appears in its place, which is distinct from the granular appearance of the rbm19 signal in subconfluent cells entering mitosis (compare Fig. 7, panel B vs. Fig. 6, prophase). In a small number of the confluent cells, we noted perinuclear foci of rbm19, consistent with egress from the nucleus within a discrete particle (Fig. 7, panel D (arrowhead)). To quantify RBM19 expression as a function of cell phenotype we performed real-time PCR of Caco-2 RNA (Fig. 7, E and F). We noted a 10-fold decline in RBM19, while SUCRASE–ISOMALTASE expression increased (Fig. 7E). Rbm19 protein levels declined after the cells reached confluence, concordant with mRNA (Fig. 7F). Neither RBM19 RNA nor protein levels ever reach undetectable levels by these assays, presumably corresponding to a persistent diffuse signal detected by immunofluorescence. We considered that the decline of nuclear rbm19 could be a consequence of cell-cycle arrest,
and tested this possibility by imposing serum-starvation on preconfluent cells or treating them with the cytostatic drug rapamycin (10–20 nM). We found no differences in rbm19 immunofluorescence after 48 h of either treatment (not shown). To determine if the loss of nuclear rbm19 is unique to this protein, we performed double immunostaining for nucleophosmin and rbm19 in newly confluent Caco-2 cells. We found that both B23 and C23 proteins are also downregulated as the cells become confluent (Fig. 8, panels A– H)). Interestingly, we noted a significant number of confluent cells negative for B23 but still positive for rbm19, but never the reverse (Fig. 8 panel A vs. B). We noted a similar result for nucleolin and rbm19 (Fig. 8 panel E vs. F). Thus, expression of rbm19 may be regulated in concert with B23 and C23. These results raised the question of whether confluent Caco-2 cells still contain nucleoli. We tested this by performing electron microscopy on cells maintained in the confluent state for 14 days. In both subconfluent and confluent cells, we readily identified nucleoli (Fig. 8, panels I and J). In the confluent cells the nucleoli were appreciably smaller and less numerous, but they appeared morphologically normal, with discernible fibrillar center, dense fibrillar component, and granular compartment. Thus, the loss of B23, C23 and rbm19
54
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
immunostaining in confluent Caco-2 cells is not a result of the disassembly of the nucleolus, to the extent that we can discern by electron microscopy.
The probe for the chick homolog of Rbm19 was made from sequence in GenBank Accession no. AJ454823. 2.3. Antibodies
1.4. Conclusions Rbm19 represents one of over 30 genes identified from 2 independent genetic screens in zebrafish that link ribosome biogenesis to intestinal development (Amsterdam et al., 2004; Mayer and Fishman, 2003). Ribosome biogenesis has been known for decades to be essential for cell growth, but even twenty years ago very little was known regarding its regulation in eukaryotic cells (Nomura et al., 1984). Recently, the site of ribosome biogenesis, the nucleolus, has emerged as a center of control for the regulation of cell growth and proliferation (Sherr, 2004). In particular, preribosomal RNA processing has been identified as a target for key mediators of cell proliferation such as c-Myc (Schlosser et al., 2003) and the tumor suppressor p19Arf (Sugimoto et al., 2003). Organogenesis is therefore likely to be regulated, at least in part, through control of ribosome biogenesis. While this hypothesis has yet to be tested explicitly, the data presented here suggest Rbm19 to be a component of this regulatory mechanism.
2. Experimental procedures 2.1. Animals, cells and human tissue Mouse wild-type strains were either CD-1 from Charles River (Wilmington, MA) or C57BL/6 from Jackson (Bar Harbor, ME). Caco-2 cells were obtained from ATCC and maintained in MEM (Eagle’s) (Invitrogen) with 20% fetal calf serum. Human samples of colon and tumor tissue were anonymous specimens obtained from the Massachusetts General Hospital Tumor Bank. Chick embryos were obtained from fertilized white leghorn chicken eggs (Spafas, CT) incubated at 37 8C with 50% humidified air and staged by day of incubation (E). All animal procedures were performed in accordance with protocols approved by the MGH and MCW Animal Care and Use Committees. 2.2. In situ probes For detection of Rbm19 message in mouse tissues, we made digoxigenin-labeled probes (sense or antisense) using a full-length mouse Rbm19 cDNA clone as the template (IMAGE #3673396). The TCF-4 probe template was generated by RT-PCR from P19 mouse teratocarcinoma cell line RNA using the following primers: SP6 (sense)CACTAGATTTAGGTGACACTATAGAAGCTTCATTAAACAAGAGACC and T7 (antisense) CACTAGTAATACGACTCACTATAGGGCACTAGCTGACGTGAA TACC.
Rabbit polyclonal antibody was elicited to a His6-tagged fragment of human rbm19 (amino acids 50–511). Antiserum was affinity-purified using the antigen linked to sepharose-4B. Anti-B23 was from Zymed (now Invitrogen) (South San Francisco, CA), anti-C23 was from Santa Cruz Biotechnology, anti-a-tubulin was from Sigma (St Louis, MO). Secondary antibodies were from Jackson ImmunoResearch (West Grove, PA) or Molecular Probes (now Invitrogen). 2.4. In situ hybridization Embryonic gut was dissected from the embryo, fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) at 4 8C overnight and then stored in methanol at K20 8C. The procedure for in situ hybridization of whole gut was as described previously for whole embryos (Nagy et al., 2003). We found the proteinase K step to be particularly critical, with 20 mg/ml incubated at 25 8C for 30 min to give the best results. For section in situs, tissues were embedded in OCT and sectioned at 10 mm thickness using a Leica cryostat onto Superfrost plus gold slides (Fisher), dried at 50 8C for 20 min and then stored at K80 8C. Slides were washed in PBS 5 min !2, then treated with proteinase K 10 mg/ml for 20 min, washed in 2 mg/ml glycine for 5 min, PBS for 5 min !2, and post-fixed in 4% paraformaldehyde in PBS for 20 min. After washing in PBS for 5 min !2, samples were acetylated by treatment with 0.1 M triethanolamine hydrochloride (TEA) pH 8.0 for 2 min, then 0.1 M TEA with 0.3% acetic anhydride (added drop wise) for 10 min. The slides were washed in PBS 5 min, water 2 min, the subjected to prehybridization for 2 h at 70 8C in the following solution: 50% formamide, 10% dextran sulfate, 10 mM Tris HCl pH 7.6, 10 mM EDTA, 0.6 M NaCl, 1! Denhardt’s solution, 0.25% sodium dodecyl sulfate (SDS) and 0.2 mg/ml torula yeast RNA. For the hybridization reaction, this solution was replaced with digoxigeninlabeled probe at a concentration of 1 ng/ml and incubated overnight at 70 8C. The next day, the slides were washed in the following solutions at 60 8C since we found that higher temperatures led to loss of sample from the slides: 6X SSC/ 50% formamide for 10 min, 2X SSC for 10 min, then 0.2X SSC 30 min !2. Slides were then cooled to 37 8C and treated with 0.1 mg/ml RNase A for 30 min, then washed in 0.1X SSC for 30 min, then PBS with 0.1% Tween (PBT) for 5 min. Samples were blocked in 10% sheep serum/1% bovine serum albumin (BSA) in PBT at 4 8C for 30 min, then treated with sheep anti-digoxigenin F(ab) fragmentlinked alkaline phosphatase (Roche) 1:1000 for 30 min. After several washes in PBT, samples were incubated in 0.1 M sodium chloride/0.1 M Tris HCl pH 9.5/50 mM
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
MgCl2/0.1% Tween 20 (NTMT) followed by BM purple (neat). Staining was allowed to proceed until the negative control (sense probe) began to develop signal. For the Rbm19 in situs, this required about 2–3 days of staining with daily replacement of BM purple solution. In selected cases the HNPP fluorescent detection kit (Roche) was used to visualize Rbm19 expression. 2.5. Immunohistochemistry Tissue was fixed in 4% PFA in PBS overnight at 4 8C, then embedded in OCT and sectioned at 10 mm thickness. Sections were fixed to the slides and then washed with PBT several times, permiablilized with 1% Triton X-100, then digested with 10 mg/ml proteinase K for 20 min for antigen retrieval. After washing with PBT for 5 min !2, tissue sections were blocked with 10% goat serum/1% BSA/PBT for 30 min, then treated with 1:100 antibody in blocking solution for 1 h. After several washes in PBT, the slides were treated with biotinylated goat anti-rabbit (1:2000) in block, washed in PBT, then developed using the Vectastain ABC elite kit (Vector Labs, Burlingame, CA) and diaminobenzidine (Sigma). The specimens were dehydrated in ethanol and citrisolve, mounted with Permount and photographed in both brightfield and Nomarski views, which were captured and merged digitally in the program Openlab. 2.6. Immunofluorescence Fresh tissue was embedded in OCT and sectioned at 10 mm thickness. Cells for immunofluorescence were plated on laminincoated wells (2 mg/cm2, from Sigma) in Tissue-Tek chamber slides. Cells or tissue were fixed by treatment with 3.2% paraformaldehyde for 10 min, 150 mM glycine for 5 min !2, 0.2% Triton X-100 for 2 min, PBS 5 min !3, blocking with 5% sheep or goat serum in PBS for 1 h at room temperature, then primary antibody (1:500) in 5% sheep serum for 1 h at room temperature. Secondary antibodies were diluted 1:500 in 5% sheep or goat serum, and cells were counterstained with DAPI. Rhodamine phalloidin was from Molecular Probes (now Invitrogen). Images were captured using a Zeiss Axioplan motorized microscope to generate images across the z-axis using Openlab, followed by deconvolution with the program Volocity (Improvision, Lexington, MA).
55
TCTGGATGCCAAAAGGA;36B4 forward GCGACCTGGAAGTCCAACTA, reverse GTGAGGTCCTCCTT GGTGAA. Products were quantified using SYBR green dye using an ABI PRISMw 7000 Sequence Detection System using the following cycle program: 94 8C !3 min, (94 8C !30 s, 58 8C !30 s, 72 8C !1 min) !40. Products were verified by post-amplification melting curve analysis and agarose gel electrophoresis. Expression levels are expressed in arbitrary units for comparison of trends. 2.8. Western blot analysis Caco-2 cells were grown in T-25 culture flasks, harvested by scraping into 1 ml ice cold PBS, pelleted, then lysed in 25 mM HEPES pH 7.4/150 mM NaCl/10% glycerol/0.5% TX-100/10 mM sodium fluoride/1 mM sodium vanadate/ protease inhibitor cocktail (PMSFCbenzamidineCaprotininCleupeptin). 25 mg of protein was loaded per lane. After electrophoresis and transfer to nitrocellulose, blots were probed with 1:2000 anti-rbm19 followed by 1:30,000 peroxidase conjugated goat anti-rabbit. 2.9. Electron microscopy Cells were washed twice with growth media without serum and then fixed in situ on ice with 2.5% glutaraldehyde in 0.1M cacodylate buffer pH 7.4 for 1 h. After fixation the cells were post-fixed with 1% osmium tetroxide for 1 h on ice, rinsed in distilled water and en-block stained with 1% aqueous uranyl acetate for 3 h on ice. Cells were then scraped, pelleted, dehydrated in graded methanol and embedded in Epon 812 epoxy resin. Ultrathin sections, 60 nm, were cut, counterstained with Reynolds lead citrate and examined by transmission electron microscopy (Hitachi H 600 TEM).
Acknowledgements We thank the MGH Cancer Center Pathology Core for technical assistance, the MGH Tumor Bank for samples of human intestinal tissue, and Drs Stephen Duncan, Colin Rudolph, and members of the Mayer lab for critical review of the manuscript. This work was supported by grants from the NIH (R03DK067176-01) and the Children’s Research Institute.
2.7. Real-time PCR RNA from Caco-2 cells was isolated with Trizol, random hexamer-primed reverse transcribed (Superscript II, Invitrogen) and then amplified with the following primers: human Rbm19 forward GGCAAGCCTCGAACCAAA, reverse AGCATCCTGCCCTGGAATA;sucrase isomaltase forward GCTGTGTATGGAGAACGGGT, reverse GAA
References Amsterdam, A., Nissen, R.M., Sun, Z., Swindell, E.C., Farrington, S., Hopkins, N., 2004. Identification of 315 genes essential for early zebrafish development. Proc. Natl Acad. Sci. USA 101, 12792–12797.
56
J.A. Lorenzen et al. / Gene Expression Patterns 6 (2005) 45–56
Andre, M., Ando, S., Ballagny, C., Durliat, M., Poupard, G., Briancon, C., et al., 2000. Intestinal fatty acid binding protein gene expression reveals the cephalocaudal patterning during zebrafish gut morphogenesis. Int. J. Dev. Biol. 44, 249–252. Bjork, P., Bauren, G., Jin, S., Tong, Y.G., Burglin, T.R., Hellman, U., et al., 2002. A novel conserved RNA-binding domain protein, RBD-1, is essential for ribosome biogenesis. Mol. Biol. Cell. 13, 3683–3695. Coulombre, A.J., Coulombre, J.L., 1958. Intestinal development. I. Morphogenesis of the villi and musculature. J. Embryol. Exp. Morphol. 6, 403–411. Dimario, P.J., 2004. Cell and molecular biology of nucleolar assembly and disassembly. Int. Rev. Cytol. 239, 99–178. Dundr, M., Meier, U.T., Lewis, N., Rekosh, D., Hammarskjold, M.L., Olson, M.O., 1997. A class of nonribosomal nucleolar components is located in chromosome periphery and in nucleolus-derived foci during anaphase and telophase. Chromosoma 105, 407–417. Farber, S.A., Pack, M., Ho, S.Y., Johnson, I.D., Wagner, D.S., Dosch, R., et al., 2001. Genetic analysis of digestive physiology using fluorescent phospholipid reporters. Science 292, 1385–1388. Fromont-Racine, M., Senger, B., Saveanu, C., Fasiolo, F., 2003. Ribosome assembly in eukaryotes. Gene 313, 17–42. Hesketh, J., Campbell, G., Piechaczyk, M., Blanchard, J.M., 1994. Targeting of c-myc and beta-globin coding sequences to cytoskeletalbound polysomes by c-myc 3 0 untranslated region. Biochem. J. 298 (Pt 1), 143–148. Jansen, R.P., 2001. mRNA localization: message on the move. Nat. Rev. Mol. Cell Biol. 2, 247–256. Jin, S.B., Zhao, J., Bjork, P., Schmekel, K., Ljungdahl, P.O., Wieslander, L., 2002. Mrd1p is required for processing of pre-rRNA and for maintenance of steady-state levels of 40 S ribosomal subunits in yeast. J. Biol. Chem. 277, 18431–18439. Klein, R.M., 1989. Small intestinal cell proliferation during development. In: Lebenthal, E. (Ed.), Human Gastrointestinal Development. Raven Press, New York, NY, pp. 367–392. Korinek, V., Barker, N., Moerer, P., van Donselaar, E., Huls, G., Peters, P.J., et al., 1998. Depletion of epithelial stem-cell compartments in the small intestine of mice lacking Tcf-4. Nat. Genet. 19, 379–383.
Mayer, A.N., Fishman, M.C., 2003. Nil per os encodes a conserved RNA recognition motif protein required for morphogenesis and cytodifferentiation of digestive organs in zebrafish. Development 130, 3917–3928. Montgomery, R.K., Mulberg, A.E., Grand, R.J., 1999. Development of the human gastrointestinal tract: twenty years of progress. Gastroenterology 116, 702–731. Nagy, A., Gertsenstein, M., Vintersten, K., Behringer, R., 2003. Manipulating the Mouse Embryo. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Nomura, M., Gourse, R., Baughman, G., 1984. Regulation of the synthesis of ribosomes and ribosomal components. Annu. Rev. Biochem. 53, 75–117. Okano, H., Imai, T., Okabe, M., 2002. Musashi: a translational regulator of cell fate. J. Cell Sci. 115, 1355–1359. Pack, M., Solnica-Krezel, L., Malicki, J., Neuhauss, S.C.F., Schier, A.F., Stemple, D.L., et al., 1996. Mutations affecting development of zebrafish digestive organs. Development 123, 321–328. Potten, C.S., Booth, C., Tudor, G.L., Booth, D., Brady, G., Hurley, P., et al., 2003. Identification of a putative intestinal stem cell and early lineage marker; musashi-1. Differentiation 71, 28–41. Rousset, M., 1986. The human colon carcinoma cell lines HT-29 and Caco2: two in vitro models for the study of intestinal differentiation. Biochimie 68, 1035–1040. Saijou, E., Fujiwara, T., Suzaki, T., Inoue, K., Sakamoto, H., 2004. RBD-1, a nucleolar RNA-binding protein, is essential for Caenorhabditis elegans early development through 18S ribosomal RNA processing. Nucleic Acids Res 32, 1028–1036. Sancho, E., Batlle, E., Clevers, H., 2004. Signaling pathways in intestinal development and cancer. Annu. Rev. Cell Dev. Biol. 20, 695–723. Schlosser, I., Holzel, M., Murnseer, M., Burtscher, H., Weidle, U.H., Eick, D., 2003. A role for c-Myc in the regulation of ribosomal RNA processing. Nucleic Acids Res. 31, 6148–6156. Sherr, C.J., 2004. Principles of tumor suppression. Cell 116, 235–246. Sugimoto, M., Kuo, M.L., Roussel, M.F., Sherr, C.J., 2003. Nucleolar Arf tumor suppressor inhibits ribosomal RNA processing. Mol. Cell 11, 415–424.