In the amphibian gastrointestine during metamorphosis, the primary (larval)epithelium undergoes apoptosis. By contrast, a small number of undifferentiated cells including stem cells actively proliferate and differentiate into the secondary (adult) epithelium that resembles the mammalian counterpart. In the present study, to clarify whether Musashi-1(Msi-1), an RNA-binding protein, serves as a marker for progenitor cells of the adult epithelium, we chronologically examined Msi-1 expression in the Xenopus laevis gastrointestine by using in situ hybridization and immunohistochemistry. Similar expression profiles of Msi-1 were observed at both mRNA and protein levels. In both the small intestine and the stomach, the transient expression of Msi-1 during metamorphosis spatio-temporally correlated well with active proliferation of the progenitor cells including stem cells of the adult epithelium but did not with apoptosis of the larval epithelium. As the adult progenitor cells differentiated into organ-specific epithelial cells after active proliferation, Msi-1 expression was rapidly downregulated. Therefore, Msi-1 is useful to identify the adult progenitor cells that actively proliferate before final differentiation in the amphibian gastrointestine. Furthermore, our culture experiments have shown that thyroid hormone (TH) organ-autonomously induces Msi-1 expression only in the adult progenitor cells of the X. laevis intestine in vitro as in vivo. However, TH could not induce Msi-1 expression in the intestinal epithelium separated from the connective tissue, where the adult epithelium never developed. These results suggest that Msi-1 expression is upregulated by TH in the adult progenitor cells under the control of the connective tissue and plays important roles in their maintenance and/or active proliferation during amphibian gastrointestinal remodeling.
Amphibian metamorphosis is triggered by a single hormone, thyroid hormone(TH) (Dodd and Dodd, 1977; Kikuyama et al., 1993), and offers us a unique model system for the study of organ remodeling(Shi, 1999). In particular, to adapt to terrestrial carnivorous life, the amphibian gastrointestine dramatically undergoes remodeling from larval to adult form during a short period. At the cellular level, the primary (larval) epithelium undergoes apoptosis (Ishizuya-Oka and Ueda,1996), whereas a small number of undifferentiated cells appear,replace the larval epithelium by active proliferation, and newly form the secondary (adult) epithelium (Hourdry and Dauca, 1977; Shi and Ishizuya-Oka, 1996). In the intestine, the adult epithelium acquires the cell renewal system along the trough-crest axis of newly formed folds (McAvoy and Dixon, 1977; Shi and Ishizuya-Oka, 1996)analogous to the mammalian crypt-villus axis(Cheng and Bjerknes, 1985; Madara and Trier, 1994), and finally differentiates into major absorptive epithelial cells expressing intestinal fatty acid-binding protein (IFABP)(Shi and Hayes, 1994; Ishizuya-Oka et al., 1997),goblet cells and endocrine cells (McAvoy and Dixon, 1978). Meanwhile, in the stomach, the adult epithelium acquires the cell renewal system where proliferating cells are localized in the neck region of glands (Oinuma et al.,1992) similar to that of the mammalian gastric glands(Lee and Leblond, 1985; Nomura et al., 1998), and finally differentiates into the surface mucous epithelium and the glandular epithelium, which consists of major glandular cells expressing pepsinogen (Pg)(Inokuchi et al., 1995; Ishizuya-Oka et al., 1998),mucous neck cells, and endocrine cells(Holmberg et al., 2001). Therefore, it is suggested that progenitor cells of the adult epithelium that appear as undifferentiated cells during metamorphosis in these organs include multipotent stem cells analogous to those in the mammalian gastrointestine. It is interesting from the standpoint of cell biology how the adult progenitor cells appear and form the mammalian-like epithelium. However, the progenitor cells of the adult gastrointestinal epithelium have thus far been only identified as morphologically undifferentiated cells stained strongly with pyronin Y (Ishizuya-Oka and Ueda,1996). The lack of biochemical markers has made it difficult to characterize the adult progenitor cells and has retarded the study of their origin and the mechanisms controlling their behavior.
Musashi-1 (Msi-1), an RNA-binding protein, was initially reported in the sensory organ of Drosophila(Nakamura et al., 1994) and has been shown to serve as a marker for proliferative neural precursor cells including stem cells in the central nervous system (CNS)(Sakakibara et al., 1996; Sakakibara and Okano, 1997). Msi-1 expression in neural precursor cells is evolutionally conserved in different species of vertebrates such as humans, mice and Xenopus laevis (Kaneko et al.,2000). Although its precise mechanisms have not yet been clarified, Msi-1 is supposed to be involved in the early asymmetric divisions that generate differentiated cells from neural stem cells(Okano, 1995). Interestingly,Msi-1 is also expressed in the murine intestine other than the neural tissues(Sakakibara et al., 1996). Moreover, it has been shown recently that Msi-1 is preferentially expressed in the predicted stem cell region of murine intestinal crypts, suggesting that Msi-1 may be a natural marker for intestinal stem cells and their immediate descendants (Booth and Potten,2000; Potten et al.,2003). These recent reports led us to study the expression of Msi-1 during amphibian gastrointestinal remodeling.
In the present study, to clarify whether the adult progenitor cells in the amphibian gastrointestine express Msi-1 as in the mammalian intestine, we examined Msi-1 expression in the X. laevis small intestine and the stomach during metamorphosis at mRNA and protein levels. We show that expression profiles of Msi-1 coincide well with active proliferation of the adult progenitor cells in both organs, suggesting that Msi-1 serves as a marker for the stem cells common to the amphibian and mammalian intestines. Moreover, we demonstrate that cell-specific Msi-1 expression can be reproduced in vitro by the inductive action of TH in the presence of the connective tissue but not in its absence.
Materials and Methods
Tadpoles of the South African clawed frog (X. laevis) were purchased from a commercial source and staged according to Nieuwkoop and Faber(Nieuwkoop and Faber,1967).
RNA isolation and reverse transcription-polymerase chain reaction(RT-PCR) analysis
Tissue fragments were isolated from the small intestine and the stomach of tadpoles at various stages, and of stage 57 tadpoles treated with or without 5 nM 3, 5, 3′-L-triiodo-thyronine (T3). Total RNA was purified from these organs using Trizol reagent (Gibco-BRL, Grand Island, NY). RNAs were prepared from a mixture of more than three tadpoles at each stage. For each reaction, 1 μg RNA was reverse transcribed to oligo(dT)-primed first-strand cDNA by using a cDNA synthesis kit (Amersham Pharmacia Biotech,Little Chalfont, UK), and the resulting cDNA was subjected to 30 cycles of PCR with Msi-1-specific primers (5′-ATGGAGACAGAAGCGCCCCAGCCCGGACTG-3′and 5′-TCAGTGGTAGCCGTTGGTGAAAGCAG-3′)(Kaneko et al., 2000) and to 25 cycles with EF-1α-specific primers(5′-CCTGAATCACCCAGGCCAGATTGGTG-3′ and 5′-GAGGGTAGTCTGAGAAGCTCTCCACG-3′)(Suzuki et al., 1993). The PCR products (5 μl) were electrophoresed through a 2% agarose gel and visualized by ethidium bromide staining.
Tissue fragments were isolated from the anterior part of the small intestine at stage 57 and were slit open lengthwise with fine forceps. Some fragments were treated with dispase (1000 units/ml; Godo, Tokyo, Japan), and their epithelial components were isolated and put on a growth factor-reduced matrigel (Becton Dickinson, Bedford, MA). They were then cultured as described previously (Ishizuya-Oka and Shimozawa,1991). Briefly, they were placed on membrane filters (type HAWP;Millipore, Bedford, MA) on stainless steel grids and were cultured in 60%Leibovitz-15 medium supplemented with 100 IU/ml of penicillin, 100 μg/ml of streptomycin (Gibco-BRL), and 10% charcoal-treated fetal bovine serum(Gibco-BRL). To induce metamorphosis, T3, insulin and hydrocortisone (Sigma) were added to the medium at 10 nM, 5 μg/ml and 0.5μg/ml, respectively. The culture medium was changed every other day for 7 days at 26°C.
In situ hybridization
A cDNA fragment of nucleotides (nt) 74-1117 was amplified by PCR with the forward primer (5′-ATGGAGACAGAAGCGCCCCAGCCCGGACTG-3′; nt 74-103),the reverse primer (5′-TCAGTGGTAGCCGTTGGTGAAAGCAG; nt 1117-1092) and the template nrp1-pSP36T, containing Xenopus Msi-1 full-length cDNA. The cDNA fragment was subcloned into pCRII vector with TOPO TA cloning kit(Invitrogen, Carlsbad, CA). Procedures for in situ hybridization were the same as those described previously(Ishizuya-Oka et al., 1994). In brief, digoxigenin (DIG)-11-UTP-labeled antisense and sense probes were prepared with DIG RNA-labeling kit (Roche Diagnostics, Mannheim, Germany)according to the manufacturer's protocol. Tissue fragments were fixed with 4%paraformaldehyde in phosphate-buffered saline (pH 7.4) at 4°C for 4 hours,frozen on dry ice, and cut at 7 μm. Sections were treated with 0.2 N HCl,digested with 1 μg/ml proteinase K (Wako, Osaka, Japan), and then fixed again with 4% paraformaldehyde. Hybridization buffer containing DIG-labeled antisense RNA probe (200 ng/ml) was applied to the pretreated sections. After hybridization at 40°C for 18 hours, the sections were treated with 20μg/ml RNase A (Wako) at 37°C for 30 minutes to remove excess unhybridized probes. They were then washed, processed for immunological detection of the hybridized DIG probes according to the manufacturer's instructions (DIG-probe Detection Kit, Roche Diagnostics), and examined microscopically. As controls, some sections were hybridized with DIG-labeled sense RNA probe (200 ng/ml).
Immunohistochemistry for IFABP, Pg and PCNA
Tissue fragments and cultured explants were fixed with 95% ethanol at 4°C for 4 hours, embedded in paraffin and cut at 5 μm. The sections were incubated at 4°C overnight with the rabbit anti-mouse Msi-1 polyclonal antibody (diluted 1:100)(Sakakibara et al., 1996) or with the rat anti-mouse Msi-1 monoclonal antibody (diluted 1:100), which has been shown to recognize the amino acid sequences highly conserved between Xenopus and mouse (Kaneko et al.,2000). They were then incubated with the biotinylated secondary antibody, followed by incubation with avidin-conjugated horseradish peroxidase(HRP) (Vector Labs, Burlingame, CA). HRP reactions were developed using 0.02%3, 3′-diamino-benzidine-4HCl (DAB) and 0.006%H2O2. Although the immunostaining with the monoclonal antibody was somewhat weaker than that with the polyclonal antibody, their expression profiles were spatio-temporally consistent (data not shown).
To distinguish between the adult progenitor cells and the remaining larval cells during stages 60-62, some sections close to the sections used for Msi-1 immunohistochemistry were stained with methyl green-pyronin Y (Muto, Tokyo,Japan) for 5 minutes. The adult progenitor cells were stained intensely red because of their RNA-rich cytoplasm, whereas the larval epithelial cells undergoing apoptosis were stained much weaker(Ishizuya-Oka and Ueda, 1996). In addition, other sections were incubated at room temperature for 1 hour with the following antibodies: the mouse anti-PCNA monoclonal antibody (diluted 1:100; Novacastra, Newcastle, UK) for proliferating cells; the rabbit anti-Xenopus IFABP antibody (diluted 1:1000) (generous gift of Dr Y.-B. Shi) (Ishizuya-Oka et al.,1997) for differentiated intestinal absorptive cells; and the rabbit anti-bullfrog Pg antibody (diluted 1:10,000) (generous gift of Dr T. Inokuchi) (Ishizuya-Oka et al.,1998) for differentiated glandular cells. They were then visualized by sequential incubation with streptavidinbiotin-peroxidase complex(Nichirei, Tokyo, Japan) and DAB/H2O2 as described above. There was no positive staining when the same concentration of pre-immune or normal serum was applied for each antiserum as the specificity control (not shown). At least three specimens were examined for each developmental stage or culture day.
Finally, to examine the relationship between Msi-1 immunoreactivity and cell proliferation more directly, some sections were double immunostained with a mixture containing the rabbit anti-Msi-1 antibody and the mouse anti-PCNA antibody at 4°C overnight. They were then incubated with a mixture containing Alexa Fluor488-conjugated anti-rabbit IgG (1:500; Molecular Probes,Eugene, OR) and Alexa Fluor568-conjugated anti-mouse IgG (1:500; Molecular Probes) and observed by fluorescence microscopy.
Expression of Msi-1 mRNA during normal and TH-induced metamorphosis
To clarify whether Msi-1 mRNA was truly expressed in the X. laevisintestine, its developmental expression was analyzed by RT-PCR. The level of Msi-1 mRNA began to increase at the onset of metamorphic climax (around stage 61) and peaked around stage 62 (Fig. 1A), when the circulating level of TH peaks(Leloup and Buscaglia, 1977). Subsequently, the level of Msi-1 mRNA decreased towards the end of metamorphosis. Similarly, in the stomach, the expression of Msi-1 mRNA was transiently upregulated around stages 61-62. In addition, to clarify whether TH upregulates Msi-1 mRNA expression in vivo, stage 57 tadpoles were treated with TH. Msi-1 mRNA was precociously detected in the both organs after 5 days of TH treatment (Fig. 1B), when metamorphic changes occur most dramatically(Shi and Ishizuya-Oka,2001).
Msi-1 mRNA expression is specific for progenitor cells of adult epithelium
To identify which type of cells express Msi-1 mRNA, we performed in situ hybridization analysis. Consistent with the RT-PCR data described above(Fig. 1A), Msi-1 mRNA was neither detected in the intestine (Fig. 2A) nor the stomach (Fig. 2F) throughout pre- and prometamorphosis. In the intestine, around the onset of metamorphic climax (stages 60-61), when progenitor cells of the adult epithelium were identified as islets stained strongly with pyronin-Y(Fig. 2B), Msi-1 mRNA became detectable only in these adult progenitor cells(Fig. 2C). By contrast, all the control hybridizations with the sense RNA probe gave only background levels of signals (Fig. 2D). Thereafter,as morphogenesis of intestinal folds proceeded, the level of Msi-1 mRNA rapidly decreased from the crest to trough of the folds(Fig. 2E). By the end of metamorphosis (stage 66), Msi-1 mRNA became hardly detectable, if any. Similarly, in the stomach, Msi-1 mRNA was transiently expressed only in pyronin-Y-positive adult progenitor cells during stages 60-61(Fig. 2G,H). Then, as the adult progenitor cells differentiated into the surface and glandular epithelia, the level of Msi-1 mRNA rapidly decreased. By stage 66, Msi-1 mRNA became hardly detectable, if any, except for nonspecific staining in the neck region of glands (Fig. 2I).
Transient expression of Msi-1 protein correlates with active proliferation of adult progenitor cells
To investigate the relationship between Msi-1 expression and development of the adult epithelium more precisely, we analyzed the expression profile of Msi-1 proteins by immunohistochemistry.
Throughout pre- and prometamorphosis, the larval epithelium remained simple columnar and negative for Msi-1 (Fig. 3A). At stage 60, when progenitor cells of the adult epithelium were first identified as pyronin-Y-positive small islets(Fig. 3B), Msi-1 immunoreactivity became weakly detectable only in these islets(Fig. 3C). Then, at stage 61,when the adult progenitor cells rapidly replaced the larval epithelium(Fig. 3D) by active proliferation (Fig. 3E), almost all of the progenitor cells were positive for Msi-1(Fig. 3F). Double immunofluorescence labeling with Msi-1 and PCNA antibodies indicated that Msi-1-positive cells have a high activity of cell proliferation(Fig. 4A). By contrast, the degenerating larval epithelial cells remained negative(Fig. 3D,F). After the completion of larval-to-adult epithelial cell replacement (stage 63)(Fig. 3G), Msi-1 immunoreactivity rapidly decreased and became weakly detectable only in the trough of newly formed folds (Fig. 3H), where proliferating cells became localized(Fig. 3I). By contrast, the immunoreactivity of IFABP, a marker for differentiated absorptive cells,increased from the crest to trough of the folds(Fig. 3J). At the end of metamorphosis, Msi-1 expression was hardly detectable, if any.
As in the intestine, Msi-1 became detectable at stage 60 only in the adult progenitor cells, which appeared in the basal region of larval glands(Fig. 5A-C) and were actively proliferating (Fig. 5D). Then,at stage 61, the proliferating adult progenitor cells rapidly replaced the degenerating larval epithelium (Fig. 5E,F). Almost all of the adult progenitor cells were strongly positive for Msi-1, whereas all of the larval epithelial cells remained negative (Fig. 5G). Msi-1-positive cells at this stage have a high activity of cell proliferation(Fig. 4B) just like in the intestine. After stage 62, when the adult epithelium completely replaced the larval epithelium and began to differentiate into the surface and glandular epithelia (Fig. 5H) expressing Pg, a marker for differentiated glands(Fig. 5I), Msi-1 expression rapidly decreased (Fig. 5J). Towards the end of metamorphosis, Msi-1 immunoreactivity became hardly detectable except for a very weak one in the neck region of adult glands(Fig. 5L), where proliferating cells were localized (Fig. 5K).
TH organ-autonomously upregulates Msi-1 expression in vitro
To clarify whether TH upregulates Msi-1 expression in vitro, we cultured larval intestines isolated from stage 57 tadpoles in the presence or absence of TH. Around day 5 in the presence of TH, progenitor cells of the adult epithelium were identified as pyronin-Y-positive islets(Fig. 6A) that were actively proliferating (Fig. 6B). Although the intensity of Msi-1 immunoreactivity was lower than in vivo, it was localized in the adult progenitor cells but not in the remaining larval cells as in vivo (Fig. 6C). Towards day 7, when the adult epithelium differentiated into the absorptive epithelium expressing IFABP (Fig. 6E), Msi-1-immunoreactive cells became undetectable again(Fig. 6F). By contrast, in the absence of TH, the differentiated larval epithelium remained negative for Msi-1 throughout the cultivation (Fig. 6D).
Furthermore, to determine whether the connective tissue surrounding the epithelium is involved in TH upregulation of Msi-1 expression, we separated the epithelium from the connective tissue and cultured it alone in the presence of TH. Around day 5, some epithelial cells became negative for pyronin-Y staining, just like degenerating larval cells in the presence of the connective tissue (Fig. 6G). However, the adult progenitor cells could be hardly distinguished by pyronin-Y staining because the other cells were stained at various intensities. In addition, the epithelial cells remained as a single layer and did not form islet structures. Thereafter, the epithelium gradually decreased in cell number and never differentiated into the adult absorptive epithelium expressing IFABP. In this condition, no Msi-1-immunoreactive cells could be detected throughout the cultivation (Fig. 6H).
Msi-1 serves as a common stem cell marker in both amphibian and mammalian intestines
In the present study, the transient expression of Msi-1 coincides well with active proliferation of progenitor cells of the adult epithelium in the intestine, as schematically summarized in Fig. 7. Similar correlation was also observed in the stomach. Although previous studies reported some TH response genes whose expression is epithelial specific in the amphibian intestine (Shi and Ishizuya-Oka,1996; Ishizuya-Oka et al.,2001), Msi-1 is the only one that is specifically expressed in adult progenitor cells that are actively proliferating, but not in differentiated adult epithelial cells nor larval cells destined to undergo apoptosis. As for other markers, pyronin-Y strongly stains the adult progenitor cells. However, it also stains the larval epithelium before apoptosis, as well as the differentiated adult epithelium, and other tissues(Ishizuya-Oka and Ueda, 1996),because the RNA-rich cytoplasm is stained with pyronin-Y regardless of cell types (Schulte et al., 1992). Similarly, PCNA immunostains not only the adult progenitor cells but also the larval epithelium and other tissues at various frequencies. Thus, Msi-1 is the first reliable marker for the adult progenitor cells including stem cells in the amphibian gastrointestine. It is noteworthy that Msi-1 expression is also expressed in the mammalian intestine(Sakakibara et al., 1996),where putative stem cells and at least the next two generations are positive for Msi-1 (Potten et al.,2003). Although Msi-1 expression became hardly detectable towards the end of metamorphosis, possibly due to a low immunoreactive sensitivity in this study, Msi-1 expression could be detectable in the region where proliferating cells were localized in both the intestine and the stomach, that is, in the trough of intestinal folds corresponding to the mammalian crypt(Cheng and Bjerknes, 1985; Potten et al., 1997) or in the neck region of gastric glands similar to that of mammalian glands(Lee and Leblond, 1985). These results indicate that the adult progenitor cells during amphibian metamorphosis include stem cells analogous to the mammalian ones, and molecular mechanisms underlying the control of these stem cells are conserved across the amphibian and mammalian gastrointestines.
Recently, in the mammalian CNS, it has been shown that Msi-1 protein binds to (G/A) UnAGU sequences in the 3′-untranslated region (3′-UTR) of Numb mRNA and represses the expression of Numb protein(Imai et al., 2001), which is a membrane-associated antagonist of Notch signaling(Wakamatsu et al., 1999). Meanwhile, in the mammalian intestine, Notch signaling has been suggested to be involved in the determination and/or maintenance of stem cells(van den Brink et al., 2001). Taken together, one possible role of Msi-1 is to maintain the adult progenitor cells through activation of Notch signaling. Alternatively, Msi-1 may bind to(G/A) UnAGU sequences of other target genes that activate proliferation of the adult progenitor cells. The nature of these target genes awaits further investigation.
TH upregulates Msi-1 expression in adult progenitor cells in vitro as in vivo
In the mammalian intestine, it has been reported that Msi-1 expression is upregulated during development of adenomas and during regeneration after irradiation (Potten et al.,2003). However, it remains unknown which molecules upregulate Msi-1 expression. In the present study, the upregulation of Msi-1 expression coincides temporally with the peak of circulating TH level during X. laevis metamorphosis (Leloup and Buscaglia, 1977). More importantly, our culture study has shown that TH upregulates organ-autonomously Msi-1 expression in the larval intestine. This means that Msi-1 gene is a TH response gene, as recently reported in the developing rat brain(Cuadrado et al., 2002). Furthermore, we have shown that Msi-1 expression in vitro is closely associated with adult progenitor cells that are actively proliferating just like in vivo. This reinforces our proposal that TH-induced Msi-1 is involved in the maintenance and/or active proliferation of adult progenitor cells.
Until now, little is known about the origin of the adult progenitor cells in the amphibian gastrointestine. In the intestinal epithelium before metamorphosis, no undifferentiated cells can be morphologically identified(Marshall and Dixon, 1978). In agreement with this, Msi-1 expression was not detected in the intestinal epithelium before metamorphosis in the present study. This implies that at least partially differentiated cells can give rise to progenitor and/or stem cells during amphibian metamorphosis. Similar cases have recently been reported in the mammalian tissues such as committed oligodendrocyte precursor cells that become multipotent stem cells(Kondo and Raff, 2000) and multinucleated myotubes that give rise to mononucleated proliferative myoblasts (Odelberg et al.,2000). Although the important roles of microenvironments known as`niche' in the control of stem cells are generally recognized(Potten et al., 1997; Blau et al., 2001; Mills and Gordon, 2001; Spradling et al., 2001), far less is known about microenvironmental factors that play key roles in reversing the differentiated state (i.e. de-differentiation into stem cells). In this study, we have shown that the adult progenitor cells expressing Msi-1 could be detected after 5 days of TH treatment in vitro in the connective tissue but not in its absence. This predicts that genes whose expression is upregulated by TH in the connective tissue within 5 days play important roles in the de-differentiation. To identify such factors, we recently isolated genes whose expression is upregulated by TH in the connective tissue by subtractive differential screening(Shimizu et al., 2002). It is worth functionally analyzing these genes as well as other TH response genes previously isolated from the X. laevis intestine(Shi and Brown, 1993; Amano and Yoshizato, 1998).
In conclusion, Msi-1 is useful as a marker for adult progenitor cells including stem cells in the amphibian gastrointestinal epithelium like in the mammalian intestinal epithelium. Amphibian gastrointestinal remodeling, where TH can induce stem cells expressing Msi-1 followed by a new epithelial formation, should provide a unique model system to clarify microenvironmental factors playing key roles in the control of stem cells.
We express our deep gratitude to Y.-B. Shi (National Institutes Health) and T. Inokuchi (Utsunomiya University) for their generous gifts of antibodies. This work was supported in part by the MEXT Grants-in-Aid for Scientific Research (C) and Research Grant of Seki Minato Foundation (to A.I.-O.).