Myb Controls Intestinal Stem Cell Genes and SelfRenewal

TISSUE-SPECIFIC STEM CELLS Myb Controls Intestinal Stem Cell Genes and Self-Renewal DANE CHEASLEY,a LLOYD PEREIRA,a,b SALLY LIGHTOWLER,a,b ELIZABETH VINCAN,c,d JORDANE MALATERRE,a,b ROBERT G. RAMSAYa,b Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia; bDepartment of Pathology and cDepartment of Anatomy and Cell Biology, University of Melbourne, Melbourne, Victoria, Australia; dVictorian Infectious Diseases Reference Laboratories, Melbourne, Victoria, Australia a

Key Words. Myb • lgr5 • bmi-1 • olfm4 • dcamkl-1 • Colon cancer • Hypomorph • Intestinal stem cells • b-Catenin • Transcription • Gene expression

ABSTRACT Rapid advances have been made in the understanding of how the highly proliferative gastrointestinal tract epithelium is regulated under homeostasis and disease. The identification of putative intestinal stem cell (ISC) genes and the ability to culture ISC capable of generating all four lineages plus the architecture of small intestinal (SI) crypts has been transformative. Here, we show that transcription factor Myb governs ISC gene expression, particularly Lgr5. Lgr5 is associated with cells that have the capacity to generate all cell lineages in SI organoid cultures and colorectal cancer cells, which overexpress Myb. Furthermore, Wnt signaling and Myb cooperate in maximal Lgr5 promoter activation while hypomorphic Myb (plt4/plt4) mice have decreased Lgr5 expression. After ionizing radia-

tion (IR), ISC genes are elevated; but in plt4/plt4 mice, this response is substantially subdued. ISC genes bmi-1 and olfm4 are expressed at subnormal levels in plt4/plt4 mice, and bmi-1 is induced with IR to half the level in mutant mice. dcamkl-1 and olfm4 failed to recover after IR in both wild-type (wt) and mutant mice. Although not considered as an ISC gene, cyclinE1 is nevertheless used to assist cells in the emergence from a quiescent state (an expectation of ISC following IR) and is overexpressed after IR in wt mice but does not change from a very low base in plt4/ plt4 mice. Self-renewal assays using organoid cultures and inducible Myb knockout studies further highlighted the dependence of ISC on Myb consistent with role in other stem cell-containing tissues. STEM CELLS 2011;29:2042–2050

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION Identification of intestinal stem cells (ISCs) and characterizing their function is reasonably thought to have direct bearing on understanding colorectal cancer (CRC) [1]. Conversely, several genes that are aberrantly expressed in CRC may not be exclusively or even normally expressed in ISC but may impart stem cell functional properties ranging from inhibition of differentiation, enhanced self-renewal, and extended proliferative capacity. Monitoring these functions necessitates phenotypic or molecular markers of which several have come to the fore and have been reviewed comprehensively [2]. One of the lead ISC markers is the Wnt-signaling potentiating R-Spondin receptor, Lgr5 [3–5]. Accordingly, the regulation of Lgr5 tracks with activated Wnt signaling [3] specifically the transcription factors b-catenin/TCF4 and ASCL2 [6]. However, studies addressing the direct transcriptional regulation of Lgr5 have not been reported. Many genes and pathways that affect intestinal cell differentiation and proliferative capacity have also been implicated

in CRC. MYB is one such gene that is overexpressed in CRC compared to most other cancers (except breast cancer) and when deleted or hypomorphic affects intestinal cell proliferation and differentiation [7]. Furthermore, clues that Myb might be important in ISC function have come from treating Myb heterozygous mice with cytotoxic agents. Such studies indicated that normal colon recovery required both Myb alleles [8]. As the proto-oncogene Myb encodes a transcription factor required for stem/progenitor function in the hematopoietic system, the finding that colon stem/progenitor cell recovery was also Myb-dependent might have been anticipated. To explore this prospect, three hypomorphic Myb mutant mice were investigated and found to have shorter colonic crypts, defects in proliferation, and all were found to be deficient in cyclinE1 expression [9], a gene required for cell cycle re-entry from quiescence [10]. This role of quiescence in maintaining ISC reserve in the intestine is a matter of great interest [2, 11-13] and the subsequent identification of a series of ISC gene opened an opportunity to investigate whether Myb might be involved in regulating stem and progenitor cells in this tissue.

Author contributions: D.C.: collection and assembly of data, manuscript writing, and data analysis and interpretation; L.P. and E.V.: collection and assembly of data and data analysis and interpretation; S.L.: collection and assembly of data; J.M.: collection and assembly of data, conception and design, manuscript writing, and data analysis and interpretation; R.R.: conception and design, financial support, manuscript writing, final approval of manuscript, and data analysis and interpretation. Correspondence: Robert G. Ramsay, Ph.D., Differentiation and Transcription Laboratory, Peter MacCallum Cancer Centre, East Melbourne, 3002 Victoria, Australia. Telephone: 613-9656-1863; Fax: 613-9656-1411; e-mail: [email protected] Received February 15, 2011; accepted for publication September 27, 2011; first published online in STEM CELLS EXPRESS October 13, 2011. C AlphaMed Press 1066-5099/2011/$30.00/0 doi: 10.1002/stem.761 V

STEM CELLS 2011;29:2042–2050 www.StemCells.com

Cheasley, Pereira, Lightowler et al.

As a point of reference in our transcription studies, we used our previously reported finding that the regulation of Wnt signaling target gene Myc in the intestines is strongly influenced by the presence of Myb [14]. Indeed, in most contexts where Myc expression is arguably elevated by Wnt signaling, Myb is also present, and when Myb expression is in haploinsufficiency, Wnt signaling-mediated adenoma formation is perturbed [14]. In view of this experience, the potential relationship between Myb and Lgr5 in the intestines was explored. Using a combination of coexpression, reporter and assays for ISC function, we report here that Myb is central to Lgr5 expression in the intestines. We show that Wnt signaling directly regulates Lgr5 but it does so more effectively when Myb is present and very poorly when Myb is functionally defective. By using radiation treatment, we further show that Myb is required for the normal and stress-mediated production of such stem/progenitor cells and markers thereof genes.

MATERIALS

AND

METHODS

Mice C57/Bl6 plt4/plt4 and plt3/plt3 hypomorphic Myb mutant mice and their colon defects have been described elsewhere [9]. Lgr5EGFP-IRES-creERT2 [3], VillinCreERT2 [15, 16], Myblox/lox mice [15, 16], and APCmin/þ mice [17] have been reported previously. Enhanced green fluorescent protein–positive (eGFP-positive) cells were isolated as single cells from colons by FACS as described [3]. Whole body irradiation (WBI) studies were conducted as described previously [8].

Cell Lines The propagation and derivation of cell lines have been described for Colo205 [18], SW620, SW480 [19], LIM1863, LIM1215, LIM2412, LIM2537 [20], HCT15 [21], HCT116 [22], NIH3T3, MC38 [23], and YAMC [24], and relevant characteristics are tabulated (Supporting Information Table 1).

Small Intestinal Organoid Cultures and Ex Vivo Irradiation Assay Crypt nests (small clusters of dispersed epithelial cells) obtained from liberated crypts enriched from small intestine (SI) were enumerated using trypan blue exclusion, and the required number was seeded in Matrigel (BD Biosciences, North Ryde, Australia) overlayed with 500 ll of Dulbecco’s modified Eagle’s medium/ F12 (Sigma, St. Louis, MO) containing 20 ng/ml EGF (BD Biosciences), 10 ng/ml basic fibroblast growth factor (Roche, Kew, Australia), 500 ng/ml R-Spondin (R&D Systems, Minneapolis, MN), 100 ng/ml Noggin (Peprotech, Rocky Hill, NJ), and B27 supplement (Invitrogen) as originally reported [25]. Cultures were established on day 0 and irradiated the next day using a c-cell irradiator. Viable organoids were scored 10 days post-irradiation following incubation in 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide. 4-Hydrotamoxifen (4OHT; Sigma) at 0.1 lM was added to organoid cultures at day 0 and harvested at day 3 for RNA analysis.

Western Blotting Cellular proteins were partitioned into nuclear and cytoplasmic fractions by differential centrifugation in 0.5% Nonident P40 lysis buffer prior to separation on 10% polyacrylamide/sodium dodecyl sulfate gels and transferred to polyvinylidene fluoride membrane [26]. Membranes were blocked in 10% nonfat skim milk powder and probed with antibodies described in Supporting Information Table 2.

www.StemCells.com

2043

Immunohistochemistry Immunohistochemistry (IHC) was performed using Envisionþ kit (Dako) according to the manufacturer’s instructions. Antigen retrieval was carried out in 10 mM Tris buffer and 1 mM EDTA pH 8 for 3 minutes in a pressure cooker at 125 C (Biocare Decloker). Sections were cut to 2 or 4 lm. Antibodies are tabulated in Supporting Information Table 2. Slides were counterstained with hematoxylin and/or periodic acid schiff reagent [9] and dehydrated in ethanol and coverslipped.

Histology Mice were sacrificed by cervical dislocation at required time points following WBI. Twelve segments of 0.5 cm in length were collected from the SI jejunum, fixed in 10% normal buffered formalin, embedded in paraffin, and sectioned transversely. Distal colon was used for analysis and sections were cut longitudinally [8]. Sections were stained with hematoxylin and eosin. Olfm4 in situ hybridization (ISH) was preformed as described elsewhere [27].

Chromatin Immunoprecipitation Assays and Transfections of NIH3T3 Cells Details are described in full in Supporting Information.

Reporter Assays and RT-PCR Analysis of Endogenous Gene Expression pCATbasic reporter constructs containing the mouse Lgr5 promoter region were generated using Lgr5 promoter primers described in Supporting Information Table 2 along with primers used in qRT-PCRs for different endogenous genes. Transactivation studies using full-length Myb, D89-b-catenin, and the Myb promoter have been described elsewhere [14].

RESULTS Coexpression of Lgr5 and Myb in Intestinal Cells In view of reports that Lgr5 is regulated by Wnt signaling [3, 6] and that Myb cooperates with activated b-catenin [14], we performed an in silico analysis of the LGR5/Lgr5 human and mouse promoters. Multiple TCF4 and Myb consensus binding sites were evident (Supporting Information Fig. 1) consistent with the Lgr5 being a Wnt target gene as well as raising the possibility that Myb may also regulate this gene. In addition, based on comparisons of epithelial tumor mRNA profiling studies (Source: Stanford University), we noted an association between MYB and LGR5 expression in colon and ovarian cancer (Supporting Information Fig. 2A), and by IHC in murine colon, we noted Myb expression in Lgr5þve cells mostly in the context of weak or undetectable nuclear b-catenin (Supporting Information Fig. 2B, 2C). As expected, we found that MYB expression was a consistent feature of CRC cell lines [7] (with a spectrum different genetic defects; Supporting Information Table 1) while Lgr5 expression was more restricted (Fig. 1A) as has been shown previously [28]. Colo205 cells that harbor amplified MYB express very high LGR5. Similarly, the murine CRC line MC38 expresses high levels of Myb compared to immortalized colon line YAMC [7] and significantly higher levels of Lgr5 mRNA (Fig. 1B). MC38 tumors showed high proliferation (proliferation cell nuclear antigen positive) and Myb [29]. Intriguingly, Lgr5 staining was not uniform (Supporting Information Fig. 3A). Nuclear b-catenin was detected rarely (not shown) and not in the same regions as Lgr5 or in colon crypts (Supporting Information Fig. 2B, 2C). Thus, there appeared to be disconnection between b-catenin nuclear translocation (and presumably Wnt signaling activation) and Lgr5

2044

Myb in GI Stem Cells

Figure 1. Lgr5/LGR5 expression tracks with Myb/MYB in human colorectal cancer and mouse intestinal cell lines. (A, B): Total RNA was extracted from human colorectal cancer cell lines, the mouse colon adenocarcinoma MC38, and immortalized colon epithelial YAMC cells, and expression of MYB/Myb and LGR5/Lgr5 was determined by qRT-PCR relative to Gapdh. (C): Chromatin immunoprecipitation studies on the Lgr5 promoter were conducted in MC38 cells demonstrating binding of acetylated histone 3 lysine 27 (AcH3K27), a marker of promoter activation status, b-catenin and its transcription factor partner TCF4 and wild type (wt) full Myb detected with three distinct antibodies (raised in mouse [Mab1.1], rabbit [Rab], and chicken [Ch]). (D): Mouse Lgr5 promoter was inserted into a reporter construct (pCATbasic), and transactivation was assessed in the presence of increasing amounts of activated (D89) b-catenin or wt Myb. (E): Cooperation between increasing amounts of b-catenin and a constant modest level of Myb. (F): Cooperation between increasing amounts of wt Myb or activated Myb (L34P) and a constant but modest level of b-catenin. Abbreviations: CAT, chloramphenicol acetyl transferase; ChIP, chromatin immunoprecipitation; qRT-PCR, quantitative reverse-transcriptase polymerase chain reaction.

expression. To further investigate the relationship between Lgr5 and Wnt signaling, we performed Western blotting and showed that MC38 cells have relatively high Lgr5 and Myb expression but have barely detectable b-catenin compared to the robust nuclear b-catenin evident in YAMC cells (Supporting Information Fig. 3B, 3C). These data indicated that Lgr5 expression does not correspond to elevated nuclear or mutant b-catenin, microsatellite instability or APC mutation status (represented by the range of CRC cell lines; Fig. 1A) but rather when Myb is present and particularly if overexpressed. To explore the Lgr5 regulation further, chromatin immunoprecipitation (ChIP) analysis of its promoter was performed showing Myb and some b-catenin engagement in MC38 but not YAMC cells (Fig. 1C). Lgr5 promoter reporter studies were then conducted in mouse 3T3 cells. These showed that activated b-catenin was a strong Lgr5 transactivator while Myb also mediated activation but not in a sustainable fashion at higher concentrations (Fig. 1D). This parabolic response for the Myb target gene MYC has been observed previously [14, 30]. Nevertheless, Myb further augmented the relatively high transactivation achieved by b-catenin alone without imposing an inhibitory effect on b-catenin (Fig. 1E). Conversely, modest levels of b-catenin reversed the parabolic

Myb dose response generating an activation exceeding 80fold (Fig. 1E). Hyperactivated Myb (L34P) that does not show a parabolic transcriptional response [31] drove Lgr5 activation more strongly than wild-type (wt) Myb but unlike wt Myb it did not synergize with b-catenin (Fig. 1F). These data suggest a functional cooperation between Myb and b-catenin; however, we have been unable to identify a direct physical association to each other (data not shown) yet these two factors work most effectively in concert.

Myb Hypomorphic Colon Crypts Express Less Lgr5 We have reported previously that Myb hypomorphic mice have shorter crypts [7]. Here, we found Lgr5 to be underexpressed in plt4/plt4 hypomorphic crypts (Fig. 2A, 2B). As Myb in plt4/plt4 mice is transcriptionally defective when assessed on an artificial reporter [32] and given that Myb can autoregulate its own promoter, we asked whether the slightly lower levels of Myb mRNA in plt4/plt4 crypts might be revealed by Myb promoter reporter studies. These studies confirm that MybPlt4 is defective but further that Myb is not driven by autoregulation alone (Fig. 2C). Crypt ChIP analysis of the Lgr5 promoter showed that Myb and b-catenin were bound in wt mice (Fig. 2D). However, a more intriguing


2045

Figure 2. Low levels of Lgr5 expression is evident in Myb hypomorphic mutant mouse crypts. (A, B): Myb and Lgr5 expression in C57Bl/6 wt, plt4/þ heterozygous, and plt4/plt4 hypomorphic mutant mice colonic crypts. (C): Autoactivation by Myb and mutant MybPLT4 of the Myb promoter. (D): Chromatin immunoprecipitation analysis of the endogenous Lgr5 promoter in wt and mutant crypts. (E, F): Transcription factor binding to the Lgr5 and control gapdh promoters in APCmin/þ colon crypts and normalized to gapdh binding. (**, p > .01; ***, p > .001). Abbreviation: wt, wild type.

aspect of these studies was that b-catenin engagement was weak compared to Myb. Furthermore, b-catenin did not compensate for the very weak MybPlt4 binding; rather both factors bound at background levels (Fig. 2D). These data are consistent with reduced DNA binding activity of Mybplt4 in electrophoretic mobility shift assay (data not shown) and that Myb may cooperate with b-catenin in engaging promoter elements. Several hypomorphic mutant forms of Myb have been reported [9]. Accordingly, we examined Lgr5 expression in colonic crypts from another mutant mouse line, plt3/plt3. These also showed reduced Lgr5 expression (Supporting Information Fig. 4A), and we found defective Lgr5 promoter activation by all three Myb hypomorphic constructs (Supporting Information Fig. 4B-4E). To assess DNA binding at the Lgr5 promoter in the context of increased Wnt signaling, colon crypts were collected from APCmin/þ mice and subjected to Lgr5 and control gapdh promoter ChIP. Note that we also found that Lgr5 expression was elevated in APCmin/þ crypts (data not shown). Robust binding of both factors was evident at higher levels than that observed for the control in crypts (Fig. 2D-2F). These data suggest that Myb binding drives, and tracks with, endogenous Lgr5 expression in the context of increased Wnt signaling.

Stem Cell Signatures in Myb Hypomorphic Mice Are Profoundly Affected by Radiation The expression of several ISC genes (dcamkl-1 [11, 33], bmi1 [13, 34], and olfm4 [27]) in addition to Lgr5 have been described. We investigated these genes in crypts isolated from wt and plt4/plt4 mice 5 days after being subjected to 13 Gy WBI. Myb expression was elevated in ionizing radiation (IR) colon but diminished in plt4/plt4 mice (Fig. 3A). As anticipated, untreated plt4/plt4 mice showed reduced Lgr5 www.StemCells.com

compared to wt but with remarkably different IR recovery (Fig. 3B). dcamkl-1 did not follow either genotype but was reduced to residual levels following IR with little evidence of restoration (Fig. 3C). By contrast, bmi-1 showed a significant induction in IR-treated colon and to some extent in SI (Fig. 3D). Interestingly bmi-1 basal levels are lower in untreated plt4/plt4 mice suggesting that Myb also has an impact on the expression of this quiescent ISC gene but it still showed induction in plt4/ plt4 mice more consistent with an indirect Myb-Bmi1 relationship. Following IR, cyclinE1 expression increased substantially in wt but not in plt4/plt4 mice at the mRNA level (Fig. 3E) or by IHC (Supporting Information Fig. 5) reinforcing the view that cyclinE1 is a Myb target gene [9]. To further assess ISC recovery, we investigated olfm4 [27]. qRT-PCR analysis of the levels of olfm4 in SI crypt samples revealed an eightfold decrease in expression in plt4/ plt4 mice (Fig. 3F). Using ISH, the characteristic restriction of olfm4 mRNA to the SI crypt base was observed. In both genotypes, the distribution of olfm4 was sporadic and considerably lowered following IR, and oflm4-positive crypts were mostly associated with reconstituted villi in wt mice (Fig. 3G). By contrast, foci of the ISH distal to the lumen were evident in the plt4/plt4 SI and reconstituted villi were rarely apparent but indicative that ISC were being recruited to generate new crypts in the much damaged mucosa. One formal possibility to explain the reduction of gastrointestinal tract (GI) stem cell genes in Myb hypomorphic mice that are also recognized Wnt target genes is that Myb may directly regulate these Wnt target genes. To test this, qRTPCR was conducted on colon and SI crypts from wt and plt4/ plt4 mice to determine tcf4, ascl2, and cyclinD1 expression (Fig. 3H). No significant differences were observed in stark contrast to the differences observed for Lgr5 and cyclinE1 expression (Fig. 3B, 3E).

2046


Figure 3. Intestinal stem cell gene expression in wt and plt4/plt4 mutant in colon and small intestine (SI) following 13 Gy c-irradiation. (A): Myb, (B) Lgr5, (C) dcamkl, (D) bmi-1, (E) cyclinE1, and (F) olfm4 expression in colon and SI with and without IR (wt: blue, untreated and purple, IR: plt4/plt4: brown, untreated and orange, IR). (G): SI sections from wt and plt4/plt4 mice were subjected to in situ hybridization using an olfm4 antisense probe and developed to reveal dark blue staining. Characteristic signal at the crypt base in unirradiated mice is evident (blue arrow). Similar localization is evident in irradiated wt mice (red arrow, different mouse SI shown) along with reconstituted villi; however, in plt4/plt4 mice crypt foci were embedded in the stroma and villi were rarely evident (*, p > .05; **, p > .01; ***, p > .001; NS, not significantly different). (H): Wnt-target genes, tcf4, ascl2, and cyclinD1 were evaluated in colon (dark blue, wt and pale blue, plt4/plt4) and SI (brown, wt and tan, plt4/plt4) to show no significant differences in expression in Myb hypomorphic mutant mice compared to wt mice. Abbreviations: IR, ionizing radiation; wt, wild type.

Cell Intrinsic Stem/Progenitor Defect in Myb Hypomorphic SI Organoids

Lgr5 Expression Is Lost Commensurate with Myb Deletion in Organoids

Primary ISC can now be cultured in vitro to generate SI organoids [35]. Using this method, cultures from wt and plt4/plt4 mice were established and challenged with IR. Under these conditions, many of the signals provided by the crypt niche are similarly present. Importantly, the presence of R-Spondin, which potentiates Wnt signaling, ensures that Wnt target genes are activated [35]. We have found that initiating these cultures with ‘‘crypt nests’’ substantially improves cloning efficiency (Fig. 4A) compared to sorting of Lgr5þve single cells (data not shown). This is likely to be due to the supporting role of Paneth cells for ISC [36, 37]. Under these conditions, wt organoids develop with approximately double the efficiency of the mutant but this difference is not exacerbated by IR (Fig. 4A, 4B). Most importantly, when organoids were dissociated into single cells and replated to assess the self-renewal capacity, the plt4/plt4 organoid-generating cells were highly impaired with or without IR (Fig. 4C). These data suggest that normal Myb function underpins organoid forming ability and subsequent self-renewal and these are the first demonstrations of a gene affecting both of these properties in ISC as modeled in Figure 4D.

To further establish the direct relationship between Myb and Lgr5, we performed IHC on sequential SI sections using antiGFP antibody as a surrogate reporter for Lgr5 expression in Lgr5-EGFP-IRES-creERT2 reporter mice along with anti-b-catenin and anti-Myb. Accordingly, colocalization of eGFP and Myb was routinely evident in basal crypt cells often in the absence of nuclear b-catenin (Fig. 5A). Strong membrane staining of b-catenin was most evident in these sections. We then isolated colonic crypts, and generated single cell suspensions from Lgr5-EGFP-IRES-creERT2 reporter mice were FACS sorted and subjected to expression analysis whereby Myb expression was clearly higher in the Lgr5þve population (Fig. 5B) using two different house keeping genes as standardization controls. Conversely, organoids were established from Mybf/f;villinCreERT2 mice which were subjected to 4-OHT treatment to maximize recombination and Myb deletion. Under these conditions, we showed that as Myb expression was lost so was Lgr5 (Fig. 5C). These data combined with those from other experiments make a strong case that Lgr5 expression is Myb-dependent.


2047

Figure 4. Myb is required for normal small intestine (SI) organoid growth and self-renewal. (A): Crypt nests were isolated from wt and plt4/ plt4 hypomorphic mice and treated with or without ionizing radiation at 2.5 or 5 Gy 1 day later. (B): Organoids formed were counted at day 10 and expressed as a percentage of crypt nests evaluated at day 1 showing that the plt4/plt4 crypt-derived cells formed less organoids under all treatment regimens. (C): Organoids were collected, disassociated, replated, and the organoid-renewal capacity was evaluated showing that the plt4 mutation in crypts impedes self-renewal. (D): Collectively, these data suggest that the plt4 mutation affects gastrointestinal tract ISC selfrenewal and the production of daughter progenitor cells. (*, p > .05; **, p > .01). Abbreviation: ISC, intestinal stem cell.

DISCUSSION Recent advances in identifying key regulators of GI stem and progenitor cell function have provided insights into GI cancer www.StemCells.com

and disease. We propose that Myb is one such regulator. The first clue in the GI came when we identified a severe morphological recovery defect in Myb6 mice following treatment with IR [8]. Notably, these mice are indistinguishable from wt littermates under steady-state situations but Myb6 colons

2048


Figure 5. eGFP-Lgr5-reporter expression tracks with Myb in vivo. (A): Consecutive sections of small intestine (SI) were subjected to immunohistochemistry staining for eGFP, b-catenin, and Myb. eGFP serves as surrogate reporter for Lgr5 expression. Boxed region encapsulates regions where there is costaining of eGFP and Myb. Black arrows highlight costaining of eGFP and Myb in close proximity to a cell with nuclear b-catenin staining. (B): eGFPþve cells were prepared from colons, selected by FACS, and subjected to qRT-PCR analysis to show enrichment for Myb expression. Data were evaluated against two independent house keeping genes g6phx and hmbs. (C): SI organoids were generated from wt and VillinCreERT2; Myblox/lox mice and propagated for 3 days in the presence of 4-OHT and then processed for RNA to show that loss of Myb expression influences Lgr5 expression. Abbreviations: eGFP, enhanced green fluorescent protein; FACS, fluorescence-activated cell sorting; 4-OHT, 4hydrotamoxifen; qRT-PCR, quantitative reverse-transcriptase polymerase chain reaction; wt, wild type.

had not recovered by day 4 post-irradiation. By contrast, plt4/ plt4 hypomorphic mutant mice, which have further reduced Myb functionality, showed compromised proliferation and reduced colonic crypt length under steady-state conditions. This and two additional Myb hypomorphs also expressed reduced cyclinE1 mRNA [9]. As CyclinE1 is not normally

required for ongoing proliferation but is essential for normal recovery from a quiescent state [10], this indicated that its dependence on Myb might be relevant to stem or progenitor cell re-entry in to the cell cycle under emergency conditions. In the course of the studies reported here, we noted that cyclinE1 mRNA and protein was also lower in the SI of plt4/plt4 mice.


By using radiation damage to drive emergency tissue recovery, it was evident that cyclinE1 induction is a feature in both the colon and SI and that plt4/plt4 mice are completely defective in this response. Although cyclinE1 is mostly dispensable when tissue has been established, it is essential for cell cycle re-entry postquiescence [10], and our data suggest that the suboptimal recovery of the plt4/plt4 GI might be related to cell cycle re-entry of cells required for tissue repair. This is likely to be related to the apparent recruitment of cells with ISC-like properties following radiation damage described by Potten et al. [2, 38]. Markers of ISC have been proposed using knockout and lineage tracing approaches. Four markers have come the fore, Lgr5 [3], Bmi1 [13, 34], Dcamkl-1 [11, 33], and olfm4 [27]. In the context of our study, Dcamkl-1 was found in equal measure in wt and plt4/plt4 mice and was essentially lost in recovering GI tissue at the mRNA level and by IHC (data not shown) in both genotypes. Perhaps, these observations indicate that dcamkl-1þve cells are profoundly radiation sensitive and would appear to be unaffected by the presence or deficit of Myb. By contrast, bmi1 expression paralleled crypt recovery and in the context of the other ISC marker and morphological recovery data bmi1 expression may be serving as a surrogate marker of ISC recovery kinetics. Nevertheless, as this polycomb repressor is described as a ISC marker in other tissue compartments and that it is expressed at crypt position þ4 in the SI, the address of the ‘‘quiescent ISC’’ [13, 34], our finding that it tracks with GI recovery was somewhat surprising. That is unless its elevated expression marks recruitment of ‘‘emergency’’ ISCs. Similarly, it was puzzling that bmi1 expression was reduced in plt4/plt4 mice and that in colon at least its expression although induced following damage its elevation was below that observed in wt crypts. Steady-state Lgr5 expression was well below normal levels in two Myb hypomorphic mouse lines and by day 5 postirradiation the SI expression levels had not recovered even in wt littermates. The colon which has faster recovery kinetics under the irradiation protocol used here showed complete restoration of Lgr5 expression but not in plt4/plt4 mice. These data indicate the profound dependence of Lgr5 expression on the status of Myb function. Lgr5 reporter studies, ChIP analysis, and expression profiling all point to Myb directly regulating Lgr5 transcription. Importantly, this regulation works most potently in cooperation with canonical Wnt signaling through the activation of b-catenin and may be the most likely context in which Lgr5 is expressed. We were, however, able to interfere with this cooperation by using mutant forms of Myb. The plt4 mutation targets the leucine-rich regulatory region (LRR) of the transcription factor which reduces its transactivation capacity (Fig. 1D) and has a modest negative effect on DNA binding (data not shown). Mice harboring two Myb alleles with this mutation have multiple crypt defects as do two other homozygous Myb hypomorphs [9]. When we used another Myb mutant protein that impinges on the LRR, MybL3,4P, we showed as expected that it was a superior transactivator to wt Myb in reporter assays [39] including the Lgr5 reporter shown here (Fig. 1F). Noting that MybPlt4 is transcriptional defective while MybL3,4P is transcriptional hyperactivate, but as both have LLR mutations, we speculate that cooperation with activated b-catenin might indirectly involve this proposed protein-protein interaction domain. This possibility is currently under exploration but it is reasonable to conclude from reporter studies conducted here that a wt LLR is required for cooperation (Fig. 1F). www.StemCells.com

2049

Oflm4 is restricted to SI in the mouse but it does allow further evaluation of ISC marker gene recovery in this tissue. Oflm4 expression kinetics presented yet another response to the loss of Myb function on the one hand where its level was sixfold lower in plt4/plt4 mice compared to wt mice but in both genotypes by day 5 post-irradiation the SI in neither genotype had restored anywhere near to normal levels. In some ways, examination of Oflm4 ISH was most instructive as this approach in combination with histological examination highlighted the relatively impaired recovery of the SI villi and crypts in mutant mice that was not as evident by examining olfm4 mRNA levels alone. Having explored the expression of a series of ISC markers it has been tempting to speculate about the order of gene expression induction following radiation-induced GI damage. Further time course studies and other mutant mice will be helpful in addressing this question with exactitude; however, our data suggest that the following order might warrant dissecting: Bmi-1> Myb > CyclinE1 > Lgr5 > Olfm4. Dcamkl-1 in this instance has not been particularly instructive. Certainly, the sequence of induction need not be entirely linear but may occur in parallel in some instances; nevertheless, the order that finds Myb upstream of Cyclin E1 and Lgr5 seems plausible. In the case of Lgr5, our transcription studies would further suggest that evaluating the activation of the Wnt pathway in this context in collaboration with Myb may provide deeper insights. All stem cell compartments are likely to be regulated at multiple levels and by parallel and perhaps interacting pathways. Here, we provide evidence that Lgr5 expression is regulated by at least two pathways, one that requires Myb and another dependent on b-catenin. Indeed, it would appear that defective Myb function impedes b-catenin binding to the Lgr5 promoter region. We further found that both of these transcription factors work in partnership to maximally induce Lgr5 expression. Until the recent advances pioneered by the Clevers group, the ability to culture primary intestinal epithelial cells has been a frustration to the field of GI stem cell biology. Using this technology, we have been able to directly demonstrate a cell intrinsic requirement for wt Myb in the formation of SI organoids and more importantly the ability to sequentially propagate the organoids by serial passage in the case of organoids established from plt4/plt4 mice. Organoids from wt mice can be progressively passaged while the plt4/plt4 organoids were essentially exhausted after the second passage. These data speak to the role of Myb in the process of selfrenewal a property that was difficult to address prior to the advent of organoid cultures. Ex vivo-induced deletion of Myb in organoids and loss of Lgr5 expression further highlights this relationship.

CONCLUSION In conclusion, our focus on MYB began with recognition that this gene is overexpressed in CRC relative to other epithelial cancer types [7] and following this that Myb mutant mice had impeded recovery from GI damage and more profoundly defective mutant mice showed defects under homeostasis. Here, we have advanced the view that Myb is important in ISC gene expression as well as affecting in vitro stem/progenitor self-renewal. These properties have implications for why Myb expression so frequently features in CRC and particularly in metastatic disease [40].


2050

ACKNOWLEDGMENTS This work was supported by a NHMRC Program Grant #487922 and NHMRC Senior Research Fellowship #APP1002117 to R.G.R.

REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

20

Barker N, Ridgway RA, van Es JH et al. Crypt stem cells as the cellsof-origin of intestinal cancer. Nature 2009;457:608–611. Potten CS, Gandara R, Mahida YR et al. The stem cells of small intestinal crypts: Where are they? Cell Prolif 2009;42:731–750. Barker N, van Es JH, Kuipers J et al. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature 2007;449: 1003–1007. Carmon KS, Gong X, Lin Q et al. R-spondins function as ligands of the orphan receptors LGR4 and LGR5 to regulate Wnt/{beta}-catenin signaling. Proc Natl Acad Sci USA 2011;108:11452–11457. de Lau W, Barker N, Low TY et al. Lgr5 homologues associate with Wnt receptors and mediate R-spondin signalling. Nature 2011;476: 293–297. van der Flier LG, van Gijn ME, Hatzis P et al. Transcription factor achaete scute-like 2 controls intestinal stem cell fate. Cell 2009;136: 903–912. Ramsay RG, Gonda TJ. MYB function in normal and cancer cells. Nat Rev Cancer 2008;8:523–534. Ramsay RG, Micallef S, Lightowler S et al. c-myb heterozygous mice are hypersensitive to 5-fluorouracil and ionizing radiation. Mol Cancer Res 2004;2:354–361. Malaterre J, Carpinelli M, Ernst M et al. c-Myb is required for progenitor cell homeostasis in colonic crypts. Proc Natl Acad Sci USA 2007;104:3829–3834. Geng Y, Yu Q, Sicinska E et al. Cyclin E ablation in the mouse. Cell 2003;114:431–443. May R, Sureban SM, Hoang N et al. DCAMKL-1 and LGR5 mark quiescent and cycling intestinal stem cells respectively. Stem Cells 2009;27:2571–2579. Umar S. Intestinal stem cells. Curr Gastroenterol Rep 2010;12: 340–348. Sangiorgi E, Capecchi MR. Bmi1 is expressed in vivo in intestinal stem cells. Nat Genet 2008;40:915–920. Ciznadija D, Tothill R, Waterman ML et al. Intestinal adenoma formation and MYC activation are regulated by cooperation between MYB and Wnt signaling. Cell Death Differ 2009;16:1530–1538. el Marjou F, Janssen KP, Chang BH et al. Tissue-specific and inducible Cre-mediated recombination in the gut epithelium. Genesis 2004; 39:186–193. Emambokus N, Vegiopoulos A, Harman B et al. Progression through key stages of haemopoiesis is dependent on distinct threshold levels of c-Myb. Embo J 2003;22:4478–4488. Moser AR, Pitot HC, Dove WF. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science 1990;247: 322–324. Alitalo K, Winqvist R, Lin CC et al. Aberrant expression of an amplified c-myb oncogene in two cell lines from a colon carcinoma. Proc Natl Acad Sci USA 1984;81:4534–4538. McNutt NS, Mak LL, Kim YS. Comparison of cell peripheries in the human colonic adenocarcinoma cell lines SW480 and SW620 grown in floating chamber culture, cover slip culture, athymic (nude) mice, and BALB/c mice. Lab Invest 1981;44:309–323. Ramsay RG, Thompson MA, Hayman JA et al. Myb expression is higher in malignant human colonic carcinoma and premalignant adenomatous polyps than in normal mucosa. Cell Growth Differ 1992;3: 723–730.

DISCLOSURE OF POTENTIAL CONFLICTS OF INTERESTS The authors indicate no potential conflict of interest.

21 Trainer DL, Kline T, McCabe FL et al. Biological characterization and oncogene expression in human colorectal carcinoma cell lines. Int J Cancer 1988;41:287–296. 22 Brattain MG, Fine WD, Khaled FM et al. Heterogeneity of malignant cells from a human colonic carcinoma. Cancer Res 1981;41:1751–1756. 23 Corbett TH, Griswold DP, Jr., Roberts BJ et al. Tumor induction relationships in development of transplantable cancers of the colon in mice for chemotherapy assays, with a note on carcinogen structure. Cancer Res 1975;35:2434–2439. 24 Ramsay RG, Ciznadija D, Sicurella C et al. Colon epithelial cell differentiation is inhibited by constitutive c-myb expression or mutant APC plus activated RAS. DNA Cell Biol 2005;24:21–29. 25 Sato T, Vries RG, Snippert HJ et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 2009;459:262–265. 26 Ramsay RG, Ishii S, Nishina Y et al. Characterization of alternate and truncated forms of murine c-myb proteins. Oncogene Res 1989;4:259–269. 27 van der Flier LG, Haegebarth A, Stange DE et al. OLFM4 is a robust marker for stem cells in human intestine and marks a subset of colorectal cancer cells. Gastroenterology 2009;137:15–17. 28 Uchida H, Yamazaki K, Fukuma M et al. Overexpression of leucinerich repeat-containing G protein-coupled receptor 5 in colorectal cancer. Cancer Sci 2010;101:1731–1737. 29 Ramsay RG, Barton AL, Gonda TJ. Targeting c-Myb expression in human disease. Expert Opin Ther Targets 2003;7:235–248. 30 Nakagoshi H, Kanei-Ishii C, Sawazaki T et al. Transcriptional activation of the c-myc gene by the c-myb and B-myb gene products. Oncogene 1992;7:1233–1240. 31 Kanei-Ishii C, MacMillan EM, Nomura T et al. Transactivation and transformation by Myb are negatively regulated by a leucine-zipper structure. Proc Natl Acad Sci USA 1992;89:3088–3092. 32 Carpinelli MR, Hilton DJ, Metcalf D et al. Suppressor screen in Mpl/ mice: c-Myb mutation causes supraphysiological production of platelets in the absence of thrombopoietin signaling. Proc Natl Acad Sci USA 2004;101:6553–6558. 33 May R, Riehl TE, Hunt C et al. Identification of a novel putative gastrointestinal stem cell and adenoma stem cell marker, doublecortin and CaM kinase-like-1, following radiation injury and in adenomatous polyposis coli/multiple intestinal neoplasia mice. Stem Cells 2008;26: 630–637. 34 Ootani A, Li X, Sangiorgi E et al. Sustained in vitro intestinal epithelial culture within a Wnt-dependent stem cell niche. Nat Med 2009; 15:701–706. 35 Sato T, Vries RG, Snippert HJ et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 2009;459:262–265. 36 Huynh D, Dai XM, Nandi S et al. Colony stimulating factor-1 dependence of paneth cell development in the mouse small intestine. Gastroenterology 2009;137:136–144. 37 Sato T, van Es JH, Snippert HJ et al. Paneth cells constitute the niche for Lgr5 stem cells in intestinal crypts. Nature 2011;469:415–418. 38 Potten CS, Loeffler M. A comprehensive model of the crypts of the small intestine of the mouse provides insight into the mechanisms of cell migration and the proliferation hierarchy. J Theor Biol 1987;127:381–391. 39 Nomura T, Sakai N, Sarai A et al. Negative autoregulation of c-Myb activity by homodimer formation through the leucine zipper. J Biol Chem 1993;268:21914–21923. 40 Biroccio A, Benassi B, D’Agnano I et al. c-Myb and Bcl-x overexpression predicts poor prognosis in colorectal cancer: Clinical and experimental findings. Am J Pathol 2001;158:1289–1299.

See www.StemCells.com for supporting information available online.