of 10

Generation of an inducible colon-specific Cre enzyme mouse line for colon cancer research

Generation of an inducible colon-specific Cre enzyme mouse line for colon cancer research Paul W. Tetteh a,b,1, Kai Kretzschmar a,b, Harry Begthel a,b, Maaike van den Born a,b, Jeroen Korving a,b, Folkert
0 views10 pages
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Generation of an inducible colon-specific Cre enzyme mouse line for colon cancer research Paul W. Tetteh a,b,1, Kai Kretzschmar a,b, Harry Begthel a,b, Maaike van den Born a,b, Jeroen Korving a,b, Folkert Morsink c, Henner Farin a,b,2, Johan H. van Es a,b, G. Johan A. Offerhaus c, and Hans Clevers a,b,3 a Hubrecht Institute for Developmental Biology and Stem Cell Research, Royal Netherlands Academy of Arts and Sciences (KNAW), 3584 CT Utrecht, The Netherlands; b University Medical Centre Utrecht, 3584 CT Utrecht, The Netherlands; and c Department of Pathology, University Medical Centre Utrecht, 3584 CX Utrecht, The Netherlands Contributed by Hans Clevers, September 12, 2016 (sent for review January 29, 2016; reviewed by Inke Nathke and Owen J. Sansom) Current mouse models for colorectal cancer often differ significantly from human colon cancer, being largely restricted to the small intestine. Here, we aim to develop a colon-specific inducible mouse model that can faithfully recapitulate human colon cancer initiation and progression. Carbonic anhydrase I (Car1) is a gene expressed uniquely in colonic epithelial cells. We generated a colon-specific inducible Car1 CreER knock-in (KI) mouse with broad Cre activity in epithelial cells of the proximal colon and cecum. Deletion of the tumor suppressor gene Apc using the Car1 CreER KI caused tumor formation in the cecum but did not yield adenomas in the proximal colon. Mutation of both Apc and Kras yielded microadenomas in both the cecum and the proximal colon, which progressed to macroadenomas with significant morbidity. Aggressive carcinomas with some invasion into lymph nodes developed upon combined induction of oncogenic mutations of Apc, Kras, p53, and Smad4. Importantly, no adenomas were observed in the small intestine. Additionally, we observed tumors from differentiated Car1-expressing cells with Apc/Kras mutations, suggesting that a top-down model of intestinal tumorigenesis can occur with multiple mutations. Our results establish the Car1 CreER KI as a valuable mouse model to study colon-specific tumorigenesis and metastasis as well as cancer-cell-of-origin questions. mouse model gastrointestinal tract colorectal cancer Car1 differentiated epithelial cells The large intestine consists of the cecum, the colon, and rectum. Its simple columnar epithelium is organized into the crypts of Lieberkühn. Unlike the small intestine, the large intestine is bereft of villi. The colon epithelium is a constantly selfrenewing tissue. The crypts of Lieberkühn contain actively cycling leucine-rich repeat-containing G-protein coupled receptor 5(Lgr5)-expressing stem cells at their base (1). These stem cells produce proliferative transit-amplifying cells that give rise to mature cells such as mucous-secreting goblet cells and shortlived absorptive enterocytes with apical microvilli (distinct from small intestine enterocytes) located at the luminal surface (1). Scattered in the large intestinal crypts are chromogranin A-positive enteroendocrine cells and tuft cells (2). M cells are also located in the Peyer s patches of the large intestine (3, 4). Human colorectal cancer (CRC) is the second leading cause of cancer-related death in the Western world (5). Generally, disease progression from benign adenomas induced by Wnt pathway gene mutations [such as in APC (adenomatous polyposis coli)] in colonic epithelial cells to invasive carcinomas involves sequential accumulation of, for example, activating mutations in RAS oncogenes, inactivating activating mutations in the SMAD family of tumor suppressor genes, and inactivating TP53 gene mutations (6). Rodent models with mutations that generate tumors in the colon, recapitulating many features of human CRC such as Apc Pirc/+ (Pirc) (7), Apc-mutant Kyoto Alpha Delta (KAD) rats (8), and Gstp-null Apc Min mice (9), have been informative in modeling cancers induced environmentally and sporadically. However, tumors are still observed in the small intestine of these models, albeit at a lower frequency, which is not seen in human CRC, limiting their ability to faithfully mimic colonspecific tumorigenesis. Controlled in vivo studies in genetic mouse models based on the Cre loxp system offer an important avenue to model the molecular etiology of CRC development via timed mutation of oncogenes and tumor suppressor genes, allowing testing of potential preventive and therapeutic interventions (10, 11). A preferred mouse model for CRC should involve Cre expression specifically in colonic epithelial cells to closely mimic the disease development observed in humans. The Carbonic anhydrase I (Car1) gene is expressed in normal colon epithelial cells as well as in colorectal tumors (12, 13). It encodes a metalloenzyme involved in hydration of carbon dioxide, ph balance, and anion exchange. Previously, a promoter/enhancer from the murine Car1 genewasusedtogenerateacre-expressing transgenic mouse (14). Although Cre expression was indeed restricted to the colon and absent from the small intestine, the usefulness of this model is limited because Cre is constitutively expressed from embryonic development onward. To overcome this limitation, we generated a mutant allele containing Cre recombinase fused to the mutated tamoxifen-inducible estrogen receptor (CreERT2) inserted into the 3 untranslated region of the Car1 gene, allowing for spatiotemporal control of Cre activity (15). Significance A major limitation of current mouse models of colorectal cancer (CRC) is that the cancer that develops is often significantly different from human colon cancer in terms of latency, intestinal location, or molecular signature. Carbonic anhydrase I (Car1) is a gene expressed by colonic epithelial cells. We generated an inducible Car1 CreER mouse model with Cre expression in the cecum and proximal colon. Mutations of genes that drive human CRC with our Car1 CreER mouse yielded tumors exclusively in the cecum and proximal colon. Differentiated colonic cells in the proximal colon with Apc/Kras mutations initiate tumors supporting a top-down model of intestinal tumorigenesis. The inducible Car1 CreER will be a useful model in studying human intestinal cancers originating from the proximal colon. Author contributions: P.W.T., H.F., and H.C. designed research; P.W.T., K.K., H.B., M.v.d.B., J.K., F.M., and J.H.v.E. performed research; P.W.T., K.K., H.F., and G.J.A.O. analyzed data; and P.W.T., K.K., H.F., and H.C. wrote the paper. Reviewers: I.N., University of Dundee; and O.J.S., Beatson Institute for Cancer Research. The authors declare no conflict of interest. 1 Present address: Department of Medical Oncology, Laboratory of Immunobiology, Dana- Farber Cancer Institute, Boston, MA Present address: Georg-Speyer-Haus Institute for Tumor Biology and Experimental Therapy, Frankfurt am Main, Germany. 3 To whom correspondence should be addressed. This article contains supporting information online at /pnas /-/DCSupplemental. CELL BIOLOGY PNAS Early Edition 1of6 by Southern blot analysis using a probe upstream of the targeted region (Fig. 1D). Accurate recombination efficiency in ES clones was 6%. Two independent positive clones were injected into C57BL/6 blastocysts according to standard procedures.the neomycin selection cassette flanked by FRT sites was excised in vivo by crossing the Car1 CreER knock-in (KI) mice with FLP1 mice. Heterozygous and homozygous Car1 CreER mice (Fig. 1E) were retrieved at the expected Mendelian ratios at birth. No discernible abnormality was observed in adult homozygous transgenic animals, which had comparable life span and fertility as wild-type littermates. To indirectly visualize expression of the Cre protein, we performed immunohistochemical staining on paraffin-embedded intestinal tissue slides with an antibody recognizing the mutant estrogen receptor domain (ERT2) fused to the Cre protein. This showed that Cre expression faithfully mirrored endogenous Car1 expression (Fig. 1 F and G). Characterization of Cre Activity of Car1 CreER KI. To further characterize Cre activity of the Car1 CreER allele and trace the behavior of Car1-expressing cells, mice were crossed to R26R LSL LacZ reporter mice where the LacZ gene is under the control of the ubiquitous Rosa26 locus (18). Eight- to 12-wk-old mice were injected with a single dose of 4OH-tamoxifen (TAM), killed at various time points (three mice per time point), and analyzed for β-galactosidase (β-gal) staining. In the proximal colon, β-gal positive cells were detected in the upper crypt (Fig. 2 A D and G). No β-gal positive cells were observed after 28 d, suggesting that Car1 labeling did not occur in colonic stem cells (Fig. 2E). Analysis of Cre activity in the cecum of Car1 CreER/+ ;R26R LSL LacZ/+ mice showed β-gal positive cells and estrogen receptor (ER) antibody-stained cells not only in differentiated cells but also at the crypt bottom, suggesting broad expression of Car1 in both stem cells and differentiated cells of the cecum (Fig. S1). β-gal positive cells were also detected in long-lived hepatocytes in the liver, which have been reported to express Car1 (Fig. 2F) (14). Fig. 1. Car1 expression in colon and generation of Car1 CreER knock-in. Comparison of Car1 mrna expression (A) to stem cell marker gene Lgr5 (B) in the proximal colon. Car1 is localized to differentiated cells at the top of the colonic crypt whereas Lgr5 is found in stem cells at the bottom of the crypt. (C) Targeting strategy. IRES CreERT2 targeting construct used to target Car1 locus. (D) Southern blotting showing successfully targeted embryonic stem cells. (E) Genotyping of Car1 CreER KI mice. (F and G) ER antibody staining showing expression of CreER fusion from the Car1 promoter in the Car1 CreER KI. CreER expression is higher in the proximal colon (F) than in the distal colon (G). (Scale bars: A and B, 100 μm; F and G, 50μm.) Results Generation of Car1 CreER Mice. Multiple studies have analyzed the organ-specific expression of Car1 (12, 13, 16, 17). In the gastrointestinal tract, Car1 is expressed in the colon but not in the small intestine, with highest expression in the proximal colon (16). We confirmed the expression of Car1 by mrna in situ hybridization with a probe specific for the Car1 gene. Comparison of Car1 mrna expression with that of the colonic stem-cell marker gene Lgr5 by in situ hybridization confirmed localization of Car1 transcripts in differentiated colonic epithelial cells and Lgr5 mrna being detectable in the stem cells at the crypt bottom (Fig. 1 A and B). To generate an inducible colon-specific Cre line, we inserted an internal ribosome entry site (IRES) CreERT2 cassette at the stop codon located in the last exon of the Car1 gene (Car1 CreER ; Fig. 1 C E). This strategy employs a poly(a) signal and the 3 UTR of the Car1 gene (Fig. 1C). Targeting arms and the IRES CreERT2 cassette were subcloned, and the construct was linearized and electroporated into embryonic stem (ES) cells. Recombined ES clones expressing the neomycin gene were selected in G418 supplemented medium. Homologous recombination was confirmed Tumor Initiation from Car1-Expressing Cells. The adenoma-carcinoma model of colon cancer initiation posits that disease progression to the metastatic/invasive stage depends on sequential or combined mutations of Apc, Kras, p53, and Smad4 genes (3, 19 21). To determine the utility of the Car1 CreER KI for colon cancer research, we crossed mice carrying mutant alleles for Apc fl/fl, Kras LSL G12D/+, p53ko, and Smad4 fl/fl with Car1 CreER KI mice and induced Cre recombination by TAM injection. The presence of tumors was confirmed by nuclear accumulation of β-catenin (Fig. 3 A and B), a hallmark of hyperactive Wnt signaling frequently found in human CRC (20). As expected, based on previous studies (20) in small intestine tumorigenesis, Apc mutation in differentiated Car1- expressing colonic cells yielded microadenomas that failed to progress to macroadenomas following TAM induction (Fig. S2). In the cecum, however, Apc deletion yielded much larger adenomas presumably because crypt bottom stem cells in the cecum express Cre (Fig. 3 A D). We next asked whether additional mutation in the Kras gene, another commonly mutated gene in CRC, can enhance the tumorigenic potential of differentiated Car1-expressing cells in the proximal colon. We observed pronounced microadenomas in Car1 CreER/+ ;Apc fl/fl ;Kras LSL G12D/+ colons 2 wk after TAM induction (Fig. 4 A and B). β-catenin antibody staining of mice 56 d following TAM injection showed that these microadenomas had progressed into large adenomas originating from differentiated Car1-expressing cells in the proximal colon and growing toward the luminal side of the intestine (Fig. 4 C E). Of note, no adenomas were ever observed in the small intestine, the distal colon, or the liver (Fig. S3). Histopathological analysis of hematoxylin and eosin (H&E)- stained sections of tumors from Car1-expressing cells revealed that these resembled conventional human adenomas (Fig. 5A and Fig. S4). Proximal colon tumors from Car1 CreER/+ ;Apc fl/fl ; 2of6 Tetteh et al. CELL BIOLOGY Fig. 2. Lineage tracing of Car1-expressing cells. Car1 CreER/+ ;R26R LSL LacZ/+ mice were induced with 5 mg/kg TAM and then killed after (A)1d,(B)2d,(C)3d,(D)4d,and 28 d (E and F). (G) Quantification of β-gal+ cells 24 h in proximal colon after TAM injection of Car1 CreER/+ ;R26R LSL LacZ/+ mice. β-gal+ cellscanbedetectedatcellposition6 from the crypt base and above, but not at the crypt bottom (n = 3; mean ± SD). Counts were made per 100 colonic crypts for each mouse. (Scale bars: 100 μm.) Kras LSL G12D/+ mice had recognizable tubular structures and severely disturbed architecture. Glands within the tumors were lined with a stratified epithelium that showed an increased nuclear/cytoplasmic ratio, hyperchromatism, prominent nucleoli, and numerous abnormal mitotic figures. Tumors derived from Car1-expressing cells were also highly proliferative (Fig. 5B) and poorly differentiated (Fig. 5C). Analysis of stem-cell marker expression revealed high abundance of Lgr5 mrna by in situ hybridization and Musashi-1 (MSI1) and Eph receptor B2 (EPHB2) protein by antibody staining in tumors, interestingly originating from differentiated Car1-expressing cells (Fig. S5). Tumor Progression upon Three or Four Oncogenic Mutations in Car1- Expressing Cells. We next generated Car1 CreER/+ ;Apc fl/fl ;Kras LSL G12D/+ ; p53ko triple-mutant mice, Car1 CreER/+ ;Kras LSL G12D/+ ; p53ko;smad4 fl/fl triple-mutant mice, and Car1 CreER/+ ;Apc fl/fl ; Kras LSL G12D/+ ;p53ko;smad4 fl/fl quadruple-mutant mice and injected TAM to mimic the effects of complex genotypes often associated with human CRC (21). Analysis after a single injection of TAM was limited to 1 mo as triple and quadruple mutations beyond this time point were lethal, mainly due to the extracolonic effects of the constitutively deleted p53 gene. Tumors in quadruple mutants (Fig. 6 A and D) showed significant Paneth cell metaplasia (Fig. 6 B, E, andf) and were more aggressive with some invasion into lymph nodes, as assessed by lymphatic vessel endothelial hyaluronan receptor 1 (LYVE1) immunostaining (Fig. 6C) (22). Due to high mortality only one animal carrying all four CRC mutations could be analyzed. Triple mutations of Apc, Kras, and p53 led to aggressive tumors that developed into carcinomas in the cecum and proximal colon (Fig. S6 A C). Combined mutations of Kras, p53, and Smad4 gave rise to benign adenomas (Fig. S6 D F). Genetic mutations can foster the development of inflammation or recruitment of immune cells that promote tumorigenesis (23). In the gut, chronic inflammation has been shown to be a risk factor for the development and progression of colitis-associated CRC as well as sporadic CRC (24). We therefore performed immunohistochemical stainings for two prominent components of the inflammatory infiltrate present in these particular CRC types [T-cell marker CD3 (25) and macrophage marker F4/80 (26)] to exclude chronic inflammation as a cancer hallmark in mice carrying Car1- specific mutations of Apc, Kras, andp53 with or without additional mutation of Smad4. Indeed, tumors arising from these mutations in Car1-expressing cells did not show significant infiltration with T cells or macrophages, suggesting absence of immunomodulation in this context (Fig. S7). Tetteh et al. PNAS Early Edition 3of6 small intestine and entire colon. Inactivation of APC with this transgene induced adenomas in the ileum, cecum, and proximal and distal colon (28, 29). Cre protein delivered by an adenoviral vector has been used to conditionally delete a floxed Apc allele inducing tumors in the distal colon (30). In contrast, our inducible Car1 CreER model generates tumors exclusively in the proximal colon. Identification of the tumor cell-of-origin is important for targeted interventions to combat cancer development and progression. We have previously reported that deletion of Apc in small intestinal crypt bottom Lgr5-positive stem cells induces macroadenoma formation (31). Loss of Apc in nonstem cells led to nuclear accumulation of β-catenin and formation of microadenomas that failed to progress into large adenomas, supporting a bottomup model of intestinal cancer initiation. Alternatively, evidence Fig. 3. Adenomas in cecum upon Apc deletion. Car1 CreER/+ ;Apc fl/fl mouse was injected with a single dose of TAM, killed after 2 mo, and analyzed. (A) H&E staining showing tubular adenomas in cecum. (B) Nuclear β-catenin staining showing Wnt hyperactivation in cecal adenomas. (C) KI67 staining showing highly proliferative cecal adenomas. (D) ER antibody staining showing Cre expression restricted in cecal adenomas, suggesting restricted expression in cancer cells fueling cancer progression (Inset). (Scale bars: 50 μm.) No primary tumors were observed in the liver in triple and quadruple mutants, suggesting that Car1-expressing hepatocytes are resistant to neoplastic transformation by combined mutations of Apc, Kras, p53, and Smad4 (Fig. S6 G I). Discussion In this study, we report the generation of an inducible proximal colon- and cecum-specific CreERT2 KI mouse line driven by the Car1 gene allowing for spatiotemporal control of Cre activity. Intestinal Cre expression occurs broadly in epithelial cells in the cecum and is more restricted to differentiating colonocytes in the proximal colon. No Cre activity was detected in the small intestine. Recently, Xue and colleagues reported the generation of a Car1-Cre (CAC) transgenic line that drives Cre expression in the colon (14). The constitutive Cre activity in this model persists from embryonic stages into adulthood, making it unsuitable to study adult-onset biological processes such as CRC. Whereas Cre expression in the Car1 CreER KI mice was restricted to differentiated colonic cells of the proximal colon, Cre activity in the highly mosaic CAC transgene was observed in all epithelial cell types in the proximal, distal, and rectal sections of the colon (14). A reason for the apparent discrepancy of Cre expression in the two models may be due to the different targeting strategies used. In our Car1 CreER KI, the IRES CreERT2 cassette is inserted at the 3 UTR of the Car1 gene without disrupting expression of the Car1 gene; thus Cre expression follows endogenous expression of the Car1 gene. In contrast, the targeting construct in the CAC transgene is inserted immediately downstream of the Car1 promoter, which leads to random integration of the DNA-targeting construct; thus, each founder can show different Cre expression due to the chromosomal context of the integration (27). Our model is also different from the previously generated colon-specific CDX2P CreERT2 transgene where Cre is expressed in both stem and differentiated epithelial cells of the ileum of the Fig. 4. Adenoma formation upon Apc and Kras mutations in Car1-expressing colonic epi
View more
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!