Tubulin glycylases are required for primary cilia, control of cell proliferation and tumor development in colon
2014; Springer Nature; Volume: 33; Issue: 19 Linguagem: Inglês
10.15252/embj.201488466
ISSN1460-2075
AutoresCecilia Rocha, Laura Papon, Wulfran Cacheux, Patricia Marques Sousa, Valeria Lascano, Olivia Tort, Tiziana Giordano, Sophie Vacher, Bénédicte Lemmers, Pascale Mariani, Didier Meseure, Jan Paul Medema, Ivan Bièche, Michael Hahne, Carsten Janke,
Tópico(s)Microtubule and mitosis dynamics
ResumoArticle1 September 2014free access Tubulin glycylases are required for primary cilia, control of cell proliferation and tumor development in colon Cecilia Rocha Cecilia Rocha Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Search for more papers by this author Laura Papon Laura Papon IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Search for more papers by this author Wulfran Cacheux Wulfran Cacheux Institut Curie Hospital, Paris, France Search for more papers by this author Patricia Marques Sousa Patricia Marques Sousa Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Search for more papers by this author Valeria Lascano Valeria Lascano Academic Medical Center, Amsterdam, The Netherlands Search for more papers by this author Olivia Tort Olivia Tort Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Institut de Biotecnologia i de Biomedicina, Department of Biochemistry and Molecular Biology, Universitat Autònoma de Barcelona, Bellaterra (Barcelona), Spain Search for more papers by this author Tiziana Giordano Tiziana Giordano Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Search for more papers by this author Sophie Vacher Sophie Vacher Institut Curie Hospital, Paris, France Search for more papers by this author Benedicte Lemmers Benedicte Lemmers IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Search for more papers by this author Pascale Mariani Pascale Mariani Institut Curie Hospital, Paris, France Search for more papers by this author Didier Meseure Didier Meseure Institut Curie Hospital, Paris, France Search for more papers by this author Jan Paul Medema Jan Paul Medema Academic Medical Center, Amsterdam, The Netherlands Search for more papers by this author Ivan Bièche Ivan Bièche Institut Curie Hospital, Paris, France Search for more papers by this author Michael Hahne Corresponding Author Michael Hahne IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Academic Medical Center, Amsterdam, The Netherlands Search for more papers by this author Carsten Janke Corresponding Author Carsten Janke Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Search for more papers by this author Cecilia Rocha Cecilia Rocha Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Search for more papers by this author Laura Papon Laura Papon IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Search for more papers by this author Wulfran Cacheux Wulfran Cacheux Institut Curie Hospital, Paris, France Search for more papers by this author Patricia Marques Sousa Patricia Marques Sousa Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Search for more papers by this author Valeria Lascano Valeria Lascano Academic Medical Center, Amsterdam, The Netherlands Search for more papers by this author Olivia Tort Olivia Tort Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Institut de Biotecnologia i de Biomedicina, Department of Biochemistry and Molecular Biology, Universitat Autònoma de Barcelona, Bellaterra (Barcelona), Spain Search for more papers by this author Tiziana Giordano Tiziana Giordano Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Search for more papers by this author Sophie Vacher Sophie Vacher Institut Curie Hospital, Paris, France Search for more papers by this author Benedicte Lemmers Benedicte Lemmers IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Search for more papers by this author Pascale Mariani Pascale Mariani Institut Curie Hospital, Paris, France Search for more papers by this author Didier Meseure Didier Meseure Institut Curie Hospital, Paris, France Search for more papers by this author Jan Paul Medema Jan Paul Medema Academic Medical Center, Amsterdam, The Netherlands Search for more papers by this author Ivan Bièche Ivan Bièche Institut Curie Hospital, Paris, France Search for more papers by this author Michael Hahne Corresponding Author Michael Hahne IGMM, CNRS UMR5535, Montpellier, France Université Montpellier Sud de France, Montpellier, France Academic Medical Center, Amsterdam, The Netherlands Search for more papers by this author Carsten Janke Corresponding Author Carsten Janke Institut Curie, Orsay, France PSL Research University, Paris, France CNRS UMR3306, Orsay, France INSERM U1005, Orsay, France Search for more papers by this author Author Information Cecilia Rocha1,2,3,4,5,6, Laura Papon5,6, Wulfran Cacheux7, Patricia Marques Sousa1,2,3,4, Valeria Lascano8, Olivia Tort1,2,3,4,9, Tiziana Giordano1,2,3,4, Sophie Vacher7, Benedicte Lemmers5,6, Pascale Mariani7, Didier Meseure7, Jan Paul Medema8, Ivan Bièche7, Michael Hahne 5,6,8,‡ and Carsten Janke 1,2,3,4,‡ 1Institut Curie, Orsay, France 2PSL Research University, Paris, France 3CNRS UMR3306, Orsay, France 4INSERM U1005, Orsay, France 5IGMM, CNRS UMR5535, Montpellier, France 6Université Montpellier Sud de France, Montpellier, France 7Institut Curie Hospital, Paris, France 8Academic Medical Center, Amsterdam, The Netherlands 9Institut de Biotecnologia i de Biomedicina, Department of Biochemistry and Molecular Biology, Universitat Autònoma de Barcelona, Bellaterra (Barcelona), Spain ‡These authors contributed equally to the work *Corresponding author. Tel: +33 1 69863127; Fax: +33 1 69863017; E-mail: [email protected] *Corresponding author. Tel: +33 4 67613639; Fax: +33 4 34359634; E-mail: [email protected] The EMBO Journal (2014)33:2247-2260https://doi.org/10.15252/embj.201488466 Correction(s) for this article Tubulin glycylases are required for primary cilia, control of cell proliferation and tumor development in colon18 November 2014 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract TTLL3 and TTLL8 are tubulin glycine ligases catalyzing posttranslational glycylation of microtubules. We show here for the first time that these enzymes are required for robust formation of primary cilia. We further discover the existence of primary cilia in colon and demonstrate that TTLL3 is the only glycylase in this organ. As a consequence, colon epithelium shows a reduced number of primary cilia accompanied by an increased rate of cell division in TTLL3-knockout mice. Strikingly, higher proliferation is compensated by faster tissue turnover in normal colon. In a mouse model for tumorigenesis, lack of TTLL3 strongly promotes tumor development. We further demonstrate that decreased levels of TTLL3 expression are linked to the development of human colorectal carcinomas. Thus, we have uncovered a novel role for tubulin glycylation in primary cilia maintenance, which controls cell proliferation of colon epithelial cells and plays an essential role in colon cancer development. Synopsis This study reports first evidence for tubulin glycylation to affect the establishment and maintenance of primary cilia. Functional abrogation of tubulin glycine ligases impacts cell proliferation and tumor onset. Glycylation and glycylases are required for robust primary cilia formation in mammals. Colon epithelial cells have primary cilia that depend on the glycylase TTLL3. Absence of TTLL3 leads to increased cell proliferation in colon. Lack of TTLL3 promotes colon carcinogenesis. Introduction Tubulin posttranslational modifications play key roles in the regulation of the microtubule (MT) cytoskeleton. Most of the known modifications occur on the C-terminal tail domains of tubulin molecules and are therefore exposed to the outer surface of MTs where essential interactions with diverse MT-associated proteins take place (reviewed in: Janke & Bulinski, 2011). Particularly complex modification patterns are generated by posttranslational addition of glutamate or glycine side chains to the C-terminal tubulin tails (Eddé et al, 1990; Redeker et al, 1994). In mammals, glutamylation has been found in a variety of cells and tissues; however, particularly high levels have been found on neuronal MTs (Audebert et al, 1994), centrioles (Bobinnec et al, 1998) and on the axonemes of cilia and flagella (Gagnon et al, 1996; Million et al, 1999). Glycylation, in contrast, has so far been found exclusively in axonemes of motile cilia and flagella in a wide variety of organisms (Bré et al, 1996). Only in some organisms, such as the unicellular ciliate Paramecium, cytoplasmic MTs are also glycylated (Bré et al, 1998). Both MT glutamylation and glycylation have been shown to play important roles in the function of motile cilia and flagella in several organisms. While glycylation has been linked to the stability and maintenance of these organelles, glutamylation has been found to play a role in the coordination of ciliary beating (Xia et al, 2000; Pathak et al, 2007, 2011; Rogowski et al, 2009; Wloga et al, 2009; Kubo et al, 2010; Suryavanshi et al, 2010; Bosch Grau et al, 2013). In contrast, less is known about the role of these two modifications in primary cilia. So far, only a link between aberrant glutamylation and defects of primary cilia has been reported (Lee et al, 2012, 2013). In mammals, tubulin glutamylation can be catalyzed by nine glutamate ligases, also referred to as glutamylases or polyglutamylases (van Dijk et al, 2007), while only two enzymes, TTLL3 and TTLL8, are able to initiate glycylation on MTs (Rogowski et al, 2009). These two glycylases appear to be partially complementary in the generation of MT glycylation, as only absence of both enzymes led to the loss of motile cilia from ependymal cells in the brain ventricles (Bosch Grau et al, 2013). This finding further suggested that glycylation is an essential stabilizer of motile cilia in mammals, but let open the question of the role of this modification in primary cilia. In the present work, we address the impact of MT glycylation on the integrity of primary cilia and study the role of glycylases in tissue homeostasis and colorectal cancer development. We found that TTLL3 is the only glycylase expressed in colon. Absence of TTLL3, synonymous for absence of glycylation in colon, leads to reduction of the number of primary cilia in colon accompanied by a strong increase in cell proliferation in the colon epithelium. In mouse embryonic fibroblasts, which express two glycylases (TTLL3, TTLL8), primary cilia are lost only upon depletion of both genes. This demonstrates that glycylation is important for primary cilia maintenance, which controls cell cycle and tissue homeostasis at least in colon epithelium. Finally, we established the relevance of TTLL3 for colon cancer development by demonstrating the impact of this gene in a mouse model for colitis-associated colon cancer, which is coherent with a significant downregulation of TTLL3 expression in patients with colorectal cancer. Results Glycylating enzymes are important for maintenance of primary cilia Glycylation has so far only been observed in motile cilia; however, nothing is known about the presence and the role of this modification in primary cilia. To investigate the role of glycylating enzymes for primary cilia, we used mouse embryonic fibroblasts (MEFs) that express both glycylating enzymes, TTLL3 and TTLL8 (Fig 1A). MEFs were grown in vitro and serum-deprived to assemble primary cilia. Cilia and their basal bodies were visualized with antibodies for acetylated α-tubulin and γ-tubulin, respectively (Fig 1B). Quantification of cilia numbers revealed that most of the cultured MEFs grow primary cilia in control and ttll3−/− MEFs (Fig 1C). This indicates that loss of the glycylase TTLL3 does not affect primary cilia in MEFs, which is similar to our observations in motile cilia of ependymal cells (Bosch Grau et al, 2013). Figure 1. Glycylating enzymes and primary cilia formation in MEFs The expression of TTLL3 and TTLL8 was analyzed by RT-PCR in samples from control and ttll3−/− MEFs. Wild-type and ttll3−/− MEFs were serum-starved for 2 days and labeled with anti-acetylated tubulin (red) and anti-γ-tubulin (green) antibodies, and nuclei were stained with DAPI (blue). Percentage of ciliated cells (analyzed from B). Mean values ± SEM are represented (control n = 4 mice; ttll3−/− n = 4 mice; 50 cells counted per mouse). MEFs as in (B) were stained with TAP952 (red) for monoglycylation. Arrowheads indicate TAP952-positive primary cilia. MEFs were transfected with vectors coding for scramble and TTLL8 shRNA and CFP and starved for 24 h. Primary cilia were visualized with anti-acetylated tubulin (red) and anti-γ-tubulin (green) antibodies. Transfected cells were identified by CFP (blue). Blue lines indicate transfected, and white lines non-transfected cells. Cilia are indicated by white arrowheads, and absence of cilia (identified by solitary basal bodies) by orange arrowheads. Percentage of transfected, ciliated MEFs after scramble shRNA (control, n = 210 cells; ttll3−/− n = 170 cells), TTLL8 shRNA_585 (control n = 130 cells; ttll3−/− n = 130 cells) and TTLL8 shRNA_729 (control n = 175 cells; ttll3−/− n = 210 cells). Data represent mean values of four independent experiments ± SEM; *P < 10−2 by two-tailed unpaired t-tests and Mann–Whitney post-test. Data information: Scale bars in (B), (D) and (E) are 10 μm. Download figure Download PowerPoint Glycylation has so far not been reported in primary cilia. Using the monoglycylation-specific antibody TAP952 (Bré et al, 1996), we detected glycylation in primary cilia of both control and ttll3−/− MEFs, confirming that glycylation in these cilia can be present, even in absence of TTLL3, indicating that the modification is carried out by TTLL8 alone (Fig 1D). The relatively weak labeling of primary cilia grown on MEFs could be due either to relatively low glycylation levels in these very short primary cilia, or to the presence of alternative modification sites on tubulin that are not detected by TAP952. Thus, TAP952 might not be an optimal marker for glycylation of primary cilia, but is sufficient to prove the presence of this modification in these organelles. We next studied the impact of the loss of both TTLL3 and TTLL8 on primary cilia in MEFs. For this, control and ttll3−/− MEFs were transfected with vectors expressing shRNA for TTLL8 and CFP. Ciliogenesis was subsequently induced by serum starvation. Transfected cells were identified by CFP fluorescence, and primary cilia were labeled with anti-acetylated tubulin and anti-γ-tubulin antibodies (Fig 1E). The percentage of CFP-positive cells that carry a primary cilium was determined (Fig 1F). Expression of scrambled shRNA had no effect on ciliogenesis in control and ttll3−/− MEFs, while depletion of TTLL8 with two different shRNA constructs reduced the number of ciliated cells by about 50% specifically in the ttll3−/− MEFs. This shows that the absence of both glycylating enzymes, TTLL3 and TTLL8, either induces ciliary loss or disturbs the assembly of primary cilia (Fig 1F). Thus, primary cilia are, similar to their motile counterparts, dependent on glycylation. This was unexpected, as glycylation had never been detected in primary cilia before. TTLL3 is the only glycylase expressed in the murine colon To investigate whether both glycylases are present throughout the mammalian organism, we analyzed the expression levels of TTLL3 and TTLL8 in a set of normal mouse tissues using reverse-transcriptase PCR (qRT-PCR). While the relative expression levels of the two glycylases varied between tissues, both enzymes were detected in most of the tissues analyzed, with the exception of colon, where only TTLL3 was found (Fig 2A). Figure 2. TTLL3 is the only glycylase expressed in colon Expression levels of TTLL3 and TTLL8 analyzed in tissues of 4-month-old wild-type mice. Five independent mRNA samples were analyzed by qRT-PCR, and mean values standardized to expression of Tbp are shown. Error bars represent SEM. Red box: note that no TTLL8 expression is detected in colon tissue. TTLL3 and TTLL8 expression analysis by RT-PCR in 4-month-old control and ttll3−/− mice. TTLL3 expression is completely abolished in all tissues tested in ttll3−/− mice, while TTLL8 expression levels are unchanged. No TTLL8 expression was detected in colon. Expression of TTLL3 detected by TTLL3-promoter–β-galactosidase (blue) on colon and testis tissue sections from 4-month-old mice. Scale bars are 100 μm. Tissues are counterstained with Nuclear Fast Red. Download figure Download PowerPoint To exclude the expression of trace amounts of TTLL8 in colon, we amplified TTLL3 and TTLL8 with RT-PCR using a very high number of PCR cycles. As controls, we used two tissues that assemble motile, highly glycylated cilia, that is, trachea and testis. Both, TTLL3 and TTLL8 are expressed in trachea and testes of wild-type mice, while no expression of TTLL8 was detected in colon, even after 40 PCR cycles (Fig 2B). The results of the PCR also corroborated the absence of TTLL3 in all tested tissues of ttll3−/− mice. To localize the expression of TTLL3 in colon tissue, we used ttll3−/− mice that carry a β-galactosidase insertion within the region of the TTLL3 gene. TTLL3 expression, visualized by staining with 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal), was detected in the epithelial cells from the bottom up to the top of the crypts. This indicates that TTLL3 is expressed throughout the colon crypts (Fig 2C). The β-galactosidase activity and thus TTLL3 expression were comparable between colon and testis confirming the qRT-PCR analysis (Fig 2A). We therefore conclude that the only enzyme available for catalyzing glycylation in colon is TTLL3. Consequently, downregulation, loss or enzymatic inactivation of TTLL3 are expected to result in the absence or at least in a decrease of glycylating activity in colon cells and should directly engender a loss of primary cilia. Absence of TTLL3 leads to reduced numbers of primary cilia in colon epithelium As primary cilia have so far not been described in colon tissue, we investigated ciliogenesis on cultured colon epithelial cells (CECs). Confluent cultured CECs from control and ttll3−/− mice were grown for 2 days under serum starvation and subsequently immunolabeled with antibodies for E-cadherin (to identify epithelial cells) and for detyrosinated α-tubulin (a specific marker of cilia and centrioles). Anti-detyrosinated tubulin staining clearly visualized short primary cilia on cultured CECs and identified the basal bodies (or centrosomes) of those cells that did not grow cilia (Fig 3A). Counting the number of ciliated CECs in densely grown regions of the cell cultures revealed a more than threefold decrease in the number of ciliated CECs from ttll3−/− mice as compared to controls (Fig 3B). Figure 3. Primary cilia in colon epithelium Colon epithelial cells (CECs) isolated from control (n = 7) and ttll3−/− (n = 6) mice at postnatal day 4 were cultured for 72–96 h and then serum-starved for 48 h. E-cadherin (green) labeling revealed adherens junctions of epithelial cells. Primary cilia and centrioles (basal body or centrosome) were visualized with the anti-detyrosinated tubulin antibody (red). Nuclei were stained with DAPI. Arrowheads indicate a representative primary cilium (control) and a basal body (or centrosome) of a non-ciliated cell (ttll3−/−). Scale bars are 10 μm. Quantification of the percentage of ciliated cells in clusters of E-cadherin-positive cells in CECs from control (n = 859) and ttll3−/− (n = 1,183) mice. Data are mean values ± SEM; *P < 10−2 by two-tailed unpaired t-tests and Mann–Whitney post-test. Confocal images (maximum projection) of cryo-sections from colon stained with the antibody Arl13b (red) and DAPI (blue). Arl13b specifically labels primary cilia (arrowheads). Cytoplasmic staining of stromal cells is non-specific, as Arl13b is a membrane protein (Cevik et al, 2010). The selected image shows a particularly high density of detectable primary cilia. Scale bars are 20 μm. Quantification of the number of clearly detectable Arl13b-positive ciliated cells per crypt. The relative numbers of crypts with 0, 1, 2, 3, 4 and more ciliated cells are represented as mean values ± SEM between individual mice (control n = 7; ttll3−/− n = 4). Total number of analyzed crypts: control = 259, ttll3−/− = 198. Average number of visualized ciliated cells per crypt. Mean values ± SEM are shown. *P < 10−2 by two-tailed unpaired t-test and Mann–Whitney post-test. Download figure Download PowerPoint To confirm the presence of primary cilia in colon epithelium in vivo, we investigated histological sections of mouse colon by immunofluorescence using different ciliary markers. We found, however, that many typical cilia markers such as antibodies for detyrosinated or acetylated tubulin failed to detect specifically primary cilia in colon tissue, which is in agreement with a previous report (Saqui-Salces et al, 2012). Only an antibody for Arl13b, a specific marker of primary cilia (Caspary et al, 2007), allowed us to detected primary cilia in the epithelial layers of the crypts (Fig 3C). Yet, only a small amount of epithelial cells were identified as being ciliated, most likely due to difficulties with antibody accessibility in the tissue. Despite these technical limitations, quantitative analysis of the number of clearly identifiable Arl13b-positive cilia demonstrated a reduction of the average number of cilia per crypt in ttll3−/− colons (Fig 3D). While the numbers of cilia per crypt varied between crypts within each colon analyzed, the mean values were strikingly similar between different animals, thus confirming a reproducible and significant decrease in the number of cilia in colon crypts of ttll3−/− mice (Fig 3E). Thus, colonic epithelial cells are able to form primary cilia, and lose cilia upon depletion of the unique glycylase TTLL3 in culture as well as in vivo. Together with the observations in MEFs (Fig 1) and ependymal cells (Bosch Grau et al, 2013), our results support a model in which the majority of cilia, primary as well as motile, depend on posttranslational glycylation. However, in the case of primary cilia, absence of glycylation leads only to a partial loss of cilia. Loss of TTLL3 results in increased cell proliferation in colon Primary cilia have been implicated in the control of cell division, tissue homeostasis and signaling (Lin et al, 2003; Croyle et al, 2011; Saqui-Salces et al, 2012; Wilson et al, 2012). To explore the impact of primary cilia loss in the absence of TTLL3, we assessed the proliferative activity and turnover of colon epithelia cells by bromodeoxyuridine (BrdU) incorporation. BrdU was intraperitoneally injected in control and ttll3−/− mice, and colons were dissected 2 h or 5 days after the injection. The number and position of BrdU-positive cells in colon crypts were assessed with anti-BrdU antibodies (Fig 4A). Two hours after injection, ttll3−/− colons displayed nearly four times increased number of BrdU-positive nuclei per crypt. Moreover, about 25% of these dividing cells were localized in the central region of the crypts, where virtually no proliferating cells were detected in wild-type colons (Fig 4B). Thus, the absence of TTLL3 deregulates cell division in colon epithelium. Figure 4. Loss of TTLL3 results in increased proliferation of colon epithelium Cell proliferation in colon crypts was analyzed by incorporation of BrdU. Representative pictures of BrdU immunohistochemistry (brown) 2 h and 5 days after intraperitoneal injection of BrdU. Paraffin-embedded colon sections from control and ttll3−/− mice were stained with anti-BrdU antibody and hematoxylin. Scale bar is 100 μm. Quantification of BrdU-positive cells in three compartments of the crypts (see scheme) 2 h and 5 days post-injection. At least 30 crypts per mouse were counted in three (2 h) and four (5 days) independent experiments (number of mice: 2 h control n = 3; 2 h ttll3−/− n = 3; 5 days control n = 4; 5 days ttll3−/− n = 4). Colon epithelial cells (CECs) from 4-month-old control and ttll3−/− mice were cultured for 48 h, and DNA labeled with 7-AAD was quantified by flow cytometry. Two representative DNA-content profiles are shown for wild-type and ttll3−/− CECs. Cells in G1 phase as well as dividing cells (S phase and mitosis; S + M) were quantified as indicated. Average number of cells in division (S + M) from quantifications as in (C). Bars represent SEM between different experiments (control n = 9; ttll3−/− n = 6). Relative increase in cell number (CECs) after 48 h in culture. Representative pictures of cyclin D1 staining (brown). Paraffin-embedded colon sections from control and ttll3−/− mice were stained with anti-cyclin D1 antibody and hematoxylin. Scale bar is 100 μm. Quantification of cyclin D1-positive cells in three compartments of the crypts (see scheme). At least 40 crypts per mouse were counted (control n = 8; ttll3−/− n = 8). Data information: Mean values with bars representing SD (B, G) or SEM (C, D) are shown. Significance values were determined by two-tailed unpaired t-tests and Mann–Whitney post test (*P < 10−2, **P < 10−3). Download figure Download PowerPoint To follow-up the fate of the faster-dividing cells, we analyzed colon tissue 5 days after BrdU injection. At this time point, the total number of BrdU-positive cells was lower in ttll3−/− as compared to control mice, and most of the BrdU-labeled cells in ttll3−/− mice were detected in the upper crypt compartment. In contrast, BrdU-positive cells were still present in the central compartment of the crypts in control animals. Thus, it appears that ttll3−/− mice display a higher proliferation rate that is coupled to a faster turnover of epithelial cells in the crypts. Most likely, the high dynamics of the colon tissue balances the increased proliferation of colonic epithelial cells by efficient shedding and thus maintains a normal architecture of the colon. This is underlined by the histochemical analysis of colon tissue, which revealed no obvious changes in differentiation of colon epithelial cells such as goblet cells (Supplementary Fig S1). To confirm these observations on the cellular level, we isolated CECs from control and ttll3−/− mice and cultivated them for 3 days in vitro. DNA was stained with 7-aminoactinomycin D (7-AAD), and the DNA content per cell was determined by flow cytometry (Fig 4C). In wild-type CECs, about 20% of the cells had a DNA content of > 2n, indicative of cells that are in S or in G2 phase/metaphase of the cell cycle (S + M). In contrast, about 45% of CECs from ttll3−/− mice were found to contain > 2n DNA (Fig 4D). This demonstrates that cultured CECs have a significantly higher mitotic activity in the absence of TTLL3. As a consequence of this higher proliferation activity in vitro, CECs from ttll3−/− mice divided more rapidly, as the significantly increased cell number after 48 h in vitro revealed (Fig 4E). We next compared the nuclear expression of cyclin D1, a proliferation marker (Tetsu & McCormick, 1999), between colon epithelial cells of control and ttll3−/− mice. The distribution of cyclin D1 (Fig 4F and G) mirrors the distribution of BrdU-positive cells 2 h after BrdU injection (Fig 4A and B), with an overall increase and a redistribution of cyclin D1-positive cells into the middle and upper part of the crypts of ttll3−/− colons. These findings confirm that absence of TTLL3 leads to increased cell proliferation and accelerated tissue turnover in colon epithelium. Loss of TTLL3 results in the amplification of tumor development in mice Despite the strongly increased proliferation rates, ttll3−/− mice have no gross abnormalities in colon tissue of young mice (2–3 months old). To assess the impact of TTLL3-deficiency on colon homeostasis in more detail, we examined colons of ten ttll3−/− mice older than 16 months. Histological analysis did not show any visible abnormality in the colon of ttll3−/− mice (Supplementary Fig S2). This indicates that colon tissue has a high compensatory potential for changes in proliferation rates. However, we speculated that the increased proliferation rate in TTLL3-deficient colon could promote cancer development. To test this possibility, we used a well-established mouse model of colitis-associated carcinogenesis (Tanaka et al, 2003; Suzuki et al, 2004). In this model, initial mutagenesis is induced by the intraperitoneal injection of azoxymethane (AOM), and subsequent inflammation in the colon by periodic administration of dextrane sodium sulfate (DSS) in the drinking water (Fig 5A). At the end of the protocol, mice were dissected and colorectal tumors were classified according to tumo
Referência(s)