Cell stress in the protein-making system controls intestinal stem cell behavior through CtBP2

To demonstrate that CtBP2 protein expression in the intestinal crypt bottom is affected by ER stress in vivo, we administered intraperitoneal injection of thapsigargin or vehicle (DMSO) in adult mice and assessed small intestinal tissue 8 h after administration. Although Ctbp2 mRNA levels were not significantly altered (Figure S3A), CtBP2 protein expression in the crypt was markedly lower (Fig. 2E, F). This implicates that intestinal CtBP2 is mainly regulated post-transcriptionally during exposure to ER stress. In line with diminished CtBP2 downstream activity, mRNA expression of T-cell lymphoma invasion and metastasis 1 (Tiam1), one of the few described targets directly activated by CtBP2 at the transcriptional level, was significantly downregulated in these mice (Figure S3B). Induction of ER stress with the highly specific ER-stress inducing agent SubAB in the human CRC cell line LS174T also lowered TIAM1 expression (Figure S3C). Next, we analyzed if loss of CtBP2 protein upon induction of ER stress was dependent on PERK-eIF2α signaling. To this end we treated LS174T cells with thapsigargin and specifically inhibited PERK activation using a specific PERK inhibitor (GSK2606414). Indeed, we observed that induction of ER stress for 24 h led to loss of CtBP2 expression that could be largely rescued upon inhibition of eIF2α signaling, indicating that CtBP2 protein levels are translationally regulated by PERK-eIF2α signaling (Fig. 2G). A similar result was found when ER stress was induced using tunicamycin, a naturally occurring antibiotic that induces ER stress in cells by interfering with protein folding (Figure S3D). To demonstrate that ER stress results in rapid PERK-dependent loss of nascent CtBP2 protein, we assessed CtBP2 expression following pulldown of nascent protein using L-azidohomoalanine (AHA) labeling. As expected, 4 h of thapsigargin resulted in near complete loss CtBP2 synthesis that could be largely rescued when PERK was inhibited (Fig. 2H). Thus, our data indicates that intestinal CtBP2 expression strongly depends on continuous global protein translation, causing a proclivity for rapid loss of expression following the translational pause as part of the activated UPR during ER stress.

Next, we assessed a functional role for CtBP2 in self renewal and stem cell function by submitting organoids with CtBP2 overexpressing to radiation-induced injury. Gamma irradiation induces damage, leading to cell death and loss of proliferation of cells. Eventually a tissue repair response in which intestinal stem cells are known to be indispensable, results in tissue restauration³³. To assess the stem cell dynamics of the response following irradiation injury, we measured Lgr5 and Olfm4 expression levels over the course of days. CtBP2 overexpression mitigated reduced mRNA expression of Lgr5 and Olfm4 and promoted restoration to normal levels (Fig. 4E). Strikingly, whilst wildtype organoids did not regenerate properly and mostly disintegrated days after irradiation, CtBP2 overexpression protected against this injury phenotype (Fig. 4F, G). Taken together, these findings support a role for CtBP2 in protection of the stem cell pool during conditions of proteotoxicity from ER stress and genotoxicity from ionizing radiation.

Finally, we investigated the effect of CtBP2 depletion in the LS174T CRC cell line, which is derived from a spontaneous Dukes B colon adenocarcinoma and has a well-described deregulated WNT/β-catenin oncogenic signature³⁷. Transduction of shRNAs against CTBP2 in these cells also resulted in strong downregulation of CtBP2 (Figure S4A). CTBP2 knockdown significantly inhibited cell growth and proliferation, assessed by crystal violet staining, EdU incorporation and reduction of stem cell markers (Fig. 5G, Figure S4B and C). Furthermore, protein expression of the WNT-associated transcription factors c-MYC, TCF4, and downstream target CD44 were completely lost after CtBP2 knockdown, indicating that CtBP2 operates upstream or at the level of TCF4, which is consistent with a previous report (Fig. 5H)³⁸. In addition, multiple intestinal stem cell markers were downregulated including LGR5, CD44, LRIG1 and BMI1 (Figure S4C) whilst differentiation markers P21 and MUC2 were increased (Figure S4D). Next to growth, cell migration which is another hallmark important for colon cancer progression associated with WNT signaling³⁹, was also impaired by CtBP2 knockdown (Fig. 5I, J). This fits with reported function of CtBP2 to promote cell migration through the downstream guanine nucleotide exchange factor (GEF) Tiam1⁴⁰. Indeed, we found that TIAM1 expression was lower in these cells (Figure S4E). Moreover, high TIAM1 expression correlated with decreased overall survival in a large dataset of CRC patients (Figure S4F). Expression was particularly high in the CMS1 and CMS4 subtypes, which are the cancers with a poor prognostic outcome (Figure S4G).

Taken together, we demonstrate that ER stress induced loss of stemness is characterized by diminished activity of various transcription factors. One of these factors, co-regulator CtBP2, was found to be important for maintenance of intestinal stem cell function. Furthermore, our data supports a crucial role for CtBP2 in stemness and growth of adenomatous and carcinomatous cells that are characterized by an oncogenic WNT signature.

Written informed consent was obtained from all donors in this study for the use of the material for research purposes. Tissues were obtained with approval of the ethical committee of the Amsterdam UMC, together with approval of the experimental procedures by the HIS Mouse Facility (Amsterdam UMC). All methods were performed in accordance with the relevant guidelines and regulations, as stated in the Amsterdam UMC Research Code, in a certified laboratory (ISO15189 accreditation M304).

Fresh adenoma and adjacent healthy tissue samples were collected and snap-frozen immediately after resection from patients in the endoscopy program for the removal of large (> 1 cm) colorectal adenomas at the University Medical Centers Amsterdam (location AMC). All patients provided written informed consent (METC2015_206). mRNA was isolated from tissue samples by mechanical disruption using the FastPrep-24-5G (MP Biomedicals) in combination with Lysis Tubes S (Qiagen, Cat No./ID: 19091) for 2 times 1 min at 6.5/s with 1 min on ice in between followed by the Allprep DNA/RNA Universal kit (Qiagen, ID 80204).

For generation of human FAP organoids, adenomatous material from the colon was collected at the University Medical Centers Amsterdam (location AMC), after being approved by the Medical Ethical Committee (normal tissue: MEC 09/146 and MEC 05/071).

The study was carried out in compliance with the ARRIVE guidelines. All animal experimentation methods were performed in accordance with the Animal Ethical Committee guidelines and regulations of the Leiden University Medical Center (Workprotocol 68215), the Netherlands. For ER stress induction, adult wild‐type (WT) C57Bl/6J mice received a single intraperitoneal injection of thapsigargin dissolved in DMSO (1 mg/kg) or DMSO alone 8 h or 24 h before being euthanized.

For CtBP2 immunostaining in homeostatic stem cells, and immunostaining in APC-driven polyps, the Lgr5-EGFP-IRES-creERT2 knock-in mouse and the Villin-Cre-ERT2-Apc^15lox/+ mouse models were used, respectively.

LS174T colorectal cancer cells (ATCC CL188) were grown in Dulbecco’s modified Eagle’s medium with 10% fetal calf serum (FCS) and 1% penicillin/streptomycin under standard culture conditions.

Mouse small intestinal epithelial organoids were cultured in ENR medium containing N2, B27 supplements (Invitrogen), 1.25 mM n-acetylcysteine, 50 ng/ml mouse EGF (Invitrogen), Noggin-Fc-conditioned medium (20%, equivalent to 200 ng/ml), Rspo1-Fc-conditioned medium, unless specifically stated otherwise. Patient-derived FAP organoids were cultured in ENR medium, with addition of 2-mercaptoethanol (Sigma-Aldrich), gentamicin (BioWhittaker), 50 mM TGF-βRI inhibitor A38-01 (Tocris Bioscience), 30 mM p38 inhibitor SB202190 (Sigma-Aldrich), and 500 µM (Leu15)-Gastrin (Sigma-Aldrich). Organoid growth was assessed after re-passaging organoids in a 48-well plate and photos were taken using a Leica DM6000 Digital Microscope equipped with LAS AF Software (Leica, Wetzler, Germany). Quantification was done by measuring the perimeter and surface area using Image-J software. For the irradiation-induced regenerative response experiments, we treated the organoid cultures, 3 days after seeding, with 6 Gy of γ-radiation as previously described⁵⁴. We made used of the following organoid models:

Inducible CtBP2 overexpression was generated as follows: in short, small intestinal tissue ~ 2 cm was collected from body wide transgenic CAG-rtTA3 mice, which were gently scraped with coverslip, washed with ice-cold PBS, and then incubated with 2 mM EDTA in PBS on a rotating wheel to obtain the epithelial fraction. Residual debris were removed by gentle shaking, the debris-containing supernatant was removed and replaced with cold PBS. This procedure was repeated until supernatant was clear. After passing the supernatant through a cell strainer and centrifuge, the pellet was resuspended in 20 µl Matrigel (BD Biosciences) at a desired crypt density and plated in a 48-wells plate. After expansion of the culture, organoids were prepared for transduction according to a previously described methodology²⁵.We used the pRetroX-Tight-Pur vector (Clontech), for expressing a gene with the Tet-On Advanced system, in which we cloned the human coding sequence of CtBP2. After puromycin selection, induction of CtBP2 was achieved using 1 μg/ml doxycycline in ENR medium.

CtBP2 knockdown in human FAP organoids was generated as follows: two subclones of polyps from a FAP patient were used, from the organoid biobank at the Amsterdam UMC according to previously described methodology²⁵. The first line carried only APC loss of heterozygosity, whereas the second line carried an additional spontaneous gain-of-function point mutation in KRAS (G > D). Stable knockdown was achieved using lentiviral transduction with the pLKO.1 vector containing 3 different shRNAs against human CTBP2 (TRCN0000013743, TRCN0000013744 and TRCN0000013746) Selection was ensured using 2 μg/ml puromycin. CTBP2 after loss of APC was studied used Villin-Cre-ERT2-Apc^15lox/15lox organoids that were recombined in vitro using 0.75 µg/ml 4-hydroxy tamoxifen for 24 h.

We used the catTFRE assay to quantitatively measure proteome-wide changes in DNA binding activity of transcription factors in response to ER stress. A detailed methodological description of the catTFRE assay has been described earlier¹⁷. In brief, LS174T cells were plated in a 6-well plate for treatment, after which cell pellets were harvested and the nuclear fraction was extracted. Subsequently, a pulldown was performed using a biotin-labeled synthetic DNA plasmid containing a concatenated tandem array of consensus transcription factor response elements (TFREs) for most known transcription factor families. After trypsin digestion, the tryptic peptides containing the enriched endogenous transcription factors were subjected to a LTQ-Orbitrap Velos mass spectrometry (Thermo Fisher Scientific) for measurement. Peptides of transcription factors with good signal response were selected for quantification by calculating the area under the curve. Proteins were identified by using the National Center for Biotechnology Information search engine against the human or mouse National Center for Biotechnology Information RefSeq protein databases. The list of candidate genes was further analyzed using DAVID Functional Annotation Tool (david.ncifcrf.gov). We assessed the top 100 genes present in both treatment groups (thapsigargin and Subtilase AB) and looked for significant enrichment of these genes in gene annotation categories of the DAVID database, setting a threshold of minimally 10 genes per category.

Cells were lysed in cell lysis buffer (Cell Signaling Technology), and heated to 95 °C for 5 min in sample buffer containing 0.25 M Tris-HCl (pH 6.8), 8% SDS, 30% glycerol, 0.02% bromophenol blue and 1% β-mercaptoethanol. Separation was done on 6–12% SDS-PAGE, and proteins were transferred to a PVDF membrane. Specific detection was done by incubating the blot overnight in TBS with 0.1% Tween-20 and 1% BSA. Antibody binding was visualized using the Lumi-Light western blotting substrate (Roche). Images were captured with a Leica DM6000 microscope using LAS AF software (Leica, Wetzlar, Germany). All western blot images were generated and exported using ImageQuant LAS 4000 software and cropped and labeled using Adobe Illustrator software version 25.2. Antibodies used: anti-CTBP2 (Cell Signaling, #13256), anti-ZMYM2 (Abcam, ab106624), anti-FOXO3 (Cell Signaling, #2497s), anti-ZFPM1 (Abcam, ab86281), anti-CDX1 (Abcam, ab126748), anti-MNX1 (Abcam, ab95926), anti-IRF8 (Cell Signaling, #5628), anti-c-MYC (Santa Cruz, sc-764), anti-CD44v6 (Biorad, #MCA1967), anti-TCF4 (Cell signaling, #2569S) anti-GRP78 (Cell Signaling #3183), anti-PERK (Cell Signaling #3192), anti-beta Actin (Sigma-Aldrich, A1978), Vinculin (Cell Signaling, #18799S).

L-Azidohomoalanine (AHA) incorporation and pulldown assays were performed to quantify nascent protein levels of CtBP2. In brief, endogenous methionine was depleted in LS174T cells by preincubation in methionine and cysteine free DMEM supplemented with glucose and glutamine. Subsequently, cells were labeled with 1 mM of AHA (Iris Biotech) and 400 nM of thapsigargin for 4 h, and controls were added for labeling (50/50 ratio of methionine and AHA) and global translation (100 mM cycloheximide). Then, after 2 washes with HBSS (w/o calcium) protein lysate was harvested in HEPES buffer (pH 7.4) containing 1% NP-40 with protease inhibitors. After measuring protein content using BCA, a click reaction was achieved by adding a mixture of the following components per 500 µg protein: 250 µM biotin-alkyne (Sigma-Aldrich), 1 mM copper sulphate (Sigma-Aldrich), 5 mM THPTA (Sigma-Aldrich) and 15 mM sodium ascorbate (Sigma-Aldrich). After incubate at room temperature for 1 h on the end-over-end shaker, 2 µl of 0.5 M EDTA was added to each sample, and lysates were loaded on a Zeba column to remove unbound Dde-alkyne biotin. 40 µl of washed NeutrAvidin beaded agarose resin (Thermo Fisher Scientific) in a 50/50 slurry with PBS was added and samples were incubated overnight on an end-over-end rotor at 4 degrees. After several washes in PBS + 1%NP40, Laemmli sample buffer supplemented with 100 mM DTT was added, following 15 min incubation at 65 degrees. This process was repeated in a 10 min incubation at 90°, after which samples were loaded on SDS-PAGE, proceeding with standard Western blot procedure (see above).

Tissue was fixed in 10% ice-cold formalin and embedded in paraffin⁶. Sections of 4 µm thickness were deparaffinized in xylene and rehydrated. Endogenous peroxidase was blocked using 0.3% H₂O₂ in methanol for 30 min. The following methods of antigen retrieval were used: Sodium citrate (slides were cooked at 100 °C for 20 min in 0.01 M sodium citrate, pH 6); Tris/EDTA (slides were cooked at 100 °C for 20 min in a Tris/EDTA buffer, 10 mM Tris, 1 mM EDTA, pH 9.0); proteinase K (slides were incubated with Proteinase K (Dako, S302030) for 5 min at room temperature). After antigen retrieval slides were blocked in PBS using 0.1% Triton X-100 and 1% bovine serum albumin (PBT) for 30 min, followed by incubation overnight at 4 °C with a primary antibody (anti-CTBP2 Cell Signaling, #13256) in PBT. Antibody binding was visualized using Powervision HRP (Immunologic) labeled secondary antibodies and diaminobenzidine (Sigma-Aldrich) for substrate development. All sections were counterstained with Mayer’s haematoxylin.

For Immunofluorescence, slides were incubated for 1 h using fluorescently labeled secondary antibodies (all Alexa Fluor secondary antibodies from Invitrogen), diluted 1:500 in PBS with 0.1% Triton X-100 and 1% bovine serum albumin at room temperature. Slides were washed and mounted with Slowfade Gold Antifade reagent with DAPI (Invitrogen, s36938). Images were obtained on a Leica DM6000 Digital Microscope equipped with LAS AF Software (Leica, Wetzler, Germany). For analysis ImageJ software was used.

RNAscope in situ hybridization (ISH) was performed according to the manufacturers’ protocol (Advanced Cell Diagnositcs, ACD). The single-plex RNAscope probe against mouse Ctbp2 (#489981 ACD Europe SRL) was used.

For migration Scratch assays LS174T cells were grown in a 6-well plate until full confluence was reached. Two scratches per well were made using a p10 pipet tip. Cells migration was filmed overnight with a Leica DMi8 inverted microscope, fitted with a humidified culture chamber maintained at 37 °C and the covered areas were quantified with Image-J software (National Institutes of Health, Bethesda, MD).

EdU staining was performed to analyze cell proliferation by flow cytometry using the Click-iT EdU Alexa Fluor 647 kit (Invitrogen). Organoids were incubated with 10 μM EdU for 3 h followed by TrypLE™ mediated cell-dissociation. Cell were fixed for 20 min with 3.7% formaldehyde, washed and incubated with Click-iT reaction cocktail according to manufacturer’s protocol. Fluorescence was quantified on a flow cytometer (LSR Fortessa) and the percentage of EdU-positive cells was calculated with FlowJo 8.0 software.

For crystal violet analyses, cells were fixed with 4% paraformaldehyde in PBS for 15 min and subsequently immersed in 5 mg/ml crystal violet (Sigma C3886) in 2% ethanol in H2O.

Statistical analysis was performed using Prism 8.0 (GraphPad Software, La Jolla, CA). All data are presented as mean ± SEM of three independent experiments of technical triplicates unless specifically stated otherwise. Results were analyzed using Student’s t test or one-way ANOVA followed by Bonferroni post-test as indicated throughout the figure legends. In one instance, the Wilcoxon signed-rank test was used to compare paired samples. Differences were considered statistically significant at a P value < 0.05.

Supplementary Information 1. Supplementary Information 2.