Fig 1.
DCLK1 expression negatively correlated with FoxD3 status in the colons of DCLK1ΔIEC mice.
A, B. Bright-field and immunofluorescence (IMF) images of Control, CR or CR+DBZ-treated mouse colon sections showing DCLK1 and FoxD3 staining. C, D. Representative heatmap channels generated for DCLK1 and FoxD3 in the indicated groups using MCD-Viewer. Boxed areas in C and D represent magnified images of DCLK1 and FoxD3 captured at the same location to highlight differences in staining. Scale bar = 200 μm; n = 2 independent experiments.
Fig 2.
DCLK1-S persists during colon inflammation and displays an inverse correlation with FOXD3.
A. Using the IMC data sets from Dclk1ΔIEC mice, DCLK1 (red) is overlayed with FoxD3 (green) in the indicated groups. B. Development of DCLK1-SOE mice. Mouse genomic fragments containing homology arms (HAs) were amplified using high fidelity Taq and were subsequently assembled into a targeting vector together with recombination sites and selection markers shown. Dclk1-Short+19-floxed mice were bred with Villin-Cre mice to generate Dclk1-Short+19-floxed; Villin-Cre mice. Mice were subsequently genotyped. Ci. WT littermates and
mice received 4% DSS in drinking water for 7 days and euthanized after 2 days of normal water. Representative H&E staining (n = 6 mice/genotype) to visualize crypt architecture. Scale bar = 150 μm. Cii. Bar graph showing histological scores. D. IMC showing DCLK1(red) staining overlayed with DNA (blue), FoxD3 (red) staining overlayed with DNA (blue) and DCLK1 (red)/FoxD3 (green) and DNA (blue) colocalization in the Control or DSS-treated
mice. Scale bars as indicated; n = 6 mice/genotype. E. Percent fold change based on staining intensity in DCLK1+ cells, fold change of DCLK1 is in blue, fold change of FoxD3 is in red, and the ratio of both DCLK1/FoxD3 is in orange.
Fig 3.
Co-localization of MMP13 with DCLK1 in various models of colitis.
A. Paraffin sections prepared from the colons of Control, CR or CR+DBZ group of Dclk1ΔIEC mice were subjected to IMC. MMP13 (green) is overlayed with DCLK1 (red) and DNA (blue). Boxed areas in CR group indicates significant co-localization of MMP13 with DCLK1. Scale bars as indicated (100 μm); 8-10 mice /group. B. Box plots of DCLK1 and MMP13 counts based on IMC data set in the Control, CR, CR+DBZ groups. C. DCLK1 and MMP13 staining from IMC in control and Dclk1-SOE group after DSS-induced colitis. Lane 1 DCLK1(red) staining is overlayed with DNA (blue), Lane 2 MMP13 (red) staining is overlayed with DNA (blue), Lane 3 DCLK1(red)/MMP13(green)/DNA(Blue) colocalization in the Dclk1-SOE mice. P values as indicated. D. MMP13 promoter-reporter activity (*, **p<0.05; n = 3 independent experiments). Ei. Western blot data of HCT116 colon cancer cells treated with PMA and PMA+DBZ. Eii. In silico molecular docking studies to predict MMP13 and DCLK1 binding. CR: Citrobacter rodentium, CR+DBZ: Citrobacter rodentium + Dibenzazepine (DBZ), PMA: Phorbol 12-Myristate 13-Acetate.
Fig 4.
MMP13 enzyme activity, subcellular distribution, DCLK1-L and DCLK1-S promoter-reporter assays.
RKO cells were treated with PMA or PMA plus selective and potent MMP13 inhibitor WAY 170523 at varying doses as indicated for 30 min followed by measurement of enzymatic activity. A. Reference curve showing relative fluorescence units (RFUs). B, C. Dose-dependent decrease in MMP13 enzymatic activity (n = 3 independent experiments; *p<0.05). D. Western blots showing relative protein abundance in three colon cancer cell lines. Boxed area represents levels of the indicated proteins in RKO cells. E. Subcellular compartmentalization of proteins from RKO cells. 1: Cytosolic Fraction, 2: Nuclear Fraction, 3: Membrane Fraction, 4: Cytoskeletal fraction (n = 3 independent experiments). F. Promoters for DCLK1-L and DCLK1-S isoforms were cloned and transfected in HEK293 cells and promoter-reporter activity assays were performed using Dual-Luciferase Reporter Assay System (E1910, Promega, Madison, WI). Luminescence was measured using a BioTek Synergy Neo luminometer. P values as indicated; n = 3 independent experiments.
Fig 5.
Molecular docking and in vitro kinase assays.
A-D. Molecular docking study showing the interaction of proMMP13 (A) and mature MMP13 (C) with DCLK1-S. B. Ser 73 and Ser 114 highlighted in green in proMMP13, represent the phosphorylation sites for DCLK1-S. D. Arg207 and Gln211 are donor sites from MMP13 in mature protein close to Thr197 (circled). E. Both rhMMP13 and rhDCLK1 were incubated in kinase buffer in presence or absence of ATP at 30 °C for 1 hr. rhMMP13 didn’t show any autophosphorylation in the same condition. Phosphorylation was detected by pan phospho-Serine/Threonine antibody. F. rhMMP13 was incubated with rhDCLK1 in presence of ATP at 30 °C for 1 hr. Phosphorylated rhMMP13 was detected with immunoblotting using pan phospho-Serine/Threonine antibody. Representative immunoblots showing the levels of phosphorylation of rhMMP13. Right panel represents the Coomassie staining (n = 3 independent experiments). G, H. Protein band intensity ratios, pMMP13/pDCLK1 and pMMP13/Total MMP13, showing rhMMP13 phosphorylation by rhDCLK1.
Fig 6.
MMP13 upregulation coincides with changes in ECM, EMT and fibrosis markers in vivo in Dclk1ΔIEC mice.
A. Tissue sections prepared from the colons of indicated groups of mice were subjected to immunostaining with antibodies against MMP13 (red), α-SMA (green), Collagen (green), E-cadherin (green) and Vimentin (green). Samples were analyzed using the Hyperion Imaging System (Standard BioTools). DAPI (blue) was used to label DNA. Scale bars = 100 μm (n = 8-10mice/group). B. Box plots of MMP13 and Collagen counts based on IMC data set in the Control, CR and CR+DBZ groups. C. An overlay of Collagen (Red) and MMP13 (green) in tissue sections of the indicated groups. The boxed area represents a magnified image of collagen accumulation in the CR+DBZ group. Scale bars = 100 μm (n = 8-10mice/group). D. t-SNE plots showing MMP13 intensity across different groups. E. Masson’s Trichrome staining of tissue sections prepared from the colons of Control, CR, and CR+DBZ mice. Please note increases in collagenous fibrous tissue (stained blue) in both CR and CR+DBZ groups. Scale bars = 200 μm; n = 8-10mice/group. CR: Citrobacter rodentium, CR+DBZ: Citrobacter rodentium + Dibenzazepine (DBZ).
Fig 7.
Loss of E-cadherin and increased deposition of Vimentin suggests an EMT phenotype with DCLK1-S driven inflammation in the colons of Dclk1ΔIEC mice.
A. Two-dimensional t-SNE plots showing the clusters and relative abundances of cells. Numbers below represent markers associated with the indicated clusters. B. Representative heatmap channels generated for E-cadherin and Vimentin in the indicated groups using MCD-Viewer. Boxed area in CR+DBZ group represents higher Vimentin intensity. Scale bars as indicated; n = 8-10mice/group. C. Box plots of Vimentin and E-cadherin counts are based on the IMC data set in the Control, CR and CR+DBZ groups. P values as indicated. D. DCLK1(red) Vimentin (green) and DNA (blue) staining in the indicated groups. Boxed area represents DCLK1-Vimentin co-localization. Scale bars as indicated; n = 8-10mice/group. E. Box plots of DCLK1 and Vimentin counts based on IMC data set in the Control, CR, CR+DBZ groups. CR: Citrobacter rodentium, CR+DBZ: Citrobacter rodentium + Dibenzazepine (DBZ).