Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Nov 14;9(1):4783.
doi: 10.1038/s41467-018-07306-7.

Colorectal Cancer Liver Metastatic Growth Depends on PAD4-driven Citrullination of the Extracellular Matrix

Affiliations
Free PMC article

Colorectal Cancer Liver Metastatic Growth Depends on PAD4-driven Citrullination of the Extracellular Matrix

A E Yuzhalin et al. Nat Commun. .
Free PMC article

Abstract

Citrullination of proteins, a post-translational conversion of arginine residues to citrulline, is recognized in rheumatoid arthritis, but largely undocumented in cancer. Here we show that citrullination of the extracellular matrix by cancer cell derived peptidylarginine deiminase 4 (PAD4) is essential for the growth of liver metastases from colorectal cancer (CRC). Using proteomics, we demonstrate that liver metastases exhibit higher levels of citrullination and PAD4 than unaffected liver, primary CRC or adjacent colonic mucosa. Functional significance for citrullination in metastatic growth is evident in murine models where inhibition of citrullination substantially reduces liver metastatic burden. Additionally, citrullination of a key matrix component collagen type I promotes greater adhesion and decreased migration of CRC cells along with increased expression of characteristic epithelial markers, suggesting a role for citrullination in promoting mesenchymal-to-epithelial transition and liver metastasis. Overall, our study reveals the potential for PAD4-dependant citrullination to drive the progression of CRC liver metastasis.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Label-free quantitative proteomics identifies the distinctive ECM composition of human CRC hepatic metastases. a Decellularized and ECM-enriched fractions from CRC hepatic metastases and paired adjacent unaffected liver tissues (n = 5 per group) were analyzed using LC-MS/MS with the following label-free quantitation. Shown is the principal component analysis of the relative protein abundance between metastasis and adjacent liver groups (P < 0.05). Percentage of variance is shown in parentheses. b Clustering analysis graph reveals categories of proteins differentially expressed in metastasis and adjacent liver groups. c Statistically significant differentially expressed proteins (287 out of 821 hits, two-way ANOVA, P < 0.05) were further classified either as ECM proteins or other proteins. d A heat map illustrating significantly different matrisome proteins (69 hits), which were selected based on a fold change threshold of > 3 and ranked in accordance with ECM category and fold change. The last gray colored table represents 9 selected other hits (i.e., non-ECM proteins yet significantly different in ECM-enriched samples). ECM extracellular matrix, P patient
Fig. 2
Fig. 2
ECM proteins in human CRC hepatic metastases are extensively citrullinated. a Schematic of PAD-mediated modification of arginine to citrulline residues. b Quantification of citrullinated peptides identified by LC-MS/MS in the enriched ECM from CRC hepatic metastases and paired adjacent unaffected liver tissues (n = 5 per group). c Quantification of citrullinated proteins (categorized as ECM proteins or other proteins) identified in the ECM from human CRC liver metastasis tissues (n = 5) and ECM of experimental liver metastases generated using HT29 (n = 2 biological replicates) and LoVo (n = 2 biological replicates) CRC cells. d Venn diagram demonstrating an overlap of citrullinated ECM proteins identified in human and experimental CRC hepatic metastases from c. Shared proteins are listed in the box to the right. e ELISA for citrullinated proteins was performed on tissue lysates of human CRC hepatic metastases, paired adjacent unaffected liver tissues (n = 12 per group), human primary CRC lesions and paired adjacent unaffected colon tissues (n = 5 per group). Representative of two experiments. f Densitometry analysis of immunoblotting for citrullinated proteins performed on tissue lysates extracted from patients with CRC hepatic metastases and paired adjacent unaffected liver tissues (n = 12 per group) and patients with primary colon adenocarcinomas and paired adjacent unaffected colon tissues (n = 5). For the densitometry, the sum of all bands was used. Densitometry results were normalized to loading control. Representative of two experiments. g, h ELISA for citrullinated proteins was performed on decellularized and biochemically enriched ECM from human CRC hepatic metastases, paired adjacent unaffected liver tissues (n = 5 per group) (g), and human primary CRC lesions and paired adjacent unaffected colon tissues (n = 5 per group) (h). Experiment was performed once. Throughout, error bars indicate range, box bounds indicate second and third quartiles, center values indicate median. *P < 0.05, **P < 0.01, ***P < 0.001, n.s. = non-significant, paired Student’s t test for b (P > 0.1 for Kolmogoro–Smirnov normality test), Mann–Whitney U test for g and h, and Kruskal–Wallis test with Dunn’s post-test for e and f. ECM extracellular matrix
Fig. 3
Fig. 3
PAD4 is produced by CRC cells and is contained in EVs. a ELISA for PAD4 was performed on tissue lysates extracted from human CRC hepatic metastases, paired adjacent liver tissues (n = 12 per group), human CRC lesions and paired adjacent colon tissues (n = 5 per group). b Densitometric analysis of immunoblotting for PAD4 performed on tissue lysates of human with CRC hepatic metastases and paired adjacent unaffected liver tissues (n = 12) and human primary CRC lesions and paired adjacent unaffected colon tissues (n = 5). For the densitometry, the most expressed 148 kDa dimer band was measured. Densitometry was normalized to loading control. c Relative (to HPRT) mRNA levels of PAD4 in CRC cells, granulocytes, other myeloid cells and the stromal cells isolated from CRC experimental hepatic metastases (n = 2 biological replicates per group). d ELISA for PAD4 was performed on the conditioned media collected from cultured CRC cells at the indicated time points and normalized to the total cell number. e Immunoblotting for PAD4 was performed on HT29, HCT116 and LoVo cells and their corresponding conditioned media 6 days post-seeding. Coomassie stain was used as a loading control. f Nanoparticle tracking analysis of particle size distribution (graph) and electron microscopy (image) of EVs isolated from HT29 cells. Scale bar = 250 nm. g Immunoblotting for indicated proteins in EVs isolated from CRC cells. Representative of two experiments. h Nanoparticle tracking analysis of particle concentration of EVs collected from HT29 cells treated with vehicle or GW4869. i Immunoblotting for indicated proteins in cells, conditioned media and EVs from cultured HT29 cells treated with vehicle or GW4869. j Representative images of co-immunostaining for PAD4 and collagen I in human CRC hepatic metastasis tissue. Scale bar = 100 μm. For a and b, error bars indicate range, box bounds indicate second and third quartiles, center values indicate median, For c and h, error bars indicate s.e.m., center values indicate mean (*P < 0.05, **P < 0.01, ***P < 0.001, n.s. = non-significant, Mann–Whitney U test for (h), and Kruskal–Wallis test with Dunn’s post-test for ac). EV extracellular vesicles, CM conditioned media, WB immunoblotting
Fig. 4
Fig. 4
Citrullination of collagen I alters adhesion, motility and epithelial mesenchymal plasticity of CRC cells. a Recombinant collagen type I alone, collagen type I pre-treated with recombinant PAD4, or collagen I pre-treated with recombinant PAD4 and BB-Cl-amidine, were incubated and subjected to LC-MS/MS analysis. Shown are the numbers of citrullinated peptides. Repeated twice. b GFP+ CRC cells were plated in wells pre-coated with either collagen type I alone, collagen type I pre-treated with recombinant PAD4 or collagen type I pre-treated with PAD4 and BB-Cl-amidine. Cells were imaged at the indicated time points using an epifluorescence microscope. Four to five fields of view per condition were taken. Shown are the cell numbers attached to the plate for each condition. c GFP+ MC38 cells were plated in wells pre-coated with either collagen type I alone, collagen type I pre-treated with recombinant PAD4, or collagen type I pre-treated with addition of PAD4 and BB-Cl-amidine. Individual cells were tracked using time-lapse microscopy. Shown is the quantification of cell median length and speed. d Representative images of cell tracks from the experiment in (c). Scale bar = 200 μm. e Cancer cells were plated in wells pre-coated with either collagen type I alone or collagen type I pre-treated with recombinant PAD4. At indicated times cells were collected and probed for the indicated proteins. Numbers indicate relative expression. f Shown is mRNA expression of different EMT genes in MC38 cells seeded on collagen type I pre-treated with PAD4 compared to those seeded on unmodified collagen type I 18 h post-seeding. Shown is average of 2 technical replicates. g Immunoblotting for the indicated proteins performed on cell lysates collected from the experiment in (f). GAPDH was used as a loading control. For a and b, error bars indicate s.e.m. center values indicate mean (*P < 0.05, **P < 0.01, Kruskal–Wallis test with Dunn’s multiple comparison post-test). For c, error bars indicate interquartile range, center values indicate median (*P < 0.05, ***P < 0.001, Kruskal–Wallis test with Dunn’s multiple comparison post-test). CTL control, BB BB-Cl-amidine
Fig. 5
Fig. 5
Pharmacological inhibition of PADs diminishes experimental hepatic metastases in vivo. a, b Immunoblotting for citrullinated proteins performed on tissue lysates extracted from experimental LoVo and HCT116 hepatic metastases from mice treated with vehicle or BB-Cl-amidine. Each lane indicates a biological replicate. GAPDH and Vinculin were used as a loading control. c Densitometric analysis for (a) and (b). d Representative images of immunostaining for citrullinated Histone H3 in experimental MC38 hepatic metastases from mice treated with vehicle (n = 5 biological replicates), or BB-Cl-amidine (n = 6 biological replicates). n = 3 sections per group, three images were analyzed per section. Scale bar = 100 μm. e Staining intensity quantification for (d). f Representative images of livers with experimental LoVo, HCT116 and MC38 hepatic metastases from mice treated with vehicle (n = 7, 6, and 5 biological replicates, respectively) or BB-Cl-amidine (n = 8, 6, and 6 biological replicates, respectively). Metastatic nodules are outlined in blue. Representative of two experiments. g Representative H&E-stained and scanned slides of dissected livers from the experiment in (f). Metastatic regions are outlined in black. Scale bar = 4 mm (scale bar of inserts = 200 μm). h Weight of excised metastasis-bearing livers from the experiment in (f). i Metastatic area based on the assessment of H&E-stained sections of livers from the experiment in (f). At least three different sections per mouse were analyzed. Throughout, error bars indicate s.e.m. center values indicate mean (**P < 0.01, n.s. = non-significant, Mann–Whitney U test). CTL control, BB BB-Cl-amidine
Fig. 6
Fig. 6
Pharmacological inhibition of PADs alters EMT markers in experimental hepatic metastases but not in subcutaneous xenografts. a Immunoblotting for the indicated EMT proteins performed on tissue lysates from the experiment in Fig. 5f. GAPDH was used as a loading control. b Representative images of staining for indicated proteins in experimental hepatic metastases from Fig. 5f. At least five biological replicates were stained per group. Scale bar = 100 μm. c Subcutaneous tumor growth curves of mice injected with indicated cells and treated with vehicle (n = 5, 5, and 6 biological replicates) or BB-Cl-amidine (n = 5, 5, and 5 biological replicates). Repeated twice. d Immunoblotting for the EMT marker proteins as indicated performed on tissue lysates from (c). For c, error bars indicate s.e.m. (*P < 0.05, n.s. = non-significant, two-way ANOVA). Ms – mouse. CTL control, BB BB-Cl-amidine
Fig. 7
Fig. 7
Genetic downregulation of PAD4 diminishes proliferation and subcutaneous tumor growth of CRC cells, and reduces experimental metastasis in vivo. a Immunoblotting for PAD4 in HT29 and HCT116 cells transfected with lentivirus harboring scrambled shRNA (CTL) or shPAD4 (shPAD4 #1 and #2). GAPDH was used as a loading control. b Immunoblotting for citrullinated proteins in conditioned media and cell lysates from control and PAD4-deficient HT29 and HCT116 cells. GAPDH was used as a loading control. c Immunoblotting analysis for PAD4 and citrullinated proteins in tissue lysates and isolated ECM from control (CTL) and PAD4 knockdown (shPAD#1) subcutaneous xenografts. GAPDH and COL5A1 were used as loading controls for tissue lysates and isolated ECM, respectively. d WST-1 proliferation and viability assay performed in cultured wild-type and PAD4-deficient HT29 and HCT116 cells (n = 5 technical replicates per group). e Tumor growth curves of mice injected subcutaneously with control and PAD4-knockdown HT29 and HCT116 cells (n = 7, 4, and 5 biological replicates for HT29 and 9, 5, and 5, for HCT116, respectively). Repeated twice. f Representative images of livers from mice injected intrasplenically with control (n = 8 biological replicates) or PAD4-deficient HT29 and HCT116 cells (n = 6 and 6 biological replicates for shPAD4 #1 and #2) 35 days postinjection. Metastatic nodules are outlined in blue. g Liver weights from the experiment in (f). h Measurement of the area of metastatic deposits in HT29 livers from the experiment in (f). For g, error bars indicate s.e.m. center values indicate mean (**P < 0.01, Kruskal–Wallis test with Dunn’s multiple comparison post-test). For d and e, error bars indicate s.e.m. center values indicate mean (***P < 0.001, two-way ANOVA). Ms mouse, CTL control, shPAD4 PAD4 knockdown, ECM extracellular matrix
Fig. 8
Fig. 8
Genetic downregulation of PAD4 in HCT116 cells diminishes spontaneous hepatic metastases in vivo. a Experimental pipeline. b Weight of excised primary cecal tumors (n = 8 biological replicates per group). c Number of visible macrometastases (n = 10 biological replicates per group). Other organs including lungs showed no evidence of metastasis. d Representative image of a GFP+ metastatic lesion counterstained with DAPI (left and middle) and the corresponding H&E image (right). Scale bar = 100 μm. e Numbers of GFP+ micrometastatic lesions observed (n = 5 biological replicates per group). f Area of micrometastases evaluated by H&E (n = 5 biological replicates per group). g Relative (to HPRT) mRNA levels of PAD4 in visible macrometastases (n = 6 and 4 for CTL and shPAD4 groups, respectively). h Relative (to HPRT) mRNA levels of PAD4 in primary cecal lesions (n = 7 and 6 for CTL and shPAD4 groups, respectively). Throughout, error bars indicate s.e.m. center values indicate mean (*P < 0.05, **P < 0.01, n.s. = non-significant, Mann–Whitney U test). CTL control, shPAD4 PAD4 knockdown, H&E hematoxylin and eosin

Similar articles

See all similar articles

Cited by 14 articles

See all "Cited by" articles

References

    1. Lu P, Weaver VM, Werb Z. The extracellular matrix: a dynamic niche in cancer progression. J. Cell. Biol. 2012;196:395–406. doi: 10.1083/jcb.201102147. - DOI - PMC - PubMed
    1. Zeltz C, Gullberg D. Post-translational modifications of integrin ligands as pathogenic mechanisms in disease. Matrix Biol. 2014;40:5–9. doi: 10.1016/j.matbio.2014.08.001. - DOI - PubMed
    1. Leeming DJ, et al. Post-translational modifications of the extracellular matrix are key events in cancer progression: opportunities for biochemical marker development. Biomarkers. 2011;16:193–205. doi: 10.3109/1354750X.2011.557440. - DOI - PubMed
    1. Barker HE, Cox TR, Erler JT. The rationale for targeting the LOX family in cancer. Nat. Rev. Cancer. 2012;12:540–552. doi: 10.1038/nrc3319. - DOI - PubMed
    1. Valesini G, et al. Citrullination and autoimmunity. Autoimmun. Rev. 2015;14:490–497. doi: 10.1016/j.autrev.2015.01.013. - DOI - PubMed

Publication types

MeSH terms

Feedback