Carcinogenic Parasite Secretes Growth Factor That Accelerates Wound Healing and Potentially Promotes Neoplasia

Abstract

Infection with the human liver fluke Opisthorchis viverrini induces cancer of the bile ducts, cholangiocarcinoma (CCA). Injury from feeding activities of this parasite within the human biliary tree causes extensive lesions, wounds that undergo protracted cycles of healing, and re-injury over years of chronic infection. We show that O. viverrini secreted proteins accelerated wound resolution in human cholangiocytes, an outcome that was compromised following silencing of expression of the fluke-derived gene encoding the granulin-like growth factor, Ov-GRN-1. Recombinant Ov-GRN-1 induced angiogenesis and accelerated mouse wound healing. Ov-GRN-1 was internalized by human cholangiocytes and induced gene and protein expression changes associated with wound healing and cancer pathways. Given the notable but seemingly paradoxical properties of liver fluke granulin in promoting not only wound healing but also a carcinogenic microenvironment, Ov-GRN-1 likely holds marked potential as a therapeutic wound-healing agent and as a vaccine against an infection-induced cancer of major public health significance in the developing world.

Fig 1. Liver fluke granulin internalized by H69 cholangiocytes.

(A) Widefield (deconvolved) micrographs showing the lateral (xy) overview of live H69 cholangiocytes imaged after 18 h incubation with Alexa Fluor 488-conjugated rOv-GRN-1 (green) and Hoescht nuclear stain (blue). (B) With further magnification of fixed cells the labeled rOv-GRN-1 was evident among the cytoskeletal actin network (red) of numerous cells with DAPI (blue) stained nuclei. (C) 3D-SIM lateral (xy) overview image of a well-separated individual cholangiocyte stained as in panel B. (D) Rendered axial (yz) view of boxed inset in (C) showing rOv-GRN-1 (green) present between the apical and basal actin filaments (red) of the cholangiocyte (DAPI channel omitted). Additional material shown in S1 Fig and S1 Movie.

Fig 2. Ov-GRN-1 stimulated wound repair in vitro.

(A) Cholangiocytes exposed to ES products from flukes where Ov-grn-1 had been silenced by RNA interference displayed significantly reduced proliferation over 36 h of co-culture. ES products (10 μg/ml) were derived from flukes that were exposed to dsRNAs for 5 days. Cell proliferation was monitored in real time using xCELLigence; every tenth data point is shown to aid visualization. Statistical comparisons were between Ov-grn-1- and luc-dsRNA-treated parasites. (B) Images of the scratch assay involving H69 cholangiocyte monolayers co-cultured in Transwell plates with Ov-grn-1 or luc-dsRNA-treated. Dotted lines denote wound edges over time. (C) Selected time points were measured from the photographs in (B); statistical comparisons were between cells cultured with Ov-grn-1-and luc-dsRNAs. (D) Wound healing scratch assay as shown in panel c but using the CCA cell line M214 (D). (E) Wound healing scratch assay as shown in panels (C) and (D) but recombinant protein applied to cells instead of co-culturing cells with live flukes. Statistical comparisons were between 20 nM rTRX and rOv-GRN-1 treatments or rTRX and PBS treatments. For all panels, data points represent the averages of two or three biological replicates with 3–5 biological replicates displayed with SEM error bars (some bars masked by data points). *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, ns = not significant. Additional data shown in S2 Fig.

Fig 3. Ov-GRN-1 stimulated wound repair in vivo.

(A) Sequential images over four days of healing wounds revealed the response of mice to treatment with recombinant rOv-GRN-1 or rTRX control; skin-deep wounds made with a 5 mm diameter biopsy punch between the ears of Balb/c mice. Minor modifications (brightness, contrast, cropping) were made to aid viewing. (B) The rate of wound healing over four days was measured; wound closure was determined electronically from photographs by measuring wound areas with ImageJ software. To aid viewing, curves have been shifted left or right marginally to minimize error bar overlap. (C) Assessment of the angiogenic properties of recombinant Ov-GRN-1 in the chorioallantoic membrane (CAM) assay. The numbers of blood vessels in quail eggs that grew on 0.5 cm² filter paper soaked in rOv-GRN-1 or vehicle (control) were ascertained after incubation for 15 hours. Data points are the averages of two experiments with 3–5 biological replicates displayed with SEM bars. * = P<0.05, ** = P<0.01, *** = P<0.001, **** = P<0.0001, ns = not significant.

Fig 4. Ov-GRN-1 stimulated responses in cholangiocytes.

(A) Clustered heatmap of 70 proteins for which expression levels were modulated between 0.5–48 h post-rOv-GRN-1 treatments. Only proteins that underwent significant changes in expression levels (> ±1.5-fold change relative to time point zero) are shown. The Euclidean distance clustering grouped proteins by translational temporal patterns and the XYZ designations distinguish the three major regulation patterns. Asterisks denote spliceosome-associated proteins. (B) Major KEGG pathways of cholangiocyte proteins whose expression levels were modulated by rOv-GRN-1. (C) Interactome of proteins whose expression levels were modulated by rOv-GRN-1. The top 25 proteins in terms of expression level changes were numbered with the average fold-upregulation; spliceosome-associated proteins shown surrounded by black circles. The A-E designations signify the major gene ontology groupings (defined in the legend) and the thickness of the lines linking different proteins represents the strength of the interactions. (D) Magnification of the boxed spliceosome grouping “E” from panel C. (E) Volcano plot of the cholangiocyte gene response to co-culture with rOv-GRN-1. Gene expression was measured using gene arrays designed to assess wound healing, oncogenesis, EMT and TLR associated transcripts detected using qPCR. (F) Heatmap depicting the changes of the significantly regulated (P < 0.05) genes shown in (E). Genes modulated with a > ±1.5-fold change denoted using bold type. Asterisks denote genes whose expression level significantly changed within one hour of co-culture with rOv-GRN-1. Genes where expression levels underwent significant changes within 24 hours (but not by one hour) of co-culture are indicated by the absence of asterisks. The color keys for panels (A) and (F) represent fold-change proportional to color intensity. Complete data sets of proteome and transcriptome changes provided in S1–S4 Tables.