Immunomics-guided biomarker discovery for human liver fluke infection and infection-associated cholangiocarcinoma

Abstract

Sensitive diagnostics are needed to improve management and surveillance of opisthorchiasis and opisthorchiasis-associated cholangiocarcinoma (CCA) throughout East Asia. Herein we generate and screen an Opisthorchis viverrini recombinant secreted proteome to identity antibody biomarkers of liver fluke infection and CCA with sera from study participants in endemic populations and evaluate their utility as point-of-care immunochromatographic tests (PoC-ICTs). We incorporate two of the most promising antigens from the proteome array screen, P1 and P9, into PoC-ICTs to further validate their diagnostic performance. The P9-IgG4 PoC-ICT is superior amongst the single recombinant antigen tests for diagnosing fluke infection as well as fluke-induced CCA, and out-performs parasite crude extract-IgG ICTs. Here we identify two biomarkers of O. viverrini infection and infection-associated CCA that could form the basis of novel antibody serodiagnostic tests for human liver fluke infection and associated cancer.

Fig. 1. Summary of immunomics approach to antigen discovery and validation using point-of-care immunochromatographic tests (PoC-ICT) for diagnosis of opisthorchiasis and opisthorchiasis-associated cholangiocarcinoma.

We used an integrated approach to identify biomarkers of Opisthorchis viverrini (Ov- curved yellow shape) infection and associated cholangiocarcinoma. A 825 total excreted/secreted (ES) proteins (water drop) were considered for investigation. B 249 proteins were selected for cell-free expression and 245 were successfully expressed. Large proteins were broken up into two ORFs for cell-free expression, resulting in a total of 278 proteins or protein fragments being printed onto a proteome microarray. Four O. viverrini ES proteins identified in previous studies and expressed in E. coli and purified were also printed onto the array as controls. C The proteome array was screened with sera from O. viverrini infected subjects, and 36 antigens were the targets of significantly (p < 0.05) elevated IgG or IgG4 responses relative to healthy (uninfected) controls. D The top nine candidates were selected for bacterial expression. E Five candidates were successfully expressed, purified and printed onto pilot immunochromatographic tests (ICTs). F The two best performing candidates were incorporated into ICTs for serological validation. Image created in Biorender. Smout, M. (2025) https://BioRender.com/2i9y022.

Fig. 2. IgG1 and IgG4 antibody responses to Opisthorchis viverrini recombinant proteins printed on a proteome microarray.

Top 40 protein hits ranked by t-test significance comparing IgG1 and IgG4 (a) responses from 100 negative subjects (50 O. viverrini fecal egg count test (FECT)-negative subjects living in an endemic region of Thailand and 50 non-endemic negative subjects residing in the US, red bars) and 50 O. viverrini FECT-positive subjects (Thailand, blue/white bars). RFU: relative fluorescence units. Bars indicate standard error of the mean (SEM). p-value comparing infected and uninfected negative control responses shown as green dots with green dashed line marking the p < 0.05 threshold. Protein accession numbers are provided on the y axis, and fluke extracts (OvES, excretory/secretory antigen; OvSo, somatic lysate antigen) were printed as controls. Immunoreactive proteins selected from the proteome array screen (b) for the serodiagnosis of opisthorchiasis and opisthorchiasis-associated cholangiocarcinoma (CCA). Protein accession codes and biological functions (families) are listed, and for ease of nomenclature designated as P1-P9. The heatmap reflects the degree of statistical significance for IgG1 or IgG4 antibody levels (relative fluorescence units) in O. viverrini FECT positive subjects (Ov + ), Clonorchis sinensis FECT positive subjects (Cs + ) or CCA patients compared to healthy controls (both endemic FECT-negative and US donors). Significance values < 0.05 are denoted by colour, scaling from least significant (yellow) to most significant (purple). “X” denotes p > 0.05 or where the healthy control signal was higher than the disease groups.

Fig. 3. Immunoreactivity of Opisthorchis viverrini recombinant proteins printed on immunochromatography tests (ICTs) for the serodiagnosis of opisthorchiasis, clonorchiasis and opisthorchiasis-associated cholangiocarcinoma (CCA).

A positive ICT reaction based upon visual assessment was afforded to a protein if a score of equal to or greater than 0.5–1 was attained; each protein has a defined cut-off as described in Supplementary Table 3. Digital scorecard (left) and ICT strips (right) with scores ranging from 0 to 4 are shown (a). C = Control band, T = test band. The IgG (b) and IgG4 (c) positivity rates (%) are shown for the 5 candidate antigens with successful ICT formulations. Ov-: 50 healthy uninfected controls from Thailand (endemic and non-endemic); Ov + : 114 O. viverrini infected subjects with known EPG (Thailand and Laos); Cs + : 50 Clonorchis sinensis infected subjects; CCA: 50 CCA patients where the causative agent was suspected to be O. viverrini infection.

Fig. 4. Association between Opisthorchis viverrini infection intensity and positivity rate of point-of-care immunochromatographic tests (PoC-ICTs) tests for P1 and P9 antigens.

P1-IgG (a), and P9-IgG4 (b) ICT reactivity scores of 114 O. viverrini infected subjects ranked by eggs per gram of feces (EPG) intensity, and endemic (Thailand) and non-endemic (Australia and Thailand) uninfected controls. Uninfected control totals vary as not all subjects were tested due to sample limitations. The median values are denoted by thick colored lines and quartiles as black dashed lines. EPG intensity was categorized as follows: Very low, 0-100 EPG; Low, 100-1000 EPG; Medium, 1000-10,000 EPG; High, >10,000 EPG. Positivity cut-off is marked with a dotted line. Two-sided Pearson correlation coefficient for P1-IgG-ICT (c) and P9-IgG4-ICT (d) scores and fecal egg counts: *** = p < 0.001. Linear regression shown by dark green line with 95% confident intervals (CI) shown as shaded light green. Positive threshold cutoff marked with dotted horizontal line. Vertical dotted lines mark low/medium/high EPG intensities. EPG values of 0 were given nominal values of 0.05–0.09 to enable plotting on log EPG axis; these values are marked by the grey shaded region. Comparison of P1-IgG, P9-IgG4 and ES-IgG4 ICTs for diagnosing O. viverrini infections (e) with Receiver-operator characteristic (ROC) curve and output table showing the superiority of the combined P1-IgG and P9-IgG4 immunochromatographic tests irrespective of infection intensity. The diagnostic accuracy of P1, P9 and P1/P9 to detect antibodies in the sera of O. viverrini infected subjects was measured by the area under the ROC curve, and the likelihood ratio for a positive diagnosis using P1/P9 was determined.

Fig. 5. Positivity rate of point-of-care immunochromatographic tests (PoC-ICTs) tests for P1, P9 and P1/P9 combined antigens for the serodiagnosis of opisthorchiasis-associated cholangiocarcinoma (CCA).

Comparison of PoC-ICT scores for serum IgG (a) and IgG4 (b) responses of Opisthorchis viverrini-associated CCA patients compared to endemic uninfected (Thailand) and non-endemic (Australia and Thailand) serum donors to P1, P9 and O. viverrini ES products. The median values are denoted by thick grey lines and quartiles as black dashed lines. Positivity cut-offs are marked with solid black lines. Receiver operating characteristic (ROC) curves and output table of IgG and IgG4 reactivity to ES products, P1, P9 and combined P1/P9 (c). The diagnostic accuracy of P1, P9 and P1/P9 to detect antibodies in the sera of CCA patients was measured by the area under the ROC curve, and the likelihood ratio for a positive diagnosis using P1/P9 was determined. The heat map illustrates the data from (a) and (b), with positive responses denoted by the red and negative responses denoted by white cells (d). Black cell = not tested. Diagnostic accuracy was also assessed based on different clinical stages of CCA, stages 0-2, 3 and 4 (e). Violin plots show median values as thick red bars with dashed line quartiles. Solid black lines denote positivity cutoffs. Positive samples/total samples and percent positive rates are reported above each plot. Cancer stages 0–2 are grouped together due to low sample numbers for each stage.