Fig 1.
Comparison of the virulence potential of isolated E. histolytica clinical strains using an animal model.
(A) Colonoscopy image of asymptomatic chronic infection. Multiple erosions localized in the cecum were identified (yellow arrows). (B) CT finding of multiple amebic liver abscesses (yellow arrows). E. histolytica clinical strain Ax19 was isolated from the aspirated pus from the abscess. (C) Experimental amebic liver abscess produced by injection of a highly virulent strain (Ax19) into Syrian hamsters. The size of the liver abscess increased according to the injected dose of Ax19. NC, negative control. (D) Each challenged hamster was euthanized 7 days after E. histolytica injection. E. histolytica infection was defined as a positive result following in vitro culture of pieces of the resected liver (* p-value < 0.05). (E) Proportion of liver abscess to whole liver in weight for the highly virulent strain (Ax19). The proportion was positively correlated with the dose of Ax19 injected in 50 μl for an average 60 g hamster. Error bars show the standard error of the mean. Statistical significance is indicated.
Fig 2.
Transcriptome profiling of RNA-seq reads in the isolated clinical strains.
(A) Workflow for the RNA-seq experimental procedure. The E. histolytica clinical strains were isolated from each clinical specimen. First, they were incubated under xenic conditions with E. coli and rice starch in Robinson’s medium for several weeks to reduce human gut bacteria gradually. After adaptation to the xenic culture, the parasites were next transferred to monoxenic culture medium with Crithidia fasciculata. Finally, the parasites were maintained in axenic culture without any bacteria. Animal experiments involved injecting axenically-cultured clinical strains to assess the parasite’s virulence in terms of liver abscess formation. Liver abscesses were successfully formed only when Ax19 (liver abscess strain) was injected. Total RNA was extracted from the trophozoites of the in vitro-cultured strains (Ax11, Ax19, and Ax22) and the animal-passaged strain (ALA19). (B) Two-dimensional (2D) plot showing principal components analysis (PCA) of the RNA-seq reads. Each data point represents a read, with the three isolated clinical strains being analyzed in triplicate. The expression pattern of the Ax19 strain is distinct from those of other two strains. (C) Hierarchical clustering based on the Spearman rank-based clustering of the three strains. (D) Heat map showing the clustering of the three strains based on the Pearson’s correlation coefficients using expression values.
Table 1.
Genotypes of the three isolated clinical strains as determined using transfer RNA-linked short tandem repeats.
Fig 3.
Systematic comparison and assessment of differentially expressed genes (DEGs) between the highly virulent strain (Ax19) and the strains of low virulence (Ax11 and Ax22).
(A–D) Volcano plot showing each DEG among the E. histolytica clinical strains. The vertical axis (y-axis) corresponds to the level of significance of each gene value at log 10 (p-value), and the horizontal axis (x-axis) displays the log 2-fold change value. The red dots represent the DEGs; the black dots represent the non-DEGs. Dotted lines indicate cutoffs; fold changes greater than 2 or less than 0.5; p-value < 0.05. The greatest number of DEGs (1,979 genes) was identified between Ax19 and Ax11. (E) Results of DEG comparisons to select E. histolytica clinical strain- and environment-specific DEGs. To investigate E. histolytica clinical strain-specific DEGs further, we selected 180 DEGs from the 6,225 E. histolytica genes using the Benjamini and Hochberg method with a false discovery rate (FDR) of 5%, followed by Tukey’s multiple comparison test, for three clinical strains. Of the 180 DEGs, each strain-specific DEG was identified as an up-regulated (top) and or down-regulated (bottom) gene using the multiple comparison method. We also selected 85 DEGs by pairwise comparisons between the in vitro-cultured Ax19 strain and the liver-passaged ALA19 strain to detect Ax19 environment-specific DEGs. (F) Gene ontology (GO) functional classification. Using PANTHER tools to analyze the biological functions of the 91 Ax19 strain-specific DEGs, we identified 17 GOs in first level categories, including one GO in cellular components (red bar), eight GOs in molecular function (yellow bars), and eight GOs in biological processes (blue bars). (G) Gene ontology (GO) functional analysis using the PANTHER tool for the 85 Ax19 environment-specific DEGs. Unlike the functional analysis of Ax19 strain-specific DEGs, only two GOs in molecular function were detected for the Ax19 environment-specific DEGs, with no statistical enrichments in biological processes and cellular components.
Table 2.
Orthologous lists of the 26 differentially expressed genes that were oppositely up- or down-regulated between Ax19 (liver abscess strain) and Ax11 (asymptomatic strain).
Fig 4.
Clustering of 15 multi-functional genes among the 26 DEGs that are inversely up- or down-regulated between strains Ax19 and Ax11.
(A) Biological process. (B) Molecular function. (C) Cellular component. (D) Protein Class. *EHI_001420 gene was identified in the gene lists for five overlapping genes between the 91 Ax19 strain-specific DEGs and the 85 Ax19 environment-specific DEGs (S9 Data).
Fig 5.
Transcriptome profiling of the E. histolytica laboratory strain (HM-1:IMSS clone 6).
(A) Experimental work flow for the three different culture conditions: HM-1 (in vitro), maintained under in vitro culture conditions for many years; HM-1 (virulent), maintained under virulent conditions by routinely passaging through an animal liver every 3 months; HM-1 (liver), collected just after liver challenge with strain HM-1. (B) The proportion of liver abscess weight to whole liver weight for strain HM-1 (virulent). The proportion was positively correlated with the dose of HM-1 (virulent), but the linear relationship of HM-1 (virulent) was relatively weak compared with that of the highly virulent Ax19 strain. (C) 2D plot of principle component analysis (PCA) of RNA-seq reads from the three different culture conditions for strain HM-1, using the two different culture conditions for the highly virulent Ax19 strain as reference gene profiles. Each data point represents a read, analyzed in triplicate. The expression profile of HM-1 (in vitro) was clearly different from that of HM-1 (virulent). The difference was greater than the changes induced by liver challenge (HM-1 (virulent) vs HM-1 (liver)). Moreover, the expression profile of HM-1 (virulent) was also distinct from that of ALA19. (D) Heat map of the three different culture conditions for the HM-1 strain. The expression profile of HM-1 (in vitro) was clearly different from those of HM-1 (virulent) and HM-1 (liver). (E and F) Volcano plots showing each differentially expressed gene (DEG) for the E. histolytica laboratory strain. Although there were 6,309 DEGs identified between HM-1 (in vitro) and HM-1 (virulent), there were no more than 565 DEGs identified between HM-1 (virulent) and HM-1 (liver). (G) Venn diagram of the 85 Ax19 environment-specific DEGs and the 565 HM-1 environment-specific DEGs. The number of overlapping genes was 21. (H) GO function analysis of the 565 HM-1 environment-specific DEGs.
Table 3.
Orthologous lists of the nine genes that were similarly up- or down-regulated among the environment-specific DEGs of strains Ax19 and HM-1 (virulent).