Use of bacterially expressed dsRNA to downregulate Entamoeba histolytica gene expression

Abstract

Background: Modern RNA interference (RNAi) methodologies using small interfering RNA (siRNA) oligonucleotide duplexes or episomally synthesized hairpin RNA are valuable tools for the analysis of gene function in the protozoan parasite Entamoeba histolytica. However, these approaches still require time-consuming procedures including transfection and drug selection, or costly synthetic molecules.

Principal findings: Here we report an efficient and handy alternative for E. histolytica gene down-regulation mediated by bacterial double-stranded RNA (dsRNA) targeting parasite genes. The Escherichia coli strain HT115 which is unable to degrade dsRNA, was genetically engineered to produce high quantities of long dsRNA segments targeting the genes that encode E. histolytica beta-tubulin and virulence factor KERP1. Trophozoites cultured in vitro were directly fed with dsRNA-expressing bacteria or soaked with purified dsRNA. Both dsRNA delivery methods resulted in significant reduction of protein expression. In vitro host cell-parasite assays showed that efficient downregulation of kerp1 gene expression mediated by bacterial dsRNA resulted in significant reduction of parasite adhesion and lytic capabilities, thus supporting a major role for KERP1 in the pathogenic process. Furthermore, treatment of trophozoites cultured in microtiter plates, with a repertoire of eighty-five distinct bacterial dsRNA segments targeting E. histolytica genes with unknown function, led to the identification of three genes potentially involved in the growth of the parasite.

Conclusions: Our results showed that the use of bacterial dsRNA is a powerful method for the study of gene function in E. histolytica. This dsRNA delivery method is also technically suitable for the study of a large number of genes, thus opening interesting perspectives for the identification of novel drug and vaccine targets.

Figure 1. Genetic interference following ingestion of dsRNA-expressing bacteria by Entamoeba histolytica.

A. General scheme for dsRNA production. The 5′-terminal portions of the genes were amplified by PCR and cloned into the bacterial vector L4440 for dsRNA transcription. Primers used for quantification of mRNA by qRT-PCR match with the 3′-terminal part of the gene, which was not cloned into the plasmid (marked “Q-PCR template” in the figure), allowing the exclusive quantification of genome-encoded, sense-orientated transcripts. The multicloning site of L4440 vector is bidirectionally flanked by T7 promoters driving the synthesis of RNA complementary strands (i.e. dsRNA). The recombinant plasmid was transfected into the bacterium HT115, allowing dsRNA purification in order to conduct the parasite soaking experiments or direct feeding of parasites with the expressing bacteria. B. Constructs performed in this work. Three independent DNA fragments were cloned and used for RNAi experiments including β-tubulin-, KERP1- and GFP-encoding genes (upper panel). Upon production of dsRNA, a nuclease-resistant dsRNA was detected in lysates of the recombinant bacterium (bottom panel).

Figure 2. Targeting β-tubulin-encoding gene by dsRNA expressed in bacteria.

Following 3, 5 and 7 days of culture of 3×10⁵ E. histolytica trophozoites in the presence of bacteria expressing dsRNA from genes encoding GFP (control) or β-tubulin (test), the parasites were counted under the microscope(A). At day 3 post-inoculation, endogenous mRNA was quantified by qRT-PCR (B) and protein levels were observed by western blot (C) according to procedures described in the materials and methods. Each graph represents the mean of 3 independent experiments.

Figure 3. Phenotype of trophozoites fed with bacteria producing kerp1 dsRNA or soaked in medium containing kerp1 dsRNA.

Experiments were performed with trophozoites incubated without (wild type, “WT”), with bacteria producing gfp or kerp1 dsRNAs or with purified dsRNA (50 µg/ml) for the indicated times. A and C. Upper frames: Western blot analysis of KERP1 protein abundance after 2-day incubation. Actin immunodetection confirms that equal numbers of amoebae were loaded in each well. Lower frames: Quantification of kerp1 gene mRNA expression and protein abundance within trophozoites after feeding for 2 days. Quantitative PCRs were normalized by quantifying the mRNA coding for L9 ribosomal protein. Western blots were normalized using immunodetection of actin. For both methods, relative quantification of the samples was achieved by linear regression using dilutions of samples as standards. Data were calculated from 3 independent experiments. B and D. Analysis of parasite growth. Tubes were inoculated with 2×10⁴ trophozoites and 2×10⁷ bacteria, when applicable. After 1 to 3 days of interaction, amoebae were harvested and counted with a hemocytometer. Data were calculated from 3 independent experiments (n = 3).

Figure 4. Study of KERP1 involvement in trophozoite interaction with LSECs.

A. Assay for adherence of modified trophozoites to LSECs. Quantification of trophozoites bound to LSECs after 30 min contact showed 50% decrease in adherence of KERP1-reduced trophozoites. Data were calculated from 3 independent experiments (n = 3). B. Assay for cytopathogenicity of modified trophozoites on LSECs. Quantification of fluorescence emitted by LSECs after 1 hr incubation with trophozoites showed a decreased cytopathogenicity of KERP1-reduced trophozoites leading to 24% reduction in fluorescence. Data obtained from 3 independent experiments (n = 3).

Figure 5. Selection of candidate genes for potential role in growth defects in E. histolytica by feeding experiments.

Quantification of trophozoite growth was performed at 24 and 48 hrs during the feeding experiment with the C3, D7, G1, G5 and G7, previously selected from a visual screening of trophozoites soaked with 85 target genes. Results are the mean of 3 independent experiments (n = 3).