Induction of virulence factors in Giardia duodenalis independent of host attachment

Abstract

Giardia duodenalis is responsible for the majority of parasitic gastroenteritis in humans worldwide. Host-parasite interaction models in vitro provide insights into disease and virulence and help us to understand pathogenesis. Using HT-29 intestinal epithelial cells (IEC) as a model we have demonstrated that initial sensitisation by host secretions reduces proclivity for trophozoite attachment, while inducing virulence factors. Host soluble factors triggered up-regulation of membrane and secreted proteins, including Tenascins, Cathepsin-B precursor, cystatin, and numerous Variant-specific Surface Proteins (VSPs). By comparison, host-cell attached trophozoites up-regulated intracellular pathways for ubiquitination, reactive oxygen species (ROS) detoxification and production of pyridoxal phosphate (PLP). We reason that these results demonstrate early pathogenesis in Giardia involves two independent host-parasite interactions. Motile trophozoites respond to soluble secreted signals, which deter attachment and induce expression of virulence factors. Trophozoites attached to host cells, in contrast, respond by up-regulating intracellular pathways involved in clearance of ROS, thus anticipating the host defence response.

Figure 1. Experimental design and TMT labelling workflow for the experiment.

(A) Summary of the experiment conditions of the control, host soluble factor and co-incubation treatments. (B) Explanation of the TMT-labelling strategy utilised in the experiments. Peptides from the triplicates of the 3 conditions and a pooled control were labeled with one each of the TMT 10plex reagents. These reagents are observed as a monoisotopic complex in the first round of MS analysis on a high resolution mass spectrometer. During MS/MS and HCD based fragmentation, the TMT labels are fragmented to produce 10 reporter ions with distinguishable masses in the low m/z range, which allow relative protein quantitation (C) Overarching experimental design and workflow. Biological triplicates of BRIS/95/HEPU/2041 were grown to confluence in parasite culture in TYI-S-33 medium, before replicates were split into a replicate of each control for cell culture conditions (Con), co-incubation with IEC monolayers (CI IEC) and incubation in host soluble factors generated by IECs (HSF). Proteins were extracted from the 9 replicates, and a pooled control was generated from equal aliquots of protein from the control triplicates. After proteolytic digestion, samples were labelled in a 10 plex TMT reaction and then pooled. The combined sample was fractionated by SCX chromatography and desalted using a C18 ZipTip prior to LC-MS/MS on a Q-Exactive Orbitrap.

Figure 2. Results of the in vitro host-cell attachment versus non-specific adherence.

(A) Rates of attachment between Giardia trophozoites incubated with HT-29 cells over 6 hours against a control for adherence (T75 flasks with media only). A ‘*’ indicates a significant difference in % attached trophozoites compared to control (designated by p-value ≤ 0.05). (B) Changes in HT-29 cell morphology induced during co-incubation with Giardia trophozoites. The arrows (▲) highlight 2 regions of affected cells throughout the 6 hour co-incubation.

Figure 3. Results for the effects of HSF on adherence and host-cell attachment during co-incubation.

The 3 treatments are as follows: trophozoites in serum-free DMEM in the first round followed by co-incubation with HT-29 (Con/CI), trophozoites exposed to HSF in serum-free DMEM and then co-incubated with HT-29 in the presence of HSF in the second round (HSF/CI) and a control of serum free DMEM in both rounds of the assay (Con/Con). Trophozoites were transferred from the first round of adherence to the second round of co-incubation. (A) Rates for adherence during the first 6 hours between trophozoites co-incubated in T75 flasks containing serum-free DMEM (Con/CI, Con/Con) and trophozoites incubated in the presence of HSF (HSF/CI). A ‘*’ indicates a significant difference in % attached trophozoites (designated by p-value ≤ 0.05). Rates of adherence in HSF-exposed trophozoites were statistically significantly lower at all 3 timepoints compared to unexposed trophozoites. (B) Images depicting the differences in density of adhered trophozoites in HSF-exposed and HSF-free flasks after 6 hours of incubation. Flasks shown were seeded with the same number of trophozoites. (C) Rates of host-cell attachment in the second 6 hours of the assay. Rates of host-cell attachment between trophozoites exposed to HT-29 monolayers (Con/CI) are compared to trophozoites incubated with HT-29 monolayers in the presence of HSF (HSF/CI). A control for adherence was also performed in triplicate in serum DMEM without HT-29 cells. A ‘*’ indicates a significant difference in percentage of attached trophozoites compared to control (designated by p-value ≤ 0.05). Rates of attachment in HSF-exposed cells was significantly lower at all time points between both the control for adherence, and the rate for host-cell attachment in unexposed trophozoites. The number of attached trophozoites co-incubated with HT-29 monolayers without HSF was statistically significantly higher compared to the control for trophozoite adherence in flasks only after 6 hours incubation.

Figure 4. Protein identification and protein quantitation summary from TMT labelling of trophozoites co-incubated with the IEC monolayer (CI IEC) and with host soluble factors alone (HSF).

(A) Outline of protein identification, differentially expressed proteins and protein quantitation FDR for the dataset. (B) Volcano plots illustrating the dual criteria for differentially expressed proteins. The x-axis represents log fold change with the vertical blue lines indicating 1.2 and 0.8 ratio, while the -log p value is plotted on the y-axis with proteins above the red horizontal line indicating significance ≤0.05. Each data point represents a single identified protein. Proteins within the upper and outer quadrants meet both the fold change and p-value cut-off, and are therefore considered as differentially expressed. (C) Proportional venn diagrams showing overlap between up-regulated and down-regulated proteins in trophozoites between CI IEC and HSF treatments.

Figure 5. Figure depicting the biphasic model of interaction between G. duodenalis trophozoites and host-cells proposed in our paper.

Proteins trends, families and pathways induced during host-parasite interactions in Giardia are distinguished between HSF-induced in non-attached, motile trophozoites (above) separate to cascades induced in host-cell attached trophozoites (below). ‘VF’ indicates induced protein groups related to known or putative virulence factors in Giardia, which were induced by host secretions. The middle of the figure shows the two distinct stages observed in early pathogenesis, where host-soluble factors lead to a switch to a non-attaching, motile population phenotype. Motile trophozoites migrate further through the gastrointestinal tract, where in the absence of host soluble factors and more optimum conditions, Giardia attaches to host cells. These two stages of host-parasite interactions induce distinct and independently protein responses.