Trophozoites of G. lamblia GS strain (ATCC #50581, ATCC, Manassas, Virginia, USA) were grown under axenic conditions in a TYI-S-33 medium (2% casein digest, 1% yeast extract, 1% glucose, 0.2% NaCl, 0.2% L-cysteine, 0.02% ascorbic acid, 0.2% K₂HPO₄, 0.06% KH₂PO₄, 10% calf serum, and 0.5 mg/ml bovine bile, pH 7.1) [13] at 37°C.

C57BL/6 mice (6 week-old) were purchased from OrientBio Co. (Seongnam, Korea), and maintained at the animal facility of Yonsei University College of Medicine under SPF conditions. All experiments were approved by the Yonsei University College of Medicine Ethics Committee (IACUC 2017-0015).

Mouse ILCs were prepared as described [14]. Briefly, the small intestines were extracted from C57BL/6 mice. After removal of the surrounding mesentery and Peyer’s patches, and fecal matters, the small intestines were minced into longitudinal pieces of 7 cm, and further cut into 1 cm pieces, followed by vigorous washes in phosphate buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 10.1 mM Na₂HPO₄, and 2 mM KH₂PO₄, pH 7.4). They were further treated with 1 mM EDTA/PBS at 37°C for 20 min for removal of the intraepithelial lymphocyte and intestinal epithelial cell fractions. Thereafter, the samples were harvested by centrifuging for 20 min at 500×g, followed by vigorous washing for 20 sec in PBS. Further, they were minced into 1–2 mm pieces, and incubated in 10 ml of pre-warmed RPMI medium supplemented with 2% fetal bovine serum containing 20 μg/ml Liberase DH (Roche, Basel, Switzerland) and 50 μg/ml DNase I (Roche) at 37°C for 30 min. The cells were subsequently passed through a 70 μm EASY strainer (BD Biosciences, San Jose, California, USA), and a 37 μm nylon filter (BD Biosciences). The flow-through cells were harvested by centrifuging for 5 min at 500×g, 4°C, and then used as an ILC fraction.

ILCs (1×10⁶ cells per well) were seeded in a 24 well culture plate with transwell membranes (Merck, USA) and incubated for 6 hr for RNA preparation and 16 hr for cytokine determination, with G. lamblia having multiplicity of infection (MOI) of 5. The culture media were collected for cytokine determination, and the ILCs were harvested for RNA preparation.

RNAs were prepared from 4 independent ILC preparations, incubated with G. lamblia, and 4 control ILCs (ILCs only) using TRIzol (Invitrogen Life Technologies, Carlsbad, California, USA), and then treated with the RNase-free DNase (Qiagen, Valencia, California, USA) according to the manufacturer’s instructions. These RNA samples were submitted to Macrogen (Seoul, Korea) for RNA sequencing by HiSeq4000 (Illumina Inc., San Diego, California, USA). The resulting sequences were mapped against the genomic DNA reference (UCSC mm10) using HISAT2 program (Johns Hopkins University, USA). After alignment, the relative abundance of the transcripts was measured with the String Tie software (Johns Hopkins University), which measures abundance of each transcript as FPKM (Fragments Per Kilobase of transcript per Million mapped reads). The 4 samples for each group (ILCs alone or ILC incubated with G. lamblia) were analyzed as a cluster to monitor similarity within the group. The numbers of clusters were independently assessed using Hierachical clustering analysis (Euclidean method, Complete Linkage) to construct a heat map. Differentially expressed transcripts were annotated and automatically categorized using KEGG pathway analysis program (http://www.kegg.jp/kegg/pathway.html). The generated in the RNA-seq is available at the Gene Expression Omnibus (GEO) database with the accession number GSE130256.

Culture media from the ILCs incubated with G. lamblia, were used for detection of IL-17A, IL-17F, IL-22, IFN-γ, IL-1β, IL-5, and IL-13, by a sandwich ELISA to detecting IL-17A, as recommended by the manufacturer (BD Biosciences). The culture media collected from the ILCs, not exposed to G. lamblia, were used as controls for the assay.

cDNA was synthesized from 4 μg of RNA using Prime-Script RT reagent kit (TaKaRa, Otsu, Japan) as directed by the manufacturer. The cDNAs were analyzed by qRT-PCR on a Light Cycler 480 II real-time PCR system (Roche Applied Science, Mannheim, Germany) using Light Cycler 480 DNA SYBR green I master kit (Roche Applied Science). The amount of each transcript was estimated using specific pair of primers; GAPDH, IL-17A, IL-17F, IL-22, IFN-γ, IL-1β, IL-5, and IL-13. The primers were made as suggested [15]. GAPDH gene was used as an internal control to normalize each transcript. Data analysis was based on the relative quantification method by determining the crossing point (Cp) value using the Light Cycler 480 II real-time PCR system software program (Roche Applied Science, version LSC480 1.5.0.39).

The collected ILCs were analyzed using the live/dead fixable dead cell stain kit (Invitrogen, Carlsbad, California, USA) to determine the percentage of dead (stained) cells. The process was followed by manufacturer’s instruction. Briefly, cells were pre-treated with anti-CD16/CD32 antibodies, and then cells were incubated with the following antibodies: FITC-conjugated anti-mouse Lineage negative (Lin-), AlexaFluor 700-conjugated anti-CD103, phycoerythrin (PE)-conjugated anti-IL-25R, allophycocyanin (APC)-conjugated anti-mouse I-A/I-E (MHC class II), AlexaFluor 700-conjugated anti-IL-17A or relevant isotype control antibodies in FACS buffer (PBS, 1% BSA, and 0.1% sodium azide). Fluorescence was measured by flow cytometry, and data were analyzed using FlowJo data analysis software (TreeStar, Ashland, Oregon, USA). All antibodies were from eBioscience (San Diego, California, USA) unless stated otherwise.

Each C57BL/6 mouse was infected orally with 5×10⁵ G. lamblia trophozoites. Similarly, control mice were orally inoculated with 0.1 ml PBS, without G. lamblia trophozoites. The mice were sacrificed at various time-points, from 1 to 11 days post-infection (n=10), and their intestinal samples were collected for experimental purpose. Sections of duodenum paraffin blocks were stained with hematoxylin and eosin (H&E) and observed under Axiovert 200 fluorescence microscope (Carl Zeiss, Jena, Germany).

In addition, 10 mice were orally infected with G. lamblia as described, and their small intestines were removed at 5 days post-infection. The same number of control mice were necropsied and pooled as one group at the end of the trial. The small intestines obtained from mice of both groups were used to isolate ILCs as described above. These ILCs were analyzed for specific markers of the ILC group, using flow cytometry.

Results are expressed as means±standard deviations of 3 independent experiments. Data were analyzed by pair-wise comparison using Student’s t-test (SYSTAT program, SigmaPlot version 9; Systat Software Inc., Chicago, Illinois, USA). Differences were considered significant if the P-value was<0.05. Data with P-value<0.01 are indicated with 2 asterisks, and data with P-values between 0.01 and 0.05 are indicated by a single asterisk.

The gene expression profiles were compared between ILCs stimulated with G. lamblia and control ILCs, using RNA-seq. For each group, 4 samples were analyzed, and their similarity within the group was monitored (Fig. 1A). Except one sample of ILC group stimulated with G. lamblia, the one-way Hierarchical Clustering analyses indicated valid tendency in the gene expression profile within the group. By applying strict selection of FPKM, i.e., P-values<0.05 and fold change >2 for all genes, 472 genes were found to be up-regulated and 498 genes were found to be down-regulated in the ILC group stimulated with G. lamblia (Fig. 1B). These genes were categorized 16 groups, according to their putative functions, which were presented as percentages of all the up- and down regulated genes (Fig. 1C). Substantial number of genes showing altered expression encoded hypothetical proteins (20–23%) and metabolic enzymes (9–13%) in both groups (up- and down-regulated clones). Eight to eleven percentage of up- and down-regulated genes were found to be involved in immune response, respectively. In addition, the expression of signaling genes (10–20%) was found altered in the ILC group stimulated with G. lamblia. Interestingly, in this group, more cell cycle-related genes (15%) was found to be down-regulated (15%) rather than up-regulated (3%).

Especially, regulation of signature genes for diverse immune response was found among the modulated genes in these RNA-seq data. As expected, several signaling pathways seemed to be activated, which included chemokine signaling, MAPK signaling, Jak-STAT, PI3K-Akt, the Rap1 signaling pathways. In addition, the data suggested activation of several immune pathways, cytokine-cytokine receptor interaction, intestine immune response for IgA production, Th17 differentiation, Th1/Th2 differentiation, and IL-17 signaling pathway. Among the various pathways found to be activated by G. lamblia, we focused on the signaling pathway for IL-17, which has been earlier reported as a main regulator for immune response against G. lamblia [12].

In this experiment, ILCs were incubated with G. lamblia in plates containing transwell membranes and examined for the secretion of the following cytokines: IFN-γ, IL-1β, IL-5, IL-13, IL-17A, IL-17F, and IL-22 (Fig. 2A). Secretion of IFN-γ, IL-1β, IL-17A, and IL-17F was found to be significantly increased. While there was moderate increase in IFN-γ, IL-1β, and IL-17F (2.5- to 3.5-fold), more than 15-fold increase in IL-17A secretion was observed. Secretion of IL-5, IL-13, and IL-22 was either decreased or not affected by G. lamblia.

These data on cytokine secretion from ILCs stimulated with Giardia, were verified by qRT-PCR (Fig. 2B). Among the 7 cytokine genes, only the IL-17A and IL-17F transcripts showed a significant increase (2.4- and 2.7-fold). Moderate increase in the transcript of IFN-γ, IL-1β, and IL-22 was observed (1.4- to 1.7-fold). Decrease in the levels of IL-5 and IL-13 transcripts was noted.

ILCs were divided into 3 groups: group 1 ILCs, group 2 ILCs, and group 3 ILCs, based on their cytokine production patterns that correspond to the helper T cell subsets, Th1, Th2, and Th17, respectively [12]. To determine which group of ILCs plays a role in G. lamblia-induced immune response, prepared murine ILCs were treated with Giardia, and assayed for expression of ILC markers using flow cytometry (Fig. 3A). In case of control cells, ILCs without G. lamblia stimulation, only a small portion of cells (0.02–0.3%) were stained with ILC1 (CD103) or ILC2 marker (IL-25R). The expression of these markers was increased to 0.2–2.3% when ILCs were stimulated with G. lamblia. Without G. lamblia stimulation, a significant portion of ILCs (25%) were stained with MHCII, an ILC3 marker. Expression of this surface protein was further increased to 53% when ILCs were incubated with G. lamblia. Along with IL-17 production, the FASC analysis indicated that among the 3 ILC groups, ILC3 played the most important role in G. lamblia-induced immune response.

Therefore, increase of ILC3 in G. lamblia-treated cells was demonstrated by FACS analysis using MHCII and IL-17A, markers for ILC3 (Fig. 3B). Co-expression of these 2 proteins was observed in only 15% of the control cells (ILCs untreated with G. lamblia), while in 46% of ILCs stimulated with G. lamblia.

Role of ILC3 in G. lamblia-induced cytokine production was examined in vivo using a mouse model. Since mice are not natural hosts for G. lamblia, infection by G. lamblia trophozoites was monitored for 11 days. Small intestines of the infected mice were extracted and prepared for H&E staining for observation of G. lamblia trophozoites, using microscopy (Fig. 4A). On the first day, few trophozoites were found in the murine intestine. Maximal number of Giardia trophozoites was observed on the 5th day post-infection, and their presence gradually decreased, till no trophozoite was observed on days 10 and 11.

Based on the infection kinetics of mice with G. lamblia, ILCs were prepared from the mice infected with G. lamblia, at 5 days post-infection, and the 3 groups of ILCs were studied using flow cytometry (Fig. 4B). Less than 1% of the cells (0.11%) were stained, either with ILC1 or ILC2 marker, in the control ILCs, isolated from the uninfected mice. ILCs obtained from G. lamblia-infected mice showed 1.9% and 0.2% staining for ILC1 and ILC2, respectively. Corroborating with our in vitro data, more ILCs were stained with the ILC3 marker, both from the uninfected mice (35%) as well as from the G. lamblia-infected mice. However, the percentage of the ILC3-positive cells increased to 66% in the ILCs isolated from mice infected with G. lamblia, indicating a role of ILC3 in G. lamblia-induced immune response, which occurred in intestine of the infected mice.