This study was approved by the National Ethics Committee for Health Research, Ministry of Health, Lao People’s Democratic Republic (PDR) and the Ethics Review Board of the University of the Ryukyus, Japan. Informed consent was obtained from each participant in the study. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later revision. Identifying information of patients of human subjects, including names, initials, addresses, or any other data that might identify patients do not be included in written descriptions in this article. Informed consent from minors was obtained from their parent before sample collection.

This cross-sectional survey was conducted at the end of each rainy season from 2005 to 2008 in villages of the Phouvong District of Attapeu Province, an area with high malaria endemicity in Lao PDR¹. The annual incidence of malaria in 2008 in this province was 14.3 cases per 1,000 people, which is the second highest incidence of malaria in the country (Jorgensen et al., 2010). Non-endemic controls (NECs) were recruited voluntarily from the population in Vientiane, the capital of Lao PDR, and Japanese healthy controls (JHCs) were recruited from volunteers in Okinawa Prefecture, Japan.

Falciparum malaria was diagnosed at the primary schools in the villages or the village head’s house. All subjects were assessed with a rapid immunochromatographic test (ICT; Paracheck Pf^®, Orchid Biomedical Laboratories, Goa, India) and a blood smear analysis. After the diagnostic tests were performed, heparinized blood samples (4 mL) were collected from the P. falciparum-positive subjects and some P. falciparum-negative volunteers (NCs) as controls. All P. falciparum-positive subjects with uncomplicated falciparum malaria, lacking signs of organ compromise or other severe symptoms, were classified as non-hospitalized uncomplicated malaria patients (UMPs), according to the World Health Organization guidelines (World Health Organization, 2006). Parasitemia and species identity were assessed by microscopic observation of the blood smears. The data from P. vivax-infected subjects and false-positive (ICT positive but microscopic observation-negative) subjects were excluded from the analysis. P. falciparum-positive subjects were treated with Coartem^® (Novartis Pharma, Basel, Switzerland), an artemisinin-based combination therapy. After the number of white blood cells was counted, the plasma was separated from the blood by centrifugation. PBMCs were obtained by Ficoll–Paque (GE Healthcare Life Sciences, Uppsala, Sweden) gradient centrifugation. The plasma and PBMCs from P. falciparum-positive and -negative subjects were initially stored in liquid nitrogen (-196°C) and at -80°C until analysis.

Enzyme-linked immunosorbent assays (ELISAs) of anti-P. falciparum IgG antibodies (Abs) were performed as below. Each well of a 96-well plate was coated with 0.5 μg of crude P. falciparum FCR-3 antigen extract (Pf Ag) in 100 μL of 50 mM Na₂CO₃ buffer, a kind gift from Dr. S. Nakazawa (Department of Protozoology, Institute of Tropical Medicine, NEKKEN, and the Global COE Program, Nagasaki University, Nagasaki, Japan). The plates were incubated overnight at 4°C to allow antigen binding, after which the wells were washed four times with phosphate-buffered saline (PBS) containing 0.05% Tween 20 (PBS/T) and then blocked for 2 h at 37°C with 150 μL of PBS/T containing 1% bovine serum albumin (BSA; Nacalai Tesque, Kyoto, Japan) per well. Test sera (100 μL aliquots), each diluted 1:400 in PBS/T containing 1% BSA, were added in duplicate, and the plates were incubated for 1 h at room temperature. The wells were washed four times with PBS/T and then incubated for 30 min at room temperature with 100 μL of horseradish-peroxidase-conjugated goat anti-human IgG (Promega, Madison, WI, United States) diluted 1:2500 in PBS/T. The wells were washed four times with PBS/T, 100 μL of tetramethylbenzidine substrate (TMB One Solution; Promega) was added, and the plates were incubated at room temperature for up to 15 min before the reaction was stopped with 100 μL of 0.5 M H₂SO₄. The optical density (OD) of the wells at 450 nm (OD₄₅₀) was determined with a microplate reader. To generate standardized units of specific IgG reactivity, a standard curve was generated to allow the OD units to be interpolated to linearized units of antibody concentration. If a sample OD₄₅₀ exceeded 2.0, the sample was diluted further as soon as possible to achieve an OD₄₅₀ within the range of 0.0–2.0.

Frozen PBMCs were thawed quickly for analysis. The surface phenotypes of the lymphocytes were determined, and the intracellular cytokine measurements were made with three- or four-color immunofluorescence tests. Fluorescein isothiocyanate (FITC), phycoerythrin (PE), or biotin-conjugated monoclonal antibodies (mAbs) were used, and biotin-conjugated reagents were developed using allophycocyanin-conjugated streptavidin (Caltag Laboratories, Burlingame, CA, United States). Anti-CD3 (145-2C11), anti-Vγ9 TCR (TM-β1), anti-αβ TCR (H57-597), anti-γδ TCR (GL3), anti-IL-10, and anti-interferon γ (IFN-γ) mAbs (BD Biosciences, Mountain View, CA, United States) and anti-Vδ1 TCR (R9.12) mAb (Beckman Coulter, Marseille, France) were used. The cells were examined with a FACSCalibur flow cytometer (BD Biosciences). Dead cells were excluded with forward scatter, side scatter, and propidium iodide gating.

Total RNA was extracted from the PBMCs of UMPs and NECs with the RNeasy Plus Mini Kit (Qiagen GmbH, Hilden, Germany). To detect the TCRγ and δ chain mRNAs, the cDNA was synthesized from 2.5 μg of RNA with random primers and Superscript VILO reverse transcriptase (Invitrogen, Carlsbad, CA, United States). The cDNA was amplified with sense primers specific for TCR Vγ or Vδ and antisense primers specific for TCR Cγ or Cδ, as previously described (Kageyama et al., 1994; Boria et al., 2008). The PCR products were visualized under UV illumination on 2% agarose gels stained with ethidium bromide.

Thawed PBMCs (1–2 × 10⁶ cells/mL) were cultured with or without interleukin 2 (IL-2, 50 U/mL) or IL-2 and Pf Ag (5 μg/mL) for 10 days at 37°C under 5% CO₂. The cultured PBMCs were restimulated with Leukocyte Activation Cocktail (BD Biosciences) containing the phorbol ester phorbol 12-myristate 13-acetate, a calcium ionophore (ionomycin), and brefeldin A for 4 h at 37°C under 5% CO₂. The cells were then fixed, permeabilized, stained with 20 μL of PE-conjugated rat anti-human IL-10 antibody or mouse anti-human IFN-γ antibody (BD Biosciences), and analyzed with a FACSCalibur flow cytometer (BD Biosciences).

The cytokines in the plasma and culture supernatants were measured using a BD^TM Cytometric Bead Array Flex Set and BD^TM Human Soluble Protein Master Buffer kit according to the manufacturer’s instructions (BD Biosciences). Aliquots (50 μL) of the test sera, diluted 1:5 with dilution buffer, were used. IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, tumor necrosis factor α, and IFN-γ were acquired on a FACSCalibur flow cytometer (BD Biosciences) and analyzed using FCAP Array software (BD Biosciences).

The study subjects were matched for age and sex. The Mann–Whitney U-test was used for two-group comparisons, the Kruskal–Wallis test was used for three-group comparisons, and paired and unpaired t-tests were used to analyze whole PBMC cultures. Spearman’s rank correlation was used for correlational analyses. All analyses were performed with Prism version 6.0 (GraphPad Software, La Jolla, CA, United States) and JMP version 8 statistical software (SAS, Cary, NC, United States).

Despite many reports of γδ T cells in malaria patients, it is unclear whether γδ T cells are induced during malarial infection (Ho et al., 1990; Roussilhon et al., 1990; Goerlich et al., 1991; Currier et al., 1992; Goodier et al., 1992; Zevering et al., 1992; Dick et al., 1996; Hviid et al., 1996, 2001; Rzepczyk et al., 1996; Goodier and Targett, 1997; Worku et al., 1997; Waterfall et al., 1998). To investigate this phenomenon in more detail, PBMCs and plasma obtained from UMPs and NCs living in malaria-endemic areas of Attapeu Province, Lao PDR and from NECs in Vientiane, the country’s capital, were analyzed during physical examination of the subjects. The hematological characteristics of the UMPs, NCs, and NECs in this study are summarized in Table 1 and Supplementary Figure S1. In agreement with the high incidence of malaria (Jorgensen et al., 2010), all subjects living in the Phouvong District of Attapeu Province had anti-P. falciparum IgG Abs in their plasma, and the subjects in the NC group had a previous history of natural malaria infection. The anti-P. falciparum IgG Ab titers were higher in the UMPs (median 827.9, range 86.9–8685.6) than those in the NCs (median 462.1, range 7.0–10,097.3; p < 0.0001). The UMPs living in the malaria-endemic area included both symptomatic (body temperature ≥ 37.5°C) and asymptomatic falciparum malaria patients (parasitemia < 250,000/μL, hematocrit > 15%) (World Health Organization, 2006).

The lymphocytes in the PBMCs were phenotypically characterized by using flow cytometry. As in other acute infections, acute falciparum malaria causes a decrease in lymphocyte counts (Supplementary Table S1 and Figure S2). A decrease in lymphocyte numbers in UMPs inhabiting the malaria-endemic area and a significant increase in white blood cell numbers in NCs were observed (Table 1 and Supplementary Figure S3); therefore, the percentage of cell populations was used for comparisons. The change in lymphocyte cell subsets was minimal, and the percentage of γδ T cells (UMPs: median 5.5%, range 1.2–13.9%; NCs: median 5.1%, range 3.5–13.4%, p = 0.6208) and αβ T cells (UMPs: 67.9%, 50.8–79.2%; NCs: 63.4%, 47.2–75.3%, p = 0.1917) was not higher in the UMPs than in the NCs (Figure 1 and Supplementary Figure S4). However, a specific γδ T cell subset, the non-Vγ9 γδ T cells (UMPs: 2.9%, 0.2–10.6%; NCs: 1.7%, 0.3–3.3%, p = 0.0018), was expanded in the UMPs (Figure 1B), and the proportion of non-Vγ9 γδ T cells within the γδ T cell subset (UMPs: 58.1%, 10.7–92.7%; NCs: 33.7%, 8.8–82.5%, p = 0.0025) was also higher in the UMPs (Figure 1C). These results differed from those of hospitalized patients with severe falciparum malaria, in whom the Vγ9 γδ T cell subset increased (Supplementary Table S1 and Figures S5A,B). Although there was no change in B cell subsets between UMPs and NCs, we found a significant increase in B cells and memory B cells (IgG+CD27-CD20+) in UMPs compared with those in NECs (Supplementary Figure S6).

γδ T cells with non-TCR Vγ9 chains were increased in the UMPs, and the other major subset in the peripheral blood had the Vδ1 chain; however, the predominant γδ T cell subset was Vγ9Vδ2. Therefore, using flow cytometry, we examined whether these non-Vγ9 γδ T cells had Vδ1 chains. The γδ T cell subsets that were elevated in the UMPs living with endemic malaria were Vγ9^-Vδ1⁺ (UMPs: 21.4%, 2.5–61.4%; NCs: 12.6%, 2.2–35.8%) and Vγ9^-Vδ1^- (UMPs: 25.3%, 3.9–53.9%; NCs: 9.5%, 2.2–45.8%). In contrast, Vγ9⁺Vδ1^- cells were the predominant subset in the NCs, but there was no significant difference in γδ TCR usage by γδ T cells between NCs and UMPs, as determined by flow cytometry (Figure 2A). Therefore, we performed further analysis of γδ TCR usage by γδ T cells from UMPs compared with that in NECs by using RT-PCR (Figure 2B). We found that Vγ and Vδ were detected at the same level in the UMPs and NECs, and the Vγ9 levels tended to be lower in the UMPs than in the NECs, whereas the levels of Vγ2/Vγ4, Vγ3/Vγ5, Vγ8, Vγ11, Vδ1 and Vδ3 tended to be higher in the UMPs. Thus, γδ TCR repertoire in the non-Vγ9 γδ T cells was highly polyclonal in the UMPs living in the malaria-endemic area, although the Vγ9Vδ2 T cell subset was predominant in the NCs and NECs.

The role of γδ T cells in malaria infection remains unclear. However, it has been reported that these cells contribute to host defense and disease pathology during infections (Stevenson and Riley, 2004; Schofield and Grau, 2005). To investigate the function of non-Vγ9 γδ T cells after P. falciparum infection, we cultured whole PBMCs from falciparum malaria patients in the presence of IL-2 and crude Pf Ag for 10 days. The cytokine concentrations were measured in the culture supernatants because it is difficult to culture non-Vγ9 γδ T cells; these cells constitute a very small population of cells in the peripheral blood, and the specific antigen recognized by the TCR remains unknown. Surprisingly, non-Vγ9 γδ T cells from NCs and UMPs in the malaria-endemic area increased 10-fold in the presence of IL-2 or IL-2 and Pf Ag after 10 days of culture (Figure 3A). Moreover, high levels of IL-10 were detected in the culture supernatants of PBMCs from the malaria-endemic groups (Figure 3B). The highest IL-10 production was observed in the supernatants of PBMCs from UMPs in the presence of IL-2 and/or Pf Ag. Interestingly, high concentrations of IFN-γ were also detected in those supernatants.

To determine whether non-Vγ9 γδ T cells produce IL-10 in patients with falciparum malaria, we used intracellular staining for IL-10 and confirmed that IL-10 was indeed produced in non-Vγ9 γδ T cells after whole PBMCs were cultured for 10 days. We found that non-Vγ9 γδ T cells still produced IL-10, and higher levels of IL-10-producing cells were present in the PBMCs from UMPs than in those from JHCs, although many other IL-10-producing cells were present (Figures 3C,D). It has been reported that IFN-γ- or IL-17-producing Vγ9 γδT cells can be identified by using the cell surface marker CD27 (Ribot et al., 2009, 2010; Ribot and Silva-Santos, 2013). Approximately 80% of Vγ9 γδ T cells in the present study were CD27-positive, but the classification of non-Vγ9 γδ T cells by CD27 expression was difficult (data not shown).

We found that the non-Vγ9 γδ T cells from patients living with recurrent malarial infections responded to IL-2 and/or crude Pf Ag, and these cells proliferated and produced both of IFN-γ and IL-10. Both of these are important cytokines for protecting the host against malarial infection. IFN-γ production by non-Vγ9 γδ T cells was also found in the JHCs with no history of malaria, although no proliferation of non-Vγ9 γδT cells or IL-10 production was observed.

To confirm whether non-Vγ9 γδ T cells correlate with levels of cytokines including IL-10 and IFN-γ in peripheral blood, we examined the plasma levels of cytokines in the infected patients living in endemic areas. The plasma levels of IL-10 were significantly higher in UMPs than in NCs (UMPs: median 28.3 pg/mL, range 11.0–664.8 pg/mL; NCs: 21.9 pg/mL, 14.5–44.5 pg/mL, p = 0.0079; Figure 4A). None of the other cytokines (IL-1β, IL-2, IL-4, IL-6, IL-8, IL-12p70, tumor necrosis factor α, and IFN-γ) were elevated in UMPs living in endemic areas (Supplementary Table S2). These results were consistent with the mild symptoms of falciparum malaria in patients living in endemic areas. When we examined the relationship between plasma levels of cytokines and non-Vγ9 γδ T cells proportion, we found that the plasma levels of IL-10 correlated with non-Vγ9 γδ T cell proportion (r = 0.3475, n = 54, p = 0.0100; Figure 4B) but that of IFN-γ did not. These data suggested that non-Vγ9 γδ T cells may be the source of IL-10 in UMPs living in endemic areas during infection.