Immunity to Malaria in Plasmodium vivax Infection: A Study in Central China

Background P. vivax infection is characterised by relapsing fever, indicating reinfection by previously hidden parasites in the host. Relapsed infection can lead to the activation of the memory T cell pool, which may lead to protective immunity. This study aims to characterise immune responses in acute P. vivax-infected patients living in an area of central China characterised by only P. vivax infection. Methodology/Principal Findings We conducted a cross-sectional immune-phenotypic analysis of adults using the following inclusion criteria: acute P. vivax infection (N = 37), a history of P. vivax infection (N = 17), and no known history of P. vivax infection (N = 21). We also conducted a 2-week longitudinal analysis following acute P. vivax infection, in which PBMC proliferation was measured in response to P. vivax and P. falciparum blood stage lysates. Using flow cytometry, we showed elevated memory T cells in the blood during acute P. vivax infection. The levels of γδ T cells were two-fold higher than those measured in naive controls. This result suggested that in the two populations, memory and γδ T cells promptly responded to P. vivax parasites. Interestingly, P. falciparum antigens stimulated T cells obtained from P. vivax-infected patients during a day 14-convalescence, whereas lymphocytes from the naïve control group responded to a lower degree of convalescence. Conclusions/Significance Cell-mediated immunity during the convalescent period of the P. vivax-infected hosts was comprised of T cells that were specifically able to recognise P. falciparum antigens. Although the magnitude of the response was only half that measured after stimulation with P. vivax antigens, the matter of cross-antigenic stimulation is of great interest.


Introduction
Malaria is a common tropical disease associated with mortality among children infected with Plasmodium falciparum, primarily in African countries [1]. In contrast, P. vivax causes relapsing infection and is rarely fatal in most areas to which it is endemic [2]. The underlying factors leading to this opposite outcome between the two malarial infections are still unclear. The unique host-parasite relationship and the immune response to P. vivax have not yet been elucidated. One reason for this lack of understanding is the difficulty in maintaining P. vivax in an in vitro culture [3,4]. Therefore, the supply of P. vivax and progress in the relevant research is limited. Recent incidences of P. vivax infection have increased gradually in tropical countries [5]. Whether this increased incidence indicates slow development of drug resistance by the P. vivax parasite remains to be verified. Several studies have been conducted in the regions in which P. falciparum is present along with P. vivax. These studies have shown that immunity to P. falciparum parasites readily exists in the host [6,7,8,9,10,11,12]. Thus, the majority of studies analysing immunological parameters and malaria severity have been performed in the regions where P. falciparum and P. vivax co-infection is normally found [13,14,15,16,17,18,19]. Recent studies suggest that immune suppression occurs in response to P. vivax infection because of elevated regulatory T cell levels [20,21]. The strain-specific and serological cross-reactive immu-nity between blood stage antigens of P. falciparum and P. vivax has been well-documented [12,22,23]. One report has shown no effect on the cell-mediated response [24]. To verify immunity to malaria in P. vivax infection without an interfering immune response caused by P. falciparum, we chose to focus on central China, where P. vivax is the only cause of malaria infection [25]. In this study, we characterise acquired cell-mediated immunity by following P. vivax infection.

Study Populations
This study was approved by the Ethical Approval Committee of the Biomedical Institute of Anhui Medical University. Written informed consent was obtained from each individual before a blood sample was taken.
Blood samples were collected from 37 patients with acute P. vivax infections (AC) at Wuhe County Hospital, Guzhen County Hospital, and The First Hospital of Bengbu in Anhui Province in China. The patients were enrolled sequentially during June and July of 2009 and 2010. All patients enrolled in this study are inhabitants of Wuhe County, Guzhen County or the Bengbu City suburbs. Malaria transmission in this region is non-stable but can lead to malaria endemic in China. In the 1960s to 1970s, there were two malaria epidemics which were primarily caused by the P. vivax parasite. P. falciparum and P. vivax parasites were found together in this region until the end of the 1980s, but P. falciparum has not been found since the early 1990s. During the first decade of this century (from 2000 to 2010), malaria in this and other regions of China was mainly caused by the P. vivax parasite.
Diagnosis of P. vivax malaria infection was based on the examination of Giemsa-stained thick blood films. To prevent the misdiagnosis of mixed infection, confirmation of P. vivax was performed by polymerase chain reaction (PCR) using speciesspecific primers [26]. Clinical characteristics of the subjects are listed in Table 1. None of the volunteers were treated with a radical cure before blood collection. After collection, all malaria patients were cured by the standard regimen recommended by the Chinese Ministry of Health [27]. All patients were asked to follow up after blood collection; however, only 10 patients consented to give the follow up blood samples on days 7 and 14.
Blood samples were collected from 17 additional healthy volunteers who were randomly selected from the same P. vivaxendemic area during the same period of time. These subjects, serving as uninfected malaria-exposed controls (UM), had P. vivax infections in the twelve months prior to the time of blood collection. Negative P. vivax parasitaemia was confirmed by PCR with species-specific primers for four strains of human malaria (P. falciparum, P. vivax, P. malariae, and P. ovale) [26]. Twenty-one healthy adults living in the Bengbu city area without previous malaria infection and with no antibodies to malaria parasites, as determined by ELISA, were recruited to serve as unexposed controls (UC).

Preparation of Peripheral Blood Mononuclear Cells (PBMCs)
Venous blood samples from P. vivax patients, uninfected malaria-exposed controls, and unexposed controls were collected in heparinised tubes, and PBMCs were freshly separated by gradient centrifugation using Ficoll-Hypaque lymphocyte isolation kits (Tian Ji Hao Yao Biological Manufacturer, Tianjing, China) according to the manufacturer's recommendations. The PBMC pellet was resuspended at 10 6 cells/mL in RPMI-1640 (GIBCO, Carlsbad, USA) supplemented with 10% heat-inactivated fetal calf serum (FCS) (Hangzhou Sijiqing Organism Engineering Materials, Hangzhou, China). The PBMCs were counted and resuspended in the RPMI-1640 for further experiments.

Parasite Cultures and Antigen Preparations
P. vivax-infected red blood cells (iRBC) enriched from the blood of acute P. vivax-infected Chinese patients were used as antigens for in vitro stimulation. Briefly, white blood cells were depleted from P. vivax infected blood by filtering through a sterile CF11 cellulose column (WhatmanH, Maidstone, UK). Red blood cells were washed with RPMI-1640 by centrifugation at 1,190 g for 5 min. The parasites were cultured for 24-30 hrs at 5% hematocrit in McCoy's 5A medium (GIBCO) supplemented with 25% human AB serum [4]. P. vivax from all samples was pooled to gain a large number of parasite antigens. For P. falciparum, the infected blood from acute P. falciparum-infected Thai patients was collected, white blood cells were removed, and the parasites were cultured at 5% hematocrit in RPMI-1640 medium supplemented with 10% human serum [28]. The isolated P. falciparum parasites from all samples were pooled to recover a large number of parasite antigens. Both P. vivax and P. falciparum parasites were maintained in an incubator containing 5% CO 2 , 5% O 2, and 90% N 2 until the parasites matured to schizont stages ($6 nuclei). The late stage iRBCs were enriched by gradient centrifugation using 60% Percoll (GE Healthcare, Uppsala, Sweden) at 1,190 g for 10 min. The range of parasite purity was 60-99% for P. vivax and 80-100% for P. falciparum.
The enriched iRBC pellets were sonicated for 40 sec at 150 watts, and the protein concentration was determined by Bradford assay (Bio-Rad, Hercules, USA). The proteins were then aliquoted and stored at 270uC until use. Uninfected RBCs were processed as above, and the RBCs with a protein concentration equal to the malaria antigens were stored at 270uC to be used as a control. To determine the level of anti-P. vivax or anti-P. falciparum antibodies in P. vivax-infected patients, uninfected malaria-exposed and unexposed control groups, we performed an ELISA as previously reported [21]. Briefly, P. vivax, P. falciparum antigens, or uninfected red blood cell lysates (10 mg/ml) were coated into a 96well plate. Plasma, at 1:100 dilutions, was added into duplicate wells and incubated for 2 hrs at 37uC. Horseradish peroxidaseconjugated goat anti-human IgG (Caltag, Burlingame, USA) was added and incubated for 1 hr at room temperature, and 2, 20azino-di-(3-ethylbenzthiazoline sulphonic acid) containing 50% hydrogen peroxide (Kirkepaard & Perry Laboratories, Gaithersberg, USA) was added and incubated for 30-60 min in the dark at room temperature. Enzyme activity was measured by an automated microplate reader. All values were subtracted from the baseline value obtained using uninfected red blood cell lysate before performing data analyses.

In vitro Stimulation
PBMCs from ten uninfected malaria-exposed controls, ten unexposed controls and ten P. vivax patients were freshly isolated and used for in vitro stimulation. PBMCs in RPMI-1640 supplemented with 25 mM HEPES, 2 mM glutamine, 40 mg/ml gentamicin and 10% heat-inactivated FCS were cultured at 2610 5 cells/well in a round-bottom, 96-well plate (Costar, Corning, USA) for 5 days at 37uC and 5% CO 2 in the presence of 1, 10, or 50 mg/ml of either the P. vivax or P. falciparum antigen (AMB47). An equivalent concentration of proteins from nRBC or medium alone was used as baseline stimulation for the lymphocytes. All stimulation values were subtracted from the median percentages of the baseline values (calculated by either values from medium alone or unstimulated RBCs). An anti-CD3 OKT antibody (2 mg/ml) was used as the positive control. All samples were performed in triplicate. After five days of activation, the cells were pulsed with 1 mCi H 3 -Thymidine for 16-18 hrs. The proliferation of PBMCs was then determined by b-counter FJ-353 (262 Factory, Xian, China), and the data are represented as counts per minute (CPM). The CPM values from the stimulation assay were calculated from the CPM of antigen-stimulated cultures divided by the values from cultures without antigen stimulation.

Determination of Surface and Intracellular Proteins by Flow Cytometric (FCM) Analysis
Two hundred microliters (ml) of whole blood were collected. RBC were lysed with RBC lysing solution at room temperature for 15 min and the remaining cells were washed with PBS before staining using various combinations of fluorochrome-conjugated mAbs (as shown in Table 2) for 30 min at 4uC. The cell pellets were analysed on a FACSCalibur using CELLQuest software (Becton Dickinson, San Jose, USA).
For intracellular staining, the cells were washed with a permeabilising solution according to the manufacturer's instructions (BioLegend, San Diego, USA). The cells were incubated in the permeabilising buffer for 20 minutes at room temperature. Alexa fluorH 488-labeled anti-Foxp3 (BioLegend, San Diego, USA) was added and incubated for 30 min at room temperature. The cells were washed with PBS and fixed with 2% paraformaldehyde in PBS for acquisition and analysis on the FACSCalibur.
See the Supplementary Figure S1 for the FACS analysis of CD4 + CD25 hi FOXP3 + T cell populations.

Data Analysis
All data were analysed using the SPSS programme (Version 11.5, Chicago, USA). Non-parametric Kruskal-Wallis tests, followed by post tests, were used for statistical comparisons among three groups ( Fig. 1 and 2). The data from phenotypic analyses of unstimulated lymphocytes in the results and figures ( Fig. 1 and 2) were expressed as medians and interquartile ranges (25 th -75 th percentile). The mean differences (MD), 95% confidence intervals (CI), and P-values from the statistical analyses are also represented. The results were considered statistically significant when P,0.05 at a 95% confidence interval. The data from the in vitro studies (Fig. 3) are represented as the median 6 standard error of the mean (SEM).

Phenotypic Analysis of Unstimulated Lymphocytes
Acute P. vivax infection modulated immune function by altering the reciprocal ratios of the major immune cells in various lymphocyte populations. The percentage of CD4 + CD45RO + memory T cells out of the total number of T cells (27%, N = 37, Fig. 1A) in the acute infection was significantly higher than that of the unexposed controls (20%, N = 21, MD = 8.8, 95%CI = 0.5-17.1, P = 0.02). However, this result was not significant when compared with the data from the uninfected malaria-exposed controls (22%, N = 17, P = 0.1). In contrast, the level of CD8 + CD45RO + memory T cells among patients with acute P. vivax infection (17%, N = 37) was similar to that of the uninfected malaria-exposed controls (18%, N = 17). These levels were higher than, but not significantly different from, the levels observed for the unexposed controls (13%, N = 21, P = 0.2, Fig. 1B).
In contrast to the percentage observed in the CD4 + CD25 hi T cells, the median percentage of Foxp3 + Treg (CD4 + CD25 hi -FOXP3 + T cells) for the total number of CD4 + T cells among the unexposed controls was high (14%, N = 21) but not significant when compared with the percentage observed for the acute infection group (7%, N = 35, P = 0.2) and the uninfected malariaexposed controls (10%, N = 16, P = 0.1). Two out of thirty-seven of the acute P. vivax infected patients had 35% FOXP3 + Treg (P = 0.001, compared to P. vivax-infected patients), and 1 out of 17 of the uninfected malaria-exposed controls had an extraordinarily high level (31%) of FOXP3 + Treg (P = 0.001, compared with the uninfected malaria-exposed control group, Fig. 2B).
The median percentage of NK cells in total lymphocytes decreased but was not significant during the acute P. vivax infection (6%, N = 37) compared with the results observed for the uninfected malariaexposed controls (15%, N = 17) and the unexposed group (16%, N = 26, P.0.05, Fig. 2C). In contrast, the percentage of NKT cells out of total number of lymphocytes (Fig. 2D) was significantly higher during acute infection (3%, N = 37) compared with the uninfected malaria-exposed group (2%, N = 17,MD = 1.4, 95%CI = 20.5-3.0, P = 0.05) and was significantly lower than the results observed for the

Proliferation of T Lymphocytes by Cross Protein Stimulation
Optimisation of PBMC stimulation was performed with PBMCs from 10 malaria patients using various concentrations of P. vivax, P. falciparum, and nRBC antigens. Stimulation with 10 mg/ml of P. vivax or P. falciparum antigens substantially elevated the level of H 3 -Thymidine uptake compared to cultures without antigen (data not shown). Therefore, further PBMC activation assays were performed using 10 mg/ml of malaria protein.
PBMCs from 10 healthy donors (unexposed) were co-cultured with the nRBC, P. vivax, and P. falciparum antigens to determine the baseline activation as shown in Fig 3. The CPM values showed a low-level response to P. vivax antigens (median6SEM = 152680 CPM). Higher proliferation was observed when PBMCs were activated with the P. falciparum antigen (11866247 CPM) compared to their activation by anti-OKT (28006923 CPM).
PBMCs from 10 uninfected malaria-exposed controls cocultured with nRBC, P. vivax, and P. falciparum antigens showed a more robust response to P. vivax antigens (29516452 CPM) than P. falciparum antigen (19636591 CPM) compared to the activation of PBMCs that was observed using anti-OKT (49686606 CPM).
PBMCs from 10 acute P. vivax-infected patients and from their follow-up periods on days 7 and 14 were co-cultured with the nRBC, P. vivax, and P. falciparum antigens to determine the cross stimulation between the two malaria species. PBMCs from acute P. vivax-infected patients stimulated with P. vivax antigen showed a higher degree of proliferation (4646417 CPM) compared to PBMC stimulation with the P. falciparum antigen (1166110 CPM) after subtracting the background measured in the nRBC stimulated culture. Higher cell levels of proliferation were observed upon activation by anti-OKT (29166552 CPM). The 7-day follow up PBMC proliferation assays from these patients showed no significant changes in proliferation upon stimulation with all three proteins. The stimulation by anti-OKT was 900861216 CPM. Interestingly, the 14-day follow up PBMC count from the infected patient group showed markedly increased proliferation upon stimulation with P. vivax antigens (1392064623 CPM) compared to stimulation with P. falciparum antigens (67726501 CPM, P = 0.03, Fig. 3). The level of stimulation induced by anti-OKT was 122096450 CPM.

Discussion
People in northern Anhui Province in Central China have had experience with only P. vivax infection. The P. falciparum parasite is absent in this area. Therefore, the existing immunity to malaria infection is shaped exclusively based on exposure to the P. vivax parasites. The T cell phenotypes in people infected with only P. vivax have not been previously investigated. Our major finding was that upon acute P. vivax infection, CD4 + memory T cells and cd T cells were elevated in the blood. CD8 + memory T cells were only detectable in the acute-infected and uninfected malaria-exposed patients. Moreover, a non-immune suppressive phenotype was found in P. vivax infection, demonstrated by the decreased expression of FOXP3 + Treg cells.
Elevation of CD4 + memory T cells in acute P. vivax-infected patients in China was consistent with a study in Thailand that showed that during both acute infection and the convalescent period, similarly high levels of CD4 + memory T cells were present [21]. However, the levels of both CD4 + and CD8 + memory T cells in acute P. vivax patients from Thailand were higher than the levels of these cells observed in China. The high level of CD4 + and CD8 + memory T cells may be due to frequent exposure to malaria parasites among the Thai population. A previous study has shown IL-4-producing CD4 + CD45RO + T cells in the circulation of people who are exposed to attenuated P. falciparum sporozoites [29]. This finding has been confirmed by recent studies showing the presence of IL-4-producing CD4 + memory T cells (CD45RO + CD27 2 ) induced by attenuated P. falciparum sporozoites. [21,30]. These memory T cells may also confer protection against P. vivax and long-term protection against P. falciparum infection. However, a recent study has demonstrated that T cells Figure 3. The activation of PBMCs from unexposed controls (N = 10), uninfected malaria-exposed controls (N = 10), and acute P. vivax-infected patients (N = 10), which was followed up on day 7 (N = 10) and day 14 (N = 10) with P. vivax antigens (10 mg/ml) (PV10) and P. falciparum antigens (10 mg/ml) (PF10). The proliferation values were expressed as counts per minute (CPM). Data are shown in median6SE. All stimulation values were subtracted from the median percentages of baseline value. doi:10.1371/journal.pone.0045971.g003 from naïve individuals (without previous exposure to malaria) can differentiate into memory T cells when exposed to cross-reacting malaria antigens. Therefore, the high level of memory T cells observed during acute P. vivax infection suggests memory T cell activation by the parasites or priming by exposure to commensal micro-organisms, pathogens and/or vaccine antigens carrying minimal T cell epitopes that cross-react with malaria proteins [31]. In comparison to the healthy group, the level of CD8 + memory T cells was readily detectable in P. vivax-infected patients. This finding has been confirmed by recent studies showing enhanced effector T-cell populations during acute P. vivax infection [32]. However, cytotoxic T cells have been shown to play a minor role in murine malaria, especially in the pre-erythrocytic stage [33]. Therefore, it is expected that regular exposure to P. vivax parasites could maintain cytotoxic T cells in uninfected malariaexposed controls. Additionally, exposure to liver-stage antigens of P. vivax during an exo-erythrocytic or a hypnozoite stage could result in the generation of long-lived CD8 + memory T cells in immune individuals.
Elevated cd T cells have been demonstrated in the current study and in various models of plasmodia infections [21,34,35,36]. A recent study showed that cd T cells reduce the severity of malaria in humans [13] and control the chronic parasitaemia of P. chabaudi infection in mice [37]. The cytotoxic role of cd T cells was shown in vitro using P. falciparum [38,39,40,41]. This result suggests that cd T cells play both cytotoxic and regulatory roles, which presumably lead to protection. However, the role of cd T cells in protection related to symptom intensity or parasitaemia during malaria infection needs to be investigated further.
Previous findings show the inhibition of B cell phenotypes in both P. falciparum and P. vivax infections [24,42,43]. In addition, our results showed the level of B cells in acute P. vivax-infected patients was unchanged. One possible explanation is that Treg may directly suppress B cells [44] or the proliferation of Th cells, resulting in the down-regulation of IL-2 or IL-4 production by responder lymphocytes and the subsequent suppression of B cell differentiation into plasma cells. In addition, we found significantly decreased levels of B cells in P. vivax-uninfected malaria-exposed controls, which was consistent with the low antibody response in P. vivax infection as previously shown [21]. A recent study in P. falciparum-exposed populations of Africa and South America showed significant changes in B cell subsets [42,45,46,47]. The difference in these findings could be consistent with the conventional understanding that P. falciparum is a polyclonal B cell activator that stimulates the production of large amounts of antibodies in the infected host during the convalescent period. However, during P. vivax infection, the patients produce lower levels of antibodies compared to the patients infected by P. falciparum [48].
Treg constitutively express CD25, which is the IL-2/a chain receptor. Co-expression of CD25 with forkhead box protein P3 (FOXP3) dictates the immune-suppressive role of Treg via the release of IL-10 and TGF-b [49,50]. The unchanged levels of FOXP3 + Treg cells among P. vivax-infected Chinese populations was in contrast to levels found in P. vivax-infected Thai and Brazilian populations, whose levels of FOXP3 + Treg cells were elevated [20,51,52]. One reason for this difference could be that patients living in the endemic areas of Thailand are likely to have experienced both P. vivax and P. falciparum infection due to the equal incidence of both species, whereas P. vivax is the only malaria species found in central China. Another reason could be that P. falciparum activates FOXP3 + Treg cells, as shown previously [53]. Taken together, these findings suggest that an activator for FOXP3 + Treg cells in the Thai patients may originate from P. falciparum parasites. However, we found some cases of P. vivaxinfected patients and uninfected malaria-exposed controls that had high levels of FOXP3 + Treg cells. This result may be caused by frequent exposure to P. vivax infection among people living in endemic areas.
The reduction of NK cells in P. vivax-infected Chinese populations compared to healthy volunteers was consistent with what was found in P. vivax-infected Thai [54] and Ethiopian populations [24]. The activation of NK cells during acute infection occurred in the host and was not due to geographical region, genetic differences, or level of malaria endemicity. The protective role of NK cells in malaria infection is supported by the finding that these cells destroy P. falciparum-infected erythrocytes via direct contact [34,55]. Moreover, a recent study showed that NK cells in peripheral blood are the primary producers of IFN-c upon exposure to the infected erythrocytes in the presence of IL-12 [34]. Therefore, NK cells may be the prime effectors against P. vivax parasites during the acute phase of infection, although only a small change in the level of NK cells was observed. However, the cumulative number of these cells present during the entire course of infection could not be calculated. Therefore, the suppressive effect of Treg and the early activation of NKT cells towards the P. vivax-infected erythrocytes were exerted at the onset of the infection. However, the lower number of NK and NKT cells found during the acute infection period compared to the number of cells observed in the malaria-unexposed controls may be explained by the incidence of lymphopoenia documented during both P. falciparum and P. vivax infections [24,56,57].
The results of stimulation of acute P. vivax-infected patients during the 7-day follow up were similar to those of the unexposed controls, suggesting drug resistance or immune suppression caused by the parasites. However, all patients were given chloroquine for 3 days plus primaquine for 8 days for treatment of P. vivax-induced malaria. Several studies have investigated the effect of chloroquine on the human immune system in vivo, with a particular focus on immune suppression [58,59]; however, the mechanisms underlying this treatment are not thoroughly understood. In our experiments, PBMCs from all day 7 follow up patients were collected 4 days after chloroquine treatment was completed. This finding suggests that there were residual effects of chloroquine on the T cells. However, the response of the patients' lymphocytes during the convalescent period in this study was not restricted to only P. vivax antigens, although P. vivax is the only cause of malaria infection in this area of central China.
The response of PBMCs from uninfected malaria-exposed controls to P. vivax and P. falciparum antigens suggested that parasites stimulated memory cell pools of P. vivax-specific lymphocytes during infection [21]. The stimulation of uninfected malaria-exposed controls and day 14 follow up PBMCs with P. falciparum antigens also resulted in greater levels of lymphocyte proliferation. This finding suggests antigenic cross-reactivity between P. vivax and P. falciparum, as recently suggested, by using the serological cross-reaction [13,23] and the accumulation of both P. vivax and P. falciparum-specific memory T cells during the convalescent period, resulting in the robust response to both P. vivax and P. falciparum antigens. This finding sheds light on a possible new strategy for malaria vaccine design, i.e., a combination of antigens derived from P. vivax and P. falciparum parasites, which may attenuate malaria severity. P. vivax parasites do not suppress the immune function of the infected host, which commonly accompanies P. falciparum infection. However, elucidation of the immune cell-parasite interaction during malaria infection using single or mixed species needs to be investigated in depth to gain a better understanding of immunity to malaria infections. Figure S1 FACS analysis of the CD4 + CD25 hi FOXP3 + T cell population (A) represents the CD3 + population. The R1 gate was used for (B), and R2 was the gating population of (C). 20000 of total evens were used for each gate. (DOCX)