Activatory Receptor NKp30 Predicts NK Cell Activation During Controlled Human Malaria Infection

Natural killer (NK) cells are known to be activated during malaria infection, exhibiting both cytokine production and cytotoxic functions. However, NK cells are heterogeneous in their expression of surface activatory and inhibitory receptors which may influence their response to malaria parasites. Here, we studied the surface marker profile and activation dynamics of NK cells during a Controlled Human Malaria Infection in 12 healthy volunteers. Although there was significant inter-patient variability in timing and magnitude of NK cell activation, we found a consistent and strong increase in expression of the activatory receptor NKp30. Moreover, high baseline NKp30 expression was associated with NK cell activation at lower parasite densities. Our data suggest that NKp30 expression may influence the NK cell response to P. falciparum, explaining inter-patient heterogeneity and suggesting a functional role for this receptor in malaria.


INTRODUCTION
Malaria infection in humans activates a broad cellular immune response involving monocytes, T cells, B cells, and NK cells. NK cells may play a functional role in protection against Plasmodium falciparum, as certain NK cell receptor genotypes are associated with decreases in malaria susceptibility and pathology [reviewed in (1)]. During the pathological blood stage of P. falciparum infection, circulating NK cells display a dual functional role, i.e., cytokine production (2)(3)(4)(5) and killing of infected blood cells both via antibody-independent (6)(7)(8) and antibody-dependent cytotoxicity (9,10). Their relative contribution to protection remains unknown.
NK cells are often considered a homogenous, unchanging population, but multicolored flow cytometry and mass cytometry have revealed that NK cells actually consist of many distinct populations, differing in their functionality against specific diseases (11)(12)(13)(14). Artavanis-Tsakonas et al. previously demonstrated that in malaria naïve donors a specific subpopulation of NK cells expressing the lectin-type receptor NKG2A are the main IFN-γ producers in response to P. falciparum-infected RBC (15). Most studies determining the NK cell response against P. falciparum demonstrate that there is large inter-donor variability (16,17). We hypothesized that this heterogeneity might at least in part be explained by differences in NK cell phenotype prior to infection.
To date most data on responsiveness of NK cells to P. falciparum has been obtained from ex vivo stimulation experiments or case-control studies in endemic areas. We took advantage of the Controlled Human Malaria Infection model to evaluate the activation and function of different NK cell subsets at multiple time points during a malaria infection. Our data show in vivo NK cell activation in all donors with an upregulation of IFN-γ and granzyme B production. There was indeed a significant variability both in the timing and magnitude of the NK cell response, and increased baseline receptor expression of NKp30 predicted a more rapid in vivo NK cell activation.

Clinical Trials
Study 1 was a single-center, open-label clinical trial in 12 malaria naïve individuals conducted at the Radboud university medical center (Nijmegen, The Netherlands) from May until June 2018. Study volunteers provided written informed consent and were screened as described previously (18). The trial was approved by the Central Committee on Research Involving Human Subjects (CCMO; NL63552.091.17) of the Netherlands, performed according to the Declaration of Helsinki and Good Clinical Practice and prospectively registered at ClinicalTrials.gov (NCT03454048). Volunteers were infected by the bites of five P. falciparum 3D7 strain-infected Anopheles mosquitoes, and followed up for parasitemia twice daily starting on day 6 post infection. Parasitemia was assessed by thick blood smear and qPCR. Volunteers were treated with a suboptimal dose of piperaquine when parasitemia reached density detectable by thick blood smear or 5,000 parasites/milliliter by qPCR, and received curative treatment if recrudescent parasitemia occurred.
Study 2 was a single-center randomized placebo controlled malaria vaccine trial (CCMO NL39541.091.12; NCT01728701) published previously (19). Only study subjects that received placebo vaccination followed by CHMI were included in the current analysis. In short, volunteers received bites from five P. falciparum NF54 strain-infected Anopheles mosquitoes, and were followed up for parasitemia twice daily starting on day 5 post infection. Parasitemia was assessed by thick blood smear and/or qPCR, and volunteers received curative treatment with atovaquone/proguanil, either when parasitemia reached levels detectable by microscopy (n = 5) or after two consecutive qPCRs >500 parasites/milliliter (n = 4).

PBMC Isolation and Cryopreservation
In study 2, blood samples for peripheral blood mononuclear cell (PBMC) isolation were taken pre-challenge, 3 days after antimalarial treatment and 35 days after challenge infection. Isolation and cryopreservation was performed as described previously (22). In short, PBMCs were isolated from citrate anti-coagulated blood using vacutainer cell preparation tubes (CPT; BD Diagnostics) by density gradient centrifugation. Cells were washed four times in ice-cold phosphate buffered saline (PBS), counted using 0·1% Trypan blue with 5% Zapo-Globin II Lytic Reagent (Beckman Coulter), cryopreserved at a concentration of 10 × 10 6 cells/ml in ice-cold fetal calf serum (Gibco)/10% DMSO (Merck), and stored in vapor-phase nitrogen.

Data Analysis and Statistics
Flow cytometry data was analyzed using Flow Jo software (version 10.0.8 for Apple OS). Statistical analysis was performed using GraphPad Prism (version 5.03 for Windows). Gating strategy and representative plots are shown in Supplementary Figure 1

Heterogeneity in NK Cell Activation After CHMI
After malaria infection, NK cell activation as defined by upregulated CD69 expression was determined daily from day 6 post-infection until 3 days after antimalarial treatment (Supplementary Figure 1). In study #1 the first activation of NK cells in a number of volunteers was observed 1 day after the first appearance of parasitemia detectable by qPCR (Figure 1).
In the absence of parasitemia, up to 3.5% of NK cells expressed CD69, therefore >3.5% CD69 expression was considered FIGURE 1 | Kinetics of parasitemia and NK cell activation during Controlled Human Malaria Infection. NK cells were analyzed by flow cytometry daily in whole venous blood from 12 volunteers undergoing Controlled Human Malaria Infection. Antimalarial treatment was initiated when parasite densities reached levels detectable by microscopy. Each graph shows the activation of NK cells (defined by CD69 surface expression) from day 6 post infection until day 3 after antimalarial treatment (blue line, left axis). The same graph shows parasitemia measured by qPCR from day 6 after infection, until day 3 after antimalarial treatment (red line, right axis). NK cell activation is first seen 1-2 days after the first appearance of parasitemia. Each graph represents the data gathered for a single volunteer (n = 12). significant NK cell activation above background (Figure 2A). There was indeed a significant heterogeneity in the timing of first NK cell activation, ranging from 1 day after the first appearance of parasitemia (i.e., volunteer 5) to 5 days after parasitemia (volunteer 7). This may be partially explained by differences in starting parasite density. Parasitemia (prior to the initiation of antimalarial treatment) correlated strongly with the degree of NK cell activation (Spearman p = 0.0017; Figure 2A). However, this does not explain the diversity entirely, as some volunteers have significant NK cell activation (defined as CD69 expression >3.5%) at very low circulating parasitemia, such as volunteer 5, while others require very high parasitemia before NK cells become activated, such as volunteer 7. This circulating parasite density prior to NK cell activation was highly variable between volunteers (mean 4,798 Pf/ml, range 25-26,152 Pf/ml), suggestive for a host-dependent factor.

Baseline NKp30 Expression Predicts Activation After CHMI
NK cell activation is dependent on a delicate balance between activatory-and inhibitory receptors, and the expressed receptor profile may relate to the observed heterogeneity during CHMI. Therefore, we next determined whether the expression of activatory receptors NKp30 or NKG2D, the inhibitory receptor NKG2A or the differentiation marker CD57 predicted an individual's response to CHMI. Indeed, higher baseline NK cell NKp30 expression correlated with activation at lower parasitemia (linear regression p = 0.047; Figure 2B). NKp30 and NKG2D were expressed on nearly 100% of NK cells for all volunteers (Supplementary Figure 1).
NK cells can be divided into distinct populations representing levels of differentiation based on their expression of CD56, CD16, NKG2A, and CD57 (11), and a previous study suggested NKG2A+ NK cells are more responsive to P. falciparum in vitro (15). We sought to determine whether this may result from differential expression of NKp30. However, while baseline expression of NKp30 varied between CD56dimNKG2A+ and CD56dimNKG2A-subsets ( Figure 3A;  Supplementary Figure 2), all NK cell subsets showed an upregulation of NKp30 ( Figure 3A). Furthermore, we did not see any differences in activation as defined by CD69 upregulation between the CD56dim subsets, though there was significantly more activation of the CD56dim subset compared to the CD56bright subset (Figures 3B,C).
NK Cell Subsets Upregulate CD69, IFN-γ, and Granzyme B During CHMI As there appears to be little activation of the CD56bright NK cell subset during the course of infection, we wanted to determine the ability of both the CD56brightCD16and CD56dimCD16+ subsets to produce granzyme B and IFN-γ and degranulate during infection, using isolated and cryopreserved peripheral blood mononuclear cells (PBMCs) from study #2 (Supplementary Figure 3). We found that both subsets increase production of granzyme B and IFN-γ and show improved degranulation during infection (Figures 4A-C;  Supplementary Figure 4).

DISCUSSION
These data show that NKp30 is a marker for the NK cell response during a Controlled Human Malaria Infection, and suggests a possible functional role in the response to infected red blood cells. We demonstrate that the expression of this receptor at baseline relates to individual NK cell responses to P. falciparum in vivo.  Furthermore, we show that NK cell activation during the course of infection is linked to an increase in NKp30 expression.
Both NKp30 and NKG2D have been shown to increase expression during NK cell activation (14,23), however, during CHMI the magnitude of NKp30 upregulation of is particularly pronounced compared to NKG2D. It has previously been demonstrated in vitro that NKp30 binds to the P. falciparum protein Pf EMP1 leading to NK cell activation (24). This supports our finding that NK cells with higher resting NKp30 expression are more sensitive to activation at lower parasitemia. However, it is important to note that other in vitro studies suggest that Pf EMP1 may be dispensable (25) and that MDA5 signaling may be essential (26) for NK cell activation in response to Pf RBC. Therefore, multiple mechanisms may be involved in NK cell activation during malaria. This is the first study with longitudinal daily samples from the initial phase of a malaria infection as parasites emerge from the liver that suggests an important role for NKp30. We thereby measured CD69 expression directly in patient blood samples, without re-stimulation, remaining close to the induced in vivo phenotype of an early natural infection. Furthermore, we show that baseline NKp30 expression is linked to a more rapid NK cell activation during subsequent infection.
Population based studies conducted in sub-Saharan Africa have identified a single nucleotide polymorphism (SNP) in the promoter for the NCR3 gene that encodes NKp30 that is associated with an increased number of clinical, uncomplicated malaria episodes in individuals over 5 years old (27)(28)(29). The combined data are highly suggestive for a potential functional role of NKp30-mediated NK cells in malaria. In our study we do not detect differences in time to parasitemia, maximum parasitemia, or parasite multiplication rate between those with high NKp30 expression and those with low NKp30 expression (data not shown). However, an important limitation of this study is that it was not designed to measure an effect on control of blood stage parasite replication. Even in volunteers with very rapid NK cell activation, this occurred only 2 days before the initiation of antimalarial treatment. This period between NK cell activation and drug treatment would be too short to measure an effect on parasite multiplication. Instead, cohort studies in endemic areas are better suited to answer this question. Future studies in endemic areas could determine NKp30 expression on NK cells at the beginning of a malaria season and during follow-up visits, and correlate this with number of clinical malaria episodes.
Broad inter-donor variability in the activation of NK cells in response to P. falciparum has been described in multiple studies (16,17,21). Our current finding suggest that baseline NK cell phenotype can play a role in this diversity. However, other immunological factors, including other activatory and inhibitory receptors not studied here, interactions with other immune cells and cytokine production likely also contribute to the NK cell response. Furthermore, parasitological factors, such as the initial starting parasitemia and parasite multiplication rate may also affect host response.
The phenotypic diversity of NK cells has been a topic of extensive study during the last decade (12,14). Since the first discovery of NK cell memory in murine CMV infection (30), specific NK cell phenotypes have been identified as the main responders in human EBV (13), CMV (31), and HIV infection (14,32) as well. Similarly, studies suggested that NKG2A+ NK cells, specifically respond to P. falciparum (15,16). Interestingly, this does not appear to be the case during controlled human malaria infection in vivo.
Nevertheless, the finding that NKp30 expression predicts the response to CHMI, underscores the potential importance of NK cell phenotype in our susceptibility to disease. The diversity of the NK cell repertoire has been implicated in the risk of HIV acquisition (14,33), and viral infections in turn have been shown to change its composition (34)(35)(36)(37). Our study suggests that NK cell phenotype affects the response to a P. falciparum infection.
The current study was limited to analysis of CD56, CD16, NKG2A, CD57, NKp30, and NKG2D. In contrast, data from studies on other diseases using cytometry by time-of-flight (CyTOF) have suggested there may be more than 100,000 NK cell phenotypes, each characterized by a distinct combination of surface receptors (12). Furthermore, the expression of diverse killer cell immunoglobulin-like receptors (KIRs) plays an important role in NK differentiation and function (11). Therefore, it is likely that additional receptors, or combinations of inhibitory and activatory receptors, are also important for the interaction between NK cells and P. falciparum parasites. Future studies looking at a larger number of receptors and cytokines could unravel both these effects in more detail.
In conclusion, this study is the first to identify the expression the NK cell activatory receptor NKp30 as a marker that predicts a rapid NK cell response to parasitemia and suggest a potential role for this receptor in NK cell functionality against P. falciparum.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Central Committee on Research Involving Human Subjects, Netherlands. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
JW and RS designed the study. JW performed the analysis and wrote the first draft of the manuscript which was supervised by RS.

FUNDING
JW and RS were supported by a grant from the Bill and Melinda Gates Foundation, OPP1091355.