Background
Malaria caused by
Plasmodium falciparum remains a major public health threat. An estimated 3.4 billion people in 92 countries are at risk of malaria. In 2018,
P. falciparum caused approximately 228 million malaria cases and 405,000 deaths worldwide, the majority among children in sub-Saharan Africa [
1]. The recent scale-up of artemisinin-based combination therapy and insecticide-treated bed nets has reduced the malaria burden in Africa [
2]. Worryingly, the trend toward declining malaria cases and deaths observed over the last decade appears to have stalled even before the global COVID-19 pandemic emerged [
1]. Now, a recent study suggests that malaria-related deaths in 2020 could increase to more than double those of 2019 if malaria-prevention activities are interrupted due to COVID-19 [
3].
Ultimately, a key tool for the control, elimination, and eventual eradication of malaria is an effective vaccine, yet the leading vaccine candidate RTS,S confers only partial, short-lived protection in African children [
4]. A consistent finding across several human malaria vaccine trials is a lower immunogenicity in malaria-exposed individuals relative to malaria-naïve individuals [
5]. The development of highly effective malaria vaccines has been hindered in part by a poor understanding of the interaction between
P. falciparum and the human immune system [
6], particularly in malaria-endemic areas, where immunoregulatory networks are induced by repeated
P. falciparum infections [
7‐
12].
In malaria-naïve individuals,
P. falciparum infections lead to high parasitemia and a strong pro-inflammatory cytokine and chemokine response, referred to as “cytokine storm”, causing fever and other malaria symptoms [
13,
14]. However, in endemic areas with repeated exposure, parasitaemia can often be controlled and asymptomatic infections are the rule rather than the exception [
15‐
17]. Although sterile immunity does not appear to be reliably achieved, protective immunity to malaria can be acquired through natural
P. falciparum infection, but only after years of repeated exposures [
18]. Once acquired, this immunity appears to wane rapidly in the absence of ongoing exposure [
19].
A crucial part of this process are dendritic cells (DCs). As sentinels of the immune system, DCs are not only important for early cytokine responses but are also essential for bridging and regulating the innate and adaptive immune responses to pathogenic infections and vaccines [
20]. DCs reside throughout the body and sample their surroundings for pathogens. Once they phagocytose or encounter pathogenic material they undergo a rapid maturation process and migrate to secondary lymphoid organs to present antigen through MHC molecules to naïve T cells and thus initiate the adaptive immune response [
20,
21]. DC maturation is typically characterized by up-regulation of MHC class II (HLA-DR), co-stimulatory markers, and secretion of chemokines and immunomodulatory cytokines such as IL-6, IL-1β, TNF and IL-10. Depending on the nature of their activation, DCs can prime naïve CD4 T cells to become Th1, Th2, or other T helper cell subsets [
22] or induce tolerance [
23].
The two major human DC subsets, plasmacytoid dendritic cells (pDCs) and myeloid DCs (mDCs), have different roles in initiating and coordinating the immune response. While pDCs secrete high levels of IFNα and TNF upon activation, mDCs up-regulate costimulatory molecules and are highly efficient antigen presenting cells [
24].
Despite their essential role in initiating adaptive immune responses to natural infections as well as vaccines, little is known about the role DCs play in the immune response to
Plasmodium, especially in human malaria infections. Studies both in humans and mouse models have been contradictory [
25]. The seminal report on human DCs and
P. falciparum suggested that DC maturation was blocked by the parasite leading to dysfunction and suboptimal allogenic T cell activation [
26]. Conclusions drawn from this report were later challenged by Elliot et al
. showing the blocking effect was parasite dose-dependent [
27]. The experiments in these studies were conducted with monocyte-derived DCs and, therefore, may not be fully representative of human primary DC functions.
Very few studies have addressed DC function using purified primary DCs. One study analysed primary DC functions ex vivo from individuals with acute
P. falciparum malaria from an endemic area in Papua, Indonesia [
28]. The authors analysed DCs within PBMCs (not isolated DCs) and showed that acute infection increased spontaneous apoptosis and dysfunction characterized by decreased costimulatory marker up-regulation, antigen uptake and allogenic T cell activation. Increased spontaneous apoptosis correlated with increased IL-10 levels in the plasma of study subjects with acute malaria. A study investigating DCs from children and adults with asymptomatic malaria from the same cohort did not analyze DC viability but showed unchanged numbers and HLA-DR expression suggesting functional DCs in these asymptomatically infected individuals [
29]. Recently, it has been shown that, in malaria-naïve US individuals, enriched primary blood DCs undergo an atypical maturation process characterized by up-regulation of costimulatory molecules and chemokines without significant secretion of cytokines. Despite the lack of inflammatory cytokine secretion, these DCs were capable of inducing a robust Th1-like CD4 T cell response in vitro [
30]. During symptomatic infection, this response is likely enhanced by host factors such as reactive oxygen species [
31].
These previous studies provide valuable insight into the effect of highly inflammatory conditions during malaria on circulating DCs and early processes during primary DC activation by the parasite in malaria-naïve individuals. However, they do not reveal how isolated DCs from malaria exposed or asymptomatically infected individuals in malaria-endemic areas respond to the parasite. Addressing this question will help us understand whether initial immunological responses to vaccines or natural infections are affected by regulatory networks in individuals living in endemic areas.
Therefore, this study aimed to address whether intense, lifelong malaria exposure or concurrent asymptomatic P. falciparum-infections in malaria-endemic Mali have an effect on DC frequency, viability and response to the parasite.
Methods
Mali study site, participants and sample collection
Both study sites, Kambila (ClinicalTrials.gov NCT00471302) and Kalifabougou (ClinicalTrials.gov NCT01322581) in Mali, have been described in detail before [
32,
33]. Both are rural villages approximately 20 km north (Kambila) and 48 km northwest (Kalifabougou) of Bamako in a region that that typically experiences intense, seasonal
P. falciparum transmission from July through December each year [
32]. For both study sites the enrollment exclusion criteria included anaemia (haemoglobin < 11 g/dL), current use of anti-malarials, corticosteroids or other immunosuppressants, fever > 37.5 °C or evidence of an acute infection, and current pregnancy. Plasma was obtained from US adult donors between the ages of 35 and 70 enrolled in a healthy donor protocol at the NIH Department of Transfusion Medicine, Clinical Center (ClinicalTrials.gov NCT00001846). Eligibility criteria for these donors included no history of malaria in the past 12 months. Further demographic and travel history data were not available from these anonymous donors, but prior
P. falciparum exposure was unlikely.
The cross-sectional study in Kambila included 35 adults (Table
1), who exhibited no symptoms of malaria and were enrolled between November and December 2016. Asymptomatic
P. falciparum infections in these individuals were detected by PCR analysis of blood spots at the NIH after in vitro assays had been conducted in Mali. Detailed methods for the detection of
P. falciparum blood-stage infection by PCR have been described before [
33]. Plasma samples from 19 children (Table
1) residing in Kalifabougou were obtained in May 2013 for the healthy baseline time point and during the following transmission season for the acute malaria and 7 days post treatment convalescence time points. Acute malaria was defined as ≥ 2500 asexual parasites/µL, an axillary temperature of ≥ 37.5 °C or self-reported fever within 24 h, and no other cause of fever discernible by physical exam. Acute malaria was treated according to the Malian National Malaria Control Programme guidelines.
Adults uninfected | 27 | 42 (37–49) | 15/12 | No |
Adults infected | 8 | 51 (46–55) | 5/3 | No |
Children | 19 | 8 (6–9) | 11/8 | |
Healthy baseline | | | | No |
Acute infected | | | | Yes |
7 days post treatment | | | | No |
Peripheral blood was collected by venipuncture into 8 mL sodium citrate-containing cell preparation tubes (BD, Vacutainer CPT Tubes) and transported to the laboratory in Bamako where PBMCs were isolated according to the manufacturer’s instructions. The blood volume was 48 mL for adults and 8 mL for children. PBMCs and dendritic cells were analyzed or isolated immediately. Plasma samples were cryopreserved at -80 °C and later shipped to the NIH. The same sample collection and processing was conducted for the US donors.
Plasmodium falciparum culture and lysate preparation
Asexual blood stage cultures of the P. falciparum strain 3D7 were maintained at 5% haematocrit in RPMI 1640, 25 mM HEPES supplemented with 10 μg/mL gentamicin, 250 μM hypoxanthine, 25 mM sodium bicarbonate, and 0.5% Albumax II under atmospheric conditions of 5% oxygen, 5% carbon dioxide, and 90% nitrogen. Late-stage P. falciparum-iRBCs (trophozoites and schizonts) were isolated using MACS Cell Separation LS Columns (Miltenyi Biotec). Plasmodium falciparum-iRBCs were washed, resuspended at 1 × 106/µL and lysed by three consecutive freeze/thaw cycles. Uninfected RBC lysates were prepared accordingly and used as negative control. Lysates were stored at − 80 °C until use.
Dendritic cell enrichment and culture
Myeloid Dendritic Cell Isolation Kits (Miltenyi Biotec) were used to negatively enrich primary peripheral blood mDCs from PBMCs, following the manufacturer’s instructions. Primary mDCs were cultured with RPMI 1640 supplemented with 10% heat-inactivated human AB serum (Valley Biomedicals) at 37 °C and 5% CO2. Enriched mDCs were seeded at 1.5 × 105 per well in 96-well tissue culture-treated plates and cultured with P. falciparum-iRBC lysates at a ratio of 1:3 (DC/iRBC) for 24 h. Uninfected RBC lysates were used as a negative control. mDCs were then harvested for FACS analysis and supernatants were stored at − 80 °C until shipment to the NIH and cytokine/chemokine analysis.
Flow cytometry
Flow cytometry was performed using a BD LSR II cytometer (BD Biosciences), and data were analysed with FlowJo 10 (Tree Star). Primary DC subsets were identified using the following labelled monoclonal antibodies: a lineage cocktail (CD3 UCHT1, CD14 HCD14, CD19 HIB19, CD20 2H7 and CD56 HCD56), HLA-DR (L243), CD303 (201A), CD1c (L161), CD141 (M80) and CD16 (3G8). DC apoptosis was quantified by staining a separate PBMC aliquot with monoclonal antibodies against lineage, HLA-DR (L243) and Annexin V as well as 7-AAD. To normalize to microlitres of blood, the total number of PBMCs was determined using a counting chamber after isolation, prior to further analysis and dendritic cell isolation. Assuming this number was close to the number of PBMCs in whole blood, FACS analysis percentages were used to determine cell #/µl of blood. mDC maturation after culture with P. falciparum-iRBC lysate was analysed using the following antibodies: HLA-DR (L243), CD80 (2D10), CD86 (IT2.2) and CD40 (5C3). All antibodies, Annexin V and 7-AAD were purchased from BioLegend or BD Biosciences.
Cytokine and chemokine analysis
To analyse cytokine and chemokine levels in plasma and culture supernatants, the following kits were used: LEGENDplex Human Inflammation Panel 1 (13-plex) (Biolegend), Cytometric Bead Array Human Inflammatory Cytokine Kit (BD Biosciences) and Cytometric Bead Array Human Chemokine Kit (BD Biosciences).
Statistical analysis
Statistical analyses were performed using Prism 8 (GraphPad Software) and JMP (SAS Institute). Depending on the experimental design, a paired or unpaired t-test with Welch’s correction was performed. For data in Fig.
6 that were mixed paired and unpaired, a linear mixed model ANOVA with Tukey post hoc tests were performed. The statistical tests are described in the figure legends.
Discussion
The biology of primary human DCs is understudied due in part to their low frequencies in peripheral blood and the technical difficulties of relevant experimental assays. The understanding of primary human DC responses to
P. falciparum in particular remains very limited (reviewed in [
44]). In light of evidence that the leading malaria vaccine candidate RTS,S confers only partial, short-lived protection in African children [
4], it is of great importance to gain a better understanding of DC responses, particularly in malaria-endemic settings. This study attempted to address this knowledge gap by analysing DC frequency, viability and responsiveness to
P. falciparum in a malaria-endemic setting.
The findings suggest that mDC numbers in circulation remain stable during asymptomatic malaria in Malian adults. pDC frequencies, however, were lower in individuals with asymptomatic
P. falciparum infections compared to uninfected individuals. Several other studies have reported lower numbers of circulating DCs in acute malaria patients in both children and adults [
28,
37,
45‐
47]. In asymptomatic
P. falciparum infections in children and adults in Papua, Indonesia, DC numbers were found to be unchanged, including pDC frequencies [
29]. This discrepancy could be explained by the distinct study sites. Malaria generally presents differently in Asia and in West Africa, likely due to a number of differences including transmission intensity and availability of resources [
48,
49]. Individuals with asymptomatic infections in our cohort were significantly older than the uninfected controls. Since pDC, but not mDC, numbers in circulation decline with age [
50‐
52], lower pDC frequencies in this group could be attributed to age rather than asymptomatic malaria. It is also unclear whether decreased numbers of DCs in circulation is due to DC apoptosis [
28], or migration of activated DCs out of circulation to target tissues. In this study, a higher percentage of apoptotic total DCs in asymptomatically infected compared to uninfected adults was not observed. Since the percentage of apoptotic pDCs was not assessed, however, the total DC data could mask a difference between groups in pDC apoptosis. The limited number of infected individuals and rather large variability of DC apoptosis limits the interpretability of these data. Additional experiments will be needed to address whether peripheral blood pDCs undergo apoptosis in asymptomatically infected individuals.
A few studies have reported an increase in CD141
+ mDCs in children with acute malaria [
38,
39,
53]. This increase was shown to be induced by Flt3 ligand, which was in turn induced by a signaling cascade staring with type I interferon. Confirming previous findings in asymptomatic adults with malaria [
29], no such increase was observed in the present study, suggesting that a threshold inflammatory response that includes type I interferon may be necessary to trigger expansion of CD141
+ mDCs through Flt3 ligand.
As observed previously with a subset of the data [
30], DCs isolated from uninfected Malian adults were able to up-regulate the activation markers HLA-DR and CD86 upon stimulation with parasite lysate. Likely due to the larger sample size, here CD80 and CD40 were also found to be significantly up-regulated. Similar to responses seen with DCs from malaria-naïve US donors [
30], mDCs obtained from Malians did not secrete significant amounts of the inflammatory cytokines IL-1β, IL-6 or TNF. Unlike the cytokine responses exhibited by DCs from US donors however, a slight but significant increase in IL-10 was measured when DCs from Malian individuals were stimulated with
P. falciparum-iRBC lysate. Whether mDCs from malaria-naïve US individuals are also capable of responding to the parasite with IL-10 secretion will have to be addressed in future experiments. The absence of a significant increase in the previous study could be due to a lower sample size and does not permit further conclusions at this time. Comparable to the response exhibited by DCs from US donors to the parasite, DCs from Malians secreted the Th1-associated chemokines CXCL9 and CXCL10 as well as CCL2 but no CCL5. Except for IL-10 secretion, in vitro DC responses to the parasite in malaria-exposed uninfected Malians seem, at least qualitatively, comparable to malaria-naïve US DCs. Future studies will have to address whether up-regulation of costimulatory molecules and chemokines in these DCs translates into effective T cell priming in malaria-endemic areas.
A number of studies have assessed cytokines in the plasma of acute malaria patients, concluding that the so called “cytokine storm”, characteristic for
P. falciparum infections, contributes to pathology [
54]. Here, an increase of most inflammatory cytokines tested during a first febrile malaria episode of the season in Malian children compared to their baseline and 7 days post treatment (convalescence) was also found. The data corroborates previous observations for IL-18, IFNγ and IL-10 in the same cohort and adds IL-6 and IFNα as being up-regulated while IL-1β, IL-17A and TNF remained unchanged [
55]. TNF levels were previously shown to be elevated during malaria in this cohort. This discrepancy is surprising and might be due to a difference in age, the present study assessed children between the ages of 6 and 9 while the previous study focused on children between the ages of 3 to 12. Another difference was the year of the blood draw, 2011 vs 2013, and kit used to analyse cytokine levels in plasma. This highlights the importance of repeated assessment of parameters at different time points to be able to draw accurate conclusions.
Plasma chemokine levels during human malaria are less well documented. Similar to other studies in Africa [
56,
57], this study reports high levels of CXCL9 and CXCL10 in the plasma of children with acute malaria and significantly higher CXCL9 levels in asymptomatically-infected compared to uninfected adults. CXCL10 has been shown to have a negative effect on the development of immunity and promotes severe outcomes in murine [
58,
59] and human [
60] malaria. Similar to this study’s findings with mDCs stimulated with
P. falciparum-iRBC lysates
, CXCL9 and CXCL10 have been reported to exhibit increased expression in murine splenic DCs during
Plasmodium chabaudi infection, with no significant increase in inflammatory cytokine gene expression [
61]. Comparable to the findings in this study, mDCs from malaria-naïve individuals secrete high levels of CXCL9 and CXCL10 in response to the parasite in vitro. Interestingly, these chemokines are induced to a greater extend by the parasite than by LPS, the gold standard DC activator [
30], suggesting that these chemokines may be characteristic of
P. falciparum infections. Similar to the “cytokine storm”, inflammatory chemokines like CXCL9 and CXCL10 are likely important to mount an effective immune response to the parasite early during the infection, but may be detrimental under certain circumstances.
In addition to CXCL9 and CXCL10, we observed that low levels of IL-10 were secreted by mDCs in Malian adults upon stimulation with the parasite. That IL-10 was detectable in the plasma of asymptomatically infected adults and at high levels in children with acute malaria is consistent with an important role for this regulatory cytokine in malaria. Indeed, Il-10 has been extensively described in murine malaria models and human studies as an important regulatory factor (reviewed in [
62]). Although commonly believed to be produced mainly by Tr1 cells, our data suggests that mDCs may be a significant contributor to the overall IL-10 response during
Plasmodium infection.
Studies that assess plasma cytokine or chemokine levels in asymptomatic malaria infections are scarce. In this study, IL-10 and CXCL9 were significantly elevated in asymptomatic
P. falciparum-infected Malian adults. While children with asymptomatic malaria have increased IFNγ, TNF and IL-4 levels [
63], IL-10 and CXCL10 were elevated in asymptomatically infected pregnant women [
64‐
66]. This study’s data suggest that asymptomatically infected adults in endemic areas exhibit elevated levels of IL-10 as well. CXCL9 was not assessed by these studies and, together with IL-10, might represent a potential biomarker to detect asymptomatic infections.
Limitations of this study include the low number of infected individuals enrolled, the cross-sectional design and the lack of data on DC-T cell interactions. Future studies should include a larger sample size, and longitudinal analysis of DC responses and functionality that include T cell activation. The Malian study site is particularly well suited for such studies due to the sharply demarcated and predictable malaria transmission season which enables the analysis of DCs from uninfected individuals before the malaria season and the same individuals as they become infected during the ensuing malaria season. Despite these limitations, the present study provides valuable insight into baselines DC frequencies and responses as well as in vivo cytokine and chemokine responses to the parasite in a malaria-endemic setting.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.