Previous malaria exposure attenuates monocyte-driven inflammation and correlates with modulation of the B cell response

Malaria remains a major global health concern, with estimates of 263 million cases and 597,000 deaths in 2023, according to WHO estimates (1). Plasmodium falciparum accounts for the vast majority of these instances and deaths. Natural immunity to malaria develops over time with repeated malaria episodes and has been historically linked to a potent antibody response against the parasite (2). Meanwhile, protection against severe malaria and immune regulation that limit immunopathology, often referred to as disease tolerance, develops more rapidly (3).

The host’s immune response is partly responsible for pathogenesis, as the parasite triggers systemic inflammation, driven by innate cells such as monocytes. Human monocytes are heterogeneous and can be classified into 3 major subsets: classical (CD14⁺CD16⁻), nonclassical (CD14^dimCD16⁺), and intermediate (CD14⁺CD16⁺) based on distinct surface protein markers (4). Monocytes are crucial for the initial sensing of P. falciparum infection through pattern recognition receptors (PRRs), such as TLR9, which detect pathogen-associated molecular patterns (PAMPs), such as unmethylated CpG DNA (5). Through their recognition of the pathogen, they play a key role in the response to the infection, indirectly by secreting proinflammatory mediators to recruit and activate other immune cells and directly by antiparasitic mechanisms of phagocytosis (6). Another effect of phagocytosis is that monocytes can present antigens to T cells, facilitating the activation of the adaptive immune response.

Monocytes play an important role in balancing protection against malaria through antiparasite mechanisms such as phagocytosis, cytokine production, and antigen presentation, while also contributing to pathology by producing inflammatory cytokines that cause systemic inflammation and vascular dysfunction (5).

Through the secretion of inflammatory mediators, monocytes can also drive inflammation and contribute to the sequestration of infected red blood cells (iRBCs) in organs such as the brain, placenta, and lungs by secreting cytokines that upregulate endothelial adhesion receptors (6). Especially, CD14⁺ monocytes are predominant cellular sources of cytokines and chemokines associated with severe malaria, such as TNF, macrophage inflammatory protein-1 α (MIP-1α), and MIP-1β, and IFN-inducible protein 10 (IP-10, encoded by the CXCL10 gene) (7, 8). In mouse models, the latter has also been associated with impaired parasite control during the blood stage of the infection (9).

An important aspect of keeping a balanced immune response is the monocytes’ ability to undergo epigenetic and transcriptional changes after initial exposure to malaria, leading to modulated responses upon subsequent infections, also termed innate training (5, 6). These changes can lead to 2 distinct and opposing states: trained immunity or tolerance. This reprogramming is induced by the engagement of various PRRs with specific microbial ligands. Trained immunity is characterized by enhanced cytokine production upon restimulation, while tolerance results in decreased cytokine production (10).

Previous research has highlighted the functional modulation of monocytes in response to repeated malaria infections, including the induction of innate immune training or tolerance programs following malaria exposure (11–17). We observed differences in monocyte abundance among individuals with naturally acquired P. falciparum infection (18). Previously exposed individuals exhibited a higher number of CD16⁺ monocytes but experienced overall reduced systemic inflammation (18). To further understand these adaptations, we investigated longitudinal transcriptional profiles of monocytes in a cohort of returning travelers with and without prior malaria exposure. We further connected the transcriptomic data with previously generated proteomic and cell profiling data from the same individuals, followed by mechanistic studies and serological analysis, to elucidate how monocytes during acute malaria correlate with modulation of the overall immune response, especially B cells.

The transcriptome of monocytes is affected by clinical P. falciparum infection. We investigated transcriptional changes in monocytes during acute malaria compared with healthy state 12-month control samples, by examining their gene expression profiles. We utilized a recently described targeted multi-omics single-cell dataset of peripheral blood mononuclear cells (PBMCs) in a cohort of returning travelers (19). The overall cohort was composed of returning travelers infected for the first time (primary infection) or with a previous history of malaria infection (previously exposed). The study participants presented with clinical malaria at the Karolinska University Hospital where they were included in the study. Samples were collected at treatment initiation or the day afterward, followed by additional sampling at 10 days and 1, 3, 6, and 12 months (Supplemental Table 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.199206DS1). For the multi-omics single-cell dataset, 2 donors with primary infection and 2 donors with previous malaria exposure were analyzed.

Based on previously annotated immune cell subsets (Figure 1A and Supplemental Figure 1A), we extracted all 14,940 monocytes for in-depth analysis (Figure 1B) originating from 12 samples of 4 individuals with 3 time points each. In this study, “acute” refers to the sampling at diagnosis/treatment of symptomatic P. falciparum malaria, “D10” to the visit 10 days after treatment, and “M12” to the convalescent visit 12 months later. The multi-omics character of the data enabled us to also assess the surface protein abundance of commonly used markers for the 11,602 CD14⁺ monocytes and the 3,338 CD16⁺ monocytes (Figure 1C and Supplemental Figure 1, B and C).

Figure 1

Transcriptomics profiles of monocytes during acute malaria. (A) UMAP representation of 73,121 CD45⁺ immune cells during and after acute malaria based on weighted nearest neighbor integrated targeted transcriptomic and epitope single-cell profiling, previously published by Lautenbach et al. (19). (B) Digital isolation of monocyte subsets of all 3 sampling time points; during acute P. falciparum malaria (Acute), approximately 10 days after treatment (D10) and at a healthy state 1 year after disease (M12). (C) Density plots for monocytes during acute malaria (only acute samples) for typical monocyte surface markers CD14 and CD16. (D) Hallmark gene expression sets enriched in monocytes of acute compared with the M12 healthy state (n_Acute = 4, n_M12 = 4). Red indicates higher in acute and blue higher in M12. FDR corrected P values were calculated using camera. See also related Supplemental Figure 1.

First, we assessed the differentially expressed genes (DEGs) for both monocyte subsets in the acute disease phase (Acute) compared with 10 days after treatment initiation and the healthy state after 12 months (M12). Due to the low number of donors, relatively few DEGs were identified using pseudobulk analysis. We did observe more DEGs in CD16⁺ compared with CD14⁺ monocytes, with higher expression of C1QB, S100A9, CD63, and AQP9 at the acute time point compared with 12 months (Supplemental Figure 1E). To further investigate the overall gene expression patterns, we assessed hallmark gene sets. Monocytes during acute malaria were associated with an upregulation of IFN-α response and protein secretion and reduced gene sets for hypoxia and glycolysis (Figure 1D).

Previous malaria exposure alters the transcriptional response during acute malaria. Next, we investigated the effect of previous malaria exposure on the transcriptomic response during acute P. falciparum malaria. We compared the summarized gene expression (pseudobulk) of monocytes from 2 primary infected and 2 previously exposed individuals during the acute disease phase. This revealed that both CD14⁺ and CD16⁺ monocytes from primary infected individuals exhibited overall higher gene expression than previously exposed donors (Figure 2A). For the CD16⁺ monocytes, however, some genes such as NKG7, ADGRE1, and VMO1 were more expressed in cells from previously exposed individuals compared with cells from primary infected individuals (Figure 2A and Supplemental Figure 1E). To determine if the changes in expression was associated with specific functional programs, we assessed hallmark gene sets (Figure 2B). Consistent with overall higher expression in primary infected individuals, more gene sets were enriched in both CD14⁺ and CD16⁺ monocytes. Among top gene sets were responses to IFN-γ and IFN-α, but also gene sets associated with inflammation and cell division (Figure 2B).

Figure 2

Previous malaria exposure affects transcriptomics profiles of monocytes during acute malaria. (A) Gene expression in individual cells of CD14⁺ and CD16⁺ monocytes at acute malaria, from primary infected individuals (n = 2) and previously exposed individuals (n = 2). Mean expression from 3,043/2,583 cells (CD14⁺ monocytes) and 640/1,426 (CD16⁺ monocytes). (B) Hallmark gene expression sets in acute CD14⁺ and CD16⁺ monocytes in previously exposed compared with primary-infected individuals. (C) Plasma protein levels of IFN-γ response hallmark genes during acute malaria for previously exposed (n = 48, orange boxes) and primary infected (n = 24, purple boxes) individuals of the same cohort (19), measured with Olink Explore1536 and shown as normalized protein expression (NPX) values. Statistical analysis for C was done by Wilcoxon rank sum with Benjamini-Hochberg correction for multiple testing with P values indicated. See also related Supplemental Figure 1.

To investigate whether the monocyte gene expression differences we observed were reflected at the systemic protein level, we repurposed a proteomic dataset generated from plasma samples from the same individuals and additional individuals from the malaria cohort of returning travelers (19). We specifically assessed the relative protein levels from genes included in the IFN-γ response gene set (Figure 2C). Out of 48 proteins, 20 were significantly different (P < 0.05, Benjamini-Hochberg correction), with all being more highly abundant in primary infected individuals (Figure 2C). These proteins included the IFN-γ response proteins CXCL9, -10, and -11; the cytokine receptors/adaptors IL-15RA, IL-18BP, and IL-2RB; cytotoxic granzyme A; and the HLA-associated molecule CD74 (Figure 2C and Supplemental Figure 1E). In summary, previous malaria exposure was associated with monocyte responses characterized by reduced production and secretion of proinflammatory mediators.

Monocyte functions are differently modulated by primary infection or previous malaria exposure. Next, we wanted to understand the functional implication of changes in the monocyte populations between primary-infected and previously exposed individuals in response to stimulation and to their microenvironment. To test this, we used iRBCs to mimic the physiological stimuli of blood-stage malaria to assess monocyte responses. We stimulated PBMCs from primary (n = 5) and previously exposed (n = 6) individuals for 24 hours using acute and convalescent (M12) samples, and we and measured intracellular TNF and CXCL9 levels and membrane levels of lysosomal-associated membrane protein-1 (LAMP-1/CD107a) as a surrogate for overall protein secretion (Supplemental Figure 2A). No differences were observed for TNF or CXCL9 (Supplemental Figure 2B) while we observed reduced CD107a levels on CD14⁺ monocytes from previously malaria-exposed individuals during clinical malaria that was normalized at convalescence (Figure 3A). Although the overall CD107a levels were high in primary-infected individuals during acute malaria, the response was further elevated in response to iRBCs compared with stimulation with uninfected RBCs (uRBCs) (Figure 3B). Assessing the effect of stimulation on monocyte numbers, there was no effect from adding uRBC or iRBC, while PMA/ionomycin led to significantly fewer monocytes (Supplemental Figure 2, C and D).

Figure 3

Previous malaria exposure is reflected in differences in plasma protein profiles. (A) CD107a⁺ monocytes after stimulation with uninfected (uRBC) or infected red blood cells (iRBC). PBMCs from primary infected (n = 5) and previously exposed individuals (n = 6) from acute and matched 12-month (convalescent) sample time points were stimulated, and degranulation was assessed via surface staining for CD107a. Data are pooled from 2 independent experiments. Primary infected individuals were compared with previously exposed individuals in A and acute samples versus convalescent samples in B. (C–E) PBMCs isolated from healthy donors (n = 11) stimulated with antibody-depleted plasma pools from the acute or convalescent time points for primary infected and previously exposed individuals. A ratio indicating geometric mean fluorescence intensity surface expression changes is shown for classical monocytes (C), intermediate monocytes (D), and nonclassical monocytes (E). Data were pooled from 2 independent experiments. Statistical analysis in A was done using a mixed effects model matched on donor, followed by Šídák’s post hoc test to correct for multiple comparisons. Statistical analysis in B was calculated using 2-tailed matched pair 2-way ANOVA followed by Šídák’s post hoc test for multiple comparisons. Statistical analyses in C–E was done using 2-tailed Student’s t tests; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. See also related Supplemental Figures 2 and 3. Data represent individual donors.

To test the effect of the microenvironment on monocyte function, we pooled acute or 12-month convalescent plasma from primary infected donors (n = 8) and previously exposed donors (n = 8) and depleted IgG antibodies to reduce antibody-mediated cytotoxic effects (20). We then cultured PBMCs isolated from healthy blood donors with 20% antibody-depleted patient plasma for 24 hours and assessed levels of CD64 (FcgRI), CD11c, HLA-DR, CD86, and CD69, as indicators of functional modulation and activation (Supplemental Figure 3A). The culture did not affect overall monocyte frequencies or cell counts between plasma pools, although there were significantly more intermediate and nonclassical monocytes detected in PBMCs cultured with acute plasma from primary infected donors (Supplemental Figure 3, B and C). To assess protein surface levels, geometric mean fluorescence intensity (geoMFI) ratios were calculated between cells stimulated with acute and convalescent plasma. The ratio was then compared between cells receiving plasma from primary infected or previously exposed individuals (Figure 3, C–E, and Supplemental Figure 3, D–F). HLA-DR and CD11c only displayed minor changes between pools in classical monocytes, with a slight increased ratio of HLA-DR and reduced ratio of CD11c in the pool from previously exposed individuals (Figure 3C). CD86 had a consistently higher ratio in the pool from previously exposed individuals for all monocyte subsets, although geoMFI changes were mainly observed for classical monocytes, while CD64 and CD69 had a consistently higher ratio and geoMFI in the pool from primary infected individuals (Figure 3, C–E).

In summary, monocytes from individuals previously exposed to malaria displayed reduced baseline activation and response to stimulation, while the plasma stimulated increased levels of the costimulatory molecule CD86. In contrast, monocytes from primary infected individuals displayed high baseline activation and exocytosis and a plasma profile promoting monocyte activation, indicated by increased surface levels of CD64 and CD69.

Previous malaria alters circulating protein profiles during acute malaria. Due to the reduced monocyte responses in previously exposed donors upon ex vivo iRBC stimulation and different effect of plasma pools from primary infected and previously exposed donors on cellular responses, we thought to further investigate differences in protein profiles associated with previous malaria exposure. For a more accurate analysis, we focused on individuals with available protein profiling data at both the acute and 12-month time points. Between primary infected (n = 10) and previously exposed (n = 21) individuals, there were no differences in age (P = 0.5), sex (P = 0.4), or parasitemia (% iRBCs; P = 0.4) (Supplemental Table 1). We identified 40 differentially abundant proteins (DAPs) between the groups at acute, compared with their 12-month follow-up, using a linear mixed-effect model approach to account for individual patient effects (Figure 4, A and B).

Figure 4

Plasma protein differences associated with malaria in individuals with primary infection versus previous exposure. Differential protein abundance between primary infected (n = 10) and previously exposed individuals (n = 21), based on a linear mixed-effects model accounting for interdonor variation, with FDR correction (Benjamini-Hochberg, 0.05) applied throughout. (A) Volcano plot comparing NPX values between primary infected and previously exposed donors (controlling for donor and time) and highlighting the top 10 differentially abundant proteins (DAPs) ranked by adjusted P value and NPX difference. (B) Protein abundance of the top 10 significant proteins (all have adjusted P ≤ 0.05) across acute and convalescent (M12) states. (C) Heatmap of ΔNPX (Acute–M12) for all significant DAPs, annotated with NPX differences and –log₁₀(adjusted P values). Parameter estimate indicates the estimated difference in protein abundance.

Overall, we observed that most proteins (29 of 40) were higher in primary infected individuals, including cytokines (TNFSF13B, TNFSF13, FASLG, CSF1), enzymes (ASAH2, GALNT10, GZMA), and receptors (PTPRF, CRLF1, CD27) (Figure 4C). In contrast, only 11 proteins were higher in previously exposed individuals, including chemokines (CXCL16, CCL14), receptors (FCRL5), enzymes (FUCA1), and enzyme inhibitors (SPINK4, SPINK6).

Among the 40 proteins, 7 (TNFSF13B, CXCL16, FUCA1, ALDH1A1, TNFSF13, AXL, and TIGAR) have been linked to monocytes and DCs via blood gene expression profiles generated by the Human Protein Atlas (proteinatlas.org) and are indicated by arrows in Figure 4C. Among them, TNFSF13B, or B cell activating factor (BAFF), showed the greatest difference between primary-infected and previously exposed individuals with 4-fold higher abundance in primary-infected compared with previously exposed individuals.

Association between monocyte and DC-produced BAFF and BAFF-R levels on B cells. The significant differences in circulating TNFSF13B (BAFF) levels associated with previous malaria exposure prompted an investigation into potential immune cell sources. To approximate this, we analyzed gene expression levels from the single-cell RNA-seq dataset in various immune cell subsets. Comparing the average gene expression per immune cell subset from all 4 donors with acute malaria, we found that myeloid cells (CD14⁺ and CD16⁺ monocytes, mDCs, and pDCs) exhibited the highest average expression among PBMCs (Supplemental Figure 4A), corroborating existing literature (21). The gene expression of TNFSF13B in mDCs, CD14⁺, and CD16⁺ monocytes showed a decreasing trend across the sampling time points (Acute, 10 days, 12 months), consistent with higher expression during acute infection (Supplemental Figure 4, B and C). To further illustrate this temporal pattern, we compared the average gene expression levels between acute and 12-month samples, which showed higher average TNFSF13B expression across samples during the acute phase, most prominently in mDCs with a 0.48-fold increase, followed by CD16⁺ monocytes at a 0.37-fold increase and CD14⁺ monocytes at a 0.35-fold increase (Supplemental Figure 4D).

BAFF is the sole ligand for the BAFF-receptor (BAFF-R), encoded by the TNFRSF13C gene, and serves as a critical prosurvival receptor for B cells (21). The TNFRSF13C gene expression was low for B cells in our cohort, while the surface protein expression was high for naive and memory B cells but was low for plasma cells (Supplemental Figure 5, A–C). To validate these observations, we downloaded and repurposed a previously published malaria single-cell RNA-seq dataset (22) including samples from individuals with acute malaria (day 0), and days 7 and 28 after treatment, as well as from noninfected controls (Supplemental Figure 5D). We observed relatively even BAFF-R expression levels over time and compared with noninfected controls. Consistent with our own dataset, TNFRSF13C expression was low in plasma cells compared with other B cells.

To further explore the potential association between BAFF and the BAFF-R in clinical malaria, we reanalyzed previously generated flow cytometry data from 50 individuals of the same traveler cohort (23), where we examined BAFF-R expression on B cell subsets from acute infection through to 12 months after infection (Supplemental Figure 6A). During acute malaria, BAFF-R levels on circulating B cells were significantly depleted, with levels returning to those of healthy controls (n = 14) approximately 3–6 months after infection (Figure 5, A and B). This pattern was similar across B cell subsets (Supplemental Figure 6, B and C). When comparing BAFF-R levels on B cells from primary (n = 17) and previously exposed (n = 33) individuals during acute infection, we observed significantly lower levels in the primary infected group (P = 0.0068) (Figure 5C). This was also reflected in the surface expression of single-cell data (Supplemental Figure 5C).

Figure 5

BAFF-R levels on B cells are reduced during acute malaria and highly negatively correlated with circulating BAFF levels. (A) Representative FACS plots of BAFF-R expression on total live lymphocytes at acute P. falciparum malaria or after 12 months (left and middle panel) and for CD19⁺ CD20⁺ B cells at all time points (right panel). (B) Longitudinal assessment of BAFF-R expression on CD19⁺CD20⁺ B cells at and after acute P. falciparum malaria. The points denote the mean with error bars showing the SEM. The gray lines indicate data from individual donors. Healthy individuals (Ctrl) as a reference with dotted line of their mean (n = 14) (C). Comparison of CD19⁺CD20⁺BAFF-R⁺ B cells during acute malaria for primary infected (purple, n = 15) and previously exposed individuals (orange, n = 28). Mean of control group as dotted line as in B. (D) Relative protein abundance NPX of BAFF (TNFSF13B) in primary infected and previously exposed individuals over the sampling time points, assessed using linear mixed-effects models accounting for repeated measures. *P < 0.05, ****P < 0.0001. (E) Association between plasma BAFF (TNFSF13B) levels and CD19⁺CD20⁺ B cell surface BAFF-R (MFI) levels during acute infection using a multiple linear regression model with BAFF and prior exposure status as predictors. The correlation coefficient (R²) was corrected for multiple testing (BAFF, exposure, BAFF × exposure). The line represents the fitted linear regression model with 95% CI. See also related Supplemental Figure 6. Data represent individual donors.

BAFF protein level dynamics in plasma over the 3 sampling time points (Acute, 10 days, and 12 months) for primary infected and previously exposed individuals confirmed that primary infected individuals exhibited highly elevated levels in response to malaria. In contrast, previously exposed individuals displayed only slightly elevated levels during acute malaria compared with at 12 months (Figure 5D). To formally assess the relationship between plasma BAFF and B cell BAFF-R levels, we fitted a multiple linear regression model including BAFF and prior exposure status as predictors (Figure 5E). Plasma BAFF was significantly associated with BAFF-R expression (β = −1,895, P = 5.8 × 10^–7), whereas prior malaria exposure was not independently associated with BAFF-R expression after adjustment for plasma BAFF (P = 0.79). Inclusion of a BAFF × exposure interaction term did not improve model fit (P = 0.30), indicating that the association between plasma BAFF and BAFF-R expression was similar in primary infected and previously exposed individuals. These observations are consistent with previous findings from a controlled human malaria infection study, where primary malaria was associated with increased plasma BAFF levels and an inverse correlation between plasma BAFF and B cell BAFF-R expression (24).

BAFF-R downregulation is mediated by soluble plasma factors. The better understand which factors that could modulate BAFF-R levels on B cells, we used in vitro experimental approaches. We first assessed if soluble factors within the patient plasma could modulate BAFF-R levels. This was done by culturing healthy PBMCs with plasma pools from primary infected or previously exposed patients at acute malaria or convalescence for 24 hours followed by gating for B cells (Figure 6A).

Figure 6

BAFF-R levels are affected by the plasma microenvironment and soluble BAFF. (A) Schematic illustration of the in vitro experimental setup for plasma stimulations of PBMCs. Plasma was pooled from primary infected (n = 8) or previously exposed (n = 8) donors while PBMCs were isolated from healthy blood donors (n = 11). (B) Geometric mean fluorescence intensity (geoMFI) of BAFF-R levels on gated B cells 24 hours after culture with indicate plasma pools. (C) Comparison of the BAFF-R geoMFI ratio for acute over 12-month plasma for primary infected and previously exposed donors. Statical analysis in B was done using a paired 2-tailed Student’s t tests and, in C, an unpaired 2-tailed Student’s t test. ***P < 0.001, ****P < 0.0001. (D) Schematic illustration of the experimental setup for recombinant BAFF stimulations of purified B cells. (E) BAFF-R geoMFI ratios between nonstimulated and BAFF, anti-Ig+CpGC, or anti-Ig+CpGC+BAFF stimulated sorted B cells after 2 days (n = 11) (E) or 4 days (n = 4) (F) of culture. Statistical analyzes in E and F were done using repeated-measures 1-way ANOVA followed by Šídák’s post hoc test with *P < 0.05, **P < 0.01. See also related Supplemental Figure 7. Data represent individual donors.

B cells from PBMCs stimulated with plasma were gated as indicated in Supplemental Figure 3 with an additional gate for BAFF-R⁺ cells (Supplemental Figure 7A). The stimulation had no effect on overall B cell numbers or BAFF-R⁺ cells (Supplemental Figure 7, B and C). In addition to BAFF-R levels, we also assessed surface expression of HLA-DR, CD11c, FcRL5, CD69, and CD86. Of these, only plasma from primary infected individuals at the acute time point could promote significant changes in B cell surface expression. FcRL5 displayed a modest upregulation compared with convalescent plasma, although the geoMFI levels remained similar to those from previously exposed plasma (Supplemental Figure 7, D and E). BAFF-R levels, however, were robustly reduced in both geoMFI and in ratio compared with plasma from previously exposed donors (Figure 6, B and C).

Since the plasma pools from primary infected individuals contained high levels of BAFF at the acute time point compared with previously exposed donors and convalescent time points (Supplemental Figure 7F), we next assessed if recombinant BAFF could be the driver of BAFF-R downregulation. We sorted total B cells from healthy individuals and stimulated them with BAFF alone, or in combination with anti-Ig (BCR ligand) and CpGC (TLR9 ligand), to mimic B cell activation during malaria infection (Figure 6D). After 2–4 days of culture, we compared BAFF-R levels between nonstimulated and stimulated cells. We observed that B cells stimulated with BAFF exhibited significant downregulation of BAFF-R (P < 0.05) compared with nonstimulated cells; however, B cells stimulated with BCR/TLR9 ligands initially (day 2) showed significant upregulation of BAFF-R (P < 0.05) compared with nonstimulated cells (Figure 6E) but then reduced overall BAFF-R levels at day 4 of culture (gated on nonplasmablasts) (Figure 6F). B cells stimulated with both BCR/TLR9 ligands and BAFF showed no significant difference in BAFF-R levels (P > 0.05) compared with cells stimulated with BCR/TLR9 ligands alone (Figure 6E). As expected from our own and previous data (25), we observed low BAFF-R levels on plasmablasts on day 4 of stimulation (Supplemental Figure 7G). Further analysis of the BAFF-R levels together with other B cell markers on day 2 and 4 of stimulation indicated that BAFF-R–low cells also displayed reduced levels of CD19 but normal levels of CD20 and no increased staining for a cell death marker (Supplemental Figure 7, H–J). This phenotype was not observed in B cells from malaria patient samples (Supplemental Figure 7I). Such a population was also seen when cultured with patient plasma, although the reduced BAFF-R levels observed were gated from the BAFF-R⁺ CD19 high population and not CD19 low cells (Supplemental Figure 7J).

In summary, the plasma from acute malaria in primary infected individuals, but not previously exposed donors, could drive BAFF-R downregulation. This effect could not be completely replicated through the addition of recombinant BAFF to B cells, indicating that additional factors or different oligomeric forms of BAFF in the plasma could be important in mediating the BAFF-R downregulation.

High levels of BAFF in circulation are associated with reduced parasite-specific antibody responses. To investigate the association between circulating BAFF and humoral immunity during acute malaria, we fitted linear mixed effects models including plasma BAFF levels, exposure group, and BAFF × exposure interaction terms and evaluated association with parasite-specific antibodies. Antibody responses were summarized using a cumulative response score (CRS) across 5 P. falciparum merozoite antigens (AMA1, MSP1-19, MSP2, MSP3, and GAMA). No significant interaction was observed for any antibody isotype (IgG: P = 0.996, IgG1: P = 0.675, IgG3: P = 0.453, IgM: P = 0.960), indicating that the association between BAFF and antibody responses was comparable across exposure groups. BAFF levels were negatively associated with CRS for IgG (R²_adj = 0.48, P = 0.00079), IgG1 (R²_adj = 0.54, P = 0,0029), IgG3 (R²_adj = 0.51, P < 0.001), and IgM (R²_adj = 0.24, P = 0.02) (Figure 7). Previous malaria exposure affected total IgG (P = 0.016) and IgG1 (P = 0.00046) — but not IgG3 and IgM. We next examined whether parasitemia could be a driving factor to explain either the BAFF (Supplemental Figure 8A) or antibody levels (Supplemental Figure 8B). However, no significant correlations were observed in either case, indicating that the negative association between circulating BAFF and antibody levels was not driven by parasitemia.

Figure 7

Circulating BAFF levels are inversely correlated with malaria-specific IgG responses. Linear regression analysis of acute BAFF levels (TNFSF13B NPX) and Cumulative Response Score (CRS) of total IgG, IgG1, IgG3, and IgM responses to 5 merozoite antigens in matching individuals (total n = 55, with primary infected n = 20 and previously exposed n = 35). BAFF × exposure interactions were determined for primary infected (purple) and previously exposed (orange) individuals. Lines represent fitted linear regression models with 95% CIs. Correlation coefficients (R²) were corrected for multiple testing (BAFF, exposure, BAFF × exposure). See also related Supplemental Figure 8. Data represent individual donors.

Together, these findings suggest that elevated BAFF during acute infection may be associated with less effective antibody responses in individuals experiencing their first malaria exposure.

Monocytes play a crucial role in balancing the antiparasite response and contributing to systemic inflammation and pathology in malaria (5). While disease-protecting immunity develops through a broad antibody repertoire, protection from severe disease arises more rapidly, likely through a combination of increasing humoral responses and immune regulatory changes that limit immunopathology (18). Here, we analyzed immune response data from individuals with acute P. falciparum malaria, focusing on the effect of previous malaria exposure. We report transcriptional changes in monocytes that influence the acute inflammatory response and immunomodulatory mechanisms that could have an effect on the B cell response during malaria.

We reported previously that individuals with prior exposure to malaria were better able to control parasitemia through the action of cytophilic (IgG1 and IgG3) antibodies in conjunction with CD16⁺ monocytes, contributing to less systemic inflammation (18). The transcriptional changes in monocytes observed here support these findings. Higher expression of CXCL16, a membrane-bound chemokine, functions as a scavenger receptor on macrophages, binding phosphatidylserine (PS) and oxidized lipoprotein (26). The exposure of PS on the plasma membrane of cells is a known signal of apoptosis, enabling phagocytes to recognize dying cells (27). Infected erythrocytes are reportedly also able to expose PS on their outer surface (28). Enhanced recognition and clearance of apoptotic cells via membrane-bound CXCL16 may help remove damage-associated molecular patterns (DAMPs) and endogenous factors that could trigger more inflammation. Conversely, higher plasma levels of soluble CXCL16, a chemoattractant for activated CXCR6⁺ T cells, may result from increased membrane-bound CXCL16 on monocytes, shed through ADAM10-mediated cleavage (29).

Transcriptomic profiles of monocytes indicated reduced CXCL10 expression in previously exposed donors compared with primary-infected individuals. CXCL10, a proinflammatory chemokine, recruits CXCR3⁺ leukocytes, contributing to inflammation and tissue damage (30). Monocytes from previously exposed individuals showed almost no CXCL10 expression, correlating with significantly lower plasma levels. CXCL10 interacts with CXCR3, which is expressed by atypical B cells (23, 31, 32), plasmacytoid dendritic cells (pDCs), and activated effector CD4⁺ T cells during acute malaria (19). Reduced CXCL10 availability may therefore decrease the activation and recruitment of these immune cells, reducing overall immune activation, while it amplifies monocyte inflammatory responses (30), potentially causing more host damage in primary infected individuals. The production and secretion of CXCL10 is IFN-γ dependent (8), suggesting that the observed lower levels of IFN-γ in previously exposed individuals potentially explain the reduced CXCL10 expression and secretion (18). IFN-γ has also been suggested to modulate myeloid cell training and tolerance responses (33), potentially explaining the lower monocyte degranulation we observed following ex vivo iRBC stimulation. Disease tolerance in malaria has also been associated with metabolic changes (34). This was potentially also reflected in our data, where previously exposed individuals had reduced plasma levels of TP53-induced glycolysis and apoptosis regulator (TIGAR), which increases during cell stress to reduce glycolysis and increase available NADPH for DNA repair (35).

Although not investigated here, epigenetic changes in monocytes have been reported to be involved in immunomodulation and innate-driven tolerance (36). Epigenetic reprogramming of human monocytes from individuals exposed to repeated P. falciparum infections can shift the monocytes to become more regulatory (11). These changes contribute to attenuated inflammatory responses during subsequent malaria exposures, reducing the risk of severe inflammatory symptoms. These observations should also be interpreted in the context of early-life and environmental exposures, which may contribute to long-lasting immune conditioning of monocytes independently of malaria (37). One such factor could be cytomegalovirus (CMV) infection, which was proportionally higher in previously exposed donors in this study. Monocytes and stem cell progenitors have been described as reservoirs for latent CMV infection (38, 39), and activation of such in vitro infected cells showed less responsiveness to IFN-γ signaling (40).

Guha et al. showed that ex vivo–stimulated monocytes of Mali adults secreted less TNF, IL-1β, and IL-6 and more IL-10 compared with unexposed US adults (11). In our cohort, we observed reduced plasma levels of TNF and IL-6 in previously exposed individuals (although not significant) but no elevated levels of IL-10 compared with primary infected individuals (18).

Other cells, rather than monocytes, might be responsible for high blood IL-10 levels. Transcriptomic profiles suggest that regulatory T (Tr1) cells could be an important source of IL-10 during acute malaria (22). CD16⁺ monocytes express high TNF levels, potentially contributing to plasma TNF (19). In contrast, we did not observe strong IL-6 expression in monocytes during acute malaria; instead, IL-6 was primarily expressed by B cells (19). Endothelial cells have also been reported to contribute to circulating IL-6 (41), although this was not analyzed here. Collectively, our findings suggest that monocyte changes, through direct and indirect effects, contribute to dampened systemic inflammation linked to previous malaria exposure. This response modulation benefits disease protection by decreasing potential host damage and symptoms and may contribute to the development of a better humoral response.

One protein that was strongly influenced by previous malaria exposure was TNFSF13B (BAFF), which was significantly more abundant in primary infected individuals. The myeloid cells upregulated TNFSF13B expression during malaria, likely contributing to the higher circulating levels as they have been described to be the major sources of BAFF (42). BAFF is critical for immature B cell maturation and mature B cell maintenance but is also associated with plasma cell differentiation and survival (21, 25, 42). IFN-γ can upregulate BAFF expression in monocytes (43), suggesting that the elevated BAFF levels might result from the significantly higher IFN-γ levels observed in primary infected individuals. High IFN-γ and other proinflammatory mediators in primary infected individuals could represent a disadvantageous inflammatory response, impairing the induction of a long-lasting B cell response. This is supported by observations in other studies where individuals with a high antibody response had lower plasma BAFF, possibly due to reduced IFN-γ levels (44).

BAFF can bind to 3 receptors, BAFF-R, TACI, and BCMA. TACI and BCMA also bind the ligand APRIL, while BAFF-R is specific for BAFF (45–47). BAFF-R is highly expressed on immature and mature B cells, where it is needed for survival and maintenance (48). As B cells become activated and differentiate to plasma cells, they downregulate BAFF-R and upregulate BCMA and TACI and become more dependent on APRIL for survival (49). BAFF-R levels can be regulated through shedding via cleavage with ADAM10 and ADAM17, as reviewed by Meinl et al. (50). The cleavage is induced via binding by BAFF trimer or 60 mer to BAFF-R. The effect can also be induced in vitro via TLR9-ligation and is then dependent on ADAM10 (51). In this study, we observed BAFF-R downregulation on B cells from patients with malaria. We could recapitulate this effect upon culture of healthy PBMCs with plasma from primary infected donors. We also observed BAFF-R downregulation in response to BCR and TLR9 stimulation, although those cells also downregulated CD19, contrasting with the ex vivo patient B cells. This could be explained by CD19 internalization, which has been described as a mechanism to regulate B cell activation. In vivo, the levels of CD19 then restore via de novo production upon cessation of the stimuli (52).

In the context of malaria, impairment of the BAFF/BAFF-R axis has been linked to disease severity and poor malaria antigen-specific B cell differentiation (53). Similar to our study, Nduati et al. observed reduced BAFF-R levels on peripheral B cells (53). However, children with higher BAFF-R levels on their B cells maintained parasite-specific IgG over 4 months, indicating more long-lasting B cell responses in individuals with reduced inflammatory responses. The study concluded that the observation was due to dysregulation of BAFF-R on B cells, in favor of short-lived antibody responses (53). Here, we could also show a negative association between antibody levels in adults during the acute inflammatory response, where high BAFF levels were associated with lower levels of parasite-specific IgG and IgM antibodies. Although individuals with previous malaria exposure could retain memory B cells, contributing to more rapid antibody production, IgG3 and IgM antibody levels are less associated with preexisting long-lived memory. The negative correlation of these with BAFF levels, therefore, more directly indicates a possible negative effect of the inflammatory response on antibody production.

However, it remains possible that rapid extrafollicular IgM and IgG3 antibody production in naive individuals and memory-derived high-affinity IgG1 antibodies could be important drivers in modulating the early monocyte and systemic inflammatory response. Acute malaria has been associated with rapid and extensive expansion of atypical B cells (23), which have been suggested to be derived from extrafollicular reactions (54). In malaria, atypical B cells are enriched for unswitched B cell receptors (55). They also have high plasma cell differentiation potential in the early stages of atypical B cell differentiation (55). However, they have also been described to require inflammatory signals to be generated (56), making it difficult to determine what comes first and what role each component contributes with to the overall modulation of the host response, leaving the temporal and mechanistic relationships between these factors to be addressed by future studies.

Although BAFF levels did not correlate with peripheral parasitemia in our cohort, this does not necessarily contradict previous findings showing an association between BAFF-R expression in splenic B cells and parasitemia in patients with malaria (57). Differences in tissue compartment analyzed, time of sampling, and disease context may contribute to these divergent observations. Together with prior CHMI study (24), these findings suggest that elevated BAFF and reduced BAFF-R expression are common features of acute malaria, while their relationship to parasitemia may depend on anatomical compartment and disease context.

In conclusion, our findings highlight the central role of monocyte-mediated immune modulation in developing tolerance and humoral immunity to malaria. The observed changes in transcriptional and cytokine profiles suggest that prior exposure to malaria can have an imprinting effect on monocytes to balance effective parasite control with reduced systemic inflammation. Implementing these insights could enhance therapeutic strategies aimed at modulating the immune response to prevent severe malaria.

Limitations of the study A key limitation of this study is the influence of host genetic variability on immune responses, which is difficult to control for in prospective study designs. Furthermore, although single-cell transcriptomic profiling was performed on a large number of cells, the data were derived from only 4 individuals (2 per exposure group), potentially limiting generalizability. These individuals were selected to avoid extremes based on prior data, but differential gene expression may not reflect broader cohort trends. We could confirm overlapping phenotypic changes using an experimental model, but validation in larger, independent cohorts is warranted. The association between circulating BAFF levels and antibody responses is based on correlations. Conclusively determining if BAFF controls antibody secretion or vice versa, therefore, necessitates further functional experiments to investigate mechanistic connections.

Sex as a biological variable. The study examined immune responses during and after malaria with both sexes included, except for the single-cell RNA-seq data, which was based on males only. This was done to reduce variability, as the dataset is limited to 4 donors.

Study design/cohort. This study is based on new and repurposed data, previously generated on the peripheral blood of patients treated for P. falciparum malaria at Karolinska University Hospital. The cohort is well characterized and has been extensively described in the context of research questions related to antibody longevity in the absence of reexposure (58), B cell profiling (23), and antibody interplay (18); identification of antibody markers for recent malaria exposure (59); and protein-centric multi-omics blood profiling (19).

Single-cell multi-omics data set. We used the previously reported targeted single-cell multi-omics data, generated on PBMCs from malaria-infected individuals within the Karolinska cohort (19). In short, CD45⁺ PBMCs from 4 individuals (2 primary infected, 2 previously exposed) with 3 time points (Acute, D10, M12) were profiled on a transcriptional level using the Immune Response Panel (targeting 399 genes), as well as on a cell surface level using 29 AbSeq oligo-conjugated antibodies (BD Biosciences). A total of 73,121 cells were annotated based on the expression of 374 genes and 29 surface markers, resulting in 27 immune cell subsets as previously described (19). Here, we focused mainly on myeloid cells, annotated as pDCs, mDCs, CD14⁺, and CD16⁺ monocytes.

All downstream analyses were performed using the Seurat R package (v4) (60). To address interindividual variability and reduce biases associated with single-cell–level differential expression analyses, we adopted a pseudobulk approach.

Raw gene counts were aggregated at the donor level to generate pseudobulk expression profiles for each cell population. Differential expression analyses were performed using an edgeR-limma workflow. Lowly expressed genes were filtered using filterByExpr(), library sizes were normalized using the trimmed mean of M-values (TMM) method, and expression data were transformed using the voom procedure. Linear models were fitted using limma (61), and differential expression between time points (Acute vs M12, Acute vs D10, and D10 vs M12) was assessed using empirical Bayes moderated statistics. Hallmark pathway enrichment analysis was subsequently performed using CAMERA with MSigDB Hallmark gene sets, accounting for inter-gene correlation within gene sets (62).

Differential protein abundance analysis. Plasma protein data previously published by Lautenbach et al. (19) comprised longitudinal measurements collected during acute Plasmodium falciparum malaria and throughout convalescence in the same cohort of returning travelers. These data were used to assess changes in protein abundance across groups and sampling time points. Differential protein abundance analysis was performed using linear mixed-effects models implemented in the lmerTest package in R (63, 64), with time point included as a fixed effect and study participant as a random effect to account for repeated measurements within individuals.

Post hoc pairwise comparisons were performed using the emmeans package (65) to identify significant differences both between sampling time points and between exposure groups at the acute time point. For group comparisons, the contrast Acute primary_infected – Acute previously_exposed was evaluated. Resulting P values were adjusted for multiple testing using the Benjamini-Hochberg procedure to control the FDR (66).

Protein annotation. Proteins were annotated using the Human Protein Atlas, version 24.0, retrieved from https://www.proteinatlas.org/ DAPs were mapped based on “Secretome location” and “Secretome function” annotations.

Stimulation of PBMC with iRBC. Flow cytometry data from a previously generated functional experiment was reanalyzed. The experiment was designed to assess functional differences in gd T cells in PBMCs from malaria-infected individuals (18). We repurposed the data and gated on CD14⁺ monocytes investigating specifically the stimulation with uninfected and infected RBCs and with PMA (phorbol 12-myristate 13-acetate) and ionomycin. The iRBCs had been lysed through several freeze-thaw cycles prior to PBMC stimulation (Supplemental Figure 2).

Plasma pooling and antibody depletion. Plasmas were pooled from 8 donors with primary infection or 8 donors with previous malaria exposure with matched samples at the acute and 12-month time points. The pools were depleted of IgG antibodies using protein-G spin columns (Ab SpinTrap, Cytiva) as described previously (20). IgG depletion was confirmed using an in-house multiplex bead assay detecting total IgG. Briefly, High-binding 96-well plates (Costar) were coated overnight at 4°C with goat anti-human IgG (Merck, I2136; 1 μg/mL in PBS, 100 μL/well). Plates were washed 4 times with PBS containing 0.05% Tween-20 (PBST) and blocked with 1% BSA and 0.05% Tween-20 in PBS (pH 7.4; 100 μL/well) for 1 hour at room temperature (RT). Plasma pools (predepletion, postdepletion, and eluted IgG fractions) were serially diluted starting at 1:100, followed by 5-fold serial dilutions (8 total), and added in duplicate (100 μL/well). Plates were incubated for 30 minutes at RT. After washing, peroxidase-conjugated goat anti-human IgG (Jackson ImmunoResearch; 1:10,000 in blocking buffer) was added (100 μL/well) and incubated for 30 minutes at RT. Plates were washed 4 times with PBST, and TMB substrate (Thermo Fisher Scientific; 75 μL/well) was added for 5–10 minutes before stopping the reaction with 0.2M H₂SO₄ (75 μL/well). Absorbance was measured at 450 nm using a microplate reader (Thermo Fisher Scientific). Depletion was confirmed by comparing EC₅₀ values, to range from 90.5% to 99.1% (Supplemental Figure 9).

Sorting of B cells. For the in vitro cultures of B cells, the cells were sorted from buffy coats from healthy donors. Buffy coats were obtained from the Karolinska University Hospital blood bank, in Stockholm, Sweden. Complete media consisted of RPMI-medium supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, 100 U/mL of penicillin, 100 mg/mL of streptomycin, 10 mM HEPES, and 0.05 mM 2-Mercaptoethanol (all from Thermo Fisher Scientific). B cells were isolated through negative selection using either the RosetteSep Human B Cell Enrichment Cocktail (Stemcell Technologies) as previously described (67) or with MACS cell separation (Miltenyi Biotec). Briefly, 25 μL of the Cell Enrichment Cocktail was added for each mL of blood, followed by incubation for 20 minutes at RT. Then 2 volumes of PBS + 2% FBS were added to the blood. In total, 15 mL Ficoll–Paque Plus (GE Healthcare) was added to SepMate (Stemcell technologies) tubes and centrifuged at 1,000g for 1 minute at RT. The blood was poured into the SepMate tubes and centrifuged at 1,200g for 10 minutes with the brake on. The cell fraction was transferred to new tubes and washed twice with PBS + 2% FBS (300g, 10 minutes, brake at 5 of 9, RT).

For the MACS cell separation, the B cells obtained from the RosettSep isolation were incubated with a Biotin-Antibody cocktail (Miltenyi Biotec) targeting non-B cells for 5 min at 4°C and then for another 10 minutes (4°C) with the anti-Biotin Microbeads (Miltenyi Biotec). The cell suspension was then passed through LS columns placed in the MACS separator (Miltenyi Biotec). The B cell fraction was saved and washed as previously described. After the final wash, the isolated cells were resuspended in PBS + 2% FBS and counted using a Countess II (Invitrogen) following a 1:1 dilution with 0.4% trypan blue. The cells were centrifuged again (300g for 10 min) and resuspended at 5 × 10⁶ cells/mL in freezing medium (FBS + 10% DMSO) and placed inside a CoolCell at –80°C overnight, after which the cells were transferred to liquid nitrogen for long-term storage. B cell purity was assessed by flow cytometry and were on average ± SD 92.3% ± 6.6%.

In vitro stimulation of sorted B cells and fresh PBMCs. For the in vitro stimulation, previously sorted B cells were thawed and moved to a 15 mL Falcon tube, and 13 mL of complete media were added dropwise, the cell suspension was carefully mixed, and rested on ice for 20 minutes. The sorted B cells were then centrifuged at 300g for 5 minutes, washed once in complete media (300g, 5 minutes), resuspended in 1 mL complete media, and counted, as previously described. Stimulations of purified B cells were done in a 96-well round-bottom tissue-culture plate (Techno Plastic Products). The stimuli included α-Ig (IgA+IgG+IgM [H+L] AffiniPureF(ab’)2 Fragment goat-anti-human [Jackson ImmunoResearch Laboratories]) at 1.25 μg/mL, CpG-C (Invivogen) at 2.5 μg/mL, and Human BAFF (PeproTech) at 125 ng/mL. The stimuli were added to wells with 1 × 10⁵ cells in a total volume of 200 μL complete media.

Fresh PBMCs were isolated from healthy donors and stimulated with pooled plasma from patients with malaria. Plasma samples were collected either during acute infection (8 primary-infected and 8 previously exposed donors) or at 12 months after treatment (donor-matched; 8 primary-infected and 8 previously exposed donors), with equal volume contributions from each donor. The total IgG from the pooled plasma have been removed by Ab SpinTrap (Cytiva). Freshly isolated PBMCs were cultured at 37°C and 5% CO₂ for 20–24 hours in RPMI-1640 medium supplemented with either 20% IgG-depleted plasma (acute or convalescent, from either primary-infected or previously exposed donors) or 20% FBS. Plates were incubated at 37°C and 5% CO₂.

Flow cytometry of stimulated B cells and stimulated fresh PBMCs. Cells were washed twice with PBS and stained with a Live/Dead viability dye (Thermo Fisher Scientific) for 20 minutes at 4°C. After washing twice with PBS containing 2% FBS, cells were incubated with an antibody cocktail for surface marker staining for 20 minutes at 4°C. For intracellular staining, cells were washed once after surface staining and fixed/permeabilized using FoxP3 Fix/Perm buffer (eBioscience) for 30 minutes at RT. Cells were then washed with FoxP3 Wash/Perm buffer and incubated with intracellular antibodies for 30 minutes at RT. Following staining, cells were washed with Wash/Perm buffer and resuspended in PBS containing 2% FBS. Stimulated B cells were acquired on a 5-laser BD LSRFortessa or a 4-laser Cytek Aurora (Cytek Biosciences) while stimulated fresh PBMCs were analyzed on a 5-laser Cytek Aurora (Cytek Biosciences) with CountBright absolute counting beads (Thermo Fisher Scientific). Analysis of flow cytometry data was done with FlowJo (version 10.10.0, BD for stimulated B cells and version 10.10.1, BD for stimulated fresh PBMCs). See Supplemental Table 2 for antibody panels.

Multiplex bead-based assay to measure parasite-specific antibodies. A multiplex bead-based antibody assay was performed to measure total IgG, IgG1, IgG3, and IgM responses against recombinant merozoite surface proteins (AMA1, MSP1, MSP2, MSP3, GAMA derived from sequences of the P. falciparum 3D7 parasite strain). The plasma samples used for this study were collected from participants at the day of diagnosis when they visited the Karolinska University Hospital. Briefly, the plasma samples were diluted at 1:250 with dilution buffer (0.1% BSA, 0.05% Tween-20 in PBS, pH 7.4). Related merozoites antigens were coupled to magnetic COOH beads on different regions (B27, B35, B44, B52, B62; Bio-Rad). Prior to antigen coupling, buffer exchange was done to phosphate-buffered saline (PBS) using Zeba spin desalting columns (Thermo Fisher Scientific). Antigen and magnetic beads coupling were performed with the Bio-Plex Amine coupling kit (Bio-Rad), EDAC (Bio-Rad), and Sulfo-NHS (Thermo Fisher Scientific). The antigen-coupled beads were diluted with dilution buffer to 50 μL, corresponding to 1,250 beads per region and mixed with 50 μL 1:250 diluted plasma in 96-well microtiter plates (Bio-Rad), followed by 1-hour incubation on a shaker at RT avoiding light. After washing 4 times with PBS+0.05% Tween-20, 100 μL of PE-conjugated Goat Anti-Human IgG (Jackson ImmunoResearch Laboratories) diluted at 1:600 or Hinge PE mouse anti-human IgG1 (diluted at 1:300), IgG3 (diluted at 1:300), or IgM (diluted 1:300) (all from Southern Biotech) was added for a 30 minutes incubation on a shaker at RT, followed by washing another 4 times with PBS+0.05% Tween-20. The beads were then resuspended with 100 μL dilution buffer for sample acquisition. Median fluorescence intensities were measured using the Bio-Plex 200 system (Bio-Rad), with a reader setting to read a minimum of 100 beads per bead region.

Statistics. All data wrangling and analysis were performed in R (v.4.3.2) using packages from the tidyverse (v.2.0.0) (68). If not stated otherwise, data were visualized using R packages ggplot2 (v.3.4.2), ComplexHeatmap (v.2.16.0), or circlerize (v.0.4.15). The figures were assembled using patchwork (v.1.1.2) or Adobe Illustrator.

Statistical analyses were performed using GraphPad prism version 11 or R and are further described in the figure legends. Briefly, analyses comparing donor responses or characteristics over time or following multiple stimulations were done using repeated measures 2-tailed 1-way or 2-way ANOVA. If the analyses included missing data, a linear mixed effects model was used. For comparison between 2 variables, 2-tailed Student’s t tests or Wilcoxon rank tests were used. P < 0.05 were considered significant.

Study approval. The prospective study was conducted at the Karolinska University Hospital in Stockholm, Sweden, with ethical approval from the Ethical Review Board in Stockholm and the Swedish Ethical Review Authority (Dnr 2006/893-31/4, 2013/550-32, 2015/2200-32, 2016/1940/32, 2018/2354-32, 2019-03436, 2020-00859, 2020-04147, 2024-00374-02). All participants gave written informed consent.

Data availability. This study did not generate new unique reagents. Data that were repurposed from previously generated/published datasets included: single-cell multi-omics data (Rhapsody target Immune response panel and AbSeq) and plasma protein profiling data (Explore1536) (19), BAFF-R expression level on B cells from patients with malaria (23), and iRBC stimulation of PBMCs from patients with malaria (18).

All code used in the analyses is available via https://github.com/SundlingLab/MalariaMonocytes

European Data Regulations, as interpreted by Karolinska Institutet, precludes open deposition of sensitive personal data into public repositories. This includes biological data that can be traced back to an individual, so-called pseudoanonymized data. The transcriptomic and proteomic data are available pending relevant permits via https://doi.org/10.48723/cmgs-aj72 This includes a metadata record describing the existing datasets. Data points in graphs derived from pseudoanonymized donors are available from the corresponding author, pending relevant permits. Values for all data points in graphs coming from anonymized donors or aggregated data are reported in the Supporting Data Values file.