Anti-phosphatidylserine antibodies (PS-Abs) are elevated in Malaysian patients with vivax and falciparum malaria and are highest in vivax malaria. In vivax and falciparum malaria, PS-Abs correlate inversely with admission and nadir hemoglobin, suggesting that PS-Abs contribute to anemia from these species.
Keywords: phosphatidylserine antibodies, anemia, malaria, Plasmodium vivax, Plasmodium falciparum, Plasmodium knowlesi, Plasmodium malariae
Abstract
Background
Anemia is a major complication of vivax malaria. Antiphosphatidylserine (PS) antibodies generated during falciparum malaria mediate phagocytosis of uninfected red blood cells that expose PS and have been linked to late malarial anemia. However, their role in anemia from non-falciparum Plasmodium species is not known, nor their role in early anemia from falciparum malaria.
Methods
We measured PS immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies in Malaysian patients with vivax, falciparum, knowlesi, and malariae malaria, and in healthy controls, and correlated antibody titres with hemoglobin. PS antibodies were also measured in volunteers experimentally infected with Plasmodium vivax and Plasmodium falciparum.
Results
PS IgM and IgG antibodies were elevated in patients with vivax, falciparum, knowlesi, and malariae malaria (P < .0001 for all comparisons with controls) and were highest in vivax malaria. In vivax and falciparum malaria, PS IgM and IgG on admission correlated inversely with admission and nadir hemoglobin, controlling for parasitemia and fever duration. PS IgM and IgG were also increased in volunteers infected with blood-stage P. vivax and P. falciparum, and were higher in P. vivax infection.
Conclusions
PS antibodies are higher in vivax than falciparum malaria, correlate inversely with hemoglobin, and may contribute to the early loss of uninfected red blood cells found in malarial anemia from both species.
Anemia is a major complication of malaria from all Plasmodium species, causing significant morbidity and mortality not only in Plasmodium falciparum malaria, but also in Plasmodium vivax malaria [1–3]. Although occurring in part due to rupture of infected red blood cells (RBCs), as well as reduced erythropoiesis, the major contributor to malarial anemia is the loss of uninfected RBCs [4], particularly in vivax malaria [5]. Mechanisms leading to the loss of uninfected RBCs are incompletely understood but likely include oxidative damage, as well as complement-mediated lysis [6]. In addition, recent evidence has emerged regarding the role of autoantibodies directed against RBC components, including phosphatidylserine (PS) [7, 8].
Phosphatidylserine is a membrane phospholipid that is normally located on the internal leaflet of the RBC lipid bilayer but that may become exposed as a result of a number of stimuli such as oxidative stress [9]. In P. falciparum, PS becomes exposed on infected RBCs during parasite maturation [10–12], enhancing phagocytosis of these cells. In addition, in P. falciparum in vitro cultures [13], and in mice infected with Plasmodium yoelii [7, 14], PS is also externalized on uninfected RBCs. The mechanisms of PS exposure on uninfected RBCs remain unclear but may relate to oxidative damage from parasite degradation with release of hemozoin or reactive oxygen species, or release and circulation of toxic heme [15, 16].
With exposure of PS on uninfected as well as infected RBCs, it has been hypothesized that autoantibodies directed against PS may contribute to malarial anemia [7]. In keeping with this, in a recent study involving the P. yoelli mouse model of malaria, infection was shown to lead to the generation of anti-PS immunoglobulin M (IgM) and immunoglobulin G (IgG) antibodies, with subsequent enhanced phagocytosis by macrophages of PS-expressing uninfected RBCs [7]. Furthermore, blocking of PS antibodies in infected mice led to faster recovery from anemia. PS IgM antibodies have been shown to be induced by P. falciparum in vitro [8], and in nonimmune humans with primary falciparum malaria total PS antibodies correlated with late postmalarial anemia [7]. However, the role of PS IgG and IgM antibodies in the anemia from non-falciparum Plasmodium species, particularly P. vivax, has not been defined, nor their role in early anemia in falciparum malaria.
In this study, we evaluated PS IgM and IgG in Malaysian patients with vivax, falciparum, knowlesi, and malariae malaria, and in healthy controls, and evaluated associations with hemoglobin, intravascular hemolysis (as a cause of oxidative stress in severe malaria), and RBC deformability. To define antibody kinetics and magnitude in primary infection with P. vivax and P. falciparum, we also measured PS antibodies longitudinally in malaria-naive human volunteers experimentally infected with these species.
METHODS
Ethics Statement
The studies conducted in Malaysia were approved by the ethics committees of the Malaysian Ministry of Health and the Menzies School of Health Research, Darwin, Australia. The malaria volunteer infection studies were approved by the QIMR Berghofer ethics committee. Informed written consent was provided by all participating adults and by the parent or guardian of any participant aged <18 years.
Malaysian Study Sites, Patients, and Study Procedures
Patients hospitalized with malaria were enrolled as part of concurrent observational studies [17, 18] and/or clinical trials [19–21] conducted at 4 study sites in Sabah, Malaysia, including a tertiary referral hospital (Queen Elizabeth Hospital [QEH]) and 3 district hospitals (Kudat, Kota Marudu, and Pitas District Hospitals) during 2010–2016. At the tertiary referral hospital site, patients were enrolled if they were within 18 hours of commencing antimalarial treatment, aged >12 years, not pregnant, and had no major comorbidities or concurrent illnesses. Inclusion criteria were the same at the district hospitals, except that all patients were enrolled prior to commencing antimalarial treatment, and all ages were included. For the current study, patients with polymerase chain reaction (PCR)–confirmed P. vivax, P. falciparum, and Plasmodium knowlesi were included from the tertiary-hospital cohort, whereas patients with PCR-confirmed P. vivax, P. falciparum, and Plasmodium malariae were included from the district hospital cohort. Patients enrolled in the observational studies [17, 18] received treatment according to hospital guidelines at the time of the studies, including artemisinin-combination therapy (ACT) for uncomplicated falciparum, knowlesi, or malariae malaria, and ACT or chloroquine plus primaquine for uncomplicated vivax malaria. Patients with knowlesi, vivax, or malariae malaria and enrolled in clinical trials received either ACT or chloroquine, according to the study protocols [19–21]. Patients with severe malaria received intravenous artesunate.
Healthy controls were visitors or relatives of malaria patients admitted to QEH, with no history of fever in the past 48 hours and with a blood film negative for malaria parasites.
Venous blood was collected in lithium heparin and centrifuged within 30 minutes, with plasma stored at –70°C. Whole blood was also collected in ethylenediaminetetraacetic acid for measurement of RBC deformability (below). At QEH, blood was collected for PS antibodies on enrollment and at days 14 and/or 28 in a subset of patients able to return for follow-up. At district hospitals, blood was collected for PS antibodies on enrollment, and on days 7 and 28 in patients with P. vivax and P. malariae. Follow-up PS antibody measurements were excluded from analysis if a patient had received a blood transfusion during admission. At all sites, hemoglobin was measured daily during admission, and at follow-up visits. In a subset of patients, plasma cell-free hemoglobin was measured by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s instructions (Bethyl Laboratories). Anti-PS IgG and IgM antibodies were detected by ELISA (Orgentec).
RBC deformability was measured on enrollment in a subset of patients by laser-assisted optical rotational cell analyzer (LORCA Mechatronics), as previously described [22]. RBC deformability was assessed at shear stresses of 1.7 Pa and 30 Pa. Shear stresses of 1.7 Pa are encountered in the capillaries [23]. Shear stresses of 30 Pa provide information on cell geometry, in particular surface area to volume ratios [24], and approximate values encountered by RBCs passing through intercellular gaps in the splenic sinusoids [25].
Malaria Volunteer Infection Studies
Malaria volunteer infection studies were conducted at QIMR Berghofer, Australia, as previously described [26, 27]. In brief, healthy malaria-naive volunteers were inoculated with approximately 2800 viable P. falciparum 3D7-infected RBCs, or approximately 1680 P. vivax–infected RBCs. Peripheral blood parasitemia was measured at least daily by quantitative PCR, and participants were treated with antimalarial drugs on day 8 (P. falciparum) or day 10 (P. vivax), when parasitemia had exceeded 5000 parasites/mL. Participants with P. falciparum and with samples available were enrolled in studies NCT02661373 (n = 13), NCT02783833 (n = 4), and NCT02573857 (n = 6). Participants with P. vivax were enrolled in study NCT02573857 (n = 8). PS IgG and IgM antibodies were measured from blood samples collected prior to infection, prior to antimalarial treatment, and at day 18 (P. vivax), or days 14–15 and day 20 (P. falciparum). Blood was collected in lithium heparin, with plasma frozen at –70°C.
Statistical Analysis
Statistical analysis was performed with Stata software (version 14). For continuous variables, intergroup differences were compared using analysis of variance or Kruskal–Wallis tests depending on distribution. Student t test or Mann–Whitney tests were used for 2-group comparisons. Categorical variables were compared using χ 2 test. Associations between continuous variables were assessed using Spearman or Pearson correlation coefficients, depending on distribution. Partial correlation was used to evaluate associations between variables after adjusting for parasitemia, fever duration, and age, with nonnormally distributed variables log-transformed to normality. Wilcoxon signed-rank test was used to compare baseline and follow-up measurements.
RESULTS
Malaysian Malaria Patients
A total of 508 malaria patients and 50 controls were included. Malaria patients included 269 patients with P. falciparum, 176 with P. vivax, 42 with P. knowlesi, and 21 with P. malariae. Baseline demographic and clinical characteristics are shown in Table 1. Overall, 375 (74%) patients were male, and median age was 26 years (interquartile range, 17–40 years). A total of 102 (20%) patients reported a previous episode of malaria. Anemia by World Heath Organization criteria [28] was common, occurring on admission in 221 (44%) patients overall, and in >90% of patients with malaria from all species during follow-up (Table 1). Moderate anemia (hemoglobin <10 g/dL) occurred in 65 (13%) patients overall on admission, and in 139 (27%) patients during follow-up (Table 1).
Table 1.
Baseline Characteristics and Phosphatidylserine Immunoglobulin M and Immunoglobulin G Antibodies in Controls and Patients With Malaria
| Characteristic | Controls (n = 50) | Plasmodium falciparum (n = 269) | Plasmodium vivax (n = 176) | Plasmodium knowlesi (n = 42) | Plasmodium malariae (n = 21) |
|---|---|---|---|---|---|
| Age, y, mean (range) | 35 (14–69) | 28 (7–78) | 24 (2–79) | 43 (17–75) | 17 (5–54) |
| Male sex, No. (%) | 34 (68) | 202 (75) | 127 (72) | 30 (77) | 16 (76) |
| Previous malaria | NA | 34 (13) | 51 (29) | 14 (34) | 51 (29) |
| Fever duration, d | … | 5 (3–7) | 5 (3–7) | 5 (4–7) | 6 (3–7) |
| Parasite count, parasites/μL | … | 10 011 (2902–32 472) | 4000 (1775–9548) | 4215 (2372–29 922) | 1313 (162–2778) |
| Severe malaria, No. (%) | … | 29 (11) | 12 (7) | 10 (26) | 0 (0) |
| Hb on enrollment, g/dL, mean (SD) | … | 12.9 (2.0) | 12.2 (2.1) | 12.4 (2.2) | 11.6 (2.2) |
| Hb nadir, g/dL, mean (SD) | … | 11.3 (1.8) | 10.8 (1.9) | 11.2 (2.3) | 10.5 (1.9) |
| Hb decline, g/dL | … | 1.4 (0.8–2.2) | 1.3 (0.7–2.1) | 1.2 (0.2–2.0) | 1.2 (0.9) |
| Anemiaa on admission, No. (%) | … | 97 (36) | 89 (51) | 21 (50) | 14 (67) |
| Anemiaa during follow-up, No. (%) | … | 251 (93) | 167 (95) | 39 (93) | 20 (95) |
| Admission Hb <10 g/dL, No. (%) | … | 23 (9) | 30 (17) | 6 (14) | 6 (29) |
| Nadir Hb <10 g/dL, No. (%) | … | 56 (21) | 61 (35) | 13 (31) | 9 (43) |
| CFHb, ng/mL | 15 146 (9641–25 256) | 34 309 (15 330–527 779) (n = 171) | 32 498 (16 813–44 489) (n = 62) | 20 042 (15 072–44 242) (n = 15) | … |
| RBC-D (at 1.7 Pa), EI | 0.197 (0.178–0.227) (n = 7) | 0.182 (0.163–0.198) (n = 90) | 0.196 (0.160–0.214) (n = 25) | 0.168 (0.147–0.190) (n = 11) | … |
| RBC-D (at 30 Pa), EI | 0.587 (0.520–0.590) (n = 7) | 0.518 (0.480–0.557) (n = 90) | 0.543 (0.518–0.572) (n = 25) | 0.492 (0.448–0.550) (n = 11) | … |
| PS IgM day 0, U/mL | 27 (19–41) | 85 (55–134) | 93 (62–181) | 60 (41–108) | 91 (58–146) |
| PS IgM, day 7, U/mL | … | … | … | … | 103 (42–142) (n = 15) |
| PS IgM, day 14, U/mL | … | 130 (80–183) (n = 32) | 143 (49–2686) (n = 15) | 81 (50–116) (n = 33) | … |
| PS IgM, day 28, U/mL | … | 83 (66–131) (n = 69) | 84 (51–124) (n = 22) | 62 (36–92) (n = 22) | 55 (39–63) (n = 14) |
| PS IgG, day 0, U/mL | 17 (14–21) | 49 (35–72) | 59 (39–92) | 44 (21–85) | 63 (41–77) |
| PS IgG, day 7, U/mL | … | … | … | … | 58 (20–85) (n = 15) |
| PS IgG, day 14, U/mL | … | 94 (64–112) (n = 32) | 62 (37–251) (n = 15) | 68 (30–93) (n = 33) | … |
| PS IgG, day 28, U/mL | … | 43 (19–63) (n = 69) | 49 (22–120) (n = 22) | 46 (26–70) (n = 22) | 38 (20–47) (n = 14) |
Data are presented as median (interquartile range) unless otherwise indicated. For PS IgG and IgM antibodies at baseline, P < .0001 for healthy controls compared to all Plasmodium species.
Abbreviations: CFHb, cell-free hemoglobin; EI, elongation index; Hb, hemoglobin; IgG, immunoglobulin G; IgM, immunoglobulin M; NA, not available; RBC-D, red blood cell deformability; PS, phosphatidylserine; SD, standard deviation.
aSelf-reported. Anemia based on World Health Organization 2011 hemoglobin measurement criteria [26]: age 6–59 months, ≤10.0 g/dL; age 5–11 years, <11.5 g/dL; age 12–14 years, <12.0 g/dL; nonpregnant women age ≥15 years, <12.0 g/dL; pregnant women, <11.0 g/dL; men age ≥15 years, <13.0 g/dL.
Phosphatidylserine IgM and IgG antibodies in Malaysian Malaria Patients
Phosphatidylserine IgM antibodies were increased on enrollment in patients with vivax, falciparum, knowlesi, and malariae malaria (P < .0001 for all comparisons with controls; Table 1, Figure 1). PS IgM antibodies were highest in patients with vivax malaria (P < .0001 for P. vivax vs P. falciparum, and for P. vivax vs P. knowlesi), and higher in falciparum compared to knowlesi malaria (P = .014). PS IgG antibodies were also higher on enrollment in vivax, falciparum, knowlesi, and malariae malaria (P < .0001 for all comparisons with controls; Figure 1), and were higher in vivax compared to falciparum (P < .0001) and knowlesi (P = .0007) malaria. For both IgM and IgG, the difference between P. vivax and P. falciparum remained significant after controlling for parasitemia, fever duration, and age (P < .0001 for both comparisons). No significant differences were seen overall in PS IgM or IgG antibodies between patients who did, or did not, report having had a previous episode of malaria.
Figure 1.
Phosphatidylserine (PS) immunoglobulin M (IgM; A) and immunoglobulin G (IgG; B) antibodies in healthy controls and in patients hospitalized with falciparum (Pf), vivax (Pv), knowlesi (Pk), and malariae (Pm) malaria. PS IgM and IgG antibodies were lower in controls than in patients with malaria from any Plasmodium species (P < .0001 for all comparisons). PS IgM and IgG antibodies were higher in P. vivax compared to both P. falciparum and P. knowlesi. PS IgM antibodies were higher in P. falciparum compared to P. knowlesi (P = .014).
Clinical Correlates of Phosphatidylserine IgG and IgM Antibodies in Malaysian Malaria Patients
In both vivax and falciparum malaria patients, PS IgM and IgG each correlated inversely with hemoglobin on enrollment, and with hemoglobin nadir (P = .001 for correlation between PS-IgM and hemoglobin on enrollment in vivax malaria; P < .0001 for all other correlations; Table 2, Supplementary Figure 1). In vivax and falciparum malaria, there was a correlation between PS IgM and fever duration (r = 0.22, P = .004 and r = 0.30, P < .0001, respectively). In patients with vivax malaria, there was also an inverse association with PS IgM and IgG antibodies and age (r = –0.21, P = .005 and r = –0.23, P = .002, respectively). No correlation was observed with parasitemia in either species. In both falciparum and vivax malaria, the correlations between PS IgM and IgG antibodies and enrollment and nadir hemoglobin remained significant after controlling for fever duration, age, and parasitemia (Supplementary Table 1).
Table 2.
Clinical Correlates of Phosphatidylserine Immunoglobulin M and Immunoglobulin G Antibodies in Patients With Malaria
| Plasmodium falciparum (n = 269) | Plasmodium vivax (n = 176) | |||||||
|---|---|---|---|---|---|---|---|---|
| PS IgM | PS IgG | PS IgM | PS IgG | |||||
| Clinical Correlate | Correlation | P Value | Correlation | P Value | Correlation | P Value | Correlation | P Value |
| Hb on enrollment | –0.27 | < .0001a | –0.26 | < .0001a | –0.30 | .001 | –0.31 | < .0001a |
| Hb nadir | –0.28 | < .0001a | –0.29 | < .0001a | –0.34 | < .0001 | –0.35 | < .0001a |
| Parasite count | –0.00 | .943 | 0.01 | .824 | 0.10 | .170 | 0.10 | .192 |
| Fever duration | 0.30 | < .0001 | 0.11 | .078 | 0.22 | .004 | 0.15 | .051 |
| Age | 0.11 | .067 | –0.06 | .358 | –0.21 | .005 | –0.23 | .002 |
| CFHbb | 0.19 | .014 | 0.27 | .027 | 0.04 | .779 | –0.07 | .601 |
| RBC-Dc (at 1.7 Pa) | –0.28 | .008 | –0.13 | .227 | 0.22 | .282 | –0.04 | .859 |
| RBC-Dc (at 30 Pa) | –0.21 | .050 | –0.18 | .085 | –0.05 | .828 | –0.28 | .176 |
Abbreviations: CFHb, cell-free hemoglobin; Hb, hemoglobin; IgG, immunoglobulin G; IgM, immunoglobulin M; PS, phosphatidylserine; RBC-D, red blood cell deformability.
aRemained significant after controlling for parasitemia, fever duration, and age (Supplementary Table 1).
bCFHb was measured in 133 patients with falciparum malaria and 57 with vivax malaria.
cRBC-D was measured in 90 patients with falciparum malaria, and 25 patients with vivax malaria.
In falciparum malaria, there was a correlation between PS IgM and PS IgG and intravascular hemolysis, as measured by cell-free hemoglobin (PS IgM: r = 0.19, P = .014; PS IgG: r = 0.27, P = .027), and an inverse correlation between PS IgM and RBC deformability (r = –0.28, P = .0008, RBC elongation index at 1.7 Pa; Table 2). Both associations with PS IgM remained significant after controlling for parasitemia.
In P. knowlesi and P. malariae malaria, there was no correlation between either PS IgM or IgG and admission or nadir hemoglobin. There was also no correlation between PS IgM or IgG and parasitemia in either species.
Longitudinal PS IgM and IgG Antibody Titers in Malaysian Malaria Patients
At the district hospital sites, in patients with vivax malaria PS IgM and IgG levels were higher at day 7 compared to enrollment (Table 1, Supplementary Figure 2), and for PS IgM this increase correlated significantly with the fractional fall in hemoglobin between day 0 and day 7 (r = 0.31, P = .006 for PS IgM, and r = 0.22, P = .059 for IgG). In patients with P. malariae, there was no significant increase in PS IgM or IgG at day 7. In district malaria patients with P. vivax and P. malariae, PS IgM and IgG had fallen by day 28, but levels remained above those of healthy controls (Table 1, Supplementary Figure 2).
For the subset of tertiary-referral patients who had PS antibodies measured at day 14 following enrollment (P. knowlesi = 33, P. falciparum = 32, P. vivax = 15), PS IgM and IgG were higher at day 14 compared to day 0, although this was only statistically significant for patients with P. knowlesi (Table 1, Supplementary Figure 3). No correlation was observed between day 14 PS IgM or PS IgG and day 14 hemoglobin. As with the district hospital patients, by day 28 PS IgG and IgM had returned to day 0 levels, but remained above the levels found in community controls (P < .0001 for all comparisons).
Phosphatidylserine IgG and IgM Antibodies in Volunteers Infected With P. falciparum and P. vivax
In participants infected with P. falciparum (n = 23), there was a significant increase in PS IgM and IgG antibodies by day 20 (Table 3, Figure 2). For both PS IgM and IgG, the magnitude of this increase correlated with peak parasitemia (r = 0.80, P < .0001 for PS IgM, and r = 0.59, P = .003 for IgG). There was an inverse correlation between PS IgM at day 20 and hemoglobin measured at day 20 (or first available day 20–28; r = –0.41, P = .054); this relationship was significant after controlling for peak parasitemia (r = –0.43, P = .045). There was also a correlation between the increase in both PS IgM and IgG between day 0 and day 20 and the fractional fall in hemoglobin (r = 0.49, P = .018 for PS IgM, and r = 0.43, P = .042 for PS IgG); however, this was not significant after controlling for peak parasitemia.
Table 3.
Antiphosphatidylserine Immunoglobulin M and Immunoglobulin G Antibodies in Volunteers With Experimental Malaria Infection
| Parameters | Plasmodium falciparum (n = 23) | Plasmodium vivax (n = 8) | P Value |
|---|---|---|---|
| Peak parasitemia, parasites/mL | 36 074 (8351–142 519) | 219 136 (112 091–308 113) | .008 |
| Baseline Hb, g/dL, mean (SD) | 14.8 (9.2) | 14.5 (7.3) | .456 |
| Hb day 18 (Pv) or day 20–28a (Pf), g/dL, mean (SD) | 13.5 (9.2) | 13.7 (7.9) | .786 |
| Hb decline, g/dL | 1.0 (0.6–1.5) | 1.0 (0.7–1.2) | .651 |
| PS IgM, day 0. U/mL | 21 (15–36) | 35 (20–70) | .124 |
| PS IgG, day 0, U/mL | 15 (6–24) | 20 (10–25) | .329 |
| PS IgM, day 18 (Pv) or day 20 (Pf), U/mL | 33 (27–52) | 71 (56–121) | .012 |
| PS IgG, day 18 (Pv) or day 20 (Pf), U/mL | 17 (3–24) | 24 (20–42) | .026 |
Data are presented as median (interquartile range) unless otherwise indicated.
Abbreviations: Hb, hemoglobin; IgG, immunoglobulin G; IgM, immunoglobulin M; Pf, Plasmodium falciparum; Pv, Plasmodium vivax; PS, phosphatidylserine; SD, standard deviation.
aFor volunteers with falciparum malaria, “Hb day 20–28” refers to hemoglobin measured on day 20, or the first available up to day 28.
Figure 2.
Phosphatidylserine (PS) immunoglobulin M (IgM) and immunoglobulin G (IgG) antibodies in participants experimentally infected with Plasmodium falciparum (A and B) and Plasmodium vivax (C and D). A total of 23 participants in 3 study cohorts were infected with P. falciparum and included in the analysis; PS antibodies were measured at baseline, days 7–8 (prior to treatment), days 14–15, and day 20. Eight participants in one study cohort were infected with P. vivax, with PS antibodies measured at baseline and at days 8, 7, 10, and 18. P values represent difference between baseline and day 20 (P. falciparum) or day 18 (P. vivax), by Wilcoxon signed-rank test. Data are presented as median and interquartile range.
In participants infected with P. vivax (n = 8), PS IgM and IgG antibodies increased by day 18. Titers of both PS IgM and IgG were higher in participants infected with P. vivax than in participants with P. falciparum (Table 3, Figure 3). In participants with P. vivax, there was no correlation between PS IgM or IgG and hemoglobin parameters, or parasitemia, although numbers were small.
Figure 3.
Phosphatidylserine (PS) immunoglobulin M (IgM; A) and immunoglobulin G (IgG; B) antibodies at day 20 in participants experimentally infected with Plasmodium falciparum and at day 18 in participants experimentally infected with Plasmodium vivax. Data are presented as median and interquartile range.
DISCUSSION
Anti-PS IgM and IgG antibodies were increased on presentation in Malaysian patients with each of the major Plasmodium species infecting humans, with both being higher in vivax malaria than in falciparum malaria. Furthermore, in both vivax and falciparum malaria, IgM and IgG PS antibody titers were inversely correlated with admission and nadir hemoglobin, with these correlations being independent of parasitemia, fever duration, and age. Findings suggest that anti-PS antibodies contribute to malarial anemia from P. vivax as well as P. falciparum, even relatively early in the disease process. These findings are supported by data from the malaria volunteer infection studies, where antibody titers increased to a greater extent following inoculation with P. vivax than P. falciparum, and in the larger cohort with falciparum malaria, correlated inversely with hemoglobin independent of parasitemia.
Our findings are consistent with an early study demonstrating increased antiphospholipid antibodies in patients with falciparum and vivax malaria [29], and another demonstrating a correlation between total PS antibodies and late anemia in nonimmune patients with primary P. falciparum malaria [7]. PS antibodies are thought to contribute to malarial anemia by binding to infected and uninfected RBCs exposing PS, and enhancing the phagocytosis of uninfected RBCs [7]. In mice with P. yoelli, PS antibodies have been shown to be produced by atypical CD11c+ T-bet+ B cells, with expansion of these cells correlating directly with parasitemia and inversely with RBC density, suggesting a role in anemia [8]. The expansion of T-bet+ B cells was shown to occur through Toll-like receptor 9 (TLR9) and interferon-γ signaling, with PS antibodies increasing shortly after the expansion of these cells. Similarly, exposure of peripheral blood mononuclear cells from healthy naive donors to P. falciparum in vitro induced expansion of T-bet+ B cells and production of PS antibodies [8]. Taken together, these data suggest that production of PS antibodies by T-bet+ B cells is stimulated directly by parasite DNA, rather than as a result of PS exposure. This may explain the finding in our study that PS IgG and IgM were already elevated on presentation in Malaysian patients with malaria, and correlated with early anemia. This early increase in antibody titers was also seen in the malaria volunteer infection studies, where PS antibodies increased by day 18 (P. vivax) or day 20 (P. falciparum), despite the low number of parasites inoculated (well below the number of merozoites released during schizont rupture following mosquito infection), and with peak parasitemias well below those seen in the clinical studies.
In this study we found that PS IgM and IgG were significantly higher in vivax compared to falciparum malaria. Plasmodium vivax has a lower pyrogenic threshold than P. falciparum and is associated with a greater inflammatory response than that seen in P. falciparum infections with a similar or greater peripheral parasitemia [30–32]. It has been postulated that this may relate to the greater GC content of the P. vivax genome, with greater stimulation of TLR9 by CpG motifs within P. vivax hemozoin [33]. As TLR9 has also been shown to mediate the expansion of T-bet+ B cells that produce PS antibodies [8], this greater GC content of P. vivax may also account for the higher titers of PS antibodies observed in this study. As PS antibodies are thought to mediate phagocytosis of uninfected RBCs, the higher antibody titers in vivax compared to falciparum malaria may explain, in part, the finding that anemia is a common complication of vivax malaria despite relatively low parasitemias, with the relative loss of uninfected RBCs to circulating infected RBCs greater in vivax compared to falciparum malaria [4, 34].
In this study we did not find a correlation between PS IgM and IgG antibody titers and hemoglobin in patients with P. malariae, despite that fact that PS antibody titers in P. malariae were comparable to those of P. vivax, and that anemia was at least as prevalent in patients with P. malariae as with the other species. Although this may have been because of the small number of patients with P. malariae, it may also be that other factors may play a relatively greater role in anemia from P. malariae. Given the morbidity associated with anemia from P. malariae [35], further studies are required to investigate mechanisms of anemia from this species.
A notable finding of our study was the higher titers of PS IgM compared to IgG in both clinical studies and volunteer infection studies. IgM antibodies have recently been shown to be rapidly induced in falciparum malaria, in both volunteer infection studies and in children and adults with clinical malaria in endemic areas [36]. In volunteer infection studies, antigen-specific IgM responses were shown to be more prevalent than IgG responses. Furthermore, IgM blocked merozoite invasion of RBCs, and was associated with a significantly reduced risk of clinical malaria in a longitudinal cohort of Papuan children [36]. These results suggest that parasite-specific IgM is an important functional antibody response targeting blood-stage malaria parasites. However, the results of the current study suggest that in addition to this contribution to malaria immunity, IgM antibodies may also be associated with pathogenic mechanisms of malarial anemia.
Our study found a correlation between PS antibodies and cell-free hemoglobin (CFHb) in patients with P. falciparum malaria. CFHb is released during intravascular hemolysis and is readily oxidized to heme, which induces lipid peroxidation in RBCs [37, 38]. Oxidative stress is a major stimulant of externalization of PS in RBCs [39], and the correlation in our study between CFHb and PS antibodies suggests that in addition to direct stimulation of T-bet+ B cells by Plasmodium DNA, hemolysis-induced exposure of PS on uninfected RBCs may also be a driver of PS antibody production.
We also demonstrated an inverse correlation between PS IgM and RBC deformability in falciparum malaria. This is consistent with a previous study demonstrating that coating of RBCs with purified anti-RBC IgG antibodies from patients with P. vivax–associated anemia increased the rigidity of RBC membranes [40]. Reduced RBC deformability has been shown to be associated with anemia in both falciparum [41] and knowlesi [42] malaria, with enhanced phagocytosis and increased splenic clearance of the more rigid RBCs both potentially contributing. It is thus possible that PS antibodies mediate anemia not only directly through phagocytosis of PS-exposing cells [7], but also through increasing the rigidity of infected and uninfected RBCs.
Our study has several limitations. It is possible that the early and marked increase in PS antibodies observed in Malaysian patients represents an epiphenomenon. However, the inverse relationships between PS antibodies and hemoglobin were independent of possible confounders such as parasitemia and duration of illness. Furthermore, the relationships between PS antibodies and hemoglobin were strongest for the 2 species (P. vivax and P. falciparum) with the highest antibody titers, and absent in P. knowlesi, where antibody responses were significantly lower. Moreover, PS IgM was also inversely associated with hemoglobin, independent of parasitemia, in the volunteer infection studies.
In conclusion, PS IgM and IgG antibodies are increased in falciparum, vivax, knowlesi, and malariae malaria. PS antibody responses were higher in P. vivax than P. falciparum infection, in both clinical disease and experimental human challenge. Both PS IgM and IgG correlated with early anemia in malaria from both species, suggesting that PS antibodies may contribute to the early loss of uninfected RBCs found in malarial anemia in both vivax and falciparum malaria.
Supplementary data
Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Acknowledgments. We thank all of the study participants in Sabah, Malaysia; the Sabah malaria research team nursing and laboratory staff; and the Malaysian Ministry of Health hospital directors and clinical staff at Queen Elizabeth Hospital and at Kota Marudu, Kudat, and Pitas district hospitals. We also recognize the support of Dr Goh Pik Pin and the Clinical Research Centre, Sabah, for logistical support, and the Director General of Health, Malaysia, for permission to publish this study. We thank the participants involved in the malaria volunteer infection studies, Q-Pharm staff, and the Medicine for Malaria Venture for funding these studies.
Financial support. This work was supported by the National Health and Medical Research Council of Australia (program grant numbers 496600 and 1037304; project grant 1045156; and fellowships to B. E. B., M. J. G., M. J. B., J. S. M., and N. M. A.). A. M. D. was supported by the Wellcome Trust of Great Britain.
Potential conflicts of interest. All authors: No reported conflicts of interest.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Presented in part: 66th Annual Meeting of the American Society of Tropical Medicine and Hygiene, Baltimore, Maryland, November 2017.
References
- 1.Douglas NM, Lampah DA, Kenangalem E, et al. Major burden of severe anemia from non-falciparum malaria species in southern Papua: a hospital-based surveillance study. PLoS Med 2013; 10:e1001575; discussion e1001575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Douglas NM, Pontororing GJ, Lampah DA, et al. Mortality attributable to Plasmodium vivax malaria: a clinical audit from Papua, Indonesia. BMC Med 2014; 12:217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Grigg MJ, William T, Barber BE, et al. Age-related clinical spectrum of Plasmodium knowlesi malaria and predictors of severity. Clin Infect Dis 2018; 67:350–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jakeman GN, Saul A, Hogarth WL, Collins WE. Anaemia of acute malaria infections in non-immune patients primarily results from destruction of uninfected erythrocytes. Parasitology 1999; 119 (Pt 2):127–33. [DOI] [PubMed] [Google Scholar]
- 5.Douglas NM, Anstey NM, Buffet PA, et al. The anaemia of Plasmodium vivax malaria. Malar J 2012; 11:135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Oyong DA, Kenangalem E, Poespoprodjo JR, et al. Loss of complement regulatory proteins on uninfected erythrocytes in vivax and falciparum malaria anemia. J Clin Invest Insight 2018; 3:124854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fernandez-Arias C, Rivera-Correa J, Gallego-Delgado J, et al. Anti-self phosphatidylserine antibodies recognize uninfected erythrocytes promoting malarial anemia. Cell Host Microbe 2016; 19:194–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rivera-Correa J, Guthmiller JJ, Vijay R, et al. Plasmodium DNA-mediated TLR9 activation of T-bet+ B cells contributes to autoimmune anaemia during malaria. Nat Commun 2017; 8:1282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lang E, Qadri SM, Lang F. Killing me softly—suicidal erythrocyte death. Int J Biochem Cell Biol 2012; 44:1236–43. [DOI] [PubMed] [Google Scholar]
- 10.Maguire PA, Prudhomme J, Sherman IW. Alterations in erythrocyte membrane phospholipid organization due to the intracellular growth of the human malaria parasite, Plasmodium falciparum. Parasitology 1991; 102:179–86. [DOI] [PubMed] [Google Scholar]
- 11.Sherman IW, Prudhomme J, Tait JF. Altered membrane phospholipid asymmetry in Plasmodium falciparum-infected erythrocytes. Parasitol Today 1997; 13:242–3. [DOI] [PubMed] [Google Scholar]
- 12.Pattanapanyasat K, Sratongno P, Chimma P, Chitjamnongchai S, Polsrila K, Chotivanich K. Febrile temperature but not proinflammatory cytokines promotes phosphatidylserine expression on Plasmodium falciparum malaria-infected red blood cells during parasite maturation. Cytometry A 2010; 77:515–23. [DOI] [PubMed] [Google Scholar]
- 13.Engelbrecht D, Coetzer TL. Plasmodium falciparum exhibits markers of regulated cell death at high population density in vitro. Parasitol Int 2016; 65:715–27. [DOI] [PubMed] [Google Scholar]
- 14.Totino PR, Magalhães AD, Silva LA, Banic DM, Daniel-Ribeiro CT, Ferreira-da-Cruz Mde F. Apoptosis of non-parasitized red blood cells in malaria: a putative mechanism involved in the pathogenesis of anaemia. Malar J 2010; 9:350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Föller M, Huber SM, Lang F. Erythrocyte programmed cell death. IUBMB Life 2008; 60:661–8. [DOI] [PubMed] [Google Scholar]
- 16.Matthews K, Duffy SP, Myrand-Lapierre ME, et al. Microfluidic analysis of red blood cell deformability as a means to assess hemin-induced oxidative stress resulting from Plasmodium falciparum intraerythrocytic parasitism. Integr Biol 2017; 9:519–28. [DOI] [PubMed] [Google Scholar]
- 17.Barber BE, Grigg MJ, William T, et al. Effects of aging on parasite biomass, inflammation, endothelial activation, microvascular dysfunction and disease severity in Plasmodium knowlesi and Plasmodium falciparum malaria. J Infect Dis 2017; 215:1908–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Barber BE, William T, Grigg MJ, et al. A prospective comparative study of knowlesi, falciparum, and vivax malaria in Sabah, Malaysia: high proportion with severe disease from Plasmodium knowlesi and Plasmodium vivax but no mortality with early referral and artesunate therapy. Clin Infect Dis 2013; 56:383–97. [DOI] [PubMed] [Google Scholar]
- 19.Grigg MJ, William T, Menon J, et al. Efficacy of artesunate-mefloquine for chloroquine-resistant Plasmodium vivax malaria in Malaysia: an open-label, randomized, controlled trial. Clin Infect Dis 2016; 62:1403–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Grigg MJ, William T, Menon J, et al. Artesunate-mefloquine versus chloroquine for treatment of uncomplicated Plasmodium knowlesi malaria in Malaysia (ACT KNOW): an open-label, randomised controlled trial. Lancet Infect Dis 2016; 16:180–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Grigg MJ, T W, Barber BE, et al. Artemether-lumefantrine versus chloroquine for the treatment of uncomplicated Plasmodium knowlesi malaria (CAN KNOW): an open-label randomized controlled trial. Clin Infect Dis 2018; 66:229–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hardeman M, Goedhart P, Dobbe J, Lettinga K. Laser-assisted optical rotational cell analyser (LORCA); I. A new instrument for measurement of various structural hemorheological parameters. Clin Hemorheol Microcirc 1994; 14:605–18. [Google Scholar]
- 23.Chien S. Physiological and pathophysiological significance of hemorheology. In: Chien S, Dormandy J, Ernst E, Matrai A, eds. Clinical hemorheology. Dordrecht, Netherlands: Springer, 1986:125–64. [Google Scholar]
- 24.Bessis M, Mohandas N, Feo C. Automated ektacytometry: a new method of measuring red cell deformability and red cell indices. Blood Cells 1980; 6:315–27. [PubMed] [Google Scholar]
- 25.Chen LT, Weiss L. The role of the sinus wall in the passage of erythrocytes through the spleen. Blood 1973; 41:529–37. [PubMed] [Google Scholar]
- 26.McCarthy JS, Griffin PM, Sekuloski S, et al. Experimentally induced blood-stage Plasmodium vivax infection in healthy volunteers. J Infect Dis 2013; 208:1688–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.McCarthy JS, Sekuloski S, Griffin PM, et al. A pilot randomised trial of induced blood-stage Plasmodium falciparum infections in healthy volunteers for testing efficacy of new antimalarial drugs. PLoS One 2011; 6:e21914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.World Health Organization. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity 2011. http://www.who.int/vmnis/indicators/haemoglobin.pdf. Accessed 15 January 2019.
- 29.Facer CA, Agiostratidou G. High levels of anti-phospholipid antibodies in uncomplicated and severe Plasmodium falciparum and in P. vivax malaria. Clin Exp Immunol 1994; 95:304–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yeo TW, Lampah DA, Tjitra E, et al. Greater endothelial activation, Weibel-Palade body release and host inflammatory response to Plasmodium vivax, compared with Plasmodium falciparum: a prospective study in Papua, Indonesia. J Infect Dis 2010; 202:109–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hemmer CJ, Holst FG, Kern P, Chiwakata CB, Dietrich M, Reisinger EC. Stronger host response per parasitized erythrocyte in Plasmodium vivax or ovale than in Plasmodium falciparum malaria. Trop Med Int Health 2006; 11:817–23. [DOI] [PubMed] [Google Scholar]
- 32.Barber BE, William T, Grigg MJ, et al. Parasite biomass-related inflammation, endothelial activation, microvascular dysfunction and disease severity in vivax malaria. PLoS Pathog 2015; 11:e1004558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Anstey NM, Russell B, Yeo TW, Price RN. The pathophysiology of vivax malaria. Trends Parasitol 2009; 25:220–7. [DOI] [PubMed] [Google Scholar]
- 34.Collins WE, Jeffery GM, Roberts JM. A retrospective examination of anemia during infection of humans with Plasmodium vivax. Am J Trop Med Hyg 2003; 68:410–2. [PubMed] [Google Scholar]
- 35.Langford S, Douglas NM, Lampah DA, et al. Plasmodium malariae infection associated with a high burden of anemia: a hospital-based surveillance study. PLoS Negl Trop Dis 2015; 9:e0004195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Boyle MJ, Chan J-A, Handayuni I, et al. IgM in human immunity to Plasmodium falciparum malaria. Sci Adv 2019. In press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Flynn TP, Allen DW, Johnson GJ, White JG. Oxidant damage of the lipids and proteins of the erythrocyte membranes in unstable hemoglobin disease. Evidence for the role of lipid peroxidation. J Clin Invest 1983; 71:1215–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Davies KJ, Goldberg AL. Oxygen radicals stimulate intracellular proteolysis and lipid peroxidation by independent mechanisms in erythrocytes. J Biol Chem 1987; 262:8220–6. [PubMed] [Google Scholar]
- 39.Eda S, Sherman IW. Cytoadherence of malaria-infected red blood cells involves exposure of phosphatidylserine. Cell Physiol Biochem 2002; 12:373–84. [DOI] [PubMed] [Google Scholar]
- 40.Mourão LC, da Silva Roma PM, Aboobacar JdSS, et al. Anti-erythrocyte antibodies may contribute to anaemia in Plasmodium vivax malaria by decreasing red blood cell deformability and increasing erythrophagocytosis. Malar J 2016; 15:397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Dondorp AM, Angus BJ, Chotivanich K, et al. Red blood cell deformability as a predictor of anemia in severe falciparum malaria. Am J Trop Med Hyg 1999; 60:733–7. [DOI] [PubMed] [Google Scholar]
- 42.Barber BE, Russell B, Grigg MJ, et al. Reduced red blood cell deformability in Plasmodium knowlesi malaria. Blood Adv 2017; 2:433–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.