Ethical statement
The study was approved by the Noguchi Memorial Institute for Medical Research Institutional Review Board (NMIMR STC Number: STC Paper 5(1) 2013–2014) and by the Ethical Review Committee of the Ghana Health Service (026/13-14). Declaration of free willingness to participate in the study and written informed consent was obtained from parents/guardians of all study participants prior to enrolment.
Study area
This study was carried out within the frame of a broader study aimed at building malaria vaccine research capacity in Ghana (MAVARECA https://mavareca.ku.dk/). The participants were enrolled in Hohoe, a town located in the Volta Region about 220 km northeast of Accra. Malaria transmission intensity in the area is high with approximately 65 infectious bites per person per year and has two seasonal peaks, a major one in April‑July and a minor one in September‑November [27].
Study participants and laboratory tests
Children 1-12 years of age and reporting with P. falciparum malaria to Hohoe Municipal Hospital were enrolled in the study between July-August 2016. A sample size of 96 children was calculated based on 40,092 malaria cases confirmed at the Hohoe Municipal hospital in 2014 [28], using a level of confidence of 95%, sampling error of 10%, and 50% expected prevalence for IgM or α2M binding. After enrolment, a project nurse and a physician completed a standardized questionnaire and performed a clinical examination. Severe malaria (SM) was defined according to the WHO criteria [29], and children were treated with artemether‑lumefantrine or quinine IV as required. Venous blood samples were taken on the day of admission to determine haemoglobin levels (Hb), ABO blood group, and for research purposes. Samples were taken daily during the hospitalization and one week after initial presentation to assess haemoglobin levels and parasitaemia. Sickle cell Hb phenotype was determined by electrophoresis and glucose‑6‑phosphate dehydrogenase (G6PDH) deficiency tested by methylene blue reduction test [30].
Field isolates and in vitro culture
After removal of plasma, the pellet containing IEs was washed twice in RPMI 1640 medium (Sigma-Aldrich, Germany) supplemented with 50 µg/mL gentamicin (Sigma-Aldrich, Germany). A 100 µL aliquot of the pellet was placed in RPMI 1640 medium supplemented with 0.5% AlbuMAX II (Gibco-Life Technologies, Denmark), 2% heat-inactivated normal human serum (NHS), 2 mM glutamine (Sigma-Aldrich, Germany), and 50 µg/mL gentamicin (referred to as 2% complete culture medium) before culturing at 2% haematocrit. The parasites were incubated at 37°C in 2% O2, 5% CO2, and 93% N2 atmosphere, before carrying out rosetting assays.
The rest of the washed pellet was gently mixed with glycerol freezing solution, aliquoted into cryotubes, and transferred to liquid nitrogen for long-term storage. Frozen samples were shipped to the University of Copenhagen, where they were thawed by standard methods before starting short-term in vitro cultures [31]. Briefly, the parasites were placed in 2% complete culture medium at 2% haematocrit and incubated at 37% in 2% O2, 5% CO2, and 93% N2 atmosphere. Parasitaemia was checked daily by Giemsa‑stained thin smears and those with normal morphology that matured to the late‑trophozoite stage were included in the study. The day before use in experiments including serum proteins, the parasites were transferred to serum-free RPMI 1640 medium (0.5% AlbuMAX II).
Rosetting assays
The rosetting of fresh isolates was assessed in the first cycle of in vitro growth when the majority of the parasites had reached the trophozoite-late stage. After staining the parasites with 7 μg/mL of ethidium bromide for two minutes, the percentage of rosettes was assessed by counting 200 ethidium bromide-stained IEs, using wet preparations and fluorescence microscopy. Rosettes were defined as IEs having two or more adhering uninfected erythrocytes.
To determine the role of non-immune IgM and a2M binding in rosetting, short-term in vitro cultures of thawed cryostabilates were used. Late trophozoite stage IEs at 2% haematocrit in serum-free RPMI 1640 medium were incubated 1h at 37°C in 2% O2, 5% CO2, and 93% N2 atmosphere with 10% NHS, or with 10 nM IgM (Sigma-Aldrich, Germany), 10 nM a2M (Sigma-Aldrich, Germany), or both. The percentage of rosettes was assessed as described above.
Measurements of non-immune IgM and α2M binding to PfEMP1
Non-immune IgM and a2M binding to IEs was detected by flow cytometry as previously described [24, 26, 32]. Binding of both proteins at 10 nM has been detected in several laboratory clones [23-26], but included two additional concentrations (1nM and 100 nM) to explore the potential binding in the field isolates. Briefly, intact and unfixed late trophozoite stage IEs purified by magnet-activated cell sorting (MACS) were incubated with either 1, 10, or 100 nM non-immune human IgM or a2M in PBS supplemented with 1% Ig-free bovine serum albumin (PBS 1% BSA). IE-bound IgM was measured with a FITC-conjugated anti-human IgM (1:150; Sigma), whereas a2M was determined with a combination of goat polyclonal anti-human α2M (1:3,000; Abcam, UK) and FITC-conjugated anti-goat IgG (1:150; Vector, UK). Non-immune IgM and a2M binding to IEs was quantified as the median fluorescence intensity (MFI) in IEs labelled with 10 µg/mL Hoechst. BD LSR II flow cytometer was used for data acquisition and FlowLogic software (Inivai Technologies, Australia) for data analysis.
Statistical analysis
Data were analysed and plotted using IBM SPSS Statistics for Macintosh, Version 26.0 (IBM Corp) and GraphPad Prism version 8.0 (GraphPad Software, San Diego, California, USA), respectively. Nominal variables were analysed using descriptive statistics. The Mann‑Whitney U or Friedman test and Kruskal-Wallis test followed by Dunn’s multiple comparison test were used to compare two and more-than-two groups, respectively. Spearman’s rank correlation (rs) was used to assess the correlation between numeric variables. Fisher’s exact test was used to compare proportions. Any p‑value < 0.05 was considered statistically significant.
Multiple linear regression models were used to evaluate the effect of potential confounders on the relationship between percentage of rosettes or protein binding (dependent variables) and relevant independent factors. Models were adjusted by clinical category (UM and SM), age, parasitaemia, haemoglobin levels at admission, ABO blood group, and days of sickness.