Sickle-trait hemoglobin reduces adhesion to both CD36 and EPCR by Plasmodium falciparum-infected erythrocytes

Abstract

Sickle-trait hemoglobin protects against severe Plasmodium falciparum malaria. Severe malaria is governed in part by the expression of the Plasmodium falciparum Erythrocyte Membrane Protein 1 (PfEMP1) that are encoded by var genes, specifically those variants that bind Endothelial Protein C Receptor (EPCR). In this study, we investigate the effect of sickle-trait on parasite var gene expression and function in vitro and in field-collected parasites. We mapped var gene reads generated from RNA sequencing in parasite cultures in normal and sickle-cell trait blood throughout the asexual lifecycle. We investigated sickle-trait effect on PfEMP1 interactions with host receptors CD36 and EPCR using static adhesion assays and flow cytometry. Var expression in vivo was compared by assembling var domains sequenced from total RNA in parasites infecting Malian children with HbAA and HbAS. Sickle-trait did not alter the abundance or type of var gene transcripts in vitro, nor the abundance of overall transcripts or of var functional domains in vivo. In adhesion assays using recombinant host receptors, sickle-trait reduced adhesion by 73-86% to CD36 and 83% to EPCR. Similarly, sickle-trait reduced the surface expression of EPCR-binding PfEMP1. In conclusion, Sickle-cell trait does not directly affect var gene transcription but does reduce the surface expression and function of PfEMP1. This provides a direct mechanism for protection against severe malaria conferred by sickle-trait hemoglobin. Trial Registration: ClinicalTrials.gov Identifier: NCT02645604.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. 3D7 var gene expression across the asexual stage in blood from HbAA and HbAS donors.

Var gene transcription relative to total RNA transcripts per million (TPM), every three hours post invasion (HPI) for 48 hours. Var genes are grouped according to their UPS classification, and individual var genes are represented by different colors according to their open reading frame (ORF). Solid lines represent average transcription of two HbAA donors, and dashed lines the average transcription of two HbAS cultures, and shaded areas are the min and max replicate value. Note difference in y-axis scales between panels.

Fig 2. Sickle-cell traits effect on parasite adhesion to CD36.

A) HbAA or HbAS RBCs infected with 3D7 reference parasite strain adhering to spots of recombinant CD36 on a petri-dish. B) Quantification of the adhesion of 3D7-infected HbAA and HbAS erythrocytes to 10 μg/ml CD36 and EPCR spots. Each condition was done as 2 protein spots on separate petri-dishes, each imaged 3 times. The assay was done 3 independent times (n = 3). Wilcoxon’s signed rank test was used to evaluate statistical significance (*** p-value = 0.0001). C) Relative adhesion across protein spots of different concentrations normalized to the mean adhesion to CD36 at 10 μg/ml. Four parameter logistic curve was fitted to the normalized data for HbAA and HbAS. D) HbAA or HbAS RBCs infected with IT4 strain parasites expressing IT4var30-PfEMP1 (IT4var30) adhering to spots of recombinant CD36 on a petri-dish. E) Quantification of the adhesion of It4var30-infected HbAA and HbAS erythrocytes to 10 μg/ml CD36 and EPCR spots. Each condition was done as 2 protein spots on separate petri-dishes, each imaged 3 times. Wilcoxon’s signed rank test was used to evaluate statistical significance (*p-value = 0.03). F) Relative adhesion across protein spots of different concentrations normalized to the mean adhesion to CD36 at 10 μg/ml. Four parameter logistic curve was fitted to the normalized data for HbAA and HbAS (n = 1).

Fig 3. Sickle-trait hemoglobin effects on erythrocyte adhesion to EPCR.

A) Erythrocytes infected with IT4 parasite-strain expressing IT4var20 adhering to spots of recombinant EPCR on a petri-dish. B) Quantification of IT4var20-infected normal (HbAA) erythrocytes and sickle-cell trait (HbAS) erythrocytes adhesion to 10 μg/ml EPCR and CD36 spots. Each condition was done as 2 protein spots on separate petri-dishes, each imaged 3 times. The assay was done 3 independent times (n = 3). Wilcoxon’s signed rank test was used to evaluate statistical significance (** p-value = 0.0002). C) Relative adhesion across protein spots at different concentrations normalized to the mean adhesion to EPCR at 10 μg/ml. Four parameter logistic curve was fitted to the normalized data for HbAA and HbAS.

Fig 4. Effects of sickle-trait hemoglobin on transcript and surface protein expression of IT4var20 PfEMP1.

A) Quantitative PCR of IT4var20 transcription in HbAA, HbAS, and HbAC converted to Transcript units relative to seryl-tRNA synthetase housekeeping gene (n = 1). B) Histogram of IT4var20 PfEMP1 expression from a representative flow cytometry experiment. The PfEMP1 fluorescence intensity was measured from infected HbAA, HbAS, and HbAC RBCs, stained with monoclonal IgG antibody targeting the CIDRα of IT4var20 and APC-conjugated anti-mouse IgG. The median flourescence is marked with a vertical line. C) Summary of the median fluorescence intensity (MFI) from separate flow cytometry experiments (n = 3). Error bars show the standard deviation.

Fig 5. Var domain expression in HbAA and HbAS children in Mali.

A) PfEMP1 domain schematic showing the domain composition of the receptor-binding headstructures of CD36-binding PfEMP1s belonging to group B, C, and B/C; EPCR-binding PfEMP1s belonging to Group A, and B/A; and CSA-binding PfEMP1s, Var2. B) Total read counts for the acidic-terminal segment (ATS) sequences for each of the 32 patient samples. Boxplots show the distribution according to hemoglobin genotype, and the individual patients samples are represented by dots. C) Normalized read counts for DBL and CIDR domains according to head structure within each patient. Read counts were normalized to ATS domain read counts for each patient. A linear model trendline is marked in blue. D) Proportion of read counts for DBLα0, DBLα1, DBLα2, and DBLpam1 domains normalized to ATS read counts in individual patient samples. Colors correspond to head structure type. E) Proportion of ATS-normalized readcounts for CIDRα2–6, CIDRα1, and CIDRpam domains. F) Normalized read counts of DBLα0, DBLα1, DBLα2, and DBLpam1 domains in children with HbAA and G) Normalized read counts for CIDRα2–6, CIDRα1, and CIDRpam domains in children with HbAA Boxplots show distribution according to hemoglobin genotype, and the individual patient samples are represented by dots. Boxplots show median, interquartile range (IQR), and whiskers show 1.5 times the IQR.