Figure 1.
Schematic diagram of positive and negative selection for P. falciparum platelet-mediated clumping.
(A) A platelet-mediated clumping assay is carried out in sterile conditions by co-incubating an unstained mature pigmented trophozoite-stage P. falciparum culture with platelets in suspension. The presence of clumps is confirmed by viewing a wet preparation of parasite culture suspension using bright-field microscopy (400× magnification). Infected erythrocytes (IEs) can be distinguished from erythrocytes (Es) by the presence of pigment (black dots) in IEs that corresponds to the parasite-generated haemozoin. The platelets cannot be seen in these unstained preparations (for visualization of platelets in clumps see references [1], [2]). (B) Magnetic beads coated with an anti-platelet antibody are added to the suspension and become bound to the platelets within the clumps. Binding of beads is confirmed by viewing a wet preparation of parasite culture suspension using bright-field microscopy (400× magnification). (C) A magnet is applied to separate beads and bead-bound clumps (positive fraction, bound to the magnet) from non-clumping IEs and Es (negative fraction, remaining in the supernatant). Each fraction is washed and transferred to a new flask with fresh uninfected Es added as required. The beads can be magnetically separated from the positive fraction the following day after the parasites have invaded fresh E to become non-adherent ring stages. The clumping frequency (CF) of the selected populations can be tested in the next cycle, with further rounds of (+) or (−) selection being performed as required.
Table 1.
Summary of clumping selection for four P. falciparum strains.
Figure 2.
Immunofluorescence assay to detects knobs.
Fixed thin blood smears of Clump+ and Clump− parasites were stained with mAb89 to KAHRP (10 µg/ml in PBS/1%BSA) followed by 1/500 dilution of Alexa Fluor 488 highly cross-absorbed goat anti-mouse IgG in PBS/1% BSA (green, left column) plus 1 µg/ml 4,6-DiAmidino-2-PhenylIndole (DAPI) to stain parasite nuclei (blue, right column). No green fluorescence was seen with the knob-negative strain Dd2 (bottom left) nor with the IgG2a isotype control mAb (not shown). In three separate IFAs on parasite cultures from separate days, HB3 Clump+ parasites consistently showed markedly brighter staining (top left) than HB3 Clump− parasites (middle left). Representative images are shown. Scale bar = 10 µM.
Table 2.
Percentage of knob-positive infected erythrocytes in P. falciparum strains selected for clumping.
Figure 3.
Inhibition of platelet-mediated clumping of P. falciparum infected erythrocytes (IEs) by antibodies to platelet receptors.
Platelets were pre-incubated with antibodies for 45 mins before being used to set up clumping assays. All antibodies were at 20 µg/ml except anti-CD36 mAb clone SMφ (also 2 and 0.2 µg/ml) and anti-CD36 mAb clone FA6-12 (10, 1, 0.1 and 0.01 µg/ml). Platelet-mediated clumping assays were set up at 10% Ht, 1% Pt, and 20% PRP, at pH7.3, and the clumping frequency (percentage of IEs in clumps out of 500 IEs counted) was assessed by fluorescence microscopy of two wet preparations at time point 1 h. To normalize between experiments with different baseline clumping frequencies, the data are shown as percentage of a control with no added antibody. Data shown are the mean and standard error (SE) of CF. (A) Parasite strain IT (six independent experiments), (B) 3D7 (two experiments), (C) HB3 (three experiments; value of SE for anti-CD41 is 29) and (D) Dd2 (two experiments). One-way ANOVA showed a significant effect of antibodies on P. falciparum clumping (p<0.005) for all strains, with CD36 antibodies showing significant inhibition in all cases (Tukey’s multiple comparison test, * p<0.05, ** p<0.005).
Figure 4.
Effect of pH on platelet-mediated clumping of P. falciparum infected erythrocytes (IEs).
Platelet-mediated clumping assays were set up at 10% Ht, 1% Pt, and 20% PRP, in binding medium at pH 7.3 or pH 6.8. The clumping frequency (CF, percentage of IEs in clumps out of 500 IEs counted) was assessed by fluorescence microscopy of two wet preparations at time point 1 h. Graphs show mean and SE of CF for each parasite strain. (A) IT (4 independent experiments), (B) 3D7 (6 experiments), (C) HB3 (2 experiments) and (D) Dd2 (2 experiments). The CF of 3D7 Clump− parasites showed a significant increase at pH 6.8 compared to 7.3 (p<0.01, paired t test). Other comparisons were not significant.
Figure 5.
Comparison of CD36-binding by P. falciparum Clump+ and Clump
− parasite strains. (A) Binding of parasite strain IT to CD36 under flow conditions (0.05 Pa). Parasite cultures at 3% Pt, 1% Ht in BM/1%BSA at either pH 7.3 or pH 6.8 were flowed over microslides coated with 5 µg/ml recombinant CD36. After washing away unbound cells, the numbers of stationary cells per 20× magnification microscope field were counted by direct microscopic observation in six separate areas on the microslide. Binding of an uninfected E control suspension to CD36 was negligible in all experiments. Data shown are mean and SE of data from eight experiments. The Clump+ parasites had 56–89% CF and the Clump− parasites had 0–11% CF. Clump+ parasites showed significantly higher binding to CD36 than Clump− parasites at both pH conditions (pH 7.3 p = 0.0247; pH 6.8, p = 0.0008; paired t test, n = 8 at each pH). (B) Binding of parasite strain 3D7 to CD36 under flow conditions (0.05 Pa). Parasite cultures at 3% Pt, 1% Ht in BM/1%BSA at pH 7.3 were flowed over microslides coated with 5 µg/ml recombinant CD36. Washing and counting were as described above. Data shown are mean and SE from six experiments. The Clump+ parasites had 22–39% CF and the Clump− parasites had 0% CF in all experiments. Clump+ parasites showed significantly higher binding than Clump− parasites (p = 0.0032, paired t test) (C) Binding of parasite strain HB3 to CD36 under flow conditions (0.05 Pa). Conditions as in (B). Data shown are mean and SE from four experiments. The Clump+ parasites had 43–62% CF and the Clump− parasites had 0–5% CF. The Clump+ parasites showed significantly higher binding than Clump− parasites (p = 0.0118, paired t test). D) Adhesion of Dd2 IEs to CD36 (25 µg/ml) under static conditions. Binding assays were performed in BM/1% BSA at 2% Ht, 3% Pt, pH 7.0 for 1 h at 37°C. Unbound cells were removed by gentle washing and the bound cells fixed with glutaraldehyde and stained with Giemsa. At least three CD36 spots were counted on at least two dishes for each parasite culture in each experiment, with 10 fields at 1000× magnification (light microscopy) being counted per spot. No adhesion of IE or E to PBS control spots was seen. Clump+ Dd2 parasites (CF 45%) showed significantly higher adhesion to CD36 than Clump− Dd2 parasites (CF 0%) (p<0.0001, paired t test. One representative experiment out of two is shown).
Figure 6.
Platelet surface marker expression on fresh and stored platelet preparations.
Platelet-rich plasma prepared from freshly drawn whole blood (“fresh platelets”) or from whole blood from the same donor stored at 4°C for one week (“stored platelets”) was diluted in modified Tyrode’s buffer at pH 7.3 or pH 6.8 and incubated for 40 min at room temperature. Surface expression of CD41 (platelet-specific constitutive marker), CD62P (platelet activation marker), and CD36 was assessed by flow cytometry under resting conditions and on platelets stimulated with 50 µM Thrombin Receptor Activator Peptide (TRAP). Specific monoclonal antibodies (mAbs) for each receptor were used along with an isotype control mAb. Data obtained from one donor are shown, and a second independent donor showed very similar results.