X-ray microanalysis investigation of the changes in Na, K, and hemoglobin concentration in plasmodium falciparum-infected red blood cells

Abstract

Plasmodium falciparum is responsible for severe malaria. During the ∼48 h duration of its asexual reproduction cycle in human red blood cells, the parasite causes profound alterations in the homeostasis of the host red cell, with reversal of the normal Na and K gradients across the host cell membrane, and a drastic fall in hemoglobin content. A question critical to our understanding of how the host cell retains its integrity for the duration of the cycle had been previously addressed by modeling the homeostasis of infected cells. The model predicted a critical contribution of excess hemoglobin consumption to cell integrity (the colloidosmotic hypothesis). Here we tested this prediction with the use of electron-probe x-ray microanalysis to measure the stage-related changes in Na, K, and Fe contents in single infected red cells and in uninfected controls. The results document a decrease in Fe signal with increased Na/K ratio. Interpreted in terms of concentrations, the results point to a sustained fall in host cell hemoglobin concentration with parasite maturation, supporting a colloidosmotic role of excess hemoglobin digestion. The results also provide, for the first time to our knowledge, comprehensive maps of the elemental distributions of Na, K, and Fe in falciparum-infected red blood cells.

Figure 1

Representative x-ray spectra of uninfected RBCs, falciparum-infected RBCs, and parasite cytoplasms. The element label in each panel is directly above the respective K_α1-transition peak. (A) Uninfected RBC, high in K and low in Na; (B) RBC host to trophozoite-stage parasite, high in Na and low in K. (C) Parasite cytoplasm, high in K and low in Na; the large P peak signals high metabolic activity. The present Ni peak of the EM grid and the C, O, and N peaks of the sample are beyond the energy range shown. The Si-peaks are an artifact stemming from the desiccant.

Figure 2

EPXMA elemental maps of the Na, K, and Fe distributions in uninfected RBCs and infected RBCs with trophozoite-stage parasite. EM images (top row) are shown vertically aligned with the corresponding EPXMA images (bottom row). Correspondence between the photographed image and the mapped field is not perfect because the transmission electron microscopy imaging camera used is not an integral part of the mapping system. The EPXMA maps are assembled as a red-green-blue overlay of the Na (green), K (red), and Fe (blue) x-ray raw peak count for each spot. Mixtures of these colors can result, e.g., in yellow (both high K and Na), or violet (both high Fe and K). (A and B) Uninfected RBC surrounded by salt crust from the dried extracellular medium. (C and D) IRBC with trophozoite-stage parasite and food vacuole. The black spots are hemozoin crystals responsible for the high Fe signal. (E and F) IRBC with trophozoite-stage parasite next to two uninfected RBCs (white starred).

Figure 3

Electron-probe-measured contents and concentrations of Na, K, and Fe in uninfected RBCs. Bars and error bars represent means and standard error of the means. (A and C) Electron-probe raw readings of Na, K, and Fe. (B and D) The Na, K, and Fe signals were converted to concentrations in mmol/(liter cell water) as explained in Materials and Methods; the Fe signal was converted to tetrameric hemoglobin concentration, [Hb]. (A and B) Cytosol composition of Na, K, and Fe of control RBCs (n = 33) and of uninfected cohort RBCs from mature trophozoite cultures (n = 5). (C and D) Cytosol composition of Na, K, and Fe of nystatin-pretreated RBCs in high-K media (n = 20) and in high-Na media (n = 25). p < 0.001 (t-test) for the differences in Na and K contents (C) and concentrations (D) between high-K and high-Na conditions; p < 0.01 for the corresponding differences in Fe (C) and [Hb] (D). (E) Comparison of the Na/K concentration ratio for the four samples above. p < 0.001 for the difference between the ratio in the high-Na nystatin condition and the other three conditions.

Figure 4

(A) Predicted stage-related changes in IRBC [Na]+[K] concentrations. (B) Electron-probe-based estimates of Na and K concentrations in the cytoplasms of host RBCs and parasites, obtained from IRBCs with mature parasites, and Hb concentrations estimates in cytoplasm of host RBCs (same IRBCs as for Na and K concentrations). (Inset of B) Na/K concentration ratio in host RBCs and parasites. (A) Stage-related changes in the [Na]+[K] concentration of IRBCs predicted by the IRBC model for different PK/PNa selectivities through NPPs: 2.3 (solid symbols) and ∼1 (open symbols). With these patterns, the value of f was estimated with [Na]+[K] = 150 mmol/Lcw for IRBCs with ring-stage parasites, and with [Na]+[K] = 156 mmol/Lcw for IRBCs with trophozoite-stage parasites and activated NPPs, as estimated from elevated Na/K concentration ratios. (B) Bars represent mean and standard error of the mean of 15 host cells and 14 parasites. The Fe signal from the red cell cytosol was converted to tetrameric Hb concentration [Hb] as explained in Materials and Methods. The Fe signal from parasite cytosol was ignored for this graph. (Inset) Na/K concentration ratio, confirming the marked difference in Na-K composition between host and parasite cytoplasms at the trophozoite developmental stage. p < 0.001 for the differences in [Na], [K], and Na/K ratio between host and parasite. The algorithm used for these simulations is described in Mauritz et al. (18). Using the nomenclature from that article, the other model parameters were: CF = 0.3, Hb_max = 0.7, t_Hb = 32 h, t_Hb = 27 h, and s_NPP = s_Hb = 3 h⁻¹.

Figure 5

Correlation between changes in hemoglobin concentration and Na/K concentration ratio in the cytoplasm of IRBCs with trophozoite stage parasites. Increases in Na/K ratio caused by progressive dissipation of Na-K gradients through the NPP-permeability pathway are assumed to reflect advancing stages in the asexual reproduction cycle of the parasite within its host red cell. The results report electron-probe-based measurements of [Hb] in 155 IRBCs from eight different cultures. The strains used for each point are shown on the figure. (Points and cross-hairs) Mean and standard error of the mean of independent measurements obtained from between 15 and 25 cells in each sample. (Open circles) IRBCs; (square) uninfected controls (n = 33); and (triangle) uninfected cohort RBCs (n = 5).

Figure 6

Model analysis of the electron-probe-measured [Hb]-decline patterns. The model of IRBC homeostasis (18,19) was used to search for the parameter variations required to provide approximate fits to the measured steep and gentle [Hb]-decline patterns reproduced here in the experimental points (A). A single parameter change, P_K/P_Na, the Na-K selectivity of the NPP permeability pathway, proved necessary and sufficient to provide the fits depicted here in panel A for the curves outlined by solid (P_K/P_Na = 2.3) and open (P_K/P_Na = 0.9) square symbols, corresponding to the gentle and steep [Hb]-decline patterns, respectively. All other parameters were the same for both simulations. Using the nomenclature in Mauritz et al. (18): CF = 0.3, Hb_max = 0.7, t_1/2(NPP) = 27 h, t_1/2(Hb) = 32 h, and s_NPP = s_Hb = 3 h⁻¹. Open circles as in Fig 5. (B–D) Model-predicted, time-dependent changes in selected model variables relevant for the understanding of the mechanism behind the P_K/P_Na-generated different patterns, analyzed in detail in Discussion. Time-dependence is reported as a function of time postinvasion. Model simulations with (solid symbols) and without (open symbols) PK/PNa selectivity.