Expression of Merozoite Surface Protein Markers by Plasmodium falciparum-Infected Erythrocytes in Peripheral Blood and Tissues of Children with Fatal Malaria

Patient samples.Postmortem samples were collected from 15 children less than 12 years old who died at a pediatric research ward between 1996 and 1998. These patients had (i) CM (with a Blantyre coma score of <3 [36], asexual parasitemia, and no other obvious cause, after correction for hypoglycemia and recovery from convulsion] (n = 7); (ii) CM and severe malarial anemia (hemoglobin level of <5 g/dl or hematocrit value of <15%) (n = 3); or (iii) coma or severe illness other than malaria (two had incidental parasitemia) (n = 5). On admission, we determined hematocrit and parasitemia levels. Thick and thin blood films were stained with Field's stain, and at least 200 leukocytes and 500 erythrocytes, respectively, were counted to record numbers and stages of parasites. Informed consent was obtained from parents or guardians, and ethical approval for the study was obtained from the investigators' institutions.

Blood parasites.For 10 of the 15 children, 5 ml of venous blood was obtained on admission. The blood was centrifuged to remove the buffy coat and plasma, and the erythrocytes were washed in sterile phosphate-buffered saline (PBS). Seven children had asexual parasites as determined by blood smears, and short-term in vitro cultures of six samples that had levels of parasitemia of >5% were set up. pRBC from three patients with fatal cases were successfully matured to express MSPs and knobs (11, 49), and three cultures failed to grow due to previous antimalarial treatment. Parasitized erythrocytes were harvested from in vitro cultures and washed twice in PBS. Cells were resuspended to 3 to 5% hematocrit in PBS, and 20- to 25-μl aliquots were placed onto wells of 12-well multispot slides (Hendley-Essex, United Kingdom). The slides were dried, packed, and stored at −20°C in self-sealing plastic bags containing silica gel as a desiccant. Residual pRBC were stored at −70°C for analysis of parasite DNA.

Immunofluorescence antibody typing (IFAT).Type-specific antibodies identify distinct allelic forms of P. falciparum MSP-1, MSP-2, and EXP-1 and thus can detect the presence of multiple-serotype mixed infections in individuals (9). The serological reagents used included monoclonal antibodies (MAbs) and polyvalent mouse antibodies specific for P. falciparum MSP-1 and -2 and EXP-1 (Table 1). Three known MSP-1 block 2 types, represented by the RO33, MAD20, and K1 isolates, were identified by specific MAbs (11, 28) and/or polyclonal antibodies (6). Other MAbs identified MSP-1 polymorphisms in blocks 3 and 4, in dimorphic regions (blocks 6 to 16) of the MAD20 type or Well/K1 type, and conserved epitopes (block 17) (9, 10). Polymorphisms between and also within the two major serogroups of MSP-2, serogroups A and B, were similarly identified by MAbs (17) or antibodies against recombinant MSP-2. The typing method described previously was used (9, 11). Briefly, slides with pRBC were fixed in acetone, and 25-μl portions of working dilutions of typing antibodies were placed in separate wells and incubated for 30 min at room temperature in a wet box. After the antibodies were removed, the wells were washed three times with PBS, and the slides were dried on a warm plate, 15 μl of fluorescein isothiocyanate (FITC)-conjugated anti-mouse immunoglobulin (1/80) was placed in each well, incubated, and washed as described above. Parasite DNA was stained with 4′,6′-diamino-2-phenylindole (DAPI) (1:100,000) for 5 min, and the blood films were counterstained with 0.1% Evans blue. After 5 min, the slides were rinsed with PBS and mounted under coverslips with Citifluor (City University, London United Kingdom) or 50% glycerol in PBS. Reactions were read by using a magnification of ×315 or ×630 and incident light at wavelengths of 450 to 490 nm for FITC fluorescence (green) or 390 to 440 nm for DAPI (blue). Serological marker epitopes are referred to below by the same code numbers as the MAbs that were used to identify them. For each isolate and MAb, the percentage of parasites exhibiting antibody-specific positive fluorescence was recorded.

Multiple-clone infections were analyzed by double-labeled (two-color) IFAT using combinations of two MAbs for different isotypes against different epitopes, as described previously (9, 10, 15) (Table 1). Briefly, FITC- and rhodamine isothiocyanate (RITC)-conjugated anti-mouse immunoglobulins (1:50) specific for the isotypes of the typing MAbs were incubated together in the second stage. The slides were not counterstained with Evans blue. Reactions were read by using incident light at wavelengths of 515 to 560 nm for RITC fluorescence (red). The proportions of schizonts showing (i) only green (and blue) fluorescence, (ii) only red (and blue) fluorescence, (iii) red and green (and blue) fluorescence, and (iv) neither red nor green (only blue) fluorescence were recorded for each pair of MAbs tested. Combined results, obtained with a series of different pairs of MAbs, resolved the number of distinct parasite clones within each isolate and the phenotypes. At least 200 schizonts per test were scored.

PCR genotyping. P. falciparum DNA extracted from 20 to 100 μl erythrocytes by the quick boiling method (18) was used in a hot-start nested PCR to analyze the multiplicity of infection (MOI). Block 2 of MSP-1 was amplified with primers O1 and O2 (outer reaction) (41) and a set of block 2 type-specific primers (inner reaction) (6). Dimorphic regions of MSP-1 and of MSP-2 were typed by the dimorphic form-specific PCR method (42).

Autopsy specimens.Postmortem specimens were obtained from 14 autopsies and a supraorbital needle sample (patient SO). Parasitemia was recorded less than 4 h before death for 13 of 15 patients, and the times between admission to the hospital and death and between death and autopsy were noted for all cases. Tissue biopsies from the cerebral hemisphere, cerebellum, lung, heart, liver, kidney, and spleen were processed in four different ways (1). Homogenates were prepared by grinding a piece of tissue (2 by 2 by 1 cm) with a pestle and mortar in PBS, washing the material twice in PBS, and placing 20 to 25 μl of the suspension onto 12-well multispot slides (2). Smears were prepared using small pieces of tissue crushed between two slides which were then pulled apart (3). Imprints were prepared by pressing a slide against a section of the organ. All slides were dried and stored at −20°C for IFAT (4). Pieces of tissue that were 7 by 3 by 3 mm were covered in Tissue-Tek optimal cutting temperature compound (Ted Pella Inc., Redding, CA) and flash frozen. Eight-micrometer frozen sections of cryopreserved tissue on slides were kindly provided by Georges Grau, Geneva, Switzerland. To assess the degree and stage of parasite sequestration, brain smears were fixed in methanol, stained with Giemsa or Field's stain, and examined with an oil immersion lens (magnification, ×100). Parasites sequestered in tissues were typed by immunofluorescence analysis, as described above (9, 15).

Detection of sequestered parasites.Using light microscopy (Fig. 1A) and IFAT, we found mature parasites sequestered in the brain, spleen, liver, and lung but not in the heart or kidney in nine CM patients who died. No sequestered parasites were detected in five non-CM patients who died. The presence of sequestered parasites in CM patients who died was confirmed by the colocalization of fluorescent staining of P. falciparum DNA and of surface proteins with antibodies. Homogenates, smears, and cryosections (Fig. 1) of brain tissues were suitable for IFAT typing. Although the morphology of the tissue was disrupted by homogenization, most capillaries containing pRBC remained intact, and smears preserved individual parasite morphology. Cryosections had the advantage that the histological structure was maintained. Estimation of the proportions of serotypes in multiple infections was easier with smears and cryosections than with homogenates, as capillaries did not clump together. Organ imprints were more appropriate than homogenates for the lung, spleen, and liver (Fig. 2).

• Open in new tab
• Download powerpoint

FIG. 1.

Detection of parasites in brains from children who died from CM. (A) Thin smear prepared from brain tissue (patient 13), methanol fixed and Giemsa stained, examined with a light microscope, showing capillaries with sequestered segmenter schizonts (S), trophozoites (T), and malaria pigment. The pink oval bodies are the nuclei (N) of the capillary endothelial cells. Magnification, ×630. (B) Brain homogenate (patient 6) showing capillaries with sequestered P. falciparum schizonts stained with DAPI. The blue dots correspond to the DNA of the multinucleated schizonts (S), and the blue oval bodies correspond to endothelial cell nuclei (N). Magnification, ×315. (C) Same sample as the sample in panel B, stained with FITC. Sequestered P. falciparum schizonts (S) were recognized by MAb 18.2, specific for a malaria knob antigen, at exactly the same positions at which blue fluorescent dots are present in panel B. Note that the lower capillary appears to be full of parasites, while the upper capillary is almost empty. N, nuclei. (D to F) Cross sections of cryopreserved brain (patient 6). (D) Capillaries containing fluorescent schizonts stained with DAPI. Blue round bodies correspond to brain cell nuclei (N) and sequestered schizonts (S). Magnification, ×315. (E) FITC staining of the sample in panel D, stained with MAb 113.2 (immunoglobulin G3) specific for MSP-2 serogroup A. Only schizonts (S) are recognized by this MAb. (F) RITC staining of the sample in panel D, stained with MAb 5.1 (immunoglobulin G1) specific for EXP-1. Both schizonts (S) and trophozoites are recognized by this MAb. For details concerning the methods and antibody reagents used see Materials and Methods and Table 1.

• Open in new tab
• Download powerpoint

FIG. 2.

Detection of parasites in organs of children who died from CM. (A, B, and C) Lung imprints. (A) Parasites detected in a lung imprint (patient 27) stained with DAPI. Trophozoites (T) and one schizont (S) are surrounded by lung cells. N, nuclei. Magnification, ×630. (B) Sample in panel A stained with FITC-conjugated MAb 111.4 (immunoglobulin G1) specific for an epitope in block 17 of MSP-1 expressed in both trophozoites (T) and schizonts (S). (C) Sample in panel A stained with RITC-conjugated MAb 123D3 (immunoglobulin G2b) specific for the K1 type (Palo Alto variant) of block 2 of MSP-1. Only one schizont (S) is recognized by this MAb. (D and E) Splenic imprints. (D) Fluorescent schizonts (patient 6) stained with DAPI. The blue dots in the center (S) correspond to schizont nuclei. The blue round bodies (N) correspond to the nuclei of spleen cells, Magnification, ×630. (E) Sample in panel D stained with FITC-conjugated MAb 9.8 (immunoglobulin G) specific for conserved epitopes of MSP-1. S, schizont. (F and G) Hepatic imprints. (F) Four parasites, two schizonts (S) and two trophozoites (T), were detected with DAPI (patient 13). Magnification, ×630. (G) Sample in panel F stained with FITC-conjugated MAb 31.1 (immunoglobulin G) specific for epitopes of the RO33 block 2 type of MSP-1. Only schizonts were detected by MSP-1-specific MAbs. Trophozoites were negative as determined by FITC fluorescence. More than one P. falciparum clone was identified in sequestered parasites from patient 13 (Table 2), and the clones were distinguished by different MSP-1 block 2 types. The FITC-fluorescent schizont (S) belonged to the RO33 block 2 type, whereas the negative unstained schizont belonged to the K1 block 2 type. (H) Fluorescent schizonts (S) detected in another spleen imprint (patient 27) stained with FITC-conjugated MAb 111.4 (immunoglobulin G1) specific for epitopes in block 17 of MSP-1. Both trophozoites and schizonts are recognized by this MAb. Magnification, ×315. (I) Spleen sample in panel H stained with RITC-conjugated MAb 123D3 (immunoglobulin G2b) specific for the K1 type (Palo Alto variant) of block 2 of MSP-1. Only schizonts (S) are recognized by this MAb. (J) Fluorescent schizonts (S) detected in another liver imprint (patient 27) stained with FITC-conjugated MAb 111.4 (immunoglobulin G1) specific for epitopes in block 17 of MSP-1. Magnification, ×630. (K) Liver sample in panel J stained with RITC-conjugated MAb 123D3 (immunoglobulin G2b) specific for the K1 type (Palo Alto variant) of block 2 of MSP-1. S, schizont.

Immunofluorescence serotyping of sequestered parasites.Mature parasites sequestered in tissues from 10 CM patients were typed using a panel of antibodies specific for MSP-1, MSP-2, and EXP-1 (Table 2). All these parasites also reacted positively with knob-specific MAbs (not shown). The antigen phenotypes of parasite clones of each isolate and their relative proportions were determined by combining the results obtained with different pairs of antibodies. Antibodies that did not react positively in any of the samples indicated antigenic types that were not expressed in the parasites examined (MAD20 type MSP-1₁₇, K1 type MSP-1_6-16, and particular polymorphic repeat epitopes of MSP-2) and are not included in Table 2.

In addition to schizonts, in most cases trophozoites were also present in the tissues, as detected by Giemsa staining and IFAT (Fig. 1 and 2). Thus, sequestered pRBC contained mostly somewhat asynchronous parasite broods, and parasites at different stages were present at each tissue site. In such cases, trophozoites were detected by DNA staining but were negative for most typing antibodies, except those recognizing EXP-1 (antibody 5.1), knob antigens, and block 17 of MSP-1 (antibody 111.4), which are expressed earlier in the blood asexual cycle and are normally present on early trophozoites. In two patients, the sequestered parasites consisted of synchronous trophozoites that were negative for the MSP-1 (except antibody 111.4) and MSP-2 antibodies. Typing of these sequestered parasites was thus restricted to EXP-1, which distinguished between parasites from patients 16 (positive) and 21 (negative).

MOI in tissues.Parasites found in tissues often belonged to more than one P. falciparum serotype (Table 2 and Fig. 3). Using double-labeled IFAT typing, we determined the proportion of each serotype in multiple infections in the cerebral hemisphere and cerebellum (Table 2) and, whenever possible, in the spleen. The spleen, liver, and lung generally contained fewer parasites than the brain, and thus it was possible to determine only the presence or absence of each antigenic type, without quantification or phenotype resolution. For the eight CM cases in which tissue parasites could be serotyped, there was one single infection, three cases with at least two serotypes, and four cases with at least three serotypes. The MOI was 2.37, which is similar to the MOI determined by PCR for the blood of 93 CM patients who did not die admitted during the study period (MOI, 2.25) (15).

• Open in new tab
• Download powerpoint

FIG. 3.

Detection of multiple-clone infections in blood and brain by double-labeled IFAT: MSP-1 serotypes of P. falciparum in peripheral blood and brain tissue of patient 13. (A and B) P. falciparum schizont-infected erythrocytes cultured from the peripheral blood (A) and homogenate from the frontal lobe of the brain (B) of the same child. A majority of schizonts (red) reacted with MSP-1 block 4-specific MAb 10-2B (immunoglobulin G2a) plus RITC-conjugated anti-immunoglobulin G2a. A minority of schizonts (green) were positive with MAb 12.1 (immunoglobulin G1) plus FITC-conjugated anti-immunoglobulin G1. The detectable serotypes or relative proportions of sequestered and circulating parasites did not differ. Magnification, ×630. (C) Thin smear prepared from a supraorbital needle sample from patient SO, showing a mixed infection of two MSP-2 serotypes in a brain capillary. Some schizonts (red) reacted with MSP-2 group B-specific MAb 8G10/48 (immunoglobulin G2b) plus RITC-conjugated anti-immunoglobulin G2b, whereas others (green) reacted with MSP-2 group A-specific MAb 12.3 (immunoglobulin G1) plus FITC-conjugated anti-immunoglobulin G1. Magnification, ×630.

Parasite serotypes in tissues and blood.Partial within-patient comparisons of tissue-sequestered serotypes in the spleen, liver, lung, and brain could be performed for seven autopsies (Table 2). Owing to the limited availability of spleen, liver, and lung specimens and the lower intensity of sequestration, a smaller panel of antibodies to dimorphic epitopes was selected for typing parasites detected in these tissues. In patient 27, the same serotypes were detected in the spleen, liver, and lung as in the brain, and they were in the same proportions in the spleen and the brain. For the other six patients only a partial comparison could be made, but we did not detect any serotype in the spleen, liver, or lung that was not also present in the brain of the same patient. In summary, in each case, the parasites detected in the spleen, liver, or lung were the same MSP-1/2 and/or EXP-1 types as the parasites in the other organs and in the brain; therefore, we concluded that within each patient there was the same qualitative distribution of sequestered P. falciparum types.

Between-patient comparisons of tissue-sequestered pRBC revealed some similarities in the serotypes detected (Table 2). Thus, all parasite clones belonged to one of two main dimorphic serotypes of MSP-1 (MAD20), none belonged to the alternative main type (K1), and most expressed the same epitopes in blocks 2 (predominantly the K1 type), 3, and 4 (Table 2). For MSP-2, most CM patients were infected with serotype A parasites (IC1-like); serotype B (FC27-like) was found in only one patient in tissues. Finally, epitope 5.1 of EXP-1 was positive in nearly all cases; the only exception was patient 27. Overall, these epitope prevalences in CM patients who died reflected the prevalences also found predominantly in the blood of children with nonfatal cerebral or uncomplicated malaria in Blantyre (15).

Quantitative comparisons of the serotype proportions in the brain and blood could be made for patients 6 (two serotypes) and 13 (three serotypes) (Fig. 3). Table 2 shows the results of IFAT typing of blood isolates that matured into schizonts in vitro in the same patient for which tissue typing was performed. In these two autopsies, the same serotypes were detected with very similar proportions in both the blood and tissues. In patient 16, blood parasites could be cultured and serotyped (three serotypes), but it was not possible to compare these parasites to parasites in the brain because sequestered stages were too young (trophozoites) to be fully typed by MSP-1/2 antibodies.

For the rest of the cases, the blood samples were too small for P. falciparum culture or the parasites failed to mature (e.g., owing to prior antimalarial treatment), and thus we determined circulating genotypes by PCR analysis. Genotyping revealed that all patients who died had P. falciparum parasites in the peripheral blood on admission, and many were infected with multiple genotypes. A more detailed account of the PCR approach was provided in a previous study (37), and here this method was used as a supplement to protein typing to validate the comparisons in cases where IFAT with blood was not available. Thus, in patients 6, 13, and 16, PCR genotyping of blood was also performed, and the genotypes found corresponded to the serotypes shown in Table 2. For patient 16, PCR proved to be more sensitive than IFAT, as IFAT with blood quantified two serotypes (K1 and RO33) whereas PCR detected three genotypes (K1, RO33, and MAD20). For three patients, serotypes of sequestered parasites could be compared only with PCR-determined genotypes of circulating parasites. For patients 11 and SO we found the same multiplicity of infection in blood and tissues, and the genotypes of MSP-1 (polymorphic block 2 and dimorphic blocks 6 to 16) and MSP-2 (dimorphic) in blood matched the corresponding serotypes in tissues. In patient 15, the peripheral parasitemia was low, but the brain and spleen were full of schizonts. For this patient, IFAT detected three MSP-1 block 2 serotypes in the brain (K1, K1, and RO33) (Table 2), whereas genotyping of blood detected four MSP-1 block 2 bands (K1, K1, RO33, and MAD20), even with the low parasitemia; a faint MAD20 PCR band in the blood was not identified as expressed protein in tissue, probably owing to the lower sensitivity of the IFAT method for detecting minority clones. In conclusion, in all cases examined, comparisons of serotypes and genotypes in tissues and blood showed that the compositions of MSP-1 and -2 types in tissues and peripheral blood were very similar and that the multiplicities of infection were consistent for parasites circulating in the blood and parasites sequestered in tissues.