ABSTRACT
Previous studies have shown that gamma interferon (IFN-γ) production in the placenta is associated with protection against placental malaria. However, it remains unknown which IFN-γ-producing cell subpopulations are involved in this protection and whether the cellular immune components of protection are the same in the peripheral and the placental blood compartments. We investigated cell subpopulations for CD4, CD8, and CD45RO memory-like T cells and CD56+/CD3− natural killer (NK) cells and for IFN-γ production by these cells in maternal peripheral and placental intervillous blood in relation to the status of malaria infection in pregnancy. Of 52 human immunodeficiency virus-negative enrolled pregnant women residing in Western Kenya, 20 had placental parasitemia. We found that the percentages of CD45RO memory-like and CD4 T cells were significantly higher in the periphery than in the placenta, while the CD56/CD3− NK-cell percentage was higher in the placenta than in the periphery, suggesting differences in immune cell profiles between the two blood compartments. Furthermore, the percentages of peripheral CD45RO memory-like and CD4 T cells were significantly elevated in aparasitemic women compared to levels in the parasitemic group, with aparasitemic multigravid women having the highest percentages of CD45RO memory-like and CD4 T cells. In contrast, at the placental level, IFN-γ production by innate NK cells was significantly increased in aparasitemic women compared to parasitemic women, regardless of gravidity. These results suggest that the elevated IFN-γ-producing NK cells in the placenta and CD45RO memory-like and CD4 T cells in peripheral blood may be involved in protection against malaria infection in pregnancy.
In regions where malaria is endemic, nonpregnant women with antimalarial immunity become more susceptible to infection during gestation than before pregnancy (3). In particular, they become at risk for placental parasitization, the placenta being the preferential site for parasite accumulation. The resulting placental malaria (PM) infection has severe consequences for both mother and child. Depending on transmission intensity, the public health consequences of malaria in pregnancy are varied. In low transmission settings, malaria infection increases the risk of maternal illness and fetal loss. In higher transmission areas, pregnant women are often asymptomatically infected but suffer preterm labor and maternal anemia and produce low-birth-weight infants who are at subsequent increased risk of neonatal and postneonatal mortality (17). In regions of intense malaria transmission, adverse consequences of malaria during pregnancy affect predominantly primigravid and secundigravid women; multigravid women are relatively less susceptible. This is true only among human immunodeficiency virus (HIV)-negative women.
Previous immunological investigations in pregnant women from regions where malaria is holoendemic have revealed an increasing development of specific humoral and cellular immune responses over successive pregnancies (7, 19). The acquisition of antibodies against placenta-binding parasites increases with an increase in parity such that primigravid women with no or low levels of anti-placental parasite antibodies have a higher burden of placental infection, while multigravid women, who have much higher levels of parasite-binding antibodies, exhibit lower PM prevalence (7). We have previously reported intervillous blood mononuclear cell gamma interferon (IFN-γ) levels with similar parity-dependent protection (19). The highest levels of IFN-γ were detected in intervillous blood mononuclear cell culture supernatants from aparasitemic multigravid pregnant women residents of Western Kenya, a region where malaria is holoendemic. In contrast, the lowest levels of IFN-γ were observed in primigravid and parasitemic multigravid women, suggesting a role for IFN-γ in protection. Additionally, HIV-infected pregnant women from the same region had impaired IFN-γ responses and were more susceptible to malaria, further suggesting an important role for IFN-γ in protection against placental infection (18).
Several studies conducted in children in different settings where malaria is endemic have shown that IFN-γ inhibits malaria parasite development (4), significantly delays reinfection (16), and is associated with protection from parasitemia, clinical malaria, and anemia (5). IFN-γ is a potent regulator of macrophage function and elicits the production of reactive oxygen species and nitrogen intermediates (8, 14, 21). Expressed by T and NK cells, IFN-γ is involved in both innate and acquired immune responses, respectively. In the above-mentioned Kenyan study, elevated levels of placental IFN-γ in multigravid aparasitemic pregnant women were reported, suggesting a role of IFN-γ in protection from PM (19). However, the specific cellular source of the IFN-γ response was not investigated. Also, it remains unknown whether the immune effectors operating in the placenta are distinct from those in the periphery in view of the fact that the placental blood compartment has been described as being immunologically unique (20). In this study, we investigated the immune cell subpopulations involved in the placental IFN-γ response and the differences between the peripheral and placental cellular immune responses in association with malaria infection during pregnancy. By using standard cell surface marker staining methods for identification of peripheral and placental immune cell subpopulations, and using intracellular cytokine staining techniques, the profiles of cell subpopulations and IFN-γ expression by CD4, CD8, and CD45RO memory-like T cells and NK cells were examined in peripheral and placental blood. The results from this study address some of the unknowns as they relate to the mechanisms of cellular immune protection against malaria infections during pregnancy in peripheral blood and in the placenta in the same individual.
MATERIALS AND METHODS
Study participants and sample collection.Placentas were obtained from mothers delivering at the Nyanza Provincial Hospital in Western Kenya, an area where malaria is holoendemic and HIV is epidemic. Placentas from singleton, uncomplicated, and term vaginal deliveries were used for this study. Intervillous blood was immediately collected (within two minutes of expulsion) by a placental prick blood sampling method as previously described (23). A peripheral blood sample was also obtained from the mother. Both peripheral and placental blood were collected into heparin-charged tubes. The mothers’ HIV statuses were determined using two rapid serology tests: a primary Determine HIV1/2 test (Abbott Laboratories, Abbott Park, IL) and a confirmatory Unigold HIV1/2 test (Trinity Biotech, Bray, Ireland). Malaria infection was determined by Giemsa-stained peripheral and placental thick and thin blood smears. Parasites were counted per 300 white blood cells, assuming 8,000 white blood cells per microliter of blood. Altogether, 52 HIV seronegative women with no other underlying disease were enrolled, 20 with PM and 32 without PM. All study methods were reviewed and approved by the Kenya National Ethical Review Committee, Nairobi, Kenya, and the Institutional Review Board of the Centers for Disease Control and Prevention, Atlanta, GA. Formal written consent was obtained from each study participant. Presumptive intermittent treatment for malaria was in place at the Nyanza Provincial Hospital antenatal clinic during the study period, in accordance with the Kenya Ministry of Health guidelines (13).
Immunophenotyping and detection of IFN-γ-producing cells.Blood samples (peripheral and placental) were diluted 1:1 with incomplete RPMI medium and activated using phorbol 12-myristate 13-acetate (15 ng/ml), ionomycin (1 μg/ml), and interleukin 2 (15 IU/ml) with monensin (1 μM/ml) as the cytokine secretion inhibitor. Cells were incubated at 37°C and 5% CO2 for four hours. Unstimulated aliquots of peripheral and placental blood incubated with monensin were processed in parallel, to serve as controls. After incubation, cell surface marker staining was carried out for the identification of the following cell subpopulations: total T (CD3+), CD4 T (CD4+ CD3+), and CD8 T (CD8+ CD3+) cells, memory-like (CD3+ CD45RO+) T cells, and NK (CD56+ CD3−) cells. One hundred microliters of the stimulated and unstimulated whole-blood samples were stained with fluorochrome-labeled surface marker antibody cocktails (5 μl of each antibody) (Pharmingen, San Diego, CA) as shown in Table 1. Erythrocytes were eliminated by lysis using BD lysing solution (BD Pharmingen, San Diego, CA), and cells were frozen at −80°C to initiate the permeabilization of the cell walls in preparation for IFN-γ intracellular staining. Frozen cells were quickly thawed at 37°C, washed, and further permeabilized using 0.1% saponin. Cells were then stained with anti-IFN-γ fluorescein isothiocyanate (FITC) antibody (125 ng/ml) (Pharmingen, San Diego, CA) and were fixed with 1% paraformaldehyde. FITC-labeled MOPC21 and mouse immunoglobulin G1 (IgG1) (Pharmingen, San Diego, CA) were used as the isotype-matched negative controls for IFN-γ and surface marker expression, respectively.
Antibody staining panel for detection of IFN-γ-producing cellsa
Cell acquisition and analysis were performed using the Cell Quest Pro software (BD Biosciences, San Diego, CA) on a BD FACSCalibur. Ten thousand gated lymphocytes were acquired and analyzed using dot plots. The percentage distribution of each T-cell subpopulation was determined by SSC-CD3+ gating (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate). Isotype controls were used to establish background fluorescence. Results for CD3, CD4, CD8, and memory-like (CD45RO+) T cells were reported as percentages of cell numbers gated in the SSC-CD3+ with percent negative control values subtracted. NK-cell percentages were evaluated from CD56+ CD3− events after subtracting the percent negative control values. The percentage of IFN-γ-expressing cells was calculated from gated cell subpopulation events with the percent negative control values subtracted.
Statistical analysis.Summary statistics were calculated to describe the patient population. A Pearson chi-square test was used to evaluate associations between categorical variables. The two-sample t test was used to compare continuous variables between PM-negative (PM−) and PM-positive (PM+) mothers. The signed rank test was used to assess whether there were significant mean percentage differences between placental and peripheral blood for various immune cellular profiles. This test, completed for all samples and for PM− and PM+ mothers separately, tests the null hypothesis that the mean percentage difference is equal to zero. The Wilcoxon two-sample rank sum test was used to compare median percentages of cell subpopulations and IFN-γ-expressing cells between PM− and PM+ women for peripheral and placental samples separately. The Kruskal-Wallis test (Monte Carlo estimation) was used to compare median percentages when the PM− and PM+ groups were further stratified by gravidity on the basis of paucigravid (primigravid and secundigravid) versus multigravid. Due to the small to moderate sample size, a Monte Carlo estimation of an exact P value was used in both the rank sum and Kruskal-Wallis tests. The data were analyzed using the SAS (SAS version 9.1; SAS Institute, Cary, NC) software package.
RESULTS
Characteristics of study participants.Table 2 shows the characteristics of the women recruited into the study. Of 52 enrolled women, 20 (38%) had Plasmodium falciparum infection of the placenta (PM+) while 32 (62%) were without placenta malaria (PM−). Among the 20 PM+ women, 17 had peripheral parasitemia. Of the 32 PM− women, 31 did not have peripheral parasitemia. Among the 20 PM+ mothers, 16 (80%) were paucigravid and 4 (20%) were multigravid; among the PM− mothers, 18 (56.3%) were paucigravid and 14 (43.8%) were multigravid (χ2 test of association; P = 0.08). The mean ages (± standard deviations [SD]) for PM− and PM+ mothers were 24 ± 7 and 21 ± 5 years, respectively (two-sample t test; P = 0.12). In this study, mean birth weights (± SD) for the babies born to PM− and PM+ mothers were similar: 3.3 ± 0.49 kg and 3.3 ± 0.47 kg, respectively (P = 0.83). Maternal hemoglobin levels between PM− and PM+ groups also were not significantly different, 12.1 ± 2.0 g/dl and 11.5 ± 1.6 g/dl, respectively (P = 0.27). Among the PM+ mothers, the median (25th percentile, 75th percentile) for peripheral parasitemia was 783 (176, 4,118) parasites/μl and for placental parasitemia was 2,160 (338, 8,451) parasites/μl (rank sum; P = 0.15).
Characteristics of study participants
Difference of cell subpopulations and IFN-γ-producing cells between peripheral and placental blood.Overall comparisons of cell subpopulation percentages revealed that CD4/CD3 and CD45RO/CD3 were higher in peripheral blood than in placental blood (P of 0.004 and 0.0002, respectively), while CD56/CD3− was higher in the placental blood than in the periphery (P = 0.004). All other cell subpopulation percentages were not significantly different for the two blood compartments. IFN-γ production by cell subpopulations did not vary between peripheral and placental blood (Table 3). When further stratified by the presence or absence of parasites, as in the overall comparison, PM− mothers had significantly higher percentages of CD4/CD3 (P = 0.005) and CD45RO/CD3 (P < 0.0001) cells in the periphery than in the placenta. In PM− mothers, CD56/CD3− cells were also higher in the placenta than in the periphery (P = 0.0025) (Table 3). In contrast, no significant differences in the cellular profiles between periphery and placenta were seen in the PM+ mothers. IFN-γ expression did not differ between peripheral and placental cell subpopulations for both PM− and PM+ mothers (Table 3).
Immune cellular profiles: differences between placental and peripheral blood
Association between cell subpopulations/IFN-γ expression and status of malaria infection.In order to determine which cell subpopulations and/or IFN-γ-expressing cell subpopulations may contribute to protection against malaria during pregnancy, comparisons between PM− and PM+ mothers were made for peripheral and placental immune responses independently. Considering peripheral immune responses alone, PM− mothers had significantly higher percentages of CD4/CD3 and CD45RO/CD3 than did the PM+ mothers (P = 0.001 and P = 0.002, respectively) (Table 4). At the placental level, comparisons of the cellular immunological parameters between PM− and PM+ revealed percentages of CD4/CD3 (P = 0.03) and CD8/CD3 cells (P = 0.04) (Table 4) that were higher in the PM− mothers than in PM+ women. Although the percentage of CD56/CD3− cells was not statistically significantly different between PM− and PM+ mothers (P = 0.09), IFN-γ production by CD56/CD3− NK cells in placental blood was significantly higher in PM− individuals than in PM+ mothers (P = 0.01) (Table 4).
Comparison of cell subpopulations and IFN-γ-producing cells in the periphery and the placenta in PM− and PM+ women
To further investigate whether or not the above five potential protective immunological parameters (CD4/CD3 and CD45RO/CD3 cells in peripheral and CD4/CD3, CD8/CD3, and IFN-γ-expressing CD56/CD3− NK cells in placental compartments) were affected by gravidity, mothers were divided into four groups based on gravidity and presence of placental parasites. Paucigravidae were considered as one group because previous work on cellular immune responses in this region of Kenya has shown that immunologic characteristics for primi- and secundigravidae are similar (19). In peripheral blood there were differences for CD4/CD3 cells (P = 0.001) and CD45RO/CD3 cells (P = 0.004) when stratified by both gravidity and parasitemia, with the highest percentages of the cells in PM− multigravid women (Fig. 1). Further pair-wise comparisons showed that there were significant differences in both CD4/CD3 (P < 0.001) and CD45RO/CD3 (P = 0.005) cells between PM− multigravid women and PM+ paucigravid women. At the placental level, IFN-γ production by CD56/CD3− cells that was significantly different by parasitemia (Table 4) showed a borderline difference when stratified by both gravidity and parasitemia; there was a trend toward a higher percentage in the PM− than in the PM+ women for both gravidity groups (P = 0.06) (Fig. 2). The percentages of CD4/CD3 and CD8/CD3 cells were not different in the placenta among the four groups (Fig. 2).
Assessment of potential protective immunological parameters in the periphery stratified by gravidity and placental parasitemia. A Kruskal-Wallis test was used to compare percentages of a median of two potential immune cellular parameters, CD4/CD3 and CD45RO/CD3 cells in among four groups of mothers stratified by both PM status and gravidity (PM− PAUCI, PM− paucigravidae [n = 18]; PM− MULTI, PM− multigravidae [n = 14]; PM+ PAUCI, PM+ paucigravidae [n = 16]; and PM+ MULTI, PM+ multigravidae [n = 4]). A P of <0.05 indicates a significant difference in at least one group. Graphs shown are box-and-whisker plots, which provide side-by-side comparisons of the median (center horizontal line), the mean (·), the 25th percentile (bottom of box), the 75th percentile (top of box), and the minimum and maximum (ends of whiskers).
Assessment of potential protective immunological parameters in placenta stratified by gravidity and placenta parasitemia. See Fig. 1 legend for details. If values are less than 25th percentile − 1.5 × IQR or greater than 75th percentile + 1.5 × IQR, where IQR is the interquartile range, the whiskers are clipped to these boundaries, with outliers represented by a small box. A Kruskal-Wallis test was used to compare percentages of three potential immune cellular parameters, CD4/CD3, CD8/CD3, and IFN-γ-expressing CD56/CD3− NK cells, among four groups of mothers.
DISCUSSION
Malaria infection in pregnancy involves two blood compartments: peripheral and placental blood. Peripheral blood feeds into the placenta; thus, both blood compartments would presumably display similar immune responses. However, the results from this study reveal clear differences both in terms of immune cell profiles between peripheral and placental blood and in their immunological association with the status of malaria infection in the two blood compartments.
We found that CD45RO memory-like and CD4 T-cell percentages were higher in the peripheral blood than in the placenta, while NK cell percentages were higher in placental blood than in the periphery in PM− but not in PM+ mothers. Our previous study conducted in the same area showed similar differences in the immunological composition of intervillous blood versus peripheral blood in PM− women (20). Taken together, results from the two studies suggest that the placental and peripheral blood compartments, while continuous, are immunologically distinct. Differences between peripheral blood and the placenta, particularly as observed in PM− mothers, may be due to two possible factors. First, the placenta in each pregnancy is a new organ with the immunological commitment of maintaining immune responses necessary for normal pregnancy progression. Immune cell profiles are therefore predominantly innate in origin, such as NK cells, providing a first line of defense against infections and pathogenic alterations in the normal homeostasis of pregnancy (10). Peripheral immune responses, on the other hand, are developed over a period of time to pathogens encountered during the course of prepregnancy infections and result in the generation of memory, in which CD45RO memory-like T cells and CD4 T cells are critical players in the regulation of immune effectors. Second, previous parasite phenotypic studies conducted by others have revealed a distinct placenta-infecting parasite that successfully establishes an infection only in the presence of a placenta. These studies also suggested specialized protective mechanisms required for placental parasite clearance distinct from those required for peripheral immunity (6). Our data suggest that PM- mothers living in areas where malaria is endemic may naturally develop different immune responses at the peripheral and placental levels. This finding further validates reports that suggest that immunity to malaria in nonpregnant women and pregnancy-associated malaria develop independently of each other (12). If our supposition is true that differences in immune responses between the placental and peripheral blood are required for protection from malaria infection in pregnancy, then PM+ women who did not display distinct cellular profiles between peripheral and placental blood may have duly been rendered susceptible to infection.
The peripheral CD45RO memory-like T-cell percentages were overall increased in PM− mothers, compared to percentages in PM+ women; particularly, the PM− multigravid women had the highest percentage of CD45RO memory-like T cells. Elevated peripheral CD45RO memory-like T-cell percentages in multigravid PM− women may be associated with protection from peripheral infection and may be part of a recall immune response that would enhance antiparasite immunity. It remains unclear why memory-like T-cell responses in peripheral blood, but not in the placenta, are associated with absence of malaria infection during pregnancy; it also remains unclear which underlying regulatory immune mechanisms are involved in this phenomenon. CD4+ CD25+ regulatory T cells play a central role in the regulatory network that benefits both pathogen and host by allowing infection to persist for the benefit of long-term memory production and subsequent resistance to reinfection while minimizing immune cell-mediated pathology (2). A recent study has shown that P. falciparum growth in adults is correlated with an up-regulation of CD4+ CD25+ T cells (28). Investigations of CD4+ CD25+ T-regulatory cells on immune effectors in pregnant women and their association with PM infection are ongoing in our laboratories.
The percentage of peripheral CD4+ T cells was also highest in PM− multigravid women compared to the percentage in paucigravid PM+ women. We propose that the decrease in peripheral CD4 T cells in PM+ women may cause a subsequent decrease in antiparasite T-helper-cell-dependent humoral immune responses in peripheral blood. Antiparasite antibodies are critical for immunity against the specialized placental parasite (7). A reduction in parasite-specific antibodies mediated by the decrease in CD4 T cells could result in parasite persistence and the establishment of a placental infection. Indeed, several studies have shown that lower levels of placental parasite-specific antibody responses in peripheral blood are associated with susceptibility to PM infection (7).
An assessment of placental immune responses between PM− and PM+ mothers revealed significantly higher percentages of IFN-γ-expressing NK cells (but not of NK cells) in PM− mothers than in PM+ women, suggesting increased NK-cell activation as opposed to NK-cell expansion. IFN-γ is a parasiticidal cytokine, and early innate IFN-γ production by NK cells is critical to the control of infection before parasite-specific immune responses are elicited (1). The elevated placental IFN-γ expression by NK cells, but not NK-cell expansion, in PM− women regardless of gravidity may be due to an innate ability to instantaneously produce IFN-γ, resulting in parasite clearance, and it would presumably then be the first line of defense against infection in the placenta. Furthermore, in vitro investigations suggest that an early IFN-γ response to P. falciparum-infected red blood cells is under the control of host genetic determinants, such as the NK-cell killer Ig-like receptor (KIR) genotypes (1, 25). It is thus conceivable that PM+ women may be genetically challenged in that their KIR genotype may inhibit the ability to rapidly express protective IFN-γ from NK cells and they are therefore susceptible to infection. This hypothesis requires further investigations. Contrary to the IFN-γ/NK results discussed here, Sartelet et al. (26) report significantly higher levels of NK cells in infected placentas than in noninfected placentas and they associate the elevated NK-cell numbers to poor birth outcomes. Sartelet's study investigated placental tissue by immunohistochemistry in a focused group of placentas obtained from women with poor birth outcomes. It is important to note here that in our current study there were no reports of adverse birth outcomes in relation to placental infection: birth weight was similar for PM− and PM+ groups. In addition, a study in Tanzania also reported results contrary to ours; they associated the presence of placental parasitemia with a total absence of NK cells in placental blood (22). The basis for this significant disparity may be attributed to differences in techniques used. Whereas the current study used highly sensitive immunophenotyping by flow cytometry to identify NK cells, the Tanzanian study used a relatively less-sensitive technique, immunohistochemical analysis of formalin-fixed paraffin-embedded tissue, which may have failed to detect NK cells.
Increased percentages of placental CD4+ T cells in PM− women were observed compared to percentages in PM+ mothers. Studies in other infectious diseases, such as influenza A, have reported that IFN-γ production by NK cells is regulated by the helper function of T cells for the resolution of infection, suggesting a close relationship between innate and acquired immune responses (11). We speculate that IFN-γ-producing, NK-cell-mediated innate immunity against PM infections may also be regulated by CD4 T cells at the placental level.
We also detected that PM− women had a higher percentage of placental CD8+ T cells than the PM+ group (P = 0.04), despite no differences in IFN-γ-producing CD8+ cells between the two groups. CD8+ T cells have generally been viewed as playing little or no role in immunity against blood-stage malaria infections, since mature erythrocytes express few class I major histocompatibility complex (MHC) molecules (27). However, several earlier studies conducted in mouse models have shown that CD8+ T cells contribute to acquired immunity to blood-stage infection (24). Additionally, human reticulocytes express a significant number of class I MHC molecules (9); during pregnancy reticulocyte numbers increase and continue to do so until parturition (15). Since P. falciparum parasites infect red blood cells of all maturities, including reticulocytes, we speculate that CD8+ T cells may exert an antiplasmodial effect on reticulocytes by a non-IFN-γ pathway at the placenta level to protect against infection. Of particular note is that some data also suggest that the NK-cell KIR receptors mentioned earlier recognize MHC class I molecules (29); therefore, the expression of MHC class I molecules on reticulocytes during pregnancy may also be critical in NK-cell-mediated destruction of malaria-infected reticulocytes.
Our study was limited in that mitogen-induced but not antigen-specific IFN-γ production by T-cell subpopulations (except for NK cells) was measured. Thus, this underestimates the role of malaria antigen-specific IFN-γ-producing T cells in protection against malaria during pregnancy. Due to operational constraints, our assessment of cell quantity allowed us to report only the percentages of cells but not the absolute cell number, thus decreasing our power to further dissect observed differences of cellular immune parameters between peripheral and placental blood compartments. Additionally, during our study period, which spanned over two years, only four PM+ HIV-negative multigravid women were recruited. This was attributable to the lower risk of malaria infection among multigravid women and higher prevalence of HIV infection in pregnant women in our study area (where malaria is holoendemic and HIV is epidemic), resulting in the extremely low number of HIV-negative multigravid women presenting at the hospital with malaria. Nevertheless, the lower number of PM+ women in the multigravid group did not affect the observed associations between the immunological parameters and the status of malaria infection in this study.
In conclusion, we found that the percentages of CD45RO memory-like and CD4 T cells were significantly higher in the periphery than in the placenta, while the CD56/CD3− NK-cell percentage was higher in the placenta than in the periphery, suggesting differences in immune cell profiles between the two blood compartments. Our study further showed that elevated percentages of peripheral CD45RO memory-like T cells and CD4 T cells were associated with absence of malaria infection in a gravidity-dependent manner, while increased placental innate IFN-γ production by NK cells was associated with absence of malaria infection regardless of gravidity, suggesting the possible involvement of the different cellular immune components in the peripheral and the placental blood compartments in relation to the control of malaria infections during pregnancy. The observed differences reported in this study between peripheral and placental blood immune cell profiles, and their immunological association with the status of malaria infection, are intriguing. Further investigations of innate and acquired immune regulatory mechanisms in both the peripheral and placental blood compartments may offer a better understanding of the susceptibility to malaria infection during pregnancy.
ACKNOWLEDGMENTS
This study was supported by TDR/WHO training grant number 980427 and by grant numbers AOT0483-PH1-2171 and HRN-A-00-04-00010-02 from the United States Agency for International Development. Julie M. Moore is supported by the NIH grant no. RO1 HD046860.
We are grateful to the mothers who participated in this study. We also thank the field and laboratory staffs of the CDC/Kenya Medical Research Institute (KEMRI) who facilitated collection of data and processing of samples. We thank Davy Koech, director of KEMRI, for his support and his approval with regard to publication of this paper.
FOOTNOTES
- Received 23 October 2007.
- Returned for modification 26 November 2007.
- Accepted 25 January 2008.
- Copyright © 2008 American Society for Microbiology
