Effects of anti-malarial prophylaxes on maternal transfer of Immunoglobulin-G (IgG) and association to immunity against Plasmodium falciparum infections among children in a Ugandan birth cohort

Background The in-utero transfer of malaria specific IgG to the fetus in Plasmodium falciparum infected pregnant women potentially plays a role in provision of immune protection against malaria in the first birth year. However, the effect of Intermittent Prophylactic Treatment in Pregnancy (IPTp) and placental malaria on the extent of in-utero antibody transfer in malaria endemic regions like Uganda remain unknown. The aim of this study was thus to establish the effect of IPTp on in-utero transfer of malaria specific IgG to the fetus and the associated immune protection against malaria in the first birth year of children born to mothers who had P. falciparum infection during pregnancy in Uganda. Methods We screened a total of 637 cord blood samples from a double blinded randomized clinical trial on Sulfadoxine-Pyrimethamine (SP) and Dihydroartemisinin-Piperaquine (DP) IPTp in a Ugandan birth cohort; study conducted from Busia, Eastern Uganda. Luminex assay was used to measure the cord levels of IgG sub-types (IgG1, IgG2, IgG3 and IgG4) against 15 different P. falciparum specific antigens, with tetanus toxoid (t.t) as a control antigen. Man- Whitney U test (non-parametric) in STATA (ver15) was used in statistical analysis of the samples. In addition, Multivariate cox regression analysis was used to determine the effect of maternal transfer of IgG on the incidence of malaria in the first birth year of children under study. Results Mothers on SP expressed higher levels of cord IgG4 against erythrocyte binding antigens (EBA140, EBA175 and EBA181) (p<0.05). Placental malaria did not affect cord levels of IgG sub-types against selected P. falciparum specific antigens (p>0.05). Children who expressed higher levels (75th percentile) of total IgG against the six key P. falciparum antigens (Pf SEA, Rh4.2, AMA1, GLURP, Etramp5Ag1 and EBA 175) had higher risk of malaria in the first birth year; AHRs: 1.092, 95% CI: 1.02-1.17 (Rh4.2); 1.32, 95% CI: 1.00-1.74 (PfSEA); 1.21, 95%CI: 0.97-1.52 (Etramp5Ag1); 1.25, 95%CI: 0.98-1.60 (AMA1); 1.83, 95%CI: 1.15-2.93 (GLURP) (GLURP), and 1.35,; 95%CI: 1.03-1.78 (EBA175). Children born to mothers categorized as poorest had the highest risk of malaria infections in the first birth year (AHR: 1.79, 95% CI: 1.31-2.4). Children born to mothers who had malaria infections during gestation had higher risk of getting malaria in the first birth year (AHR 1.30; 95% CI: 0.97-1.7). Conclusion Malaria prophylaxis in pregnant mothers using either DP or SP does not affect expression of antibodies against P. falciparum specific antigens in the cord blood. Poverty and malaria infections during pregnancy are key risk factors of malaria infections in the first birth year of growth of children. Antibodies against P. falciparum specific antigens does not protect against parasitemia and malaria infections in the first birth year of children born in malaria endemic areas. Blood samples for the Neonates were collected from Masafu General Hospital (the main Referral health facility within Busia District). Pregnant mothers in the first trimester were enrolled from across 40 (forty) villages located within 30 Kms radius away from Masafu Hospital. Participants were drawn from the Sub-counties of Buteba, Busitema, Masaba, Lunyo, Bulumbi Masafu and Lumino. Demographic information and clinical data of study participants were captured(Mother and baby pair) (Table 2). The samples for this study were obtained from a previous study done among pregnant mothers. Cord blood samples were screened to assess the levels of malaria specific antibodies. Details of participants’ demographics, socio-economic status and clinical information is contained in a published study. The current study obtained samples archived from the primary study, demographic and clinical details of participants can be found in a previous study [18]. Collection of the cord blood by the primary study took place at the time of delivery. Blood collection was done by a midwife, medical officer or obstetrician who conducted delivery or cesarian section. The blood was then centrifuged to obtain serum, archived at appropriate temperature until analysis. Archival was done and random-access numbers generated by the computer to help in tracking the samples in the biobank. Buffer A was prepared following a method previously reported in a study by I.ssewanyana et al. Briefly, 1L of phosphate buffer solution was measured and put in a conical flask after which 500μl of tween (Sigma-Aldrich; USA) was dispensed into the PBS to form PBS-tween solution. 5g of PVP powder (Sigma-Aldrich; USA) was weighed using a weighing balance in aluminum foil and added to the above solution after which equal amount of PVA powder was weighed and added to the mixture; proper agitation and mixing was done using a vortexer. 5mls of BSA (Thermofisher Scientific;USA) was pipetted and added to the solution; finally, 0.2g of Sodium azide (Sigma-Aldrich; USA) was weighed under a biosafety hood in an aluminum foil, added to the solution, properly vortexed and mixed. The final solution was labelled with day of preparation and names of those who prepared. A method by [19] was followed in the preparation of buffer B. Briefly, 1000mls of phosphate buffer solution (PBS) was transferred to a sterile glass tube. To this was added, 500ul of tween solution, 5g of polyvinyl alcohol (PVA) (Sigma-Aldrich; USA), 5mls of bovine serum albumin solution (BSA), 0.2g Sodium azide and 3mls of lyophilized E. coli solution (Avanti Polar lipids 100600C, Sigma-Aldrich-USA)). The mixture was incubated overnight at 4°C. 693μls of buffer B was transferred into deep well plates using multi-channel pipette. 7μl plasma was then dispensed to the same well. The procedure was repeated for all plasma samples and mixed by shaking. The final sample dilution used was 1 in 100. Diluted samples were then incubated in buffer B over night at 4°C to allow E. coli extract mop out anti-E coli antibodies to minimize any background response that may be due to E. coli protein contamination of the antigens. The P. falciparum antigens were expressed in E. coli. A total of 16 recombinant P. falciparum pre-erythrocytic, erythrocytic and infected RBCs antigens, Tetanus toxoid (t.t-non malaria control) (Microcoat GmbH, Germany) coupled on Luminex beads were used in this experiment. Antigens were batched in three groups; (i) infected red blood cells associated antigens, (ii) merozoite apical complex expressed proteins, and (iii) merozoite surface antigens (Table 1). A standard protocol was used to make the dilution [19]. Briefly, 5mls of buffer A was transferred to a falcon tube. To this, 8μl of each bead region was added and mixed by shaking. This was repeated for all the 16 coupled beads antigenic regions. IgG sub-types 1–4 against 15 P.falciparum blood specific antigens and Tetanus Toxoid (TT) was measured in plasma diluted in buffer B. 50μl of bead suspension was added to each well (1,000 beads/region/well) of the 96 plate (Bio-plex pro-flat bottom). The plate was washed, placed on a magnetic block for 2 minutes and the supernatant was then poured off. The beads were washed twice with PBS tween buffer, laid on the magnetic separator for 2 minutes and supernatant poured off. Using a multichannel pipette, 50μl of prepared plasma at a dilution of 1in1000 in buffer B and pooled hyper immune control (27 adults from malaria endemic areas) was added to the beads; two wells had PBS added in them as blanks. The plates were then put on a shaker at 600 rpm for 90 minutes after covering the plates with aluminum foils at room temperature. After incubation, the plates were placed on a magnetic separator for 2 minutes, supernatant poured off by a rapid inversion with a sharp shake followed by a gentle blot on a paper towel. The plates were then washed three times using 200ul wash buffer. 50μl of a secondary antibody specific for an IgG sub-type (IgG1, IgG2, IgG3 and IgG4) with concentration of 1in1000 for IgG1,IgG2 and IgG4 while secondary antibody against IgG3 was diluted at 1in 2000 in dilution buffer and added to each well containing bound primary antibody-antigen complex, blank wells, positive control wells and negative control wells. Incubation was done for 60 minutes in a shaker. The plate was put on a magnetic separator for 2 minutes, supernatant poured off and washed three times using wash buffer to remove unbound antibodies. Using a multichannel pipette, 50μl of 1in 200 R-Phycoerythrin-conjugate AffinPure F (ab’) goat anti-human IgG (abcam,Boston,MA,USA) diluted in buffer A was added. The plate was then incubated at room temperature in a shaker at 600 rpm for 60 minutes. After the incubation, the plate was then placed on a magnetic separator for 2 minutes and supernatant poured off, washed three times using wash buffer. 100μl of plain PBS was then added to the plates. Incubation was done at room temperature by putting it on a shaker at 600rpm for 30 minutes. The plates were read on a MagPix platform (Luminex Corp, USA), acquiring at least 50 beads/region/well. The results were expressed as median fluorescence intensity (MFI). The blank well MFI (background effects) was deducted from each well to determine the net MFI (IgG positive result) Luminex assay was done following modified manufacturer’s guidelines. Positive control plasma samples were obtained from 27 adults’ resident in malaria areas with known episodes of malaria and thought to be hyper immune. Negative controls samples were serum samples obtained from whole blood of 6 Caucasians from the United Kingdom with no known history of exposure to P. falciparum. Dilution curves were developed using positive control serum samples. For each plate run, negative and positive controls would be included. To ensure quality of results, weekly calibration of MagPix machine was done, sample probe blown and sonicated every morning prior to use, routine stringent cleaning using bleach and sodium hydroxide done to avoid clogging of the probe. A plate water run was done at the end of the day’s sample run and a plate would be re-prepared if more than 10 wells had the bead counts less than 50. Other quality assurance measures included preparing fresh reagents after every three days or if there was evidence of contamination All Net fluorescence Index (Net MFI) were log transformed prior to analysis. Part of the net MFI was also normalized to minimize effects of extreme values on mean. Graphical representation using box plot was used to compare mean levels of cord IgG sub types against P. falciparum signature antigens of infants whose mothers were on DP versus those on SP. Man-Whitney U test in STATA (ver 15); a non -parametric test was used to compare means of IgG1, IgG2, IgG3 and IgG4 levels against signature P.falciparum antigens for DP group compared to SP group. A statistically significant difference in the means of the two groups was confirmed if the P-value was less or = 0.05. To evaluate effects of placental malaria on cord expression of IgG sub-types, Man Whitney U-test and box plots were used to compare the difference in the mean of cord IgG sub-class levels of infants whose mothers had placental malaria versus mothers who never had. Only IgG specific to one antigen at a time was used in building the Cox regression model while adjusting for non-IgG covariates at each time. The Cox regression model was built using back ward elimination method. For each IgG, the levels were grouped into percentiles, that is 25th, 50th and 75th. The effects of different levels of the IgG (25th, 50th and 75th percentiles) on the incidence of malaria among study participants were assessed. The levels of the IgG at the 75th percentile was found to consistently have higher Hazard ratios and significant confidence intervals. Therefore, the levels of the IgG at the 75th percentile were used in building the Cox regression model. The P-values of the IgG variables in the Cox regression models were adjusted for Multiple testing using the Benjamin-Hoceberg correction of multiple testing in excel. The non-IgG covariates included were birth weight (<2500 grams, >2500grams), maternal malaria incidence (No malaria detected, and malaria detected), Maternal wealth index (poorest, middle, and least poor), gravidity category (1–3 and 4+). Cofounding was assessed in the Cox regression model and any variable that had the percentage difference of more than 10% between the crude and adjusted Hazard ratios was considered a cofounder. In the model building, the covariates that had p-value of ≤ 0.2 in the bivariate analysis were included in the model. The primary study was approved by the Research and Ethics Committee, School of Biomedical Sciences, Makerere University College of Health Sciences; Uganda National Council for Science and Technology (SBS 114; S1 Appendix). Informed consent was obtained after explaining possible risks and benefits of the research to participants. They were requested to sign after understanding the contents before maternal and cord blood draws. The study sought for permission to archive participants’ samples for future use in research; those who declined to this had their samples discarded while keeping for those who consented. This current study was also approved by the Makerere University School of Biomedical Sciences Research and Ethics Committee (SBS 1012; S2 Appendix)