Effect of Plasmodium Infection during Pregnancy on Passive Neonatal Immunity against Tetanus Toxoid and Rotavirus

ABSTRACT.

Passive immunity acquired through transplacental IgG transport is essential to protect infants against pathogens as childhood vaccination programs begins. Diarrhea caused by rotavirus and neonatal tetanus are common and potentially fatal childhood infections that can be prevented by transplacental IgG. However, it is not known whether maternal infections in pregnancy can reduce the transfer of these antibodies to the fetus. This study evaluated the effect of submicroscopic Plasmodium infection during pregnancy on the transfer of maternal IgG antibodies against rotavirus (anti-RV) and tetanus toxoid (anti-TT) to newborns of pregnant women residing in Puerto Libertador and Tierralta, Colombia. Expression of different immune mediators and levels of IgG against rotavirus and tetanus toxoid were quantified in pregnant women with and without Plasmodium infection during pregnancy. Submicroscopic infection at the time of delivery was associated with a cord-to-maternal ratio (CMR) > 1 for anti-RV and < 1 for anti-TT IgG, as well as with an increase in the expression of immune mediators of inflammation (IFN-γ), anti-inflammation (IL-10, TGF-β), and regulation (FoxP3, CTLA-4). When compared by species, these findings (CMR > 1 for anti-RV and < 1 for anti-TT IgG) were conserved in submicroscopic Plasmodium vivax infections at delivery. The impact of Plasmodium infections on neonatal susceptibility to other infections warrants further exploration.

INTRODUCTION

Neonates are exposed to a large number of pathogenic organisms and have an immune system without experience and with a tolerogenic profile that makes them vulnerable to infections.^1,2 Therefore, passive immunity acquired through transplacental IgG transport is essential to provide immunity against infections in early life.³

Vaccination during pregnancy is one strategy that aims to protect both pregnant women and their infants by increasing the concentration of specific maternal IgG that are passed to the fetus by active transport through the placenta,⁴ as well as the transport of IgG against different antigens to which the mother had been exposed to naturally. The tetanus toxoid vaccine is administered during pregnancy in the last trimester, when maternal transplacental transport of antibodies increases. The vaccine consists of tetanus toxoid, a protein antigen that induces IgG1 antibodies, the subclass of IgG transported most efficiently through the placenta.⁵ The increase of antibodies by vaccination improves the titer of antibodies transferred to the fetus, reaching protective levels in the mother and the newborn at the end of pregnancy.⁶

This mechanism allows infants to have protection against pathogens that cause serious infections as the childhood vaccination program begins. For example, diarrheal disease is the second leading cause of death in children worldwide, and rotavirus is the most common cause of severe diarrhea in infants and young children.⁷ More than 90% of rotavirus deaths occur in low–to middle-income countries⁸; in Latin America, rotavirus diarrhea alone caused more than 70 thousand hospitalizations per year and 15,000 deaths between 1990 and 2009.⁹ In Colombia, vaccination against rotavirus is provided at age 2 months. This means there is a period of susceptibility to rotavirus infection in neonates unless infants are born with protective antibodies from their mother. However, studies indicate that higher levels of transplacental rotavirus-specific IgG from naturally exposed mothers contribute to reduced vaccine seroconversion in their infants.^10,11

Several maternal factors can reduce transplacental passive immunity. Gestational age determines the levels of antibodies that pass through the placenta. During the first trimester of pregnancy, the transfer of antibodies is minimal,¹² but as of the second trimester, the percentage of antibodies that are transferred increases over gestational weeks. By full term, fetal IgG levels usually exceed maternal ones by 20% to 30%.^12–14 Other maternal factors associated with reduced passive transplacental immunity include hypergammaglobulinemia,^15,16 malnutrition,¹⁷ diabetes mellitus,¹⁸ and chronic infections such as HIV and microscopic placental malaria due to Plasmodium falciparum.^{15,16,19–23}

Different studies have explored the effect of microscopic placental malaria by P. falciparum on the transplacental transfer of maternal antibodies against other pathogens. Although some results are discordant, an association was reported between microscopic placental malaria and a reduction of 81%²² and 72%¹⁵ in the passage of maternal antibodies against the measles virus, and reduction by 69%, 58%, and 55% in the passage of maternal antibodies against the herpes simplex virus 1, respiratory syncytial virus, and varicella-zoster virus, respectively.¹⁶ Similarly, there was an 82% decrease in the passage of antibodies against Streptococcus pneumoniae polysaccharides²² and a decrease in the passage of maternal antibodies against tetanus toxoid.^20,21 In contrast, other studies did not report a decrease in maternal antibodies transferred via the umbilical cord against measles,¹² respiratory syncytial virus,¹⁹ and tetanus toxoid in cases of placental malaria.^16,22

Submicroscopic P. falciparum infections are also common in pregnant women,²⁴ but their effects on the transfer of maternal antibodies are not known. Likewise, the effects of Plasmodium vivax infection have not been studied. In the northwestern region of Colombia, both P. falciparum and P. vivax are endemic, and a high frequency of pregnancy-associated submicroscopic plasmodial infections have been reported in peripheral blood during the course of pregnancy (between 23% and 49%) and in placental blood (between 5% and 57%).^3,4 These submicroscopic infections are not diagnosed by routine methods and they are usually asymptomatic; as a result, they are not treated.²⁵

The mechanisms by which placental malaria infection can affect passive neonatal immunity have not been studied. Here we evaluated the influence of malaria in pregnancy on the transfer of maternal IgG antibodies against rotavirus and tetanus toxoid to newborns of Colombian pregnant women. We hypothesized that alterations in the placental transcriptional profiles of pro–and anti-inflammatory immune mediators already described during plasmodial infections^26,27 can alter the molecular balance at the maternal–fetal interface and alter processes such as neonatal passive immunity.

METHODS

Study site.

Women were enrolled between September 2013 and May 2016 (32 months) as part of a larger study in the municipalities of Puerto Libertador (07°53′35′ N, 75°40″16″ W) and Tierralta (8°10′22″N 76°03′34″O), Department of Córdoba, in the Uraba-Sinu-San Jorge-Bajo Cauca region of Colombia. This region has an estimated area of 43,506 km², and a population of 2.5 million at risk of malaria.²⁸ The region has stable malaria transmission intensity and is homogeneous in terms of eco-epidemiology and malaria transmission. The region has a high malarial incidence with a mean annual parasite index of 35.8 cases/1,000 inhabitants. P. vivax and P. falciparum coexist, but P. vivax prevails (60–70% of the total).²⁹

Study design and sample selection.

The study was part of a larger project aimed at exploring the epidemiology, clinical aspects, and immunopathology of malaria in pregnancy and placental malaria in northwest Colombia, from which partial results have been reported elsewhere.^30,31 A total of 610 pregnant women were recruited sequentially in the main study. On the basis of the availability and quality of the material collected from maternal peripheral blood, umbilical cord blood, and placental tissue, a subset of women was selected for this study to explore the influence of malaria in pregnancy on the transfer of maternal IgG antibodies against tetanus toxoid and rotavirus to the neonate. A sample of 125 women was selected in this study: 50 with confirmed submicroscopic plasmodial infection at delivery (SPID group); 38 with confirmed history of plasmodial infection (submicroscopic or microscopic) at different times in pregnancy, but negative at delivery (plasmodial infection in pregnancy [PIP] group); and 25 women without microscopic or submicroscopic plasmodial infection during pregnancy nor at delivery (no-PI group) (Figure 1). The status of infection was determined at monthly antenatal visits. Plasmodial infection was diagnosed by thick blood smear (TBS) and quantitative real-time polymerase chain reaction (qPCR) in peripheral blood collected at each antenatal visit and in peripheral and placental blood at delivery. Submicroscopic infection was defined as a positive result by qPCR and a negative result by TBS. Women in the no-PI group had negative results with both tests (qPCR and TBS) in peripheral blood collected during pregnancy and at delivery, as well as in placental blood.

Figure 1.

Sample collection and selection diagram.

Inclusion and exclusion criteria.

Inclusion criteria for the main study were voluntary acceptance and informed consent, permanent residency (> 1 year) in the malaria-endemic region, no history of pre-eclampsia, and negative HIV and TORCH tests. The exclusion criterion was withdrawal of consent.

Data and specimen collection.

After enrollment, a survey of maternal information was completed, which recorded data including age, number of pregnancies (parity), gestational age, and hemoglobin level. Newborn information only included the birthweight.

During pregnancy, peripheral blood samples were taken at each antenatal visit. Placental and cord blood were taken immediately after delivery and maternal peripheral blood was obtained within 24 hours. All peripheral blood samples were collected in separate tubes without anticoagulant or with EDTA to obtain serum, plasma, packed red blood cells (RBC), and buffy coat. The cord blood samples were collected in separate tubes with EDTA to obtain plasma. TBS and blood spots on Whatman #3 filter paper were prepared from each sample. Tubes were centrifuged and cryovials with serum, packed RBC, and buffy coat homogenized with Trizol (Invitrogen, Carlsbad, CA) (1:4) were stored in liquid nitrogen or at –20°C and –80°C until processing.

Placentas were processed within 8 hours of delivery as follows: they were cleaned with saline, after which a small (∼1 cm³) section of placenta was removed from the maternal side and the pooled blood was collected by pipet aspiration. Placental bloods were used to prepare TBS and dried blood spots (∼100 µL of blood on Whatman #3 filter paper) to diagnose malaria by microscopy and qPCR, respectively. For the histopathological study, two tissue fragments were obtained by sectioning the placenta (∼2 cm² surface area through the entire thickness), which were fixed in 10% neutral buffered formalin and paraffin-embedded within 48 hours. One fragment corresponded to an area immediately next to the umbilical cord insertion (central fragment), and the other fragment corresponded to an area between the cord insertion and the placental edge (middle fragment).²⁶

Diagnosis of infection by Plasmodium.

Peripheral and placental blood were tested for Plasmodium infection by microscopy of Field-stained thick films and qPCR. An experienced microscopist based at the field laboratory determined the presence of infection by counting the number of parasites per 200 leukocytes, based on a mean count of 8,000 leukocytes per microliter of blood. Samples were negative when no parasites were detected in 200 high power (100×) fields. For the qPCR, DNA was extracted from placental blood spots on filter paper using the Chelex method described by Plowe et al.³² A qPCR reaction was performed as described previously.³³ Briefly, samples were first tested for Plasmodium using a genus-specific set of primers and probe. The reaction was performed in a final volume of 25 μL containing 5 μL of DNA, 12.5 μL of TaqMan universal master mix (Applied Biosystems, Bedford, MA), 200 nM of each primer (Plasmo1 and Plasmo2), 50 nM of Plasprobe on the ABI 7500 FAST platform, under universal cycling conditions as published. Samples with a cycle threshold (Ct) value under 45 were tested in a duplex species-specific real-time PCR reaction for P. falciparum and P. vivax.³³

Expression analysis of immune mediators in placental tissue.

A reverse-transcription real-time PCR assay with relative quantification (qRT-PCR) was used to evaluate the expression of immune mediators of inflammation (TNF, IFN-γ, IL-8, and CD54), anti-inflammation (IL-10, TGF-β, and IL-13), regulation (FOXP3, CTLA4, TNFRII, CD163, and PD-L1), and co-stimulation (CD86 and CD40) of the immune response in placental tissue. Total RNA was extracted using QIAamp RNA Blood Mini (Qiagen, Maryland, USA) from the placental tissue stored at 4°C in RNALater (Qiagen) following the manufacturer’s instructions. The EXPRESS OneStep Superscript qRT-PCR kit (Invitrogen) was used for reverse transcription and amplification of each molecule, using a StepOnePlus real-time PCR system (Applied Biosystem) and the primers and probes listed in Table 1. The OneStep Software V2.3 was used for data analysis. Relative quantification was calculated using the 2 – ΔΔCt method³⁴ and normalized with β-actin as the reference gene. A pool of RNA from placental tissue of healthy pregnant women was used as a calibrator. All experiments included a no-template control and samples were tested in triplicate.

Table 1.

Primers and probes used in this study

Gene	Forward primer (5'–3')		Reverse primer (5'–3')	Probe (5'–3')	Product size (bp)
IL-10		CCTGGAGGAGGTGATGCCCCA	CAGCGCCGTAGCCTCAGCC	CAAGGCGCATGTGAACTCCCTG	131
IL-13		GGAGCTGGTCAACATCACCC	CGTTGATCAGGGATTCCAGG	GGAGCATCAACCTGACAGCTGGC	116
TGF-β		TCAGAGCTCCGAGAAGCGGTA	GTTGCTGTATTTCTGGTACAT	CCGGGCAGAGCTGCGTCTGCTGA	92
IFN-γ		GAAGAATTGGAAAGAGGAGAGTGA	TGGACATTCAAGTCAGTTACCG	TTCCTTGATGGTCTCCACACTCTTTTGG	218
TNF		GCCCAGGCAGTCAGATCA	GCTTGAGGGTTTGCTACAACA	CCCGAGTGACAAGCCTGTAGCCC	74
FOXP3		GAGAAGCTGAGTGCCATGCA	GGAGCCCTTGTCGGATGAT	CCACCTGGCTGGGAAAATGGCAC	87
CTLA4		GCTCAGCTGAACCTGGCTAC	CGTGCATTGCTTTGCAGAAGAC	CCTGCACTCTCCTGTTTTTTC	88
PD-L1		CTGTGAAAGTCAATGCCCCATAC	CAGTTCATGTTCAGAGGTGACTG	CCAAAGAATTTTGGTTGTGGAT	80
TNFRII		CTGCCATGGTGTGTCCCTC	GGCAGGTCACAGAGAGTCAG	CATGGACGTTCGGGGCATGCT	88
IL-8		CAGCTCTGTGTGAAGGTGC	GGTGGAAAGGTTTGGAGTATGTC	AGTTTTGCCAAGGAGTGCTAAAGAACT	87
CD-86		GTCAGTGCTTGCTAACTTCAGTC	CTCATCTTCTTAGGTTCTGGGTAAC	CAGAAAATGTGTACATAAATTTGACC	120
CD-40		TCTCACCTCGCTATGGTTCGT	GATGGACAGCGGTCAGCAA	TGCCTCTGCAGTGCGTCCTCTGG	70
CD-163		GATCACATGTGACAACAAGATAAGAC	GGAACCTCCATGCCAGATCT	GGACCCACTTCCTGTTCTGGACG	83
CD-54		GCAGACAGTGACCATCTACAGCTT	CTTCTGAGACCTCTGGCTTCGT	CCGGCGCCCAACGTGATTCT	68
β-actin		CGAGCGCGGCTACAGCTT	CCTTAATGTCACGCACGATT	ACCACCACGGCCGAGCGG	58

Quantification of IgG against tetanus toxoid.

Levels of anti–tetanus toxoid (TT) IgG were quantified by ELISA in the serum samples obtained from maternal peripheral blood at the time of delivery and in the umbilical cord blood. The antibodies were quantified with the kit Tetanus ELISA IgG (Vircell S.L., Granada, Spain) following the manufacturer’s instructions. The samples were evaluated in triplicate and diluted 1:2 with the test diluent. Plates were read in a Labsystems Multiskan microplate reader (Thermo Scientific, Waltham, MA, USA) at 450 nm. The calculations and regression analyses were performed manually in Microsoft Excel 2013.

ELISA for measuring rotavirus specific IgG in serum.

Rotavirus specific IgG was determined in serum of peripheral maternal and umbilical cord blood by ELISA, as described previously,³⁵ with some modifications. For total anti-rotavirus (anti-RV) IgG, Immulon 2 HB Flat Bottom MicroTiter Plates were coated with 70 μL of 1/10 dilution (in phosphate-buffered saline [PBS] pH 7.4) of supernatant from rotavirus 89-12 strain virus-infected MA104 cells [10⁷plaque forming units (PFU)/mL] or the supernatant of mock-infected MA104 cells (negative control) in PBS pH 7.4 and incubated overnight at 4°C (donated by Dr. Manuel Antonio Franco of the Universidad Javeriana, Bogotá, Colombia). After three washes with PBS-Tween 0.05%, 150 μL of PBS-Tween 1% bovine serum albumin [BSA] 1% was used to block nonspecific binding sites and incubated at 37°C for 1 h. Serum samples were diluted 1:400 in PBS-Tween 0.01%–BSA 0.1%, and 70 μL of the dilution was added to each well. After 2 h of incubation at 37°C, the plates were washed five times with wash solution and incubated at 37°C for 1 h with 70 μL of anti-IgG-HRP (A0170 Sigma) diluted 1:6000 in PBS-Tween 0.01% BSA 0.1%. After three washes, 100 μL of TMB (T0440 Sigma-Aldrich St. Louis, MO, USA) was added and the reaction was stopped with 50 μL of 1M sulfuric acid. Plates were read in a Labsystems Multiskan microplate reader (Thermo Scientific) at 450 nm. A serial dilution of pooled sera of children with anti-RV IgG was used as a positive control. Serum from an infant without evidence of previous rotavirus infection or placental transfer of maternal IgG anti-RV was used as a negative control. Optical densities (OD) were converted to arbitrary units (AU)³⁶ relative to a positive control that was run on each plate, according to the formula: AU = [(OD test sera – OD mock test sera)/(OD positive control – OD mock positive control)] * 100. Mock test serum corresponds to the incubation of serum in the plate coated with supernatant of mock-infected MA104 cells; mock positive control corresponds to positive control sera incubated in a plate coated with supernatant of mock-infected MA104 cells. Cutoffs for anti-RV IgG were determined on the basis of the OD of the negative control plus 2 SD and converted to AU against the positive control on the same plate. The calculations and regression analyses were performed manually in Microsoft Excel 2013.

Histologic study of the placental tissue.

All placental biopsy specimens were examined by trained staff without prior knowledge of the maternal characteristics, pregnancy outcomes, or malaria episodes in pregnancy. Paraffin-embedded placental specimens were cut to yield at least three sections at a thickness of 5 mm, stained with hematoxylin and eosin, and examined by microscopy under 40× (magnification = 400) and 100× (magnification = 1,000; high-power field).¹³ Histological analysis of the decidua, the intervillous space (IVS) and the villi was based on specific variables established by the researchers (Table 2). Two slides were prepared from each placenta, and each slide corresponded to the central and the middle fragments. Variables were evaluated in two ways: qualitatively (present/absent) and quantitatively (for each variable, the absence of the condition corresponded to a value of zero; the presence of the condition was measured by adding the number of conditions in each of the 40 fields). The variables evaluated quantitatively were immune cells, fibrin deposits, syncytial nodes, placental villi, and fetal capillaries. The variables evaluated qualitatively were atherosis, decidual necrosis, abruption, edema, villous infarction, hemorrhage, thrombus, and calcifications.

Table 2.

Characterization of variables for the histological study

Decidua*
Immune cells	Total cells in the immune infiltrates
Atherosis	Change in arteries of the decidua, specifically the thickening of the arterial endothelium
Decidual necrosis	Ischemic area with degenerative lesions in the decidua
Abruption	Heavy hemorrhage observed in the decidua
Villi*†
Immune cells	Total cells in the immune infiltrates
Fibrinoid or fibrin deposits	Accumulation of fibrin in the villous stroma or around the villi
Syncytial nodes	Small areas of terminal villi with thinning of the syncytiotrophoblast that covers them and thickening of the basement membrane of the trophoblast
Edema	Fluid accumulation in stroma of the chorionic villous, characterized by expansion or swelling of the villous and presence of voids in the stroma
Villous infarction	Wide nodules of fibrinoid degeneration surrounding the villi
Placental villi	Placental structure, formed by trophoblast cells, through which the exchange between maternal and fetal blood takes place.
Fetal capillaries	The capillaries of the villi are terminal branches of the umbilical blood vessels.
IVS*†
Immune cells	Total cells in the immune infiltrates
Hemorrhage:	Increase in red blood cells in the IVS
Thrombus	Blood clot as a result of bleeding
Calcifications	Calcium salt deposition in tissue, causing tissue degeneration

IVS = intervillous space.

* The presence or absence of the histological event and the amount of it in decidua, villi, and IVS were recorded.

^†The presence, quantity, and location of infected erythrocytes and hemozoin deposits were evaluated.

Statistical analysis.

Most of our data were not normally distributed, based on Kolmogorov–Smirnov test; thus, the nonparametric Mann–Whitney U and Kruskal–Wallis tests were performed to evaluate differences between the groups. IBM SPSS Statistics (version 24) was used. Comparisons were made between the noninfected group (no-PI) versus the infected groups (SPID and PIP), and additional comparisons were made between the Plasmodium species (P. vivax versus P. falciparum). Significance was accepted for all analyses at P < 0.05. Spearman’s rho was used to measure the correlation between the variables. After a significant Kruskal–Wallis test, Dunn’s test of multiple comparisons was performed for adjustment. Placental transfer was measured as the ratio of antibody levels in the cord blood over the maternal blood (cord-to-mother ratio [CMR]).

Ethics.

The study protocol was reviewed and approved by the Ethics Committee of the Instituto de Investigaciones Médicas of Universidad de Antioquia (endorsement March 31, 2016). Each participant gave full informed consent according to the Declaration of Helsinki convention and the Colombian regulations for this type of research. Each subject voluntarily agreed to participate in the study.

RESULTS

General characteristics of pregnant women and newborns.

The general characteristics of the study women and their newborns are presented in Table 3. The ages of women and gestational ages were similar across all groups (SPID, PIP, and no-PI). In general, the pregnant women were young (median 23 years, interquartile range [IQR] 16–28 years) and had full-term deliveries. The parity was different between the groups based on the Kruskall–Wallis test; however, the Dunn’s test showed differences between the groups associated with infection (greater parity in the SPID group than in the PIP group), but there were no differences with the control group (non-PI). Hemoglobin levels were significantly different between the groups evaluated with the Kruskal–Wallis test (Table 3), and the Dunn’s test showed lower levels of hemoglobin in the women with SPID with respect to the control group (P < 0.01).

Table 3.

General characteristics of pregnant women and newborns according to infection status by quantitative polymerase chain reaction (qPCR)

Variable	SPID			PIP			No-PI			P*
	n	Med.	IQR 25–75%	n	Median	IQR 25–75%	n	Med.	IQR 25–75%
Age (years)	50	23	19–27	37	20	16–25	25	22	18–28	0.2017
Parity	50	2	2–4	37	1	1–2	25	2	1-5	0.0039
Gestational age (weeks)	50	39	37.8–39	37	39	38–39	25	38	38–39.5	0.3135
Hemoglobin (g/dL)	46	10.8	10.2–11.9	19	11.4	10.4–12.3	25	11.9	11.2–12.4	0.0016
Birth weight (g)	50	3000	2,675–3,300	36	3100	2,900–3,475	25	3,100	2,975–3,500	0.1392

IQR = interquartile range; Med. = median.

^*P value based on the Kruskal–Wallis statistical test; comparison of submicroscopic plasmodial infection at delivery (SPID) group, plasmodial infection during pregnancy (PIP) group, and no plasmodial infection (no-PI).

When the birth weight of the neonates was analyzed, no difference was found between the groups (Table 3). Only 5% of newborns had low birth weight (< 2,500 g), and these cases only occurred in the groups with malaria in pregnancy (two in SPID and two in PIP groups).

In the SPID group (n = 50) 50% of women had infection in peripheral blood and 50% had placental infection detected by qPCR. Of these samples, 40% (20/50) were infected with P. falciparum, 54% (27/50) with P. vivax, and 6% (3/50) had mixed infections. Women included in the PIP group had microscopic (50%) and submicroscopic (50%) infections detected at one or more antenatal visits at any time in pregnancy. The distribution of the infecting species in the PIP group was 66% (25/38) P. vivax infection, 32% (12/38) P. falciparum infection, and 2% (1/50) mixed infection.

Anti-TT and anti-RV IgG levels in peripheral maternal and umbilical cord blood.

The levels of anti-TT and anti-RV IgG were similar in maternal peripheral blood among the groups, but in umbilical cord blood, higher anti-TT IgG levels were observed in the no-PI group compared with the infected groups SPID (P < 0.001) and PIP (P < 0.05).

In the SPID group, there were significant differences in the levels of anti-RV and anti-TT IgG, with a greater amount of anti-RV IgG in cord blood (median 66.3 UI/mL, IQR 9.4–91.0 UI/mL) than maternal blood (median 47.3 UI/mL, IQR 26.1–65.0 UI/mL; P = <0.0062) (Figure 2A). In contrast, there was more anti-TT IgG in maternal blood (median 4.9, IQR 4.05–5.3 UI/mL) than in cord blood (median 3.7 UI/mL, IQR 2.2–4.7 UI/mL; P = 0.0003) (Figure 2B). In the PIP group, there were no differences in the levels of either of the two types of antibodies (Figure 1A and B). In the no-PI group, there was only a difference in anti-TT IgG (P = 0.0108), with a greater amount of anti-TT IgG in the cord blood (median 5.3 UI/mL, IQR 4.85–5.85 UI/mL) than in the maternal blood (median 4.8 UI/mL, IQR 4.2–5.3 UI/mL) (Figure 2B). This suggests a higher amount of anti-RV IgG in the cord blood in the SPID and PIP groups compared with no-PI group and a lower amount of anti-TT IgG in the cord blood in the SPID group compared with the other groups (Figure 2C and D, Table 4).

Figure 2.

Anti–tetanus toxoid (TT) and anti-retrovirus (RV) IgG levels in maternal peripheral and umbilical cord blood. Median and interquartile range (25–75%) of anti-TT IgG expressed in international units per milliliter (IU/mL) (A) and anti-RV IgG expressed in arbitrary units (UA) (B) in maternal peripheral blood and umbilical cord blood, in each of the groups. P value by Mann–Whitney test. Mean and 95% confidence interval (CI) of the ratio of the levels of anti-RV IgG (C) and anti-TT IgG (D) cord blood-to-mother blood (CMR) in the evaluated groups. Mean and 95% CI of the ratio of the levels of anti-RV IgG (E) and anti-TT IgG (F) CMR blood in the submicroscopic plasmodial infection at delivery (SPID) and plasmodial infection in pregnancy (PIP) groups by species.

Table 4.

Cord-to-maternal ratio (CMR) of anti-RV and anti-TT in the study groups

	SPID		PIP		no-PI
	CMR*	95% CI	CMR†	95% CI	CMR‡	95% CI
anti-RV	9.630	−5.986 to 25.25	1.440	1.108 to 1.772	1.248	0.832 to 1.664
anti-TT	0.687	0.594 to 0.780	1.033	0.897 to 1.169	1.083	1.040 to 1.126

CI = confidence interval; CMR = Cord-to-maternal ratio; No-PI = no plasmodial infection; PIP = plasmodial infection in pregnancy; RV = retrovirus; SPID = submicroscopic plasmodial infection at delivery; TT = tetanus toxoid. Maternal and cord blood plasma samples were tested for the presence of anti-RV–and anti-TT-specific antibodies.

^*Mean of the ratio of cord blood IgG with respect to maternal blood in the SPID group.

^†Mean of the ratio of cord blood IgG with respect to maternal blood in the PIP group.

^‡Mean of the ratio of cord blood IgG with respect to maternal blood in the no-PI group.

The CMR in the SPID and PIP groups was evaluated according to the Plasmodium species (Figure 2E and F, Table 5). The SPID-Pf and SPID-Pv groups has CMR > 1 with a mean of 3.305 and 2.289 for anti-RV, respectively. For anti-TT, the mean of CMR was < 1 in both groups (SPID-Pf [0.971] and SPID-Pv [0.770]), being lower in the SPID-Pv group (Figure 2E and F, Table 5). These values of CMR were only significant for the SPID-Pv group.

Table 5.

Cord to maternal ratio (CMR) of anti-RV and anti-TT in the SPID and PIP groups by species

	SPID Pf		SPID Pv
	CMR*	95% CI	CMR*	95% CI
Anti-RV	3.305	0.665–5.945	2.289	1.107–3.470
Anti-TT	0.971	0.672–1.270	0.770	0.553–0.988

	PIP Pf		PIP Pv
	CMR†	95% CI	CMR†	95% CI
Anti-RV	1.015	0.710–1.320	1.212	0.915–1.510
Anti-TT	1.096	0.226–1.965	1.486	0.820–2.152

CI = confidence interval; CMR = cord-to-maternal ratio; No-PI = no plasmodial infection; PIP = plasmodial infection in pregnancy; Pf = Plasmodium falciparum; Pv = Plasmodium vivax; RV = retrovirus; SPID = submicroscopic plasmodial infection at delivery; TT = tetanus toxoid. Maternal and cord blood plasma samples were tested for the presence of anti-RV- and anti-TT-specific antibodies.

^*Mean of the ratio of cord blood IgG with respect to maternal blood in the SPID group by species.

^†Mean of the ratio of cord blood IgG with respect to maternal blood in the PIP group by species.

Effect of malaria in pregnancy on the expression of immune mediators in placental tissue.

To determine the transcriptional changes resulting from malaria in pregnancy and their association with the passage of maternal antibodies via the transplacental route, the expression of different mediators of the immune response in placental tissue was measured in the three study groups.

Statistically significant differences were observed in the study groups for all immune mediators associated with anti-inflammation and regulation: IL-10 (P < 0.0001), TGF-β (P = 0.0002), FoxP3 (P = 0.035), CTLA-4 (P = 0.0004), PD-L1 (P = 0.0265), TNF-RII (P = 0.0387), and CD-163 (P = 0.0227) (Figure 3A–G). There was a significant increase in the expression of IL-10 (Figure 3A), TGF-β (Figure 3B), FoxP3 (Figure 3D), and CTLA-4 (Figure 3E) in the SPID group compared with the no-PI group. No significant differences in the expression of these genes were observed between the PIP group and the no-PI group.

Figure 3.

Relative quantification of the expression of immune mediators in placental tissues. Expression levels of immune mediators in placental tissue in study groups. The measure for all variables is the median and interquartile range 25% to 75% of the relative units (RU) of expression in which the basal expression of the constitutive gene was compared with the expression of each cytokine according to the cycle threshold (CT) value. Significant P values according to Dunn’s test for multiple comparisons. No-PI = no plasmodial infection; PIP = plasmodial infection in pregnancy; SPID = submicroscopic plasmodial infection at delivery. The dotted line corresponds to the median expression of each immune mediator in the control group. *P < 0.05; **P < 0.01; ***P < 0.001.

The expression of the pro-inflammatory molecules IFN-γ (P < 0.0001) and TNF (P < 0.0001) was significantly different between the groups (Figure 3H–I), specifically we observed an increase in the expression of IFN-γ in the SPID and PIP groups compared with the no-PI group (Figure 3H).

No significant difference was observed in the expression of the costimulatory molecules CD-86 and CD-40, the adhesion molecule ICAM (CD-54), and chemokine IL-8 among the groups.

Given the possibility that other independent variables support the findings found in the immune profile of the study groups, the modification of the effect was evaluated by multivariate analysis, taking into account two criteria: first, the association with the dependent variable (study group), and second, the association with independent variables. On this basis, in the SPID group, parity was positively correlated with IFN-γ expression. When performing this same analysis in the PIP group, parity was not correlated with the expression of immune mediators. However, when these same correlations were analyzed in the non-PI group, the parity variable was negatively correlated with TGF expression and positively correlated with CTLA-4 and TNF-RII expression. Those results show great heterogeneity in the correlations presented with the independent variable parity among the study groups; as such, there was no modification of the effect by the independent variables.

Expression of these modulators was compared in the SPID and PIP groups based on the infecting species of Plasmodium. In the SPID group, IFN-γ expression was higher in P. falciparum infection and TNF expression was higher in P. vivax infection. In the PIP group, no differences were observed in the expression of immune mediators according to the Plasmodium species (Table 6).

Table 6.

Relative quantification of the expression of immune mediators in placental tissues in the SPID and PIP groups by species

	Plasmodium falciparum		Plasmodium vivax		P*
	Median	IQR 25–75%	Median	IQR 25–75%
SPID
IL-8	0.6	0.2–2.1	0.4	0.2–1.7	0.8644
IL-10	3.5	3.2–4.0	3.3	1.6–4.1	0.5315
IFN-γ	10.3	9.3–11.2	9.1	8.2–9.4	0.0009
TNF	10.2	9.3–11	11.4	9.4–12	0.0442
TGF-β	2.9	2.5–2.9	2.2	0.6–2.6	0.0592
FoxP3	0.9	0.25–1	0.8	0.4–1.0	0.9249
CTLA-4	2.7	2.05–7.95	3.0	1.5–5.3	0.7051
PD-L1	0.3	0.2–1.4	0.2	0.1–1	0.5735
TNF-RII	0.8	0.2–11.2	0.55	0.3–4.2	> 0.9999
CD-86	0.4	0.2–17.2	0.95	0.4–3.07	0.3516
CD-40	1.65	0.22–8.1	1.4	0.7–6.2	0.5826
CD-163	0.25	0.1–1.05	0.3	0.2–0.9	0.3512
CD-54	0.80	0.4–1.1	0.8	0.7-1.1	0.4079
PIP
IL-8	1.5	0.4–3.25	0.65	0.17–1.62	0.2059
IL-10	0.5	0.1–1.7	0.55	0.17–1.22	0.9787
IFN-γ	3.0	0.1–7.1	4.2	3.2–9.0	0.1044
TNF	0.6	0.2–2.65	1.7	0.6–4.65	0.2211
TGF-β	0.7	0.15–0.95	0.6	0.1–1.6	0.8125
FoxP3	0.4	0.3–2.05	0.65	0.3–1.2	0.7164
CTLA-4	1.3	1.05–2.9	1.2	0.7–1.7	0.2486
PD-L1	0.5	0.2–1.65	0.6	0.17–1.3	0.7933
TNF-RII	0.3	0.2–0.5	0.3	0.2–1.7	0.4931
CD-86	0.3	0.25–0.65	0.5	0.2–3.6	0.4349
CD-40	0.9	0.5–1.4	0.6	0.3–1.2	0.4465
CD-163	0.4	0.25–1.75	1.25	0.27–2.42	0.4931
CD-54	0.9	0.75–1.4	0.95	0.77–1.4	0.9528

IQR = interquartile range; PIP = plasmodial infection in pregnancy; SPID = submicroscopic plasmodial infection at delivery. Expression levels of immune mediators in placental tissue in study groups. The measure for all variables is the median and IQR of the relative units (RU) of expression in which the basal expression of the constitutive gene was compared with the expression of each cytokine according to the cycle threshold (CT) value.

^*P value by the Mann–Whitney test.

Bold values denote statistically significant difference.

Placental histological findings.

Decidual necrosis was detected exclusively in the presence of malaria in pregnancy, while almost absent in the no-PI group. The number of immune cells in the decidua, villi, and IVS, as well as the number of syncytial nodes, edema, atherosis, hemorrhage, placental villi, and capillaries per villus was significantly different between the groups. All histopathological findings were observed in placentas infected by P. vivax or P. falciparum.

The number of immune cells in the villi, IVS, and decidua was greater in malaria in pregnancy (SPID and PIP groups) compared with the no-PI group (Figure 4A–C). The number of placental villi and syncytial nodes were greater in the PIP group with respect to the no-PI group (Figure 4D and F). The fetal capillaries per villi, atherosis, and hemorrhage were lower in the PIP group compared with the no-PI group (Figure 4E, G, and I). Edema was statistically different among groups with the Kruskal–Wallis test, but no significant differences were observed between the groups in the post hoc test (Figure 4H).

Figure 4.

Histological findings with statistical significance using the Kruskal–Wallis test. Median and interquartile range (25–75%) of the variables in 40 fields. Significant P values according to Dunn’s test for multiple comparisons. No-PI = no plasmodial infection; PIP = plasmodial infection in pregnancy; SPID = submicroscopic plasmodial infection at delivery. *P < 0.05; **P < 0.01; ***P < 0.001.

Of the 50 women with SPID, 20% (10/50) of placentas had deposits of hemozoin, compared with only 5% (2/38) among placentas from women with PIP. Infected erythrocytes were not observed in any of the placentas.

The SPID and PIP groups were compared based on the infecting species of Plasmodium. In the SPID group, the number of fetal capillaries per villi was higher with P. falciparum infection. In the PIP group, the villi immune cells, decidua immune cells and the number of fetal capillaries per villi were higher with P. vivax infection (Table 7).

Table 7.

Histological findings in placental tissues in the SPID and PIP groups by species

	Plasmodium falciparum		Plasmodium vivax		P*
	Median	IQR, 25–75%	Median	IQR, 25–75%
SPID
Intervillous space immune cells	107.5	76.2–177.3	107	65–194	0.6051
Villi immune cells	41.5	22.2–70.5	34	22–42	0.2634
Decidua immune cells	42.5	26.5–68.2	34	4–58	0.2537
Placental villi	357.5	299–404	358	290–397	0.7888
Fetal capillaries per villi	7.77	6.4–9.3	5.3	4.7–8.34	0.0052
Syncytial nodes	119.5	84.5–172.8	97	70–139	0.1370
Edema	4.5	0–11.7	4	0–9	0.8831
Hemorrhage	10.5	4–17.5	9	5–18	0.8517
PIP
Intervillous space immune cells	103	28.5–214.5	159	110.3–182.8	0.2592
Villi immune cells	39	15–54.5	55	34.2–94.5	0.0632
Decidua immune cells	5	2–10	10.5	5.5–18.75	0.0188
Placental villi	406	322.5–523.5	495.5	107–590.5	1.1660
Fetal capillaries per villi	4.49	4.13–6.07	6.11	5.15–7.1	0.0310
Syncytial nodes	132	108.5–188	132	114–151.3	> 0.999
Edema	3	0–7.5	5	1.25–8.5	0.5939
Hemorrhage	7	2.5–23	7	3–9.75	0.5337

IQR = interquartile range; PIP = plasmodial infection in pregnancy; SPID = submicroscopic plasmodial infection at delivery. Histological findings in placental tissues in 40 fields. The measure for all variables is the median and IQR of the variables in 40 fields.

^*P value by Mann–Whitney test.

Bold values denote statistically significant difference.

Significant correlations between anti-RV and anti-TT IgG in umbilical cord with anti-RV and anti-TT IgG in maternal blood and placental histological findings.

On the basis of the hypothesis that placental histological changes caused by malaria in pregnancy interfere with maternal to fetal transfer of antibodies, we explored correlations between anti-TT and anti-RV IgG levels in umbilical cord blood with placental histological findings in the three study groups.

In the SPID group, the levels of anti-TT IgG in maternal and umbilical cord blood were strongly correlated (rho = 0.895, P < 0.0001) (Figure 5A). This positive correlation indicates that the amount of anti-TT IgG in maternal blood is directly proportional to the amount of IgG found in umbilical cord blood. In the same group, there was a moderate positive correlation between the levels of anti-RV IgG in maternal and umbilical cord blood (rho = 0.470, P = 0.0007) (Figure 5B). Also, a weak negative correlation was observed between the levels of anti-RV IgG from umbilical cord blood and the number of immune cells in the decidua (rho = –0.321, P = 0.0241) (Figure 5C). In this case the levels of anti-RV IgG in the umbilical cord blood were inversely proportional to the number of immune cells in the decidua.

Figure 5.

Correlations between anti-retroviral (RV) and anti–tetanus toxoid (TT) IgG in umbilical cord, maternal blood, and in relation to placental histological findings. Significant correlations (P < 0.05) based on Spearman’s rho test. IVS = intervillous space; No-PI = no plasmodial infection; PIP = plasmodial infection in pregnancy; SPID = submicroscopic plasmodial infection at deliver.

In the PIP group, the levels of anti-TT IgG in maternal and umbilical cord blood were positively correlated (rho = 0.408, P = 0.018) (Figure 5D) and a weak positive correlation was observed between the levels of anti-RV IgG in the umbilical cord blood and the number of immune cells in the IVS (rho = –0.321, P = 0.0241) (Figure 5E). A strong positive correlation was observed in the no-PI group between the levels of anti-TT IgG in maternal and umbilical cord blood (rho = 0.939, P < 0.0001) (Figure 5F).

Significant correlations between CMR of anti-RV and CMR of anti-TT in umbilical cord with the expression of immune mediators and placental histological findings.

In the SPID group, the CMR of anti-TT was weakly correlated with the immune mediators CD54, CD163, TGF-β, and IL8 (Table 8). In the PIP group, only one correlation was found, the CMR of anti-TT in cord blood was weakly correlated with the expression of CTLA-4 (Table 8). These positive correlations indicate that the ratio of anti-TT IgG in cord blood is directly proportional to the expression of these immune mediators in the placental tissue. No correlations were observed in the no-PI group nor for the CMR of anti-RV.

Table 8.

Correlations between the CMR of anti-TT IgG in umbilical cord and the expression of immune mediators in placental tissue

Variable 1	Variable 2	P value	Correlation
SPID group
CMR of anti-TT	CD-54	0.004	0.409
	CD-163	0.050	0.284
	TGF-β	0.025	0.323
	IL-8	0.004	0.421
PIP group
CMR of anti-TT	CTLA-4	0.004	0.497

CMR = cord-to-maternal ratio; PIP = plasmodial infection in pregnancy; SPID = submicroscopic plasmodial infection at delivery; TT = tetanus toxoid. Significant correlations (P < 0.05) based on Spearman’s rho test.

DISCUSSION

This study tested the hypothesis that exposure to malaria in pregnancy would interfere with the transplacental transfer of anti-RV and anti-TT IgG from the mother to the neonate (CMR ≥ 1.0) (Table 4). For anti-TT IgG, we observed significant differences between the antibody levels in umbilical cord compared with maternal blood in the control group (no-PI). This increase corresponded to 8.3% greater amount of anti-TT IgG in cord blood, a finding that is supported by the literature.¹⁴ No significant differences were observed for anti-RV IgG levels in this group. In the SPID group, there were significant differences in the passage of both antibodies between maternal and cord blood, but they correspond to the fact that anti-TT levels were 31.3% lower in cord blood and, in contrast, anti-RV levels were 8.6 times higher in cord. It should be noted that previous studies demonstrated a decrease in the passage of IgG antibodies against tetanus toxoid in cases of microscopic placental malaria caused by P. falciparum.^20,21 Here it was found that even submicroscopic infections affect the passage of antibodies against tetanus toxoid. In the PIP group, anti-RV IgG levels were 44% higher in cord blood and for anti-TT IgG, the concentration of antibodies in maternal and cord blood was not different.

This was also the first study in Colombia to evaluate levels of rotavirus antibodies in pregnant women and transplacental neonatal passive immunity against rotavirus. Considering that in pregnant women the levels of anti-RV IgG were acquired naturally by infection with the virus, a high percentage of seropositivity was observed (94%). Antibody levels were highly variable, and neither maternal nor cord anti-RV IgG levels differed between study groups.

We observed different results between the passage of maternal anti-TT IgG and anti-RV IgG to the cord, particularly in the SPID group. There are several possible explanations for the increase in passive transfer of anti-RV IgG in the SPID group, in contrast to that observed for anti-TT IgG. Anti-RV IgG levels in pregnant women are highly variable. Anti-RV IgG are acquired in natural rotavirus infections throughout life, causing them to vary from individual to individual, whereas for TT, antibody levels are homogeneous due to vaccination before and during pregnancy. It is also possible that the concentration of anti-RV IgG is low enough that the antibodies do not saturate the FcRn receptors, in contrast to anti-TT IgG resulting from vaccination during pregnancy. The anti-RV IgG may be better at competing for binding to the finite FcRn receptors and thus are easily transported across the placenta. Furthermore, pregnancy is associated with increased galactosylation and sialylation of the Fc and Fab regions and these modifications to IgG can affect binding affinity.^37–39 The precise mechanisms of differential transfer of antibodies need to be investigated further.

Different results were observed in the CMR of the SPID group when analyzed by species: the CMR of the anti-TT was considerably lower in the SPID-Pv compared with the SPID-Pf group. For the CMR of the anti-RV in submicroscopic infection, a high transfer of antibodies is observed, this being more efficient and significant in the SPID-Pv group compared with SPID-Pf. In the comparison of the CMR in the PIP group by species, no differences were observed. Despite the small number of samples in the study groups, these results indicate that during submicroscopic infection with P. falciparum, anti-RV IgG are transferred more efficiently and that during infection with P. vivax there is a greater reduction in the transfer of anti-TT IgG. No previous studies were identified that evaluated the transfer of anti-TT across the placenta in maternal infections due to P. vivax. Nor were there any studies that evaluated the passage of anti-RV IgG through the placenta in maternal P. vivax or P. falciparum infections. A previous study reported a decrease in the passage of anti-TT IgG in maternal microscopic P. falciparum infections;²¹ in our study, submicroscopic P. falciparum infections did not significantly affect the passage of antibodies to the cord. Surprisingly, in submicroscopic infection by P. falciparum, anti-RV IgG was transferred more efficiently. It is possible that there is less inflammation in the placenta due to the presence of antibodies against VAR2CSA, which may limit the parasite burden. Alternatively, prior immunity to Plasmodium could interfere with this analysis.

Other studies that evaluated the effect of plasmodial infection on transplacental neonatal passive immunity^{12,15,16,19–22} did not take into account the placental changes produced by the infection and its possible association with the passage of maternal antibodies. In the present study, the transcriptional changes caused by malaria infection during pregnancy were evaluated. It was expected that a greater number of transcriptional changes would be observed in the SPID group, since the infection in these women was present at the time of measurement (at delivery). In this group, compared with the control group without infection, an increase in anti-inflammatory (IL-10 and TGF-β) and pro-inflammatory (IFN-γ and TNF) cytokines was observed. These data are consistent with previous studies showing that submicroscopic placental infections cause increased expression of pro-inflammatory cytokines.^26,27 The evidence of inflammation in the placenta could be associated with the decrease in the passage of anti-TT IgG in the SPID group. It is important to highlight that the expression of IFN-γ negatively regulated the constitutive expression of the FcRn receptor in epithelial cells, THP-1 cells and peripheral blood mononuclear cells evaluated by real-time RT-PCR and Western blot.⁴⁰ Few placental transcriptional changes were observed in the PIP group; only an increase in IFN-γ and FoxP3 expression were observed. Despite being negative at the time of sample collection, the history of malaria infection during pregnancy also altered the placental transcriptional profile.

On the other hand, there were few differences in the expression of immune mediators between the groups of women infected by P. vivax and by P. falciparum, which suggests that both species have a similar impact on the placental transcriptional profile and supports the idea that both species have comparable consequences during pregnancy.⁴¹ However, the differences in the expression of cytokines such as TNF and IFN-γ in P. vivax infections can be explained by the different pathophysiological mechanisms of malaria by these two species. The increase in TNF observed with P. vivax compared with P. falciparum is mainly associated with feverish paroxysms.⁴² The increase in IFN-γ expression in P. falciparum infections compared with P. vivax is associated with the higher frequency of severe malaria.^43,44

This histological evaluation showed a significant increase in the number of immune cells in the villi, in the intervillous space and in the maternal decidua in the two groups associated with infection (SPID and PIP). A previous study showed similar histopathological alterations, such as a decrease in the number of capillaries per villus of the placenta and an increase in the infiltration of immune cells during submicroscopic plasmodial infections.^45,46 The increased infiltration of immune cells into the placenta is consistent with the increased expression of pro-inflammatory cytokines in the tissue. The inflammation could interfere with the transfer of antibodies to the cord. Additionally, the histological findings in the number of villi and fetal capillaries per villi that were associated with the Plasmodium species could indicate a decrease in the surface area, limiting exchange from the mother to the fetus.

The increase in the expression of pro-inflammatory mediators such as IFN-ɣ and TNF could explain the increase in immune cells in the placenta associated with infection. Some studies have suggested that maternal infections may also influence the quality of transplacental antibody transfer from mother to fetus.^{15,16,19–21,47} However, our study indicates that infection and inflammation could directly influence the transplacental passage of IgG to the cord. Because our understanding of the mechanisms underlying transplacental transfer of IgG is poor, this model remains speculative.

This study has several limitations, including the heterogeneity of the infections of the SPID group that included pregnant women with placental or peripheral infection or both. There were limited numbers of samples per species of Plasmodium in the study groups. The statistically significant difference in the parity of the pregnant women between the SPID and PIP groups and the protection against placental malaria reported for P. falciparum may have influenced comparisons of immune mediators, histological changes, and passive immunity.