INTRODUCTION

Submicroscopic placental plasmodial infection (SPPI) by Plasmodium falciparum is frequent in endemic countries and contributes to the development of maternal anemia and low birth weight ¹. SPPI by P. vivax (SPPI-vivax) is almost unknown in the world.

The standard technique for malaria diagnosis in endemic areas is the thick blood smear, which, by definition, does not detect submicroscopic infections, which are usually asymptomatic. These two conditions, submicroscopic and asymptomatic, had kept SPPI in absolute oblivion until the beginning of the 21st century. But there are already strong arguments suggesting that asymptomatic infections have important consequences for health and society, and should be renamed as chronic malaria infections ².

The immunopathogenesis of placental malaria (PM) should consider pregnancy to be a unique physiological state in which the maternal immune system must protect the mother against infection and other noxae, while modulating her immune response to prevent rejection by the semi-allogeneic fetus ³. The presence of Plasmodium and its byproducts (such as hemozoin) in the placental tissue essentially alters the immune environment that regulates the placenta ⁴.

In the case of gestational and placental malaria, increased cytokines TNF-α, IFN-γ and IL-10 are associated with trophoblastic damage, low birth weight and prematurity ⁵. The expression of placental proinflammatory cytokines is high in women with PM and the expression of anti-inflammatory cytokines is low ^{5 - 9}.

PM alters angiogenic remodeling ^{10 , 11}. Mononuclear-cell infiltration and its inflammatory products are associated with low birth weight; and a shorter and less impactful inflammatory process in multigravid women, when comparing with primigravid women. These differences may explain the different outcomes that these women and their offspring undergo ^{12 - 14}.

In normal pregnancy, the evident physiological processes of apoptosis, inflammation, hypoxia, vasculogenesis and angiogenesis, among others, are strictly controlled. In these processes the role of cellular communication (CC) markers/mediators is decisive ^{15 - 19}. The CC system is a complex network between cells, that sends signals to other cells, the latter with receptor functions, producing a physiological response and a change in cellular activity. In CC, there are many markers of physiological or physiopathological processes and they have different origins (hormones, growth factors, cytokines, prostaglandins, leukotrienes, etc.) ²⁰.

The objective of the study was to determine the association of submicroscopic infection by P. vivax or P. falciparum with histopathological events, the behavior of immune cells, cytokine-associated gene expression, hypoxia, angiogenesis and inflammation in placentas.

MATERIALS AND METHODS

Quantification of the expression of genes associated with markers/mediators

A tissue fragment preserved with RNA Later® (Qiagen) at 4 °C was used to quantify the expression of genes associated with markers/mediators and cytokines in placental tissue. The mediators were grouped as follows: proinflammatory (IL-2, TNF, IFNɣ, COX-1, COX-2, C5a), anti-inflammatory (IL-10, IL-4), angiogenic (VEGF, VEGFR-1) and hypoxic (HIF).

Different markers, except C5a, were measured by relative quantification of mRNA by qPCR. Relative quantification was done to determine the expression levels of the study mediators in relation to the expression levels of the constituent gene and to obtain the relationship between the gene of interest and the constituent gene. The Pfaffl procedure was applied to determine delta-delta TC as follows ²⁶:

By following the manufacturer’s recommendations, the commercial kit Human C5a Elisa Kit was used to quantify the complement C5a fraction in placental serum samples.

RESULTS

From the total, 95% (38/40) of PM cases are submicroscopic infections. From the pregnant women total, 25% (15/59) had history of malaria during the current pregnancy.

The age range was from 14 to 41 years and the average in the three groups was similar (PM-no: 22; PM-F: 25; PM-V: 23). The average gestational age was 38.6 weeks (range from 36 to 41). Previous pregnancies averaged 2.8. From the total, 36% were in their first gestation, 20% in their second and 44% were multi-gestational (3 to 9). Of the 59 deliveries, 5 were by caesarean section.

At the time of delivery, the average hemoglobin value was 11.1 g/dL. Women with PM had lower hemoglobin levels than uninfected women (10.86 g/dL and 11.69 g/dL, respectively). Hemoglobin was lower in the PM-V group (10.75 g/dL) than in the PM-F group (10.97 g/dL), but without significant difference (p=0.506). The mean neonatal weight in the PM-no group was 2,974 g; in the PM-F group, 2,852 g; and in the PM-V group, 2,737 g. The difference in mean neonatal weight was 237 g between children in the PM-V and PM-no groups, and 115 g between children in the PM-V and PM-F groups.

The control group (PM-no) had absence of necrosis and higher frequency of atherosis, abruptio and thrombus than the other two. The PM-V group presented similar results to PM-F in atherosis, necrosis, infarction, fibrin deposits and thrombi, but less hemozoin and infected erythrocytes. Fibrin deposits were observed in all placentas. In the intervillous space, 95-100% of placentas showed hemorrhage and 6 out of 10 placentas showed thrombi and calcifications. There was significant difference (p < 0.05) between the three groups regarding abruption, syncytial nodes and hemozoin. The PM-F group generated this difference. In the species comparison, there was no significant difference (p > 0.05) (Table 1).

Table 1 Frequency of placental histological findings by study group 

In the PM-no group, atherosis and abruption are rare, averaging 1.3 and 2.2 events, respectively. In the same order; infarction and villous edema had 9.1 and 8.7 events on average. In the same PM-no group, the amount of fibrin and syncytial node deposits is 72.5 and 109 on average. Placentas without infection have 366 villi, 1463 capillaries and 3.97 capillaries per villi. The average number of hemorrhages was 18, while the average number of thrombi and calcifications were 2 (they were in 60-70% of them) (Table 2).

Table 2 Magnitude of placental histological events, according to the presence of plasmodial infection 

^a Indicates whether the Dunn test showed a difference between each pair of comparisons (PM-no vs. PM-V; PM-V vs. PM-F) with p < 0.05.

^b DIC: Deciduous immune cells. VIC: villous immune cells. ISIC: immune cells in intervillous space. SD: standard deviation

In PM-F and PM-V groups, the appearance of placental infection reduces the amount of atherosis, villous edema, capillary hemorrhage and thrombi, but increases villous infarction, calcifications and immune cells. Only a few events show significant difference (Table 2).

Immune cells were found in all placentas, in all three compartments, and regardless of infection. The quantities were different, in the PM-no group, 9 immune cells in decidua, 22 in villous and 50 in intervillous space were observed; in the PM-V group, 58, 34 and 130 were observed respectively; in the PM-F group the highest amount of immune cells were found: 68, 45 and 157, respectively (Table 2).

Table 3 shows the expression of inflammatory, angiogenic and hypoxic mediators by group. Inflammation mediators (COX-1, COX-2, IL-10, IFNɣ, TNF, C5a) were significantly higher in PM. IL-2 and IL-4 varied little among the three groups. Although C5a did not show statistically significant difference between the groups, it had higher values in PM-V compared to the two other groups. Regarding angiogenesis, VEGF and VEGFR-1 showed no significant difference between groups, but their expression increased in groups with PM. Regarding hypoxia, HIF-1α was significantly different and with higher values in groups with PM. In general, the mediators that showed no significant difference between the three groups were IL-2, VEGF, VEGFR-1 and C5a. For VEGFR-1 and C5a, the high variability within each group is probably a factor affecting statistical significance when comparing the three groups.

Table 3 Expression of inflammation, angiogenesis and hypoxia mediators according to study group 

^a Indicates whether Dunn’s test showed a difference between each pair of comparisons (PM-no vs. PM-V; PM-V vs. PM-F), with p < 0.05.

^b For C5a, concentration in blood was measured and not the expression of the associated gene.

Figures 1, 2 and 3 show the bivariate linear correlations between histological events and process mediators. In the group without infection, the predominance of negative and weak (p < 0.10) significant correlations (SC) is clear. Fibrin deposits, syncytial nodes and thrombi have the highest amount of SC. The 6 SC that fibrinoid deposits have are made with 7 process mediators, which are inflammation promoters (COX-2 [but not COX-1], C5a, IFNɣ [but not TNF], VEGF) and hypoxia (HIF) and angiogenesis mediators (VEGF, but not its receptor). The 5 SC of the syncytial nodes are made with 3 of the mediators that are also present in the relations with the fibrinoid deposits (COX-2, HIF and VEGF) and with cytokines 2 (pro-inflammatory) and 10 (anti-inflammatory). The 4 SC of thrombi occur with the inflammation mediators, TNF, IL-2, VEGF and the hypoxia mediator HIF.

Figure 1 Significant bivariate linear correlations (SC) between placental histological events 

Figure 2 Significant bivariate linear correlations (CS) between mediators of placental processes of angiogenesis, inflammation and hypoxia. 

Figure 3 Linear correlations between histological events and process mediators in the placenta. 

On the other hand, it is necessary to highlight the SC that several mediators establish with some histological events. This is the case with HIF, which, in addition to being associated with syncytial nodes, fibrinoid deposits and thrombi, it is also associated with infarction. TNF is associated with abruptio, thrombi and villous capillaries.

In the PM-F group, SC reduce and the polarity changes; in this case, the poles are not the syncytial nodes, fibrinoid deposits and thrombi, but the only pole that stands out with the villous capillaries (three SC: with TNF, HIF and IL-10). TNF had SC with edema, infarction, the number of villi and the number of capillaries per villi; IL-2 had SC with abruption, atherosis, thrombi and fibrinoid deposits.

For the PM-V group, SC abound as in the absence of infection, but with different focuses: abruption, thrombus, calcifications. The SC of abruption are formed with COX-1, C5a and HIF-1α; those of thrombi and calcifications with TNF and IL-2. Thrombi are also associated with VEGF, whereas calcifications are related to VEGFR-1.

DISCUSSION

This study found important differences in the placental effects of the two plasmodial species evaluated. The amount of abruption, syncytial nodes, villi, capillaries and capillaries by villi is significantly lower when P. vivax is present, compared to P. falciparum. On the other hand, placentas infected with P. vivax, when compared to placentas without infection, generated more infarction and less abruption, edema, villi, capillaries and capillaries by villi. There is evidence of the pathogenic nature of PM-SPPI due to P. falciparum in pregnant women and their children ¹. Even in studies in the same geographical area where this study took place ^{27 - 29}, submicroscopic placental infection by P. vivax causes harmful effects.

Comparisons of the P. vivax group against the P. falciparum group or against the group without infection strongly support the placental pathogenic capacity of P. vivax. These findings provide a basis for explaining its effects by a mechanism that does not necessarily involve the cytoadhesion of infected erythrocytes to the placental tissue, as occurs in P. falciparum infections.

There is scarce information on the correlations between histological events and process mediators. When there is no infection, SC between histological events and process mediators are predominantly less than 10%. This can be interpreted as a homeostatic state, in which no specific response predominates, but all of them are active, in basal and balanced states. Otherwise gestation would be at risk.

When there is a plasmodial infection, the panorama for SC changes radically, both in the case of placental plasmodial infection by P. falciparum (SPPI-F), but more so in the case of placental plasmodial infection by P. vivax (SPPI-V). SC in the non-infection state disappear and new ones arise in plasmodial infections. The centers or poles of SC that were the syncytial nodes and fibrinoid deposits disappear and are replaced by poles that now correspond to the process mediators, mainly inflammation. IL-2, TNF and IL-10 are poles in SPPI-F, while IL-2, TNF and C5a are poles in SPPI-V. In the absence of infection, the villi and capillaries are virtually CS-free; when infection occurs, they emerge as poles. All this would express the exacerbated inflammatory state of the placenta as a consequence of the infection.

In this study, no hormones were measured at any point in gestation. Progesterone, estrogens, androgens and glucocorticoids are involved in gestation from implantation to delivery; their biosynthesis and metabolism are the result of complex pathways involving the fetus, the placenta and the mother ³⁰. Many sexual hormones clearly interact with the immune system and many mediators of physiological processes also interact with the immune system. There is, therefore, a complex physiological network between hormones, mediators of physiological processes and the immune system.

One of the strengths of this study is the comprehensive approach to PM, addressing the relationship between histopathological events and mediators of inflammation, angiogenesis and hypoxia. To our knowledge these relationships have not been studied before. We had a well-defined control group, which was negative for Plasmodium according to highly sensitive and specific technique (qPCR), without the presence of TORCH syndrome, HIV, eclampsia/preeclampsia, and diabetes. However, the presence of intestinal parasites and malnutrition could not be ruled out.

We consider the small sample size and the selection by convenience as limitations due to time and money available. The mediators evaluated basically represent the inflammatory process of the placenta at full-term, other process mediators were not sufficiently measured. The analysis of the correlations between histological findings and process mediators is superficial because there is little or no information available.

In conclusion, when SPPI occurs, there are placental tissue changes as well as changes in the expression of inflammatory process mediators, whether the causal agent is P. falciparum or P. vivax. If the parasite is pathogenic to the placenta, it is to be expected that when parasitemia increases, effects will be stronger. Placental infection with P. vivax contributes to the increase in placental histological findings associated with tissue damage and deterioration. The alteration of the placental structure was mainly associated with the decrease of villi and the number of capillaries per villi, as well as with the increase of ischemic degenerative lesions associated with calcifications and infarction. These results could be correlated with low birth weight and low hemoglobin levels.