Malaria in pregnancy alters l-arginine bioavailability and placental vascular development

Reducing adverse birth outcomes due to malaria in pregnancy (MIP) is a global health priority. However, there are few safe and effective interventions. l-Arginine is an essential amino acid in pregnancy and an immediate precursor in the biosynthesis of nitric oxide (NO), but there are limited data on the impact of MIP on NO biogenesis. We hypothesized that hypoarginemia contributes to the pathophysiology of MIP and that l-arginine supplementation would improve birth outcomes. In a prospective study of pregnant Malawian women, we show that MIP was associated with lower concentrations of l-arginine and higher concentrations of endogenous inhibitors of NO biosynthesis, asymmetric and symmetric dimethylarginine, which were associated with adverse birth outcomes. In a model of experimental MIP, l-arginine supplementation in dams improved birth outcomes (decreased stillbirth and increased birth weight) compared with controls. The mechanism of action was via normalized angiogenic pathways and enhanced placental vascular development, as visualized by placental microcomputerized tomography imaging. These data define a role for dysregulation of NO biosynthetic pathways in the pathogenesis of MIP and support the evaluation of interventions to enhance l-arginine bioavailability as strategies to improve birth outcomes. The objective of the clinical study was to quantify plasma concentrations of L-arginine, ADMA, and SDMA in a cohort of pregnant women in association with malaria infection. Samples were collected as part of a multisite, open-label, two-arm, randomized superiority trial in southern Malawi (Pan African Clinical Trials Registry PACTR20110300280319 and ISRCTN Registry ISRCTN69800930), which took place between 2011 and 2013, as previously described (40). Briefly, eligibility criteria included HIV-negative women with an estimated gestational age between 16 and 28 weeks of gestation by ultrasound, last menstrual period (LMP), or both; hemoglobin >7 g/dl at baseline; a willingness to deliver in hospital; and not having received a dose of SP in pregnancy. Women were randomized to receive one of the following over the second and third trimester of pregnancy: (i) three or four doses of SP (IPTp-SP) or (ii) screening with malaria rapid diagnostic tests (RDT) (First Response Malaria pLDH/ HRP-2 Combo Test, Premier Medical Corporation Ltd.) and treatment of RDT-positive women with a standard 3-day course of DP (ISTp-DP; 40 mg/320 mg of tablets; Eurartesim, Sigma-Tau). We randomly selected 384 primigravidae for the assessment of L-arginine, SDMA, and ADMA provided they met the following inclusion criteria: live birth with known birth weight and singleton delivery. Of the 384 women included, 379 had an enrollment sample tested and 94 had multiple samples tested over pregnancy for longitudinal assessment of L-arginine, SDMA, and ADMA. Written informed consent was obtained for all study participants. This study was reviewed and approved by the Liverpool School of Tropical Medicine, the Malawian National Health Science Research Committee, and the University Health Network Research Ethics Committee. Our primary endpoint for the human cohort study was the association between the arginine pathway and adverse birth outcomes in primigravidae. Using pilot data from the enrollment visit, we estimated a sample size of 323 women, assuming a mean difference in ADMA of 8 ng/ml and an SD of 19, with 20% of women expected to have an adverse birth outcome (β = 0.80, α = 0.05). In case the data were not normally distributed, we adjusted our sample size upward by 15% to generate a final minimum sample size of 372 women. EDTA plasma samples were tested for L-arginine, ADMA, or SDMA using high-pressure liquid chromatography electrospray tandem mass spectrometry, as described below. The coefficients of variation for arginine testing were 5.2% for L-arginine, 2.0% for SDMA, and 1.4% for ADMA. Concentrations of L-arginine, ADMA, and SDMA were quantified as nanograms per milliliter, and the ratios are expressed as L-arginine/ADMA, L-arginine/SDMA, and ADMA/SDMA (63, 64). All samples were analyzed blinded to the malaria infection status of the participants. For the human study, relative risk was calculated using a log-binomial model, including all variables with P < 0.20 by bivariate analysis. In addition, treatment arm, maternal age, and malaria status at enrollment (by microscopy) were included in the model. To compare the association between markers of NO biosynthesis and nutritional status (maternal BMI, MUAC, and hemoglobin), we used linear regression, adjusting for maternal age and gestational age at enrollment. For longitudinal analysis, we used linear mixed-effects modeling with the lme4 (65) package in R (66) to evaluate the relationship between longitudinal ADMA concentrations and the SGA outcome. We first constructed a null model with six fixed effects: the linear effect of gestational age, maternal age, enrollment BMI, enrollment malaria status, socioeconomic status, and the interaction between gestational age and treatment arm. This interaction term adjusted for the possibility that the rate of change of ADMA was affected by either treatment. Using likelihood ratio tests, we then assessed whether adding SGA as a fixed effect significantly improved the model fit, followed by adding the interaction between SGA and gestational age (table S1). For random effects, all models included a by-participant intercept and a by-participant slope for the effect of gestational age. Biomarker concentrations were transformed using the natural logarithm to stabilize their variance. No deviation from homoscedasticity or normality was apparent on the residual plots. Similarly, but without adjusting for other covariates, linear mixed-effects (LME) models were used to assess the relationship between malaria status at enrollment (by microscopy and PCR) and gestational changes in ADMA, SDMA, and L-arginine concentrations. The objectives of the studies using the EMIP model were to examine the impact of L-arginine supplementation on in utero development (viability and weight) in malaria-infected dams, as well as the impact of L-arginine supplementation on placental vascular development. The EMIP model used in this study is a validated murine model of MIP, which replicates key pathogenic factors of human MIP (41). Female wild-type BALB/c mice between 6 and 8 weeks of age were mated with male wild-type BALB/c mice (8 to 9 weeks of age, obtained from the Jackson Laboratory). Naturally mated pregnant mice were infected on G13 with 106 P. berghei ANKA (PbA)–infected erythrocytes in RPMI 1640 (Gibco) via injection into the lateral tail vein. Control pregnant females were injected on G13 with RPMI 1640 alone. Thin blood smears were taken daily and stained with Giemsa stain (Protocol Hema3 Stain Set, Sigma-Aldrich) to monitor parasitemia. Investigators were not blinded to the experimental group during treatment because the investigators had to prepare the inoculum and L-arginine–supplemented water. However, investigators were blinded during sample processing and outcome assessment, including tissue collection (G19, assessment of weight and viability), processing of samples [placental tissue for reverse transcription PCR (RT-PCR) and serum for mass spectrometry], and assessment of vascular development by micro-CT. All experimental protocols were approved by the University Health Network Animal Care Committee and performed in accordance with current institutional regulations. On the day of pairing, mice were randomly assigned to one of the following treatment groups: (i) vehicle control (regular drinking water) or (ii) 1.2% L-arginine in drinking water (L-arginine monohydrochloride A6969, Sigma-Aldrich). Mice received L-arginine–supplemented drinking water (or vehicle control) beginning before pregnancy and a minimum of 13 days before malaria infection (depending on what day they became pregnant after pairing). A dose of 1.2% was selected because it represents about twice the daily intake of L-arginine in regular chow (50 mg/day, assuming a daily intake of 3 to 5 g of chow with 1% L-arginine), based on the assumption that mice drink 5 to 6 ml of water per day (60 mg/day intake via supplemented water). There was no difference in the daily intake of water between dams receiving the vehicle control and L-arginine–supplemented water at a dose of 1.2% L-arginine. All mice received treatment via ad libitum access to bottled drinking water throughout pregnancy. All supplementation treatments were given in autoclaved water and water bottles. Dams that received L-arginine–deficient chow were placed on a diet of exclusively deficient chow (Harlan Laboratories) beginning at G9 (confirmation of pregnancy) until tissue collection. Dams were kept on their regular chow (Harlan Teklad) diet until this time (G9) to minimize disruptions to their environment (change in diet) during pairing and early pregnancy. Mice were assigned to the treatment groups, as defined above. The EMIP model followed the protocol outlined above. Dams were sacrificed at G19 using carbon dioxide inhalation, yolk sacs were dissected from uteri, fetuses were removed and weighed, and placentas were snap-frozen and stored at −80°C until analysis. Fetal viability was determined by assessing pedal withdrawal reflex. Nonviable fetuses (lacking the pedal withdrawal reflex) were considered stillbirths. All fetuses were weighed at this time. RNA extraction was performed on snap-frozen fetal placenta tissue collected at G19. Serum from mice was collected from cardiac punch and stored at −80°C until analysis. Only placentas collected from viable fetuses were used in the transcript analysis. Tissue was homogenized in TRIzol (0.5 ml/100 mg tissue; Invitrogen) according to the manufacturer’s protocol, and RNA was extracted. Extracted RNA (2 µg per sample) was then treated with DNase (deoxyribonuclease) I (Ambion) and reverse-transcribed to complementary DNA (cDNA) with SuperScript III (Invitrogen) in the presence of oligo(dT)18 primers (primer sequences in table S5) (Fermentas). Residual RNA was degraded with RNase (ribonuclease) H (Invitrogen). Sample cDNA was amplified in triplicate with SYBR Green Master mix (Roche) in the presence of forward and reverse primers (1 µ M both) in a Light Cycler 480 (Roche). Transcript number was calculated on the basis of Ct (cycle threshold) compared to the standard curve of mouse genomic DNA included on each plate by Light Cycler 480 software (Roche) and was normalized to the geometric average of the expression of the housekeeping genes Gapdh and Hrpt. Concentrations of L-arginine, ADMA, and SDMA were assayed by mass spectrometry, as previously described (67). Briefly, the chromatographic conditions included a 125 × 3 mm Nucleosil 100-5 silica column with a 4 × 2 mm silica filter insert. Mobile phase A consisted of 1 liter of water mixed with 0.25 ml of trifluoroacetic acid and 10 ml of propionic acid. Mobile phase B consisted of 1 liter of acetonitrile mixed with 0.25 ml of trifluoroacetic acid and 10 ml of propionic acid. Isocratic elution with one part mobile phase A and nine parts mobile phase B was delivered at a flow rate of 0.5 ml/min at a temperature of 30°C. Samples were prepared with 60 µl of serum and 20 µl of the respective internal standard. Samples (10 µl) were injected automatically, and the electrospray ion source run time duration was 3 to 6.5 min under the following conditions: 32 (arbitrary units); auxiliary gas, 20 (arbitrary units); needle voltage, +4.5 kV; capillary temperature, 300°C. Detailed methods for preparing the fetoplacental vasculature for micro-CT imaging have been described previously (68). Briefly, uteri were extracted from dams at G18 and anesthetized via hypothermia [immersion in ice-cold phosphate-buffered saline (PBS)]. Each individual fetus was then extracted from the uterus while maintaining the vascular connection to the placenta. The embryo was briefly resuscitated via immersion in warm PBS to resume blood circulation. Embryos that could not be resuscitated were not perfused and were removed from the study. A catheter was then inserted into the umbilical artery, and the fetus was perfused with saline [with heparin (100 U/ml)], followed by radiopaque silicone rubber contrast agent (Microfil, Flow Technology). After perfusion, specimens were postfixed with 10% formalin and imaged using micro-CT. Specimens were scanned at 7.1 µm resolution for 1 hour using a Bruker SkyScan 1172 high-resolution micro-CT scanner. A total of 996 views were acquired via 180° rotation with an x-ray source at 54 kVp (kilovolt peak) and 185 µA. Three-dimensional micro-CT data were reconstructed using SkyScan NRecon software. The structure of the vasculature was identified automatically using a segmentation algorithm, as previously described in detail (69). The leaves of the vascular tree were pruned to 0.035 mm (threshold diameter) to improve data consistency. Analysis was performed on wild-type [unexposed (n = 7) and malaria-exposed (n = 8)] offspring of control (nonsupplemented) dams and unexposed (n = 7) and malaria-exposed (n = 7) offspring of L-arginine–supplemented dams. Each group contained a minimum of three dams per group and one to three specimens per litter. Statistical analysis was performed using Stata v14 (StataCorp), R v3.2.1 (R Core Team, 2015, R Foundation for Statistical Computing), and GraphPad Prism v6 (GraphPad Software Inc.). Student’s t test, oneway analysis of variance (ANOVA) (nonparametric Kruskal-Wallis, P < 0.05), post test (Tukey test), independent samples t test, χ2 test, and relative risk were used to examine the statistical significance of differences between experimental groups. Analysis of the cumulative distribution of vessel diameters for each placenta was fit with a natural spline with eight degrees of freedom. A two-way ANOVA was conducted to determine whether there was an effect of treatment group on the spline parameters. There was a significant interaction between spline coefficient and group (P < 0.001), and therefore, a post hoc analysis was performed to compare pairs of treatment groups. Post tests on all groups were conducted using Dunn’s multiple comparison test (P < 0.05).