Objective
Intrauterine growth restriction (IUGR) is associated with increased inflammatory responses. We sought to investigate whether magnesium (Mg) attenuates inflammation and IUGR in a rat model.
Study design
Pregnant Wistar rats (12 weeks, gestational day 18) were randomly assigned to 1 of 4 groups: normal diet with bilateral uterine artery ligation (BL) (n = 6) or sham surgery (SH) (n = 5); and Mg chloride (MgCl 2 ) 1% (wt/vol) in the drinking water throughout gestation + BL (MgBL) (n = 6) or SH (MgSH) (n = 5). Dams were euthanized 24 hours postsurgery (gestational day 19). Maternal plasma, fetal plasma (pooled), individual amniotic fluid (AF) samples, and placentas (PL) were collected and assessed from live fetal pups only (BL, n = 36; SH, n = 20; MgBL, n = 20; MgSH, n = 20). All samples were analyzed for cytokines/chemokines (interleukin [IL]-6, IL-1β, chemokine [C-X-C motif] ligand 1 [CXCL1], chemokine [C-C motif] ligand 2 [CCL2], and tumor necrosis factor [TNF-α] sensitivity <3 pg/mL) using a multiplex platform. Data were analyzed using Mann Whitney, analysis of variance, and Fisher exact tests.
Results
The incidence of IUGR (pup weight <10th percentile of SH) in the MgBL group was significantly lower (31%) than the BL group (86.3%) (relative risk, 0.36; 95% confidence interval, 0.2–0.6; P < .0001). BL significantly increased AF levels of IL-6, IL-1β, TNF-α ( P < .05), and CCL2 ( P < .001) vs SH and PL levels of IL-6, IL-1β, CCL2 and CXCL1 ( P < .001), and TNF-α ( P < .05) vs SH. Maternal MgCl 2 supplementation significantly decreased IL-1β, TNF-α, and CCL2 levels in AF and IL-1β in PL tissues of MgBL vs BL rats ( P < .0001).
Conclusion
Maternal oral MgCl 2 supplementation reduced BL-induced IUGR by 64% and suppressed cytokine/chemokine levels in the AF and PL.
Intrauterine growth restriction (IUGR), is defined as fetal weight <10th percentile of a given population at the same gestational age. Approximately 25-30% of all nonanomalous IUGR fetuses are the result of placental (PL) insufficiency. IUGR is associated with increased preterm birth, fetal and neonatal death, low Apgar scores, hypoxemia, acidosis at birth, sepsis, intracranial hemorrhage, necrotizing enterocolitis, and respiratory complications. In addition, IUGR impacts fetal programming and is associated with numerous long-term health sequelae, including cognitive disabilities, schizophrenia, cerebral palsy, bronchopulmonary dysplasia, and an increased risk for cardiovascular, metabolic, and renal diseases in adulthood. To date there are no antenatal treatments that improve neonatal outcomes once IUGR is diagnosed. The only current recommendation is close antenatal surveillance to determine the optimal timing of delivery.
The excessive production of inflammatory cytokines and chemokines during hypoxic events may lead to tissue injury. Elevated cytokine levels have been found in both the amniotic fluid (AF) and umbilical cord serum (interleukin [IL]-6, IL-1β, and tumor necrosis factor [TNF-α]) and increased PL cytokine messenger RNA expression (IL-8, IL-6, interferon-γ, and TNF-α) has been observed among small-for-gestational-age (SGA) neonates. These inflammatory mediators are believed to contribute, in part, to the short- and long-term adverse consequences of IUGR.
Magnesium (Mg) is reported to have a neuroprotective effect on preterm fetuses. A randomized clinical trial by Rouse et al showed a significantly lower rate of cerebral palsy among infants born to women at 24-32 weeks’ gestation who received Mg sulfate (MgSO 4 ) before delivery. A more recent metaanalysis confirmed the neuroprotective effects of maternal MgSO 4 administration. While the exact mechanism for the beneficial effect of Mg is not known, numerous studies support its antiinflammatory and vasodilatory activities. Research in our laboratory revealed that MgSO 4 attenuates cytokine production by lipopolysaccharide (LPS)-stimulated endothelial cells and PL explants. Further studies by our group and by Burd et al demonstrated the potent antiinflammatory and fetal neuroprotective effects of maternal MgSO 4 using rodent models of maternal infection. The vasodilatory effect of Mg has been investigated in several vessels. Both in vivo and in vitro studies have shown vasorelaxation on rat aortic rings, mesenteric vessels and cerebral microcirculation, human PL, and uterine arteries (UAs) after MgSO 4 exposure.
Animal models of IUGR by UA ligation have been used to study IUGR-related metabolic, behavioral, and neurological conditions. While little is known regarding the maternal and fetal inflammatory responses following UA ligation, we hypothesized that the production of numerous proinflammatory mediators would ensue within the maternal and fetal compartments following ligation. Furthermore, given the antiinflammatory and vasodilatory activities of Mg, we examined the effects of maternal oral Mg supplementation on maternal and fetal inflammation, as well as on fetal and PL weights using a rat model of IUGR induced following bilateral UA ligation (BL).
Materials and Methods
Animal
The Institutional Animal Care and Use Committee approved all animal studies prior to animal experimentation. Female Wistar rats (12-15 weeks old) (Taconic Farms, Germantown, NY) were initially acclimatized under normal environmental conditions and allowed free access to standard rat chow and tap water for at least 72 hours prior to mating with normal male Wistar rats (Taconic Farms). Pregnancy was assessed by a positive vaginal plug and/or sperm in vaginal lavage fluids. On gestational day (GD) 1, dams were randomly assigned to 1 of 4 treatment groups: normal chow/tap water with BL (n = 6) or sham surgery (SH) (n = 5); and normal chow/tap water containing Mg chloride (MgCl 2 ) 1% (wt/vol) + BL (MgBL) (n = 6) or SH (MgSH) (n = 5). Dams were individually housed and normal chow and tap water ±1% MgCl 2 were provided throughout gestation. Oral MgCl 2 supplementation given early and throughout pregnancy was chosen to mimic human prophylactic therapy. MgCl 2, was administered at 1% (wt/vol) because in a pilot trial this was the best tolerated dose (ie, it did not cause diarrhea and resulted in maternal weight gain and pup weights that were similar to dams given normal tap water). Rats were weighed immediately prior to mating and then on GD1, GD5, GD11, GD18, and GD19. On GD18 dams were anesthetized using inhaled isoflurane and underwent laparotomy for the BL procedure under sterile conditions. The uterine horns were exteriorized and the numbers of pups were determined; the UAs were identified and sterile steel wire (0.1-mm diameter) was placed parallel along the top of the both UAs to achieve consistent arterial occlusion upon ligation using 3-0 Vicryl suture material. Following occlusion, the wires were gently removed and the abdominal wound was closed with 3-0 Vicryl. Sham rats underwent the same surgical procedures without ligation of the UAs. Rats recovered within a few minutes and had ad libitum access to food and water (±MgCl 2 ). Dams were euthanized 24 hours after the surgical procedure by carbon-dioxide inhalation followed by exsanguination via cardiac puncture using heparinized needles/syringes. Fetuses, delivered by cesarean section, were euthanized by decapitation and fetal blood was collected into heparinized capillary tubes. PL weights, litter number, fetal weight, and fetal mortality were assessed. PL and fetal samples were collected from live fetuses only. Maternal blood and fetal blood were centrifuged to obtain plasma. Maternal plasma (MP) and fetal plasma (FP) (pooled), AF (from individual sacs), and PL samples were flash frozen in liquid nitrogen and stored at –80°C until processing for analysis of inflammatory mediators.
Preparation of biological samples and assessment of inflammatory mediators
After an initial centrifugation step, MP, FP, and AF were analyzed for inflammatory mediators using a multiplex platform (described below). Frozen PL samples were homogenized using freshly prepared lysis buffer (150 mmol/L sodium chloride [NaCl], 20 mmol/L Tris [pH 7.5], 0.25% Triton X-100, 10 mmol/L sodium fluoride [NaF], and phosphatase/protease inhibitor cocktail [Thermo Scientific, Waltham, MA]). After centrifuging, cell-free homogenates were analyzed for inflammatory mediators using a multiplex platform (described below). Inflammatory mediator levels in the maternal and fetal samples were determined using customized rat cytokine 5-plex kits for TNF-α, IL-6, chemokine (C-X-C motif) ligand 1 (CXCL1), IL-1β, and chemokine (C-C motif) ligand 2 (CCL2), with sensitivity <3 pg/mL (purchased from Meso Scale Discovery, Rockville, MD), according to the manufacturer’s directions. The Meso Scale Discovery plate was analyzed using the Meso Scale Discovery Sector Imager 2400 plate reader. The raw data were measured as electrochemiluminescence signals detected by photodetectors and analyzed using Discovery Workbench 3.0 software (Meso Scale Discovery). A 4-parameter logistic fit curve was generated for each cytokine using the standards and used to determine the concentration of the individual cytokines in each sample. PL cytokine data were normalized for protein concentrations (as determined using the Bradford method, Bio-Rad Laboratories, Hercules, CA). For analyses, fetal samples from the BL and MgBL groups (except fetal blood, which was pooled) were selected from fetal pups weighing ≤10th percentile of the SH group; SH and MgSH samples were selected from fetal pups weighing >10th percentile. A larger number of AF and PL samples were evaluated in the BL group (n = 36) to assess the impact of fetal position within the uterine horn (with respect to the BL procedure) and proximity to fetal demise.
Assessment of MP and AF Mg 2+ levels
MP and pooled AF samples (for each dam) were assayed for total Mg 2+ levels by the Core Laboratories of the North Shore-LIJ Health System, Manhasset, NY.
Statistical analysis
Maternal and fetal outcome data (weight gain, fetal weight, PL weight, and Mg levels), presented as means and SD, were compared using the Kruskal-Wallis test (nonparametric analysis of variance) with post hoc testing. Fetal mortality and incidence of IUGR (pups <10th percentile of normal SH pups), presented as rates, were analyzed using the Fisher exact test. Cytokine data are shown as means and SEM and analyzed with Student t test. Maternal and fetal cytokine data were analyzed first by comparing the SH vs BL groups. Subsequently, to test the efficacy of the MgCl 2 treatment, BL cytokine data were compared to the MgBL group. Statistical significance was set at P < .05.
Results
Maternal Mg supplementation in an IUGR model improves fetal pup weight
Dams given normal drinking water and those given MgCl 2 1% (wt/vol) in their drinking water throughout gestation (GD1-GD18) gained similar weights (83.54 ± 13 g vs 79.26 ± 14.48 g, respectively; P = .52). There were no significant differences in litter sizes among the 4 groups (11.25 ± 1.7 [SH], 11.8 ± 1.3 [MgSH], 12.1 ± 1.7 [BL}, and 12 ± 2.3 [MgBL]; P = .6). The average fetal weight among BL dams was significantly lower than the SH dams (0.93 ± 0.2 g [BL] vs 1.22 ± 0.08 g [SH], P < .001) ( Table 1 ). Likewise, PL weights obtained from the BL dams were significantly lower than those obtained from the SH dams (0.32 ± 0.05 g [BL] vs 0.42 ± 0.08 g [SH], P < .001) ( Table 1 ). By contrast, there were no differences in fetal pup weights (1.24 ± 0.27 g vs 1.20 ± 0.06 g, P = .8) or PL weights (0.35 ± 0.04 g vs 0.36 ± 0.05 g, P = .4) between the MgBL vs MgSH groups, respectively ( Table 1 ). Overall fetal mortality was not significantly different between the BL vs MgBL groups (31/82 [38%] vs 32/74 [43%], respectively) and no fetal mortality was observed in either the SH (0/56 [0%]) or MgSH (0/60 [0%]) groups ( Table 1 ). The position of the fetal pups within the uterine horn had no impact on fetal or PL weight following BL or SH procedures. The 10th percentile for fetal pup weight among the control (SH) group was ≤1.127 g. The incidence of IUGR (fetal pup weight ≤10th percentile of the control group) among the BL vs MgBL groups was 86.3% vs 31%, respectively (relative risk, 0.36; confidence interval, 0.2–0.6; P < .0001]). Total Mg 2+ concentrations, according to 2 methods, found in the MP and AF were similar among all groups ( Table 2 ).
| Variable | SH, n = 56 | BL, n = 51 | MSH, n = 60 | MBL, n = 42 | P value ANOVA A a : SH vs BL B a : BL vs MBL |
|---|---|---|---|---|---|
| Pups, g | 1.22 ± 0.08 | 0.933 ± 0.2 | 1.2 ± 0.06 | 1.24 ± 0.27 | A: < .001 B: < .001 |
| Placenta, g | 0.42 ± 0.08 | 0.32 ± 0.05 | 0.35 ± 0.04 | 0.36 ± 0.05 | A: < .001 B: < .01 |
| Fetal mortality | 0/56 (0%) | 31/82 (38%) | 0/60 (0%) | 32/74 (43%) | A: .52 b |
a Kruskal-Wallis (nonparametric ANOVA) with post hoc test.
| Variable | SH, n = 5 | BL, n = 6 | MSH, n = 5 | MBL, n = 6 | P value ANOVA |
|---|---|---|---|---|---|
| Maternal serum, mg/dL | 2.7 ± 0.25 | 2.58 ± 0.24 | 2.64 ± 0.18 | 2.76 ± 0.56 | .6223 |
| Amniotic fluid, mg/dL | 3.02 ± 0.13 | 3.08 ± 0.23 | 2.92 ± 0.24 | 3.18 ± 0.23 | .1637 |
Maternal Mg supplementation blocks BL-induced inflammatory mediator levels in the AF and IL-1β in the PL
First, we examined the effect of BL on inflammatory mediator levels in the MP and FP 24 hours postsurgery. Interestingly, we found that even though IL-1β, IL-6, TNF-α, CCL2, and CXCL1 levels in the MP and FP levels were not significantly different between the SH vs BL groups, all inflammatory mediator levels (with the exception of IL-6) were significantly decreased by maternal Mg supplementation ( Figure 1 ). In contrast, IL-6 levels in the FP and PL were significantly higher in the MgBL group compared to SH and MgSH groups.
Next, we examined the effect of BL (vs SH) on inflammatory mediator production in the AF and PL 24 hours postsurgery. Levels of IL-6, IL-1β, TNF-α, CCL2, and CXCL1 were significantly increased in both the BL-AF and BL-PL tissues when compared to the SH-AF and SH-PL tissues, respectively ( Figure 2 ). Maternal oral Mg supplementation significantly decreased BL-induced IL-1β, TNF-α, and CCL2 ( P < .001) levels in the AF and significantly reduced BL-induced IL-1β levels in the PL tissues ( P < .001) ( Figure 2 ). IL-1β, TNF-α, and CCL2 levels in the AF and IL-1β, TNF-α, and CXCL1 levels in the PL tissues of the MgBL group were not significantly different when compared with AF and PL tissues obtained from the MgSH group, respectively ( Figure 2 ). IL-6 and CXCL1 levels were decreased in the AF samples obtained from the MgBL group when compared to the BL group; however, these differences were not statistically significant ( Figure 2 ). Fetal position within the uterine horn and proximity to fetal demise had no impact on inflammatory mediator levels in the various compartments (data not shown).
