The purpose of this study was to investigate whether magnesium sulfate (MgSO 4 ) prevents fetal brain injury in inflammation-associated preterm birth (PTB).
Using a mouse model of PTB, lipopolysaccharide (LPS) or normal saline solution (NS)–exposed mice were randomized to intraperitoneal treatment with MgSO 4 or NS by intrauterine injection. From the 4 treatment groups (NS + NS; LPS + NS; LPS + MgSO 4 ; and NS + MgSO 4 ), fetal brains were collected for quantitative polymerase chain reaction studies and primary neuronal cultures. Messenger RNA expression of cytokines, cell death, and markers of neuronal and glial differentiation were assessed. Immunocytochemistry and confocal microscopy were performed.
There was no difference between the LPS + NS and LPS + MgSO 4 groups in the expression of proinflammatory cytokines, cell death markers, and markers of prooligodendrocyte and astrocyte development ( P > .05 for all). Neuronal cultures from the LPS + NS group demonstrated morphologic changes; this neuronal injury was prevented by MgSO 4 ( P < .001).
Amelioration of neuronal injury in inflammation-associated PTB may be a key mechanism by which MgSO 4 prevents cerebral palsy.
Cerebral palsy (CP) is a nonprogressive motor impairment syndrome that occurs in 1-3.6 of 1000 live births. Almost 8% of ex-preterm children who were born at <28 weeks of gestation are affected by CP. Despite the advances in perinatology and neonatology and the dramatic reduction in the mortality rate of high-risk infants, there has been no reduction in the prevalence of CP.
See related editorial, page 207
Magnesium sulfate (MgSO 4 ) has been investigated in several clinical trials and in systemic reviews and metaanalyses as a possible therapeutic agent to reduce CP in “at-risk“ fetuses. In most of these trials, at-risk fetuses were those likely to be born preterm from spontaneous preterm birth, which is a condition that frequently is associated with intrauterine inflammation. Although these studies demonstrated that MgSO 4 may prevent CP at 2 years of life, there have been some concerns raised with the use of antenatal MgSO 4 . Although a metaanalysis did not show an increase in neonatal death associated with MgSO 4 , there still has been some concern raised regarding a possible increased risk of neonatal death from MgSO 4 . Furthermore, in an in vivo model, maternal administration of high doses of MgSO 4 was shown to lead to cell death in a developing mouse brain.
Based on the latest Cochrane review, the neuroprotective role for antenatal MgSO 4 that was given to women who were at risk of preterm birth for the preterm fetus has now been established. Despite these recommendations, the mechanism by which MgSO 4 serves as a neuroprotective agent in the preterm brain has not been elucidated.
From models of hypoxic-ischemic (HI) and traumatic brain injury, the protective effect of MgSO 4 is believed to be through its action as a noncompetitive antagonist of the N-methyl-D-aspartic acid receptor. However, other animal data suggest that MgSO 4 may serve an antiapoptotic role and prevent neuronal cell loss. To date, there are no animal trials that have investigated the use of MgSO 4 as a neuroprotective agent in the setting of prenatal inflammation.
These studies sought to determine whether MgSO 4 that is administered to the mother can prevent fetal brain injury as a possible mechanism by which MgSO 4 appears to be neuroprotective in human clinical trials of preterm infants. Using a mouse, we have demonstrated that intrauterine inflammation results in a cytokine response in the fetal brain, in white matter damage (WMD), and in neuronal injury. Using this established model, we investigated the ability of MgSO 4 , when administered to the mother, to prevent fetal brain injury. The objectives of these studies were to investigate (1) whether the administration of MgSO 4 altered the proinflammatory response in fetal brain, (2) whether the administration of MgSO 4 altered apoptotic or necrotic pathways in the fetal brain, and, most importantly, (3) whether MgSO 4 , when administered in vivo, could prevent fetal brain injury and, specifically, neuronal injury.
Materials and Methods
Mouse model of intrauterine inflammation
CD-1 out-bred, timed pregnant mice (Charles River Laboratories, Wilmington, MA) were used in an established model of inflammation-induced preterm birth. Because approximately 85% of spontaneous preterm births at <28 weeks of gestation are associated with intrauterine inflammation, as demonstrated by the presence of histologic chorioamnionitis, this mouse model aptly mimics this common clinical scenario that occurs in many cases of spontaneous preterm birth. Furthermore, using this model of preterm birth, we demonstrated that, before preterm delivery, exposure to intrauterine inflammation results in fetal brain injury; thus, this model is also useful for the assessment of interventions that might ameliorate or reduce adverse neonatal outcomes from inflammation-associated preterm birth. Survival surgery and intrauterine injections of lipopolysaccharide (LPS, 250 μg/dam in 100 μL phosphate-buffered saline solution; from Escherichia coli , 055:B5; Sigma Chemical Co, St. Louis, MO) were performed on embryonic day 15 of gestation (term is 19 days) as previously reported. Briefly, anesthesia was obtained by a continuous flow of isofluorane/oxygen, supplied by a mask that fits over the mouse’s head. After deep anesthesia was reached, a minilaparotomy was performed in the lower abdomen. The right uterine horn was identified and LPS or saline solution was infused into the uterus between the first and the second gestational sacs. Routine closure was performed, and the dams recovered in 3-5 minutes. Dams were killed humanely 6 hours after surgery with carbon dioxide. Three dams were used for each treatment group. Immediately after death, 4 fetuses per dam were taken from lower uterine horns; as such, all fetuses from all dams were in the same proximity to location at which the LPS was infused. Fetal brains were collected for messenger RNA (mRNA) studies and for primary cortical neuronal cultures. Guidelines for the care and use of animals were approved by the University of Pennsylvania.
After intrauterine infusion of LPS or normal saline solution (NS; as previously described), dams were randomized to intraperitoneal treatment with MgSO 4. The maternal MgSO 4 injection protocol involved an intraperitoneal dose of 270 mg/kg followed by 27 mg/kg every 20 minutes for 4 hours; injections were given in a volume of 0.1 mL. A second dose of 270 mg/kg was given at the end of the 4-hour period. An average dam weight at embryonic day 15 is 40 g. Control mice were injected with same volume of NS and at the same timing schedule. The selected protocol followed the procedure of Hallak et al. A previous report that used this protocol in mice demonstrated that, 30 minutes after the injection of MgSO 4, the magnesium values in the mothers’ blood samples were approximately double the normal values. The use of this protocol in rats resulted in a 125% increase in magnesium level in the fetal forebrain after 4 hours. This level of MgSO 4 was demonstrated to prevent fetal brain damage that is associated with HI brain injury in a rodent model. Hence, with this protocol, the following 4 treatment groups were compared in these studies: (1) NS and NS (negative control), (2) LPS and NS (positive control), (3) LPS and MgSO 4 , and (4) NS and MgSO 4 . Three dams from each treatment group were used for these experiments; from each dam, 4 fetal brains were used.
Primary cortical neuronal cultures
With a sterile technique, embryonic day 15 fetal brains were harvested 4.5-6 hours after the intrauterine randomization and placed into Petri dishes that contained cold Ca ++ /Mg ++ -free Hanks Balanced Salt Solution (Invitrogen, Carlsbad, CA), pH 7.4. The cortex, which is a part of the fetal brain, was separated from meninges, olfactory bulbs, brain stem, and cerebellum. Each cortex was minced, placed in 4 mL neurobasal medium (Invitrogen) that contained 0.03% trypsin (Invitrogen), and incubated for 15 minutes at 37°C and 5% carbon dioxide. Brain tissue was removed and placed in 4.5 mL neurobasal medium that contained 10% fetal bovine serum and allowed to settle to inactivate the trypsin. The medium was decanted and replaced with neurobasal medium that was supplemented with B-27 vitamin (Invitrogen) and 0.5 mmol/L L-glutamine, and the cells were dissociated by trituration. This media combination, neurobasal medium in the absence of fetal bovine serum, allows for the select growth of neurons and not glia (astrocytes or microglia). Cells were plated at low density (10 4 cells/mL) on poly-L-lysine (1 mg/mL; Sigma-Aldrich, St. Louis, MO)–coated glass coverslips on 12-well culture plates. Twelve fetal brains from 3 dams per treatment group (4 fetal brains per dam) were used for the analysis of neuronal morphologic condition per each treatment group. Cells were plated to equal density for each experiment. All experiments were performed in triplicate to ensure the consistency of the results.
Cortical cell cultures were fixed and stained at days in vitro 3 and 10 to assess morphologic changes between the treatment groups, with the use of double immunofluorescence as previously reported. A mouse monoclonal antibody to microtubule-associated protein 2 (MAP2; Sigma-Aldrich) was used to identify dendrites and cell bodies at a dilution of 1:100. A rabbit polyclonal antibody to 200 kd neurofilament protein (NF-200; Sigma-Aldrich) was used to label the entire cell at a dilution of 1:400. Confocal microscopy (Leica SP2 Confocal; Leica Microsystems Inc, Bannockburn, IL) was used for the morphologic evaluation of the neurons.
Quantitative analysis of dendritic processes from cortical cultures experiments
Dendritic processes that emanated from neuronal cell body were analyzed at day in vitro 3 with the use of previously described techniques. Briefly, cells were selected at random with at least 3 coverslips for each condition (3 different cultures or 3 different dams). At least 3 experiments were performed for the condition. To quantify processes that emanated from each cell body, 30 neurons from each treatment group were evaluated at a final image magnification of ×400. Individual neurons were selected if they were clearly defined and not overlapping with other neurons. Fluorescent images were recorded and analyzed with a computer (Latitude D620; Dell Inc, Round Rock, TX), with an image-processing program (Image J 1.37v; National Institutes of Health, Bethesda, MD).
Quantitative polymerase chain reaction for the expression of proinflammatory cytokines, neuronal and glial differentiation markers, and cell death–associated genes
Whole embryonic day 15 fetal brains were harvested after the intrauterine randomization for mRNA, by the Trizol method (Invitrogen). Fetal brains from each dam were pooled; RNA was extracted, and complementary DNA was created as per protocol. Quantitative polymerase chain reaction was performed as previously reported for the evaluation of (1) proinflammatory cytokines (interleukin [IL]-1β, IL-6, and tumor necrosis factor [TNF]-α); (2) neuronal differentiation markers (MAP2 and nestin); (3) markers of WMD, glial fibrillary acidic protein, and prooligodendrocyte marker; and (4) cell death-associated genes (caspase 1, 3, 8, and 9). Twelve fetal brains from 3 dams per treatment group (4 fetal brains per dam) were used for the analysis and the comparison of the mRNA expression.
Statistical analyses were performed with the SigmaStat software program (Aspire Software International, Ashburn, VA). For the comparison of the mRNA expression results and the number of dendritic processes between groups, 1-way analysis of variance (ANOVA) or ANOVA on ranks (for nonparametric data) was used. If significance was reached, pair-wise comparison was then performed with Student-Newman-Keuls (SNK) or Dunn methods. One-way ANOVA was used for all of the comparisons of the mRNA expression because the data for these studies were distributed normally. ANOVA on ranks was used for the comparison of the dendritic processes.
Quantitative polymerase chain reaction for expression of proinflammatory cytokines, neuronal and glial differentiation markers, and cell death–associated genes
In whole fetal brains, IL-1β mRNA levels were significantly different among the treatment groups ( P = .008; 1-way ANOVA; Figure 1 ). In LPS + NS–treated fetal brains, IL-1β mRNA was increased 9-fold, compared with NS + NS treatment ( P = .184; SNK). Similarly, IL-1β mRNA levels in the LPS + MgSO 4 group were 23-fold increased, compared with the NS + NS group ( P = .007; SNK), and were 24-fold increased, compared with the NS + MgSO 4 group ( P = .012; SNK. IL-1β mRNA expression was not significantly different between the LPS + NS and LPS + MgSO 4 groups ( P = .06; SNK). The proinflammatory cytokines, IL-6 and TNF-α, were not expressed differentially among the groups ( P > .05; 1-way ANOVA for both; Figure 1 ).