|Home | About | Journals | Submit | Contact Us | Français|
To elucidate possible mechanisms of fetal neuronal injury in inflammation-induced preterm birth.
Utilizing mouse model of preterm birth, primary cultures were prepared from fetal brains: 1) control neurons (CN); 2) LPS-exposed neurons (LN); 3) control co-culture (CCC), consisting of neurons and glia; 4) LPS-exposed co-culture (LCC), consisting of LPS-exposed neurons and glia. CN and LN were treated with culture media from CN, LN, CCC and LCC after 24 hours in vitro. Immunocytochemistry was performed for culture characterization and neuronal morphology. Quantitative PCR was performed for neuronal differentiation marker, MAP2, and for cell death mediators, Caspases 1, 3 and 9.
LPS exposure in vivo did not influence neuronal or glial content in co-cultures but decreased expression of MAP2 in LN. Media from LN and LCC induced morphological changes in control neurons comparable to LN. The neuronal damage caused by in vivo exposure (LN) could not be reversed by media from control groups.
LPS-induced preterm birth may be responsible for irreversible neuronal injury.
Preterm birth has been demonstrated to be a major cause of adverse neurological outcome.1-3 Cerebral palsy (CP), believed to arise from “white matter damage” (WMD) has long been considered the main neurological outcome of concern in ex-preterm infants. However, recent studies demonstrated that ex-preterm children are at risk not only for motor disorders such as CP but also for significant cognitive and behavioral deficits.4-12 WMD, which involves loss of oligodendrocytes and proliferation of astrocytes, may be an insufficient paradigm to explain the increasing prevalence of these adverse outcomes. Corollary to that, we have demonstrated in our laboratory that inflammation-induced preterm birth results in neuronal injury.13 The mechanisms by which intrauterine inflammation in a preterm birth results in neuronal injury have not been investigated.
In many neuroinflammatory diseases, including multiple sclerosis and Alzheimer's disease, a brain inflammatory response results in cellular interactions through the release of cytokines, and of trophic and/or cytotoxic signals.14 Neuronal injury occurs consequent to these neuron-neuron and glial-neuronal interactions.15
The relatively new focus on neuronal injury13, 16 that accompanies the WMD in the fetal brain mandates new goals to identify the mechanisms of neuronal injury in the setting of the inflammation-induced preterm birth. We hypothesize that the mechanisms involved in fetal brain injury in preterm birth are similar to other neuroinflammatory disorders which involve neuron-neuron and glial-neuronal communication. First, we hypothesize that following an inflammatory challenge, injured fetal neurons, and possibly glia, produce sufficient neurotoxic factors that are capable of inducing a neuronal injury in ‘normal’ neurons, thus propagating neuronal injury. Second, as glia produce neurotrophic and neuroprotective factors, we hypothesize that neuronal injury from intrauterine inflammation could be reversed or ameliorated by the normal neuronal and/or glial milieu.
CD-1 out-bred, timed pregnant mice (Charles River Laboratories, Wilmington, MA) were utilized. Guidelines for the care and use of animals were approved by the University of Pennsylvania. Dams were randomized into 2 groups: controls which received no intervention and lipopolysaccharide (LPS)-exposed. Survival surgery and intrauterine injections of LPS were performed on day 15 of gestation (E15) of a 19 day gestation, as previously reported.17-21 With this model, preterm birth occurs >95% of the time, occurring between 8−20 hours after infusion of LPS.17-21
For this model, a continuous isofluorane/oxygen anesthesia was supplied by a mask that fits over the dam's face. After a deep anesthesia was reached, a mini-laparotomy was performed in the lower abdomen. The right uterine horn was isolated and LPS (250 μg/dam in 100 μL PBS; from Escherichia coli, 055:B5, Sigma Chemical Co., St. Louis, MO). Routine closure was performed and the dams were recovered. Dams were humanely euthanized 4−6 hours after surgery by utilizing CO2. Fetuses were taken from the dams immediately after euthanasia and primary cultures were obtained from fetal brains (cortex only) as described in detail below. As we have previously reported that in fetal brains from animals receiving anesthesia, surgery and intrauterine infusion of saline that there is no evidence of neuronal injury, we used control dams for these described experiments.13 In contrast to those experiments, where the use of ‘sham’ animals is an absolute necessity to demonstrate the interventions themselves do not cause neuronal injury, in these experiments, normal neurons—exposed to no manipulation (anesthesia or surgery), are required. For the purpose of these experiments, we are investigating the possible effect of injured neurons and glia on normal neurons and normal glia and neurons on injured neurons. As such, unaffected and untreated neurons and glia are required not shams. Therefore, the following treatment groups were utilized for these experiments: 1) fetal corteces from pups exposed to intrauterine LPS for 6 hours and 2) fetal corteces from gestationally-matched dams receiving no intervention or anesthesia. To assess neuronal morphology, co-culture characterization, and evaluation of mRNA, 9 dams were utilized for the experimental arm and 9 dams served as control.
Using sterile technique, E15 fetal brains were harvested from fetuses by incising and peeling off the calvaria and placed into Petri dishes containing cold Ca++/Mg++-free Hanks Balanced Salt Solution (HBSS; Invitrogen, Carlsbad, CA), pH 7.4. The cortex, 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 (NBM; Invitrogen, Carlsbad, CA) containing 0.03% trypsin (Invitrogen, Carlsbad, CA) and incubated for 15 minutes at 37°C and 5% CO2. Brain tissue was removed and placed in 4.5 ml NBM containing 10% fetal bovine serum (FBS) and allowed to settle to inactivate the trypsin. The medium was decanted and replaced with NBM supplemented with B-27 vitamin (Invitrogen, Carlsbad, CA) and 0.5mM L-glutamine and cells were dissociated by trituration. This media combination, NBM in the absence of fetal bovine serum, allows for the select growth of neurons and not glia (astrocytes or microglia). 22-23 In these media conditions, greater than 95% of the cells are neurons as demonstrated by positive microtubule-associated protein 2 (MAP2) staining, a neuronal somatodendritic specific marker and a marker of neuronal differentiation.24
Cell concentration was determined using 1:10 dilution. Cells were plated at low density (4 × 104 cells/ml) on poly-L-lysine (1 mg/ml; Sigma-Aldrich, St. Louis, MO) coated glass coverslips, using 6- and 12-well culture plates. Groups were plated to equal density for each experiment. For each experiment, 3−4 fetal brains from one dam constituted one culture. For each experimental trial, 3 dams per treatment group were utilized. The experiment was repeated 3 separate times, providing 9 dams per treatment group.
Co-cultures (glial-neuronal cultures) were performed as above with removal of the fetal cortex. Glia is a collective term for astrocytes and microglia. Mature oligodendrocytes are not present at this time of the mouse brain development.
Tissues were then placed in 4.5 ml minimal essential medium (MEM; Invitrogen, Carlsbad, CA) containing 10% fetal bovine serum (FBS) and allowed to settle for trypsin inactivation. The medium was decanted and replaced with MEM supplemented with 10% FBS, B-27 vitamin and 2mM L-glutamine and cells were dissociated by trituration. This media condition allows for the growth of neurons as well as glia (astrocytes and microglia). In contrast to neurons, which may grow in the absence of 10% of FBS, the 10% of FBS is necessary to sustain glial growth in a cell culture.
Cell concentration was determined using 1:10 dilution. Cells were plated at density of 2.5 × 105 cells/ml on poly-L-lysine coated glass coverslips, using 6- and 12-well culture plates. Groups were plated to equal density for each experiment. For each experiment, 3−4 fetal brains from one dam constituted one culture. For each experimental trial, 3 dams per treatment group were utilized. The experiment was repeated 3 separate times, providing 9 dams per treatment group.
Cortical cell cultures were fixed and stained at division days (DD) 3, 10, 14 to assess morphologic changes between the treatment groups, using double immunofluorescence. A mouse monoclonal antibody to Microtubule-associated protein 2 (MAP2; Sigma-Aldrich, St. Louis, MO) was used to identify dendrites and cell bodies at dilution of 1:100. A rabbit polyclonal antibody to 200 kDa Neurofilament protein (NF-200, Sigma-Aldrich, St. Louis, MO) was used to label the entire cell at dilution of 1:400. Glial-neuronal cultures were characterized at DD 7 (when cultures reached confluence) by triple-label immunofluorescence. A rabbit polyclonal antibody to glial fibrillary acidic protein (GFAP; Sigma-Aldrich, St. Louis, MO) was used to identify astrocytes at dilution of 1:200. Neurons were identified with mouse monoclonal antibody to NeuN (Sigma-Aldrich, St. Louis, MO) at a dilution of 1:200. Bio-tinilated A4B5 (Sigma-Aldrich, St. Louis, MO) identified microglia at a dilution of 1:500. Alexa Fluor goat anti-mouse 488 (Invitrogen, Carlsbad, CA) and Alexa Fluor goat anti-rabbit 568 (Invitrogen, Carlsbad, CA) were used for immunofluorescence at 1:500 dilution. Confocal microscopy (Leica SP2 Confocal) was utilized for the morphological evaluation of the neurons and glia. For co-culture characterization, 20 high power (400X) fields were investigated on 3 different coverslips.
Total RNA was harvested from cortical cultures and co-culture with Trizol (Invitrogen), purified with the Qiagen RNeasy midi kit (Qiagen), and cDNA was generated with high capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA). CDNA was created. QPCR was performed, as previously reported,19-21 for evaluation of MAP2, a neuronal differentiation marker, and mediators of cell death (Caspases 1, 3 and 9). Briefly, primer sets, conjugated to Taqman MGB probes, were used for QPCR (Applied Biosystems). QPCR reactions were carried out with equivalent dilutions of each cDNA sample on the Applied Biosystems Model 7900 sequence detector PCR machine, as previously reported from our laboratory.19-20 The relative abundance of the target of interest was divided by the relative abundance of 18S in each sample to generate a normalized abundance for the target of interest. All samples were analyzed in triplicate.
Dendrite processes were analyzed at DD 3 using previously described techniques.25 Briefly, cells were selected at random using at least 3 coverslips for each condition. One coverslip represented 3 fetal brains from 1 dam and three different dams were used for each condition. At least 3 experiments were performed for the condition. To quantify processes emanating from each cell body, 30 neurons from each treatment group were evaluated at a final image magnification of 400X. Individual neurons were selected if they were clearly defined and not overlapping with other neurons. Fluorescent images were recorded and analyzed using a Dell Latitude D620, using an image processing program (Image J 1.37v).
In these experiments, two treatment groups were utilized. Fetal corteces were extracted from dams exposed to LPS for six hours and controls at E15. Primary cultures of either neurons or neurons plus glia (co-cultures) were created from the fetal corteces as described previously.
At same gestational age, fetal brains were harvested from control dams or dams exposed to intrauterine LPS and four primary cultures were created: 1) LPS-exposed neuronal culture (LN), 2) LPS-exposed co-culture (LCC), 3) Control neuronal culture (CN) and 4) Control co-culture (CCC) (Figure 1).
To test our first hypothesis, we created primary neuronal cultures from control fetuses. After 24 hours in culture, control neurons were exposed to media from LPS-exposed neurons (LN), LPS exposed co-culture (LCC), control neurons (CN) and control co-culture (CCC) (“Damage experiment”) (Figure 1A). To test our 2nd hypothesis, we created primary neuronal cultures from fetal brains exposed to LPS in vivo (LN). After 24 hours in culture, these primary cultures were exposed to media from LPS-exposed neurons (LN), LPS exposed co-culture (LCC), control neurons (CN) and control co-culture (CCC)(“Rescue experiment”) (Figure 1B).
In the ‘damage’ and ‘rescue’ experiments, the outcomes assessed were 1) neuronal morphology evaluated by immunocytochemistry at DD3, DD10 and 14; and 2) quantitative assessment of dendritic processes at DD3 which was performed 48 hours after media treatment.
For characterization of co-cultures, Student t-test was utilized to compare percent of cell types (astrocytes, microglia or neurons) in the LPS-exposed and control co-cultures. Similarly, Student t-test was used for the comparison of QPCR MAP2 mRNA expression between LPS-exposed cultures and controls. For QPCR analysis of cell mediators of cell death, the mean mRNA from the treatment groups was statistically analyzed between the 4 treatment groups (LN, CN, LCC and CCC) with One-way ANOVA. Pair-wise comparison was then performed, utilizing Student-Newman-Keuls (SNK) method. For the investigation of the number of dendritic processes, One-way ANOVA was used to compare values between the different treatment groups. Pair-wise comparison was then performed using SNK.
Culture characterization at DD7, showed no significant differences in glial or neuronal content between control cultures (29% neurons; 70% astrocytes and 1% microglia) and LPS-exposed co-cultures (66% astrocytes, 33% neurons and 1% microglia) (P>0.5, Student t-test for each cell type) (Figure 2).
To investigate whether in vivo exposure to LPS altered neuronal content in the cortical and co-cultures, we assessed the expression MAP2, a neuronal differentiation marker. MAP2 mRNA expression was significantly decreased in neuronal cultures exposed to LPS than in control (Figure 3). Yet, there was no statistically significant difference in MAP2 expression between co-culture controls and co-cultures from the LPS-exposed fetal brains.
To determine if in vivo exposure to LPS activated mediators involved in cell death pathways, we assessed expression of Caspases 1, 3 and 9. Caspase-3 mRNA expression was significantly different between the treatment groups (One-way ANOVA, P=0.049). However, there was no significant difference between LPS-exposed and control cortical or co-cultures with pair-wise comparison (SNK). Neither Caspase-1 nor caspase-9 mRNA expression was significantly different between LPS-exposed or control neuronal cultures and co-cultures.
Unlike the control neuronal cells which demonstrated numerous processes present with discrete cellular morphology, the LPS-exposed neurons (from fetal brains exposed in vivo) in both cortical and co-culture demonstrated decreased growth of the dendritic processes, abnormal morphology and overall decreased MAP2 staining, a phenomenon known as MAP2 beading,26 as they continued to grow (Figure 4).
Similarly, in the ”damage experiments” abnormal neuronal morphology was induced in control neurons 1) exposed to the media from LPS-exposed neurons (LN) (Figure 5) and 2) exposed to media from LPS treated co-cultures (LCC). Media from LN and LCC induced similar morphological changes in neurons as LN with an absence of uniform cytoskeletal staining, marked by a decrease in MAP2 staining. In the “rescue experiments,” the abnormal morphology of LPS-exposed neurons (LN) remained and could not be ameliorated after treatment with media from 2 control groups, CCC and CN, suggesting that the normal ‘milieu’ of neurons or neurons and glia were unable to reverse neuronal damage from in vivo exposure to LPS.
To quantify the observed neuronal injury from the “damage” and “rescue” experiments, we evaluated the dendritic growth. In “damage experiments,” the number of dendrites was evaluated in control neurons exposed to conditioned media. As evidenced by the number of dendrites at 48 hours after treatment (DD3), media from LPS-exposed neurons and LPS-exposed co-cultures induced quantifiable and statistically significant morphological changes in control neurons comparable to those of LPS –exposed neurons when looking at dendritic arborization as determined by the number of dendritic processes (Figure 6). In contrast, in “rescue experiments,” the neuronal damage caused by an in vivo exposure to inflammation in the LPS-exposed neurons could not be reversed by media from control neurons or control co-culture as the number of dendrites did not significantly change (Figure 7).
Neuronal injury in inflammation-induced preterm birth may be a critical mechanism for adverse neurological outcomes in the ex-preterm children. In previous work in our laboratory, we have demonstrated the presence of neuronal injury in preterm fetal brains using an LPS-exposed mouse model. 16 These current studies now demonstrated that this neuronal injury can be propagated by the injured neurons. This suggests that while the trigger of inflammatory pathways may be ‘removed’ after the preterm birth that neuronal injury may continue into the neonatal period and beyond, with the presence of injured neurons. If valid in vivo, this finding may imply that a chronic neuronal injury may be present in the brains of ex-preterm neonates. These findings and their implications mandate a paradigm shift in the conceptualization of fetal brain injury in a preterm birth.
A number of animal models have been developed to elucidate the mechanisms of inflammation induced brain damage.20, 27-31 While all animal models have inherent limitations, these studies and such mechanistic questions cannot be assessed in humans. It is noted that CNS development in the rodent differs from the human in regards to oligodendrocytes development. In the E15 fetal mouse brain, myelination is not completed. While pro-oligodendrocytes are present, mature oligodendrocytes are not. We have previously demonstrated that intrauterine inflammation results in a loss of these pro-oligodendrocytes. 20 Notably, for these studies, we are interested in neuron-neuron and neuron-glial interactions and hence the lack of mature oligodendrocytes for these studies is not a limitation. The strengths of the study are that this model utilizes an in utero or local model which is believed to mirror the inflammation associated with human preterm birth. 18 Furthermore, the molecular events revealed in the studies most likely represent what occurs in vivo in the fetal brain with intact paracrine effects of neuron-neuron and glial-neuronal mechanisms.
A potential limitation of the study; if only as an unknown factor, is the utilization of CO2 for euthanasia. The possible consequences of rising CO2 on health of an embryo are well realized. However, since both control and LPS-exposed groups were euthanized with the same amount of CO2, the effect would be equal across treatment groups and hence is unlikely to be the reason neuronal injury was observed
Animal models, used to elucidate the mechanisms by which preterm birth promotes fetal brain injury,27,28,31 have, to date, concentrated on WMD and astrogliosis as the outcomes of interest. However, recent studies suggest that adverse neurological outcomes in ex-preterm infants encompass many disorders beyond CP 4-11 and neuronal injury is known to mechanistically contribute to such deficits as found in other diseases involving neurotoxicity.32,33 As demonstrated from our laboratory, an intrauterine LPS induces cytokine production in the fetal brain20 and subsequent fetal neuronal injury.13 This study provides further understanding into the mechanisms promoting the neuronal injury.
Neuroinflammation and neurotoxicity are frequently considered distinct but common hallmarks of several neuroinflammatory disorders, including Parkinson's disease, amyotrophic lateral sclerosis, multiple sclerosis, and Alzheimer's disease. Neuron-neuron and glial-neuronal interactions have been demonstrated to be critical mechanisms in these disease states.14 Microglial cells, the brain resident macrophages, and astrocytes, the most prevalent type of cell in brain, are actively involved in the control of neuronal activities both in fetus and adult.34 At the same time, neurons influence glial and neuronal functions, through direct cell-to-cell interactions as well as the release of soluble mediators.35 Similarly, these studies herein demonstrate that neurons participate in cross talk in fetal brain injury in the setting of intrauterine inflammation. Most importantly, this study has demonstrated that neurons injured during fetal life are sufficient to cause further neuronal injury.
This work suggests that mechanisms of injury in fetal brain damage in preterm birth are similar to other neuroinflammatory disorders and that neuron-neuron communication are implicated in the observed injury. Further work is required to determine the specific contribution of glia, specifically astrocytes, to fetal brain injury. It could be speculated, based on our observation of similar MAP2 expression in control and LPS-exposed co-cultures, as opposed to a differential expression in neuronal cultures, that glia (present in co-culture) may serve a neuroprotective role through paracrine interactions.
It appears self-evident that alterations in neuronal morphology, as we have now demonstrated, is present in fetal brains exposed to intrauterine inflammation, is detrimental for neuronal function in that the presence of fewer synaptic spines implies a diminished neuronal information processing.36 Dendritic morphology has been demonstrated to be vital to normal neuronal information processing and allows the neuron to receive messages and permits cortical connectivity.37 A decreased MAP2 staining in cultured neurons is demonstrated to be associated with fragility and be a result of excitotoxicity.38, 39 Our findings, which demonstrate that inflammation-induced preterm birth results in diminished MAP2 staining and MAP2 expression, suggest that intrauterine inflammation may result in abnormal neuronal processing and synaptic communication. Our results further indicate that the neuronal injury appears to be present in inflammation-induced preterm birth, accompanies WMD, 20 and may explain the adverse neurological outcomes, including cognitive and behavioral deficits.
In these studies, the observed phenomenon that injured neurons alone are sufficient to induce the neuronal injury in previously unaffected neurons implies that local paracrine effects are important in are important in development of a global brain injury and propagation of neuronal morphological change. These neuron-neuron interactions have immense implications for human outcomes. An important finding in our study is that the co-culture characterization did not reveal any changes in percentage of glia or neurons nor did we observe an increase in Caspase expression, suggesting that astrogliosis and neuronal death may be later findings in neurotoxicity. The inability of the normal neuronal or glial-neuronal milieu (from control cortical or co-cultures) to reverse the neuronal injury suggests the irreversibility of neuronal injury possibly due to excitotoxic pathways. Alternatively, this may be a finding limited to an ex vivo study and, in fact, neuronal injury may be reversible in vivo as neuroprotective mechanisms may require paracrine effects which are present in the intact fetal brain.
Consequently, these results have immediate consequences both for research and for the development of novel therapies. As there are no proposed mechanisms of neuronal injury in the setting of preterm birth, these data are crucial for directing new research directions. With the increasing prevalence of adverse neurological outcomes in ex-preterm infants, understanding the mechanism of fetal brain injury in a preterm birth is critical if we are to decrease both acute and long term adverse outcomes for these children. Based on our recent work, targeting neuronal injury warrants active investigations. Specifically, future studies are required to determine if the neuronal-neuronal injury continues into the postnatal period or whether there are innate neuroprotective mechanisms, such as neurotrophins, that limit injury. If neuronal-neuronal injury is a perpetuating source of brain injury in the ex-preterm infant, these events become critical therapeutic targets.
This project was supported by the NIH: 5-RO1-HD046544-0 (ME) and supported in part by the Institute for Translational Medicine and Therapeutics of the University of Pennsylvania. The project described was supported by Grant Number UL1RR024134 from the National Center For Research Resources. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center For Research Resources or the National Institutes of Health.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Presented as an Oral presentation #21 at Society for Maternal-Fetal Medicine in San Diego, CA, January 2009
Inflammation-induced preterm birth results in irreversible neuronal injury in the fetal brain.