- Email: [email protected]
Maternal infection regulates BDNF and NGF expression in fetal and neonatal brain and maternal–fetal unit of the rat
Journal of Neuroimmunology 138 (2003) 49 – 55 www.elsevier.com/locate/jneuroim
Maternal infection regulates BDNF and NGF expression in fetal and neonatal brain and maternal–fetal unit of the rat John H. Gilmore *, L. Fredrik Jarskog, Swarooparani Vadlamudi Department of Psychiatry, CB #7160, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7160, USA Received 24 January 2003; received in revised form 3 March 2003; accepted 5 March 2003
Abstract Maternal infection during pregnancy is associated with increased risk for neurodevelopmental disorders. Lipopolysaccharide (LPS) or saline was administered to rats to model maternal infection, and levels of brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) in maternal plasma, placenta, amniotic fluid, fetal liver/spleen, fetal brain, and cerebral cortex after birth were determined by ELISA or semiquantitative Western blot analysis. BDNF expression was significantly increased in the fetal brain ( p = 0.039); NGF expression was significantly increased in neonatal cortex ( p = 0.0009). Neurotrophic factor expression was also altered in other tissues of the maternal – fetal unit. Abnormal expression of neurotrophic factors represents a potential mechanism through which maternal infection increases risk for neurodevelopmental disorders. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Lipopolysaccharide; Tumor necrosis factor-a; Placenta; Amniotic fluid
1. Introduction Prenatal exposure to maternal infection is a risk factor for neurodevelopmental disorders including schizophrenia (Brown et al., 2000), mental retardation (Rantakallio and Von Wendt, 1985), and autism (Ciaranello and Ciaranello, 1995). Little is known about the mechanisms through which exposure to maternal infection act on the developing brain and how this increases risk for neurodevelopmental disorders. It has been proposed that inflammatory cytokines, generated by the maternal or fetal immune system, play a key mechanistic role in the association between maternal infection, especially chorioamnionitis, and white matter damage (Dammann and Leviton, 1998; Gomez et al., 1998). Cytokines play an important role in the normal development of neurons, including neuron proliferation, survival, differentiation, and axodendritic outgrowth and synapse regulation. For example, we have shown that interleukin 1h (IL-1h), interleukin 6 (IL-6), and tumor necrosis factor-a (TNFa) decrease in vitro survival of embryonic dopamine and serotonin neurons (Jarskog et al., 1997) and embryonic cortical neurons (Marx et al., * Corresponding author. Tel.: +1-919-966-6971; fax: +1-919-9669604. E-mail address: [email protected] (J.H. Gilmore).
2001). Thus, maternal infection may alter the development of neurons in a way that increases risk for neurodevelopmental disorders. There is increasing evidence of interactions between inflammatory cytokines and neurotrophins in the nervous system. Neurotrophins, including brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) play critical roles in the development of the nervous system, regulating neuron survival and differentiation, as well as synaptic development and maintenance. Systemically administered IL-1h decreases BDNF mRNA in the hippocampus (Lapchak et al., 1993); neuron death after NGF deprivation in vitro is mediated by TNFa (Barker et al., 2001). In addition, IL-1h and TNFa increase NGF expression in microglia (Heese et al., 1998) and astrocytes (Juric et al., 2001). Transgenic mice overexpressing TNFa have alterations in both BDNF and NGF expression in several brain areas (Aloe et al., 1999). Cytokine regulation of neurotrophic factor expression in the developing brain represents a potential mechanism through which early exposure to infection could alter neuron development, including neuron survival and the establishment and maintenance of synaptic connectivity, processes that have been implicated in neurodevelopmental disorders. Little is known about the expression of BDNF and NGF during pregnancy, how these neurotrophic factors are regu-
0165-5728/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0165-5728(03)00095-X
50
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
lated by maternal infection, and how maternal infection ultimately alters their expression in the fetal and neonatal brain. NGF levels in cord serum are low in children with microcephaly (Haddad et al., 1994). We found that levels of NGF and BDNF are decreased in the amniotic fluid of human pregnancies with evidence of infection; amniotic fluid NGF was decreased in pregnancies in which the fetus had CNS abnormalities (Marx et al., 1999). BDNF levels are increased in the blood of newborns that go on to develop autism or mental retardation (Nelson et al., 2001). Therefore, neurotrophic factor levels in the blood or amniotic fluid appear to be altered by maternal infection and may be associated with abnormal brain development. We studied the regulation of BDNF and NGF expression in maternal plasma, placenta, amniotic fluid, fetal liver/spleen, fetal brain, and neonatal cortex in a rodent model of maternal infection that utilizes E. coli lipopolysaccharide (LPS), a cell wall endotoxin that stimulates an immune response. LPS administered to pregnant mice increases IL-1a, IL-6, and TNF-a in the maternal circulation and IL-1a and IL-6 in amniotic fluid (Fidel et al., 1994). We have shown that LPS increases IL-1h, IL-6, and TNFa expression in the placenta and amniotic fluid in rats and decreases fetal brain TNFa in rats (Urakubo et al., 2001). Maternal LPS administration in rats and hamsters produces a variety of CNS abnormalities including enlarged ventricles, microcephaly, leukoencepalopathy, and neuronal necrosis (Ornoy and Altschuler, 1976). We hypothesized that maternal LPS exposure will regulate neurotrophic factor expression in the fetal and neonatal brain, as well as in the placenta and amniotic fluid. Levels of TNFa protein in the maternal plasma and fetal liver/ spleen were determined as a marker of the immune response in the mother and fetus.
spleen, two samples from each litter were pooled because of small sample weight, giving an N of 3 per group. For the subacute study, timed-pregnant rats were injected i.p. daily with 0.1 mg/kg of LPS or saline on E14, E15, and E16. Pups were decapitated on day 6 after birth; frontal cortex was dissected and frozen on dry ice. Three male and three female pups from three separate litters were used for the treatment and control groups. 2.2. Sample preparation All specimens were stored at 80 jC. Brain tissue, placenta, and liver/spleen was placed in 10– 15 volumes of 50 mM Tris –HCl buffer (pH 7.4) with 0.6 M NaCl, 0.2% Triton X-100, 0.5% BSA containing freshly dissolved protease inhibitors: 1 mM benzamidine, 0.1 mM benzethonium chloride, and 0.1 mM phenylmethylsulfonyl fluoride. Samples were homogenized (PowerGen 125, Fisher Scientific, Pittsburgh, PA) on ice for 30 s and sonicated (Sonic Dismembrator 60, Fisher Scientific) for 10 s at 10 mV. Samples were centrifuged at 12,000 rpm for 20 min at 4 jC and the supernatants were aliquoted and frozen at 80 jC until assays were performed. Enzyme-linked immunosorbent assay (ELISA) was used to measure BDNF in all tissues and NGF in brain and amniotic fluid. BDNF was detected in all samples except amniotic fluid with ELISA. Because the NGF ELISA uses an anti-rat IgG-horseradish peroxidase (HRP) conjugate that has the potential to cross-react with IgG in samples with high IgG levels, a semiquantitative Western blot analysis was used to determine the relative amounts of NGF in maternal plasma, placenta, and fetal liver/spleen. NGF was detected in fetal brain and amniotic fluid with ELISA and in fetal liver/spleen and placenta with Western blots (Fig. 1). The active NGF isoform (approximately 14 kDa) were not detectable in maternal plasma using Western blot.
2. Materials and methods 2.3. ELISA 2.1. Animals Experimental protocols were approved by the UNC Institutional Animal Care and Use Committee. For the acute study, gestational day 16 timed-pregnant Sprague – Dawley rats (Charles River, Raleigh, NC) were injected i.p. with 0.5 mg/kg of LPS (E. coli, 055:B5; Sigma, St. Louis, MO) or saline. After injection (2, 8, and 24 h), the animals were anesthetized with ether and decapitated after blood was obtained by cardiac puncture. The uterine horns containing embryonic day 16 pups were surgically removed. Amniotic fluid was aspirated with a syringe; placenta, fetal liver/ spleen, and fetal whole brain were dissected and immediately frozen on dry ice. Maternal plasma was obtained after centrifuging EDTA-treated blood samples for 6 min at 2200 rpm. Two pups from three separate litters (n = 6) were used for each treatment and control group; there were three samples per group for maternal plasma. For the fetal liver/
Neurotrophin levels were determined using a two-site ELISA according to the directions of the manufacturer for BDNF, NGF (Emax Immunoassay System, Promega, Madison, WI) and rat TNFa (Biosource, Camarillo, CA). Preliminary experiments were done to determine the optimal sample dilution for the standard curve of each assay-fetal
Fig. 1. Representative Western blots of NGF protein in the (A) placenta 2 h after maternal LPS exposure and (B) fetal liver/spleen 8 h after maternal LPS exposure. Lanes 2, 4, and 6 are saline exposed; lanes 3, 5, and 7 are LPS (0.5 mg/kg) exposed. Lane 1 is control and lane 8 is a pooled sample.
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
51
brain: 1:15 for BDNF and NGF; placenta: 1:10 for BDNF; liver/spleen: 1:15 for BDNF and TNFa; amniotic fluid: neat for BDNF and NGF; maternal plasma: neat for BDNF and TNFa; neonatal cortex: 1:15 for BDNF and NGF. For the BDNF and NGF ELISAs, samples were added to a 96-well microplate coated with a primary antibody (Ab) followed by a secondary Ab (NGF: primary polyclonal goat anti-NGF Ab, monoclonal rat anti-NGF secondary Ab; BDNF: primary monoclonal mouse anti-BDNF Ab, secondary polyclonal chicken anti-BDNF Ab), followed by a horseradish peroxidase (HRP) conjugate and the chromogen tetramethyl benzidine (TMB). For the TNFa ELISA, samples were added to a 96-well microplate coated with a primary rabbit polyclonal anti-TNFa Ab followed by a secondary biotinylated rabbit polyclonal anti-TNFa. Streptavidin-peroxidase was added, followed by TMB. In each assay, the reaction was stopped and the optical density was measured at 450 nm using a microplate reader (Vmax, Molecular Devices, Sunnyvale, CA). The limits of detection for BDNF, NGF, and TNFa were 15.6, 15.6, and 4 pg/ml, respectively. The mean intraassay coefficients of variance for BDNF, NGF and TNFa were 2.4%, 3.3%, and 5.0%, respectively. Samples and standards were run in duplicate. LPS and control samples were run in the same assay on the same plate. 2.4. Western blot analysis Total protein in the samples were measured by using a BCA Protein Assay Kit (Pierce, Rockford, IL). Samples
Fig. 3. BDNF and NGF protein levels in the cortex of day 6 rat pups after maternal exposure to 0.1 mg/kg LPS or saline on E14, E15, and E16.
Fig. 2. BDNF and NGF protein levels in the fetal brain and placenta 2, 8, and 24 h after maternal injection of 0.5 mg/kg LPS or saline.
52
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
were separated on 16% Tris – glycine polyacrylamide gels using a mini cell electrophoresis unit (Xcell II, NOVEX, San Diego), as previously described (Jarskog and Gilmore, 2000) with some minor modifications. Briefly, 40 Ag of protein from each sample was mixed with 15 –20 Al of sample buffer, boiled for 5 min, quickly cooled, and loaded to the wells. Along with the samples, low-range molecular weight ladder (Rainbow MW Marker, Amersham-Pharmacia) and 0.25 ng of rat h-NGF (R&D systems) or 15 ng of BDNF (Santa Cruz Biotechnology) were used as controls. Separated proteins were transferred onto PVDF membrane at 25 V for 90 min. Uniformity of protein loading and transfer was ascertained by staining duplicate membranes with Ponceau S. Membranes were blocked for 1 h at room temperature,
with 5% nonfat dry milk in TBS containing 0.1% Tween-20 (TBST). Immunodetection was done by primary antibody incubation with rabbit polyclonal anti-NGF (1:100, Santa Cruz Biotechnology) overnight at 4 jC and a rabbit polyclonal BDNF (1:200, Santa Cruz Biotechnology) for 1 h at room temperature. After washing with TBST, the membranes were incubated for 1 h at room temperature with anti-rabbit-HRP (1:5000, Amersham-Pharmacia). Membranes were developed using chemiluminescence (ECL, Amersham-Pharmacia) and protein bands were detected on radiographic film (Hyperfilm, Amersham-Pharmacia). The optical densities of the bands were measured using a NIHImage 1.62. Band densities were normalized to the density of a pooled sample, which was loaded on all gels.
Fig. 4. BDNF, NGF, and TNFa protein levels in the fetal liver/spleen, maternal plasma, and amniotic fluid 2, 8, and 24 h after maternal injection of 0.5 mg/kg LPS or saline. BDNF was undetectable in amniotic fluid and NGF was undetectable in maternal plasma.
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
2.5. Statistical analysis Statistical analyses were performed using Prism 3.0 (GraphPad Software, San Diego, CA). A two-way ANOVA was used to determine the effect of both treatment and time in the acute study and treatment and gender for neonatal cortex. Bonferroni post-tests were performed if overall treatment effects were significant. Significance was set at 0.05 using a two-tailed test for all analyses.
3. Results In the fetal brain, acute LPS exposure significantly increased BDNF protein levels (treatment p = 0.039), while there was no treatment effect on NGF levels (treatment p = 0.82; Fig. 2). There were significant effects of time on both BDNF (time p = 0.0007) and NGF ( p = 0.0015), with levels for both neurotrophins increasing over the 24 h after injection in both LPS exposed and control fetuses. In day 6 pups, cortical NGF was significantly increased (treatment p = 0.0009) while BDNF was unchanged ( p = 0.1199); there was no gender effect for either NGF or BDNF (Fig. 3). In the liver/spleen, there was a significant interaction between time and treatment for NGF (interaction p = 0.0033), with NGF levels being decreased at 8 h after injection (Fig. 4). TNFa levels were significantly decreased overall (treatment p = 0.013), with a significant decrease at 2 h after injection ( p < 0.05). There was no treatment effect on BDNF ( p = 0.94). In the placenta, LPS significantly increased NGF expression (treatment p = 0.0029), while there was a trend for an increase of BDNF ( p = 0.078; Fig. 4). In amniotic fluid, there was a significant interaction between time and treatment for NGF (interaction p = 0.0088), with NGF levels being higher than controls at 2 and 8 h, and lower at 24 h after injection. BDNF was undetectable in all amniotic fluid samples. In maternal plasma, TNFa was robustly increased after LPS exposure though this did not reach statistical significance due to the small sample size (treatment p = 0.068; Fig. 4). BDNF in maternal plasma was significantly decreased overall (treatment p = 0.0022), and was significantly decreased 24 h after LPS injection ( p < 0.05; Fig. 4).
4. Discussion This study indicates maternal infection can regulate expression of BDNF and NGF protein in the fetal brain and neonatal cortex. There was a significant increase in BDNF in the fetal brain after acute maternal LPS exposure. This increase represents a unique response of the fetal brain to maternal infection as there was no change of BDNF in the fetal liver/spleen. Increased BDNF in the developing cortex after exposure to maternal infection could ultimately change
53
the number and specificity of synaptic contacts between neurons altering connectivity in functional circuits that underlie the symptoms associated with schizophrenia and other neurodevelopmental disorders. For example, BDNF can inhibit ocular dominance formation (Cabelli et al., 1995) and can inhibit dendritic growth in layer 6 of the visual cortex while stimulating dendritic growth in layer 4 (McAllister et al., 1997). The precise impact of abnormal neurotrophic factor expression in the developing cortex would depend on the severity and timing of infection in relation to region-specific neurodevelopmental events. In our model, subacute maternal infection caused an increase of NGF in neonatal cortex at day 6, indicating that prenatal exposure to maternal infection can cause longlasting changes in neurotrophic factor expression in the developing cortex. The increased NGF in the neonatal cortex observed in this study may be the result of activated astrocytes or microglia. Activated microglia and astrocytes express NGF (Heese et al., 1998), and prenatal maternal LPS exposure increases glial acidic fibrillary protein expression at day 8 after birth, an indication of astrogliosis (Cai et al., 2000). Though further study is needed, this persistent inflammatory response in the developing cortex after maternal infection would likely impact developing cortical circuits. TNFa and NGF levels in the fetal liver/spleen decreased after maternal LPS exposure. We previously found that TNFa was decreased in the fetal brain and increased in the placenta and amniotic fluid after maternal LPS exposure (Urakubo et al., 2001). The decrease of TNFa in fetal liver/ spleen (and fetal brain) is unexpected and in contrast to increases of TNFa mRNA in the liver and spleen (Turrin et al., 2001) and in the brain (Nadeau and Rivest, 1999) of adult rodents after LPS exposure. LPS does not enter the fetal circulation (Goto et al., 1994); thus, the response of TNFa, BDNF and NGF in the fetus would be the result of a unique regulation by inflammatory signals generated by the mother after exposure. While a direct comparison is not possible, the increase of BDNF in fetal brain after maternal infection is in contrast to the decrease of BDNF described in the adult hippocampus after systemic IL-1h administration (Lapchak et al., 1993) and indicates that the fetal brain responds to inflammatory stimuli in a different manner than the mature brain. Our study indicates that it is not possible to predict the response of the fetal immune system or the fetal brain to infection based on studies in mature animals. Maternal infection significantly increases expression of NGF in the placenta. NGF mRNA is found in human placenta (MacGrogan et al., 1992), and BDNF and NGF is expressed in human amniotic epithelial tissue (Uchida et al., 2000). It has been proposed that the placenta and amniotic fluid is a source of neurotrophic factors for the developing fetus (Uchida et al., 2000). It is therefore possible that abnormal expression of NGF or other neurotrophic factors by the placenta in the setting of infection would have a significant impact on the developing fetal brain.
54
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
NGF in amniotic fluid was acutely increased after maternal LPS exposure, then decreased compared to controls. In human pregnancies complicated by infection, we found a decrease of NGF and BDNF in amniotic fluid (Marx et al., 1999). The source(s) of NGF in the amniotic fluid is unknown, though it is likely that placental amniotic epithelial tissue is one source (Uchida et al., 2000). The increase of amniotic fluid NGF in our study was similar to that observed in the placenta and the opposite of the overall decreased levels observed in the fetal liver/spleen, indicating an extra-fetal source. Interestingly, BDNF was significantly and persistently decreased in the maternal plasma after LPS exposure. To our knowledge, this is the first study of the effect of infection on plasma BDNF. There is evidence that some maternally derived growth factors cross the placenta and are active in the fetus; if maternally derived BDNF contributes to fetal brain development, then a decrease in maternal levels after infection could have a significant impact on fetal brain development. In summary, maternal infection causes significant alterations in BDNF and NGF in the fetal and neonatal brain, representing a potential mechanism through which maternal infection can alter neuron and synaptic development and increase risk for neurodevelopmental disorders such as schizophrenia and mental retardation. In the setting of infection, the neurotrophic factor expression response may represent a neuroprotective response to the toxicity of inflammatory mediators (Stadelmann et al., 2002). The precise impact of abnormal neurotrophic expression in the developing cortex after infection would depend on the severity and timing of infection in relation to specific neurodevelopmental events. Maternal infection also decreased BDNF in the maternal plasma, and increased NGF in the amniotic fluid and placenta. Alterations of these systemically generated neurotrophic factors after maternal infection may contribute to abnormal brain development and risk for neurodevelopmental disorders. Further research is required to understand cytokine – neurotrophic factor interactions in the developing brain and their regulation by exposure to maternal infection.
Acknowledgements This study was supported by NIMH grant MH60352 (JHG).
References Aloe, L., Properzi, F., Probert, L., Akassoglou, K., Kassiotis, G., Micera, A., Fiore, M., 1999. Learning abilities, NGF and BDNF brain levels in two lines of TNF-a transgenic mice, one characterized by neurological disorders, the other phenotypically normal. Brain Res. 840, 125 – 137. Barker, V., Middleton, G., Davey, F., Davies, A.M., 2001. TNFa contrib-
utes to the death of NGF-dependent neurons during development. Nat. Neurosci. 4, 1194 – 1198. Brown, A.S., Schaefer, C.A., Wyatt, R.J., Goetz, R., Begg, M.D., Gorman, J.M., Susser, E.S., 2000. Maternal exposure to respiratory infections and adult schizophrenia: a prospective birth cohort study. Schizophr. Bull. 26, 287 – 295. Cabelli, R.J., Hohn, A., Shatz, C.J., 1995. Inhibition of ocular dominance column formation by infusion of NT-4/5 or BDNF. Science 267, 1662 – 1666. Cai, Z., Pan, Z.L., Pang, Y., Evans, O.B., Rhodes, P.G., 2000. Cytokine induction in fetal brains and brain injury in neonatal rats after maternal lipopolysaccharide administration. Pediatr. Res. 47, 64 – 72. Ciaranello, A.L., Ciaranello, R.D., 1995. The neurobiology of autism. Annu. Rev. Neurosci. 18, 101 – 128. Dammann, O., Leviton, A., 1998. Infection remote from the brain, neonatal white matter damage, and cerebral palsy in the preterm infant. Semin. Pediatr. Neurol. 5, 190 – 201. Fidel, P.L., Romero, R., Wolf, N., Cutright, J., Ramirez, M., Araneda, H., Cotton, D.B., 1994. Systemic and local cytokine profiles in endotoxininduced preterm parturition in mice. Am. J. Obstet. Gynecol. 170, 1467 – 1475. Gomez, R., Romero, R., Ghezzi, F., Yoon, B.H., Mazor, M., Berry, S.M., 1998. The fetal inflammatory response syndrome. Am. J. Obstet. Gynecol. 179, 194 – 202. Goto, M., Yoshioka, T., Ravindranath, T., Battelino, T., Young, R.I., Zeller, M.P., 1994. LPS injected into the pregnant rat late in gestation does not induce fetal endotoxemia. Res. Commun. Mol. Pathol. Pharmacol. 85, 109 – 112. Haddad, J., Vilge, V., Juif, J.G., Maitre, M., Donato, L., Messer, J., Mark, J., 1994. Beta-nerve growth factor levels in newborn cord sera. Pediatr. Res. 35, 637 – 639. Heese, K., Hock, C., Otten, U., 1998. Inflammatory signals induce neurotrophin expression in human microglia cells. J. Neurochem. 70, 699 – 707. Jarskog, L.F., Gilmore, J.H., 2000. Developmental expression of Bcl-2 protein in human cortex. Dev. Brain Res. 119, 225 – 230. Jarskog, L.F., Xiao, H., Wilkie, M.B., Lauder, J.M., Gilmore, J.H., 1997. Cytokine regulation of fetal dopaminergic and serotonergic neuron survival in vitro. Int. J. Dev. Neurosci. 15, 711 – 716. Juric, D.M., Carman-Krzan, M., 2001. Interleukin-1h, but not IL-1a, mediates nerve growth factor secretion from rat astrocytes via type I IL-1 receptor. Int. J. Dev. Neurosci. 19, 675 – 683. Lapchak, P.A., Araujo, D.M., Hefti, F., 1993. Systemic interleukin-1h decreases brain-derived neurotrophic factor messenger RNA expression in the rat hippocampal formation. Neuroscience 53, 297 – 301. MacGrogan, D., Saint-Andre, J.P., Dicou, E., 1992. Expression of nerve growth factor and nerve growth factor receptor genes in human tissues and in prostatic adenocarcinoma cell lines. J. Neurochem. 59, 1381 – 1391. Marx, C.E., Vance, B.J., Jarskog, L.F., Chescheir, N.C., Gilmore, J.H., 1999. Nerve growth factor, brain-derived neurotrophic factor, and neurotrophin-3 levels in human amniotic fluid. Am. J. Obstet. Gynecol. 181, 1125 – 1230. Marx, C.E., Jarskog, L.F., Lauder, J.M., Lieberman, J.A., Gilmore, J.H., 2001. Cytokine effects on cortical neuron MAP-2 immunoreactivity: implications for schizophrenia. Biol. Psychiatry 50, 743 – 749. McAllister, A.K., Katz, L.C., Lo, D.C., 1997. Opposing roles for endogenous BDNF and NT-3 in regulating cortical dendritic growth. Neuron 18, 767 – 778. Nadeau, S., Rivest, S., 1999. Regulation of the gene encoding tumor necrosis factor alpha (TNF-a) in the rat brain and pituitary in response to different models of systemic immune challenge. J. Neuropathol. Exp. Neurol. 58, 61 – 77. Nelson, K.B., Grether, J.K., Croen, L.A., Dambrosia, J.M., Dickens, B.F., Jelliffe, L.L., Hansen, R.L., Phillips, T.M., 2001. Neuropeptides and neurotrophins in the neonatal blood of children with autism or mental retardation. Ann. Neurol. 49, 597 – 606.
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55 Ornoy, A., Altschuler, G., 1976. Maternal endotoxemia, fetal anomalies, and central nervous system damage: a rat model of a human problem. Am. J. Obstet. Gynecol. 124, 196 – 204. Rantakallio, P., Von Wendt, L., 1985. Risk factors for mental retardation. Arch. Dis. Child. 60, 946 – 952. Stadelmann, C., Kerschensteiner, M., Misgeld, T., Bruck, W., Hohlfeld, R., Lassmann, H., 2002. BDNF and gp145trkB in multiple sclerosis brain lesions: neuroprotective interactions between immune and neuronal cells? Brain 125, 75 – 85. Turrin, N.P., Gayle, D., Ilyin, S.E., Flynn, M.C., Langhans, W., Schwartz,
55
G.J., Plata-Salaman, C.R., 2001. Pro-inflammatory and anti-inflammatory cytokine mRNA production in the periphery and brain following intraperitoneal administration of bacterial lipopolysaccharide. Brain Res. Bull. 54, 443 – 453. Uchida, S., Inanaga, Y., Kobayashi, M., Hurukawa, S., Araie, M., Sakuragawa, N., 2000. Neurotrophic function of conditioned medium from human amniotic epithelial cells. J. Neurosci. Res. 62, 585 – 590. Urakubo, A., Jarskog, L.F., Lieberman, J.A., Gilmore, J.H., 2001. Prenatal exposure to maternal infection alters cytokine expression in the placenta, amniotic fluid, and fetal brain. Schizophr. Res. 47, 27 – 36.
Maternal infection regulates BDNF and NGF expression in fetal and neonatal brain and maternal–fetal unit of the rat John H. Gilmore *, L. Fredrik Jarskog, Swarooparani Vadlamudi Department of Psychiatry, CB #7160, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7160, USA Received 24 January 2003; received in revised form 3 March 2003; accepted 5 March 2003
Abstract Maternal infection during pregnancy is associated with increased risk for neurodevelopmental disorders. Lipopolysaccharide (LPS) or saline was administered to rats to model maternal infection, and levels of brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) in maternal plasma, placenta, amniotic fluid, fetal liver/spleen, fetal brain, and cerebral cortex after birth were determined by ELISA or semiquantitative Western blot analysis. BDNF expression was significantly increased in the fetal brain ( p = 0.039); NGF expression was significantly increased in neonatal cortex ( p = 0.0009). Neurotrophic factor expression was also altered in other tissues of the maternal – fetal unit. Abnormal expression of neurotrophic factors represents a potential mechanism through which maternal infection increases risk for neurodevelopmental disorders. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Lipopolysaccharide; Tumor necrosis factor-a; Placenta; Amniotic fluid
1. Introduction Prenatal exposure to maternal infection is a risk factor for neurodevelopmental disorders including schizophrenia (Brown et al., 2000), mental retardation (Rantakallio and Von Wendt, 1985), and autism (Ciaranello and Ciaranello, 1995). Little is known about the mechanisms through which exposure to maternal infection act on the developing brain and how this increases risk for neurodevelopmental disorders. It has been proposed that inflammatory cytokines, generated by the maternal or fetal immune system, play a key mechanistic role in the association between maternal infection, especially chorioamnionitis, and white matter damage (Dammann and Leviton, 1998; Gomez et al., 1998). Cytokines play an important role in the normal development of neurons, including neuron proliferation, survival, differentiation, and axodendritic outgrowth and synapse regulation. For example, we have shown that interleukin 1h (IL-1h), interleukin 6 (IL-6), and tumor necrosis factor-a (TNFa) decrease in vitro survival of embryonic dopamine and serotonin neurons (Jarskog et al., 1997) and embryonic cortical neurons (Marx et al., * Corresponding author. Tel.: +1-919-966-6971; fax: +1-919-9669604. E-mail address: [email protected] (J.H. Gilmore).
2001). Thus, maternal infection may alter the development of neurons in a way that increases risk for neurodevelopmental disorders. There is increasing evidence of interactions between inflammatory cytokines and neurotrophins in the nervous system. Neurotrophins, including brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) play critical roles in the development of the nervous system, regulating neuron survival and differentiation, as well as synaptic development and maintenance. Systemically administered IL-1h decreases BDNF mRNA in the hippocampus (Lapchak et al., 1993); neuron death after NGF deprivation in vitro is mediated by TNFa (Barker et al., 2001). In addition, IL-1h and TNFa increase NGF expression in microglia (Heese et al., 1998) and astrocytes (Juric et al., 2001). Transgenic mice overexpressing TNFa have alterations in both BDNF and NGF expression in several brain areas (Aloe et al., 1999). Cytokine regulation of neurotrophic factor expression in the developing brain represents a potential mechanism through which early exposure to infection could alter neuron development, including neuron survival and the establishment and maintenance of synaptic connectivity, processes that have been implicated in neurodevelopmental disorders. Little is known about the expression of BDNF and NGF during pregnancy, how these neurotrophic factors are regu-
0165-5728/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0165-5728(03)00095-X
50
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
lated by maternal infection, and how maternal infection ultimately alters their expression in the fetal and neonatal brain. NGF levels in cord serum are low in children with microcephaly (Haddad et al., 1994). We found that levels of NGF and BDNF are decreased in the amniotic fluid of human pregnancies with evidence of infection; amniotic fluid NGF was decreased in pregnancies in which the fetus had CNS abnormalities (Marx et al., 1999). BDNF levels are increased in the blood of newborns that go on to develop autism or mental retardation (Nelson et al., 2001). Therefore, neurotrophic factor levels in the blood or amniotic fluid appear to be altered by maternal infection and may be associated with abnormal brain development. We studied the regulation of BDNF and NGF expression in maternal plasma, placenta, amniotic fluid, fetal liver/spleen, fetal brain, and neonatal cortex in a rodent model of maternal infection that utilizes E. coli lipopolysaccharide (LPS), a cell wall endotoxin that stimulates an immune response. LPS administered to pregnant mice increases IL-1a, IL-6, and TNF-a in the maternal circulation and IL-1a and IL-6 in amniotic fluid (Fidel et al., 1994). We have shown that LPS increases IL-1h, IL-6, and TNFa expression in the placenta and amniotic fluid in rats and decreases fetal brain TNFa in rats (Urakubo et al., 2001). Maternal LPS administration in rats and hamsters produces a variety of CNS abnormalities including enlarged ventricles, microcephaly, leukoencepalopathy, and neuronal necrosis (Ornoy and Altschuler, 1976). We hypothesized that maternal LPS exposure will regulate neurotrophic factor expression in the fetal and neonatal brain, as well as in the placenta and amniotic fluid. Levels of TNFa protein in the maternal plasma and fetal liver/ spleen were determined as a marker of the immune response in the mother and fetus.
spleen, two samples from each litter were pooled because of small sample weight, giving an N of 3 per group. For the subacute study, timed-pregnant rats were injected i.p. daily with 0.1 mg/kg of LPS or saline on E14, E15, and E16. Pups were decapitated on day 6 after birth; frontal cortex was dissected and frozen on dry ice. Three male and three female pups from three separate litters were used for the treatment and control groups. 2.2. Sample preparation All specimens were stored at 80 jC. Brain tissue, placenta, and liver/spleen was placed in 10– 15 volumes of 50 mM Tris –HCl buffer (pH 7.4) with 0.6 M NaCl, 0.2% Triton X-100, 0.5% BSA containing freshly dissolved protease inhibitors: 1 mM benzamidine, 0.1 mM benzethonium chloride, and 0.1 mM phenylmethylsulfonyl fluoride. Samples were homogenized (PowerGen 125, Fisher Scientific, Pittsburgh, PA) on ice for 30 s and sonicated (Sonic Dismembrator 60, Fisher Scientific) for 10 s at 10 mV. Samples were centrifuged at 12,000 rpm for 20 min at 4 jC and the supernatants were aliquoted and frozen at 80 jC until assays were performed. Enzyme-linked immunosorbent assay (ELISA) was used to measure BDNF in all tissues and NGF in brain and amniotic fluid. BDNF was detected in all samples except amniotic fluid with ELISA. Because the NGF ELISA uses an anti-rat IgG-horseradish peroxidase (HRP) conjugate that has the potential to cross-react with IgG in samples with high IgG levels, a semiquantitative Western blot analysis was used to determine the relative amounts of NGF in maternal plasma, placenta, and fetal liver/spleen. NGF was detected in fetal brain and amniotic fluid with ELISA and in fetal liver/spleen and placenta with Western blots (Fig. 1). The active NGF isoform (approximately 14 kDa) were not detectable in maternal plasma using Western blot.
2. Materials and methods 2.3. ELISA 2.1. Animals Experimental protocols were approved by the UNC Institutional Animal Care and Use Committee. For the acute study, gestational day 16 timed-pregnant Sprague – Dawley rats (Charles River, Raleigh, NC) were injected i.p. with 0.5 mg/kg of LPS (E. coli, 055:B5; Sigma, St. Louis, MO) or saline. After injection (2, 8, and 24 h), the animals were anesthetized with ether and decapitated after blood was obtained by cardiac puncture. The uterine horns containing embryonic day 16 pups were surgically removed. Amniotic fluid was aspirated with a syringe; placenta, fetal liver/ spleen, and fetal whole brain were dissected and immediately frozen on dry ice. Maternal plasma was obtained after centrifuging EDTA-treated blood samples for 6 min at 2200 rpm. Two pups from three separate litters (n = 6) were used for each treatment and control group; there were three samples per group for maternal plasma. For the fetal liver/
Neurotrophin levels were determined using a two-site ELISA according to the directions of the manufacturer for BDNF, NGF (Emax Immunoassay System, Promega, Madison, WI) and rat TNFa (Biosource, Camarillo, CA). Preliminary experiments were done to determine the optimal sample dilution for the standard curve of each assay-fetal
Fig. 1. Representative Western blots of NGF protein in the (A) placenta 2 h after maternal LPS exposure and (B) fetal liver/spleen 8 h after maternal LPS exposure. Lanes 2, 4, and 6 are saline exposed; lanes 3, 5, and 7 are LPS (0.5 mg/kg) exposed. Lane 1 is control and lane 8 is a pooled sample.
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
51
brain: 1:15 for BDNF and NGF; placenta: 1:10 for BDNF; liver/spleen: 1:15 for BDNF and TNFa; amniotic fluid: neat for BDNF and NGF; maternal plasma: neat for BDNF and TNFa; neonatal cortex: 1:15 for BDNF and NGF. For the BDNF and NGF ELISAs, samples were added to a 96-well microplate coated with a primary antibody (Ab) followed by a secondary Ab (NGF: primary polyclonal goat anti-NGF Ab, monoclonal rat anti-NGF secondary Ab; BDNF: primary monoclonal mouse anti-BDNF Ab, secondary polyclonal chicken anti-BDNF Ab), followed by a horseradish peroxidase (HRP) conjugate and the chromogen tetramethyl benzidine (TMB). For the TNFa ELISA, samples were added to a 96-well microplate coated with a primary rabbit polyclonal anti-TNFa Ab followed by a secondary biotinylated rabbit polyclonal anti-TNFa. Streptavidin-peroxidase was added, followed by TMB. In each assay, the reaction was stopped and the optical density was measured at 450 nm using a microplate reader (Vmax, Molecular Devices, Sunnyvale, CA). The limits of detection for BDNF, NGF, and TNFa were 15.6, 15.6, and 4 pg/ml, respectively. The mean intraassay coefficients of variance for BDNF, NGF and TNFa were 2.4%, 3.3%, and 5.0%, respectively. Samples and standards were run in duplicate. LPS and control samples were run in the same assay on the same plate. 2.4. Western blot analysis Total protein in the samples were measured by using a BCA Protein Assay Kit (Pierce, Rockford, IL). Samples
Fig. 3. BDNF and NGF protein levels in the cortex of day 6 rat pups after maternal exposure to 0.1 mg/kg LPS or saline on E14, E15, and E16.
Fig. 2. BDNF and NGF protein levels in the fetal brain and placenta 2, 8, and 24 h after maternal injection of 0.5 mg/kg LPS or saline.
52
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
were separated on 16% Tris – glycine polyacrylamide gels using a mini cell electrophoresis unit (Xcell II, NOVEX, San Diego), as previously described (Jarskog and Gilmore, 2000) with some minor modifications. Briefly, 40 Ag of protein from each sample was mixed with 15 –20 Al of sample buffer, boiled for 5 min, quickly cooled, and loaded to the wells. Along with the samples, low-range molecular weight ladder (Rainbow MW Marker, Amersham-Pharmacia) and 0.25 ng of rat h-NGF (R&D systems) or 15 ng of BDNF (Santa Cruz Biotechnology) were used as controls. Separated proteins were transferred onto PVDF membrane at 25 V for 90 min. Uniformity of protein loading and transfer was ascertained by staining duplicate membranes with Ponceau S. Membranes were blocked for 1 h at room temperature,
with 5% nonfat dry milk in TBS containing 0.1% Tween-20 (TBST). Immunodetection was done by primary antibody incubation with rabbit polyclonal anti-NGF (1:100, Santa Cruz Biotechnology) overnight at 4 jC and a rabbit polyclonal BDNF (1:200, Santa Cruz Biotechnology) for 1 h at room temperature. After washing with TBST, the membranes were incubated for 1 h at room temperature with anti-rabbit-HRP (1:5000, Amersham-Pharmacia). Membranes were developed using chemiluminescence (ECL, Amersham-Pharmacia) and protein bands were detected on radiographic film (Hyperfilm, Amersham-Pharmacia). The optical densities of the bands were measured using a NIHImage 1.62. Band densities were normalized to the density of a pooled sample, which was loaded on all gels.
Fig. 4. BDNF, NGF, and TNFa protein levels in the fetal liver/spleen, maternal plasma, and amniotic fluid 2, 8, and 24 h after maternal injection of 0.5 mg/kg LPS or saline. BDNF was undetectable in amniotic fluid and NGF was undetectable in maternal plasma.
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
2.5. Statistical analysis Statistical analyses were performed using Prism 3.0 (GraphPad Software, San Diego, CA). A two-way ANOVA was used to determine the effect of both treatment and time in the acute study and treatment and gender for neonatal cortex. Bonferroni post-tests were performed if overall treatment effects were significant. Significance was set at 0.05 using a two-tailed test for all analyses.
3. Results In the fetal brain, acute LPS exposure significantly increased BDNF protein levels (treatment p = 0.039), while there was no treatment effect on NGF levels (treatment p = 0.82; Fig. 2). There were significant effects of time on both BDNF (time p = 0.0007) and NGF ( p = 0.0015), with levels for both neurotrophins increasing over the 24 h after injection in both LPS exposed and control fetuses. In day 6 pups, cortical NGF was significantly increased (treatment p = 0.0009) while BDNF was unchanged ( p = 0.1199); there was no gender effect for either NGF or BDNF (Fig. 3). In the liver/spleen, there was a significant interaction between time and treatment for NGF (interaction p = 0.0033), with NGF levels being decreased at 8 h after injection (Fig. 4). TNFa levels were significantly decreased overall (treatment p = 0.013), with a significant decrease at 2 h after injection ( p < 0.05). There was no treatment effect on BDNF ( p = 0.94). In the placenta, LPS significantly increased NGF expression (treatment p = 0.0029), while there was a trend for an increase of BDNF ( p = 0.078; Fig. 4). In amniotic fluid, there was a significant interaction between time and treatment for NGF (interaction p = 0.0088), with NGF levels being higher than controls at 2 and 8 h, and lower at 24 h after injection. BDNF was undetectable in all amniotic fluid samples. In maternal plasma, TNFa was robustly increased after LPS exposure though this did not reach statistical significance due to the small sample size (treatment p = 0.068; Fig. 4). BDNF in maternal plasma was significantly decreased overall (treatment p = 0.0022), and was significantly decreased 24 h after LPS injection ( p < 0.05; Fig. 4).
4. Discussion This study indicates maternal infection can regulate expression of BDNF and NGF protein in the fetal brain and neonatal cortex. There was a significant increase in BDNF in the fetal brain after acute maternal LPS exposure. This increase represents a unique response of the fetal brain to maternal infection as there was no change of BDNF in the fetal liver/spleen. Increased BDNF in the developing cortex after exposure to maternal infection could ultimately change
53
the number and specificity of synaptic contacts between neurons altering connectivity in functional circuits that underlie the symptoms associated with schizophrenia and other neurodevelopmental disorders. For example, BDNF can inhibit ocular dominance formation (Cabelli et al., 1995) and can inhibit dendritic growth in layer 6 of the visual cortex while stimulating dendritic growth in layer 4 (McAllister et al., 1997). The precise impact of abnormal neurotrophic factor expression in the developing cortex would depend on the severity and timing of infection in relation to region-specific neurodevelopmental events. In our model, subacute maternal infection caused an increase of NGF in neonatal cortex at day 6, indicating that prenatal exposure to maternal infection can cause longlasting changes in neurotrophic factor expression in the developing cortex. The increased NGF in the neonatal cortex observed in this study may be the result of activated astrocytes or microglia. Activated microglia and astrocytes express NGF (Heese et al., 1998), and prenatal maternal LPS exposure increases glial acidic fibrillary protein expression at day 8 after birth, an indication of astrogliosis (Cai et al., 2000). Though further study is needed, this persistent inflammatory response in the developing cortex after maternal infection would likely impact developing cortical circuits. TNFa and NGF levels in the fetal liver/spleen decreased after maternal LPS exposure. We previously found that TNFa was decreased in the fetal brain and increased in the placenta and amniotic fluid after maternal LPS exposure (Urakubo et al., 2001). The decrease of TNFa in fetal liver/ spleen (and fetal brain) is unexpected and in contrast to increases of TNFa mRNA in the liver and spleen (Turrin et al., 2001) and in the brain (Nadeau and Rivest, 1999) of adult rodents after LPS exposure. LPS does not enter the fetal circulation (Goto et al., 1994); thus, the response of TNFa, BDNF and NGF in the fetus would be the result of a unique regulation by inflammatory signals generated by the mother after exposure. While a direct comparison is not possible, the increase of BDNF in fetal brain after maternal infection is in contrast to the decrease of BDNF described in the adult hippocampus after systemic IL-1h administration (Lapchak et al., 1993) and indicates that the fetal brain responds to inflammatory stimuli in a different manner than the mature brain. Our study indicates that it is not possible to predict the response of the fetal immune system or the fetal brain to infection based on studies in mature animals. Maternal infection significantly increases expression of NGF in the placenta. NGF mRNA is found in human placenta (MacGrogan et al., 1992), and BDNF and NGF is expressed in human amniotic epithelial tissue (Uchida et al., 2000). It has been proposed that the placenta and amniotic fluid is a source of neurotrophic factors for the developing fetus (Uchida et al., 2000). It is therefore possible that abnormal expression of NGF or other neurotrophic factors by the placenta in the setting of infection would have a significant impact on the developing fetal brain.
54
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55
NGF in amniotic fluid was acutely increased after maternal LPS exposure, then decreased compared to controls. In human pregnancies complicated by infection, we found a decrease of NGF and BDNF in amniotic fluid (Marx et al., 1999). The source(s) of NGF in the amniotic fluid is unknown, though it is likely that placental amniotic epithelial tissue is one source (Uchida et al., 2000). The increase of amniotic fluid NGF in our study was similar to that observed in the placenta and the opposite of the overall decreased levels observed in the fetal liver/spleen, indicating an extra-fetal source. Interestingly, BDNF was significantly and persistently decreased in the maternal plasma after LPS exposure. To our knowledge, this is the first study of the effect of infection on plasma BDNF. There is evidence that some maternally derived growth factors cross the placenta and are active in the fetus; if maternally derived BDNF contributes to fetal brain development, then a decrease in maternal levels after infection could have a significant impact on fetal brain development. In summary, maternal infection causes significant alterations in BDNF and NGF in the fetal and neonatal brain, representing a potential mechanism through which maternal infection can alter neuron and synaptic development and increase risk for neurodevelopmental disorders such as schizophrenia and mental retardation. In the setting of infection, the neurotrophic factor expression response may represent a neuroprotective response to the toxicity of inflammatory mediators (Stadelmann et al., 2002). The precise impact of abnormal neurotrophic expression in the developing cortex after infection would depend on the severity and timing of infection in relation to specific neurodevelopmental events. Maternal infection also decreased BDNF in the maternal plasma, and increased NGF in the amniotic fluid and placenta. Alterations of these systemically generated neurotrophic factors after maternal infection may contribute to abnormal brain development and risk for neurodevelopmental disorders. Further research is required to understand cytokine – neurotrophic factor interactions in the developing brain and their regulation by exposure to maternal infection.
Acknowledgements This study was supported by NIMH grant MH60352 (JHG).
References Aloe, L., Properzi, F., Probert, L., Akassoglou, K., Kassiotis, G., Micera, A., Fiore, M., 1999. Learning abilities, NGF and BDNF brain levels in two lines of TNF-a transgenic mice, one characterized by neurological disorders, the other phenotypically normal. Brain Res. 840, 125 – 137. Barker, V., Middleton, G., Davey, F., Davies, A.M., 2001. TNFa contrib-
utes to the death of NGF-dependent neurons during development. Nat. Neurosci. 4, 1194 – 1198. Brown, A.S., Schaefer, C.A., Wyatt, R.J., Goetz, R., Begg, M.D., Gorman, J.M., Susser, E.S., 2000. Maternal exposure to respiratory infections and adult schizophrenia: a prospective birth cohort study. Schizophr. Bull. 26, 287 – 295. Cabelli, R.J., Hohn, A., Shatz, C.J., 1995. Inhibition of ocular dominance column formation by infusion of NT-4/5 or BDNF. Science 267, 1662 – 1666. Cai, Z., Pan, Z.L., Pang, Y., Evans, O.B., Rhodes, P.G., 2000. Cytokine induction in fetal brains and brain injury in neonatal rats after maternal lipopolysaccharide administration. Pediatr. Res. 47, 64 – 72. Ciaranello, A.L., Ciaranello, R.D., 1995. The neurobiology of autism. Annu. Rev. Neurosci. 18, 101 – 128. Dammann, O., Leviton, A., 1998. Infection remote from the brain, neonatal white matter damage, and cerebral palsy in the preterm infant. Semin. Pediatr. Neurol. 5, 190 – 201. Fidel, P.L., Romero, R., Wolf, N., Cutright, J., Ramirez, M., Araneda, H., Cotton, D.B., 1994. Systemic and local cytokine profiles in endotoxininduced preterm parturition in mice. Am. J. Obstet. Gynecol. 170, 1467 – 1475. Gomez, R., Romero, R., Ghezzi, F., Yoon, B.H., Mazor, M., Berry, S.M., 1998. The fetal inflammatory response syndrome. Am. J. Obstet. Gynecol. 179, 194 – 202. Goto, M., Yoshioka, T., Ravindranath, T., Battelino, T., Young, R.I., Zeller, M.P., 1994. LPS injected into the pregnant rat late in gestation does not induce fetal endotoxemia. Res. Commun. Mol. Pathol. Pharmacol. 85, 109 – 112. Haddad, J., Vilge, V., Juif, J.G., Maitre, M., Donato, L., Messer, J., Mark, J., 1994. Beta-nerve growth factor levels in newborn cord sera. Pediatr. Res. 35, 637 – 639. Heese, K., Hock, C., Otten, U., 1998. Inflammatory signals induce neurotrophin expression in human microglia cells. J. Neurochem. 70, 699 – 707. Jarskog, L.F., Gilmore, J.H., 2000. Developmental expression of Bcl-2 protein in human cortex. Dev. Brain Res. 119, 225 – 230. Jarskog, L.F., Xiao, H., Wilkie, M.B., Lauder, J.M., Gilmore, J.H., 1997. Cytokine regulation of fetal dopaminergic and serotonergic neuron survival in vitro. Int. J. Dev. Neurosci. 15, 711 – 716. Juric, D.M., Carman-Krzan, M., 2001. Interleukin-1h, but not IL-1a, mediates nerve growth factor secretion from rat astrocytes via type I IL-1 receptor. Int. J. Dev. Neurosci. 19, 675 – 683. Lapchak, P.A., Araujo, D.M., Hefti, F., 1993. Systemic interleukin-1h decreases brain-derived neurotrophic factor messenger RNA expression in the rat hippocampal formation. Neuroscience 53, 297 – 301. MacGrogan, D., Saint-Andre, J.P., Dicou, E., 1992. Expression of nerve growth factor and nerve growth factor receptor genes in human tissues and in prostatic adenocarcinoma cell lines. J. Neurochem. 59, 1381 – 1391. Marx, C.E., Vance, B.J., Jarskog, L.F., Chescheir, N.C., Gilmore, J.H., 1999. Nerve growth factor, brain-derived neurotrophic factor, and neurotrophin-3 levels in human amniotic fluid. Am. J. Obstet. Gynecol. 181, 1125 – 1230. Marx, C.E., Jarskog, L.F., Lauder, J.M., Lieberman, J.A., Gilmore, J.H., 2001. Cytokine effects on cortical neuron MAP-2 immunoreactivity: implications for schizophrenia. Biol. Psychiatry 50, 743 – 749. McAllister, A.K., Katz, L.C., Lo, D.C., 1997. Opposing roles for endogenous BDNF and NT-3 in regulating cortical dendritic growth. Neuron 18, 767 – 778. Nadeau, S., Rivest, S., 1999. Regulation of the gene encoding tumor necrosis factor alpha (TNF-a) in the rat brain and pituitary in response to different models of systemic immune challenge. J. Neuropathol. Exp. Neurol. 58, 61 – 77. Nelson, K.B., Grether, J.K., Croen, L.A., Dambrosia, J.M., Dickens, B.F., Jelliffe, L.L., Hansen, R.L., Phillips, T.M., 2001. Neuropeptides and neurotrophins in the neonatal blood of children with autism or mental retardation. Ann. Neurol. 49, 597 – 606.
J.H. Gilmore et al. / Journal of Neuroimmunology 138 (2003) 49–55 Ornoy, A., Altschuler, G., 1976. Maternal endotoxemia, fetal anomalies, and central nervous system damage: a rat model of a human problem. Am. J. Obstet. Gynecol. 124, 196 – 204. Rantakallio, P., Von Wendt, L., 1985. Risk factors for mental retardation. Arch. Dis. Child. 60, 946 – 952. Stadelmann, C., Kerschensteiner, M., Misgeld, T., Bruck, W., Hohlfeld, R., Lassmann, H., 2002. BDNF and gp145trkB in multiple sclerosis brain lesions: neuroprotective interactions between immune and neuronal cells? Brain 125, 75 – 85. Turrin, N.P., Gayle, D., Ilyin, S.E., Flynn, M.C., Langhans, W., Schwartz,
55
G.J., Plata-Salaman, C.R., 2001. Pro-inflammatory and anti-inflammatory cytokine mRNA production in the periphery and brain following intraperitoneal administration of bacterial lipopolysaccharide. Brain Res. Bull. 54, 443 – 453. Uchida, S., Inanaga, Y., Kobayashi, M., Hurukawa, S., Araie, M., Sakuragawa, N., 2000. Neurotrophic function of conditioned medium from human amniotic epithelial cells. J. Neurosci. Res. 62, 585 – 590. Urakubo, A., Jarskog, L.F., Lieberman, J.A., Gilmore, J.H., 2001. Prenatal exposure to maternal infection alters cytokine expression in the placenta, amniotic fluid, and fetal brain. Schizophr. Res. 47, 27 – 36.