INTRODUCTION
Hypoxia–ischemia (HI) is a common cause of perinatal brain damage, and contributes to later life neurologic impairment including cerebral palsy, epilepsy, learning disorders and intellectual disability (Vannucci, Reference Vannucci2000; Volpe, Reference Volpe2009). Both term and preterm infants are vulnerable to HI, although the pathophysiology observed is dependent on the age of the fetus and the developmental stage of the brain (Volpe, Reference Volpe2009). In the United States, approximately 63,000 infants are born annually with very low birth weight (VLBW ≤ 1500 g) (Martin et al., Reference Martin, Kung, Mathews, Hoyert, Strobino and Guyer2008; Volpe, Reference Volpe2009). The incidence of neurological sequelae is common, with 25–50% of VLBW infants suffering long-term cognitive, behavioral or attention deficits, while major motor deficits including cerebral palsy are present in 5–10% (Bayless and Stevenson, Reference Bayless and Stevenson2007; Allin et al., Reference Allin, Walshe, Fern, Nosarti, Cuddy and Rifkin2008; Kobaly et al., Reference Kobaly, Schluchter, Minich, Friedman, Taylor and Wilson-Costello2008; Larroque et al., Reference Larroque, Ancel, Marret, Marchand, Andre and Arnaud2008; Volpe, Reference Volpe2009). The encephalopathy of prematurity is diverse, but the most common manifestation is periventricular leukomalacia (PVL), although other neuronal/axonal deficits are also observed in the thalamus, basal ganglia, cerebral cortex, brainstem and cerebellum (Volpe, Reference Volpe2009).
Developmentally, white matter is most vulnerable to HI injury through midgestation and the preterm period, when premyelinating OLs (preOLs) are most prevalent (Kinney and Back, Reference Kinney and Back1998; Back et al., Reference Back, Luo, Borenstein, Levine, Volpe and Kinney2001). Multiple factors have suggested that the analogous developmental period to this window of susceptibility in the human is at the end of the first postnatal week in the rodent (Silbereis et al., 2010; Jensen, Reference Jensen2006; Talos et al., Reference Talos, Fishman, Park, Folkerth, Follett and Volpe2006a, Reference Talos, Follett, Folkerth, Fishman, Trachtenberg and Volpeb). In the rodent, the subcortical white matter and specifically oligodendrocytes (OLs) appear to be highly susceptible to injury between postnatal day (P) 5–7 (Follett et al., Reference Follett, Deng, Dai, Talos, Massillon and Rosenberg2004; Jantzie et al., Reference Jantzie, Cheung and Todd2005; Back et al., Reference Back, Riddle and McClure2007).
The mechanisms of injury to the premature brain are multifactorial, and depend on the maturational state of antioxidant defenses, inflammatory cascades, cerebral vasculature and cell-specific receptors for the excitatory neurotransmitter glutamate (Ferriero, Reference Ferriero2004; Folkerth et al., Reference Folkerth, Haynes, Borenstein, Belliveau, Trachtenberg and Rosenberg2004; Jensen, Reference Jensen2006; Volpe, Reference Volpe2009). HI causes increased accumulation of extracellular glutamate and initiates excitotoxic injury and cell death (Benveniste, Reference Benveniste1991; Follett et al., Reference Follett, Deng, Dai, Talos, Massillon and Rosenberg2004; Jensen, Reference Jensen2006; Manning et al., Reference Manning, Talos, Zhou, Selip, Park and Park2008). Oligodendrocytes (OLs) are the primary cell type in developing white matter and express ionotropic glutamate receptors (iGluRs) including N-methyl-d-aspartate (NMDA) and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, although these subtypes are also expressed on white matter microglia, astrocytes and neurons (Fern and Moller, Reference Fern and Moller2000; Follett et al., Reference Follett, Rosenberg, Volpe and Jensen2000, Reference Follett, Deng, Dai, Talos, Massillon and Rosenberg2004; Manning et al., Reference Manning, Talos, Zhou, Selip, Park and Park2008).
In addition to iGLuRs, OLs also express metabotropic GluRs (mGluRs) (Luyt et al., Reference Luyt, Varadi and Molnar2003; Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004), which are G-protein coupled and are classified into groups I–III based on signal transduction pathways and pharmacology (Byrnes et al., Reference Byrnes, Loane and Faden2009a). While iGluRs mediate fast glutamatergic synaptic transmission, mGluRs activate second messenger signaling cascades including those modulating glutamatergic transmission (Maiese et al., Reference Maiese, Chong and Li2005; Alexander and Godwin, Reference Alexander and Godwin2006; Byrnes et al., Reference Byrnes, Loane and Faden2009a).
mGluR expression is developmentally regulated on OLs in vitro with upregulated expression on immature and premyelinating OLs (preOLs) compared to mature OLs (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004; Luyt et al., Reference Luyt, Varadi, Durant and Molnar2006). mGluR1 and mGluR5 are variably expressed on OLs, astrocytes and endothelial cells (Maiese et al., Reference Maiese, Chong and Li2005; Ferraguti and Shigemoto, Reference Ferraguti and Shigemoto2006; Byrnes et al., Reference Byrnes, Loane and Faden2009a). Conversely, microglia express primarily mGluR5, which can regulate their activation (Biber et al., Reference Biber, Laurie, Berthele, Sommer, Tolle and Gebicke-Harter1999; Byrnes et al., Reference Byrnes, Stoica, Loane, Riccio, Davis and Faden2009b).
Relevant to PVL, activation of group I mGluRs with the broad spectrum agonist (1S,3R) 1-aminocyclopentane-trans-1,3,-dicarboxylic acid (ACPD) and the specific group I agonist (R,S)-3,5-dihydroxyphenylglycine (DHPG) attenuated kainate and oxygen-glucose deprivation (OGD)-induced preOL death in vitro. This protective effect of ACPD and DHPG on OLs occurred independently of endocytosis of AMPA/kainate receptors from the cell surface (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004), unlike what has been observed in cultured adult hippocampal neurons (Snyder et al., Reference Snyder, Philpot, Huber, Dong, Fallon and Bear2001). In contrast, the protective efficacy of group I mGluR agonists on OLs appeared to be through activation of PCKα leading to maintenance of intracellular glutathione and decreased production of reactive oxygen species (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004).
While OL excitotoxicity and oxidative stress have been shown to play a major role in the pathogenesis of PVL and other cerebral white matter disorders (Haynes et al., Reference Haynes, Folkerth, Keefe, Sung, Swzeda and Rosenberg2003; Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004; Gonsette, Reference Gonsette2008), the role of group I mGluRs located on OLs in vivo and in perinatal human tissue remains undefined. Here we evaluated the expression patterns of mGluR1 and mGluR5 on OLs in developing white matter in both rodent and human brain, and tested the protective efficacy of a CNS-penetrating mGluR agonist, ACPD, in a model of PVL.
OBJECTIVE
Based on previous results with in vitro OLs, we hypothesized that group I mGluRs were upregulated on immature OLs in situ in white matter during the preterm period, and that HI-mediated white matter injury is in part mediated by the expression of group I mGluRs on OLs. Thus, the present study investigated the spatial and temporal expression of group I mGluRs on OLs in rodent and human white matter during development. Further, we examined the protective efficacy of the CNS-penetrating mGluR agonist ACPD on white matter injury in a rodent model of PVL. These studies aimed to identify novel mechanisms of group I mGluRs in the white matter of premature brains in response to HI.
METHODS
All procedures presented were approved and in accordance with the guidelines set forth by the Institutional Animal Care and Use Committee at Children's Hospital Boston. Unless otherwise indicated, all chemicals and reagents were obtained from Sigma (St. Louis, MO, USA).
Animals and surgical procedures
We used a previously published model of HI cerebral injury induced by unilateral common carotid artery occlusion and systemic hypoxia in P6 Long-Evans rat pups (Follett et al., Reference Follett, Deng, Dai, Talos, Massillon and Rosenberg2004; Manning et al., Reference Manning, Talos, Zhou, Selip, Park and Park2008). At P6, male pups were anesthetized with ether. A small lateral skin incision was made on the anterior ventral surface of the neck, and the left common carotid artery was exposed, isolated and permanently ligated by cauterization. Following recovery, pups were placed in a hypoxia chamber held at 6% O2 balanced N2 for 1 h. Normothermia (32–34°C) was maintained in the hypoxia chamber with a heating pad. During hypoxia, pups were monitored for behavioral seizures/convulsions (pawing, nodding, myoclonic jerks, tonic–clonic head and limb movement, loss of posture), and seizure duration was recorded for each animal, in every ACPD treatment paradigm (Rakhade et al., Reference Rakhade, Zhou, Aujla, Fishman, Sucher and Jensen2008).
Assessment of protective efficacy of systemic ACPD treatment in a rat model of PVL
Rat pups received 100 µl injections of 5, 10 or 20 mg kg−1 of the broad spectrum group I/II mGluR agonist ACPD (Tocris), or phosphate buffered saline (PBS), intraperitoneally (i.p.) 1 h prior to carotid ligation (n = 10 per group). Additional doses of ACPD were then administered immediately following hypoxia and every 12 h for the first 48 h post-HI, consistent with the reported half-life of drugs in this class (Hodge et al., Reference Hodge, Miles, Sharko, Stevenson, Hillmann and Lepoutre2006; Carroll, Reference Carroll2008). Chemical modulators of mGluRs, such as ACPD and other glycine derivatives, have half-lives between 1 and 4 h and thus pre-treatment was required to ensure adequate drug delivery and absorption by the time cerebral injury was initiated during hypoxia. Ninety-six hours post-HI, pups were euthanized via lethal injection of sodium pentobarbital (100 mg kg−1, i.p.), and transcardially perfused with PBS, followed by 4% paraformaldehyde (PFA). Brains were extracted, post-fixed in PFA overnight and then cryoprotected in 30% sucrose at 4°C. Brains were then frozen and 16 µm sections were obtained using a cryostat. Sections were mounted on slides, and stored at −20°C until immunochemical analyses were performed.
Developmental profile of group I mGluRs in the immature rodent
The normal developmental expression of group I mGluRs were assessed in naïve, male rat pups. P3, P6, P9 and P14 (n = 3 per group) pups were euthanized and transcardially perfused as described above. Brains were extracted, post-fixed and cryoprotected, and then sectioned at 50 µm thickness for free-floating sections or 12 µm for slide mounted sections, and were subsequently processed for immunohistochemistry.
Developmental profile of group I mGluRs in human white matter
Human parietal–occipital lobe specimens were collected from the fetal, neonatal, pediatric and adult autopsy populations of Children's Hospital Boston and Brigham & Women's Hospital, Boston. Cases ranged from 19 post-conceptional weeks (PCW) to 57 years of age (n = 17; 12 males and 5 females). Additional cases were obtained from the University of Maryland Brain and Tissue Bank for Developmental Disorders (n = 4; three males and one female) (Table 1). The samples were obtained from standard diagnostic postmortem examinations, and all procedures and experiments were conducted under guidelines approved by the Clinical Research Committee at all institutions. The causes of death are listed in Table 1. We divided our study cases into those with no recognizable abnormalities on microscopic examination (n = 16), and those with minimal white matter changes (mild diffuse white matter gliosis [DWMG]) (n = 4). When possible, the postmortem interval was limited to ≤24 h.
Clinical characteristics of sample population*.
* ACA, acute chorioamnionitis; DWMG, diffuse white matter gliosis; GA, gestational age; GM, germinal matrix; IUGR, intrauterine growth restriction; MCA, multiple congenital anomalies; NEC, necrotizing enterocolitis; PCA, post-conceptional age; PH, pulmonary hypoplasia; PMI, postmortem interval; PNA, postnatal age; NA, information not available; M, male; F, female.
Immunochemical and histological analyses
Free-floating 50 or 12 µm slide mounted sections were processed and double labeled with antibodies against mGluR1 or mGluR5, in combination with cell-specific markers. Briefly, sections were rinsed and blocked for 2 h in a 5% normal goat serum solution. Primary antibodies against preOLs (anti-O4, 1:500, gift from S. Pfeiffer, University of Connecticut, Farmington, CT), immature OLs (anti-galactocerebroside C (GalC), 1:100, Millipore), neurons (anti-neuronal nuclei (NeuN), 1:500, Millipore), astrocytes (anti-glial fibrillary acidic protein (GFAP, SMI 22), 1:500, Covance), axons (anti-neurofilament 160 kDa, Millipore, 1:200), and activated microglia/macrophages (anti-CD68, 1:500, Serotec) were then applied and sections were allowed to incubate overnight at 4°C. The following day, sections were washed, and a species-appropriate fluorescent secondary antibody conjugated to Alexa-Fluor 568 (1:1000, Invitrogen) was applied for 1 h. Sections were then washed, blocked and incubated with antibodies against mGluR1 or mGluR5 (anti-mGluR1/5, 1:500, Millipore) overnight at 4°C. Subsequently, a biotinylated anti-rabbit secondary antibody was applied, followed by a fluorescent avidin conjugate. Sections were then washed and coverslipped with a Dapi containing mountant. Appropriate negative control (no primary antibody) and blocking peptide experiments were run in parallel and as per the procedure described above.
Brain sections were also evaluated for myelin basic protein (MBP) via immunohistochemistry. Serial coronal sections were blocked and incubated overnight with anti-MBP (1:1000, SMI-99, Sternberger Monoclonals). The following day, an anti-mouse Alexa-Fluor 488/568 secondary antibody was applied for 1 h. Slides were then rinsed and coverslipped. An investigator blinded to all treatment conditions assessed lesion size. Using a semiquantitative scale as previously described (Follett et al., Reference Follett, Rosenberg, Volpe and Jensen2000, Reference Follett, Deng, Dai, Talos, Massillon and Rosenberg2004; Manning et al., Reference Manning, Talos, Zhou, Selip, Park and Park2008), MBP immunostaining/white matter injury following HI was rated on a scale of 0 (no injury) to 5 (severe injury). These scores were assigned and compared with the white matter in the hemisphere contralateral to carotid ligation in three stereotactically similar sections per animal.
Western blot analyses
Western blot analyses were performed on tissue homogenates from regions of dissected human parietal-occipital white matter and cortex (Table 1). Adult cases were used as indices of maturity (n = 3) and as the standard for protein expression. After membrane protein extraction, a Bradford protein assay was performed to determine total protein amount, and 30 µg of protein/sample was loaded on precast 7.5% Tris glycine gels. Following transfer to polyvinyldifluoride membranes, membranes were blocked for 1 h in 5% non-fat dry milk in Tris buffered saline solution containing 0.1% Tween-20. Membranes were then incubated in mGluR5 primary antibody solution (1:500, Millipore) overnight at 4°C. An HRP conjugated anti-rabbit IgG secondary antibody was then applied for 1 h and after appropriate washes protein bands were visualized using SuperSignal West Femto Maximum Sensitivity ECL substrate. Relative optical density was measured for each band using FujiFilm software.
Statistical analyses
All data presented are expressed as mean ± standard error of the mean (SEM). ANOVA analysis was used for multiple comparisons. Normally distributed data between two independent groups were analyzed with Student's t-test. The general convention of a probability value of P < 0.05 was used to establish statistical significance.
RESULTS
Group I mGluRs are developmentally expressed on immature OLs in the neonatal rodent brain
Given our previous data indicating that mGluR1 and mGluR5 are expressed on developing OLs in vitro (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004), we first assessed cell specific expression of mGluR1 and mGluR5 on OLs, neurons, astrocytes and activated microglia in immature rats from P3 to P14. In OLs, the regional and subcellular localization varied with age. In white matter, mGluR1 was prominently localized on GalC-positive OL cell somas at earlier ages (P3–P6, Fig. 1A–F), while in later postnatal development (P14) mGluR1 was almost exclusively localized to cell processes (Fig. 1G–I). At P14, there appeared to be a ventral–dorsal gradient with the highest density of process staining localized to the gray:white junction, and on white matter fibers radiating through the lower layers of the cortex and subplate (Fig. 1G–I). mGluR5 was also maximally expressed on GalC-positive OL cell bodies from P3 to P6 in white matter (Fig. 2A–F). By P14, mGluR5/GalC colocalization was minimal, and limited to restricted patches of OL processes (Fig. 2G–I). Overall at P14, mGLuR5 labeling of OL processes was of lower density than that of mGluR1. Co-labeling with the O4 OL marker revealed similar patterns at all ages (data not shown).
Developmental expression of mGluR1 in rodent callosal white matter. At postnatal day 6 (P6), mGluR1 (E, green) is highly expressed on GalC-positive OL cell bodies and processes (red) in callosal white matter (B–F). mGluR1 (green) is also strongly co-expressed on GalC-positive OL cell bodies and processes (red) from P3 (A–C) through P14 (G–I).
Developmental expression of mGluR5 in rodent callosal white matter. At postnatal day 6 (P6), mGluR5 (E, green) is highly expressed on the cell bodies of GalC-positive OLs (red) in callosal white matter (D–F). In contrast to the consistent GluR1 expression pattern throughout development (Fig. 1), the peak of mGluR5 (green) on GalC-positive OLs (red) is between P3 (C) and P6 (F). By P14 (H), mGLuR5-GalC co-expression is reduced and restricted to patches of OL processes throughout the corpus callosum and is similar to that observed in older animals.
Both mGluR1 and mGluR5 were also developmentally regulated on non-OL components of developing white matter. At P6, in contrast to OLs that highly expressed both mGluR1 and mGluR5, group I mGluRs were minimally present on CD68-positive activated microglia and GFAP-positive cell bodies (see supplementary Fig. 1A–D online). However, with advancing age (P9–P14) both group I mGluR subunits were increasingly expressed on astrocytic processes, while mGluR5 expression increased on activated microglia. Interestingly, this increased expression of mGluR5 on astrocytes and microglia in the white matter in the second postnatal week is occurring during a time that this receptor is diminishing in expression on GalC-positive OLs.
Neuronal expression followed a different pattern with both mGluR1 and mGluR5 present in the cortex prior to P9, although at lower levels compared to OLs. Consistent with previous reports, group I mGluRs then gradually increased during the second postnatal week on neuropil and dendrites (data not shown) (Boer et al., Reference Boer, Encha-Razavi, Sinico and Aronica2010). In addition, both mGluR1 and mGluR5 were minimally present on neurofilament 160 kDa-positive axons in white matter and cortex at P6 (see supplementary Fig. 1E,F online), corroborating prior evidence of very low levels of group I mGluRs on axons and consistent with the literature reporting that the majority of mGluR1 and mGluR5 receptors (i.e. >90%) are located postsynaptically on dendritic spines (Romano et al., Reference Romano, Sesma, McDonald, O'Malley, Van Den Pol and Olney1995; Lujan et al., Reference Lujan, Roberts, Shigemoto, Ohishi and Somogyi1997; Olive, Reference Olive2009).
Administration of the group I/II mGluR agonist, ACPD, attenuates white matter injury in a rodent model of PVL
Given the high expression of mGluR1 and mGluR5 on developing OLs in rodent white matter throughout the first postnatal week, and the protective effects of mGluR agonists against OL injury in vitro (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004), we next examined the protective efficacy of ACPD against HI-mediated white matter injury in our P6 rat model of PVL. Administration of ACPD in four divided doses during P6–P8 significantly attenuated callosal and periventricular white matter loss as assessed by MBP protein expression. HI in vehicle-treated rats resulted in significant loss of MBP and unilateral decreases in white matter in the hemisphere ipsilateral to carotid ligation compared to MBP expression in the contralateral hemisphere (Fig. 3A,B; 3.6 ± 0.17 versus 2.2 ± 0.20, P < 0.0001). In contrast, the administration of ACPD at each of the three doses examined (5, 10 or 20 mg kg−1), significantly attenuated the loss of MBP and protected white matter when MBP injury was assessed 96 h post-HI. ACPD treatment preserved myelin processes radiating into the cortex and pericallosal process bundles compared with vehicle-treated controls (Fig. 3C–F). Semiquantitative scoring by a blinded investigator confirmed a significant protection of ACPD on white matter injury (5 mg kg−1: 1.4 ± 0.43 versus 0.5 ± 0.17, P < 0.001; 10 mg kg−1: 1.6 ± 0.4 versus 1.1 ± 0.31, P < 0.001; 20 mg kg−1: 1.7 ± 0.47 versus 1.1 ± 0.31, P = 0.001, Fig. 3G).
Administration of ACPD attenuates white matter loss after HI in postnatal day 6 (P6) Long-Evans rats. HI results in the significant loss of MBP in the hemisphere ipsilateral to carotid ligation (A), compared to MBP expression in the uninjured, contralateral hemisphere (B). Systemic administration of ACPD for 48 h significantly attenuates callosal and periventricular white matter loss 96 h following HI when administered at 5 mg kg−1 (C,D) or 20 mg kg−1 (E,F). Treatment with ACPD preserves myelin processes radiating into the cortex and pericallosal process bundles at each dose examined (C,E) compared to vehicle-treated controls, and significantly decreases overall MBP injury scores (G).
While ACPD appeared to be protective in developing white matter, this agent has been shown to be proconvulsant at older ages, when neuronal expression of group I mGluRs is higher in cortex. The convulsant effect of group I mGluR agonists (McDonald et al., Reference McDonald, Fix, Tizzano and Schoepp1993) has been attributed to presynaptic actions of glutamate (Zhong et al., Reference Zhong, Gerber, Kojic and Randic2000; Ure et al., Reference Ure, Baudry and Perassolo2006). We thus evaluated whether administration of ACPD was proconvulsant in our model of PVL in P6 Long-Evans rats. Importantly, there were no significant differences in either percentage of animal seized per treatment group (77% vehicle-treated HI animals versus 50% 5 mg kg−1 ACPD versus 70% 10 mg kg−1 ACPD versus 80% 20 mg kg−1 ACPD, P > 0.05) or cumulative seizure duration (19.7 min vehicle-treated HI animals versus 18.3 min 5 mg kg−1 ACPD versus 19.1 min 10 mg kg−1 ACPD versus 20.8 min 20 mg kg−1, P > 0.05). Consistent with our results showing low mGluR expression on neurons at this age, these data suggest that mGluRs do not significantly contribute to HI-induced seizure activity in this model.
Group I mGluRs are localized on immature OLs in human white matter
We next determined whether group I mGluRs were developmentally regulated in human brain tissue in the premature period. Using the O4 marker for immature OLs, we evaluated the expression of mGluRs in postmortem parietal–occipital lobe tissue from cases 19 to 38 PCW (Table 1). Throughout the investigated time course, mGluR1 was highly expressed on O4-positive OLs (Fig. 4). In contrast, mGluR5 was minimally expressed on OLs at 19 PCW (Fig. 5A–C), and then increased by 29 PCW (Fig. 5D–F), but returned to lower levels at 38 PCW (Fig. 5G–I). The transient upregulation of mGluR5 was similar to that observed in the rodent. Also corroborating rodent staining patterns, mGluR5 was expressed at very low levels on GFAP-positive astrocytes prior to and at 36 PCW, but then increased beginning at 38 PCW (see supplementary Fig. 2 online). In addition, neuronal expression of mGluR5 gradually increased with age from low levels in scattered cortical neurons during the preterm period to widespread expression in most neuronal cell bodies and dendrites at term birth (data not shown), consistent with published data (Byrnes et al., Reference Byrnes, Loane and Faden2009a; Boer et al., Reference Boer, Encha-Razavi, Sinico and Aronica2010) and similar in pattern to that seen in the rodent. Western blot analyses performed on human parietal–occipital white matter confirmed the transient increase in white matter mGluR5 expression (Fig. 6), whereby mGluR5 was strongly expressed in white matter from 20 to 38 PCW, compared to a marked postnatal decrease (76 PCW-7 years) and very low levels sustained through adulthood.
Developmental expression of mGluR1 on human OLs. Expression of mGluR1 (green) on O4-positive preOLs (red) in parietal–occipital white matter is consistent throughout gestation. O4-mGluR1 co-expression (yellow) is high through 19 post-conceptional weeks (PCW, A–C), 29 PCW (D–F) and 38 PCW (G–I).
Developmental expression of mGluR5 on human OLs. A biphasic expression pattern is observed with mGluR5 (green) on human O4-positive OLs (red) whereby co-expression (yellow) is low at 19 post-conceptional weeks (PCW, A–C) and 38 PCW (G–I). The peak of mGluR5 expression on O4-positive preOLs is observed at 29 PCW.
Developmental expression of mGluR5 in human subcortical white matter. Western blot analyses reveal that mGluR5 protein is strongly expressed in human subcortical, parietal–occipital white matter from 20 to 38 PCW, compared to a marked postnatal decrease (76 PCW-7 years) and very low levels sustained through adulthood.
CONCLUSIONS
• The expression of mGluR1 and mGluR5 in the white matter of neonatal rodents is developmentally regulated. mGluR1 is strongly expressed on GalC-positive OLs from P3 to P14, whereas mGLuR5 is maximally expressed on GalC-positive cell bodies from P3 to P6. In contrast, both mGluR1 and mGluR5 are minimally expressed on astrocytes, microglia, neurons and axons at this early stage of development, but increase during the second postnatal week. Hence, group I mGluRs are present primarily on OLs during the peak period of vulnerability to HI-mediated white matter injury.
• Patterns of mGluR1 and mGluR5 in human white matter appear to follow the same developmental trends as the rodent. While mGluR1 is expressed on O4-positive OLs throughout gestation (19–38 PCW), mGluR5 is maximally expressed on O4-positive OLs through the preterm period. With advancing gestational age, mGluR5 is increasingly expressed on cortical neurons, dendrites and GFAP-positive astrocytes in the white matter. This expression pattern validates group I mGluRs as targets on premyelinating OLs, and confirms group I mGluR expression during the peak period of vulnerability to HI-mediated white matter injury.
• The systemically administrable mGluR agonist ACPD significantly attenuated white matter injury in a rat model of PVL, during a period of high mGluR expression on preOLs. These data are consistent with prior evidence of protective efficacy of selective group I agonists in in vitro models of preOL injury. Given the primary expression pattern of group I mGluRs on OLs described above in the immature brain, these data suggest that this protective efficacy is at least in part related to direct modulation of mGluRs on OLs.
DISCUSSION
There is substantial literature regarding the developmental regulation of group I mGluRs on neurons and their role in synaptic plasticity, synaptogenesis and neuronal excitability (Di Giorgi Gerevini et al., Reference Di Giorgi Gerevini, Caruso, Cappuccio, Ricci Vitiani, Romeo and Della Rocca2004; Bear, Reference Bear2005; Catania et al., Reference Catania, D'Antoni, Bonaccorso, Aronica, Bear and Nicoletti2007), as well as their role in the pathogenesis of glutamate-mediated cell death in disorders of the adult brain, including epilepsy, stroke, chronic pain, schizophrenia and Alzheimer's disease (Catania et al., Reference Catania, D'Antoni, Bonaccorso, Aronica, Bear and Nicoletti2007; Luscher and Huber, Reference Luscher and Huber2010; Ribeiro et al., Reference Ribeiro, Paquet, Cregan and Ferguson2010). Their relevance as a therapeutic target has been suggested as group I mGluR agonists have been shown to be protective in models of stroke in adult rodent brain (Bao et al., Reference Bao, Williams, Faden and Tortella2001; Kelland and Toms, Reference Kelland and Toms2001).
However, the role of mGluRs in developmental disorders in the immature brain or their expression on non-neuronal cell populations has not been completely described in vivo. Here we report for the first time that the systemically administrable group I/II mGluR agonist, ACPD, protects against white matter injury in an animal model of PVL. Importantly, both mGluR1 and mGluR5 are prominently expressed on immature OLs in rodent and human subcortical developing white matter. The protection in vivo is consistent with our prior findings of a direct protective effect of mGluR agonists on OLs in vitro. Given the presence of group I mGluRs on human OLs, this protective strategy may have future clinical potential in the treatment of perinatal HI-mediated white matter injury.
Similar patterns of expression of group I mGluRs on OLs in developing rodent and human subcortical white matter
It is well established that iGluRs are expressed on all cell populations in the developing white matter of both rodents and humans (Jensen, Reference Jensen2006; Talos et al., Reference Talos, Fishman, Park, Folkerth, Follett and Volpe2006a, Reference Talos, Follett, Folkerth, Fishman, Trachtenberg and Volpeb). The developmental expression of iGluRs critically regulates neuronal, synaptic and glial differentiation (Butt, Reference Butt2006). It has been postulated that glutamate also modulates myelination by OLs, although the mechanism whereby iGluRs might regulate this process is unclear (Karadottir and Attwell, Reference Karadottir and Attwell2007; Wake et al., Reference Wake, Lee and Fields2011). However, the transient upregulation of these receptors during development render cells vulnerable to excitotoxicity in diseases characterized by excessive glutamate release (McDonald and Johnston, Reference McDonald and Johnston1990; Jensen, Reference Jensen2006).
Group I mGluRs are also of interest in HI injury, because they are present on many cell types in the brain and their activation may influence outcome, as related to modulation of downstream protein kinases involved in cell growth, differentiation and survival, and their direct effects on endothelium, neuroinflammation and apoptosis (Luyt et al., Reference Luyt, Varadi, Durant and Molnar2006; Catania et al., Reference Catania, D'Antoni, Bonaccorso, Aronica, Bear and Nicoletti2007; Byrnes et al., Reference Byrnes, Loane and Faden2009a; Ribeiro et al., Reference Ribeiro, Paquet, Cregan and Ferguson2010). Indeed, we found both mGluR1 and mGluR5 to be developmentally regulated on glial cells in rodent and human white matter. mGluR1 was persistently present on immature rodent and human OLs, while mGluR5 was transiently increased throughout this time course of selective vulnerability to injury. The transient overexpression of mGluR5 on OLs in human and rodent white matter in the preterm period may promote OL survival, as mGluR5 has been shown to be important for OL proliferation, migration, neuron-glia signaling and the prevention of apoptosis (Luyt et al., Reference Luyt, Varadi, Durant and Molnar2006). We observed mGluR1 and mGluR5 on OL processes and cell bodies throughout the first two postnatal weeks and in the premature human brain. While mGluRs on OLs may be the primary target of the ACPD effects observed here, we document expression of mGluR1 and mGluR5 on astrocytes and microglia, although at lower levels, throughout this period of susceptibility. mGluR5 receptors are upregulated on reactive astrocytes in ALS, MS and epilepsy, and it is hypothesized that the interaction between OLs, astrocytes and microglia influence the response mediated by mGluRs under pathologic conditions characterized by neuroinflammation (D'Antoni et al., Reference D'Antoni, Berretta, Bonaccorso, Bruno, Aronica and Nicoletti2008; Drouin-Ouellet et al., Reference Drouin-Ouellet, Bownell, Saint-Pierre, Fasano, Emond and Trudeau2011). In culture, mGluR activation modulates microglial reactivity and decreases the production of proinflammatory mediators, enhancing OL progenitor survival (Farso et al., Reference Farso, O'Shea and Beart2009; Taylor et al., Reference Taylor, Pirianov, Holland, McGinnity, Norman and Reali2010). Specific to the developing rodent brain, neuroinflammation induces mGluR5 expression on astrocytes and microglia, and administration of minocycline significantly protects against ibotenic acid-induced excitotoxicity (Drouin-Ouellet et al., Reference Drouin-Ouellet, Bownell, Saint-Pierre, Fasano, Emond and Trudeau2011). Interestingly, we have also shown that doxycycline and minocycline inhibit microglial activation and protect neurons and OLs from HI in neonatal rodents (Lechpammer et al., Reference Lechpammer, Manning, Samonte, Nelligan, Sabo and Talos2008; Jantzie and Todd, Reference Jantzie and Todd2010). Hence, while this report focused on the actions of group I mGluRs on OLs, there is likely to be other important mechanisms involving microglia and astrocytes.
Protective efficacy of ACPD in an animal model of PVL
A novel finding in this study is the protective effect of group I mGluR agonists in models of white matter injury in vivo. Prior studies using premature HI injury models have shown that glutamate excitotoxicity is mediated by iGluRs present on preOLs in immature white matter, and blockade of the AMPA or NMDA subtypes of iGluRs protects the white matter when administered after HI to neonatal rats (Follett et al., Reference Follett, Rosenberg, Volpe and Jensen2000, Reference Follett, Deng, Dai, Talos, Massillon and Rosenberg2004; Manning et al., Reference Manning, Talos, Zhou, Selip, Park and Park2008; Liu et al., Reference Liu, Lin, Wu and Qiu2009). Given the prominent role glutamate appears to play in this pathophysiology, we reasoned that mGluRs may also modulate preOL injury. Here, importantly, ACPD was protective and significantly attenuated callosal and periventricular white matter loss after preterm HI. The in vivo protection observed here is consistent with our previous finding that the specific mGluR1 agonist DHPG was protective against OGD and kainate-induced excitotoxicity, and that developing OLs in vitro express both mGluR1 and mGluR5 (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004). In that study, no protection was seen with DCG-IV or L-AP4, group II or III specific mGluR agonists, indicating the protective effects on OL injury were primarily related to actions on group I mGluRs. Specifically, ACPD and DHPG significantly protected preOLs, and the protective effects were reversed by (S)-methyl-4-carboxyphenylglycine (MCPG, a broad-spectrum mGluR antagonist) (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004). In addition, the protective effects of DHPG were partially prevented by 7-hydroxyiminocyclopropanchromen-1a-carboxylic acid ethyl ester (CPCCOEt, selective mGluR1 antagonist) or 2-methyl-6-(phenylethynyl)-pyridine (MPEP, selective mGluR5 antagonist) and fully abolished by the combination of the two agents, whereas the antagonists per se (MCPG, CPCCOEt, MPEP or CPCCOEt plus MPEP) had no effect (Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004). These data in combination with our in vivo study here indicate that ACPD was effective via modulation of group I mGluRs with agonists. While mGluR agonists are protective in our model, there have been no reports of the opposite effects of mGluR antagonists. MPEP administration to preOLs in vitro or to naïve prepubertal rodents has not been reported to induce cell death or white matter injury (Luyt et al., Reference Luyt, Varadi and Molnar2003; Deng et al., Reference Deng, Wang, Rosenberg, Volpe and Jensen2004; Lojkova and Mares, Reference Lojkova and Mares2005).
The protective effects of ACPD were maximal at the lowest dose examined, without additional benefit at higher doses, and the timing of ACPD administration corresponded with peak levels of mGluR1 and mGluR5 expression on OLs during the preterm period, the peak period of OL cell death following HI, and the pharmacokinetics of glycine-derived mGluR modulators (Hodge et al., Reference Hodge, Miles, Sharko, Stevenson, Hillmann and Lepoutre2006; Carroll, Reference Carroll2008). This developmental expression afforded the ability to preferentially target OLs during their peak vulnerability to HI, prior to the window of maximal neuronal mGluR expression, and likely avoiding proconvulsant effects (McDonald et al., Reference McDonald, Fix, Tizzano and Schoepp1993). Given the prior literature in P7 Sprague Dawley rats (McDonald et al., Reference McDonald, Fix, Tizzano and Schoepp1993), it was prudent to examine behavioral seizure activity in our model, as it is well established that P6 Long-Evans rats have seizures during hypoxia. However, as reported, seizure activity was not altered by ACPD. Of note, Sprague Dawley rats are 2 days more mature than Long-Evans rats, due to a longer gestational period (Talos et al., Reference Talos, Fishman, Park, Folkerth, Follett and Volpe2006a, Reference Talos, Follett, Folkerth, Fishman, Trachtenberg and Volpeb; Dean et al., Reference Dean, Moravec, Grafe, Abend, Ren and Gong2011), and hence were significantly older than those in our study (approximately analogous to a P9–P11 Long-Evans rat). Further, McDonald and colleagues tested ACPD at doses outside our range of 5–10 mg kg−1. Given the primary expression of group I mGluRs on OLs at P6 and markedly lower expression levels on neurons, astrocytes and microglia at this time point, our results indicate that ACPD's protective effect is in part related to direct modulation of mGluRs on OLs.
In conclusion, the selective expression of mGluRs on OLs in premature white matter may represent an age-specific therapeutic target. Their developmental expression on both rodent and human OLs may also be a subject for further studies of neuron–glia and glia–glia interactions (Maiese et al., Reference Maiese, Chong and Li2005).
ACKNOWLEDGEMENTS
The authors are very grateful for the assistance provided by Bela Kosaras M.D., Jocelyn Lippman Bell Ph.D., Mirna Lechpammer M.D., Ph.D. and the support contributed by the Hearst Foundation (L.L.J.), the Heart and Stroke Foundation of Canada (L.L.J.), the Alberta Heritage Foundation for Medical Research (L.L.J.), PO1-NS038475 (F.E.J.), RO1-NS31718 (F.E.J.) and DP1-OD003347 (F.E.J.). The authors are also very grateful for the human tissue obtained from the NICHD Brain and Tissue Bank for Developmental Disorders at the University of Maryland, Baltimore, MD.
Statement of interest
None.
Supplementary material
The supplementary material referred to in this article can be found online at journals.cambridge.org/ngb.
