Search

Brain injury during the last trimester to the first 1-4 years in humans is now thought to trigger an array of intellectual and emotional problems later in life, including disorders such as schizophrenia. In adult schizophrenic brains, there is a specific loss of neurons that co-express glutamic acid decarboxylase-parvalbumin (GAD67-PV). Loss of this phenotype is thought to occur in mature animals previously exposed to N-methyl-D: -aspartate receptor (NMDAR) antagonists during late gestation or at postnatal day 7 (P7). However, in similarly treated animals, we have previously shown that GAD67 and PV are unaltered in the first 24 h. To more precisely define when changes in these markers first occur, we exposed rat pups (P7 or P6-P10) to the NMDAR antagonist MK801 and at P11 co-stained brain sections for GAD67 or PV. In the cingulate cortex, we found evidence for a reduction in PV (GAD67 levels were very low to undetectable). In contrast, in the somatosensory cortex, we found that expression of GAD67 was reduced, but PV remained stable. Further, repeated but not single doses of MK801 were necessary to see such changes. Thus, depending on the region, NMDAR antagonism appears to influence expression of PV or GAD67, but not both. These observations could not have been predicted by previous studies and raise important questions as to how the GAD67-PV phenotype is lost once animals reach maturity. More importantly, such differential effects may be of great clinical importance, given that cognitive deficits are seen in children exposed to anesthetics that act by blocking the NMDAR.

N-methyl-D-aspartate receptor blockade promotes apoptosis at postnatal day 7 (P7) and is linked to loss of glutamic acid decarboxylase 67 (GAD67) expression in older animals. To more fully appreciate this relationship we must first understand how GAD67 is regulated postnatally. Thus, the brains of P7, P14 and P21 rats were examined for expression of GAD67 protein and we found that levels of this GABAergic marker increased steadily with age, such that by P21 there was as much as a 6-fold increase compared to P7 animals and a 1.5- to 2-fold increase compared to P14 animals, depending on the region sampled. At P7, GAD67 was almost exclusively detected in puncta, with very few cell bodies displaying this marker. In contrast, at P14 and especially P21, both puncta and cell bodies were robustly labeled. Our data indicate that adult-like expression of GAD67 emerges quite late in the postnatal period.

The acoustic startle reflex in rats can be inhibited if a prepulse stimulus is presented just before the startle stimulus (prepulse inhibition; PPI). When postnatal day 7 (P7) rats are exposed to agents that block the NMDA receptor (NMDAR), robust apoptosis is observed within hours and is thought to be followed at later ages by a significant loss of PPI. To understand these observations further, we exposed rat pups to vehicle or the NMDAR antagonist MK801 (1 mg/kg) at P6, P8, and P10. We then examined animals for PPI at P28 and P56. Compared to vehicle controls, we found no evidence for PPI deficits in the MK801-treated group, although we did observe prepulse-induced delay in response time at P56 (but not at P28). In a parallel study, we also performed histological analysis of brain sections for evidence of the pro-apoptotic marker activated caspase-3, 8 hours after vehicle or MK801 injection into P6 animals. We found that there was a robust increase in this marker of cell death in the inferior colliculus of MK801 compared to vehicle treated animals. Thus, transient blockade of the NMDAR during the postnatal period not only promotes early apoptosis in a brain region critical for acoustic processing but also leads to auditory deficits at a later age, suggesting that injury-induced loss of collicular neurons leads to network reorganization in the auditory system that is progressive in nature.

We have previously shown that P7 rat pups injected with the N-methyl- d -aspartate receptor (NMDAR) blocker MK801 displayed robust apoptotic injury within hours after injection. Further studies from our lab suggest that loss of calcium cannot be compensated for when vulnerable neurons lack calcium buffering capabilities. Thus, to elevate calcium in these neurons prior to MK801 exposure, we injected P7 rats with the calcium channel agonist BayK 8644. Whereas BayK 8644 did not induce apoptosis by itself, it was found to block MK801-induced injury in a dose-dependent manner. Reversal of MK801 toxicity was complete in the caudate–putamen, partial in the somatosensory cortex but was not observed in the retrosplenial cortex. These results suggest that postnatal brain injury resulting from agents that block the NMDAR, which include commonly used anesthetics as well as drugs of abuse, may be prevented in vulnerable neurons by compensatory increases in calcium prior to exposure to these antagonists.

An abundance of evidence exists that shows calcium channel blockade promotes injury in cultured neurons. However, few studies have addressed the in vivo toxicity of such agents. We now show that the L-type calcium channel antagonist nimodipine promotes widespread and robust injury throughout the neonatal rat brain, in an age-dependent manner. Using both isolated neuronal as well as brain slice approaches, we address mechanisms behind such injury. These expanded studies show a consistent pattern of injury using a variety of agents that lower intracellular calcium. Collectively, these observations indicate that postnatal brain development represents a transitional period for still developing neurons, from being highly sensitive to reductions in intracellular calcium to being less vulnerable to such changes. These observations directly relate to current therapeutic strategies targeting neonatal brain injury.

Schizophrenia, a progressive disorder displaying widespread pathological changes, is associated with the loss of glutamatergic function and selective loss of cytoskeletal proteins, such as MAP2, in regions severely affected by this disease. As schizophrenia is associated with perinatal brain trauma, we monitored changes in several functionally different proteins following injury-promoting MK801 blockade of N-methyl-D-aspartate receptors in neonatal rats. Within the somatosensory cortex, MK801 triggered robust, caspase-3-dependent apoptotic injury, reduced expression of cytoskeletal proteins MAP2 and tau, and increased synapse associated protein SNAP25. Thus, both neuronal injury and loss of structural elements important for successful cell-cell contact may reorganize brain circuitry, which at later ages could promote similar behavioral changes observed in schizophrenia.

To identify binding partners of the A1AR (A1 adenosine receptor), yeast two-hybrid screening of a rat embryonic cDNA library was performed. This procedure led to the identification of erythrocyte membrane cytoskeletal protein (represented as 4.1G) as an A1AR-binding partner. Truncation studies revealed that the C-terminal domain of 4.1G was essential for binding to A1ARs and that the C-terminal domain of 4.1G and the third intracellular loop of A1ARs interacted. A1AR–4.1G interaction was also confirmed in studies using brain tissue. Studies in HEK-293 (human embryonic kidney 293) cells and Chinese-hamster ovary cells showed that 4.1G interfered with A1AR signal transduction, as 4.1G reduced A1AR-mediated inhibition of cAMP accumulation and intracellular calcium release. 4.1G also altered cell-surface A1AR expression. These observations identify 4.1G as a novel A1AR-binding partner that can regulate adenosine action.

The neurotransmitter glutamate influences intracellular Ca(2+) levels and plays an essential role in maintaining neuronal viability during early development. Blockade of NMDA receptors induces cell death in the neonatal forebrain via mechanisms that are not understood. Other neuromodulators that can influence intracellular Ca(2+) levels include the nucleoside adenosine, which acts via A(1) adenosine receptors subtypes (A(1)ARs). Because A(1)AR activation inhibits glutamate release and action, A(1)AR activation may also contribute to neonatal brain injury. To examine this possibility, we treated primary neuronal cultures with the A(1)AR agonist CPA, the NMDAR antagonist MK801, or CPA + MK801. Combined MK801 + CPA treatment resulted in profound cellular injury, exceeding that seen in other groups. In keeping with the hypothesis that altered Ca(2+) signaling mediates CPA + MK801 injury, reduction of Ca(2+) levels with EGTA, thapsigargin, or BAPTA-AM enhanced CPA + MK801-induced neuronal damage. In contrast, increasing intracellular Ca(2+) using ionomycin reversed CPA + MK801 toxicity. Direct visualization of intracellular Ca(2+) by confocal microscopy revealed that CPA + MK801 inhibited KCl-evoked increases in intracellular Ca(2+). Supporting the concept that A(1)AR activation and NMDAR blockade results in brain injury, neonatal rats injected with A(1)AR agonists + MK801 showed widespread apoptosis in many brain regions. These observations show that A(1)AR activation and NMDAR blockade lead to early postnatal cell injury by mechanisms that involve inhibition of intracellular Ca(2+) signaling.

Studies using animal models have shown that general anesthetics such as ketamine trigger widespread and robust apoptosis in the infant rodent brain. Recent clinical evidence suggests that the use of general anesthetics on young children (at ages equivalent to those used in rodent studies) can promote learning deficits as they mature. Thus, there is a growing need to develop strategies to prevent this injury. In this study, we describe a number of independent approaches to address therapeutic intervention. Postnatal day 7 (P7) rats were injected with vehicle (sterile PBS) or the NMDAR antagonist ketamine (20 mg/kg). After 8 h, we prepared brains for immunohistochemical detection of the pro-apoptotic enzyme activated caspase-3 (AC3). Focusing on the somatosensory cortex, AC3-positive cells were then counted in a non-biased stereological manner. We found AC3 levels were markedly increased in ketamine-treated animals. In one study, microarray analysis of the somatosensory cortex from ketamine-treated P7 pups revealed that expression of activity dependent neuroprotective protein (ADNP) was enhanced. Thus, we injected P7 animals with the ADNP peptide fragment NAPVSIPQ (NAP) 15 min before ketamine administration and found we could dose-dependently reverse the injury. In separate studies, pretreatment of P6 animals with 20 mg/kg vitamin D3 or a nontoxic dose of ketamine (5 mg/kg) also prevented ketamine-induced apoptosis at P7. In contrast, pretreatment of P7 animals with aspirin (30 mg/kg) 15 min before ketamine administration actually increased AC3 counts in some regions. These data show that a number of unique approaches can be taken to address anesthesia-induced neurotoxicity in the infant brain, thus providing MDs with a variety of alternative strategies that enhance therapeutic flexibility.

General anesthetics have long been thought to be relatively safe but recent clinical studies have revealed that exposure of very young children (4 years or less) to agents that act by blocking the N-methyl-D-aspartate receptor (NMDAR) can lead to cognitive deficits as they mature. In rodent and non-human primate studies, blockade of this receptor during the perinatal period leads to a number of molecular, cellular and behavioral pathologies. Despite the overwhelming evidence from such studies, doubt remains as to their clinical relevance. A key issue is whether the primary injury (apoptotic cell death) is specific to receptor blockade or due to non-specific, patho-physiological changes. Principal to this argument is that loss of core body temperature following NMDAR blockade could explain why injury is observed hours later. We therefore examined the neurotoxicity of the general anesthetic ketamine in P7, P14 and P21 rats while monitoring core body temperature. We found that, at P7, ketamine induced the pro-apoptotic enzyme activated caspase-3 in a dose-dependent manner. As expected, injury was greatly diminished by P14 and absent by P21. However, contrary to expectations, we found that core body temperature was not a factor in determining injury. Our data imply that injury is directly related to receptor blockade and is unlikely to be overcome by artificially changing core body temperature.

Blockade of the N-methyl-D-aspartate receptor (NMDAR) in postnatal day 7 (P7) rats can promote rapid and robust induction of the pro-apoptotic marker activated caspase-3 (AC3) and loss of the GABAergic marker GAD67 at P56. Thus, we hypothesized that NMDAR blockade-induced AC3 occurs in GAD67 positive cells at P7. To test this idea, we injected P7 rat pups with vehicle or MK801 and after 8 hours (peak of AC3 induction) we examined brain sections for both AC3 and GAD67. Compared to vehicle, MK801 profoundly induced AC3 in all brain regions examined but co-expression of GAD67 in the same cells was not observed. However, in brain regions where punctate (synaptic) GAD67 was abundant (for example, layer IV of the somatosensory cortex), AC3 was robust. These data suggest that whereas somatic expression of AC3 and GAD67 may be non-overlapping, areas that exhibit punctate GAD67 (and are high in synaptic turnover) may be more vulnerable to MK801 exposure.

MK801-induced activation of caspase-3 is developmentally regulated, peaking at postnatal day (P) 7 and decreasing with increasing postnatal age thereafter. Further, at P7, cells displaying activation of caspase-3 lack expression of calcium binding proteins (CaBPs). To further explore this relationship, we investigated postnatal expression of calbindin (CB), calretinin (CR) and parvalbumin (PV) in two brain regions susceptible to MK801-induced injury, the somatosensory cortex (S1) and layer II/III of motor cortex (M1/M2). Expression of CB and especially PV was low to absent prior to P7 but substantially increased from P7 through to P21 and adulthood. In contrast, CR expression was more variable at early developmental ages, stabilized to lower levels after P7 and showed a marked decline by P21. The results suggest that not only does calcium buffering capacity increase developmentally but also acquisition of enhanced buffering may be one mechanism by which neurons survive agent-induced alterations in calcium homeostasis.

Glutamate and norepinephrine (NE) are believed to mediate the long-lasting synaptic plasticity in the accessory olfactory bulb (AOB) that underlies pheromone recognition memory. The mechanisms by which these neurotransmitters bring about the synaptic changes are not clearly understood. In order to study signals that mediate synaptic plasticity in the AOB, we used AOB neurons in primary culture as a model system. Because induction of pheromone memory requires coincident glutamatergic and noradrenergic input to the AOB, and requires new protein synthesis, we reasoned that glutamate and NE must induce gene expression in the AOB. We used a combination of agonists that stimulate alpha1 and alpha2 adrenergic receptors in combination with N-methyl-d-aspartic acid and tested expression of the immediate-early gene (IEG) c-Fos. We found that the glutamatergic and noradrenergic stimulation caused significant induction of c-Fos mRNA and protein. Induction of c-Fos was significantly reduced in the presence of inhibitors of protein kinase C, mitogen-activated protein kinase (MAPK) and phospholipase C. These results suggest that glutamate and NE induce gene expression in the AOB through a signaling pathway mediated by protein kinase C and MAPK.

It has become increasingly clear that agents that disrupt calcium homeostasis may also be toxic to developing neurons. Using isolated primary neurons, we sought to understand the neurotoxicity of agents such as MK801 (which blocks ligand-gated calcium entry), BAPTA (which chelates intracellular calcium), and thapsigargin (TG; which inhibits the endoplasmic reticulum Ca(2+)-ATPase pump). Thus, E18 rat cortical neurons were grown for 1 day in vitro (DIV) and then exposed to vehicle (0.1% DMSO), MK801 (0.01-20 microM), BAPTA (0.1-20 microM), or TG (0.001-1 microM) for 24 h. We found that all three agents could profoundly influence early neuronal maturation (growth cone expansion, neurite length, neurite complexity), with the order of potency being MK801BAPTATG. We next asked if cultures exposed to these agents were able to re-establish their developmental program once the agent was removed. When we examined network maturity at 4 and 7 DIV, the order of recovery was MK801BAPTATG. Thus, mechanistically distinct ways of disrupting calcium homeostasis differentially influenced both short-term and long-term neuronal maturation. These observations suggest that agents that act by altering intracellular calcium and are used in obstetrics or neonatology may be quite harmful to the still-developing human brain.

Age-dependent, MK801-induced, activated caspase-3 expression in the postnatal brain is generally not observed in neurons expressing calcium-binding proteins (CaBPs), suggesting that apoptosis and calcium buffering are inversely related. In regions such as the cingulate and retrosplenial cortex, injury peaks at postnatal Day 7 (P7) and rapidly diminishes thereafter, whereas expression of calbindin (CB) and calretinin (CR) was relatively low from P0 to P7 and steadily increased from P7 to P14. At ages thereafter, CB and CR expression either remained stable then declined or rapidly declined. Parvalbumin (PV) was generally low-absent prior to P7 but expression dramatically increased from P10 onwards, peaking at P21. These studies suggest calcium entry (through N-methyl-D-aspartate receptor (NMDARs)) and buffering (by CaBPs) are integral to normal CNS maturation. Because schizophrenia is associated with glutamate hypo-function, developmental injury, and aberrant CaBP expression, our data indicate that this postnatal brain injury model may offer important insights into the nature of this disorder.

Loss of neuronal calcium is associated with later apoptotic injury but observing reduced calcium and increased apoptosis in the same cell would provide more definitive proof of this apparent correlation. Thus, following exposure to vehicle or the calcium chelator, BAPTA (1-20 microM), primary cortical neurons were labeled with Calcium Green-1 which was then cross-linked with EDAC, prior to immuno-staining for various proteins. We found that BAPTA-induced changes in calcium were highly correlated with changes in expression of activated caspase-3 as well as the calcium binding proteins calbindin, calretinin, and parvalbumin. Additionally, in brain slices from P7 neonatal rats, BAPTA induced significant loss of calcium in a brain region we have previously shown to express only moderate levels of calcium binding proteins as well as display robust apoptosis following calcium entry blockade. In contrast, BAPTA had little influence on calcium levels in a brain region we have previously shown to express robust calcium binding proteins as well as display far less apoptosis following calcium entry blockade. These data suggest that the ability of developing neurons to buffer changes in calcium may be critical to their long-term survival.

Age-dependent, neuronal apoptosis following N-methyl-D-aspartate receptor blockade has been linked to loss of calcium. To further explore this relationship, we examined expression of activated caspase-3, as well as the calcium binding proteins, calbindin-D 28K, calretinin and parvalbumin, following injection of vehicle or the N-methyl-D-aspartate receptor blocker, MK801, in postnatal day 7 or 21 rats. At postnatal day 7, MK801-induced activated caspase-3 expression was most frequently found in mutually exclusive cell populations to those expressing any of the three calcium binding proteins. For example, in the somatosensory cortex, most immunoreactivity for activated caspase-3 was found in layers IV/V, layered between areas of high calbindin or calretinin expression. Further, in the caudate putamen, activated caspase-3 rarely invaded zones of intense calbindin immunoreactivity. Suggesting expression patterns of these proteins were inversely related, these same brain regions no longer displayed MK801-induced activated caspase-3 at postnatal day 21, but instead robustly expressed calcium binding proteins. This later surge in expression was especially true for parvalbumin in regions such as the somatosensory and retrosplenial cortex, as well as the subicular complex. Calbindin-D 28K was also found to increase in the same regions though not as impressively as parvalbumin. Thus, developmental regulation of calcium binding protein expression may be a critical factor in age-dependent sensitivity to agents that disrupt calcium homeostasis in maturing neurons, providing a possible mechanistic explanation for age-dependent MK801 toxicity.

The cellular mechanisms that lead to neuronal death following glucose deprivation are not known, although it is recognized that hypoglycemia can lead to perturbations in intracellular calcium ([Ca 2 + ] i ) levels. Recently, activation of At adenosine receptors (A 1 AR) has been shown to alter [Ca 2 + ] i and promote neuronal death. Thus, we examined if A 1 AR activation contributes to hypoglycemia-induced neuronal injury using rat cortical neurons. First, we observed that hypoglycemia was associated with large increases in neuronal adenosine release. Next, decreased neuronal viability was seen with progressive reduction in glucose concentration (25, 6, 3, 0.75 and 0 mM). Using the calcium-sensitive dye, Fluo-3, we observed both acute and long-term changes in relative [Ca 2 + ] i during hypoglycemic conditions. Demonstrating a role for adenosine in this process, both the loss in neuronal viability and the early changes in [Ca 2 + ] i were reversed by treatment with A 1 AR antagonists (8-cyclopentyl, 1,3-dipropylxanthine; 9-chloro-2-(2-furyl)(1,2,4)-triazolo(1,5-c)quinazolin-5-amine; and N-cyclopentyl-9-methyladenine). We also found that hypoglycemia induced the expression of the pro-apoptotic enzyme, caspase-3, and that A 1 AR antagonism reversed hypoglycemia-induced caspase-3 activity. Collectively, these data show that hypoglycemia induces A 1 ARs activation leading to alterations in [Ca 2 + ] i , which plays a prominent role in leading to hypoglycemia-induced neuronal death.

Periventricular leukomalacia is characterized by a reduction in brain matter and secondary ventriculomegaly and is a major cause of developmental delay and cerebral palsy in prematurely born infants. Currently, our understanding of the pathogenesis of this condition is limited. In animal models, features of periventricular leukomalacia can be induced by hypoxia and activation of A 1 adenosine receptors (A 1 ARs). Using mice that are deficient in the A 1 AR gene ( A 1 AR –/–), we show that A 1 ARs play a prominent role in the development of hypoxia-induced ventriculomegaly in neonates. Supporting a role for adenosine in the pathogenesis of developmental brain injury, ventriculomegaly was also observed in mice lacking the enzyme adenosine deaminase, which degrades adenosine. Thus, adenosine acting on A 1 ARs appears to mediate hypoxia-induced brain injury ventriculomegaly during early postnatal development.

Few signaling molecules have the potential to influence the developing mammal as the nucleoside adenosine. In contrast to most neurotransmitters, adenosine is released by all cells and is present in all tissues. The adenosinergic system is therefore not dependent on the presence of mature synaptic structures or an intact autonomic nervous system for its release. However, similar to other signaling molecules, adenosine levels are dynamically regulated and increase with increased tissue activity, hypoxia, or stress. Local adenosine concentrations thus provide a “humoral barometer” of acute changes in cellular physiology. The receptors that transduce adenosine action include A 1 , A 2a , A 2b , and A 3 adenosine receptors. These receptors differ in their affinities for adenosine and in patterns of tissues expression. During development A 1 adenosine receptors (A1ARs) are especially important, and A1ARs are among the earliest receptors expressed in the embryonic brain and heart. In the developing heart, the adenosinergic system is the dominant regulator of fetal cardiac function and A1AR activation inhibits cardiac cell division leading to cardiac hypoplasia. In the forming central nervous system, A1AR activation potently inhibits the development of axons and can lead to leukomalacia. These recent data suggest that adenosine is an important modulator of mammalian development.

The initial aim of this study was to determine if the HSP70 (the main inducible heat shock protein), HO-1 (heme oxygenase-1, HSP32) and HSP47 (a collagen chaperone) stress proteins were induced in the same focal regions of rat brain following experimental subarachnoid hemorrhage (SAH). The next objective was to determine whether anti-oxidants prevented the stress gene expression in the focal regions. Lysed blood (150 microliter) was injected into the subarachnoid space of adult, female Sprague-Dawley rats via the cisterna magna. Animals were sacrificed 24 h later. Immunocytochemistry showed focal regions of stress gene induction in most animals (13/21), HSP70 and HO-1 proteins being expressed in neurons, microglia and astrocytes and HSP47 being expressed in microglia. Co-induction of the same three stress proteins was observed in focal areas in the striatum and cerebellum as well. In the 13 animals with focal regions of stress gene induction there were 8.1+/-1.8 foci in cortex, 5.5+/-0.9 foci in striatum, and 11.7+/-7.3 foci in cerebellum in the brain of each animal. The focal regions of stress gene induction varied in size from 200 micrometer to 7 mm in diameter. Systemic administration of the tirilazad-like anti-oxidants U101033E (n=8) and U74389G (n=7) completely blocked stress protein induction in focal brain regions normally produced by cisternal injections of lysed blood. There were fewer drug treated animals (0/15) with focal areas of stress gene induction compared to non-drug (13/21) treated animals following the cisternal lysed blood injections (p0.01 using Fisher's probability test). This study shows that anti-oxidants prevent focal regions of injury as assessed by heat shock protein expression in a rat model of SAH.

It has been hypothesized that the main olfactory bulb, with its relatively rich source of neurotrophins, may provide trophic support for neurons that project to the bulb. We monitored expression of the common, low affinity receptor for neurotrophins, p75 NGFR , in the olfactory bulb and basal forebrain of unilaterally bulbectomized and sham-treated rats, 1–16 weeks post-surgery, using the monoclonal antibody MAb192. An induction of p75 NGFR -immunoreactivity was observed in both the glomerular and olfactory nerve layers of the right, contralateral main olfactory bulb of lesioned animals. The naturally occurring regeneration taking place in the olfactory neuroepithelium is known to be altered by olfactory bulbectomy, with subsequent changes in the sensory input to the remaining bulb. These changes in expression of p75 NGFR in the olfactory bulb support the hypothesis we have developed in previous papers, that changes in the extent of the peripheral input from the olfactory neuroepithelium to the main olfactory bulb regulate p75 NGFR expression in both the glomerular and the olfactory nerve layers. Expression of p75 NGFR in the basal forebrain of bulbectomized animals was found to be no different than sham-treated controls and does not support the hypothesis that the olfactory bulb provides trophic support to this region of the central nervous system.

The heme oxygenase-1 gene, HO-1, induced by heme, ischemia, and heat shock, metabolizes heme to biliverdin, free iron, and carbon monoxide. Though the distribution of HO-1 has been described in normal rat brain, little is known about how extracellular heme proteins in the subarachnoid space distribute in brain. To address this issue, hemoglobin was injected into the cisterna magna of adult rats. Expression of HO-1 in these animals was compared with saline-injected, BSA-injected, and uninjected controls. Western blot analysis showed that 24 hours after injection oxyhemoglobin increased HO-1 levels approximately four- to fivefold in all brain regions studied compared with saline-injected and BSA-injected controls. In the brain, HO-1 immunoreactivity was evident at 4 hours and peaked at 24 hours after oxyhemoglobin injections, returning to control levels by 4 to 8 days. This HO-1 induction was detected mainly in cells with small, rounded somas bearing two to four truncated processes, a morphology consistent with that of microglia. These cells were double-stained with the microglial marker, OX42, in every brain region examined. It is proposed that subarachnoid hemoglobin may be taken up into microglia wherein heme induces HO-1.

✓ Induction of the hemeoxygenase-1 (ho-1) stress gene is of importance for rapid heme metabolism and protection against oxidative injury in vitro and in vivo. Although ho-1 expression is observed in glia following exposure to whole blood and oxyhemoglobin, expression is mild, and other stress genes are not induced simultaneously in this setting. Hemeoxygenase-1 can be induced by several other physiological stresses in addition to heme. In the brain, ho-1 induction has been observed in the penumbra following focal cerebral ischemia. Because lysed blood is a spasmogen, the authors studied focal hyperexpression of the ho-1 gene after injection of lysed blood, whole blood, or saline into the cisterna magna of adult rats. Immunocytochemical analysis of HO-1 was performed at 1, 2, 3, and 4 days after the injections. Because the 70-kD inducible heat shock protein (HSP70) is induced by cellular stress, alternate sections were immunostained for HSP70 to assess whether focal hyperexpression was a stress phenomenon. An oligonucleotide probe was also used for in situ hybridization to demonstrate that ho-1 messenger (m)RNA was present. Focal HO-1 immunostained areas were observed after lysed blood injection only and were located mainly in the basal cortex and cerebellar hemisphere, although focal hyperexpression was also found in many other regions. The intensity of staining and the number of regions were maximum at 1 day. Double-labeled immunofluorescence revealed that many HO-1—immunoreactive cells were microglia. The HSP70 immunostaining of adjacent sections from the same animals demonstrated focal regions of immunoreactivity whose topography corresponded exactly with the topography of the HO-1—immunostained areas. Conventional histology in regions of HO-1 hyperexpression was often normal. In situ hybridization using the same oligonucleotide demonstrated that ho-1 mRNA was induced in focal areas of forebrain and in large regions of cerebellum within 6 hours of injection. These results demonstrate that focal hyperexpression of the ho-1 stress gene occurs after lysed blood injection and appears to be an indicator of cellular stress and injury in regions in which infarction does not occur. These results also suggest that cellular injury that occurs after injection of lysed blood may go undetected using conventional histology. Although direct heme metabolism was not investigated, our results indicate that rapid metabolism of heme, both intracellular and extracellular, may prove to be beneficial after subarachnoid hemorrhage.

The heme released following subarachnoid hemorrhage is metabolized by heme-oxygenase (HO) to biliverdin and carbon monoxide (CO) with the release of iron. The HO reaction is important since heme may contribute to vasospasm and increase oxidative stress in cells. HO is comprised of at least two isozymes, HO-2 and HO-1. HO-1, also known as heat shock protein HSP32, is inducible by many factors including heme and heat shock. HO-2 does not respond to these stresses. To begin to examine HO activity following subarachnoid hemorrhage (SAH), the expression of HO-1 and HO-2 was investigated after experimental SAH in adult rats. Immunocytochemistry for HO-1, HO-2 and HSP70 proteins was performed at 1, 2, 3 and 4 days after injections of lysed blood, whole blood, oxyhemoglobin and saline into the cisterna magna. A large increase in HO-1 immunoreactivity was seen in cells throughout brain following injections of lysed blood, whole blood, and oxyhemoglobin but not saline. Lysed blood, whole blood and oxyhemoglobin induced HO-1 in all of the cortex, hippocampus, striatum, thalamus, forebrain white matter and in cerebellar cortex. HO-1 immunoreactivity was greatest in those regions adjacent to the basal subarachnoid cisterns where blood and oxyhemoglobin concentrations were likely highest. Double immunofluorescence studies showed the HO-1 positive cells to be predominately microglia, though HO-1 was induced in some astrocytes. HO-1 expression resolved by 48 h. HO-2 immunoreactivity was abundant but did not change following injections of blood. A generalized induction of HSP70 heat shock protein was not observed following injections of lysed blood, whole blood, oxyhemoglobin, or saline. These results suggest that HO-1 is induced in microglia throughout rat brain as a general, parenchymal response to the presence of oxyhemoglobin in the subarachnoid space and not as a stress response. This microglial HO-1 response could be protective against the lipid peroxidation and vasospasm induced by hemoglobin, by increasing heme clearance and iron sequestration, and enhancing the production of the antioxidant bilirubin.

We have disrupted the integrity of the rat olfactory neuroepithelium using intranasally applied TX-100, a procedure known to reversibly eliminate the sensory neuron input from the neuroepithelium to the olfactory bulb [Margolis et al. (1974) Denervation in the primary olfactory pathway of mice: biochemical and morphological effects. Brain Res.81, 469–483]. One week after TX-100 exposure, we observed a disruption of the pseudo-stratified organization of the neuroepithelium which was accompanied by a 60% reduction in neuroepithelial width, compared to saline-treated controls. Full recovery of the neuroepithelium was not observed until 16 weeks post-lesion. During this post-lesion period, we monitored the expression of the low affinity receptor for neurotrophins, p75NGFR, in the olfactory bulb of saline- and TX-100-treated animals, using the monoclonal antibody, MAb192. In saline-treated animals, p75NGFR-immunoreactivity (p75NGFR-ir) was localized to individual glomeruli in the olfactory bulb, with little or undetectable p75NGFR-ir in the olfactory nerve layer. We have previously reported that pre-lesioned levels of p75NGFR-ir in the glomerular layer were dramatically reduced while an induction of p75NGFR-ir was observed in the olfactory nerve layer, one and two weeks after intranasal exposure to TX-100 [Turner & Perez-Polo (1992) Regulation of the low affinity receptor for nerve growth factor, p75NGFR, in the olfactory system of neonatal and adult rat. Int. J. Devl Neurosci.10, 343–359]. In this paper, we demonstrate that this previously reported reduction in glomerular p75NGFR-ir took 16 weeks to fully recover and was, thus, coincident with the post-lesion recovery of the neuroepithelium. In the olfactory nerve layer, the return of p75NGFR-ir to pre-lesioned levels took only four weeks. No changes in neuroepithelial width and integrity or alterations in p75NGFR-ir in the olfactory bulb were observed in saline-treated animals. Thus, the TX-100-induced removal of the peripheral input to the olfactory bulb resulted in a reversible change in expression of p75NGFR-ir in the bulb. We believe that these changes are a reflection of the regenerative capacity of the olfactory system.

Using MAb192, a monoclonal antibody to the rat low affinity receptor for nerve growth factor (p75NGFR), we determined the expression of p75NGFR in rat neonatal and adult olfactory system. In neonates and adults, we observed discrete p75NGFR-immunoreactivity (p75NGFR-ir) in the glomerular layer of the main olfactory bulb. The intensity and organization of glomerular p75NGFR-ir increased with age. This was in keeping with the general ontogeny of the main olfactory bulb. Generally, granule cells, mitral cells and periglomerular cells of the main olfactory bulb were not specifically stained. However, in early neonates, granule cells close to the lateral olfactory tract exhibited p75NGFR-ir. Additional specific staining was found in the olfactory receptor neurons of neonatal and adult olfactory neuroepithelium, the olfactory fascicles and in the glomeruli of the accessory olfactory bulb. The intensity, but not the organization, of specific staining in the accessory olfactory bulb increased as the animal matured. We believe that p75NGFR-ir in the olfactory system is associated with its unique capacity to regenerate its peripheral input to the main olfactory bulb. The presence of p75NGFR-ir in the accessory olfactory bulb would suggest a broader role for this protein. Here we discuss the implications of these findings with regards to nerve growth factor, other trophic molecules, and their receptors. The data presented provide a foundation for studies involving manipulation of regenerative phenomena while monitoring the expression of neurotrophic factors and their receptors.