Persistent cognitive deficits are one of the most important sequelae of head injury in humans. In an effort to model some of the structural and neuropharmacological changes that occur in chronic postinjury brains, we examined the longitudinal effects of moderate vertical controlled cortical impact (CCI) on place learning and memory using the Morris water maze (MWM) test, morphology, and vesicular acetylcholine (ACh) transporter (VAChT) and muscarinic receptor subtype 2 (M2) immunohistochemistry. Vertical CCI (left parietal cortex, 4 m/sec, 2.5 mm; n = 10) or craniotomy (sham) was produced in male Sprague-Dawley rats (n = 10). Place learning was tested at 2 weeks, 4 weeks, 3 months, 6 months, and 12 months postinjury with the escape platform in a different maze quadrant for each time point. At each interval, rats received 5 days of water maze acquisition (latency to find hidden platform), a probe trial to measure place memory, and 2 days of visible platform trials to control for nonspecific deficits. At 3 weeks, half the animals were sacrificed for histology. At these injury parameters, CCI produced no significant differences in place learning between injured and sham rats at 2 weeks, 4 weeks, or 6 months after injury. However, at 3 and 12 months, the injured rats took significantly longer to find the hidden platform than the sham rats. Probe trial performance differed only at 12 months postinjury between injured (25.73+/-2.1%, standard error of the mean) and sham rats (44.09+/-7.0%, p < 0.05). The maze deficits at 1 year were not due to a worsening of performance, but may have resulted from a reduced ability of injured rats to benefit from previous water maze experience. Hemispheric loss of 30.4+/-5.5 mm3 was seen at 3 weeks after injury (versus respective sham). However, hemispheric loss almost doubled by 1 year after injury (51.5+/-8.5 mm3, p < 0.05 versus all other groups). Progressive tissue loss was also reflected by a three- to fourfold increase in ipsilateral ventricular volume between 3 weeks and 1 year after injury. At 1 year after injury, immunostaining for VAChT was dramatically increased in all sectors of the hippocampus and cortex after injury. Muscarinic receptor subtype 2 (M2) immunoreactivity was dramatically decreased in the ipsilateral hippocampus. This suggests a compensatory response of cholinergic neurons to increase the efficiency of ACh neurotransmission. Moderate CCI in rats produces subtle MWM performance deficits accompanied by persistent alteration in M2 and VAChT immunohistochemistry and progressive tissue atrophy. The inability of injured rats to benefit from repeated exposures to the MWM may represent a deficit in procedural memory that is independent of changes in hippocampal cholinergic systems.

Previous work in our laboratory and others using the weight drop (WD) model of traumatic brain injury (TBI) has shown that neutrophils accumulate in brain tissue during the initial 24 h posttrauma as measured by myeloperoxidase (MPO) activity and immunohistochemistry. This study compares the acute inflammatory response to TBI over time, as measured by MPO activity, in the WD and controlled cortical impact (CCI) models. Anesthetized adult Sprague-Dawley rats were traumatized using WD (10-g weight dropped 5 cm) or CCI (4 m/sec, 2.5 mm depth). At 2, 24, 48, or 168 h after trauma, rats (n = 4-5/group at each time) were anesthetized and killed, the brains were removed, and 6-mm coronal slices from traumatized and contralateral hemispheres were assayed for MPO activity. Nontraumatized rats (n = 4) served as controls. Three additional rats underwent a more severe CCI (3 mm depth) with MPO activity assayed at 24 h. A separate group of rats (n = 6) was subjected to WD trauma and killed at 2 weeks after injury for analysis of lesion volume. MPO activity in the traumatized hemisphere was demonstrated at 24 and 48 h in both the WD (0.3152 +/- 0.0472 and 0.3017 +/- 0.0228 U/g, respectively, p < 0.05 vs controls) and CCI (0.1866 +/- 0.0225 and 0.1937 +/- 0.0772 U/g, respectively, p < 0.05 vs controls) models. MPO activity was below the sensitivity of the assay in the control, 2 h, and 168 h groups in both models.(ABSTRACT TRUNCATED AT 250 WORDS)

Traumatic brain injury (TBI) is often accompanied by an acute inflammatory reaction mediated initially by neutrophils. Adhesion molecules expressed on vascular endothelium are requisite elements during recruitment of leukocytes at sites of inflammation. In a rat model of TBI the induction and persistent expression of E‐selectin (CD62E) on cerebrovascular endothelium ipsilateral, but not contralateral, to the site of contusion was demonstrated (P < 0.05 at 4 and 48 h posttrauma). In addition, these studies confirmed up‐regulation and prolonged expression of ICAM‐1 (CD54) on endothelium in the traumatized hemisphere (P < 0.05 at 4, 24, 48, and 72 h posttrauma). It is of interest that increased expression of CD54 was noted on blood vessels in the contralateral, non‐traumatized hemisphere 48 h posttrauma. Expression of a third endothelial adhesion molecule, PECAM‐1 (CD31), was unchanged following trauma. Administration of a murine monoclonal antibody (TM‐8) that inhibits the adhesive function of CD54 blocked a significant portion (37.9%) of neutrophil recruitment 24 h posttrauma (P = 0.04). Employing immunocytochemistry and a monoclonal antibody specific for rat neutrophils (RP‐3), peak infiltration of neutrophils was shown to occur 48 h after trauma. In contrast to emigration of neutrophils from blood vessels within the contusion, however, entry of neutrophils occurred from the surrounding leptomeninges and choroidal vessels. These studies demonstrate the relevance of CD54 (ICAM‐1) in recruitment of neutrophils following TBI. However, the majority of neutrophil influx relies on endothelial adhesion molecules other than CD54. Because emigration of neutrophils was shown to occur predominantly from vessels within the leptomeninges and choroid plexus, intrathecal delivery of agents that inhibit the adhesive interactions between neutrophils, endothelial CD54, and other endothelial adhesion molecules to be defined may offer a novel form of therapy to prevent the acute inflammatory response that follows TBI. J. Leukoc. Biol. 61: 279–285; 1997.

Controlled cortical impact (CCI) is a contemporary model of experimental cerebral contusion. We examined the cerebrovascular and neuropathologic effects of a severe CCI in rats. The utility of magnetic resonance imaging (MRI) for the assessment of contusion volume after severe CCI was also established. Severe CCI (3.0 mm depth, 4 m/sec velocity) to the left (L) parietal cortex was produced in anesthetized (isoflurane/N2O/O2), intubated, and mechanically ventilated male Sprague-Dawley rats (n = 58). Physiologic parameters were controlled. The time course of alterations in edema [L-R% brain water (% BW) in 3-mm coronal sections through injured and contralateral hemispheres, wet-dry weight] was evaluated at 2 h, 24 h, 48 h, and 7 days posttrauma. Local cerebral blood flow (ICBF, measured in 8 structures in each hemisphere by autoradiography) was evaluated at 2 h, 24 h, and 7 days. Contusion volume (measured by histology and image analysis) was assessed at 14 days and measured in 6 rats by both MRI and histology. The survival rate after severe CCI was 96.2%. The L-R difference in % BW increased to 1.69 +/- 0.18% at 2 h, 3.00 +/- 0.08% at 24 h, 2.69 +/- 0.09% at 48 h, and 0.94 +/- 0.21% at 7 days. These values all differed from the control (p < 0.05). The % BW was greater at 24 h and 48 h than at 2 h and 7 days (p < 0.05). Marked reductions in ICBF were limited to structures in the injured hemisphere and were observed in the parietal cortex (2 and 24 h), subcortical white matter (2 and 24 h), and hippocampus (2 h), (p < 0.05) vs control rats. In the contusion core, ICBF was 19.4 +/- 8.8 mL 100 g-1 min-1 at 24 h (p = 0.011 vs normal). Necrosis was seen in large portions of the parietal cortex and subcortical white matter, and portions of the hippocampus and thalamus. Contusion volume was 47.8 +/- 9.2 mm3, which represented 14.4 +/- 2.1% of the traumatized hemisphere. Estimates of contusion volume by MRI and histology were closely correlated (r = 0.941, p < 0.017). Severe CCI in rats is accompanied by contusion, reproducible edema, and marked hypoperfusion, involving over 14% of the injured hemisphere, and can be produced with minimal mortality. T2-weighted MRI successfully and noninvasively identifies contusion volume in this model.

The influence of anesthetic agents on cerebral blood flow (CBF) was tested in normal rats. CBF is quantified with arterial spinlabeled MRI in rats anesthetized with either an opiate (fentanyl), a potent inhalation anesthetic agent (isoflurane), or a barbiturate (pentobarbital) using doses commonly employed in experimental paradigms. CBF values were found to be about 2. Despite the frequent use of anesthetics in experimental MRI in rats using techniques such as arterial spin labeling (ASL) and blood oxygenation level-dependent (BOLD) signal changes, quantitative assessment of the effect of commonly used anesthetics on cerebral blood flow (CBF) in these protocols is limited (1). Assessment of the effect on CBF of fentanyl, a potent narcotic analgesic agent frequently administered in the clinical neuroanesthesia and neurointensive care setting, is limited even using conventional techniques applied to normal rats (2,3). To this end, we quantified CBF using the continuous ASL method (CASL) in rats administered fentanyl (10 g kg Ϫ1 bolus, then 50 g kg Ϫ1 h Ϫ1 intravenously (IV), during ventilation with N 2 O/O 2 , 2:1) and compared these findings to those observed using isoflurane (1% by inhalation in N 2 O/O 2 , 1:1) or pentobarbital (50 mg kg Ϫ1 h Ϫ1 IV, during ventilation with N 2 /O 2 , 1:1) anesthetic regimens. METHODS Animal ModelThe study protocol was approved by the University of Pittsburgh Animal Care and Use Committee. Twelve mature male Sprague-Dawley rats (332-390 g) were anesthetized with isoflurane and N 2 O/O 2 , 1:1, endotracheally intubated and mechanically ventilated. A femoral arterial catheter and a double lumen venous catheter were surgically inserted for blood sampling, for pressure monitoring, and for administration of agents. Rats were assigned to one of three anesthetic groups for CBF measurements: fentanyl, isoflurane, or pentobarbital (n ϭ 4 per group). Isoflurane was then continued at 1% by inhalation (isoflurane group), or discontinued and either fentanyl (10 g kg Ϫ1 bolus, then 50 g kg Ϫ1 h Ϫ1) or pentobarbital (50 mg kg Ϫ1 h Ϫ1 ) was administered IV. N 2 O/O 2 was continued in fentanyl (2:1) and isoflurane (1:1) groups. Rats were ventilated with N 2 /O 2 , 1:1 in the pentobarbital group. Fentanyl or isoflurane anesthesia was administered with N 2 O because this approach is frequently used in clinical neuroanesthesia (4) and in experimental models (5,6). Similarly, to mirror clinical (7) and laboratory (8) usage, pentobarbital was administered without N 2 O. In the pentobarbital group, N 2 O was replaced with N 2 to match the fraction of inspired oxygen concentration in the isoflurane-anesthetized group. After a 30-min equilibration period, perfusion images were acquired in duplicate. In all studies pancuronium bromide (0.1 mg kg Ϫ1 h Ϫ1) was administered IV for paralysis. Rectal temperature was continually monitored and maintained at 37°C with a heated water blanket. Arterial blood gases and mean arterial blood pressure (MABP) were also monitored throughout the NMR studies. Respiratory t...