Original Paper
Pharmacology 1992;45:142-153
R. Rischkea A. Ram ia U. Bachmanna A. Rahiéh J. Krieglsteina Institut für Pharmakologie und Toxikologie, Fachbereich Pharmazie und Lebensmittelchemie der Philipps-Universität, Ketzerbach, Marburg, BRD; Laboratoire de Neurobiologie Endocrinologique, URA 1197 du CNRS, Université Montpellier II, Montpellier, France
Activated Astrocytes, but not Pyramidal Cells, Increase Glucose Utilization in Rat Hippocampal CA1 Subfield after Ischemia
Abstract
Received: April 17.1991 Accepted: December 2,1991
Dr. A. Rami Institut für Pharmakologie und Toxikologie Fachbereich Pharmazie und Lebensmittelchemie. Philipps-Universität. Ketzerbach 63. D-W-3550 Marburg (FRG)
©1992 S. Karger AG. Basel 0031-7012/92/ 0453—0142$2.75/0
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
KeyWords Ischemia Rat Hippocampus Local cerebral glucose utilization Ketamine Synaptophysin Microtubule associated protein 2 Glial fibrillary acidic protein
The local cerebral glucose utilization (CMRgic) in the damaged rat hippo campal CAl subfield increases 7 days after lOmin of cerebral ischemia. We have used the N-methyl-D-aspartate antagonist (NMDA antagonist) ketamine in rats 7 days after sham operation or cerebral ischemia to deter mine whether the elevated postischemic CMRgic of the CA1 subfield is due to long-lasting hyperexcitation of surviving or injured neurons, or, alterna tively, to the metabolism of other cell types. The autoradiographic data were interpreted with the aid of histochcmical analysis of the postischemic hippocampal cell changes. Anesthetic doses of ketamine significantly reduced the CMRg|Cin the CAl strata oriens, pyramidale and radiatum of sham-operated rats, while the postischemic increases in CMRgic in these hippocampal CA 1 strata were not affected by ketamine. In addition, there were ketamine-induced increases in the CMRg|Cof the CAl stratum lacunosum moleculare of both sham-operated and postischemic rats. The immunoreactivity of the microtubule-associated protein 2 (MAP2), a postsynaptic protein marker, was decreased markedly in the CAl subficld in postischemic rats, while the prcsynaptic protein marker, synaptophysin, remained the same in sham-operated and postischemic rats. The glial fibrillary acidic protein (GFAP) immunoreactivity of astrocytes raised markedly in the ischemically damaged CAl subfield. Although it could be demonstrated that presynaptic terminals remain intact in the postisch emic damaged CAl subfield, the lacking ketamine effect on CAl pyrami dal neurons indicated that the increase in CMRgic in this brain area is not due to postsynaptic neural hyperexcitation, but probably has to be attrib uted to astrocytes activated by neuronal damage.
There is general agreement that the pyram idal neurons in the hippocampal CA1 subfield are particularly vulnerable to transient isch emia [1,2]. About 70-90% of the neurons in the hippocampal CA1 subfield of rats were damaged 7 days after a 10-min forebrain isch emia [2, 3], However, despite this postischemic damage, recent autoradiographic mea surements using [l4C]-labeled 2-deoxy-Z)-glucose have revealed an increase in local cere bral glucose utilization (CMRg|C) in different strata of the CA1 subfield [4, 5]. CMRgic is considered to be closely coupled to the func tional activity of neuronal elements in the normal brain, but the cause of increased hip pocampal CMRgic in the ischemically dam aged CA1 subfield remains controversial. While there is evidence that activation of astrocytes [6-8] produces an increase in CMRg|Cof the CA1 subfield after ischemia [5, 9], it is always possible that long-lasting hy perexcitation of the damaged neurons may contribute to this phenomenon. We have, therefore, studied the pathophys iology underlying the increased CMRgic of the CA1 subfield 7 days after ischemia by mea suring the CMRgic in rats whose hippocampal neuronal activity was modified by anesthetic doses of ketamine 7 days after sham operation or ischemia. The resulting autoradiographic data were interpreted in conjunction with his tological examination of the hippocampal neurons, astrocytes and macrophages, plus hippocampal pre- and postsynaptic struc tures.
Materials and Methods Animals
Male Wistar rats (Ivanovas, Kisslcgg, FRG) weigh ing 250-300 g were maintained under controlled light ing and environmental conditions (12 h dark/light cy
cle. 23 ± 1 °C. 55 ± 5% relative humidity) and fed a standard diet (Altromin. Lage. FRG) and tap water ad libitum. The rats were fasted overnight preceding the ischemic procedure. Materials
[l4C]-2-deoxy-£)-glucose (=DOG; specific activity 51.1-52.5 mCi/mmol) was purchased from NEN (Dreieich. FRG). Mouse monoclonal antibodies di rected against glial fibrillary acidic protein (GFAP). rabbit antimouse y-globulin, mouse peroxidase-antihorscradish peroxidase complex (PAP) and 3,3'-diaminobenzidine (DAB) were obtained from Serva (Hei delberg, FRG). The anti-microtubule associated pro tein 2 (MAP2) serum was a gift from Dr. A.M. Hill (Gif-sur-Yvette, France). The second antibody, goat anlirabbit y-globulin and the rabbit PAP complex were also obtained from Serva. Monoclonal antibodies di rected against synaptophvsin were purchased from Boeringer Mannheim (FRG). Biotinylated sheep antimouse y-globulin and streptavidin-peroxidase com plex were obtained from Amersham (Braunschweig, FRG). All other chemicals were of reagent grade. The brains were fixed by immersion in a mixture of 60% ethanol, 30% chloroform and 10% acetic acid (Carnoy’s solution). Induction o f Ischemia
Ischemia was induced according to Smith et al. [3], The rats were anesthetized with 3.5% halothane and connected to a Starling-type respirator delivering 0.8 % halothane in a 2:1 mixture of nitrous oxide and oxy gen. The common carotid arteries were isolated via a cervical incision and ligatures were placed around them. A polyethylene cannula was inserted into the tail artery for blood sampling and blood pressure record ing, and a silicone catheter was advanced into the infe rior caval vein via the right jugular vein. Heparin was injected i.v. (200 IU/kg) to prevent coagulation, and muscle paralysis was maintained with 5 mg/kg i.v. sux amethonium chloride every 15-20 min. Halothane was discontinued 30 min before starting ischemia to reach a preischemic steady state. Trimethaphan cam phor sulfonate (5 mg/kg) was infused and forebrain ischemia was induced by carotid clamping and exsanguination to a blood pressure of 40 mm Hg. Blood pressure was restored after 10 min of ischemia by removing the carotid clamps and reinfusing the re moved blood. A solution of NaHCOj (0.6 mmol/1) was then injected (0.5 ml), and the wound were sutured. The rats were ventilated for about 45 min until they regained consciousness, when all catheters were re moved. Body temperature was kept at 37 °C with a
143
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Introduction
Local Cerebral Glucose Utilization Seven days after sham-operation or cerebral isch emia. CMRrf.- was measured in ketamine treated and untreated rats according to the 2-|llC]-deoxy glucose method described by Sokoloff ct al. 110). I he femoral vessels were cannulated with polyethylene catheters under light anesthesia with a 2:1 mixture of nitrous oxidc/oxygcn containing 0.8% halothane and the wound was packed with lidocaine gel and closed. The animals were immobilized with plaster cats, and fixed on a preparation desk, where they were allowed to recover from anesthesia for 2 h before measuring CMR*!«.. Body temperature was held at 3 7 °C with a healing lamp. The physiological variables were deter mined before infusion of the radiotracer. CMR,k. was detei mined by infusing 120 pCi/kg DOG in physiolog ical saline via the femoral vein over 25 s. A total o f 16 arterial blood samples (-80 pi) wctc collected during 45 nun from the beginning o f the experiment for mea suring glucose and DOG. Immediately after the lust sample had been taken (46 nun) the rats were decapi tated. the brains dissected out and immediately frozen in isopentane ( - 5 0 * 0 . I he brains were sectioned (20 pm) and the sections exposed to Osray M3 film (Agfa-Gcvacrt. Leverkusen. FRG) for 10 days. I he optical densities o f the autoradiograms were measured with a computer-based image analyzer (IBAS. Kontron. Hchingcn. FRG). Ketamine Administration A total of 60 mg/Kg of ketamine was administered. I he infusion schedule was designed to maintain anes thetic blood levels throughout the 45 min of the auto radiographic experiments 111). A bolus i.v. injection of 30 mg/kg ketamine was first given \ ia the femoral vein over a I-min period. The physiological parameters were determined immediately theralter. and IXXi was infused over the next 25 s. A second booster dose of 30 mg/kg ketamine was infused over the following I 5 min using an infusion pump (Original Perfusor. Bruun-Mctsungcn. FRG). Controls were given physio logical saline by the same protocol.
144
Histology Seven days after sham operation or ischemia ) rats were killed by decapitation. The brain were carefully removed, fixed in Camoy’s solution, and embedded in paraffin. Sections (5 pm) w-cre cut 6 mm rostral of the teritoiiul incision, and mounted on chromc-alum-gelatinc-coatcd slides. Adjacent sections from each brain were stained by each of the procedures described below.
Histochemtcal Procedures Neuronal damage was assessed by staining the brain sections w ith acidic fuchsin/cclcsline blue. Dam aged neurons are stained violet |I2). Intact and dam aged neurons were counted in entire hippocampal sub fields and neuronal damage was calculated as the per centage of damaged pyramidal cells. An adjacent sec tion was stained with hematoxylin and eosin (HE) to visualize macrophages.
Imniunohistocheniical Procedures Astrocytes were immunosiained by the PAP method [131. Paraffin was removed from the sections with toluene and they were rchydrated. washed in phosphate-buffered saline (PBS: I0pmol/1. pH 7.4) and flooded for 16h at with diluted mouse pri mary antibody (1:5 in PBS containing 1% normal swine serum. PBS-NSS). The sections were then carelully washed w ith PBS and flooded w ith rabbit second ary antibody (1:50 in PBS-NSS) for 30 min. washed again with PBS. and covered with the mouse PAP (1:50 in PBS-NSS) for 20 min. Finally, the sections were washed in PBS. incubated for 10-15 min with 0.05% DAB containing 0.01 % H ;0; in PBS. washed several times with PBS and mounted with Permount (Fisher Scientific Company. USA). MAP2 was immunolocalizcd using the same procedure with a rabbit antiserum [I I]. The second antibody and the rabbit PAP were diluted 1:50 and 1:100 in PBS containing 5% bovine serum albumin (BSA). Immunohistochcmical staining o f synaptophysin was performed by using the biotin-streptavidin system (Arnersham). Tris buffer (0.05 mol/l Tris. 0.12 mol/l N a d . pi I 7.6) was used throughout. The dcparafTincd. rehydrated sections were flooded for 16 li al 4 “C with diluted monoclonal antibody (1:5 in Tris buffer con taining 5% BSA. I ris-BSA). The sections were then carefully washed with buffer and flooded with biotiny lated sheep antimouse -/-globulin (diluted 1:130 in Tris-BSA) for 30 min. The sections were washed with I ris-BSA and covered with the strcptavidin-piroxidasc complex (diluted 1:200 in Tris-BSA) for 20 min. The slides were washed in Tris. incubated for 10-
Rischke/Rami/Bachmann/Rabii/
Kncglstcm
Ischemia Resistance o f Hippocampal Prcsynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
healing lamp. Blood gases, blood pH (Corning 178. Corning Medical. Giessen. FRG). blood pressure(Statham P23DB. Hcio Rev. Puerto Rico: IISF-Flcktromanometer. Hugstetten, FRG), and plasma glucose (Beckman Glucose Analyzer. Munich. FRG) were de termined before induction o f ischemia. Sham-operated control rats underwent all the ex perimental procedures except occlusion o f the carotid arteries and exsanguination.
Pharmacology, vol. 45 S. Karger. Basel
Rixchke/ Rami/Bachmantt/Rabie/ Kncglstcm
Plaie I
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 1. Representative autoradiograms o f (,4C]-deoxy-£>-glucosc metabo lism in rat hippocampus of ketamine-treated and untreated rats 7 days after sham operation or ischemia, a Sham operation, untreated, b Sham operation, ketamine treated, c Ischemia, untreated, d Ischemia, ketamine-treated. Keta mine was administered by a bolus injection (30 mg/kg i.v.) immediately before tracer infusion followed by a booster dose (30 mg/kg i.v.) during the next 15 min after the tracer infusion to maintain ketamine anesthesia.
Table 1. Physiological variables
Before sham operation (n = 12) paCK mm Hg PaCCT, mm Hg Arterial pH Plasma glucose, mg/l Blood pressure, mm Hg Temperature, °C
7 days after ischemia
sham operation
ischemia
( n - 14)
untreated (n = 6)
untreated ( n - 7)
143 ±17 137 ± 10 37±2 38 ±2 7.40 ±0.02 7.39 ±0.01 1,190 ± 120 1.270± 150 132 ±13 129 ± 11 37.0±0.2 36.9 ±0.1
ketamine (n = 6)
98 ±10 99 ± 12 39 ± 1 42 ±2* 7.40 ±0.12 7.38 ±0.09 1,630 ±90 l.580± 140 145 ±15 131 ± 12 37.0 ±0.1 37.0±0.2
ketamine (n = 7)
95 ±12 96 ± 8 37 ± 2 41 ± 2* 7.41 ±0.16 7.39 ± 0.07 1.730 ±150 1.650 ± 110 140 ±12 132 ± 10 37.0 ±0.2 37.0± 0.2
!5min with 0.05% DAB containing 0.01% HjO? in Tris, washed several times with Tris and finally mounted in Permount. Control sections were incubated with Tris instead of primary antibody. They showed no positive reac tion to GFAP, MAP2 or synaptophysin antibodies. The immunohistochemical stainings were assessed qualitatively by light microscopy.
Results
The physiological variables determined be fore sham operation or ischemia were in the normal range for both the control and test groups of rats (table 1). The physiological pa rameters measured 7 days after sham opera tion or cerebral ischemia were normal for the 2 untreated groups; intravenous 30 mg/kg ke tamine caused a significant increase in the PaCCb of all treated rats (table 1). The behav ior of the rats during ketamine administration was the same for the sham-operated and isch emic rats. There were occasional head and limb movements at the start of the bolus injection, but the rats remained motionless
during infusion of the booster dose and for a 30 min period. CMRgic was determined in 6 defined hip pocampal areas of untreated rats and rats treated with ketamine 7 days after sham oper ation and ischemia (table 2). The strata oriens and pyramidale of the CA1 subfield were ana lyzed together, because it was not possible to differentiate between them on the autoradio grams (see fig. 6). The treatment of rats with anesthetic doses of ketamine 7 days after sham-operation induced uneven changes in CMRg|C. The CMRgic of the CA1 strata oriens. pyramidale and radiatum was reduced during ketamine anesthesia, while the CMRg|t. of the CA1 stratum lacunosum moleculare in creased markedly (fig. la.b. 6). The CMRg|C of the CA3 and CA4 subfields and the dentate gyrus did not change significantly under these conditions (fig. la.b). As demonstrated previously, CMRgic in the CA1 strata oriens and pyramidale increased 7 days after forebrain ischemia (fig. la.c). In contrast to the measurements performed 7 days after sham operation, administration of
145
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
The physiological variables were determined 30 min after beginning artificial ventilation, before sham opera tion or ischemia in both ketamine-treated and untreated rats, and 7 days after sham operation or ischemia. Values are the means ± SD of n experiments. * p < 0.05 (Mann-Whitney U test).
Table 2. The effect of ketamine on CMRgic of the rat hippocampal subfields 7 days after sham operation or
ischemia Hippocampal area
CAl strata Oriens and pyramidale Radiatum Lacunosum moleculare CA3 CA4 Dentate gyrus
Sham operation
Ischemia
untreated (n -6 )
ketamine (n - 6)
untreated (n = 7)
ketamine (n= 7)
480±50 570± 100 600 ±90 570 ±80 470 ±90 490 ±100
4I0±60* 430 ±50* 1,070 ±130** 490 ±40 470 ±40 600 ±70
610 ± 120* 540± 110 630 ±120 560 ±120 500± 110 470± 110
560± 110 550± 110 740 ± 110+ 550±80 480 ±80 490 ±90
CMRgic is given as pmol kg"1 min-1. Ketamine was given intravenously 7 days after sham operation or cerebral ischemia. Anesthetic plasma levels of ketamine throughout the period of tracer uptake were obtained by giving a 30-mg/kg bolus injection, followed by immediate tracer infusion, and a second infusion of ketamine (30 mg/kg) over the next 15 min. The data were analyzed using the Mann-Whitncy U test. Values are means ± SD of n experiments. * p < 0.05; ** p < 0.01: compared with the vehicle group after sham operation. + p < 0.05: compared with the corresponding untreated ischemic control.
146
fields of the hippocampus in sham-operated controls (fig. 4a.b), but much less MAP2 im munostaining in the CA1 subfield of postischemic rats (fig. 4c.d). However, synaptophysin-immunostaining was the same in hippo campal sections of sham-operated controls and postischemic rats (fig. 5a-d).
Discussion
The histological appearance of the rat hip pocampus confirms earlier findings that a transient forebrain ischemia mainly injures the pyramidal neurons in the CA1 subfield [1. 2], The loss of immunoreactivity for the postsynaptic somal and dendritic protein marker MAP2 [15] from theCAl subfield 7 days after ischemia also clearly shows the selective vul nerability of pyramidal cells in this brain area (fig. 4a-d. 6). Interestingly, visualization of presynaptic terminals by synaptophysin im-
Rischke/Rami/Bachmann/Rabic.' Krieglstein
Ischemia Resistance of Hippocampal Presynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
ketamine 7 days after recirculation did not reduce the CMRgic of the CA1 strata oriens, pyramidale and radiatum (fig. lc.d). The ke tamine-induced increase in the CMRgic in the CA1 stratum lacunosum moleculare of shamoperated rats also occurred in rats given the anesthetic 7 days after ischemia (fig. lc.d). No damaged neurons could be seen in the hippocampal subfields of sham-operated con trols (fig. 2a), but the 10-min forebrain isch emia damaged 83% of pyramidal cells in the CA1 subfield (fig. 2b), 2% of neurons in the CA3 subfield, and 8% of CA4 and hilar neu rons. No macrophages were seen in HEstained control sections or in sections of brains 7 days after ischemia. GFAP immunocytochemical staining was scattered through out the hippocampus 2 days postischemia (fig. 3a). In contrast, there was intense antiGFAP immunoreactivity in astrocytes in the CA1 subfield 7 days after ischemia (fig. 3b). There was MAP2 immunostaining in all sub
munostaining [16] revealed no differences be tween the hippocampal CA1 areas of shamoperated and postischemic rats (fig. 5a-d). This agrees with the findings of Kirino et al. [7], who demonstrated by electron micros copy that presynaptic terminals survive and maintain their structural integrity in the post ischemic CA1 subfield of the gerbil. As described recently [4. 5], the postisch emic neuronal damage in the CA1 subfield is associated with an increase in CMRg|C (fig. 1a,c). The effects of ketamine anesthesia on CMRgic in sham-operated and in postisch emic rats should help to show whether the
increased CMRgic in the postischemic CA1 subfield is due to excitation of remaining neu rons or to activation of other cells. Ketamine acts as a noncompetitive antagonist on the excitatory NMDA receptor [17]. high concen trations of which are present in the hippocam pal CA1 subfield [18]. Inhibiting the NMDA receptor with ketamine causes a specific blockade of postsynaptic neurons in the hip pocampal CA1 subfield [19]. Thus, if the increased CMRg|c of the CA1 subfield 7 days after ischemia was a result of enhanced neural excitation, there should be a ketamine-in duced reduction in the CMRg|C.
147
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 2. Light photomicrographs show the hippocampal CA1 subfield 7 days after sham operation (a) and ischemia (b). Arrows indi cate some damaged neurons. X 300.
The half-life of ketamine is 10 min in the plasma and 8 min in the brain of the rat [20]. resulting in an 8-min period of hypnosis when a single intravenous injection of 30 mg/kg ke tamine is given [20]. However, a single injec tion of ketamine makes the measurement of CMRg|C, which requires steady state condi tions, difficult. In oder to measure the CMRgic during an unchanged anesthetic state, we de cided to give a loading dose of 30 mg/kg fol lowed by infusion of a further 30 mg/kg dose. The rat’s behavior, as mentioned in the
148
present study, indicated dissociative anes thetic conditions for a 30-min period, suggest ing that the drug was present in the brain throughout the tracer uptake. Under these conditions, ketamine caused a significant reduction in CMRg|C of the CA1 strata oriens, pyramidale and radiatum of sham-operated control rats (fig. la,b: table 2). This agrees with the finding that ketamine depresses synaptic excitation of CA1 neurons [19] by blocking the NMDA receptor [17], The fact that the NMDA receptor is not
Rischkc/Rami/Bachmann/Rabi6' Krieglstein
Ischemia Resistance of Hippocampal Presynaplic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 3. GFAP-positive astro glial cells in the rat hippocampus 2 days after ischemia (a) or 7 days after ischemia (b). X 300.
a
w V
c
149
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 4. Immunocytochemical localization of 'mi crotubule-associated protein 2 in the rat hippocampus and CA1 subficld 7 days after sham operation (a, b) or ischemia (c,d). a,c X 31: b, d X 300. Immunorcactivitv against the postsynaptie protein marker is lost from the CA1 subfield 7 days after ischemia.
a
b
c
Fig. 5. Photomicrograph of rat hippocampus and CAI subficld stained with 'synaptophysin' antibodies 7 days after sham operation (a, b) or ischemia (c,d). a,c X 31;b,d X 300. The immunostaining of the presynaptic protein marker, synaptophysin. was the same in sham-operated controls and postischemic rats.
Rischke/Rami/Bachmann/Rabic/ Krieglstein
Ischemia Resistance of Hippocampal Prcsynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
150
d
C A 1 -S U B F IE L D
located on hippocampal glial cells [21, 22] suggests that this measured reduction of CMRgic reflects a specific action of the drug on hippocampal neurons. It is difficult to compare the present CMRgic data in keta mine-treated rats with data from previous reports. Some studies were only qualitative [23, 24], while several other studies were per formed after a single injection of ketamine [25, 26]. Davis et al. [11], who used steady state conditions, only measured the brain glu cose uptake. This may be why no previous study reported a decrease in the CMRg|Cin the CAI strata oriens, pyramidal and radiatum of ketamine-treated, sham-operated rats. Most of these reports only recorded an increase in
the CMRgic of the hippocampal stratum lacu nosum moleculare [11,23, 24. 26], which has been suggested to be caused by activation of the perforant path [11. 23]. However, these findings are in good agreement with our find ing that ketamine markedly elevated the CMRgic in the CAI stratum lacunosum mo leculare of sham-operated rats (fig. la.b; ta ble 2). The CMRgic of the CA1 strata oriens. pyramidale and radiatum of postischemic rats were not reduced by ketamine treatment (fig. lb.c: table 2). We therefore conclude that the increased CMRgk. in the postischemic CAI subfield is not due to residual activity of postsynaptic neuronal structures. This con clusion agrees with the finding that the hippo
151
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 6. Strata of the hippocampal CAI subfield. O = Stratum oricns: P = stratum pyramidale: R = stratum radiatum; L = stratum lacunosum moleculare; DG = dentate gyrus: g = granular layer: H = hilus. Axons coming from the entorhinal cortex make synapses on the dendrites of CAI pyramidal cells in the stratum lacunosum moleculare (perforant path). Axons from the ipsilateral CA3 neurons (Schaffer collaterals) and from the contralateral CA3 neurons (commissurals) project to the dendrites of pyramidal cells in area CAI. Rat hippo campus was stained with antisynaptophysin (fig. 5).
152
In summary, the increased CMRg|Cin the strata oriens and pyramidale of the ischemi cally damaged CAI subfield did not seem to be due to hyperexcitation of surviving or damaged neural structures, but may be caused by activated astroglial cells. There is also a persistent ketamine-induced increase in CMRgic in stratum lacunosum moleculare of the CAI subfield, an area where presynaptic fibers of the perforant path terminate in the postischemic hippocampus. This, together with the immunohistochemical finding that the amount of presynaptic protein marker synaptophysin was the same in hippocampal sections of sham-operated and postischemic rats, suggests that presynaptic terminals re main intact in the postischemic CAI sub field.
Acknowledgements This research was supported by a grant (Kr 354/132) from the Deutsche Forschungsgemeinschaft and by a scholarship (R. Rischke) from the Philipps-Universi tät Marburg. A. Rami is a fellow of the Alexander von Humboldt-Stiftung. The authors are grateful to Mrs. R. Hartmannsgruber. Miss S. Engel and Mr. F. Caruso for their skilful technical assistance.
Rischke/Rami/Bachmann/Rabie/ Krieglstein
Ischemia Resistance of Hippocampal Presynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
campal theta-rhythm, which is the character istic frequency of the neuronal electrical ac tivity in the hippocampal CA1 subfield, de creased from day 2 onward after 20 min of cerebral ischemia in the rat [6]. We recently reported that the time course in the postischemic increase in CMRg)c of the CA1 subfield coincides with intense astroglial activation which has been most prominent in stratum pyramidale [5]. The astrocyte-immunostaining results also indicate that the increased hippocampal CMRg|C7 days after ischemia is due to astro glial metabolism (fig. 3b). Cooper et al. [9] pointed out that it is important to consider glial contribution to DOG uptake in rats sub jected to brain damage. It thus seems very likely that the postischemic increase in CMRgii- in the selectively vulnerable CA1 sub field is due to the metabolism of activated astrocytes. Interestingly, the CMRgic of ketaminetreated rats also increased in the stratum lacunosum moleculare of the ischemically damaged CAI subfield as well as in the in tact hippocampal tissue of control rats. The most plausible explanation for this effect could be an inhibition of presynaptic NMDA receptors by ketamine leading to an increased glutamate release and. thereby, to the increased CMRgic of presynaptic termi nals. At the moment we do not have clear evidence for that although Fink et al. [27] reported on presynaptic NMDA receptors in rat cerebral cortex which stimulated nor adrenaline release. On the other hand, the disinhibition of neurons by ketamine in the entorhinal cortex [11, 23] or of interneurons seem to be more speculative than the afore mentioned hypothesis. In any case, the keta mine-induced increase in CMRgjc in the CAI stratum lacunosum moleculare after isch emia indicates that the surviving presynaptic fibers are still excitable.
References 10 Sokoloff L. Reivich M. Kennedy C. Des Rosiers MH. Patlek CS, Sakurada O, Shinohara M: The [l4C]deoxyglucosc method for mea surement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neuro chem 1977;28:897-916. 11 Davis DW. Mans AM. Biebuyck JF, Hawkins RA: The influence of keta mine on regional brain glucose use. Anesthesiol 1988:69:199-205. 12 Auer RN, Olsson Y. Siesjo BK: Hy poglycemic brain injury in the rat: correlation of density of brain dam age with the EEG isoelectric time. Diabelis 1984;33:1090-1098. 13 Sternberger LA: Immunocytochcmistry. New York, Wiley. 1979, pp I104. 14 Hill AM, Cassoly R. Chetrite G, Pantaloni D: High molecular weight microtubule-associated proteins from pig brain arc immunologicallv related to human erythrocyte mem brane protein spectrin, ankyrin pro teins 4.1 and 4.2. Biol Cell 1985:53: 148-153. 15 Kitagawa K, Matsumoto M. Niinobe M, Mikoshiba K. Hata R, Ueda H. Handa N, Fukunaga R. Isaka Y. Kimura K. KamadaT: Mi crotubule-associated protein 2 as a sensitive marker for cerebral isch emic damage. Immunohistochemi cal investigation of dendritic dam age. Neuroscience 1989:31:401411. 16 Wiedemann B. Franke WW: Identi fication and localization of synaptophysin. an integral membrane glyco protein of Mr 38.000. characteristic of presynaptic vesicles. Cell 1985; 41:1017-1028. 17 Coan EJ. CollingridgeGL: Effects of phencyclidine. SKF 10.047 and re lated psychotomimetic agents on Nmethyl-O-aspartate receptor me diated synaptic responses in :at hip pocampal slices. Br J Pharmacol 1987;91:547-556. 18 Monaghan DT. Yao D. Oherman HJ, Watkins JC, Cotman CW: Auto radiography of D-2-[3H]amino-5phosphonopentanoatc binding sites in rat brain. Neurosci Lett 1982;52: 253-258.
19 Davies SN, Alford ST, Coan EJ, Lester RAJ. Collingridge GL: Keta mine blocks an NMDA receptormediated component of synaptic transmission in rat hippocampus in a voltagc-dcpcndcnt manner. Neu rosci Lett 1988:92:213-217. 20 Marietta MP. White PF. Pudwill CR. Way WL. Trevor J: Self-induc tion of metabolism. J Pharmacol Exp Ther 1976;196:536-544. 21 Backhus KH, Kettenmann H. Schachner M: Pharmacological characterization of the glutamate re ceptor in cultured astrocytes. J Neu rosci Res 1989;22:274-282. 22 Lehmann A. Hansson E: Kainateinduced stimulation of amino acid release from primary' astroglial cul tures of the rat hippocampus. Neu rochem Int 1988:13:557-561. 23 Hammer RP. Herkenham M, Pert CB, Quirion R: Correlation of re gional brain metabolism with recep tor localization during ketamine anesthesia: Combined autoradio graphic 2-[’H]deoxy-D-glucose re ceptor binding technique. Proc Natl AcadSci USA 1982;79:3067-3070. 24 Nelson SR. Howard RB. Cross RS. Samson F: Ketamine-induced changes in regional glucose utiliza tion in the rat brain. Anesthesiology 1980;52:330-334. 25 Cavazutti M. Porro CA, Biral GP. Benassi C. Barbiéri GC: Ketamine effects on local cerebral blood flow and metabolism in the rat. J Cereb Blood Flow Metab 1987.7:806811. 26 Crosby G, Crane AM. Sokoloff L: Local changes in cerebral glucose utilization during ketamine anesthe sia. Anesthesiology 1982:56:437— 443. 27 Fink K, Bonisch H. Gdthert M: Pre synaptic NMDA receptors stimulate noradrenaline release in the cerebral cortex. Eur J Pharmacol 1990:185: 115-117.
153
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
1 KirinoT: Delayed neuronal death in the gcrbil hippocampus following ischemia. Brain Res 1982:239:57— 69. 2 Pulsinelli WA. Bricrly JB. Plum F: Temporal profile of neuronal dam age in a model of transient forebrain ischemia. Ann Neurol 1982:11:491498. 3 Smith ML. Bendek G. Dahlgren N. Rosen I, Wieloch T, Siesjo BK: Models for studying long-term re covery following forebrain ischemia in the rat. A 2-vessel occlusion mod el. Acta Neurol Scand 1984:69:385— 401. 4 Rischke R. Krieglstein J: Increased LCGU and decreased LCBF in rat hippocampus 7 days after ischemia. J Neurochem 1989:52:S56B. 5 Rischke R. Krieglstein J: Postischcmic neuronal damage causes astro glial activation and increase in local cerebral glucose utilization of rat hippocampus. J Cereb Blood Flow Metab 1991:11:106-113. 6 Monmaur P. Thomson MA, Harzi MM: Temporal changes in hippo campal theta activity following twenty minutes of forebrain isch emia in the chronic rat. Brain Res 1986:378:262-273. 7 Kirino T, Tamura A, Sano K: Chronic maintenance of prcsynaptic terminals in gliotic hippocampus following ischemia. Brain Res 1990: 510:17-25. 8 Yamamoto K. Yoshimine T. Hornburger HA, Yanagihara T: Immunohistochcmical investigation of re gional cerebral ischemia in the gerbil: Occlusion of the posterior com municating artery. Brain Res 1986: 371:244-252. 9 Cooper RM, Thurlow GA, Rooney BJ: 2-Dcoxvglucose uptake and his tologic changes in rat thalamus after neocortical ablations. Exp Neurol 1984:83:134-143.
Pharmacology 1992;45:142-153
R. Rischkea A. Ram ia U. Bachmanna A. Rahiéh J. Krieglsteina Institut für Pharmakologie und Toxikologie, Fachbereich Pharmazie und Lebensmittelchemie der Philipps-Universität, Ketzerbach, Marburg, BRD; Laboratoire de Neurobiologie Endocrinologique, URA 1197 du CNRS, Université Montpellier II, Montpellier, France
Activated Astrocytes, but not Pyramidal Cells, Increase Glucose Utilization in Rat Hippocampal CA1 Subfield after Ischemia
Abstract
Received: April 17.1991 Accepted: December 2,1991
Dr. A. Rami Institut für Pharmakologie und Toxikologie Fachbereich Pharmazie und Lebensmittelchemie. Philipps-Universität. Ketzerbach 63. D-W-3550 Marburg (FRG)
©1992 S. Karger AG. Basel 0031-7012/92/ 0453—0142$2.75/0
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
KeyWords Ischemia Rat Hippocampus Local cerebral glucose utilization Ketamine Synaptophysin Microtubule associated protein 2 Glial fibrillary acidic protein
The local cerebral glucose utilization (CMRgic) in the damaged rat hippo campal CAl subfield increases 7 days after lOmin of cerebral ischemia. We have used the N-methyl-D-aspartate antagonist (NMDA antagonist) ketamine in rats 7 days after sham operation or cerebral ischemia to deter mine whether the elevated postischemic CMRgic of the CA1 subfield is due to long-lasting hyperexcitation of surviving or injured neurons, or, alterna tively, to the metabolism of other cell types. The autoradiographic data were interpreted with the aid of histochcmical analysis of the postischemic hippocampal cell changes. Anesthetic doses of ketamine significantly reduced the CMRg|Cin the CAl strata oriens, pyramidale and radiatum of sham-operated rats, while the postischemic increases in CMRgic in these hippocampal CA 1 strata were not affected by ketamine. In addition, there were ketamine-induced increases in the CMRg|Cof the CAl stratum lacunosum moleculare of both sham-operated and postischemic rats. The immunoreactivity of the microtubule-associated protein 2 (MAP2), a postsynaptic protein marker, was decreased markedly in the CAl subficld in postischemic rats, while the prcsynaptic protein marker, synaptophysin, remained the same in sham-operated and postischemic rats. The glial fibrillary acidic protein (GFAP) immunoreactivity of astrocytes raised markedly in the ischemically damaged CAl subfield. Although it could be demonstrated that presynaptic terminals remain intact in the postisch emic damaged CAl subfield, the lacking ketamine effect on CAl pyrami dal neurons indicated that the increase in CMRgic in this brain area is not due to postsynaptic neural hyperexcitation, but probably has to be attrib uted to astrocytes activated by neuronal damage.
There is general agreement that the pyram idal neurons in the hippocampal CA1 subfield are particularly vulnerable to transient isch emia [1,2]. About 70-90% of the neurons in the hippocampal CA1 subfield of rats were damaged 7 days after a 10-min forebrain isch emia [2, 3], However, despite this postischemic damage, recent autoradiographic mea surements using [l4C]-labeled 2-deoxy-Z)-glucose have revealed an increase in local cere bral glucose utilization (CMRg|C) in different strata of the CA1 subfield [4, 5]. CMRgic is considered to be closely coupled to the func tional activity of neuronal elements in the normal brain, but the cause of increased hip pocampal CMRgic in the ischemically dam aged CA1 subfield remains controversial. While there is evidence that activation of astrocytes [6-8] produces an increase in CMRg|Cof the CA1 subfield after ischemia [5, 9], it is always possible that long-lasting hy perexcitation of the damaged neurons may contribute to this phenomenon. We have, therefore, studied the pathophys iology underlying the increased CMRgic of the CA1 subfield 7 days after ischemia by mea suring the CMRgic in rats whose hippocampal neuronal activity was modified by anesthetic doses of ketamine 7 days after sham operation or ischemia. The resulting autoradiographic data were interpreted in conjunction with his tological examination of the hippocampal neurons, astrocytes and macrophages, plus hippocampal pre- and postsynaptic struc tures.
Materials and Methods Animals
Male Wistar rats (Ivanovas, Kisslcgg, FRG) weigh ing 250-300 g were maintained under controlled light ing and environmental conditions (12 h dark/light cy
cle. 23 ± 1 °C. 55 ± 5% relative humidity) and fed a standard diet (Altromin. Lage. FRG) and tap water ad libitum. The rats were fasted overnight preceding the ischemic procedure. Materials
[l4C]-2-deoxy-£)-glucose (=DOG; specific activity 51.1-52.5 mCi/mmol) was purchased from NEN (Dreieich. FRG). Mouse monoclonal antibodies di rected against glial fibrillary acidic protein (GFAP). rabbit antimouse y-globulin, mouse peroxidase-antihorscradish peroxidase complex (PAP) and 3,3'-diaminobenzidine (DAB) were obtained from Serva (Hei delberg, FRG). The anti-microtubule associated pro tein 2 (MAP2) serum was a gift from Dr. A.M. Hill (Gif-sur-Yvette, France). The second antibody, goat anlirabbit y-globulin and the rabbit PAP complex were also obtained from Serva. Monoclonal antibodies di rected against synaptophvsin were purchased from Boeringer Mannheim (FRG). Biotinylated sheep antimouse y-globulin and streptavidin-peroxidase com plex were obtained from Amersham (Braunschweig, FRG). All other chemicals were of reagent grade. The brains were fixed by immersion in a mixture of 60% ethanol, 30% chloroform and 10% acetic acid (Carnoy’s solution). Induction o f Ischemia
Ischemia was induced according to Smith et al. [3], The rats were anesthetized with 3.5% halothane and connected to a Starling-type respirator delivering 0.8 % halothane in a 2:1 mixture of nitrous oxide and oxy gen. The common carotid arteries were isolated via a cervical incision and ligatures were placed around them. A polyethylene cannula was inserted into the tail artery for blood sampling and blood pressure record ing, and a silicone catheter was advanced into the infe rior caval vein via the right jugular vein. Heparin was injected i.v. (200 IU/kg) to prevent coagulation, and muscle paralysis was maintained with 5 mg/kg i.v. sux amethonium chloride every 15-20 min. Halothane was discontinued 30 min before starting ischemia to reach a preischemic steady state. Trimethaphan cam phor sulfonate (5 mg/kg) was infused and forebrain ischemia was induced by carotid clamping and exsanguination to a blood pressure of 40 mm Hg. Blood pressure was restored after 10 min of ischemia by removing the carotid clamps and reinfusing the re moved blood. A solution of NaHCOj (0.6 mmol/1) was then injected (0.5 ml), and the wound were sutured. The rats were ventilated for about 45 min until they regained consciousness, when all catheters were re moved. Body temperature was kept at 37 °C with a
143
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Introduction
Local Cerebral Glucose Utilization Seven days after sham-operation or cerebral isch emia. CMRrf.- was measured in ketamine treated and untreated rats according to the 2-|llC]-deoxy glucose method described by Sokoloff ct al. 110). I he femoral vessels were cannulated with polyethylene catheters under light anesthesia with a 2:1 mixture of nitrous oxidc/oxygcn containing 0.8% halothane and the wound was packed with lidocaine gel and closed. The animals were immobilized with plaster cats, and fixed on a preparation desk, where they were allowed to recover from anesthesia for 2 h before measuring CMR*!«.. Body temperature was held at 3 7 °C with a healing lamp. The physiological variables were deter mined before infusion of the radiotracer. CMR,k. was detei mined by infusing 120 pCi/kg DOG in physiolog ical saline via the femoral vein over 25 s. A total o f 16 arterial blood samples (-80 pi) wctc collected during 45 nun from the beginning o f the experiment for mea suring glucose and DOG. Immediately after the lust sample had been taken (46 nun) the rats were decapi tated. the brains dissected out and immediately frozen in isopentane ( - 5 0 * 0 . I he brains were sectioned (20 pm) and the sections exposed to Osray M3 film (Agfa-Gcvacrt. Leverkusen. FRG) for 10 days. I he optical densities o f the autoradiograms were measured with a computer-based image analyzer (IBAS. Kontron. Hchingcn. FRG). Ketamine Administration A total of 60 mg/Kg of ketamine was administered. I he infusion schedule was designed to maintain anes thetic blood levels throughout the 45 min of the auto radiographic experiments 111). A bolus i.v. injection of 30 mg/kg ketamine was first given \ ia the femoral vein over a I-min period. The physiological parameters were determined immediately theralter. and IXXi was infused over the next 25 s. A second booster dose of 30 mg/kg ketamine was infused over the following I 5 min using an infusion pump (Original Perfusor. Bruun-Mctsungcn. FRG). Controls were given physio logical saline by the same protocol.
144
Histology Seven days after sham operation or ischemia ) rats were killed by decapitation. The brain were carefully removed, fixed in Camoy’s solution, and embedded in paraffin. Sections (5 pm) w-cre cut 6 mm rostral of the teritoiiul incision, and mounted on chromc-alum-gelatinc-coatcd slides. Adjacent sections from each brain were stained by each of the procedures described below.
Histochemtcal Procedures Neuronal damage was assessed by staining the brain sections w ith acidic fuchsin/cclcsline blue. Dam aged neurons are stained violet |I2). Intact and dam aged neurons were counted in entire hippocampal sub fields and neuronal damage was calculated as the per centage of damaged pyramidal cells. An adjacent sec tion was stained with hematoxylin and eosin (HE) to visualize macrophages.
Imniunohistocheniical Procedures Astrocytes were immunosiained by the PAP method [131. Paraffin was removed from the sections with toluene and they were rchydrated. washed in phosphate-buffered saline (PBS: I0pmol/1. pH 7.4) and flooded for 16h at with diluted mouse pri mary antibody (1:5 in PBS containing 1% normal swine serum. PBS-NSS). The sections were then carelully washed w ith PBS and flooded w ith rabbit second ary antibody (1:50 in PBS-NSS) for 30 min. washed again with PBS. and covered with the mouse PAP (1:50 in PBS-NSS) for 20 min. Finally, the sections were washed in PBS. incubated for 10-15 min with 0.05% DAB containing 0.01 % H ;0; in PBS. washed several times with PBS and mounted with Permount (Fisher Scientific Company. USA). MAP2 was immunolocalizcd using the same procedure with a rabbit antiserum [I I]. The second antibody and the rabbit PAP were diluted 1:50 and 1:100 in PBS containing 5% bovine serum albumin (BSA). Immunohistochcmical staining o f synaptophysin was performed by using the biotin-streptavidin system (Arnersham). Tris buffer (0.05 mol/l Tris. 0.12 mol/l N a d . pi I 7.6) was used throughout. The dcparafTincd. rehydrated sections were flooded for 16 li al 4 “C with diluted monoclonal antibody (1:5 in Tris buffer con taining 5% BSA. I ris-BSA). The sections were then carefully washed with buffer and flooded with biotiny lated sheep antimouse -/-globulin (diluted 1:130 in Tris-BSA) for 30 min. The sections were washed with I ris-BSA and covered with the strcptavidin-piroxidasc complex (diluted 1:200 in Tris-BSA) for 20 min. The slides were washed in Tris. incubated for 10-
Rischke/Rami/Bachmann/Rabii/
Kncglstcm
Ischemia Resistance o f Hippocampal Prcsynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
healing lamp. Blood gases, blood pH (Corning 178. Corning Medical. Giessen. FRG). blood pressure(Statham P23DB. Hcio Rev. Puerto Rico: IISF-Flcktromanometer. Hugstetten, FRG), and plasma glucose (Beckman Glucose Analyzer. Munich. FRG) were de termined before induction o f ischemia. Sham-operated control rats underwent all the ex perimental procedures except occlusion o f the carotid arteries and exsanguination.
Pharmacology, vol. 45 S. Karger. Basel
Rixchke/ Rami/Bachmantt/Rabie/ Kncglstcm
Plaie I
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 1. Representative autoradiograms o f (,4C]-deoxy-£>-glucosc metabo lism in rat hippocampus of ketamine-treated and untreated rats 7 days after sham operation or ischemia, a Sham operation, untreated, b Sham operation, ketamine treated, c Ischemia, untreated, d Ischemia, ketamine-treated. Keta mine was administered by a bolus injection (30 mg/kg i.v.) immediately before tracer infusion followed by a booster dose (30 mg/kg i.v.) during the next 15 min after the tracer infusion to maintain ketamine anesthesia.
Table 1. Physiological variables
Before sham operation (n = 12) paCK mm Hg PaCCT, mm Hg Arterial pH Plasma glucose, mg/l Blood pressure, mm Hg Temperature, °C
7 days after ischemia
sham operation
ischemia
( n - 14)
untreated (n = 6)
untreated ( n - 7)
143 ±17 137 ± 10 37±2 38 ±2 7.40 ±0.02 7.39 ±0.01 1,190 ± 120 1.270± 150 132 ±13 129 ± 11 37.0±0.2 36.9 ±0.1
ketamine (n = 6)
98 ±10 99 ± 12 39 ± 1 42 ±2* 7.40 ±0.12 7.38 ±0.09 1,630 ±90 l.580± 140 145 ±15 131 ± 12 37.0 ±0.1 37.0±0.2
ketamine (n = 7)
95 ±12 96 ± 8 37 ± 2 41 ± 2* 7.41 ±0.16 7.39 ± 0.07 1.730 ±150 1.650 ± 110 140 ±12 132 ± 10 37.0 ±0.2 37.0± 0.2
!5min with 0.05% DAB containing 0.01% HjO? in Tris, washed several times with Tris and finally mounted in Permount. Control sections were incubated with Tris instead of primary antibody. They showed no positive reac tion to GFAP, MAP2 or synaptophysin antibodies. The immunohistochemical stainings were assessed qualitatively by light microscopy.
Results
The physiological variables determined be fore sham operation or ischemia were in the normal range for both the control and test groups of rats (table 1). The physiological pa rameters measured 7 days after sham opera tion or cerebral ischemia were normal for the 2 untreated groups; intravenous 30 mg/kg ke tamine caused a significant increase in the PaCCb of all treated rats (table 1). The behav ior of the rats during ketamine administration was the same for the sham-operated and isch emic rats. There were occasional head and limb movements at the start of the bolus injection, but the rats remained motionless
during infusion of the booster dose and for a 30 min period. CMRgic was determined in 6 defined hip pocampal areas of untreated rats and rats treated with ketamine 7 days after sham oper ation and ischemia (table 2). The strata oriens and pyramidale of the CA1 subfield were ana lyzed together, because it was not possible to differentiate between them on the autoradio grams (see fig. 6). The treatment of rats with anesthetic doses of ketamine 7 days after sham-operation induced uneven changes in CMRg|C. The CMRgic of the CA1 strata oriens. pyramidale and radiatum was reduced during ketamine anesthesia, while the CMRg|t. of the CA1 stratum lacunosum moleculare in creased markedly (fig. la.b. 6). The CMRg|C of the CA3 and CA4 subfields and the dentate gyrus did not change significantly under these conditions (fig. la.b). As demonstrated previously, CMRgic in the CA1 strata oriens and pyramidale increased 7 days after forebrain ischemia (fig. la.c). In contrast to the measurements performed 7 days after sham operation, administration of
145
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
The physiological variables were determined 30 min after beginning artificial ventilation, before sham opera tion or ischemia in both ketamine-treated and untreated rats, and 7 days after sham operation or ischemia. Values are the means ± SD of n experiments. * p < 0.05 (Mann-Whitney U test).
Table 2. The effect of ketamine on CMRgic of the rat hippocampal subfields 7 days after sham operation or
ischemia Hippocampal area
CAl strata Oriens and pyramidale Radiatum Lacunosum moleculare CA3 CA4 Dentate gyrus
Sham operation
Ischemia
untreated (n -6 )
ketamine (n - 6)
untreated (n = 7)
ketamine (n= 7)
480±50 570± 100 600 ±90 570 ±80 470 ±90 490 ±100
4I0±60* 430 ±50* 1,070 ±130** 490 ±40 470 ±40 600 ±70
610 ± 120* 540± 110 630 ±120 560 ±120 500± 110 470± 110
560± 110 550± 110 740 ± 110+ 550±80 480 ±80 490 ±90
CMRgic is given as pmol kg"1 min-1. Ketamine was given intravenously 7 days after sham operation or cerebral ischemia. Anesthetic plasma levels of ketamine throughout the period of tracer uptake were obtained by giving a 30-mg/kg bolus injection, followed by immediate tracer infusion, and a second infusion of ketamine (30 mg/kg) over the next 15 min. The data were analyzed using the Mann-Whitncy U test. Values are means ± SD of n experiments. * p < 0.05; ** p < 0.01: compared with the vehicle group after sham operation. + p < 0.05: compared with the corresponding untreated ischemic control.
146
fields of the hippocampus in sham-operated controls (fig. 4a.b), but much less MAP2 im munostaining in the CA1 subfield of postischemic rats (fig. 4c.d). However, synaptophysin-immunostaining was the same in hippo campal sections of sham-operated controls and postischemic rats (fig. 5a-d).
Discussion
The histological appearance of the rat hip pocampus confirms earlier findings that a transient forebrain ischemia mainly injures the pyramidal neurons in the CA1 subfield [1. 2], The loss of immunoreactivity for the postsynaptic somal and dendritic protein marker MAP2 [15] from theCAl subfield 7 days after ischemia also clearly shows the selective vul nerability of pyramidal cells in this brain area (fig. 4a-d. 6). Interestingly, visualization of presynaptic terminals by synaptophysin im-
Rischke/Rami/Bachmann/Rabic.' Krieglstein
Ischemia Resistance of Hippocampal Presynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
ketamine 7 days after recirculation did not reduce the CMRgic of the CA1 strata oriens, pyramidale and radiatum (fig. lc.d). The ke tamine-induced increase in the CMRgic in the CA1 stratum lacunosum moleculare of shamoperated rats also occurred in rats given the anesthetic 7 days after ischemia (fig. lc.d). No damaged neurons could be seen in the hippocampal subfields of sham-operated con trols (fig. 2a), but the 10-min forebrain isch emia damaged 83% of pyramidal cells in the CA1 subfield (fig. 2b), 2% of neurons in the CA3 subfield, and 8% of CA4 and hilar neu rons. No macrophages were seen in HEstained control sections or in sections of brains 7 days after ischemia. GFAP immunocytochemical staining was scattered through out the hippocampus 2 days postischemia (fig. 3a). In contrast, there was intense antiGFAP immunoreactivity in astrocytes in the CA1 subfield 7 days after ischemia (fig. 3b). There was MAP2 immunostaining in all sub
munostaining [16] revealed no differences be tween the hippocampal CA1 areas of shamoperated and postischemic rats (fig. 5a-d). This agrees with the findings of Kirino et al. [7], who demonstrated by electron micros copy that presynaptic terminals survive and maintain their structural integrity in the post ischemic CA1 subfield of the gerbil. As described recently [4. 5], the postisch emic neuronal damage in the CA1 subfield is associated with an increase in CMRg|C (fig. 1a,c). The effects of ketamine anesthesia on CMRgic in sham-operated and in postisch emic rats should help to show whether the
increased CMRgic in the postischemic CA1 subfield is due to excitation of remaining neu rons or to activation of other cells. Ketamine acts as a noncompetitive antagonist on the excitatory NMDA receptor [17]. high concen trations of which are present in the hippocam pal CA1 subfield [18]. Inhibiting the NMDA receptor with ketamine causes a specific blockade of postsynaptic neurons in the hip pocampal CA1 subfield [19]. Thus, if the increased CMRg|c of the CA1 subfield 7 days after ischemia was a result of enhanced neural excitation, there should be a ketamine-in duced reduction in the CMRg|C.
147
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 2. Light photomicrographs show the hippocampal CA1 subfield 7 days after sham operation (a) and ischemia (b). Arrows indi cate some damaged neurons. X 300.
The half-life of ketamine is 10 min in the plasma and 8 min in the brain of the rat [20]. resulting in an 8-min period of hypnosis when a single intravenous injection of 30 mg/kg ke tamine is given [20]. However, a single injec tion of ketamine makes the measurement of CMRg|C, which requires steady state condi tions, difficult. In oder to measure the CMRgic during an unchanged anesthetic state, we de cided to give a loading dose of 30 mg/kg fol lowed by infusion of a further 30 mg/kg dose. The rat’s behavior, as mentioned in the
148
present study, indicated dissociative anes thetic conditions for a 30-min period, suggest ing that the drug was present in the brain throughout the tracer uptake. Under these conditions, ketamine caused a significant reduction in CMRg|C of the CA1 strata oriens, pyramidale and radiatum of sham-operated control rats (fig. la,b: table 2). This agrees with the finding that ketamine depresses synaptic excitation of CA1 neurons [19] by blocking the NMDA receptor [17], The fact that the NMDA receptor is not
Rischkc/Rami/Bachmann/Rabi6' Krieglstein
Ischemia Resistance of Hippocampal Presynaplic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 3. GFAP-positive astro glial cells in the rat hippocampus 2 days after ischemia (a) or 7 days after ischemia (b). X 300.
a
w V
c
149
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 4. Immunocytochemical localization of 'mi crotubule-associated protein 2 in the rat hippocampus and CA1 subficld 7 days after sham operation (a, b) or ischemia (c,d). a,c X 31: b, d X 300. Immunorcactivitv against the postsynaptie protein marker is lost from the CA1 subfield 7 days after ischemia.
a
b
c
Fig. 5. Photomicrograph of rat hippocampus and CAI subficld stained with 'synaptophysin' antibodies 7 days after sham operation (a, b) or ischemia (c,d). a,c X 31;b,d X 300. The immunostaining of the presynaptic protein marker, synaptophysin. was the same in sham-operated controls and postischemic rats.
Rischke/Rami/Bachmann/Rabic/ Krieglstein
Ischemia Resistance of Hippocampal Prcsynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
150
d
C A 1 -S U B F IE L D
located on hippocampal glial cells [21, 22] suggests that this measured reduction of CMRgic reflects a specific action of the drug on hippocampal neurons. It is difficult to compare the present CMRgic data in keta mine-treated rats with data from previous reports. Some studies were only qualitative [23, 24], while several other studies were per formed after a single injection of ketamine [25, 26]. Davis et al. [11], who used steady state conditions, only measured the brain glu cose uptake. This may be why no previous study reported a decrease in the CMRg|Cin the CAI strata oriens, pyramidal and radiatum of ketamine-treated, sham-operated rats. Most of these reports only recorded an increase in
the CMRgic of the hippocampal stratum lacu nosum moleculare [11,23, 24. 26], which has been suggested to be caused by activation of the perforant path [11. 23]. However, these findings are in good agreement with our find ing that ketamine markedly elevated the CMRgic in the CAI stratum lacunosum mo leculare of sham-operated rats (fig. la.b; ta ble 2). The CMRgic of the CA1 strata oriens. pyramidale and radiatum of postischemic rats were not reduced by ketamine treatment (fig. lb.c: table 2). We therefore conclude that the increased CMRgk. in the postischemic CAI subfield is not due to residual activity of postsynaptic neuronal structures. This con clusion agrees with the finding that the hippo
151
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
Fig. 6. Strata of the hippocampal CAI subfield. O = Stratum oricns: P = stratum pyramidale: R = stratum radiatum; L = stratum lacunosum moleculare; DG = dentate gyrus: g = granular layer: H = hilus. Axons coming from the entorhinal cortex make synapses on the dendrites of CAI pyramidal cells in the stratum lacunosum moleculare (perforant path). Axons from the ipsilateral CA3 neurons (Schaffer collaterals) and from the contralateral CA3 neurons (commissurals) project to the dendrites of pyramidal cells in area CAI. Rat hippo campus was stained with antisynaptophysin (fig. 5).
152
In summary, the increased CMRg|Cin the strata oriens and pyramidale of the ischemi cally damaged CAI subfield did not seem to be due to hyperexcitation of surviving or damaged neural structures, but may be caused by activated astroglial cells. There is also a persistent ketamine-induced increase in CMRgic in stratum lacunosum moleculare of the CAI subfield, an area where presynaptic fibers of the perforant path terminate in the postischemic hippocampus. This, together with the immunohistochemical finding that the amount of presynaptic protein marker synaptophysin was the same in hippocampal sections of sham-operated and postischemic rats, suggests that presynaptic terminals re main intact in the postischemic CAI sub field.
Acknowledgements This research was supported by a grant (Kr 354/132) from the Deutsche Forschungsgemeinschaft and by a scholarship (R. Rischke) from the Philipps-Universi tät Marburg. A. Rami is a fellow of the Alexander von Humboldt-Stiftung. The authors are grateful to Mrs. R. Hartmannsgruber. Miss S. Engel and Mr. F. Caruso for their skilful technical assistance.
Rischke/Rami/Bachmann/Rabie/ Krieglstein
Ischemia Resistance of Hippocampal Presynaptic Fibers
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
campal theta-rhythm, which is the character istic frequency of the neuronal electrical ac tivity in the hippocampal CA1 subfield, de creased from day 2 onward after 20 min of cerebral ischemia in the rat [6]. We recently reported that the time course in the postischemic increase in CMRg)c of the CA1 subfield coincides with intense astroglial activation which has been most prominent in stratum pyramidale [5]. The astrocyte-immunostaining results also indicate that the increased hippocampal CMRg|C7 days after ischemia is due to astro glial metabolism (fig. 3b). Cooper et al. [9] pointed out that it is important to consider glial contribution to DOG uptake in rats sub jected to brain damage. It thus seems very likely that the postischemic increase in CMRgii- in the selectively vulnerable CA1 sub field is due to the metabolism of activated astrocytes. Interestingly, the CMRgic of ketaminetreated rats also increased in the stratum lacunosum moleculare of the ischemically damaged CAI subfield as well as in the in tact hippocampal tissue of control rats. The most plausible explanation for this effect could be an inhibition of presynaptic NMDA receptors by ketamine leading to an increased glutamate release and. thereby, to the increased CMRgic of presynaptic termi nals. At the moment we do not have clear evidence for that although Fink et al. [27] reported on presynaptic NMDA receptors in rat cerebral cortex which stimulated nor adrenaline release. On the other hand, the disinhibition of neurons by ketamine in the entorhinal cortex [11, 23] or of interneurons seem to be more speculative than the afore mentioned hypothesis. In any case, the keta mine-induced increase in CMRgjc in the CAI stratum lacunosum moleculare after isch emia indicates that the surviving presynaptic fibers are still excitable.
References 10 Sokoloff L. Reivich M. Kennedy C. Des Rosiers MH. Patlek CS, Sakurada O, Shinohara M: The [l4C]deoxyglucosc method for mea surement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neuro chem 1977;28:897-916. 11 Davis DW. Mans AM. Biebuyck JF, Hawkins RA: The influence of keta mine on regional brain glucose use. Anesthesiol 1988:69:199-205. 12 Auer RN, Olsson Y. Siesjo BK: Hy poglycemic brain injury in the rat: correlation of density of brain dam age with the EEG isoelectric time. Diabelis 1984;33:1090-1098. 13 Sternberger LA: Immunocytochcmistry. New York, Wiley. 1979, pp I104. 14 Hill AM, Cassoly R. Chetrite G, Pantaloni D: High molecular weight microtubule-associated proteins from pig brain arc immunologicallv related to human erythrocyte mem brane protein spectrin, ankyrin pro teins 4.1 and 4.2. Biol Cell 1985:53: 148-153. 15 Kitagawa K, Matsumoto M. Niinobe M, Mikoshiba K. Hata R, Ueda H. Handa N, Fukunaga R. Isaka Y. Kimura K. KamadaT: Mi crotubule-associated protein 2 as a sensitive marker for cerebral isch emic damage. Immunohistochemi cal investigation of dendritic dam age. Neuroscience 1989:31:401411. 16 Wiedemann B. Franke WW: Identi fication and localization of synaptophysin. an integral membrane glyco protein of Mr 38.000. characteristic of presynaptic vesicles. Cell 1985; 41:1017-1028. 17 Coan EJ. CollingridgeGL: Effects of phencyclidine. SKF 10.047 and re lated psychotomimetic agents on Nmethyl-O-aspartate receptor me diated synaptic responses in :at hip pocampal slices. Br J Pharmacol 1987;91:547-556. 18 Monaghan DT. Yao D. Oherman HJ, Watkins JC, Cotman CW: Auto radiography of D-2-[3H]amino-5phosphonopentanoatc binding sites in rat brain. Neurosci Lett 1982;52: 253-258.
19 Davies SN, Alford ST, Coan EJ, Lester RAJ. Collingridge GL: Keta mine blocks an NMDA receptormediated component of synaptic transmission in rat hippocampus in a voltagc-dcpcndcnt manner. Neu rosci Lett 1988:92:213-217. 20 Marietta MP. White PF. Pudwill CR. Way WL. Trevor J: Self-induc tion of metabolism. J Pharmacol Exp Ther 1976;196:536-544. 21 Backhus KH, Kettenmann H. Schachner M: Pharmacological characterization of the glutamate re ceptor in cultured astrocytes. J Neu rosci Res 1989;22:274-282. 22 Lehmann A. Hansson E: Kainateinduced stimulation of amino acid release from primary' astroglial cul tures of the rat hippocampus. Neu rochem Int 1988:13:557-561. 23 Hammer RP. Herkenham M, Pert CB, Quirion R: Correlation of re gional brain metabolism with recep tor localization during ketamine anesthesia: Combined autoradio graphic 2-[’H]deoxy-D-glucose re ceptor binding technique. Proc Natl AcadSci USA 1982;79:3067-3070. 24 Nelson SR. Howard RB. Cross RS. Samson F: Ketamine-induced changes in regional glucose utiliza tion in the rat brain. Anesthesiology 1980;52:330-334. 25 Cavazutti M. Porro CA, Biral GP. Benassi C. Barbiéri GC: Ketamine effects on local cerebral blood flow and metabolism in the rat. J Cereb Blood Flow Metab 1987.7:806811. 26 Crosby G, Crane AM. Sokoloff L: Local changes in cerebral glucose utilization during ketamine anesthe sia. Anesthesiology 1982:56:437— 443. 27 Fink K, Bonisch H. Gdthert M: Pre synaptic NMDA receptors stimulate noradrenaline release in the cerebral cortex. Eur J Pharmacol 1990:185: 115-117.
153
Downloaded by: Stockholm University Library 130.237.165.40 - 12/18/2018 2:34:46 PM
1 KirinoT: Delayed neuronal death in the gcrbil hippocampus following ischemia. Brain Res 1982:239:57— 69. 2 Pulsinelli WA. Bricrly JB. Plum F: Temporal profile of neuronal dam age in a model of transient forebrain ischemia. Ann Neurol 1982:11:491498. 3 Smith ML. Bendek G. Dahlgren N. Rosen I, Wieloch T, Siesjo BK: Models for studying long-term re covery following forebrain ischemia in the rat. A 2-vessel occlusion mod el. Acta Neurol Scand 1984:69:385— 401. 4 Rischke R. Krieglstein J: Increased LCGU and decreased LCBF in rat hippocampus 7 days after ischemia. J Neurochem 1989:52:S56B. 5 Rischke R. Krieglstein J: Postischcmic neuronal damage causes astro glial activation and increase in local cerebral glucose utilization of rat hippocampus. J Cereb Blood Flow Metab 1991:11:106-113. 6 Monmaur P. Thomson MA, Harzi MM: Temporal changes in hippo campal theta activity following twenty minutes of forebrain isch emia in the chronic rat. Brain Res 1986:378:262-273. 7 Kirino T, Tamura A, Sano K: Chronic maintenance of prcsynaptic terminals in gliotic hippocampus following ischemia. Brain Res 1990: 510:17-25. 8 Yamamoto K. Yoshimine T. Hornburger HA, Yanagihara T: Immunohistochcmical investigation of re gional cerebral ischemia in the gerbil: Occlusion of the posterior com municating artery. Brain Res 1986: 371:244-252. 9 Cooper RM, Thurlow GA, Rooney BJ: 2-Dcoxvglucose uptake and his tologic changes in rat thalamus after neocortical ablations. Exp Neurol 1984:83:134-143.
