Pathophysiological aspects of acute experimental allergic encephalomyelitis

Supplementum No. 119

. Vol.

78

. 1988

Pathophysiological aspects of acute experimental allergic encephalomyelitis Marianne Juhler Department of Neurology, Rigshospitalet , Copenhagen, Denmark

Neurologica xanamavica MUNKSGAARD

COPENHAGEN

MUNKSGAARD International Booksellers and Publishers Ltd. 35 Nsrre Ssgade, DK 1370 Copenhagen K Denmark ISBN: 87-16-064441-0 ISSN: 0065-1427

Printed in Denmark by A. Backhausen, Holger J . Serrensen aps, Horsens

PATHOPHYSIOLOGY OF EAE

Summarv J

Traditionally, research in experimental allergic encephalomyelitis (EAE) has focussed on immunological and histopathological aspects. The present review introduces a physiological approach to EAE. As EAE is characterized by many small, focal lesions in the central nervous system (CNS), methods with a high spatial resolution should be used to conduct studies on regional pathophysiology in the condition. Quantitative autoradiography seems an ideal method as it offers, 1) high regional resolution (approximately 50 um), 2) precise quantitation and, 3) a direct correlation between regional histopathology and pathophysiology. By the use of this method, the author has performed studies on 1) regional blood-brain barrier (BBB) permeability, and 2) regional metabolism of energy substrate and related subjects, (i.e. regional cerebral blood flow, regional cerebral glucose metabolic rate and regional pH). Corresponding to the EAE lesions (lymphocytic accumulations), there is a considerable increase in BBB permeability. Metabolism of energy substrate at the lesion sites is severelyderanged, which is expressed in a CBF/CMR ratio of3 ml/pmol compared to the normal 1.5 ml/pmol. No changes in regional pH are seen in the lesions. Unrelated to the lesion sites there is a 50% decrease in blood flow in cerebral cortex. This observation probably reflects a functional decrease in cortical flow due to sensory motor impairment. Key words: experimental allergic encephalomyelitis (EAE) - pathophysiology - bloodbrain barrier (BBB) permeability - regional cerebral blood flow (rCBF) - regional cerebral glucose metabolic rate (rCMR) - regional cerebral pH (r-pH) - methodological considerations.

3

4

M. JUHLER

Dansk resum6 Traditionelt har forskning vedrarende eksperimentel allergisk encephalomyelitis (EAE) beskaeftiget sig med immunologiske og histopatologiske forandringer. Nzrvzrende oversigt introducerer en fysiologisk synsvinkel pa EAE. Da EAE er kendetegnet ved mange sm8 fokale liesioner i centralnervesystemet (CNS), bor man anvende methoder med en h a j regional oplaselighed for at beskrive patofysiologien ved tilstanden. Kvantitativ autoradiografi synes at viere en ideel metode, idet den yder

1) h a j regional oploselighed (i starrelsesorden 50 mikrometer), 2) praecis kvantitering af fysiologiske parametre og 3) direkte korrelation mellem regional histopatologi og patofysiologi. Ved anvendelse af denne metode har forfatteren ti1 nzrvierende afhandling udfort studier vedrarende regional blod-hjerne-barriere (BBB) permeabilitet, og 2) regional omsztning af energirige substrater og tilhmende fysiologiske processer (dvs. regional cerebral blodgennemstromning, regional cerebral glucosemetabolisme og regionalt cerebralt pH. Svarende ti1 EAE lzsionerne er der en vzsentlig agning i BBB permeabiliteten, hvilket i naervaerende afhandling er relateret ti1 immunologiske aktive substansers adgang ti1 centralnervesystemet, og derigennem ti1 igangsaettelsesmekanisrnen for selve sygdomsprocessen. Omsaetningen af energisubstrater i laesionerne er svaert pavirket, hvilket eksemplificeres i en gennemblodnings/glucose metabolisme ratio pa 3 ml/pmol sammenlignet med normalt 1,5 ml/pmol. Der er ingen tilsvarende forandringer i cerebralt pH. Disse sammenhaenge er tolket som primaer forstyrrelse i mitokondriefunktionen. Uafhzngigt af selve de histopatologiske lzsioner findes en 50% reduktion in blodgennemstramningen i hjernebarken. Denne observation er blevet tolket som en funktionsbetinget nedsaettelse i blodgennemstrornningen sekundiert ti1 de svSere lammelser hos dyrene.

Marianne Juhler

PATHOPHYSIOLOGY OF EAE

5

Acknowledgements U

The author is indebted to Drs. O.B. Paulson, N.H. Diemer and R.G. Blasberg for their support and inspiration, to all my co-authors for their help and interest in the individual studies, to Ms Bodil Kjax, Kirsten Kyhn and Helle Pedersen for patient and expert secretarial assistance and to Ms Annemette Elle and Mr. John T. Christensen for putting up with me in the laboratory. Special thanks is owed to Drs. Annette Nordenbo and Bente Pakkenberg for valuable comments and suggestions during the preparation of this review. The studies were supported by The Danish Medical Research Council, Danish Multiple Sclerosis Society, The Warwara Larsen Foundation, The Hede Nielsen Foundation and The Hestehandler Ole Jacobsen Foundation.

Ti1mor og far

6

M. JUHLER

The present review is based on the following papers: 1. Juhler M, Barry DI, Offner H, Konat G, Klinken L, Paulson OB. Blood brain and blood spinal cord barrier permeability during the course of experimental allergic encephalomyelitis. Brain Research 1984:302:347-355. 2. Juhler M, Blasberg RG, Fenstermacher JD, Patlak CS, Paulson OB. A spatial analysis of the blood-brain barrier damage in experimental allergic encephalomyelitis. J Cereb Blood Flow Metabol 19855.345-553. 3. Juhler M, Barry DI, Laursen H. The distribution of immunoglobulins and albumin in the central nervous system in acute experimental allergic encephalomyelitis. Acta Neurol Scand 1986363: 272-278. 5 . Juhler M. Simultaneous determination of regional cerebral blood flow, glucose metabolism and

pH in actue experimental allergic encephalomyelitis. J Cereb Blood Flow Metabol 1987:7378-584. 6. Juhler M, Diemer NH. A method of I4C and 'H double label autoradiography. J Cereb Blood Flow Metabol 1987:7.372-577.

PATHOPHYSIOLOGY OF EAE

7

Introduction Experimental allergic encephalomyelitis (EAE) is traditionally the object of immunological and histopathological research, and apart from a great number of reports of blood-brain barrier (BBB) damage in EAE, there are only few studies on the pathophysiology of EAE. The papers constituting the basis of the present review are an attempt to introduce a physiological approach to EAE, which might add to the understanding ofthe still unclarified pathogenesis. For this purpose, three main subjects are discussed in the following chapters: i) the sensitivity of a method must be taken into account before it is applied to the study ofa given condition. Because ofits highly focal nature, EAE places special requirements on the regional resolution of a method (chapter 4); ii) the immunesystem is virtually excluded from the central nervous system (CNS) by the blood-brain barrier (immunological seclusion of the CNS - see ref. 21). As EAE results from systemic immunization, BBB pathophysiology is a central issue in the initial pathogenesis of EAE (chapter 2); iii) studies on cerebral blood flow and an blood flow-energy metabolism coupling have added greatly to the understanding of ischemic brain disease. It has been suggested that ischemia secondary to inflammatory edema could contribute to tissue damage in EAE. Therefore studies on cerebral blood flow and energy metabolism may provide new insight into the pathophysiology ofEAE (chapter 3). The author has selected to study acute EAE in a rat model because of the predictability of the time course of the condition and its suitability for studying both the initial pathogenesis and the special methodological requirements. For completeness, a short chapter on neuropathological and immunological studies in EAE is included in this review. EAE also exists in a chronic, relapsing form, which has been described as resembling multiple sclerosis clinically by many relapses and remissions and pathologically by plaquelike areas of extensive demyelination. Unless stated otherwise, EAE in the following will refer to acute EAE in guinea pig or rat.

Chapter I

Clinics, pathology and immunology of EAE EAE has been known since 1933, when it was shown, that repeated subcutaneous injections of central nervous system (CNS) tissues resulted in an acute immunological reaction against the CNS of the recipient (1). Later it was shown that a single injection of the CNS antigen sufficed if a non-specific immune activator (adjuvant) was injected simultaneously (2). Patterson et al. (3) have described the clinical evolution in detail, and the characteristic ascending flaccid paresis/paralysis which starts approximately 2 weeks after inoculation in the tail and may eventually progress as far cranially as toinvolve the front limbs, has since been reproduced by others (4,5,6,7). Histologically, EAE is characterized by many perivenular foci of mononuclear cellular inflammation primarily in spinal cord and brainstem (3,4,5,8,9). The infiltrating cells have been identified as mainly lymphocytes (10) but plasmacells (9) and macrophages (9,11,12)are also present. Another histopathological feature of EAE is demyelination. In acute EAE, demyelination is slight, evidenced by myelin debris in macrophages (9,11,12) and/or by a few demyelinating axons, which are never found outside the areas covered by cellular inflammation (1 3). Electron microscopically two types of demyelination differing by morphology may be found probably representing unrolling of myelin lamellae by infiltrating cells and lysis (vesicular disruption) by soluble factors, respectively (9,13). The histopathological description of EAE is closely related to the immunological studies, which together constitute the bulk of EAE research (14,15).The predominance of mononuclear cells in the exudate resembles the histopathology of cell mediated immune (CMI) reactions (delayed hypersensitivity) (15). The concept of EAE as a pure CMI-phenomenon is supported by Patterson’s observation that EAE may be passively transferred by lymphnode

8

M. JUHLER

cells from sensitized donors (16), and by Condie et al.’s observation that removal oflymphnodes draining the injection site prevented inoculated animals from developing EAE (17). Cellular immunity is principally a function of T-lymphocytes which are activated upon contact with the appropriate antigen. Specifically the helper - inducer T-cell subset has been implicated in EAE (10,18,19,20). T-cell activation requires the antigen to be presented in conjunction with the Major Histocompatibility Complex (MHC) by antigen presenting cells (APC). (21). This requirement is met in EAE, where monocytes, astrocytes and endothelial cells have all been suggested as APC (22,23,24,25). It is particularly interesting that the latter cell-type could function as APC, as an immunological reaction taking place on/within the endothelial cell might represent a possible breach in the immunological seclusion of the CNS (21; see also chapter 2). Although the quoted studies leave little doubt of the great importance of T-cells in the pathogenesis of EAE, they do not exclude that a humoral mechanism may be equally important (26). Humoral immunity is a result of systemically circulating antibodies secreted by B-lymphocytes in response to antigenic challenge. Antibodies may directly combine with and neutralize the antigen or may require complement to deal with the antigen. Deposits of IgG in the parenchyma visualized by immunohistochemical methods always accompany the invasion of the CNS by T-cells (10) or may even precede cellular invasion (11). EAE can be passively transferred by cell free lymph (27) which indicates that immunoglobulins or other humoral factors may initiate the pathogenesis. Immunoglobulin deficient (IgG and IgM) rats developed neither histological nor clinical signs of EAE (28). Specific suppression of B-cell function leaving the T-cell function intact prevents rats from developing clinical EAE, and impairs demyelination, but does not diminish the degree of cellular infiltration compared the infiltration in to rats with intact B-cell function (29). The quoted evidence in favour of a pathogenetic humoral factor does not contradict or exclude the existence of cell mediated mechanisms. Although cellular and humoral immunity are classically described as entirely different entities, they share

common features and often interact (21). It is well known that T-cell subsets have regulatory influence on B-cell activity and that humorally mediated immunoresponses may require the ”co-operation” of both T-cells and macrophages (21,30). The observation that depletion of helper T-lymphocytes produces a milder form of EAE (18) indicates that such a cooperation may be important in EAE. Studies on the mechanism of demyelination (mostly performed in chronic relapsing EAE) show that both cellular infiltration and humoral factors - possibly a complement-fixing antibody - are necessary for primary demyelination to take place (31,32,33,34,35). The observations that 1) macrophage depletion impedes both clinical and histological EAE (36) and 2) electron microscopic evidence of macrophages splitting myelin lamellae (9,13) seem to place this cell as the actual effector of demyelination. The second issue to address when dealing with an immune mediated process, is the definition of the antigen. Several antigens may be responsible for triggering the events. Kies (37) found that myelin basic protein (MBP) had encephalitogenic properties, and it is still regarded a major encephalitogenic factor in EAE (31). The encephalitogenicity is dependent on certain regions (epitopes) ofthe MBP molecule as well as on the strain of experimental animal (38,39). In animals sensitized to MBP instead of crude CNS homogenate, only slight demyelination is seen corresponding to the perivascular cellular infiltrates (31,40) whereas sensibilisation to galactocerebroside (31) or proteolipid (41, Hartley guineapigs) greatly enhances demyelination. This shows that not only MBP, but probably several antigens are responsible for the expression of all aspects of EAE (31,41, 42). The observation has been explained by the asymmetric localization ofantigens in myelin, where cerebrosides which are located at the surface are readily accessible for attack, whereas MBP is not, as it is deeply embedded in the lipid bilayers (43). In summary, the immune mechanisms behind EAE involves a complicated interaction between elements of both cellular and humoral immunity. Full blown histological and clinical EAE evolves in response to several CNS antigens of both protein and lipid nature.

PATHOPHYSIOLOGY OF EAE

Chapter II Blood-brain barrier permeability in EAE

9

EAE lesions tend to cluster at the spinal and cranial nerve root entry zones (8,12,40) and at the floor of the IVth ventricle (8). By analyzing this microtopography of lesion distribution, Juhler et al. (8) found a striking congruence between the predilection sites ofthe lesions and the areas, where the BBB is lacking in the normal CNS (44,45,46,53), i.e. spinal/trigeminal root entry zones, area postrema in the floor of the IVth ventricle and attachment zones of there choroid plexus. (Figs. 1,2). In these areas, the exchange between plasma solutes (including antibodies and other humoral substances) and CNS extracellular fluid (including myelin degradation products from the normal turnover) occurs more freely than in other areas of the CNS, and it is conceivablethat the initiating pathogenetic events of EAE might take place in these permeable areas (8,ll). In support of this hypothesis, it has been found that non specific damage to the BBB (i.e. by ischemic/anoxic, chemical or thermal lesions) and subsequent passive induction of EAE by transfer of donor lymphocytes will cause typical

The vasculature of the normal CNS is lined by endothelial cellsjoined by so-called tight junctions whereby a barrier is formed which greatly limits the exchange between water soluble plasma substances and formed elements on one side and CNS extracellular fluid on the other (44,45,46). This phenomenon is termed the blood-brain barrier (BBB). The existence of a BBB means that antibodies, other humoral substances and immunoreactive cells do not readily gain access to the CNS. This phenomenon is referred to as the immunological seclusion of the CNS (21). The question how these elements reach and recognize their antigen on the other side of the BBB to initiate the pathogenesis of EAE, can not be answered from the studies referred to in the previous chapter. To address this question, studies must be designed to describe any breach in the BBB, whereby contact between the circulating immune system and CNS can be established. NUMBER OF LESIONS t S E M There is a turnover of myelin proteins in normal I animals (47,48) and it has been shown that there is slow transport ofproteins from the CNS into the circulation (49). Immunofluorescence studies have shown that MBP accumulates perivascularly very early during the evolution of the EAE (50). On this basis Wiesniewsky (40,51)has put forward the hypothesis that the endothelial cells present the antigen to the circulating immunoreactive cells making them adhere to and subsequently penetrate the endothelium. If this were the full explanation, EAE lesions (perivascular cellular infiltrates and demyelination) would be expected to be scattered throughout the CNS being especially numerous where capillary density is greatest, for instance in the cerebral cortex. However, the distribution of lesions in the CNS is quite different with predilection of lesions in Fig. 7. The number of EAE lesions (lymphocytic accumuthe spinal cord and brain stem (3,4,5,8,12,52). lations) per tissue section at different levels of the neuraxis. Lesions are only rarely seen in the forebrain (8) and A: spinal cord; B: brainstem (pons); C: brainstem (entry when they do occur, they are found in periventrizone of Vth nerve); D: brainstem (cerebellar peduncle); E: cular areas (3,5,8,40)and choroid plexus (3). It has forebrain. (Reproduced with kind permission of Raven been found that within the mentioned regions, Press; J Cereb Blood Flow Metab 1985:5: 545-553).

+

10

M. JUHLER

Fig. 2. Schematic representation of “leaky” areas of the CNS (left);predilection sites of EAE lesions (middle); and the topographical correlation between leaky areas and lesion distribution (right).

EAE lesions to occur in the ”leaky”areas and not in other places (54). Electrophysiologicalstudies on rabbits with EAE have pinpointed conduction abnormalities of demyelinating sensory nerve fibers to the dorsal root ganglia which further emphasizes the role of these permeable areas in the pathogenesis of EAE (55). However, BBB permeability in EAE is not restricted to the extent seen in normal animals. Leaky vessels occur in the parenchyma at a distance from the naturally leaky areas radiating from these towards the centre of the tissue (11). This suggests that some permeability increasing factor is created in the permeable areas and diffuses along the vessels into the tissues. The nature of such a factor has not been established, but it could be a lymphokine or the result of antigen-antibody interaction (15,27,28, 35) which is known to be able to cause vascular permeability and to attract leukocytes (21). In search of an answer to assess the relative importance of cellular versus humoral immunity in the pathogenesis of EAE, much effort has been put into describing the time course of the BBB changes in relation to the occurrence of cellular infiltrates (4,5,11,52,56,57,58,59). It seems natural to distinguish between qualitative and quantitative BBB studies because of differences in spatial resolution and because of the somewhat discrepant results

when comparing one type of method to the other (8). Qualitative methods demonstrating extravasation of dye markers bound to albumin were the first to be used (56,60). More sophisticated immunohistochemical studies visualize extravasation of naturally occurring plasma components, e.g. albumin (11,59), fibrinogen (58), IgG (11,50,58). (Fig. 3). Extravasation of radioactively labelled tracers (thorotrast (13), 1251bovine albumin (61,62), 1251-gammaglobulin(57)) can be visualized autoradiographically by dipping thin tissue sections in photographic emulsion (63). These methods generally demonstrate that increased permeability of capillaries and venules occurs several days before the cellular infiltrates (11,58,61). Most quantitative studies on BBB permeability in EAE have used a tissue uptake technique (4,5,52,62),i.e. determination of the amount of circulating radioactive tracer taken up by various, macroscopically dissectable CNS regions (64). Such studies show a much shorter (1-2 days) time interval between significantly increased BBB permeability and the cellular inflammation (4,52) and some studies fail to detect a time lag at a11(5,62). Juhler et al. (8) have analyzed the discrepancy between the results obtained by the two types of methods and find that BBB damage is indeed an

PATHOPHYSIOLOGY OF EAE

11

Fig. 3. Immunohistochemical visualization of IgC distribution in spinal cord of rat with EAE. 18 days after immunization. A. and B: diffuse staining with some focal (perivascular) accumulation. C and D: perivascular lymphocytic infiltrations o n grey/white matter border and in white matter, respectively. Earlier stages of EAE show only focal perivascular albumin or IgG reaction products without lymphocytic lesions (not shown in figure). (Reproduced with kind permission from Acta Neurol Scand 1986:73:119-124).

12

M. JUHLER

early phenomenon in EAE, but the sensitivity (i.e. spatial resolution) of tissue uptake methods is not sufficient to detect early BBB damage in EAE (4,8). This is further discussed in the chapter dealing with methodological considerations. Another question to address is the nature of BBB damage in EAE. Substances may permeate the endothelium either by being transported in microvesicles (65) or by passing through channels/slits in the cells (66).The first mechanism allows molecules to pass the BBB irrespectable of their molecular size, whereas the latter exerts a sieve-like effect

letting small molecules through in larger quantities than large ones (66). Electron microscopic studies using the tracer horse radish peroxidase indicate that the vesicular transport as well as trans- or interendothelial passage may take place (67,68). By simultaneously administering four tracers of different molecular sizes (Na+, C1-, sucrose and inulin) in a study using the tissue uptake technique, Juhler et al. (4)found an inverse correlation between molecular size and BBB permeability in EAE (Fig 4) indicating that slits or pores in the endothelium are responsible for

cm3q" s-') Regional PS values ( on each of the experimental days

c1-1

Na*:o

'O1

0

Regional PS values ( 10-5cm3g - ' s-'1 on e a c h of the e x p e r i m e n t a l days

Occipital cartex

Sucrose z e

'O1 01..

I

I

I

I

1

1

'O1

1

1

1

Inulin = 0 Occipital cortex

5-1

- 7

Cerebellum r

'"1 old4

I

I

I

Cerebellum

I

I

n I

,

,

I 1

,

Hindbrain

l01

0'4:

I

I

i

I

I

401

I

I

I

I

I

Spinal cord segment 3

2ol 10

Spinal cord segment 3 T

B 1

401 -I

Spinal cord segment 7

20

EZYS

I

8

10

12

201

Spinal cord segment I

1

14

16

21

Days

Days

Fig. 4. Regional BBB permeability to 24Na+,36Cl-(small tracers), 3H-sucrose (medium sized tracer) and 14C-inulin (large tracer) on each of the experimental days. The increase in permeability is greater and occurs earlier with the small tracer. (Reproduced with kind permission from Elsevier Pub1 Co, Brain Research 1984:302:347-355).

PATHOPHYSIOLOGY OF EAE

the permeability increase. A time lag of one day between the occurrence ofalbumin (MW 65.000) and IgG (Mw 150.000)found in an immunohistochemical study (11)supports this concept. However, other factors, for instance electrical charge or stereochemical configuration of the permeating molecule, could also influence the permeability (69). That such mechanisms may also be operative in the BBB dysfunction in EAE has been suggested on the basis of the mentioned difference in permeability between albumin (positively charged) and IgG (electrically neutral) (11). Summarizing this chapter, it may be concluded, 1) that naturally permeable regions of the CNS seem to play an essential role in the initiating pathogenesis of EAE, 2) that BBB permeability is an early event, preceding the occurrence of cellular infiltrates by several days and, 3) that the BBB dysfunction in EAE seems to be a selective phenomenon.

Chapter 111

Metabolic changes in EAE This chapter will focus on energy metabolism and related physiological variables in EAE. Glucose and oxygen are the only energy sources ofquantitative significancein the CNS, and normally cerebral blood flow (CBF) is closely coupled to cerebral metabolic rate (CMR) of glucose and oxygen (70,71).It is therefore natural to consider studies of CMR and CBF together. The rationale for studying energy metabolism and CBF/CMR coupling in EAE is twofold: first, it is conceivable that a general metabolic derangement may exist in the CNS if it is being attacked by noxious substances and ifthere is a general change in its biochemical composition. Second, methods are presently available for studying both CBF and CMR in humans, and EAE might extend our knowledge of the tissue damage in multiple sclerosis, if experimental and clinical studies of that type could be compared. At present, two entirely different hypothesis on cerebral glucose metabolism in EAE are in existence. Both have emerged from observations of in-

13

creased lactate production in the CNS, but the use of different approaches to provide a description of accompanying physiological changes is probably the reason, why two so basically discrepant theories have been proposed. Smith (72) incubated tissue slices from the CNS of EAE animals with radioactive glucose and found that glucose consumption was unaltered in the presence of a decreased CO, production and an increased lactate production. The described in vitro system is independent of tissue perfusion, and the author arrived at the conclusion that the metabolic steps were impaired from the breakdown of lactate and onwards, i.e. the oxidative metabolism involving the tricarboxylic acid (Krebs) cycle and the respiratory chain enzymes, which are both located in the mitochondria. The alternative theory put forward by Simmons et al. (7) states that lactic acid production is caused by anaerobic glycolysis due to ischemic hypoxia in the inflamed, edematous tissue, where flow might at least focally - be mechanically obstructed. Indirect support for focal anaerobic glycolysis at the lesion sites might be inferred from an - equally qualitative - study, where uptake ofthe CMR tracer 2-deoxy glucose (2-DG) was found to be higher corresponding to the lesions than in the surrounding, non lesion tissue (73). This could indicate that anaerobic glycolysis was taking place, as the unfavorable energy yield of anaerobic glucose metabolism would result in a need for breakdown of more glucose molecules to keep up production of energy rich compounds. By using 14C-antipyrine as a tracer for CBF in a tissue uptake study, Simmons et al. (7)were unable to support their theory of hypoperfusion. According to Juhler & Paulson (74) the 'experimental conditions chosen in the mentioned study may not have been optimal for detecting focal flow changes in small areas (please refer to chapter IV on methodology for details). By optimizing the experimental design for capturing such changes, it was demonstrated that CBFis focally increased by approximately 100% corresponding to the histological lesions (74) (Table 1). This observation would seem to contradict a theory of ischemic hypoxia and resulting anaerobic glycolysis, unless the reported focal accumulations of 2-DG in the lesions could be shown to represent an increase in CMR of more than 100%. This question was

14

M. JUHLER

Table 1 Values of rCBF (ml g-' min-') f SD in selected CNS regions. rCBF in lesion-free areas in the EAE animals is significantly different from control values in parietal cortex (~(0.05)and brainstem nuclei (~(0.025).rCBF in EAE lesions (lymphocytic accumulations is approximately three times the normal value (pCO.00 1). (Reproduced with kind permission from Elsevier Pub1 Co, Brain Research 1986:363:272-278. Region

Control

EAE animals Lesion-free areas

Parietal cortex Corpus callosum Ventral medial nucleus of thalamus Hippocampus, gyrus dentatus Hippocampus, Ammon's horn Internal capsule Hypothalamus Cerebellum, cortex Cerebellum, peduncules Brainstem, nuclei Brainstem, tracts

2.68k0.92 0.53k0.06 1.40k0.22 0.91k0.07 0.84k0.01 0.45k0.04 0.79k0.02 0.84k0.05 0.37k0.04 1.25+0.30 0.41k0.06

1.46k0.44 0.52k0.11 1.54k0.36 1.20k0.24 1.28k0.28 0.48f0.12 1.25k0.40 0.9950.24 0.46k0.16 I .51*0.29 0.46k0.16

approached by quantitative measurements of CBF and CMR by a method which allows measurement of both parameters simultaneously in the same region. Because of the possibility of focal lactacidosis at the sites of the lesions, measurement of regional pH together with CBF and CMR would be preferable. Juhler (75) designed a triple label quantitative autoradiographic study where all three parameters could be measured in the same region in the same animal and correlated to the exact regional occurrence of lesions. This study reproduced both the previously mentioned flow increases (74) and the focally increased 2-DG uptake (73) in the lesions; but CBF was almost double that

EAE animals Lesions 0.52k0.05

0.63f0.19

1.19k0.46 1.35kO.10

of normal compared to a modest approximately 15% increase in CMR. This finding may also be expressed as a rise in the CBFICMR ratio from the normal value of about 1.5 rnl/prnol to a value of almost 3 ml/pmol in the lesions (Table 2). On the basis of this, it was concluded, that ischemic tissue hypoxia is not responsible for the changes in energy metabolism, but that some primary disturbance in metabolism (which might exist at the mitochondrial level) probably results in a secondary increase in CBF. As no concomitant changes in regional pH were found in the study, accumulation of vasodilating metabolites other than H+ could cause the increase in blood flow, e.g. K+ or adenosine (76).

Table 2 Mean values of CBF (ml100 g-' min-I), CMK ( p o l 100 g-' min-I), CBF/CMK index (ml/prnol), and p H c! SEM. Measurements in cerebellar cortex are shown to allow comparison with values obtained by other investigators. Measurements on EAE lesions were performed in brainstem nuclei at the floor of the IVth ventricle and are compared to values from the same areas in control animals (t-statistics; *:p(0.002;**:p(0.05). Reproduced with kind permission of Raven Press, J Cereb Blood Flow Metabol 1987:7:578-584. CBF (m1/100e/min\

CMR (umo1/100e/min\

CBF/CMR (ml/urnoll

PH

Cerebellar cortex

76k2

50k4

1.5k0.2

6.83f0.14

Brainstem nuclei

81k3

45k4

1.8kO.2

6.74k0.05

EAE lesions

145fl*

51k3**

2.8f0.2,'

6.71k0.06

PATHOPHYSIOLOGY OF EAE

An alternative hypothesis proposed by the author was that the CBF increase could reflect inflammatory hyperemia and the focal increase in CMR could be due metabolism by the lymphocytes (75). The uniformity in regional tissue pH in EAE is in accordance with the observation that the lactate level in the CNS returns to normal at fulminant clinical disease (7). This might be due to the increased CBF which could wash out lactate or might even stimulate any functional reserve in the mitochondria by increased oxygen delivery (75). The interpretation that oxidative metabolism in the CNS in EAE may be profoundly disturbed, is supported by observations that the level of enzymes participating in the tricarboxylic acid (Krebs) cycle and in the respiratory chain are decreased in EAE brains resulting in uncoupling of oxidative phosphorylation (77). These enzymes are all located in the mitochondria, and a decreased mitochondria1 protein content (77) together with an abnormal morphology of mitochondria (78) in the CNS in EAE is further evidence in favour of a basic metabolic derangement. Pathophysiological abnormalities are also present without direct correlation to lesions. Juhler & Paulson (74) found a decrease in cortical CBF in EAE to 50% that of normal values (Table 1). In agreement with the general pattern of lesion distribution, the animals used for the study did not exhibit a single cortical lesiofi. The authors interpreted this change in CBF as a functional decrease due to the severe sensory motor impairment in the paralytic animals. That this could well be the case, is illustrated by electrophysiologicalstudies, which have demonstrated the presence of decreased conduction of impulses in the spinal root ganglia, which would mean impaired afferent conduction of sensory impulses (55). In summary, disturbed metabolism of energy substrate is present in EAE. Most likely it is due to mitochondrial dysfunction resulting in a basic disturbance in glucose metabolism illustrated by uncoupling of the close CBF/CMR correlation which normally exists. Changes in CNS blood flow in EAE - both the focal increases in the lesions and the decrease in regions remote from any lesions - must be seen in context with other physiological changes such as altered glucose metabolism or impaired functional capacity in the paralyzed animals.

15

Chapter IV

Methodological considerations This chapter is divided into two main sections. The first focusses on general methodological aspects of studies on physiology/pathophysiology.The second is dedicated to methodological considerations specifically relating to studies on pathophysiology of EAE and similar conditions. The general principles which emerge from this last section of the chapter may also be applicable to clinical studies in patients with multifocal diseases ofthe CNS such as multiple sclerosis.

General considerations A widely used principle for the study of physiological variables is based on the tracer concept (79). A tracer is a molecule which is introduced into the organism being studied, where 1) it can be detected ("traced"), because it is recognizable e.g. by radioactive or other labelling, and where 2) it is distributed, metabolized/accumulated and excreted like the mother substance, which the experimentator has selected to study. The latter of the two tracer principles is referred to as compartmentation. Compartments can be illustrated as a number of small rooms or steps, which the tracer must pass on its way from being introduced into the organism to being located somewhere in the organism. The compartments may be connected in series, in parallel or interconnected in more complicated ways (80). The exchange between the compartments is described by kinetic constants, and this description is in many ways similar to the characterization of enzymatic processes by kinetic constants (81$2). The simplest compartmentation is the two compartment model. In CNS physiology, it may be exemplified by blood-brain barrier (BBB) studies where the blood/plasma is one and the CNS the other compartment, and where the exchange is described by an influx constant (blood to brain) (K1)and an efflux constant (k2)in the reverse direction (83,84).When dealing with compounds which are metabolized by brain cells, more complicated models with three or more compartments are required, which additionally incorporate the kinetic constants for metabolism. In brain physiology, the

16

M. JUHLER

three compartment situation can be illustrated by studies of glucose metabolism, where the kinetic constant for glucose phosphorylation (K3) is taken into account together with the BBB permeation constants (81,82,85,86).A prerequisite fortheuseof tracer methodology is that the tracer does not disturb the physiological variable being studied for instance by blocking or disturbing competition for enzymatic processes. When quantitating physiological processes - in the CNS as anywhere else - the appropriate kinetic constants and the blood- or plasma concentration of the tracer must be measured or known in advance (79,81,82,83,85,86).Qualitative methods can be used to asses changes in distribution of a given physiological variable, but they ordinarily allow neither comparison between individual animals or groups of animals nor conclusions on increase or decrease from normal values (87). The validity of this statement may be seen from the following arguments: 1) a diffuse decreasehncrease would never be detected because it does not alter the distribution pattern of the tracer, 2) focal accumulations of tracer could be wrongly interpreted, because they may represent either focal increases on a background of physiologically normal tissue or they may represent normal "islands" of tissue on an abnormal background. These general considerations have all been taken into account by the author when designing and performing the reported studies on pathophysiological changes in EAE.

Physiological studies in EAE Until recently, radiotracer studies on regional physiology were based on the tissue uptake principles; i.e. the amount of tracer taken up by macroscopically dissectable regions, transformed into a meaningful physiological expression through the knowledge of tracer concentration in plasma and various constants. In chapter 2, it was briefly mentioned that the sensitivity of this type of method is limited. This will now be discussed in detail on the basis of BBB studies in EAE. In a tissue uptake study on the time course of BBB alterations during the evolution of EAE (4), it was found that BBB permeability (permeabilitysurface area, PS product) to small tracers increases

significantly from day 12 (Fig 4). However, already from day 10, the standard deviation of the calculated PS product increased manyfold, which indicates that the system was becoming instable. The authors interpreted this finding as the occurrence of a few small areas with increased permeability against a background of normal permeability. Such focal changes would not suffice to increase the average permeability in a block of tissue with otherwise normal permeability, but would make the tissue inhomogeneous, and thus create a larger standard deviation. By analyzing the discrepant results when comparing qualitative to quantitative studies on BBB permeability, the authors found that the difference in regional resolution could explain why qualitative methods with a microscopic regional resolution e.g. immunohistochemistry, qualitative autoradiography (11,50,58) demonstrate BBB leakage to be present several days before changes could be found with quantitative (i.e. tissue uptake) methods. In the latter, the regional resolution is limited to macroscopically dissectable regions (4,5,52), whereas the first deal with microscopically defined regions. Although no comment is made on the different properties of the methods used, this argument is clearly illustrated in papers by Leibowitz (61,62), where both types of methods were used, and where obvious extravasation of radioiodinated albumin was autoradiographically visualized some days before cellular infiltration, whereas a tissue uptake technique failed to reproduce any time interval between the occurrence of cellular infiltration and increased BBB permeability. O n the basis of the methodological analysis and their own results, Juhler et al. (4) concluded that methods combining high regional resolution and quantitation must be used for the appropriate study of the small, focal lesions which are characteristic of EAE. Another disadvantage of tissue uptake methods, is that histological analysis must be performed on other tissue samples than those used for physiological studies, whereby a direct correlation of regional histopathology and regional pathophysiology is precluded. Quantitative autoradiography (QAR) is an autoradiographic method allowing quantitation of tracer content and thus quantitation of physiological

PATHOPHYSIOLOGY OF EAE

variables in small areas of thin tissue sections (88). QAR has a theoretical lower limit in spatial resolution of approximately 50 um (89). The method also has the advantage of direct correlation between regional histopathology and regional pathophysiology, if the same or a neighboring tissue section is stained for histological examination after film exposure (8,74,75,88). QAR should thus be ideal for the study of small focal abnormalities such as EAE lesions. This is clearly illustrated in a study on regional BBB permeability in EAE (8) where it was found that the BBB damage in EAE is highly focal, corresponding to the histologicallesions (Fig. 5). By measuring the average permeability of entire sections, the study evaluated the sensitivity of conventional tissue sampling methods as applied to small, widely separated lesions. As expected, the difference between permeability in the actual lesion and the permeability of an entire section was highly significant. The authors concluded that macroscopic tissue sampling techniques will greatly underestimate the magnitude of localized changes or may even overlook them, ifthey are small and infrequent. Although QAR is well suited for the study of EAE and other multifocal conditions, it does have some drawbacks. When dealing with small areas of interest, diffusion away from the focal lesion into the surrounding parenchyma may become a major problem (8,74) (Table 3). This may result in an under-

Table 3 Average area and calculated radius of the histologicallydefined lesions and of the areas with increased permeability. The area of histologicallydefined lesionswas determined by computerized analysis of optical density in histological preparations. This could be done as the lesions(lymphocytic accumulations) were considerably darker than the surrounding neural tissue (Fig. 5). Areas of permeability (K,) defined lesions were determined as areas with a K, value F3SDs above the K, value of areas with normal permeability distant from lesions in the same animal. (With kind permission from Raven Press, J Cereb Blood Flow Metabol 1985:5:545-553). Definition of lesion Measured average area (mm’) Calculated radius (mm)

Histologically defined

K defined

0.05f0.03 0.13

1.03f0.10 0.57

17

Fig. 5. Histological section and corresponding digitized autoradiogram; BBB tracer C-alpha amino isobyturic acid (AIB). The increase in BBB permeability corresponds to the distribution of lesions. The increase is maximal at the actual (histologically defined) lesions, gradually reaching normal values further away.

estimation of the magnitude of the pathophysiological change and an overestimation of the area where the change is present. This problem is particularly pertinent when dealing with highly diffusible tracers, which is the case when measuring blood flow (74). Thus, both the multifocally of the condition and the characteristics of the tracer in terms of diffusibility and in terms of experimental time required to give reliable measurements should be considered in the experimental design, when planning studies ofregional physiology in EAE. This is illustrated by a study on regional cerebral blood flow in EAE (74),

18

M. JUHLER

where tracer diffusion in the parenchyma was minimized a) by shortening the experimental period to the shortest time possible for obtaining reliable values (20 sec) and b) by rapid removal and freezing of the brain; and where contrast between the focal high-flow areas (lesions)and the surrounding, normally perfused non-lesion tissue was enhanced by employing an infusion schedule giving steadily increasing arterial tracer concentration. Because of the inhomogeneity of the tissue and variation of focal pathology from animal to animal in EAE and other multifocal abnormalities of the CNS, it may be problematic to perform inter animal/intergroup comparisons. Such comparisons are necessary if several physiological parameters are studied separately in different groups of animals. In these cases, it therefore seems particularly advantageous to use methods, which allow simultaneous determination of more than one parameter plus histology on the same tissue section e.g. by double (90,91,92,93) or triple (75, 94,95) label autoradiography. In summary, quantitative physiological studies require an appropriate compartment model and the knowledge of appropriate kinetic constants. Generally, qualitative studies are not recommendable as they do not yield absolute measurements and may preclude any evaluation of interanimal/ intergroup comparison, and may even lead to erroneous conclusions. When designing studies of regional pathophysiology in EAE and other experimental or clinical multifocal conditions, the spatial resolution of the method is critical. Methods of macroscopic resolution generally are not sensitive enough and will underestimate or even overlook focal changes, which "drownn in the majority of normal tissue. When designing studies by methods of high regional resolution such as QAR, the kinetics of the tracer/tracers used must be taken into consideration, and the experimental design should a) minimize diffusion, b) maximize contrast between the focally altered tissue and its normal surroundings and c) yield reliable, reproducible results. When dealing with multifocal conditions either in the experimental or in the clinical setting, where the tissue is inhomogeneous, techniques allowing simultaneous measurements of 2-3 physiologic variables in the same sample and a direct correla-

tion to regional pathology may be particularly feasible.

Future research Many issues are still unresolved in EAE. Although the pathogenetic importance of increased BBB permeability in EAE has been much debated, the author feels convinced that is has a crucial role in initiating the disease process in the tissue because it allows entry into the CNS of immunoreactive agents normally excluded. The interaction of cellular and humoral immune reactions and their relative importance in the initial stages of EAE may be different than in later stages of EAE, where extensive BBB damage and tissue destruction (demyelination) is seen. This aspect is currently being intensely pursued by many researchers. The metabolic changes in the CNS of EAE animals is new on the arena of EAE research. There can be little doubt that the metabolic derangement is an epiphenomenon. EAE is not primarily a mitochondrial disease - its etiology remains immunological. However, the disturbed metabolism can certainly add to CNS dysfunction and promote tissue damage or even tissue destruction. Time course studies on the CBF increase in the EAE lesions could clarify whether they result from lymphocytic secretion of inflammatory vasoactive mediators, or from factors related to the disturbed metabolism of the nervous tissue. Studies of the tissue content of energy rich phosphates - e.g. by magnetic resonance spectroscopy - might further consolidate the hypothesis of secondary mitochondria1 dysfunction in EAE. EAE is often referred to as an "animal model of multiple sclerosis" and many studies on EAE have been inspired by adesire to solve the mystery of MS. There are many similarities, but important immunological differences between EAE and MS indicate that the analogy is not complete (96). Nevertheless, EAE is still an important tool in MS research because it provides possibilities of studying the process of ongoing demyelination directly. Furthermore, it seems to show that immune reactions in the CNS may have profound effects on tissue metabolism, and thus probably on CNS function - an observation that may add to understanding of CNS physiology in MS.

PATHOPHYSIOLOGY OF EAE

References 1. Rivers TM, Sprunt DH, Berry GP. Observations on attempts to produce acute disseminated encephalomyelitis in monkeys. J Exp Med 1933 2. Kabat EA, WolffA, BezerAE. Rap disseminated encephalomyelitis in rhesus mankeys by injektion of brain tissue with adjuvants. Science 1946:104: 362-363. 3. Paterson PY, Drobish DG, Hansen MA, Jacobs AF. Induction of experimental allergic encephalomyelitis in Lewis rats. Int Arch All 1970:27:26-40. 4. Juhler M, Barry DI, Offner H, Konat G, Klinken L, Paulson OB. Blood-brain and blood-spinal cord barrier permeability during the course of experimental allergic encephalomyelitis in the rat. Brain Research 1984:302:347-355. 5. Oldendorf WH, Towner HF. Blood-brain barrier and DNA changes during the evolution of experimental allergic encephalomyelitis. J Neuropath Exp Neurol 1974:33:616-631. 6.Simmons RD, Bernard CCA, Kerlero de Rosbo N, Carnegie PR. Criteria for an adequate explanation of typical clinical signs of EAE in rodents. In EC Alvord, MW Kies, AJ Suckling (eds.). Experimental allergic encephalomyelitis - a useful model for multiple sclerosis. Progress in clinical and biological research, vol. 146, Alan R. Liss, New York 1984:23-29. 7. Simmons RD, Bernard CCA, Singer G, Carnegie PR. Experimental autoimmune encephalomyelitis. An anatomically based explanation of clinical progression in rodents. J Neuroimmunol 1982:3:307-318. 8. Juhler M, Blasberg RG, Fenstermacher JD, Patlak CS, Paulson OB. A spatial analysisofthe blood-brain barrier damage in experimental allergic encephalomyelitis. J Cerebral Blood Flow Metabol 1985:5:545-553. 9. Lampert PW. Pathological implications of immunological disease in the central nervous system. In B.H. Waksman (Ed.) Immunoneuropathology,Saunders, London, Philadelphia, Toronto 1982:347-369. 10. Traugott U, Shevach E, Chiba J, Stone SH, Raine CS. Acute experimental autoimmune encephalomyelitis: Tand B-cell distribution within the target organ. Cell Immunol 1982: 70:345-356. 11.Juhler M, Laursen H, Barry DI. The distribution of immunoglobulins and albumin in the central nervous system in acute experimental encephalomyelitis. Acta Neurol Scand 1985:1986:73:119-124. 12. Kristenson K, Wisniewsky HM, Bornstein MB. About demyelinating properties of humoral antibodies in experimental allergic encephalomyelitis. Acta neuropath 1976:36:307-314. 13. Lampert PW, Carpenters. Electron microscopicstudies on the vascular permeability and the mechanism of demyelination in experimental allergic encephalomyelitis. J Neuropath Exp Neurol 1965:24:11-24. 14. Alvord EC, Kies MW, Suckling AJ. Experimental allergic encephalomyelitis. A useful model for multiple sclerosis. Progress in clinical and biological research, vol146, Alan R. Liss, Inc., New York 1984.

19

15. WaksmanBH(ed.). Immunoneuropathology, Clinicsin immunology and allergy, vol. 24, W.B. Saunders, London, Philadelphia, Toronto, 1982. 16. Paterson PY. Transfer of allergic encephalomyelitis in rats by means oflymph node cells.J Exp Med 1960:111:119-135. 17. Condie RM, Good RA. Experimental allergic encephalomyelitis. Its production, prevention and pathology as studied by light and electron microscopy. I: Kerey SR, ed. The biology of myelin. New York:Hoeber-Harper 1959:pp 321-332. 18. Sriram S, Roberts CA. Treatment ofestablished chronic relapsing ecperimental allergic encephalomyelitis with anti-L,T, antibodies. J Immunol 1986:136:4464-4469. 19. Gonatas NK, Howard JC. Inhibition of experimental allergic encephalomyelitis in rats severely depleted of T-cells. Science, 1974186:839. 20. Swanborg RH. Autoimmune effector cells. V.A. monoclonal antibody specific for rat helper T-lymphocytes inhibits adoptive transfer of autoimmune encephalomyelitis. J lmmunol 1983:130:1503-1505. 21. Juhler M, Neuwelt EA. the blood-brain barrier and the immunesystem. In: Neuwelt EA (ed). The impact of the blood-brain barrier and its manipulation. In press, Plenum Press. 22. Matsumoto Y, Fujiwara M. In situ localisation og class I and I1 major histocompatibility antigens in the rat central nervous system during experimental allergic encephalomyelitis. An immunohistochemical study.J Neuroimmunol 1986: 12:265-277. 23. Vass K, Lassmann H, Wekerle H, Wisniewski HM. The distribution ofIa antigenin thelesionsofrat acuteexperimental allergic encephalomyelitis. Acta neuropathol (Berlin) 1986:70: 149-160. 24. Vandenbark AA, RausJCMY (eds). Immunoregulatory process in experimental allergic encephalomyelitis and multiple sclerosis. Elsevier Sci Pub1 Co, Amsterdam 1984. 25. Hickey WF, Osborn JP, Kirby WM. Expression of Ia molecules by astrocytes during acute experimental allergic encephalomyelitis in the Lewis rat. Cell Immunol 1985:91:528-535. 26. Werdelin 0,McCluskey RT. The nature and specificity of mononuclear cells in experimental autoimmune inflammation and the mechanism leading to their accumulation. J Exp Med 1971:133:1242-1249. 27. Willenborg DO. Transfer of lesions of allergic encephalomyelitis in sheep with cell-free lymph from animals sensitized with homologous spinal cord plus adjuvant. Scand J Immunol 1982:16:437-441. 28. Willenborg DO, Prowse SJ. Immunoglobulin-deficient rats fail to develop experimental allergic encephalomyelitis. J Neuroimmunol 1983:5:99-109. 29. Gausas J, Paterson PY, Day ED, dal Canto MC. Intact B-cell activity is essential for complete expression of experimental allergic encephalomyelitis in Lewis rats. Cell Immunol 1982:72:360-366. 30. Lemire JM, Weigle WO. Passive transfer ofexperimental allergic encephalomyelitis by myelin basic protein specific L3T,4 T-cell clones possessing several functions.

20

M. JUHLER

J Immunol 1986:137:3169-3174. 31. Brosnan CF, Traugott U, Raine CS. Analysis ofhumoral and cellular events and the role of lipid haptens during CNS demyelination. Acta Neuropathol. (Berl.), 1983:~~ppl. 9:59-70. 32. Liu WT, Vanguri P, Shin ML. Studies on demyelination in vitro: the requirement of membrane attack components of the complement system. J Immunol 1983:131:772-778. 33. Stoner GL, Brosnan CF, Wisniewsky HM, Bloom BR. Studies on demyelination by activated lymphocytes in the rabbit eye. J Immunol 1977:118:2094-2102. 34. Mehta PD, Lassmann H, Wisniewski HM. Immunologic studies of chronic relapsing EAE in guineapigs: similarities to multiple sclerosis. J Immunol 1981:127:334-338. 35. Grundke-Iqbal I, Lassmann H, Wisniewsky HM. Chronic relapsing experimental allergic encephalomyelitis. Immuno-histochemical studies. Arch Neurol 1980:37:65 1-656. 36. Brosnan CF, Bornstein MB, Bloom BR. The effects of macrophage depletion on the clinical and pathologic expression of experimental allergic encephalomyelitis. J Immunol 1981:126: 614-620. 37. Kies MW. Chemical studies on an encephalitogenic protein from guinea pig brain. Am. N.Y. Acad Sci 1965:122:161-169. 38. Vandenbark AA, Offner H, ReslefT, Fritz R, Chon CH, Cohen IR. Specificity ofT-lymphocyte lines for peptides of myelin basic protein. J Immunol 1985:135:229-233. 39. Hashim GA. Neural antigens and the development of antoimmunity. J Immunol 1985:135 (suppl):838-842. 40. Wisniewsky HM, Lassmann H , Brosnan CF, Mehta PD, Lidsky AA, Madrid RE. Multiple Sclerosis: Immunological and experimental aspects. In: Matthews W.B., Glaser, C.H. (eds). Recent advances in neurology 3, Churchill Livingstone, Edinburgh, London, Melbourne, New York 1982:pp 95-124. 41. Yoshimura T, Kumishita T , Sakai K, Endoh M, Namikawa T, Tabira T. Chronic experimental allergic encephalomyelitis in guineapigs induced by proteolipid protein. J Neurol Sci 1985:69:47-58. 42. YonezawaT. Circulating myelinotoxic factors in human and experimental demyelinative tissue. Acta Neuropath (Berl.), 1983: suppl9:47-58. 43. Maggio B, Cumar FA, Roth GA, Montferran CG, Fidelio GD. Neurochemical and model membrane studiesin demyelinating diseases. Acta Neuropathol., (Berl.) 1983: Suppl9:71-85. 44. Bradbury M. Theconcept ofablood-brain barrier, John Wiley. Chichester-New York-Brisbane-Toronto, 1979: pp. 84-137. 45. Brightman MW, ReeseTS.Junctionsbetween intimately opposed cell membranes in thevertebrate brain. J Cell Bio 1969:40:648-677. 46. Rapoport SI. Blood-brain barrier in physiology and medicine. New York, Raven Press 1976: pp 43-127. 47. Lajtha A. Protein metabolism of the nervous system. Neurobiology 1964:1-98.

48. Smith ME. The turnover of myelin proteins. Neurobiology, 1972:2:35-40. 49. Klatzo I, Miquel J, Ferris PJ, Prokop JD and Smith DE. Observations on the passage of flourescin labelled proteins (FLSP) from the cerebrospinal fluid. J Neuropath Exp Neurol 1964:23:18-35. 50. SimonJ, Anzil AP. Immunohistological evidence ofperivascular localization of basic protein in early development of experimental allergic encephalomyelitis. Acta Neuropath. (Berl.), 1974:27:33-42. 51. Wiesniewsky HM, Lassmann H. Etiology and pathogenesis of monophasic and relapsing inflammatory demyelination human and experimental. Acta Neuropath. (Berlin) 1983: suppl. IX:21-31. 52. Daniel PM, Lam DKC, Pratt OE. Changes in the effectiveness of the blood-brain and blood-spinal cord bamers in experimental allergic encephalomyelitis.J Neurol Sci 1981:51:2119. 53. Olsson Y. Topographical differences in the vascular permeability of the peripheral nervous system. Acta Neuropath 1968:10:26-33. 54. Levine S, Sowinsky R. Experimental allergic encephalomyelitis: simple method for producing the localized form. J Immunol Methods 1982:54:355-360. 55. Pender MP, SearsTA. The pathophysiology ofacuteexperimental allergic encephalomyelitis in the rabbit. Brain 1984: 107:699-726. 56. Barlow CF. A study of abnormal blood-brain barrier permeability in experimental allergic encephalomyelitis. J Neuropath Exp Neurol 1956:15:196-208. 57. Cutler RWP, Lorenzo AV, Barlow CF. Brain vascular permeability ot I125 gamma globulin and leucocytes in allergic encephalomyelitis. J Neuropath Exp Neurol 1967~26~558-71. 58. Oldstone MD, Dixon FJ. Immunohistochemical study ofallergic encephalomyelitis. Amer J Path 1968:52:251263. 59. Vulpe MD, Hawkins MA, Rodzilsky B. Permeability of cerebral blood vessels in experimental allergic encephalomyelitis studied by radioactive iodinated bovine albumine by radio iodinated bovine albumin. J Neurol 1960:10171-177. 60. Broman T. The Permeability of the Cerebrospinal Vessels in Normal and PathologicalConditions. Munksgaard, Copenhagen, 1949: pp.61-69. 61. Leibowitz S. Cerebral vascular permeability in experimental allergic encephalomyelitis. Neuropath Pol 1969:7:303-309. 62. Leibowitz S, Kennedy L. Cerebral vascular permeability and cellular infiltration in experimental allergic encephalomyelitis. Immunol 1972:22:859-869. 63. Rogers AW. Practical autoradiography. Amsherham Reviews 20, Buckinghamshire, 1979. 64. Ohno K, Pettigrew KD, Rapoport SI. Lower limits of cerebrovascular permeability to nonelectrolytes in the conscious rat. Amer J Physiol 1978:253:H299-H307. 65. Westergaard E. The blood-brain barrier to horseradish peroxidase under normal and experimental conditions. Acta Neuropath., (Berlin) 1977:39:181-188.

PATHOPHYSIOLOGY OF EAE

66. Pappenheimer JR. In: Crone C, Lassen NA (Eds.). Capillary Permeability: the Transfer of Molecules Between Capillary Blood and Tissue, Munksgaard, Copenhagen, 1970: pp 278-290. 67. Hirano A, Dembitzer HM, Becker NH, Levine S, Zimmermann HM. Fine structural alterations of the bloodbrain barrier in experimental allergic encephalomyelitis. J Neuropath Exp Neural 1970:29:432-440. 68. Lassmann H. Comparative neuropathology of chronic experimental allergic encephalomyelitis and multiple sclerosis. Schriftenreihe Neurologie, vol. 25. Berlin, Springer 1983. 69. Wisniewsky HM, Lossinsky AS, Moretz RC, Vorbrodt AW, Lassmann H, Carp RI. Increased blood-brain barrier permeability in scrapie infected mice. J Neuropath Exp Neural 1983:42:615-626. 70. Siesjo BK. Brain energy metabolism. New York, John Wiley and Sons, 1978. 71. Sokoloff L. Localization of functional activity in thecentral nervous system by measurement of glucose utilization with radioactive deoxyglucose.J Cereb Blood Flow Metabol 1981:1:7-36. 72. Smith ME. Glucose metabolism of central nervous tissues in rats with experimental allergic encephalomyelitis. Nature 1966209:1031-1032. 73. di Rocco RJ, Hashim GA. Increased glucose utilization associated with inflammatory brain lesions of experimental allergic encephalomyelitis. Neurosci Lett 1983:37:105-10. 74. Juhler M, Paulson OB. Regional cerebral blood flow in acute experimental allergic encephalomyelitis. Brain Research 1986:363:272-278. 75. Juhler M. Simultaneous determination of regional cerebral blood flow, glucose metabolism and pH in acute experimental allergic encephalomyelitis. J Cereb Blood Flow Metabol 1987:7:578-584. 76. Pulsinelli WA, Levy DE, Duffy TE. Regional cerebral blood flow and glucose metabolism following transient forebrain ischemia. Ann Neural 1982:11:499-509. 77. Saragea M, Clopotaru M, SicaM, Vladutiu A, NegruT, Rotaru N. Biochemical changes occumng in animals with EAE. Med Pharmacol Exp 1965:13:74-80. 78. Roizin L. Some basic principles of molecular pathology. J Neuropath Exp Neural 1964:23:209-252. 79. Lassen NA, Per1 W. Tracer Kinetic methods in medical physiology. Raven Press, New York 1979. 80. Patlak CS, Blasberg RG, Fenstermacher JD. Graphical evaluation of blood-to-brain transfer constants from multipletime uptake data.J Cereb Blood Flow Metabol 1983:3:119-135. 81. Gjedde A. Calculation of cerebral glucose phosphorylation from brain uptake of glucose analogs in viva: a reexamination. Brain Research Reviews, 19824237274. 82. Lassen NA, Gjedde A. Kinetic analysis of the uptake of glucose and some of its analogs in the brain using the singlecapillary model: comments on some points of controversy. In: Tracer kinetics and physiologic modelling.

21

Lecture notes in biomathematics. Val. 48 (Lambrecht, R.M., Resigno, A.,eds.) Berlin, Springer 1983: pp.384407. 83. Blasberg RG, Fenstermacher JD, Patlak CS. Transport ofalpha-aminoisobutyric acid acrossbrain capillaryand cellular membranes. J Cereb Blood Flow Metabol 1983:3:8-32. 84. Fenstermacher JD, Blasberg RG, Patlak CS. Methods for quantifying the transport of drugs across brain barrier systems. Pharmac Ther 1981:14:217-48. 85. Sokoloff L, Reivich M, Kennedy C, des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada 0, Shinohara M. The 14-C deoxyglucose method for measurement of local cerebral glucose utilization: Theory, procedure and normal values in the conscious and anesthetized albino rat. J Neurochem 1977:28:897-916. 86. Sokoloff L. The 14-C deoxyglucose method: four years later. In: Cerebral blood flow and metabolism. Acta Neurol Scand Suppl (Gotoh F, Nagai H, Tazaki Y, eds.) Copenhagen, Munksgaard 1979: pp 640-649. 87. Kelly PAT, McCulloch JA. Critical appraisal of semiquantitative analysisof2-deoxyglucoseautoradiograms. Brain Research, 1983:269:165-167. 88. Blasberg RG, Groothuis D, Molnar P. Application of quantitative autoradiographic measurements in experimental brain tumor models. Sem Neural 1981:1:203221. 89. Smith CB. Localization of activity-associated changes in metabolism of the central nervous system with the deoxyglucosemeth0d:prospects for cellular resolution. InBarkerJLand McKelveyJF(eds).Current methodsin cellular neurobiology. New York, John Wiley & Sons 1983: pp 270-317. 90. Juhler M, Diemer NH. A Method for I4Cand 'H double label autoradiography. J Cereb Blood Flow Metabol 1987:7:572-577. 91. Diemer NH, Rosencarn J. Determination of local cerebral blood flow and glucose metabolism or transferby means of a double autoradiographic method. J Cereb Blood Flow Metabol suppl 1981:1:S72-S73. 92. Lear JL, Jones SC, Greenberg JM, Fedora TJ, Reivich M. Use of 123-1and 14-Cin a double radionuclide autoradiographic technique for simultaneous measurement of LCBF and LCMR gl. Theory and method. Stroke 1981:12~589-597. 93. Mies G, Niebuhr I, Hossmann KA. Simultaneous measurement of blood flow and glucose metabolism by autoradiographic techniques. Stroke 1981:12581-588. 94. Mies G, Bodsch W, Paschen W, Hossmann KA. Triple tracer autoradiography of cerebral blood flow, glucose utilization andprotein synthesis in rat brain. J Cereb Blood Flow Metabol 1986:6:59-70. 95. Nedergaard M, Gjedde A, Diemer NH. Focal ischemia ofthe rat brain: autoradiographic determination ofcerebra1glucoseutilization, glucose content, and blood flow. J Cereb Blood Flow Metabol 1986:6:414-424. 96. Arnason, BGW. Neuroimmunology. New Engl J Med 1987:316:406-408.