Journal of Affective Disorders, 22 (1991) 159-164 0 1991 Elsevier Science Publishers B.V. 0165-0327/91/$03.50 ADONIS 016503279100103T

159

JAD 00816

Relationship between observed NMR changes and brain water content following electroconvulsive stimulation in the rat John A.O. Besson and Ian C. Reid * Department of Mental Health, University of Aberdeen, Aberdeen AB9 220,

U.K.

(Received 14 December 1990) (Revision received 27 February 1991) (Accepted 3 April 1991)

Summary

The administration of electroconvulsive stimuli to anaesthetised rats results in changes in the relaxation times and water content of grey and white matter, but not in mid brain, hind brain or cerebellum. White matter changes occur in both T, and T,, are biphasic in character and related to water content. Grey matter changes are confined to T, and water content only. It is suggested that these changes are related to altered compartmentalisation of water, manifesting in a different manner in the two tissues probably due to their different cellular and biochemical composition.

Key words: Electroconvulsive

stimulation; Nuclear magnetic resonance; Brain water content; (Rat)

Introduction

The effect of electroconvulsive therapy (ECT) of depressed patients has been reported to result in a transient increase in white matter proton T, relaxation time which returns to the pretreatment value (Mander et al., 1987). It has been suggested that this may be due to an increase in water

Address for correspondence: Dr John A.O. Besson, Department of Mental Health, University Medical Buildings, Foresterhill, Aberdeen AB9 2ZD, U.K. * Present address: Department Edinburgh, Edinburgh, U.K.

of Psychiatry,

University

of

content resulting from transient alterations in blood barrier permeability, which have been observed using other techniques (Bolwig, 1984). Changes in nuclear magnetic resonance (NMR) relaxation times in fact represent a composite of both water content and water structuring in the tissue under study (Mathur De Vre, 19841, and are obtained by exciting the protons of water molecules by a radio frequency pulse while the protons are aligned in a static magnetic field. There are two types of relaxation time measures: T, (spin lattice relaxation time) and T, (spin-spin relaxation time); and these reflect different aspects of the behaviour of protons relative to their environment. Regarding water content, Bell et al. (1987) have shown that in the case of brain tu-

160

mours treated with mannitol and dexamethasone, T, measured in vivo predicts tissue water; and similarly Besson et al. (1991) has shown that white matter T, and T, changes in the brains of patients with Alzheimer’s disease measured post mortem are related to water content measured by dehydration. However T,, T, and water content measures are not always so related. In the brains of rats fed an alcohol-‘enhanced’ diet for 6 months for example, T, increases occur in the absence of T, and water content changes (Besson et al., 1989). It follows then that the only way in which the relationship between T,, T, and water content can be accurately determined and interpreted is by studying these measures directly in animals. The aim of this study is to examine the effects of a series of electroconvulsive stimulations on rat brain water content, and to relate any changes found to T, and T, measures obtained in vitro. Method Subjects Sixty-six male adult Sprague-Dawley rats (average weight 300 g) were used. Subjects were caged in groups of four and fed a normal diet with water ad libitum. Protocol Rats were randomly allocated to one of 11 groups of six subjects, five groups subjected to repeated electroconvulsive seizures (ECS) under brief inhalational halothane anaesthesia, and five groups repeatedly anaesthetised with halothane alone. The remaining group was not subjected to repeated exposure to either anaesthesia or ECS. Following six exposures to either condition, delivered twice daily over 3 days, groups of animals were killed by decapitation either immediately after the final (sixth) treatment, or at intervals of 4, 16, 24 and 72 h post treatment and the brains immediately removed and prepared for NMR analysis. Electroconvulsive stimulation Single electric shocks were applied via bilateral silver ear clip electrodes, using a purposebuilt stimulator delivering 150 V for 1 s (50 Hz

AC sine wave). This stimulation resulted reliably in a grand ma1 convulsion, after which rats were permitted to recover before return to their home cage. Brain dissection The brains of all rats were rapidly removed. The meninges and large blood vessels were stripped off with fine forceps, the cerebellum removed at its peduncular connection with the brain stem and the grey matter sliced from the surface of the cerebral hemispheres. The hind brain and mid brain were excised using fine scissors. The remaining white matter was also retained for further analysis. NMR spectrometry Samples were placed in NMR tubes and T, and T, were determined on a purpose-built 2.5 MHz proton NMR relaxation spectrometer. T, was measured using an inversion recovery/Hahn spin echo sequence with 15 T values (180 “-90 o pulse intervals) over a range of three times T,. T, was measured using a Carr-Purcell-Meiboom-Gill spin echo sequence collecting 100 echoes and analysing from 40 of these. All measurements were carried out at 30 o C. After measurement of relaxation times, sample tubes were unplugged. The tubes were weighed and placed in a 60°C oven for 3 weeks until dried to constant weight, and by deduction the original water content was obtained and expressed as a percentage of the pellet mass. All relaxation times given in this study are calculated using a mono-exponential model for relaxation behaviour. This model has been verified for T, for both white and grey brain by calculation of the correlation coefficient of fit of the weighed linear regression line on an inversion recovery data set from collection of 30 values of different 7 (180 “-90 o pulse interval) over a range of three times T,. For white brain this yielded a T, value of 311 ms with a correlation coefficient of 0.9995 (chi-squared value 6.829 with 19 degrees of freedom) and for grey brain a T, of 351 ms with a correlation coefficient of 0.9999 (chi-squared value 0.972 with 19 degrees of freedom). T, showed a greater tendency towards multi-

161

exponentiality, as shown by correlation coefficients for white brain of 0.998 (T, = 75 ms, chisquared 128) and for grey brain of 0.994 CT2 = 81 ms, chi-squared 790). For the purposes of the study however, it was not deemed necessary to attempt separation of the T, relaxation components since in the great majority of imaging studies (to which this directly relates) T, is treated as a mono-exponential. Results The T,, T, and water content estimations in cerebellum, hind brain and mid brain following treatment with .halothane alone and halothane + ECS showed no significant differences from one another over the various time periods, or from those of subjects who received neither anaesthesia nor ECS (Tables l-3). In white matter samples (Table 4, Fig. 1) there were no significant changes in T,, T, or water content in response to halothane anaesthesia, but following halothane + ECS there was a biphasic response of T,, T, and water content with an initial reduction immediately following the seizure which returned to baseline over the following few hours, falling again at 16 and 24 h post treatment. In grey matter, no significant changes were noted following halothane anaesthesia alone, but a biphasic change similar to that seen in white

TABLE

1

Controls 0 4 16 24 72 0 4 16 24 72

Differences ney U-test.

WATER CONTENT (MEANSkSD) OF DURING TREATMENT WITH HALOAND HALOTHANE + ECS (ECS)

T, (ms)

T, (ms)

H,O

321*9 325k4 332i7 331+6 329+4 328k9 325i4 326k8 328+4 324+8 322k8

82 k2.3 81.3k 1.5 82.1 k 2.5 79.4 * 2.1 81.6k 1.6 78.6 f 1.4 80.8 * 0.9 81.1f2.2 80.9 f 1.5 80.8 f 2.4 78 +3.8

79.05 i 0.24 78.87 f 0.17 78.71 * 0.24 78.38 k 0.3 78.47 + 1.73 78.57 k 0.43 78.4 f0.31 79.19kO.26 78.47 + 0.42 78.77 f 0.31 78.67 + 0.19

between

groups

non-significant

2

T,, T, AND WATER CONTENT (MEANS + SD) OF HIND BRAIN DURING TREATMENT WITH HALOTHANE (HAL) AND HALOTHANE + ECS (ECS)

Controls HAL 0 HAL 4 HAL 16 HAL 24 HAL 72 ECS 0 ECS 4 ECS 16 ECS 24 ECS 72 Differences ney U-test.

T, (ms)

H,O

(% dry weight)

284+ 5 285 k 10 291+ 8 280 f 12 291+ 10 284k 11 287& 8 286k12 287k 4.8 285 f 10 284+ 12

78.6+ 1.7 77.1 f 2.8 77.4 f 2.6 76.8 + 2.0 76.5 f 1.4 76.5 + 1.9 76.7 +_1.7 75.7 * 3.0 76.5 k 1.1 75.2 + 2.3 74.0 f 4.9

73.69 73.96 73.37 73.25 73.94 73.63 73.11 74.15 74.21 74.29 74.39

k 0.2 k 0.44 f 0.47 + 0.41 + 0.64 k 0.38 io.33 * 0.41 k 0.99 i_ 0.6 f 0.66

between

groups

non-significant

on Mann-Whit-

T, un-

Discussion The absence of significant changes in relaxation times and the stability of measures in the cerebellum, hind brain and mid brain act as an internal quality control over the reproducibility of the measures by this technique. In grey and white

3

T,, T, AND WATER CONTENT (MEANS BRAIN DURING TREATMENT WITH (HAL) AND HALOTHANE + ECS (ECS)

& SD) OF MID HALOTHANE

T, (ms)

T2 (ms)

Hz0

(% dry weight)

Controls HAL 0 HAL 4 HAL 16 HAL 24 HAL 72 ECS 0 ECS 4 ECS 16 ECS 24 ECS 72

29Ok 13 292k 9 305 * 13 288 * 20 284 k 24 295k 9 294+17 291k 6 285 k 10 294k 16 285i 6

76.4+ 76.4+ 77.2* 75.0* 75.8+ 75.9+ 75.0* 76.6k 76.4 k 75.5 * 75.85

77.5 77.18 77.14 76.97 77.36 77.51 76.42 77.77 76.50 77.44 77.25

+0.3 + 0.72 f 0.36 k 0.44 k 0.82 * 0.66 i 0.5 1 * 0.43 k 0.25 * 0.5 k 0.35

Differences ney C/-test.

between

(% dry weight)

on Mann-Whit-

T, (ms)

matter was noted in T, and water content. values in this tissue, however, remained changed (see Table 5, Fig. 2).

TABLE

T,, T, AND CEREBELLUM THANE (HAL)

HAL HAL HAL HAL HAL ECS ECS ECS ECS ECS

TABLE

groups

1.2 1.7 1.6 2.1 2.1 0.6 7.4 6.2 10.6 16.6 6.6

non-significant

on Mann-Whit-

162 TABLE

4

T,, T, AND WATER CONTENT (MEANS+SD) OF WHITE MATTER DURING TREATMENT WITH HALOTHANE (HAL) AND HALOTHANE + ECS (ECS)

Controls HALO HAL 4 HAL 16 HAL 24 HAL 72 ECS 0 ECS 4 ECS 16 ECS 24 ECS 12

T, (ms)

T, (ms)

H ,O (% dry weight)

293+ 8 293+ 7 294& 9 289+ 7 299+ 7 295+ 3 272k 9* 292k 11 284 f 10 279fll* 283+ 8

72.5 + 2.6 74.7 + 1.7 74.4+ 1.5 73.1+ 1.4 73.5 f 1.8 72.4 f 1.3 69.7,4.1* 72.2 + 2.9 72.0 f 3.3 70.0+5.0 * 69.1 k 4.3

76.58 + 0.37 76.58 f 0.56 76.33 f 0.64 75.87+0.2 77.14+0.4 76.75 +0.39 74.78 f 1.06 * 76.36 + 0.74 75.24kO.95 75.65 + 0.68 75.71 k 0.89

* ECS values significantly different from corresponding values and controls at P < 0.05 on Mann-Whitney

T, Controls HAL 0 HAL 4 HAL 16 HAL 24 HAL 72 ECS 0 ECS 4 ECS 16 ECS 24 ECS 72

(ms)

324k8 326*7 333+5 331*5 331+4 326*7 310+5 ** 325f9 316+8 * 312*9** 326rf-6

H,O

75.9 + 1.2 78 +1.6 78 + 1.6 79 +1.7 77 * 1.5 76.2+ 1.8 75.4 + 2.4 76.2 k 1.7 75.7+ 1.3 76.3 + 3.8 73.3 f 2.1

79.99 f 0.21 80.11 kO.32 80.19kO.18 79.87 + 0.26 80.22 * 0.2 80.10 + 0.47 79.43 * 0.45 * * 80.81 f 1.0 79.41 f 0.35 * 79.58 & 0.42 * * 79.91 kO.13

ECS values significantly different from values and controls on Mann-Whitney ** P < 0.01.

;

260 I!

0 c

77 50 7625

,”

63.75L

only

! I 10

I 20

I 30

I 50

40

I 60

I 70

, 60

HAL U-test.

CONTENT (MEANS k SD) OF GREY TREATMENT WITH HALOTHANE + ECS (ECS)

T, (ms)

270 265

u)

5

T,, T, AND WATER MATTER DURING AND HALOTHANE

;

0

matter, water content did not change in response to halothane administration alone, while repeated electroconvulsive seizures resulted in an acute reduction in water content immediately after the last seizure, normalisation at 4 h, followed by a transient reduction between 16 and 24 h, with subsequent normalisation at 72 h. It would seem likely that these changes are specifically related to the seizure activity. In grey matter, the reductions in water content were associated with reductions in T, , but not T, TABLE

Halothane -----________

(% dry weight)

corresponding HAL U-test: * P < 0.05,

1 0

0

76.ooo-c

Ln

77500., 77 000

-

76.500

-

10

20

30

40

Hours

post

50

60

70

80

1

76 000

-

75 500

-

75 ooo74 500

-

74 000

-

73 500

treatment

Fig. 1. Changes in (al T, value, (b) T, value and (c) water content following treatment with halothane alone and halothane + ECS in white matter. *P < 0.05.

values, while in white matter reductions occurred in both measures. This probably reflects the differential response to ECS of tissues in which biochemical composition and cellular organisation vary. White matter elements can, for example, be divided into myelin and non-myelin components on this basis. The non-myelin component most resembles grey matter in terms of water, protein and lipid content, while the myelin component is rather different, containing half the water content, slightly more protein and about five times the lipid content of its non-myelin counterpart (Norton, 1975). These variations may

163

~~,othane

f

EC5

---------__________

72

,l

,” 71

0

10

10

20

30

20 tiours

40

30

50

40 post

50

60

60

70

70

( 80

80

treatment

Fig. 2. Changes in (a) T, value, (b) T, value and (c) water content following treatment with halothane alone and halothane + ECS in grey matter. *P < 0.05;* *P < 0.01.

explain the differences in T, and T2 values obtained following similar water content changes in different tissues, given that water structuring and compartmentalisation, determined by tissue composition, may differentially affect NMR indices in addition to total water content. Models have been devised to test the effects of changing water content and compartmentalisation on T, and T, measures. These suggest that at a critical point in the structuring of water as determined by the pore size of gel compartments, T, and T, measures may change rapidly at a

constant water content (Murase and Watanabe, 1989). While this simple structural model cannot be directly compared with the sophistication of the cell membrane, nevertheless it illustrates the complexity of the relationships between water content, compartmentalisation and relaxation times. We would propose that the differential effects on relaxation times in the two tissues following ECS may be additionally related to altered compartmentalisation of water, and in this respect white and grey brain behave differently. The T, change observed in white matter may reflect altered water compartmentalisation, given that water content changes are not invariably associated with T, changes (as seen in the grey matter samples here). We speculate that changing compartmentalisation of tissue water may reflect, directly or indirectly, changes in the ordering of the cell membrane components, temporary in nature, and related to the ictal process. These changes may be linked to a number of processes that are believed to be modulated by cell membranes and relevant to their functions (Cullis and De Kruijff, 1980). Bolwig et al. (1977) have suggested that the increase in blood brain barrier permeability following ECS may be related to the increase in cerebral blood flow reported during ECT-induced seizures (Brodersen et al., 1973; Bolwig et al., 1977). Mander et al. (1987) in turn postulate that this may account for the rise in T, times seen in their subjects. In our study, however, the earliest effects seen were falls in T, time and water content, followed by a transient rise in values over the following 4 h. It is possible that the changes noted by Mander et al. represent this later short-lived increase in values, though the increase reported was in excess of pre-ECT levels, in contrast to our own findings. Comparison between this and Mander et al.‘s clinical study is complicated by numerous obvious methodological differences, and this possibly accounts for the lack of concordance between detailed findings. Our results probably reflect more accurately the time course of changes following ECS in whole normal brain tissue, but cannot take account of brain T,, T, and water content changes consequent on, for example, depressive illness in human subjects. In depressed patients

164

both relaxation times (Rangel-Guerra et al., 1982) and brain water (Shaw et al., 1969) have been shown to be increased, and this may modify changes observed following ECT. It would appear, then, that there is no clear-cut relationship between regional cerebral blood flow (rCBF), changes in blood brain barrier permeability and observed NMR indices across studies. Sources of variation include Silfverskiold’s finding (1986), using the ‘33Xe technique, that rCBF is substantially reduced 2 h following ECT in human subjects and that this change is reduced with the number of treatments. In addition, Preskorn et al. (1981) have shown that repeated ECS blunts blood brain barrier permeability responses to altered CBF measured 15 min post seizure. In conclusion, in vitro animal models provide a useful technique of establishing directly the relationship between relaxation times and water content and as such are able to complement in vivo studies in man. References Bell, B.A., Smith, M.A., Kean, D.M., McGhee, C.N.J.. MacDonald, H.L., Miller, D., Barnett, G.H., Tocher, J.L., Douglas, R.H.B. and Best, I.J.K. (1987) Brain water measurement by magnetic resonance imaging: correlation with direct estimation and changes after mannitol and dexamethasone. Lance1 i, 66-69. Besson, J.A.O., Greentree, S.G., Foster, M.A. and Rimmington, J.E. (1989) Effects of ethanol on rat brain relaxation times - acute administration, dependency and chronic long term effects. Br. J. Psychiatry 155, 818-821. Besson, J.A.O., Best, P.V. and Skinner, E.R. (1991) Post mortem proton NMR spectrometric measures in brain regions of patients with a pathological diagnosis of

Alzheimer’s disease and multi-infarct dementia. Br. J. Psychiatry tin press). Bolwig, T.G. (1984) The influence of electrically induced seizures on deep brain structures. In: B. Lerer, R.D. Weiner and R.H. Belmaker (Eds.1, E.C.T.: Basic Mechanisms. John Libbey, London. pp. 1322138. Bolwig, T.G., Hertz. M.M. and Holm-Jensen. A. (1977) Blood brain barrier during electroshock seizures in the rat. Eur. J. Clin. Invest. 7, 95-100. Brodersen, P., Paulson, 0.9.. Bolwig, T.G.. Rogon. Z.E.. Rafaelson, O.J. and Lassen, N.A. (1973) Cerebral hyperanaemia in electrically induced epileptic seizures. Arch. Neural. 28, 334-338. Cullis, P.R. and De Kruijff, B. (1980) Lipid polymorphism and the functional roles of lipids in biological membranes. Biochim. Biophys. Acta 559, 339-419. Mander, A.J.. Whitfield, A. and Kean, D.M. (19871 Cerebral and brain stem changes after ECT revealed by nuclear magnetic resonance imaging. Br. J. Psychiatry 151. 69-71. Mathur De Vre, R. (1984) Biomedical implications of the relaxation behaviour of water related to NMR imaging. Br. J. Radiol. 57. 1145-1148. Murase, N. and Watanabe, T. (1989) Nuclear magnetic relaxation studies of the compartmentalised water in cross linked polymer gels. Magn. Reson. Med. 9, l-7. Norton. W.T. (1975) In: D.B. Tower (Ed.). The Nervous System, Vol. 1. Raven Press, New York, NY, p. 467. Preskorn, S.II., Irwin. G.H., Simpson, S., Friesen. D.. Rinne, J. and Jerkovich, G. (1981) Medical therapies for mood disorder alter the blood-brain barrier. Science 213, 46Y471. Rangel-Guerra, R.A.. Perez-Payou. I. and Todd, L.E. (1982) Nuclear magnetic resonance in bipolar affective disorder. Magnet. Reson. Imaging 1, 229. Shaw, D.M., Friail, D. and Camps, F.E. (1969) Brain electrolytes in depressive and alcoholic suicides. Br. J. Psychiatry 115, 69-79. Silfverskiold. P.. Gustafson, L., Risberg. J. and Ingmar, R. (1986) Acute and late effects of electroconvulsive therapy: clinical outcome, regional cerebral blood flow and electroencephalogram. Ann. NY Acad. Sci. 462, 2366248.

Relationship between observed NMR changes and brain water content following electroconvulsive stimulation in the rat.

The administration of electroconvulsive stimuli to anaesthetised rats results in changes in the relaxation times and water content of grey and white m...
519KB Sizes 0 Downloads 0 Views