Neuroscience Letters, 130 (1991) 133-136 Elsevier Scientific Publishers Ireland Ltd. ADONIS 030439409100508T
133
NSL 08028
Membrane bilayer instability and the pathogenesis of disorders of myelin Lionel Ginsberg* and N o r m a n L. Gershfeld Laboratory of Physical Biology, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD 20892 (U.S.A.,) (Received 31 January 1991; Revised version received 21 March 1991; Accepted 7 June 1991)
Key words: Myelin; Membrane; Metachromatic leukodystrophy; Multiple sclerosis; Lipid bilayer We have previously shown that total lipid extracts from normal nervous tissues spontaneously form a structure in vitro resembling the cell membrane bilayer, but only at a critical temperature, T*, equal to the 'physiological' temperature of the original tissues. In the present study, we found T* for normal human myelin lipids was 37°C, in agreement with the concept that lipid metabolic pools maintain a critical composition in vivo which permits spontaneous formation of the (myelin) membrane bilayer at normal body temperature. But T* for myelin lipids from a patient with metachromatic leukodystrophy was < 30°C. Thus, myelin lipid composition was inappropriate for normal bilayer stability at this patient's core temperature, suggesting a mechanism whereby defective lipid metabolism in this disease could produce pathological myelin. The shift in T* in this patient was unlikely to be simply secondary to myelin destruction, as myelin lipids from a patient with advanced multiple sclerosis yielded a normal value for T* of 37°C, even when extracted from areas of extensive demyelination.
Metachromatic leukodystrophy (MLD) and multiple sclerosis (MS) are diseases in which the main pathological impact is delivered to myelin. In MLD, an inborn error of metabolism leads to an accumulation of cerebroside sulfate [15], a lipid which is found in most significant quantities in myelin [19, 20]. But it is not clear how this metabolic defect produces abnormal myelin. In MS, the role of abnormal central nervous system lipid composition in the pathogenesis of the disease has been much studied [16]. A logical site where a lipid defect could affect normal function is the lipid bilayer of cell membranes. We report here experiments indicating that the link between abnormal lipids and dysmyelination is readily provided by a newly established physicochemical theory of membrane bilayer assembly and stability [8lO]. The salient conclusion from this theory is that the major determinants of membrane bilayer stability, a prerequisite for normal cellular function, are lipid composition and temperature, which are interdependent. This conclusion arose from a systematic thermodynamic study in which it was shown that lipids dispersed in water only form an isolated bilayer state, equivalent to
*Present address: Department of Neurology, University of Cambridge Clinical School, Addenbrooke's Hospital, Cambridge CB2 2QQ, U.K. Correspondence: N.L. Gershfeld, Laboratory of Physical Biology, NIAMS, Building 6, Room 139, National Institutes of Health, Bethesda, MD 20892, U.S.A.
their structure in cell membranes, such as myelin, at a single critical temperature, T*. At other temperatures the equilibrium dispersion will generally form a multilamellar (multibilayer) state [9, 10]. For total lipid extracts from normal biological membranes, T* measured in vitro exactly equals the ambient temperature of the original living membranes [8, 14]. This implies that cellular metabolic pools normally maintain a critical lipid composition which permits the membrane bilayer to assemble spontaneously at the cell's 'physiological' temperature [8, 10, 14]. If the critical conditions of temperature or lipid composition are no longer met, the theory indicates that a membrane bilayer will necessarily degenerate with potentially serious consequences for the cell [10, 1l, 14]. The relevance of this mechanism of membrane degeneration to myelin disorders is here assessed in patients with MLD and MS. A preliminary report of this work has been published in abstract form [13]. Cerebral white matter was dissected from the occipital pole of a patient who had expired from adult-onset MLD. The diagnosis in this patient, who had died aged 32, had been established in life by enzyme assay and confirmed post mortem by neuropathological examination. Thinqayer chromatography of the extracted myelin lipids [4] revealed the predicted qualitative excess of cerebroside sulfate and deficiency of cerebroside. White matter was also dissected from the frontal lobes of a patient who had died from MS, and of an age- and sex-matched control individual who had died from non-neurological
134
disease. Tissue was obtained in the MS case separately from areas containing obvious plaques of demyelination and from areas which were macroscopically plaque-free (normal-appearing white matter). Tissue samples were protected from autolysis by storage before use at - 8 0 ° C . There was no histological evidence of autolysis, nor did the myelin lipid class patterns, demonstrated by unidimensional and two-dimensional thin-layer chromatography [4], suggest that this had occurred. Myelin was prepared from these tissue samples [18]; the yield of myelin from the M L D sample was < 1% of normal, as previously observed [19, 20]. Total lipid extracts were obtained from the isolated myelin and purified [14]. In view of the greater proportion of proteolipid protein in myelin compared to other membranes we have studied [7], an additional step was introduced into our lipid purification regimen [14]. Initial chloroform stock solutions of myelin lipids were evaporated under nitrogen. The dried extracts were then suspended in freshly prepared water, three-times distilled from quartz, heating under nitrogen to 60°C for 10 min to facilitate protein denaturation, and eluted with methanol through Supelclean LC-Si solid phase extraction tubes (Supelco, Inc., Bellefonte, PA) [14]. This first eluate was re-evaporated, dissolved in methanol and eluted again with methanol through a second Supelclean tube. Critical temperarues, T*, for the purified lipid extracts were measured as described previously [8, 10, 14]. Normal human myelin lipids yielded the expected value for T* of 37°C (Fig. 1A). This was consistent with our previous observations of the close correspondence between T* for lipids from normal tissues and the 'physiological' temperature of the source membranes, suggesting that the 'critical bilayer state' [10] can form under existing in vivo conditions, indeed that this process is central to the mechanism of biological membrane assembly [8, 10, 14]. The corollary to these observations is that lipid metabolism, in this case in oligodendroglia, the cells responsible for the synthesis and maintenance of myelin in the central nervous system, is normally geared to permit the (myelin) bilayer to assemble spontaneously at physiological temperatures. For myelin lipids from the patient with MLD, T* was less than 30°C (Fig. 1B), significantly below the normal ambient temperature of 37°C in this patient's tissues. Hospital records provide no evidence for hypothermia even terminally in this patient. Under these circumtAlthough myelin is frequently described as multilamellar, the distinction between an individual oligodendroglial plasma membrane process wrapped many times round an axon to form the myelin sheath and the multiple concentric closed bilayers of a multilamellar vesicle is not only topographical ('jellyroll' vs 'onionskin') but also reflects a difference in physical state [10].
i ;0
B
A
i
i
E k
i 10
;
I
' I
C
i
I
II I
T
l ;
I
D ¸
v
37,
o
3,'
I I
JL
i 20
30
40
5'0
, 20
3'0
i__ 40
50
TEMPERATURE °C
Fig. 1. Estimation of the critical temperature, T*, for total lipid extracts from human myelin in health and disease. Two general conditions must be met to show that a given aqueous dispersion of lipid is able to form the critical bilayer state spontaneously: (i) the equilibrium surface pressure (~e)-temperature phase diagram must pass through a maximum, and (ii) in the temperature range encompassing the surface pressure maximum, the bulk lipid must form a lamellar liquid-crystalline phase [8 10, 14]. The temperature of the surface pressure maximum then gives T*. Condition (ii) was satisfied for all the myelin samples as evidenced by the appearance of liposomes and other lamellar liquid-crystalline structures when the dried lipid extracts were allowed to swell in water at an appropriate temperature and viewed under phase-contrast (Zetopan microscope, A/O Reichert, Buffalo, NY, equipped with constant-temperature stage, Cambion, Cambridge, MA). The values of T* for the myelin lipid extracts were then given by the locations of the maxima in the phase diagrams: (A) normal control, (B) metachromatic leukodystrophy, (C) multiple sclerosis: normal-appearing white matter, (D) multiple sclerosis: plaque and periplaque. Each curve was drawn from data obtained with a single lipid preparation; temperatures were selected at random to avoid systematic errors. For each point where error bars are drawn 3 individual films were used.
stances, the critical state theory predicts that a membrane bilayer will necessarily transform into a multilamellar state, creating membrane structural defects, ultimately with potentially serious consequences at cellular level [11, 14]. Thus, the theory provides a mechanism whereby the lipid metabolic defect in MLD could produce pathological myelin I. The patient's oligodendroglial lipid metabolic pool is envisaged as being constrained by the worsening excess of cerebroside sulfate and deficiency of cerebroside, hence being unable to provide a composition appropriate for normal myelin bilayer stability at 37°C. The profound metabolic disturbance in MLD might have been expected to prevent oligodendroglial maturation ab initio, particularly as excessive cerebroside sulfate has been demonstrated in the central nervous system of a fetus with MLD [2]. Yet the disease may develop at any stage, from prenatal [6] to late adult [3]. Much of this variation in age of onset may be explicable in terms of
135 degree of deficiency of arylsulfatase A, the enzyme responsible for catabolism of cerebroside sulfate [22]. But application of the critical state theory provides a more complete explanation for the ability of oligodendroglia initially to develop normally in the face of this metabolic insult, with disease only becoming manifest later in life. Early in development, only a small excess of cerebroside sulfate and deficiency of cerebroside would be expected. Possibly, the composition of oligodendroglial lipid metabolic pools can initially adjust to counteract this imbalance, therefore still permitting normal myelin bilayer assembly, that is T* remains at 37°C. The ability of more than one composition of a given combination of lipids (differing mole fractions of the same components) to yield the same value for T* is a direct consequence of the Gibbs phase rule [9, 10]. The deviations from normal fatty acid composition observed in M L D white matter lipids [21] are consistent with this concept of a homeostatic mechanism. Eventually, however, the cells' ability to compensate for the metabolic defect would be exceeded. The theory then predicts slow degradation of the myelin bilayer, the rate of this process increasing as the difference between body temperature and T* widens [10, 11, 14]. Total lipid extracts from MS myelin yielded a normal value for T* of 37°C, both for the normal-appearing white matter sample and for the sample taken from an area containing obvious plaques of demyelination (Fig. 1C,D resp.). Thus, the shift in T* observed for M L D lipids was unlikely simply to be secondary to demyelination. Beyond acting as a control for M L D , the results in Fig. 1C,D provide some insight into the pathogenesis of MS. Various early hypotheses postulated a primary lipid biochemical abnormality in MS patients, predisposing them to demyelination [5, 25]. Although MS plaque lipid composition differs from normal white matter, and lipid abnormalities have even been detected in normalappearing white matter from MS patients [12, 17, 27], no single lipid metabolic defect has been identified [1, 24]. Current theories of MS pathogenesis involve myelin damage by immune-mediated mechanisms [23, 26] and do not depend on a primary lipid disturbance. The abnormal lipid composition of MS white matter is thought
2The apparent paradox of finding abnormal lipid composition in MS white matter [12, 17, 27] yet no change in T* is again resolved by application of the Gibbs phase rule [9, 10]. The value ofT* for normal myelin lipids and for lipids from other cellular membranes in normal white matter will necessarilybe the same as that for a total lipid extract from whole (unfractionated) white matter, i.e. 37°C. If the proportions of myelin and the other membranes change due to disease, the total lipid composition of white matter will also change but the composition of the individual membranes will be unaltered and T* will be unaffected [9, 10].
simply to reflect myelin loss and replacement by other cell types. Our observations are entirely consistent with these latter views as myelin destruction by non-lipid dependent routes would not necessarily be expected to affect T* (see footnote 2). The critical state theory of membrane assembly and stability therefore provides a mechanism for faulty myelin formation in a disorder with a known primary disturbance of lipid metabolism, and measurement of T* can distinguish this disease from demyelination due to nonlipid dependent processes. The membrane destabilisation mechanism may be relevant to the pathogenesis of other inherited lipid metabolic defects, and may help identify an underlying primary disturbance of lipid chemistry in neurological diseases with etiologies which are presently obscure. Tissue specimens were obtained from the National Neurological Research Bank, V A M C Wadsworth Division, Los Angeles, CA 90073, U.S.A., which is sponsored by N I N D S / N I M H , NMSS, H D Foundation, Comprehensive Epilepsy Program, Tourette Syndrome Association, Dystonia Research Foundation, and Veterans Administration. Myelin extractions were conducted in the laboratory of Dr. R.H. Quarles, N I N D S , N I H , Bethesda, M D 20892, U . S . A . L . G . was Visiting Scientist at the National Institutes of Health, 1988-1989. I Alling, C., Vanier, M.-T. and Svennerholm, L., Lipid alterations in apparently normal white matter in multiple sclerosis, Brain Res., 35 (1971) 325 336. 2 Baier, W. and Harzer, K., Sulfatides in prenatal metachromatic leukodystrophy, J. Neurochem., 41 (1983) 176(~1768. 3 Bosch, E.P. and Hart, M.N., Late adult-onset metachromatic leukodystrophy. Dementia and polyneuropathy in a 63-year-old man, Arch. Neurol., 35 (1978) 475-477. 4 Christie, W.W., Lipid Analysis, Pergamon, Oxford, 1973. 5 Dick, G., The etiology of multiple sclerosis, Proc. R. Soc. Med., 69 (1976) 611~15. 6 Feigin, I., Diffusecerebral sclerosis(metachromaticleuko-encephalopathy), Am. J. Pathol., 30 (1954) 715 737. 7 Folch-Pi, J. and Stoffyn, P.J., Proteolipids from membrane systems, Ann. N.Y. Acad. Sci., 195 (1972) 86-107. 8 Gershfeld, N.L., Phospholipid surface bilayers at the air-water interface. III. Relation between surface bilayer formation and lipid bilayer assemblyin cell membranes, Biophys.J., 50 (1986) 457-461. 9 Gershfeld, N.L., Spontaneous assembly of a phospholipid bilayer as a critical phenomenon: influence of temperature, composition and physical state, J. Phys. Chem., 93 (1989) 5256-5261. 10 Gershfeld, N.L., The critical unilamellar lipid state; a perspective for membrane bilayer assembly, Biochim. Biophys. Acta Rev. Biomembr., 988 (1989) 335 350. 11 Gershfeld, N.L. and Murayama, M., Thermal instability of red blood cell membrane bilayers: temperature dependence of hemolysis, J. Membr. Biol., 101 (1988) 67-72. 12 Gerstl, B., Kahnke, M.J., Smith, J.K., Tavaststjerna, M.G. and Hayman, R.B., Brain lipids in multiple sclerosis and other diseases, Brain, 84 (1961) 310-319.
136 13 Ginsberg, L. and Gershfeld, N.L., Membrane bilayer instability and the pathogenesis of myelin disorders, J. Neurol., 237 (1990) $49. 14 Ginsberg, L., Gilbert, D.L. and Gershfeld, N.L., Membrane bilayer assembly in neural tissue of rat and squid as a critical phenomenon: influence of temperature and membrane proteins, J. Membr. Biol., 119 (1991) 65 73. 15 Jatzkewitz, H., Zwei Typen von Cerebrosid-schwefels~ureestern als sog. 'Pr~ilipoide' und Speichersubstanzen bei der Leukodystrophie, Typ Scholz (metachromatische Form der diffusen Sklerose), Z. Physiol. Chem., 311 (1958) 279-282. 16 Matthews, W.B., Acheson, E.D., Batchelor, J.R. and Weller, R.O., McAlpine's Multiple Sclerosis, Churchill Livingstone, Edinburgh, 1985. 17 Neu, I. and Woelk, H., Investigations of the lipid metabolism of the white matter in multiple sclerosis: changes in glycerophosphatides and lipid-splitting enzymes, Neurochem. Res., 7 (1982) 727 735. 18 Norton, W.T. and Poduslo, S.E., Myelination in rat brain: method of rnyelin isolation, J. Neurochem., 21 (1973) 749 757. 19 Norton, W.T. and Poduslo, S.E., Biochemical studies of metachromatic leukodystrophy in three siblings, Acta Neuropathol., 57 (1982) 188-196. 20 O'Brien, J.S. and Sampson, E.L., Myelin membrane: a molecular abnormality, Science, 150 (1965) 1613 1614.
21 Pilz, H. and Heipertz, R., The fatty acid composition of cerebrosides and sulfatides in a case of adult metachromatic leukodystrophy, Z. Neurol., 206 (1974) 203 208. 22 Polten, A., Fluharty, A.L., Fluharty, C.B., Kappler, J., von Figura, K. and Gieselmann, V., Molecular basis of different forms of metachromatic leukodystrophy, N. Engl. J. Med., 324 (1991) 1822. 23 Scolding, N.J., Morgan, B.P., Houston, W.A.J., Linington, C., Campbell, A.K. and Compston, D.A.S., Vesicular removal by oligodendrocytes of membrane attack complexes formed by activated complement, Nature, 339 (1989) 62ff622. 24 Suzuki, K., Kamoshita, S., Eto, Y., Tourtellotte, W.W. and Gonatas, J.O., Myelin in multiple sclerosis. Composition of myelin from normal-appearing white matter, Arch. Neurol., 28 (1973) 293 297. 25 Thompson, R.H.S., Fatty acid metabolism in multiple sclerosis, Biochem. Soc. Symp., 35 (1973) 103 111. 26 Wren, D.R. and Noble, M., Oligodendrocytes and oligodendrocyte/type-2 astrocyte progenitor cells of adult rats are specifically susceptible to the lytic effects of complement in absence of antibody, Proc. Natl. Acad. Sci. U.S.A., 86 (1989) 9025-9029. 27 Yanagihara, T. and Cumings, J.N., Alterations of phospholipids, particularly plasmalogens, in the demyelination of multiple sclerosis as compared with that of cerebral oedema, Brain, 92 (1969) 59 70.
133
NSL 08028
Membrane bilayer instability and the pathogenesis of disorders of myelin Lionel Ginsberg* and N o r m a n L. Gershfeld Laboratory of Physical Biology, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD 20892 (U.S.A.,) (Received 31 January 1991; Revised version received 21 March 1991; Accepted 7 June 1991)
Key words: Myelin; Membrane; Metachromatic leukodystrophy; Multiple sclerosis; Lipid bilayer We have previously shown that total lipid extracts from normal nervous tissues spontaneously form a structure in vitro resembling the cell membrane bilayer, but only at a critical temperature, T*, equal to the 'physiological' temperature of the original tissues. In the present study, we found T* for normal human myelin lipids was 37°C, in agreement with the concept that lipid metabolic pools maintain a critical composition in vivo which permits spontaneous formation of the (myelin) membrane bilayer at normal body temperature. But T* for myelin lipids from a patient with metachromatic leukodystrophy was < 30°C. Thus, myelin lipid composition was inappropriate for normal bilayer stability at this patient's core temperature, suggesting a mechanism whereby defective lipid metabolism in this disease could produce pathological myelin. The shift in T* in this patient was unlikely to be simply secondary to myelin destruction, as myelin lipids from a patient with advanced multiple sclerosis yielded a normal value for T* of 37°C, even when extracted from areas of extensive demyelination.
Metachromatic leukodystrophy (MLD) and multiple sclerosis (MS) are diseases in which the main pathological impact is delivered to myelin. In MLD, an inborn error of metabolism leads to an accumulation of cerebroside sulfate [15], a lipid which is found in most significant quantities in myelin [19, 20]. But it is not clear how this metabolic defect produces abnormal myelin. In MS, the role of abnormal central nervous system lipid composition in the pathogenesis of the disease has been much studied [16]. A logical site where a lipid defect could affect normal function is the lipid bilayer of cell membranes. We report here experiments indicating that the link between abnormal lipids and dysmyelination is readily provided by a newly established physicochemical theory of membrane bilayer assembly and stability [8lO]. The salient conclusion from this theory is that the major determinants of membrane bilayer stability, a prerequisite for normal cellular function, are lipid composition and temperature, which are interdependent. This conclusion arose from a systematic thermodynamic study in which it was shown that lipids dispersed in water only form an isolated bilayer state, equivalent to
*Present address: Department of Neurology, University of Cambridge Clinical School, Addenbrooke's Hospital, Cambridge CB2 2QQ, U.K. Correspondence: N.L. Gershfeld, Laboratory of Physical Biology, NIAMS, Building 6, Room 139, National Institutes of Health, Bethesda, MD 20892, U.S.A.
their structure in cell membranes, such as myelin, at a single critical temperature, T*. At other temperatures the equilibrium dispersion will generally form a multilamellar (multibilayer) state [9, 10]. For total lipid extracts from normal biological membranes, T* measured in vitro exactly equals the ambient temperature of the original living membranes [8, 14]. This implies that cellular metabolic pools normally maintain a critical lipid composition which permits the membrane bilayer to assemble spontaneously at the cell's 'physiological' temperature [8, 10, 14]. If the critical conditions of temperature or lipid composition are no longer met, the theory indicates that a membrane bilayer will necessarily degenerate with potentially serious consequences for the cell [10, 1l, 14]. The relevance of this mechanism of membrane degeneration to myelin disorders is here assessed in patients with MLD and MS. A preliminary report of this work has been published in abstract form [13]. Cerebral white matter was dissected from the occipital pole of a patient who had expired from adult-onset MLD. The diagnosis in this patient, who had died aged 32, had been established in life by enzyme assay and confirmed post mortem by neuropathological examination. Thinqayer chromatography of the extracted myelin lipids [4] revealed the predicted qualitative excess of cerebroside sulfate and deficiency of cerebroside. White matter was also dissected from the frontal lobes of a patient who had died from MS, and of an age- and sex-matched control individual who had died from non-neurological
134
disease. Tissue was obtained in the MS case separately from areas containing obvious plaques of demyelination and from areas which were macroscopically plaque-free (normal-appearing white matter). Tissue samples were protected from autolysis by storage before use at - 8 0 ° C . There was no histological evidence of autolysis, nor did the myelin lipid class patterns, demonstrated by unidimensional and two-dimensional thin-layer chromatography [4], suggest that this had occurred. Myelin was prepared from these tissue samples [18]; the yield of myelin from the M L D sample was < 1% of normal, as previously observed [19, 20]. Total lipid extracts were obtained from the isolated myelin and purified [14]. In view of the greater proportion of proteolipid protein in myelin compared to other membranes we have studied [7], an additional step was introduced into our lipid purification regimen [14]. Initial chloroform stock solutions of myelin lipids were evaporated under nitrogen. The dried extracts were then suspended in freshly prepared water, three-times distilled from quartz, heating under nitrogen to 60°C for 10 min to facilitate protein denaturation, and eluted with methanol through Supelclean LC-Si solid phase extraction tubes (Supelco, Inc., Bellefonte, PA) [14]. This first eluate was re-evaporated, dissolved in methanol and eluted again with methanol through a second Supelclean tube. Critical temperarues, T*, for the purified lipid extracts were measured as described previously [8, 10, 14]. Normal human myelin lipids yielded the expected value for T* of 37°C (Fig. 1A). This was consistent with our previous observations of the close correspondence between T* for lipids from normal tissues and the 'physiological' temperature of the source membranes, suggesting that the 'critical bilayer state' [10] can form under existing in vivo conditions, indeed that this process is central to the mechanism of biological membrane assembly [8, 10, 14]. The corollary to these observations is that lipid metabolism, in this case in oligodendroglia, the cells responsible for the synthesis and maintenance of myelin in the central nervous system, is normally geared to permit the (myelin) bilayer to assemble spontaneously at physiological temperatures. For myelin lipids from the patient with MLD, T* was less than 30°C (Fig. 1B), significantly below the normal ambient temperature of 37°C in this patient's tissues. Hospital records provide no evidence for hypothermia even terminally in this patient. Under these circumtAlthough myelin is frequently described as multilamellar, the distinction between an individual oligodendroglial plasma membrane process wrapped many times round an axon to form the myelin sheath and the multiple concentric closed bilayers of a multilamellar vesicle is not only topographical ('jellyroll' vs 'onionskin') but also reflects a difference in physical state [10].
i ;0
B
A
i
i
E k
i 10
;
I
' I
C
i
I
II I
T
l ;
I
D ¸
v
37,
o
3,'
I I
JL
i 20
30
40
5'0
, 20
3'0
i__ 40
50
TEMPERATURE °C
Fig. 1. Estimation of the critical temperature, T*, for total lipid extracts from human myelin in health and disease. Two general conditions must be met to show that a given aqueous dispersion of lipid is able to form the critical bilayer state spontaneously: (i) the equilibrium surface pressure (~e)-temperature phase diagram must pass through a maximum, and (ii) in the temperature range encompassing the surface pressure maximum, the bulk lipid must form a lamellar liquid-crystalline phase [8 10, 14]. The temperature of the surface pressure maximum then gives T*. Condition (ii) was satisfied for all the myelin samples as evidenced by the appearance of liposomes and other lamellar liquid-crystalline structures when the dried lipid extracts were allowed to swell in water at an appropriate temperature and viewed under phase-contrast (Zetopan microscope, A/O Reichert, Buffalo, NY, equipped with constant-temperature stage, Cambion, Cambridge, MA). The values of T* for the myelin lipid extracts were then given by the locations of the maxima in the phase diagrams: (A) normal control, (B) metachromatic leukodystrophy, (C) multiple sclerosis: normal-appearing white matter, (D) multiple sclerosis: plaque and periplaque. Each curve was drawn from data obtained with a single lipid preparation; temperatures were selected at random to avoid systematic errors. For each point where error bars are drawn 3 individual films were used.
stances, the critical state theory predicts that a membrane bilayer will necessarily transform into a multilamellar state, creating membrane structural defects, ultimately with potentially serious consequences at cellular level [11, 14]. Thus, the theory provides a mechanism whereby the lipid metabolic defect in MLD could produce pathological myelin I. The patient's oligodendroglial lipid metabolic pool is envisaged as being constrained by the worsening excess of cerebroside sulfate and deficiency of cerebroside, hence being unable to provide a composition appropriate for normal myelin bilayer stability at 37°C. The profound metabolic disturbance in MLD might have been expected to prevent oligodendroglial maturation ab initio, particularly as excessive cerebroside sulfate has been demonstrated in the central nervous system of a fetus with MLD [2]. Yet the disease may develop at any stage, from prenatal [6] to late adult [3]. Much of this variation in age of onset may be explicable in terms of
135 degree of deficiency of arylsulfatase A, the enzyme responsible for catabolism of cerebroside sulfate [22]. But application of the critical state theory provides a more complete explanation for the ability of oligodendroglia initially to develop normally in the face of this metabolic insult, with disease only becoming manifest later in life. Early in development, only a small excess of cerebroside sulfate and deficiency of cerebroside would be expected. Possibly, the composition of oligodendroglial lipid metabolic pools can initially adjust to counteract this imbalance, therefore still permitting normal myelin bilayer assembly, that is T* remains at 37°C. The ability of more than one composition of a given combination of lipids (differing mole fractions of the same components) to yield the same value for T* is a direct consequence of the Gibbs phase rule [9, 10]. The deviations from normal fatty acid composition observed in M L D white matter lipids [21] are consistent with this concept of a homeostatic mechanism. Eventually, however, the cells' ability to compensate for the metabolic defect would be exceeded. The theory then predicts slow degradation of the myelin bilayer, the rate of this process increasing as the difference between body temperature and T* widens [10, 11, 14]. Total lipid extracts from MS myelin yielded a normal value for T* of 37°C, both for the normal-appearing white matter sample and for the sample taken from an area containing obvious plaques of demyelination (Fig. 1C,D resp.). Thus, the shift in T* observed for M L D lipids was unlikely simply to be secondary to demyelination. Beyond acting as a control for M L D , the results in Fig. 1C,D provide some insight into the pathogenesis of MS. Various early hypotheses postulated a primary lipid biochemical abnormality in MS patients, predisposing them to demyelination [5, 25]. Although MS plaque lipid composition differs from normal white matter, and lipid abnormalities have even been detected in normalappearing white matter from MS patients [12, 17, 27], no single lipid metabolic defect has been identified [1, 24]. Current theories of MS pathogenesis involve myelin damage by immune-mediated mechanisms [23, 26] and do not depend on a primary lipid disturbance. The abnormal lipid composition of MS white matter is thought
2The apparent paradox of finding abnormal lipid composition in MS white matter [12, 17, 27] yet no change in T* is again resolved by application of the Gibbs phase rule [9, 10]. The value ofT* for normal myelin lipids and for lipids from other cellular membranes in normal white matter will necessarilybe the same as that for a total lipid extract from whole (unfractionated) white matter, i.e. 37°C. If the proportions of myelin and the other membranes change due to disease, the total lipid composition of white matter will also change but the composition of the individual membranes will be unaltered and T* will be unaffected [9, 10].
simply to reflect myelin loss and replacement by other cell types. Our observations are entirely consistent with these latter views as myelin destruction by non-lipid dependent routes would not necessarily be expected to affect T* (see footnote 2). The critical state theory of membrane assembly and stability therefore provides a mechanism for faulty myelin formation in a disorder with a known primary disturbance of lipid metabolism, and measurement of T* can distinguish this disease from demyelination due to nonlipid dependent processes. The membrane destabilisation mechanism may be relevant to the pathogenesis of other inherited lipid metabolic defects, and may help identify an underlying primary disturbance of lipid chemistry in neurological diseases with etiologies which are presently obscure. Tissue specimens were obtained from the National Neurological Research Bank, V A M C Wadsworth Division, Los Angeles, CA 90073, U.S.A., which is sponsored by N I N D S / N I M H , NMSS, H D Foundation, Comprehensive Epilepsy Program, Tourette Syndrome Association, Dystonia Research Foundation, and Veterans Administration. Myelin extractions were conducted in the laboratory of Dr. R.H. Quarles, N I N D S , N I H , Bethesda, M D 20892, U . S . A . L . G . was Visiting Scientist at the National Institutes of Health, 1988-1989. I Alling, C., Vanier, M.-T. and Svennerholm, L., Lipid alterations in apparently normal white matter in multiple sclerosis, Brain Res., 35 (1971) 325 336. 2 Baier, W. and Harzer, K., Sulfatides in prenatal metachromatic leukodystrophy, J. Neurochem., 41 (1983) 176(~1768. 3 Bosch, E.P. and Hart, M.N., Late adult-onset metachromatic leukodystrophy. Dementia and polyneuropathy in a 63-year-old man, Arch. Neurol., 35 (1978) 475-477. 4 Christie, W.W., Lipid Analysis, Pergamon, Oxford, 1973. 5 Dick, G., The etiology of multiple sclerosis, Proc. R. Soc. Med., 69 (1976) 611~15. 6 Feigin, I., Diffusecerebral sclerosis(metachromaticleuko-encephalopathy), Am. J. Pathol., 30 (1954) 715 737. 7 Folch-Pi, J. and Stoffyn, P.J., Proteolipids from membrane systems, Ann. N.Y. Acad. Sci., 195 (1972) 86-107. 8 Gershfeld, N.L., Phospholipid surface bilayers at the air-water interface. III. Relation between surface bilayer formation and lipid bilayer assemblyin cell membranes, Biophys.J., 50 (1986) 457-461. 9 Gershfeld, N.L., Spontaneous assembly of a phospholipid bilayer as a critical phenomenon: influence of temperature, composition and physical state, J. Phys. Chem., 93 (1989) 5256-5261. 10 Gershfeld, N.L., The critical unilamellar lipid state; a perspective for membrane bilayer assembly, Biochim. Biophys. Acta Rev. Biomembr., 988 (1989) 335 350. 11 Gershfeld, N.L. and Murayama, M., Thermal instability of red blood cell membrane bilayers: temperature dependence of hemolysis, J. Membr. Biol., 101 (1988) 67-72. 12 Gerstl, B., Kahnke, M.J., Smith, J.K., Tavaststjerna, M.G. and Hayman, R.B., Brain lipids in multiple sclerosis and other diseases, Brain, 84 (1961) 310-319.
136 13 Ginsberg, L. and Gershfeld, N.L., Membrane bilayer instability and the pathogenesis of myelin disorders, J. Neurol., 237 (1990) $49. 14 Ginsberg, L., Gilbert, D.L. and Gershfeld, N.L., Membrane bilayer assembly in neural tissue of rat and squid as a critical phenomenon: influence of temperature and membrane proteins, J. Membr. Biol., 119 (1991) 65 73. 15 Jatzkewitz, H., Zwei Typen von Cerebrosid-schwefels~ureestern als sog. 'Pr~ilipoide' und Speichersubstanzen bei der Leukodystrophie, Typ Scholz (metachromatische Form der diffusen Sklerose), Z. Physiol. Chem., 311 (1958) 279-282. 16 Matthews, W.B., Acheson, E.D., Batchelor, J.R. and Weller, R.O., McAlpine's Multiple Sclerosis, Churchill Livingstone, Edinburgh, 1985. 17 Neu, I. and Woelk, H., Investigations of the lipid metabolism of the white matter in multiple sclerosis: changes in glycerophosphatides and lipid-splitting enzymes, Neurochem. Res., 7 (1982) 727 735. 18 Norton, W.T. and Poduslo, S.E., Myelination in rat brain: method of rnyelin isolation, J. Neurochem., 21 (1973) 749 757. 19 Norton, W.T. and Poduslo, S.E., Biochemical studies of metachromatic leukodystrophy in three siblings, Acta Neuropathol., 57 (1982) 188-196. 20 O'Brien, J.S. and Sampson, E.L., Myelin membrane: a molecular abnormality, Science, 150 (1965) 1613 1614.
21 Pilz, H. and Heipertz, R., The fatty acid composition of cerebrosides and sulfatides in a case of adult metachromatic leukodystrophy, Z. Neurol., 206 (1974) 203 208. 22 Polten, A., Fluharty, A.L., Fluharty, C.B., Kappler, J., von Figura, K. and Gieselmann, V., Molecular basis of different forms of metachromatic leukodystrophy, N. Engl. J. Med., 324 (1991) 1822. 23 Scolding, N.J., Morgan, B.P., Houston, W.A.J., Linington, C., Campbell, A.K. and Compston, D.A.S., Vesicular removal by oligodendrocytes of membrane attack complexes formed by activated complement, Nature, 339 (1989) 62ff622. 24 Suzuki, K., Kamoshita, S., Eto, Y., Tourtellotte, W.W. and Gonatas, J.O., Myelin in multiple sclerosis. Composition of myelin from normal-appearing white matter, Arch. Neurol., 28 (1973) 293 297. 25 Thompson, R.H.S., Fatty acid metabolism in multiple sclerosis, Biochem. Soc. Symp., 35 (1973) 103 111. 26 Wren, D.R. and Noble, M., Oligodendrocytes and oligodendrocyte/type-2 astrocyte progenitor cells of adult rats are specifically susceptible to the lytic effects of complement in absence of antibody, Proc. Natl. Acad. Sci. U.S.A., 86 (1989) 9025-9029. 27 Yanagihara, T. and Cumings, J.N., Alterations of phospholipids, particularly plasmalogens, in the demyelination of multiple sclerosis as compared with that of cerebral oedema, Brain, 92 (1969) 59 70.
