References 1.

2.

3.

4.

5.

6.

- I .

8.

9.

10.

11.

12.

13.

14.

15.

Dyck PJ, Prineas J, Pollard J. Chronic inflammatory demyelinat- ing polyradiculoneuropachy. In: Dyck PJ, Thomas PK, Griffin JW, et al, eds. Peripheral neuropathy. 3rd ed. Philadelphia: Saunders, 1993: 1498-1 517 Hartung HP, Stoll G, Toyka KV. Immune reactions in the pe- ripheral nervous system. In: Dyck PJ, Thomas PK, Griffin JW, et al, eds. Peripheral neuropathy. 3rd ed. Philadelphia: Saun- ders, 1993:418-444 Hartung HP, Toyka KV. T-cell and macrophage activation in experimental autoimmune neuritis and Guillain-Barre syn- drome. Ann Neurol 1990;27(suppl):S57-S63 Bevilacqua MP, Pober JS, Mendrick DL, et al. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc Natl Acad Sci USA 1987;84:9238-9242 Osborn L. Leukocyte adhesion to endothelium in inflammation. Cell 1990;62:3-6 Bevilacqua MP. Endothelid-leukocyte adhesion molecules. Annu Rev Immunol 1993;11:?6?-804 Pigott R, Dillon LP, Hemingway IH. Gearing AJH. Soluble forms of E-selectin, ICAM-1 and VCAM-1 are present in the supernatants of cytokine-activated cultured enduthelial cells. Biochem Biophys Rcs Commun 1992;187:584-589 Sharief MK, McLean B, Thompson EJ. Elevated serum levels of tumor necrosis factor-a in Guillain-Barre syndrome. Ann Neurol 199’73 3:59 1-596 Kuroki S, Saida T, Nukina M, et al. Campylobacter jejuni strains from patients with Guillain-Barri. syndrome belong mostly to penner serogroup 19 and contain P-N-acerylglucosamine resi- dues. Ann Neurol 1993;33:243-247 Selmaj KW, Raine CS. Tumor necrosis fictor mediates myelin and oligodendrucyte damage in vitro. Ann Neurol 1988;23:

Ryan J, Brett J. Tijburg P, et al. Tumor necrosis factor-induced endothelid tissue factor is associated with subendothelial matrix vesicles but is not expressed on the apical surface. Blood 1992; 49:261-272 Hughes RAC. Immunological mechanisms of demyelination. J R Soc Med 1992;85:53-57 Cronstein BN, Kimmel SC, Levin RI, er al. A mechanism for the antiinflammatory effects of corticosteroids: the glucocorticoid receptor regulates leukocyte adhesion to endothelial cells and expression of endothehal-leukocyte adhesion molecule 1 and intercellular adhesion molecule 1. Proc Narl Acad Sci USA

Sharief MK, Noori MA, Ciardi M, et al. Increased levels of circulating ICAM-I in serum and cerebrospinal fluid of patients with active multiple sclerosis. Correlation with TNF-a and blood-brain barrier damage. J Neuroimmunol 1993;43:15-22 Tsukada N, Miyagi K, Matsuda M, Yanagisawa N. Increased levels of circulating intercellular adhesion molecule- 1 in multiple sclerosis and human T-lymphotropic virus type I-associated my- elopathy. Ann Neurol 1993;33:646-649

339-346

1992;89:9991-9995

Cerebellar a-Ketoglutarate Dehvdrogenase Activitv Is Red;ced;n Spinocerebellar AtaxiaType 1 Frank Mastrogiacomo, BSc, and Stephen J. Kish, PhD

We measured the activity of the thiamine pyrophos- phate-dependent enzyme a-ketoglutarate dehydroge- nase complex in postmortem brain of 12 patients with the spinocerebellar ataxia type 1 form of olivopontocere- bellar atrophy. a-Ketoglutarate dehydrogenase complex activity measured in the absence of thiamine pyrophos- phate was markedly reduced ( - 72%) in olivopontocere- bellar atrophy cerebellar cortex. Decreased activity of this key rate-limiting Krebs cycle enzyme could compro- mise cerebellar energy metabolism and excitatory amino acid synthesis and thereby contribute to the brain dys- function of olivopontocerebellar atrophy.

Mastrogiacomo F, Kish SJ. Cerebellar a-ketoglutarate dehydrogenase activity is reduced in spinocerebellar ataxia type 1.

Ann Neurol 1974;35:624-626

Reduced levels of thiamine and its phosphates have been reported in cerebrospinal fluid of patients with spinocerebellar damage, including patients with domi- nantly inherited olivopontocerebellar atrophy (OPCA; cf. [I}). Experimental animal data suggest that the mechanism by which thiamine deficiency produces brain dysfunction and damage could involve depletion of energy stores consequent to diminished activity of thiamine pyrophosphate (TPP)-dependent enzymes, including especially a-ketoglutarate dehydrogenase complex (aKGDHC), a key rate-limiting enzyme of the Krebs cycle [2]. To our knowledge, no informa- tion is available on the status of aKGDHC in brain of patients with OPCA. Therefore, we measured aKGDHC activity, both in the presence and absence of TPP, in brain of 12 patients with the spinocerebellar ataxia type I form of OPCA.

Patients and Methods Autopsied brain was obtained from 12 clinically end-stage patients with the spinocerebellar ataxia type 1 form [ 3 ] of OPCA and 13 neurologically normal control subjects

From the Human Neurochemical Pathology Laboratory, Clarke In- stitute of Psychiatry, Toronto, Ontario, Canada.

Received Sep 2, 1993, and in revised form Oct 4. Accepted for publication Oct 5 , 1993.

Address correspondence to Dr Kish, Human Neurochemical Pathol- ogy Laboratory, Clarke Institute of Psychiatry, 250 College Street. Toronto, Ontario, Canada M5T 1R8.

624 Copyright 0 1994 by the American Neurological Association

matched ( p > 0.05) with respect to age (OPCA, 40 * 3 yr; controls, 40 2 5 yr), postmortem interval (OPCA, 9 ? 2 hr; controls, 13 -t 1 hr), and, as an index of premortem agonal status (cf. [4]), cerebral (frontal) cortical p H (OPCA, 6.2 ? 0.1; controls, 6.2 2 0.1). The affected gene for this form of OPCA contains an unstable CAG trinucleotide re- peat on chromosome 6 [3}. The diagnosis of OPCA was based upon the presence of clinical signs (limb and gait ataxia, dysarthria, dysphagia) and characteristic neuropathological changes, severe in all of the cases, consisting of neuronal cell loss and gliosis in the Purkinje cell and molecular layers of the cerebellar cortex, basal pontine nuclei, and inferior olives. Semiquantitative histopathological analysis of cerebral cortex showed no evidence of neuronal cell loss or gliosis with the exception of slight neuronal cell loss and gliosis in parietal cortex only of 2 OPCA patients. Most of the OPCA patients had died of bronchopneumonia and had likely received anti- biotics and corticosteroids just prior to death. Activities of a K G D H C and citrate synthase (CS), a marker of rnitochon- drial mass, and brain p H were determined as previously de- scribed [4] . Aspartate and glutamate levels, previously re- ported on 6 of the 12 OPCA patients [5 ) , were determined by a high-performance liquid chromatographic-fluorometric procedure [ S ] .

Results As shown in the Table, aKGDHC activity determined in the presence of a maximally stimulatory concentra- tion of TPP was markedly reduced ( - 50%) in OPCA cerebellar cortex (just missing statistical significance; p = 0.058) but was normal in both frontal and occipital cortices. In the absence of TPP, (YKGDHC levels were significantly and more markedly reduced ( - 72gb) in cerebellar cortex with a nonsignificant trend for de- creased ( - 33%) levels in occipital cortex ( p > 0.05 for frontal cortex). Mean CS activity was slightly but significantly decreased in OPCA frontal (~ 6%), oc- cipital ( - 2393, and cerebellar ( - 20%) cortices. In OPCA cerebellar cortex, activity of CS-corrected

aKGDHC activity in the absence of TPP was signifi- cantly reduced ( - 65%)) with a nonsignificant trend for reduction ( - 39%) when measured in the presence of TPP ( p > 0.05 for other brain areas). The mean differ- ence between aKGDHC or CS-corrected aKGDHC activity in the presence versus absence of TPP (i.e., TPP effect and CS-corrected TPP effect, respectively) was significantly greater in both cerebellar and occipital cortices of the OPCA patients compared with the con- trols, with the exception of the TPP effect in OPCA cerebellar cortex, which just missed statistical signifi- cance ( p = 0.062). In frontal cortex, neither the TPP nor CS-corrected TPP effect between the OPCA and control patients was significantly different ( p > 0.05). Levels of aspartate and, to a lesser extent, glutamate, were significantly reduced in all three brain areas with the greatest reductions occurring in cerebellar cortex.

Discussion Our observation that TPP produced in OPCA above normal stimulation of aKGDHC and CS-corrected aKGDHC activity in both occipital and cerebellar cor- tices provides some indirect support for the presence of lowered endogenous levels of TPP in OPCA brain. This possibility is consistent with the report of de- creased thiamine levels in cerebrospinal fluid of pa- tients with OPCA El]. Although this TPP effect could be explained by reduced dietary thiamine intake in the end-stage OPCA patient, the finding that, in OPCA, blood thiamine levels (which should reflect dietary changes) are normal [l} argues against this interpreta- tion. The observation that aKGDHC activity in OPCA cerebellum remained below normal even in the presence of a maximally stimulatory concentration of TPP could be explained by irreversible inactivation of the enzyme due to chronic deficiency of its cofactor TPP, which has been postulated to have an enzyme

Mitochondria1 Enzyme ActivitieJ and Amino Acid Levels in Postmortem Brain of Conlrol Subjec-tJ and Patients with Olivopontocevebellar Atrophy

CS-"Corrected" aKGDHC + TPP aKGDHC - TPP TPP Effccr CS (QKGDIIC t TPPYCS (aKGDIIC - TPP:/CS TPP Effect Aspartatt Gluramare

Cerebellar cortex Controls OPCA ci of change

occlpltd iurrex Lonrrols OPCA 6 of change

Frontal conex Controls OPCA $$ ot change

6 34 5 1.50 3.20 ? 0.4.i

- 5 0

7 20 ? 1.9? 6.92 f 1.21

- 4

780 2 1.13 8.31 ? 1.30

+ 7

5 33 2 1 4 9 1 4 8 f 0.24a

- 72

7 59 t 2.2s 5.09 2 0.96

-33

7.12 t 1 36 7.42 I 1.52

- 4

1.01 f 0.12 206 i i 0.031 ? 0 00' 1 7 2 i 0 27 165 i 3" 0.019 I 0.003

-20 - 19

039 7 0.4: 213 I 9 0.034 z 0.010 1.83 2 0.17' 164 2 4' 0.041 I 0.00'

- 23 +21

0 68 f 0.12 188 i 1 0.042 t 0.006 0 89 i 3.49 176 I 2' 0 035 I 0.908

-6 1-

0.026 z 0.00:

- 65 0.009 2 0.001

0.016 z 0011 0.030 0.005

- 1'

0.038 I 0 00' 0.03i ? 0.008

- 18

0.005 0.001 21.7 3 1.6 90.2 I 5.2 0.010 z i) 002' 4 ? 0.Y 50.8 I 3 Ra

- 66 - 44

-0.002 0.002 20.2 t 2 2 75.5 z 4.1 (!.Oil 3 0.002' 12.4 3 l .>a 58.8 2 5.3b

- 39 - 22

0.004 f 0.002 16.6 f 1.6 88.1 f 3.2 (j.004 2 0.002 12.11 2 i . i b 75.2 3 . ~

- 28 -15

Valuer represenr mean f SE of 13 control rublcctr and 12 parienrs wirh dominanrly inherited olivopontocerebellar armphv (OPCA) for enzyme acti>ities and 8 to 11 OPCA patients for aminn acid levelc, in lrcntal cortex (Brodmann uea lo), occipital correx (area 171, and cerebellar currex. uKGDHC + TPP = acrivir) (nmollmmlrng of proiem) of u-kemglutardtr dchydrogmasc complch laKGDHC) in the presence of thiamine pgrophosphare iTPP1 added to rhe reacrion mixture; aKGDHC - TPP = activity ofaKGOHC in the absence ofTPP; TPP Effect = mean diffcrrnie between mKGDHC acrivirg m the presence vs absence of TPP CS = acrivity (nmollminimy: of prore~nl of titrate rynthase ICS), iolKGDHC + TPPYCS = artiwy of uKGDHL in the prrsentc of TPP corrected for CS activity, (aKGDHC - TPP1:CS = activity of aKGDHC In the abrencr uf TPP corrected for CS acnviry. CSXorrecred" TPP Lfferr = mean ddfcrcnce k t w c c n Cb corrcired aKGDHC amvxy ~n rhe presence vs absence of TPP; apartate a d gluramate valuer (pmol'gm of protein) represenr levels ot the rerpectlw amino a d 5

' p < 0.02, ' p < 0.05 (two-railed Studenr's f resr).

Brief Communication: Mastrogiacomo and Kish: Brain aKGDHC in OPCA 625

stabilization role (cf. {GI). Decreased activity of aKGDHC, a mitochondrial enzyme, is unlikely to be due in toto to loss of mitochondrid mass, since the degree of enzyme deficiency greatly exceeded the slight reduction in activity of CS, a marker of mito- chondrial content, and cerebellar cortical aKGDHC levels in OPCA, corrected for CS, still showed a statis- tically significant reduction. Reduced clKGDHC levels could be explained at least in part by neuronal loss, since enzyme activity was markedly reduced only in cerebellar cortex, with a slight and nonsignificant de- crease (occipital cortex) or no reduction (frontal cortex) in the morphologically spared cerebral cortex; this pat- tern is not observed in Alzheimer’s disease brain, which is characterized by markedly reduced cerebral cortical aKGDHC activity (cf. {4}). Should aKGDHC be preferentially contained in a subgroup of neurons more severely affected in OPCA (eg., Purkinje cell neurons), then neuronal loss could, in fact, account for the enzyme reduction.

Reduced aKGDHC activity in OPCA cerebellum could have important functional consequences involv- ing both brain energy metabolism and excitatory amino acid production. A 72% reduction in cerebellum of aKGDHC, a rate-limiting Krebs cycle enzyme, might lead to depletion of energy stores, which could affect both brain function and susceptibility to excitotoxic neuronal damage { i} . Although no information ap- pears to be available in the adult, a 7 5 to 95% and apparently biochemically restricted aKGDHC defi- ciency (assessed in cultured fibroblasts) is associated with severe central nervous system developmental ab- normalities in humans 18, 91. Since normal activity of the Krebs cycle is necessary for excitatory amino acid synthesis, a chronic deficiency of aKGDHC might also affect excitatory amino acid levels and neurotransmis- sion. Although we could not detect by regression anal- ysis any statistically significant correlation ( p > 0.05) in OPCA brain between enzyme activity and either aspartate or glutamate levels, our observation that, among the three brain regions examined, aspartate and aKGDHC levels (without TPP) were reduced gener- ally in parallel (see Table) supports this possibility.

In conclusion, we have demonstrated a marked re- duction in the activity of a thiamine-dependent Krebs cycle enzyme in cerebellum of patients with one form of OPCA. Further studies will be required to provide the neuronal/biochemical basis for this finding.

This study was supported by US NIH NINDS grant no. NS26034.

References 1. Pedraza OL, Botez MI. Thiamine scatus in inherited ataxias.

J Neurol Neurosurg Psychiatry 1992;55:136-137 2. Butterworth RF, Heroux M. Effect of pyrithiamine treatment and

subsequent thiamine rehabilitation on regional cerebral amino

’

3.

4.

5 .

6.

I .

8.

9.

acids and thiamine-dependent enzymes. J Neurochem 1989;52:

Orr HT, Chung AM-Y, Banfi S, et al. Expansion of an unstable trinucleotide CAG repeat in spinocerebellar ataxia type 1. Nature Genet 1093;4:22 1-226 Masrrogiacomo F, Bergeron C, Kish SJ. Brain a-ketoglucarate dehydrogenase complex activity in Alzheimer’s disease. J Neuro- chem 1993;61:2007-2014 Kish SJ, Robitaille Y, El-Awar M, et al. Brain amino acid reduc- tions in one family with chromosome 6p-linked dominantly inher- ited olivopontocerebellar atrophy. Ann Neurol 1991;30:780- 784 Blass JP, Gleason P, Brush D, et al. Thiamine and Alzheimer’s disease. Arch Neurol 1988;45:833-835 Bed MF, Hyman BT, Koroshetz W. Do effects in mitochondrial energy metabolism underlie the pathology of neurodegenerative diseases? Trends Neurosci 1993;16:125-13 1 Kahlschutter A, Behbehani A, Langenbeck U, et al. A familial progressive neurodegenerative disease with 2-oxoglutaric acidu- ria. Eur J Pediatr 1982;138:32-37 Bonnefont J-P, Chretien D, Rustin P, et al. Alpha-ketoglutarate dehydrogenase deficiency presenting as congenital lactic acidosis. J Pediatr 1992; 12 1.2 5 5-258

1079-1084

Absence of Postoperative Hyponatremia Syndrome in -Young, Healthy Females Eelco F. M. Wijdicks, MD, and Timothy S. Larson, MD

Young and previously healthy females have been re- ported to develop severe postoperative hyponatremia with a fatal outcome. The clinical presentation is dra- matic, with seizures, respiratory arrest, and permanent, often catastrophic, brain damage. The true incidence is unknown. We report a survey of 290,815 surgical proce- dures on females at the Mayo Clinic from 1976 to 1772. Postoperatively 1,498 females had cardiopulmonary ar- rest, 255 had a metabolic encephalopathy, 32 had new- onset seizures, and 6 had central pontine myelinolysis. We failed to identify any association of respiratory arrest with postoperative hyponatremia. Our findings indicate that the postoperative hyponatremia syndrome in young, healthy females with respiratory arrest is ex- tremely uncommon.

Wijdicks EFM, Larson TS. Absence of postoperative hyponatremia syndrome in young, healthy

females. Ann Neurol 1994;35:626-628

Of the numerous postoperative complications, hypona- tremia is common and is most often associated with

From the Mayo Clinic, Rochester, MN. Received Oct 6, 1093, and in revised form Nov 16. Accepted for publication Nov 18, 1993.

Address correspondence to Dr Wijdicks, Department of Neurology (Neurology Critical Care Service), Mayo Clinic and Mayo Founda- tion, 200 First Street, SW, Rochester, MN 55905.

626 Copyright 0 1994 by the American Neurological Association