- Email: [email protected]
Humoral response to hsp 65 and hsp 70 in cerebrospinal fluid in Parkinson's disease
JOURNAL
OF THE
NEUROLOGICAL SCIENCES ELSEVIER
Journal of the Neurological Sciences I39 (1996) 66-70
Humoral response to hsp 65 and hsp 70 in cerebrospinal fluid in Parkinson’s disease U. Fiszer a3* , S. Fredrikson b, A. Czlonkowska a h Dir,ision
’ lnsitute of Psychiatry and Neurology. Warsaw, Poland of Neurology. Karolinska Institute, Huddingr Hospital, Huddinge,
Sweden
Received 13 October 1994;revised 7 December 1995;accepted 18 December 1995
Abstract Parkinson’s disease (PD) is an age-related neurodegenerative movement disorder of unknown etiology. In PD immune abnormalities were reported, but the cause of such abnormalities has not been resolved. Recently, the increased proportion of yS+ T cells out of all T cells has been found in patients with PD. Heat shock proteins (hsps) could be targets for ~6~ T lymphocytes. We examined serum and CSF of patients with PD, age-matched patients with other non-inflammatory neurological diseases (OND old), young patients with other non-inflammatory neurological diseases (OND young), and donors of blood (DB). Antibodies were detected using the enzyme-linked immunosorbent assay. The plates were coated with recombinant mycobacterial hsp 65 and hsp 70. The present study showed that the mean ELISA ratio of CSF from patients with PD was significantly greater than that of CSF from patients with OND old (tested against IgG anti-hsp 65 and IgG anti-hsp 70) and OND young (tested against IgG anti-hsp 70). There was no difference between the mean ELISA ratio of sera from patients with PD, OND old and OND young (tested against IgG anti-hsp 65 and IgG anti-hsp 70). The significance of hsps immmunity is not completly clear. Increased hsps expression, which is induced by stress, provides cells with protection against the environmental insults. Alternatively, the antibodies may be present as a consequence of prior infections. Keywords;
Parkinson’s disease;Cerebrospinal fluid: Heat shock protein
1. Introduction Parkinson’s disease (PD) is an age-related neurodegenerative movement disorder of unknown etiology. The massive loss of pigmented dopaminergic nerve cells takes place in the affected brain areas. Lewy bodies are now considered a prerequisite for the definite diagnosis of PD, although their origin has remained unclear. Immune abnormalities have been broadly described, including the occurrence of autoantibodies against neuronal structures (Pouplard et al., 1979), reduced numbers of peripheral blood lymphocytes and reduced mitogen responses (Marttila et al., 1984), and complement-activated oligodendroglial cells (Yamada et al., 1990). The HLA-DR method highlights phagocytic activity surrounding dead and dying neurons which is a prominent feature in idiopathic PD (McGeer et al., 1988). A higher HLA-DR expression on CSF mono-
* Corresponding author. II Department of Neurology, Institute of Psychiatry and Neurology, Al. Sobieskiego l/9, 02.957 Warsaw, Poland, Tel: (+ 48-22) 42-40-23, Fax: (+ 48.22) 42-40-23, ( + 48-2) 642-53-75. 0022-510X/96/$15.00 P/I SOO22-5
0 1996 Elsevier Science B.V. All rights reserved
10X(96)00002-0
cytes in comparison with blood monocytes has been found (Fiszer et al., 1994a). Moreover, the elevated proportion of y6+ T cells out of all T cells in CSF and blood has been found in PD (Fiszer et al., 1994b). Heat shock proteins (hsps) could be targets for y6+ T lymphocytes (O’Brien and Born, 1991). Hsps or stress proteins comprise about two dozen proteins belonging to several families (Kaufmann, 1990, Craig et al., 1993). The induction of hsps expression may serve several purposes at a cellular level. Little is currently known regarding the role of hsps in the central nervous system; here hsps could have a critical function (Chopp, 1993). Variable expression of hsps by human glial cells has been shown by Freedman et al. (1992). Gao et al. (1994) suggest that antibodies reactive with hsp 60/65 are present in CSF in a number of chronic neurodegenerative conditions (Alzheimer’s disease, amyatrophic lateral sclerosis, PD). Since the effective cause of immune abnormalities in these patients is not entirely clear, we attempted to identify changes in anti-hsp 65 and 70 immunity in serum and CSF
U. Fiszer et al. / Journal
of the
Neurological
of patients with PD and patients with other neurological diseases (OND young and OND old, in relation to the aging process) and healthy donors of blood (DB). 2. Materials
and methods
2.1. Patients We studied 22 patients (15 men, 7 women) with PD and without evidence of any other associated neurological disorders. The diagnosis was established on the basis of clinical examination (subjects presented with typical PD symptoms and signs). All patients received a computed tomographic scan. Their ages ranged from 47 to 83 years (mean 71 years). The duration of PD varied between 1 and 18 years (mean 5.4 years). Fourteen patients were treated with levodopa at the time of the present study. CSF was obtained from 20 patients. A mononuclear pleocytosis ( > 5 X lo6 cells/l) was found in one patient and elevated CSF/serum albumin ratio (Tibbling et al., 1977) reflecting blood-CSF barrier damage in 4 patients. The IgG index equal to (CSF/serum IgG)/(CSF/serum/albumin) (Tibbling et al., 1977) was normal ( < 0.7) in all patients, and none of them had oligoclonal IgG bands. Three groups of control patients were included in the study. The first control group consisted of 15 patients (8 women) with OND young. Their ages varied between 16 and 47 years (mean 31.5 years). Seven of these patients had epilepsy, 3 had neurosis, 3 had intervertebral disc disease, 1 patient had neuropathy, and 1 patient had headache. None of them revealed any signs or symptoms of other diseases, and they displayed normal routine CSF findings. The second control group consisted of 13 age-matched patients (8 women) with OND old. Their ages varied between 42 and 84 years (mean 56.2 years) (with no statistically significant difference between PD and OND old). Five of these patients had cerebrovascular disease, 2 had epilepsy, 2 had spinal cord tumor, 1 had neuropathy, 1 had hereditary ataxia, 1 had intervertebral disc disease, and 1 had headache. None of them revealed any signs or symptoms of other diseases, and they displayed normal routine CSF findings. The third control group consisted of 17 (2 women) healthy DB. Their ages varied between, 19 and 54 years (mean 34.5 years); blood of those persons was collected from the local blood bank. Serum and CSF was stored at - 70°C. 2.2. Assay of antibodies to hsp 65 and hsp 70 Antibodies were detected using an enzyme-linked immunoabsorbent assay (ELISA) according to the method described by Danieli et al. (1992) (the method was modified for the purposes of our work). 96-well microtitre plates (Costar, Cambridge, MA,
Sciences
139 (19961 66-70
61
USA) were coated with recombinant mycobacterial hsp 65 (Mycobacterium bovis BCG, batch MA-12A) and hsp 70 (Mycobacterium tuberculosum, batch MT70-6A, a gift from Dr. J.D.A. van Embden, National Institute of Public Health and Environmental Protection, Bilthoven, Netherlands), 1 kg/ml in PBS; 100 ~1 per well was added at 4°C overnight. The plates were coated with BPS containing 0.5% Tween 20 (PBS-T) and 1% BSA for 30 min at room temperature to block nonspecific ligands. After washing, 100 pl of test serum, diluted 1:lO and l:lOO, and 2:l concentrated CSF using Minicon (Amicon, Beverly, MA, USA) and non-concentrated CSF (for analysis a titration curve - see below), were added in duplicate to antigen coated wells. After 45 min incubation with rocking at room temperature, the plates were washed with PBS-T. Then 100 pl of peroxidase conjugated goat anti-human IgG or IgM antibody (Sigma, St. Louis, MO, USA) was added to each well at a I:500 and 1:lOOO dilution respectively, in BPS-T containing 0.1% BSA. The plates were incubated for an additional 45 min and, after washing, 100 ~1 ATBS [citrate phospate buffer (0.1 mol/l, ph 4.2) 0.35 mg/ml 2,2’-azinodi (3-ethylbenzothiazoline-6-sulfonic acid, 0.003% H,O,] was added to the wells. After incubation for 30 min, absorbance was determined at 405 nm on a Stat-Fax 2100 microplate reader (Awareness Technology Inc., Palm City, FL, USA). Antigen and conjugate dilutions, incubation times and the blocking needed to reduce background were determined in preliminary experiments. Background control wells and control sera (positive and negative determined in preliminary experiments) were included in each plate. The samples marked for examination were applied in serial dilutions (normal serum, dilution l:lO, 1: 100, 1:500, 1:5000). After analysis of the titration curve for 5 control and 5 patient sera tested by ELISA against hsp 65 and 70, all sera were assayed at a dilution of 1: 10 and 1: 100 (according to the optimal part of the curves). Data are shown for dilution 1: 10. Results are expressed as a ratio of the absorbance value of the test serum or CSF to the mean absorbance value of three negative control sera (ELISA ratio). The analysis of the titration curve (4-fold concentration, 2-fold concentration, normal concentration, dilution 1:2) for 5 control and 5 patient CSF tested by ELISA against hsp 65 and 70 was performed, and subsequently all CSF were assayed at 2-fold concentration and normal concentration (according to the optimal part of the curves). Concentration of CSF was prepared in accordance with the general formula established by Reiber and Felgenhauer (1987). The calculations were made for a 2-fold concentration and normal concentration. Ratio value greater than 2 SD above the mean of the ratio obtained in sera from DB was considered positive in sera; a ratio value greater than 2 SD above the mean of ratio obtained in CSF from OND young was considered positive in CSF.
68
U. Finer et al. /Journal
of the Neurological Sciences 139 (1996) 66- 70
3. Statistical analysis Since data were not distributed normally, nonparametric statistics were used. Differences between the three groups were tested by one-factor analysis of variance (ANOVA procedure, Kruskal-Wallis test). Differences between pairs of groups were tested by Mann-Whitney’s U-test. Differences in incidence of positive antibodies in the patient group were tested by chi-square test (Kendall’s Tau B). Correlation between ages of patients and ratios of ELISA was proven by Spearman’s rank order test. P value of 0.05 was chosen as the level of significance.
4. Results Details of the laboratory findings are presented in Tables 1 and 2 and Figs. 1 and 2. In regard to the sera of patients, when tested against hsp 65, the IgM ELISA ratio of patients with PD (2.4 & 1.6, median 2.3) was greater ( p < 0.005) than that of sera from OND young (1.4 k 0.5, median 1.4) OND old (1.2 f 0.3, median 1. l), and DB (1.3 f 0.5, median 2.3) (Fig. 1). When tested against hsp 70, only the IgG ELISA ratio from patients with PD (1.3 -t 0.3, median 1.1) was greater ( p < 0.005) than that of sera from DB (0.9 * 0.3, median 0.9) (Fig. 2). Similarly, incidence of positive antibodies (determined by the cut-off, i.e. 2 SD above the mean value for DB) was higher in PD patients than in other studied groups; this was observed for IgM antibodies (for hsp 65, p < 0.0001) (Fig. 1) and for IgG antibodies (for hsp 70, p < 0.05) (Fig. 2). Due to the insufficient amount of CSF, the ELISA test
Table 2 CSF IgG and IgM antibody titres to hsp 70 in patients with Parkinson’s disease (PD) and in patients with other neurological diseases (OND young, OND old). The data are expressed as the ELISA ratio and incidence of positive antibodies Patient groups
kG
I@
2.fold cont. normal cont. 2-fold cont. normal cont. PD Mean (SD) Median No pas/No exam OND young Mean (SD) Median Mean + 2SD Upper quartile No pas/No exam OND old Mean (SD) Median No pas/No exam
2.7 (1.31 a~’ 1.4 (0.51= 2.8 1.2 2/S d 8/20 ’
1.1 (0.4) 1.0 O/4
1.3 (0.3) 1.3 l/11
1.4 (0.91 1.4 3.2 1.6 O/ 15
1.o (0.3) 1.2 1.6 1.6 l/15
1.1 (0.3) 1.2 1.7 1.4 o/15
1.1 (0.41 1.o 1.9 1.3 l/15
1.3 (0.4) 1.4 O/ 11
1.2 (0.5) 1.3 2/13
1.0 (0.2) 1.0 O/9
1.o (0.3) 1.0 o/10
Mann-Whitney’s U test, ’ PD vs. OND young, p < 0.05, b PD vs. OND old, p < 0.02. Kruskall-Wallis test for three groups, ’ p < 0.02. Chi-square test (Kendall’s Tau B), d p < 0.05, ’ p = 0.06.
could not be performed in all patients for 2-fold concentration. When tested against hsp 65, only the IgG ELISA ratio of CSF from patients with PD (1.7 & 0.6, median 1.4) was greater ( p < 0.02) than that of CSF from OND old (1.0 k 0.4, median 1.O) (for a 2-fold concentration) (Table 1). When tested against hsp 70 the IgG ELISA ratio of CSF from patients- with PD (2.7 $- 1.3, median 2.8) was
HSP 65 serum
Table 1 CSF IgG and IgM antibody tines to hsp 65 in patients with Parkinson’s disease (PD) and in patients with other neurological diseases (OND young, OND old). The data are expressed as the ELISA ratio and incidence of positive antibodies Patient groups
PD Mean (SD) Median No pas/No exam OND young Mean (SD) Median Mean + 2SD Upper quartile No pas/No exam OND old Mean (SD) Median No pas/No exam
5k
4 0 ‘i;
a
33
w kM 2-fold cont. normal cont. 2-fold cont. normal cont. 1.7 (0.6) ab 1.3 (0.4) 1.4 1.2 2/8 O/16
1.1 (0.7) 1.0 O/4
1.4 (0.4) 1.3 o/14
1.5 (0.3) 1.4 2.1 1.6 l/15
1.4 (0.4) 1.4 2.2 1.6 o/15
1.3 (0.4) 1.2 2.1 1.4 2/15
I .2 (0.4) 1.2 2.0 1.3 l/15
1.o (0.4) 1.0 O/ 11
1.1 (0.5) 1.2 o/12
1.1 (0.4) 1.1 o/10
1.2 (0.31 1.1 o/10
Mann-Whitney’s U test, a PD vs.OND old p < 0.02. Kruskall-Wallis test for three groups, b p < 0.005.
W
PD OND OND DB young old 2122 l/l5 O/l3 2117
w
PD OND OND DE young old lOl22 l/l5 o/13 0117
Fig. 1. Serum IgG and IgM antibody titres to hsp 65 in patients with Parkinson’s disease (PD), in patients with other neurological diseases (OND young, OND old) and donors of blood (DB). The data are expressed as the ELISA ratio. Columns represent the ELISA ratio (mean values and standard deviations). Incidence of positive antibodies is indicated separately. For IgM: Mann-Whitney’s U test, PD vs. OND old, p < 0.005 PD vs. DB, p < 0.005 Kruskal-Wallis test for four groups, p < 0.05 chi-square test (Kendall’s Tau B), p < 0.0001.
U. Fiszer et al. /Journal
of the Neurological Sciences 139 (1996) 66-70
5,
4I
0 bG PD OND OND DB young old 6121 o/15 l/l3 l/l?
w.4
PD OND OND DB young old o/21
o/15
o/13
Ill?
Fig. 2. Serum IgG and IgM antibody titres to hsp 70 in patients with Parkinson’s disease (PD), in patients with other neurological diseases (OND young, OND old) and donors of blood (DB). The data are expressed as the ELISA ratio. Columns represent the ELISA ratio (mean values and standard deviations). Incidence of positive antibodies is indicated separately. For IgG: Mann-Whitney’s U test, PD vs. DB, p < 0.005 chi-square test (Kendall’s Tau B), p < 0.05.
greater ( p < 0.05) than that of CSF from OND young (1.4 f 0.9, median 1.4) as well as OND old (1.3 + 0.4, median 1.4) (for a 2-fold concentration). Moreover, the IgG ELISA ratio of CSF from patients with PD (1.4 Ifr.0.5, median 1.2) was greater than that of CSF from OND young (1.0 + 0.3, median 1.2) for normal concentration (p < 0.05) (Table 2). The incidence of positive antibodies (determined by the cut-off, i.e. 2 SD above the mean value of OND young) was higher ( p < 0.05) in PD patients than in other studied groups; this was observed only for IgG antibodies (for hsp 70) in 2-fold concentrated CSF (Table 2). There was no correlation between patient’s age and Elisa ratio tested by Spearman’s rank order test. Also, there were no significant differences in ELISA ratios of IgG and IgM (to hsp 65 and hsp 70) in blood and CSF between control groups (OND young and OND old).
5. Discussion The present study showed that the mean ELISA ratio of CSF from patients with PD was significantly greater than that of CSF from patients with OND old (tested against IgG anti-hsp 65 and anti-hsp 70) and OND young (tested against IgG anti-hsp 70). There were no significant differences between mean ELISA ratio of sera from patients with PD, OND old and OND young (tested against IgG anti-hsp 65 and IgG anti-hsp 70). Increased IgG immunity in CSF to hsps (particularly to hsp 70) occurs in some patients with PD. Results of the present study confirm suggestion made by Gao et al. (1994). We also found
69
increased IgM immunity to hsp 65 in serum compared to OND old and DB. Those immune abnormalities in PD are not to be attributed to the aging process. The IgG response to hsps may reflect cross reactivity to proteins released by a wide variety of bacteria (possibly from breakdown in the gut) or recognition of other immunodominant antigens with high levels of cross reactivity. Increased hsps expression, which is induced by stress, provides cells with protection against environmental insults. Alternatively, the antibodies may be present as a consequence of prior infections. An increased level of antibodies against hsps has been found in several diseases and it is acknowledged that hsps are involved in pathogenesis of different neurological conditions. Selmaj et al. (1991) found that hsp 65 positive oligodendrocytes colocalized with y6+ T cells in multiple sclerosis plaques. Hamos et al. (1991) showed that hsp 72, which was present at very low levels in control brains, increased dramatically in Alzheimer’s disease patients, and was localized exlusively in neuritic plaques and neurofibrillary tangles. The authors suggested that ‘hsps participation in the pathological process may be protective, a function identified for hsps in several systems, or they may facilitate proteolysis, directing abnormal proteins into a degradative pathway’. Hsps are produced in response to many common stressors. The expression of some hsps, especially those in the 70 to 80 kDa range, is increased in response to a variety of known insults including hyperthermia, ischemia and exposure to free radicals (Thomas et al., 1981, Burdon, 1986). Carr and Farbman (1993) reported that administration of ketamine to rats leads to transient enhancement of immunoreactivity to monoclonal antibodies against two stress proteins, ubiquitin and human 70 kDa hsp in rat olfactory mucosa. The Lewy bodies that occur in idiopathic Parkinson’s disease are composed principally of neurofilament proteins, and stain heavily with monoclonal antibodies against ubiquitin (Jellinger, 1989). The 70 kDa hsp family is known to be an important participant in several intracellular events, including protein folding and trafficking. It binds to many, if not all proteins during their translation and maintains its association with some protein complexes as a subunit. It is assumed that transcriptonal factor for the hsp 72 gene is propably a ‘zinc finger’. The experiments indicated that the gene expression for the ‘zinc finger’ precedes induction of hsp 72 messenger RNA (Abe et al., 1991). Chopp (1993) suggested that there is a provocative coupling between growth factors and hsps, because the ‘zinc finger’ is important in cell growth regulation and differentation, and its transcription factor can be induced by nerve growth factors. Hsp 70 plays a role in antigen processing (Lakey et al., 1987, Vanbuskirk et al., 1991). The results of studies of Manara et al. (1993) performed on mice show a potential
70
U. Fiszer et al. / Journal
of the Neurological
role of a protein related to human hsp 70 kDa family in antigen processing and/or presentation. All of PD patients showed numerous HLA-DR positive microglia surrounding dead and dying neurons (McGeer et al., 1988) and, probably, such phenomenon is linked with an increased hsp 70 expression in that disease. Still, it remains unknown whether there is a direct cause of the elevated yS+ T cell population found in PD (Fiszer et al., 1994b). There was no correlation between the percentages of yS+ T cells in CSF and ELISA ratios of IgG to hsp 70 in PD patients. A further study on hsps expression and y6+ T cell with use of immunohistochemistry of frozen autopsy material from brain of PD patients, especially around areas such as the substantia nigra, and in association with Lewy bodies, could allow: for (i) an exact localization of such elements in tissue, and (ii) an assessment of their role in pathogenesis of PD.
Acknowledgements This investigation received financial support from the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR) and State Committee for Scientific Research in Poland.
References Abe, K., Kawagoe Jun-Ichi, &to, S., Sahara, M. and Kogure, K. (1991) Induction of the ‘zinc finger’ gene after transient focal ischemia in rat cerebral cortex. Neurosci. Lett., 123: 248-250. Burden, R.H. (1986) Heat shock and the heat shock proteins. Biochem. J., 240: 313-324. Chopp, M. (1993) The roles of heat shock proteins and immediate early genes in central nervous system normal function and pathology. Curr. Opin. Neural. Neurosurg., 6: 6-10. Craig, E.A., Gambill, B.B. and Nelson, R.J. (1993) Heat shock proteins: molecular chaperones of protein biogenesis. Microbial. Rev., 57: 402-414. Carr, V.M. and Farbman, A.I. (1993) Effect of ketamine on stress protein immunoreactivities in rat olfactory mucosa. NeuroReport, 5:, 197-200. Danieli, M.G., Candela, M., Ricciatti, A.M., Reginelli, R., Danieli, G., Cohen, I.R. and Gabrielli, A. (1992) Antibodies to Mycobacterial 65 kDa heat shock protein in systemic sclerosis. J. Autoimmunity, 5: 443-452. Gao, Y.L., Raine, C.S. and Brosnan, C. (1994) Humoral response to hsp 65 in multiple sclerosis and other neurological conditions. Neurology, 44: 941-946. Hamos, J.E., Oblas, B., Pulaski-Sale, B.S., Welch, W.J., Bole, D.G. and Drachman, D.A. (1991) Expression of heat shock proteins in Alzheimer’s disease. Neurology, 41: 345-350.
Sciences
139 (1996)
66-70
Fiszer, U., Mix, E., Fredrikson, S., Kostulas, V. and Link, H. (1994al Parkinson’s disease and immunological abnormalities: increase of HLA-DR expression on monocytes in cerebrospinal fluid and of CD45RO+ T cells in peripheral blood. Acta Neural. Stand., 90: 160- 166. Fiszer, U., Mix, E., Fredrikson, S., Kostulas, V., Olsson, T. and Link, H. (1994bl y6+ T cells are increased in patients with Parkinson’s disease. J. Neural. Sci., 121: 39-45. Freedman, M.S., Buu, N.N., Ruijs, T.C.J., Wiliams, K. and Ante], J.P. (1992) Differential expression of heat shock proteins by human glial cells. J. Neuroimmunol., 41: 231-238. Jellinger, K. (1989) Cytoskeletal pathology in parkinsonism and aging brain. In: P.B. Caine, G. Comi, D. Crippo, E. Horowski and M. Traboucchi (Eds.), Parkinsonism and Aging, Raven Press, New York, NY, pp. 35-56. Kaufmann, S.H.E. (1990) Heat shock proteins and the immune response. Immunol. Today, 11: 129-136. Lakey, E.K., Margoliash, E. and Pierce, S.K. (1987) Identification of a peptide binding protein that plays a role in antigen presentation, Proc. Nat]. Acad. Sci. USA, 84: 1659-1663. Manara, G.C., Sansoni, P., Badiali De Giorgi, L., Gallinella, G., Ferrari, C., Brianti, V., Fagnoni, F.F., Ruegg, C.L, De Panfilis, G. and Pasquinelli, G. (1993) New insights suggesting a possible role of a heat shock protein 70 kD family related protein in antigen processing/ presentation phenomenon in humans, Blood, 82: 2865-287 1. Marttila, R.J., Eskola, J., Plivarinta, M. and Rinne, U.K. (1984) Immune functions in Parkinson’s disease. Adv. Neural., 40: 315-323. McGeer, P.L., Itagaki, S., Boyes, B.E. and McGeer, E.G. (1988) Reactive microglia are positive for HLA-DR in the substantia nigra of Parkinson’s and Alzheimer’s disease brains, Neurology, 38: 1285-1291. O’Brien, R.L. and Born, W. (1991) Heat shock proteins as antigens for yS T cells. Semin. Immunol., 3: 81-87. Pouplard, A., Emile, J., Pouplard, F. and Hurez, D. (1979) Parkinsonism and autoimmunity: antibody against human sympathetic ganglion cells in Parkinson’s disease. Adv. Neural., 24: 32 l-326. Reiber, H. and Felgenhauer, K. (1987) Protein transfer at the blood cerebrospinal fluid barrier and the quantification of the humoral immune response within the central nervous system. Clin. Chim. Acta, 163: 319-328. Selmaj, K., Brosnan, C.F. and Raine, C.S. (1991) Colocalization of lymphocytes bearing y8 T-cell receptor and heat shock proteins hsp65+ oligodendrocytes in multiple sclerosis Proc. Nat]. Acad. Sci., 88: 6452-6456. Tibbling, G., Link, H. and Ghman, S. (1977) Princilpes of albumin and IgG analyses in neurological disorders, I. Establishment of reference values. Stand. J. Clin. Lab. Invest., 37: 385-390. Thomas, G.P., Welch, W.J., Mathews, M.B. and Feramisco J.R. (1981) Molecular and cellular effects of heat-shock and related treatments of mammalian tissue-culture cells. Cold. Spring Harbor Symp. Quant. Biol., 46: 985-996. Vanbuskirk, A.M., DeNagel, D.C., Guagliardi, L.D., Brodsky, F.M. and Pierce, S.K. (1991) Cellular and subcellular distribution of PBP72/74, a peptide-binding protein that plays a role in antigen processing. J. Immunol., 146: 500-506. Yamada, T., Akiyama M. and McGeer, P.L. (1990) Complement activated oligodendroglia: a new pathogenic entity identified by immunostaining with antibodies to human complement proteins C3d and C4d. Neurosci. Lett., 112: 161-166.
OF THE
NEUROLOGICAL SCIENCES ELSEVIER
Journal of the Neurological Sciences I39 (1996) 66-70
Humoral response to hsp 65 and hsp 70 in cerebrospinal fluid in Parkinson’s disease U. Fiszer a3* , S. Fredrikson b, A. Czlonkowska a h Dir,ision
’ lnsitute of Psychiatry and Neurology. Warsaw, Poland of Neurology. Karolinska Institute, Huddingr Hospital, Huddinge,
Sweden
Received 13 October 1994;revised 7 December 1995;accepted 18 December 1995
Abstract Parkinson’s disease (PD) is an age-related neurodegenerative movement disorder of unknown etiology. In PD immune abnormalities were reported, but the cause of such abnormalities has not been resolved. Recently, the increased proportion of yS+ T cells out of all T cells has been found in patients with PD. Heat shock proteins (hsps) could be targets for ~6~ T lymphocytes. We examined serum and CSF of patients with PD, age-matched patients with other non-inflammatory neurological diseases (OND old), young patients with other non-inflammatory neurological diseases (OND young), and donors of blood (DB). Antibodies were detected using the enzyme-linked immunosorbent assay. The plates were coated with recombinant mycobacterial hsp 65 and hsp 70. The present study showed that the mean ELISA ratio of CSF from patients with PD was significantly greater than that of CSF from patients with OND old (tested against IgG anti-hsp 65 and IgG anti-hsp 70) and OND young (tested against IgG anti-hsp 70). There was no difference between the mean ELISA ratio of sera from patients with PD, OND old and OND young (tested against IgG anti-hsp 65 and IgG anti-hsp 70). The significance of hsps immmunity is not completly clear. Increased hsps expression, which is induced by stress, provides cells with protection against the environmental insults. Alternatively, the antibodies may be present as a consequence of prior infections. Keywords;
Parkinson’s disease;Cerebrospinal fluid: Heat shock protein
1. Introduction Parkinson’s disease (PD) is an age-related neurodegenerative movement disorder of unknown etiology. The massive loss of pigmented dopaminergic nerve cells takes place in the affected brain areas. Lewy bodies are now considered a prerequisite for the definite diagnosis of PD, although their origin has remained unclear. Immune abnormalities have been broadly described, including the occurrence of autoantibodies against neuronal structures (Pouplard et al., 1979), reduced numbers of peripheral blood lymphocytes and reduced mitogen responses (Marttila et al., 1984), and complement-activated oligodendroglial cells (Yamada et al., 1990). The HLA-DR method highlights phagocytic activity surrounding dead and dying neurons which is a prominent feature in idiopathic PD (McGeer et al., 1988). A higher HLA-DR expression on CSF mono-
* Corresponding author. II Department of Neurology, Institute of Psychiatry and Neurology, Al. Sobieskiego l/9, 02.957 Warsaw, Poland, Tel: (+ 48-22) 42-40-23, Fax: (+ 48.22) 42-40-23, ( + 48-2) 642-53-75. 0022-510X/96/$15.00 P/I SOO22-5
0 1996 Elsevier Science B.V. All rights reserved
10X(96)00002-0
cytes in comparison with blood monocytes has been found (Fiszer et al., 1994a). Moreover, the elevated proportion of y6+ T cells out of all T cells in CSF and blood has been found in PD (Fiszer et al., 1994b). Heat shock proteins (hsps) could be targets for y6+ T lymphocytes (O’Brien and Born, 1991). Hsps or stress proteins comprise about two dozen proteins belonging to several families (Kaufmann, 1990, Craig et al., 1993). The induction of hsps expression may serve several purposes at a cellular level. Little is currently known regarding the role of hsps in the central nervous system; here hsps could have a critical function (Chopp, 1993). Variable expression of hsps by human glial cells has been shown by Freedman et al. (1992). Gao et al. (1994) suggest that antibodies reactive with hsp 60/65 are present in CSF in a number of chronic neurodegenerative conditions (Alzheimer’s disease, amyatrophic lateral sclerosis, PD). Since the effective cause of immune abnormalities in these patients is not entirely clear, we attempted to identify changes in anti-hsp 65 and 70 immunity in serum and CSF
U. Fiszer et al. / Journal
of the
Neurological
of patients with PD and patients with other neurological diseases (OND young and OND old, in relation to the aging process) and healthy donors of blood (DB). 2. Materials
and methods
2.1. Patients We studied 22 patients (15 men, 7 women) with PD and without evidence of any other associated neurological disorders. The diagnosis was established on the basis of clinical examination (subjects presented with typical PD symptoms and signs). All patients received a computed tomographic scan. Their ages ranged from 47 to 83 years (mean 71 years). The duration of PD varied between 1 and 18 years (mean 5.4 years). Fourteen patients were treated with levodopa at the time of the present study. CSF was obtained from 20 patients. A mononuclear pleocytosis ( > 5 X lo6 cells/l) was found in one patient and elevated CSF/serum albumin ratio (Tibbling et al., 1977) reflecting blood-CSF barrier damage in 4 patients. The IgG index equal to (CSF/serum IgG)/(CSF/serum/albumin) (Tibbling et al., 1977) was normal ( < 0.7) in all patients, and none of them had oligoclonal IgG bands. Three groups of control patients were included in the study. The first control group consisted of 15 patients (8 women) with OND young. Their ages varied between 16 and 47 years (mean 31.5 years). Seven of these patients had epilepsy, 3 had neurosis, 3 had intervertebral disc disease, 1 patient had neuropathy, and 1 patient had headache. None of them revealed any signs or symptoms of other diseases, and they displayed normal routine CSF findings. The second control group consisted of 13 age-matched patients (8 women) with OND old. Their ages varied between 42 and 84 years (mean 56.2 years) (with no statistically significant difference between PD and OND old). Five of these patients had cerebrovascular disease, 2 had epilepsy, 2 had spinal cord tumor, 1 had neuropathy, 1 had hereditary ataxia, 1 had intervertebral disc disease, and 1 had headache. None of them revealed any signs or symptoms of other diseases, and they displayed normal routine CSF findings. The third control group consisted of 17 (2 women) healthy DB. Their ages varied between, 19 and 54 years (mean 34.5 years); blood of those persons was collected from the local blood bank. Serum and CSF was stored at - 70°C. 2.2. Assay of antibodies to hsp 65 and hsp 70 Antibodies were detected using an enzyme-linked immunoabsorbent assay (ELISA) according to the method described by Danieli et al. (1992) (the method was modified for the purposes of our work). 96-well microtitre plates (Costar, Cambridge, MA,
Sciences
139 (19961 66-70
61
USA) were coated with recombinant mycobacterial hsp 65 (Mycobacterium bovis BCG, batch MA-12A) and hsp 70 (Mycobacterium tuberculosum, batch MT70-6A, a gift from Dr. J.D.A. van Embden, National Institute of Public Health and Environmental Protection, Bilthoven, Netherlands), 1 kg/ml in PBS; 100 ~1 per well was added at 4°C overnight. The plates were coated with BPS containing 0.5% Tween 20 (PBS-T) and 1% BSA for 30 min at room temperature to block nonspecific ligands. After washing, 100 pl of test serum, diluted 1:lO and l:lOO, and 2:l concentrated CSF using Minicon (Amicon, Beverly, MA, USA) and non-concentrated CSF (for analysis a titration curve - see below), were added in duplicate to antigen coated wells. After 45 min incubation with rocking at room temperature, the plates were washed with PBS-T. Then 100 pl of peroxidase conjugated goat anti-human IgG or IgM antibody (Sigma, St. Louis, MO, USA) was added to each well at a I:500 and 1:lOOO dilution respectively, in BPS-T containing 0.1% BSA. The plates were incubated for an additional 45 min and, after washing, 100 ~1 ATBS [citrate phospate buffer (0.1 mol/l, ph 4.2) 0.35 mg/ml 2,2’-azinodi (3-ethylbenzothiazoline-6-sulfonic acid, 0.003% H,O,] was added to the wells. After incubation for 30 min, absorbance was determined at 405 nm on a Stat-Fax 2100 microplate reader (Awareness Technology Inc., Palm City, FL, USA). Antigen and conjugate dilutions, incubation times and the blocking needed to reduce background were determined in preliminary experiments. Background control wells and control sera (positive and negative determined in preliminary experiments) were included in each plate. The samples marked for examination were applied in serial dilutions (normal serum, dilution l:lO, 1: 100, 1:500, 1:5000). After analysis of the titration curve for 5 control and 5 patient sera tested by ELISA against hsp 65 and 70, all sera were assayed at a dilution of 1: 10 and 1: 100 (according to the optimal part of the curves). Data are shown for dilution 1: 10. Results are expressed as a ratio of the absorbance value of the test serum or CSF to the mean absorbance value of three negative control sera (ELISA ratio). The analysis of the titration curve (4-fold concentration, 2-fold concentration, normal concentration, dilution 1:2) for 5 control and 5 patient CSF tested by ELISA against hsp 65 and 70 was performed, and subsequently all CSF were assayed at 2-fold concentration and normal concentration (according to the optimal part of the curves). Concentration of CSF was prepared in accordance with the general formula established by Reiber and Felgenhauer (1987). The calculations were made for a 2-fold concentration and normal concentration. Ratio value greater than 2 SD above the mean of the ratio obtained in sera from DB was considered positive in sera; a ratio value greater than 2 SD above the mean of ratio obtained in CSF from OND young was considered positive in CSF.
68
U. Finer et al. /Journal
of the Neurological Sciences 139 (1996) 66- 70
3. Statistical analysis Since data were not distributed normally, nonparametric statistics were used. Differences between the three groups were tested by one-factor analysis of variance (ANOVA procedure, Kruskal-Wallis test). Differences between pairs of groups were tested by Mann-Whitney’s U-test. Differences in incidence of positive antibodies in the patient group were tested by chi-square test (Kendall’s Tau B). Correlation between ages of patients and ratios of ELISA was proven by Spearman’s rank order test. P value of 0.05 was chosen as the level of significance.
4. Results Details of the laboratory findings are presented in Tables 1 and 2 and Figs. 1 and 2. In regard to the sera of patients, when tested against hsp 65, the IgM ELISA ratio of patients with PD (2.4 & 1.6, median 2.3) was greater ( p < 0.005) than that of sera from OND young (1.4 k 0.5, median 1.4) OND old (1.2 f 0.3, median 1. l), and DB (1.3 f 0.5, median 2.3) (Fig. 1). When tested against hsp 70, only the IgG ELISA ratio from patients with PD (1.3 -t 0.3, median 1.1) was greater ( p < 0.005) than that of sera from DB (0.9 * 0.3, median 0.9) (Fig. 2). Similarly, incidence of positive antibodies (determined by the cut-off, i.e. 2 SD above the mean value for DB) was higher in PD patients than in other studied groups; this was observed for IgM antibodies (for hsp 65, p < 0.0001) (Fig. 1) and for IgG antibodies (for hsp 70, p < 0.05) (Fig. 2). Due to the insufficient amount of CSF, the ELISA test
Table 2 CSF IgG and IgM antibody titres to hsp 70 in patients with Parkinson’s disease (PD) and in patients with other neurological diseases (OND young, OND old). The data are expressed as the ELISA ratio and incidence of positive antibodies Patient groups
kG
I@
2.fold cont. normal cont. 2-fold cont. normal cont. PD Mean (SD) Median No pas/No exam OND young Mean (SD) Median Mean + 2SD Upper quartile No pas/No exam OND old Mean (SD) Median No pas/No exam
2.7 (1.31 a~’ 1.4 (0.51= 2.8 1.2 2/S d 8/20 ’
1.1 (0.4) 1.0 O/4
1.3 (0.3) 1.3 l/11
1.4 (0.91 1.4 3.2 1.6 O/ 15
1.o (0.3) 1.2 1.6 1.6 l/15
1.1 (0.3) 1.2 1.7 1.4 o/15
1.1 (0.41 1.o 1.9 1.3 l/15
1.3 (0.4) 1.4 O/ 11
1.2 (0.5) 1.3 2/13
1.0 (0.2) 1.0 O/9
1.o (0.3) 1.0 o/10
Mann-Whitney’s U test, ’ PD vs. OND young, p < 0.05, b PD vs. OND old, p < 0.02. Kruskall-Wallis test for three groups, ’ p < 0.02. Chi-square test (Kendall’s Tau B), d p < 0.05, ’ p = 0.06.
could not be performed in all patients for 2-fold concentration. When tested against hsp 65, only the IgG ELISA ratio of CSF from patients with PD (1.7 & 0.6, median 1.4) was greater ( p < 0.02) than that of CSF from OND old (1.0 k 0.4, median 1.O) (for a 2-fold concentration) (Table 1). When tested against hsp 70 the IgG ELISA ratio of CSF from patients- with PD (2.7 $- 1.3, median 2.8) was
HSP 65 serum
Table 1 CSF IgG and IgM antibody tines to hsp 65 in patients with Parkinson’s disease (PD) and in patients with other neurological diseases (OND young, OND old). The data are expressed as the ELISA ratio and incidence of positive antibodies Patient groups
PD Mean (SD) Median No pas/No exam OND young Mean (SD) Median Mean + 2SD Upper quartile No pas/No exam OND old Mean (SD) Median No pas/No exam
5k
4 0 ‘i;
a
33
w kM 2-fold cont. normal cont. 2-fold cont. normal cont. 1.7 (0.6) ab 1.3 (0.4) 1.4 1.2 2/8 O/16
1.1 (0.7) 1.0 O/4
1.4 (0.4) 1.3 o/14
1.5 (0.3) 1.4 2.1 1.6 l/15
1.4 (0.4) 1.4 2.2 1.6 o/15
1.3 (0.4) 1.2 2.1 1.4 2/15
I .2 (0.4) 1.2 2.0 1.3 l/15
1.o (0.4) 1.0 O/ 11
1.1 (0.5) 1.2 o/12
1.1 (0.4) 1.1 o/10
1.2 (0.31 1.1 o/10
Mann-Whitney’s U test, a PD vs.OND old p < 0.02. Kruskall-Wallis test for three groups, b p < 0.005.
W
PD OND OND DB young old 2122 l/l5 O/l3 2117
w
PD OND OND DE young old lOl22 l/l5 o/13 0117
Fig. 1. Serum IgG and IgM antibody titres to hsp 65 in patients with Parkinson’s disease (PD), in patients with other neurological diseases (OND young, OND old) and donors of blood (DB). The data are expressed as the ELISA ratio. Columns represent the ELISA ratio (mean values and standard deviations). Incidence of positive antibodies is indicated separately. For IgM: Mann-Whitney’s U test, PD vs. OND old, p < 0.005 PD vs. DB, p < 0.005 Kruskal-Wallis test for four groups, p < 0.05 chi-square test (Kendall’s Tau B), p < 0.0001.
U. Fiszer et al. /Journal
of the Neurological Sciences 139 (1996) 66-70
5,
4I
0 bG PD OND OND DB young old 6121 o/15 l/l3 l/l?
w.4
PD OND OND DB young old o/21
o/15
o/13
Ill?
Fig. 2. Serum IgG and IgM antibody titres to hsp 70 in patients with Parkinson’s disease (PD), in patients with other neurological diseases (OND young, OND old) and donors of blood (DB). The data are expressed as the ELISA ratio. Columns represent the ELISA ratio (mean values and standard deviations). Incidence of positive antibodies is indicated separately. For IgG: Mann-Whitney’s U test, PD vs. DB, p < 0.005 chi-square test (Kendall’s Tau B), p < 0.05.
greater ( p < 0.05) than that of CSF from OND young (1.4 f 0.9, median 1.4) as well as OND old (1.3 + 0.4, median 1.4) (for a 2-fold concentration). Moreover, the IgG ELISA ratio of CSF from patients with PD (1.4 Ifr.0.5, median 1.2) was greater than that of CSF from OND young (1.0 + 0.3, median 1.2) for normal concentration (p < 0.05) (Table 2). The incidence of positive antibodies (determined by the cut-off, i.e. 2 SD above the mean value of OND young) was higher ( p < 0.05) in PD patients than in other studied groups; this was observed only for IgG antibodies (for hsp 70) in 2-fold concentrated CSF (Table 2). There was no correlation between patient’s age and Elisa ratio tested by Spearman’s rank order test. Also, there were no significant differences in ELISA ratios of IgG and IgM (to hsp 65 and hsp 70) in blood and CSF between control groups (OND young and OND old).
5. Discussion The present study showed that the mean ELISA ratio of CSF from patients with PD was significantly greater than that of CSF from patients with OND old (tested against IgG anti-hsp 65 and anti-hsp 70) and OND young (tested against IgG anti-hsp 70). There were no significant differences between mean ELISA ratio of sera from patients with PD, OND old and OND young (tested against IgG anti-hsp 65 and IgG anti-hsp 70). Increased IgG immunity in CSF to hsps (particularly to hsp 70) occurs in some patients with PD. Results of the present study confirm suggestion made by Gao et al. (1994). We also found
69
increased IgM immunity to hsp 65 in serum compared to OND old and DB. Those immune abnormalities in PD are not to be attributed to the aging process. The IgG response to hsps may reflect cross reactivity to proteins released by a wide variety of bacteria (possibly from breakdown in the gut) or recognition of other immunodominant antigens with high levels of cross reactivity. Increased hsps expression, which is induced by stress, provides cells with protection against environmental insults. Alternatively, the antibodies may be present as a consequence of prior infections. An increased level of antibodies against hsps has been found in several diseases and it is acknowledged that hsps are involved in pathogenesis of different neurological conditions. Selmaj et al. (1991) found that hsp 65 positive oligodendrocytes colocalized with y6+ T cells in multiple sclerosis plaques. Hamos et al. (1991) showed that hsp 72, which was present at very low levels in control brains, increased dramatically in Alzheimer’s disease patients, and was localized exlusively in neuritic plaques and neurofibrillary tangles. The authors suggested that ‘hsps participation in the pathological process may be protective, a function identified for hsps in several systems, or they may facilitate proteolysis, directing abnormal proteins into a degradative pathway’. Hsps are produced in response to many common stressors. The expression of some hsps, especially those in the 70 to 80 kDa range, is increased in response to a variety of known insults including hyperthermia, ischemia and exposure to free radicals (Thomas et al., 1981, Burdon, 1986). Carr and Farbman (1993) reported that administration of ketamine to rats leads to transient enhancement of immunoreactivity to monoclonal antibodies against two stress proteins, ubiquitin and human 70 kDa hsp in rat olfactory mucosa. The Lewy bodies that occur in idiopathic Parkinson’s disease are composed principally of neurofilament proteins, and stain heavily with monoclonal antibodies against ubiquitin (Jellinger, 1989). The 70 kDa hsp family is known to be an important participant in several intracellular events, including protein folding and trafficking. It binds to many, if not all proteins during their translation and maintains its association with some protein complexes as a subunit. It is assumed that transcriptonal factor for the hsp 72 gene is propably a ‘zinc finger’. The experiments indicated that the gene expression for the ‘zinc finger’ precedes induction of hsp 72 messenger RNA (Abe et al., 1991). Chopp (1993) suggested that there is a provocative coupling between growth factors and hsps, because the ‘zinc finger’ is important in cell growth regulation and differentation, and its transcription factor can be induced by nerve growth factors. Hsp 70 plays a role in antigen processing (Lakey et al., 1987, Vanbuskirk et al., 1991). The results of studies of Manara et al. (1993) performed on mice show a potential
70
U. Fiszer et al. / Journal
of the Neurological
role of a protein related to human hsp 70 kDa family in antigen processing and/or presentation. All of PD patients showed numerous HLA-DR positive microglia surrounding dead and dying neurons (McGeer et al., 1988) and, probably, such phenomenon is linked with an increased hsp 70 expression in that disease. Still, it remains unknown whether there is a direct cause of the elevated yS+ T cell population found in PD (Fiszer et al., 1994b). There was no correlation between the percentages of yS+ T cells in CSF and ELISA ratios of IgG to hsp 70 in PD patients. A further study on hsps expression and y6+ T cell with use of immunohistochemistry of frozen autopsy material from brain of PD patients, especially around areas such as the substantia nigra, and in association with Lewy bodies, could allow: for (i) an exact localization of such elements in tissue, and (ii) an assessment of their role in pathogenesis of PD.
Acknowledgements This investigation received financial support from the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR) and State Committee for Scientific Research in Poland.
References Abe, K., Kawagoe Jun-Ichi, &to, S., Sahara, M. and Kogure, K. (1991) Induction of the ‘zinc finger’ gene after transient focal ischemia in rat cerebral cortex. Neurosci. Lett., 123: 248-250. Burden, R.H. (1986) Heat shock and the heat shock proteins. Biochem. J., 240: 313-324. Chopp, M. (1993) The roles of heat shock proteins and immediate early genes in central nervous system normal function and pathology. Curr. Opin. Neural. Neurosurg., 6: 6-10. Craig, E.A., Gambill, B.B. and Nelson, R.J. (1993) Heat shock proteins: molecular chaperones of protein biogenesis. Microbial. Rev., 57: 402-414. Carr, V.M. and Farbman, A.I. (1993) Effect of ketamine on stress protein immunoreactivities in rat olfactory mucosa. NeuroReport, 5:, 197-200. Danieli, M.G., Candela, M., Ricciatti, A.M., Reginelli, R., Danieli, G., Cohen, I.R. and Gabrielli, A. (1992) Antibodies to Mycobacterial 65 kDa heat shock protein in systemic sclerosis. J. Autoimmunity, 5: 443-452. Gao, Y.L., Raine, C.S. and Brosnan, C. (1994) Humoral response to hsp 65 in multiple sclerosis and other neurological conditions. Neurology, 44: 941-946. Hamos, J.E., Oblas, B., Pulaski-Sale, B.S., Welch, W.J., Bole, D.G. and Drachman, D.A. (1991) Expression of heat shock proteins in Alzheimer’s disease. Neurology, 41: 345-350.
Sciences
139 (1996)
66-70
Fiszer, U., Mix, E., Fredrikson, S., Kostulas, V. and Link, H. (1994al Parkinson’s disease and immunological abnormalities: increase of HLA-DR expression on monocytes in cerebrospinal fluid and of CD45RO+ T cells in peripheral blood. Acta Neural. Stand., 90: 160- 166. Fiszer, U., Mix, E., Fredrikson, S., Kostulas, V., Olsson, T. and Link, H. (1994bl y6+ T cells are increased in patients with Parkinson’s disease. J. Neural. Sci., 121: 39-45. Freedman, M.S., Buu, N.N., Ruijs, T.C.J., Wiliams, K. and Ante], J.P. (1992) Differential expression of heat shock proteins by human glial cells. J. Neuroimmunol., 41: 231-238. Jellinger, K. (1989) Cytoskeletal pathology in parkinsonism and aging brain. In: P.B. Caine, G. Comi, D. Crippo, E. Horowski and M. Traboucchi (Eds.), Parkinsonism and Aging, Raven Press, New York, NY, pp. 35-56. Kaufmann, S.H.E. (1990) Heat shock proteins and the immune response. Immunol. Today, 11: 129-136. Lakey, E.K., Margoliash, E. and Pierce, S.K. (1987) Identification of a peptide binding protein that plays a role in antigen presentation, Proc. Nat]. Acad. Sci. USA, 84: 1659-1663. Manara, G.C., Sansoni, P., Badiali De Giorgi, L., Gallinella, G., Ferrari, C., Brianti, V., Fagnoni, F.F., Ruegg, C.L, De Panfilis, G. and Pasquinelli, G. (1993) New insights suggesting a possible role of a heat shock protein 70 kD family related protein in antigen processing/ presentation phenomenon in humans, Blood, 82: 2865-287 1. Marttila, R.J., Eskola, J., Plivarinta, M. and Rinne, U.K. (1984) Immune functions in Parkinson’s disease. Adv. Neural., 40: 315-323. McGeer, P.L., Itagaki, S., Boyes, B.E. and McGeer, E.G. (1988) Reactive microglia are positive for HLA-DR in the substantia nigra of Parkinson’s and Alzheimer’s disease brains, Neurology, 38: 1285-1291. O’Brien, R.L. and Born, W. (1991) Heat shock proteins as antigens for yS T cells. Semin. Immunol., 3: 81-87. Pouplard, A., Emile, J., Pouplard, F. and Hurez, D. (1979) Parkinsonism and autoimmunity: antibody against human sympathetic ganglion cells in Parkinson’s disease. Adv. Neural., 24: 32 l-326. Reiber, H. and Felgenhauer, K. (1987) Protein transfer at the blood cerebrospinal fluid barrier and the quantification of the humoral immune response within the central nervous system. Clin. Chim. Acta, 163: 319-328. Selmaj, K., Brosnan, C.F. and Raine, C.S. (1991) Colocalization of lymphocytes bearing y8 T-cell receptor and heat shock proteins hsp65+ oligodendrocytes in multiple sclerosis Proc. Nat]. Acad. Sci., 88: 6452-6456. Tibbling, G., Link, H. and Ghman, S. (1977) Princilpes of albumin and IgG analyses in neurological disorders, I. Establishment of reference values. Stand. J. Clin. Lab. Invest., 37: 385-390. Thomas, G.P., Welch, W.J., Mathews, M.B. and Feramisco J.R. (1981) Molecular and cellular effects of heat-shock and related treatments of mammalian tissue-culture cells. Cold. Spring Harbor Symp. Quant. Biol., 46: 985-996. Vanbuskirk, A.M., DeNagel, D.C., Guagliardi, L.D., Brodsky, F.M. and Pierce, S.K. (1991) Cellular and subcellular distribution of PBP72/74, a peptide-binding protein that plays a role in antigen processing. J. Immunol., 146: 500-506. Yamada, T., Akiyama M. and McGeer, P.L. (1990) Complement activated oligodendroglia: a new pathogenic entity identified by immunostaining with antibodies to human complement proteins C3d and C4d. Neurosci. Lett., 112: 161-166.