вход по аккаунту


Borrelia burgdorferiЧspecific and autoreactive T-cell lines from cerebrospinal fluid in lyme radiculomyelitis.

код для вставкиСкачать
Bomeh bwghrferi-specific
and Autoreactive T-cell Lnes
from Cerebrospinal Fluid in
Lyme Radiculomyelitis
Roland Martin, MD,* Johannes Ortlauf, MSt,' Veronika Sticht-Groh, MD,t Ulrich Bogdahn, MD,'
Shraga F. Goldmann, MD,S and Hans Georg Mertens, MD'
In 3 patients with Lyme radiculomyelitis, cellular immune reactions of cerebrospinal fluid (CSF) lymphocytes were
analyzed. Phenotypic analysis of CSF cells demonstrated that the majority were T cells (CD3 +) of the helperlinducer
subset (CD4 + ). These T cells were directly expanded from the CSF by limiting dilution. A total of 505 T-cell lines
were tested for Bowelia bargdo~jie9-i(Bb)-specific proliferation and also partly tested for reactivity to a panel of central
and peripheral nervous system antigens. Proliferative assays revealed 33 of them to be Bb specific, 16 to be specific for
myelin basic protein, 16 to be specific for peripheral myelin, 1 to be specific for cardiolipin, and 2 to be specific for
galactocerebrosides. The antigen-specific proliferation was restricted by autologous human leukocyte antigen (HLA)
class 11molecules. The majority of CSF-derived T-cell lines stained positively for CD3, CD4, and HLA class I1 antigens
and negatively for CD8 (cytotoxiclsuppressor subset). One T-cell line provided help for the production of specific I g G
by autologous B cells and secreted gamma-interferon upon stimulation with Bb antigen in the presence of autologous
antigen-presenting cells. These data show that in patients with severe neurological manifestations of late Lyme disease,
not only Bb-specific T-cell lines but also T cells reactive to central or peripheral nervous system autoantigens can be
Martin R, Ortlauf J, Sticht-Groh V, Bogdahn U, Goldmann SF, Mertens HG. Bowelia burgdorferi-specific
and autoreactive T-cell lines from cerebrospinal fluid in Lyme radiculomyelitis.
Ann Neurol 1988;24:509-516
Lyme disease is caused by Borrelia burgdorferi (Bb) [l}
and transmitted by a group of closely related ticks
(Ixodes dummini, I . pacificus, or I . ricinus). Neurological
manifestations include aseptic meningitis, meningoradiculomyelitis, or even progressive encephalomyelitis and demyelinating encephalopathy E2-91. In addition to nonspecific alterations of immune functions
[10- 121, Bb-specific T lymphocytes have been found
in peripheral blood, joint fluid, and cerebrospinal fluid
KSF) C13, 141.
Prompted by recent studies of a patient with chronic
rubella panencephalitis [lS], we addressed the question of whether and to what extent Bb-specific T-cell
lines (TCLs) can be found in the CSF Of Patients suffering from LYme meningoradiculomyelitis. Beyond that,
it W V ~of
S particular interest whether autoreactive T-cell
reactions could be induced by B. burgdorferi. For this
purpose, CSF T cells were directly expanded under
From the *Department of Neurology and the ?Institute of Hygiene
and Microbiology, University of Wiirzburg, Wiirzburg, FRG, and
lmmunologyy Dm
Bank, Oberer Eselsberg, Ulm, FRG.
limiting dilution conditions and analyzed for surface
marker expression and antigen-specific proliferation in
response to Bb antigen and to peripheral and central
nervous system (PNS) (CNS) autoantigens. Using
these techniques, it could be demonstrated that in patients with Lyme disease showing signs of CNS involvement, Bb-specific T cells are present in the CSF,
as are T cells reactive to PNS and CNS autoantigens.
Our data indicate a possible link between chronic bacterial infections and demyelinating diseases of humans.
and Methods
The cellular immune responses of peripheral blood lymphocytes (PBLs) and CSF lymphocytes from 3 patients suffering
from chronic Lyme radiculomyelitis were investigated. Relevant clinical data are given in Table 1. Serological and important CSF findings are summarized in Table 2. The same in-
Received Oct 2, 1987, and in revised form Dec 31, 1987, and Apr
13, 1988. Accepted for publication Apr 13, 1988.
Address correspondence to Dr Martin, Department of Neurology,
University of Wiirzburg, Josef-Schneider-Str 11, 8700 Wiinburg,
Copyright 0 1988 b y the American Neurological Association 509
Table 1. Anamnestic and Clinical Data on 3 Patients with
Chronic Lyme Radiculomyelitis and 1 Control Patient with a
Lymphocytic Meningitis of Unknown Origin
Table 2. Laboratory Findings (CSF and Serum) in 3 Patients
with Chronic Lyme Radicalomyelitis and 1 Control Patient
with a Lymphocytic Meningitis of- Unknown Origin
Age (yr)
ECM/tick bite
Treatment coursesb
Duration of
symptoms (mo)
"Pretreatment with doxycycline 2 x 100 mg/day for 21 days.
bTreatment with penicillin G 20 x lo6 unidday for 14 days.
ECM = erythema chronicum migrans; ND = not done.
vestigations were performed with PBLs and CSF cells from 1
patient suffering from a lyrnphocytic meningitis of unknown
origin for control purposes (see Tables 1 and 2).
Lymphocytes were isolated from freshly drawn CSF as described previously 157, cryopreserved at 5 x lo4 cellslml in
50% fetal calf serum (FCS) (Boehringer, Mannheim, FRG)
and 10% dirnethylsulfoxide (Sigma, Taufkirchen, FRG), and
stored in liquid nitrogen until use. PBLs were drawn on
several dates and cryopreserved at 5 to 10 x lo6 cells/ml.
All blood and CSF samples were obtained with informed
consent of the patients. Human leukocyte antigen (HLA)-A,
-B, -C, -DR, and -DQ typing was performed by S. F. Goldmann, Department of Transplantation Immunology, Red
Cross Blood Service, Urn, FRG. The HLA types of the
patients were as follows:
Patient 1: A2lA2, Bw44/Bw61, Cw2/-, DRwlUDRw6,
DQw 1lDQw3
Patient 2: Al/A3, Bw57/-, Cw6l-, DR2/DR7, DQwll
Patient 3: A26/A29, B45IBw70, Cw2lCw6, DR.w 11/DRwl3,
Total protein
Albumin (mg/dl)
IgG (mg/dl)
IgA (mg/dl)
IgM (rng/dl)
IgG index"
I I T titer (IgG)
Total leukocytes
Plasma cells
Other cells
IgG (mg/dl)
Albumin (mgldl)
IFT titer (I&)
1,170 1,170
4,360 4,000
1:256 c l : 1 6
"Calculatedaccording to Delpech and Lichtblau [ 1 >a]; normal values
range from 0.3 to 0.7.
CSF = cerebrospinal fluid; Ig = immunoglobulin; IFI' =: immunofluorescence test.
protein (MBP) was prepared according to the procedure of
Banik and Davison [17] from white matter of a nonneurological patient and used at 5 pg/ml. Whole human peripheral myelin (PM) was prepared from sciatic nerves of
nonneurological patients by repeated centrifugation over sucrose gradients and delipidation in ethedethanol (4/3 viv)
before the material was added at 5 pg/ml. Galactocerebrosides (Gal/Cs) type I1 (Sigma, Taufkirchen, FRG) were
initially dissolved in 50% ethanol and then diluted in phosphate-buffered saline until a final concentration of 5%
ethanol was reached, sonicated on ice, and added at 10 pg/
ml. Cardiolipin (Sigma) was used at 10 pg/ml.
The control patient was not HLA typed.
Phenotypic AnaIysis of Cells
Bb antigen was derived from Lyme-disease spirochetes
(strain W1) (Rocky Mountain Lab, Hamilton, MT, USA)
isolated from the CSF of a child suffering from Lyme meningitis by V. Sticht-Groh [16] and typed by T. Schwan [l6] at
the National Institutes of Health. The antigen preparation
was then washed and sonicated as described previously {13].
Antigen was used for proliferative assays at 5 pglrnl. Leptospiral antigen, Treponema phagedenis antigen (Bio Mkrieux,
Mercy Etoile, France), and tetanus toxoid (Behring Werke,
Marburg, FRG) were used at 5 pg/ml. Human myelin basic
510 Annals of Neurology
Vol 24
No 4
October 1988
Uncloned CSF cells and TCLs were stained by indirect irnmunofluorescence using monoclonal antibodies Leu 4 (pan,
T-cell marker; CD3), Leu 3 (helperlinducer T cells; CD4),
Leu 2 (cytotoxic/suppressor T cells; CD8; Beckton Dick-.
inson, Mountainview, CA), OKIa (HLA class 11 antigens),
OKM (monocytes; Ortho, Raritan, NJ), and anti-1g.M (B
cells; Dakopatts, Hamburg, FRG) and a fluorescein isothiocyanate-conjugated goat anti-mouse F(abI2 (Tago, Burlingame, CA) as a second reagent.
For fluorescent-activated flow cytometric analysis, at least
2 x lo* cells were stained by indirect immunofluorescence
as described above and quantitatively analyzed using an
Ortho Cytofluorograph (Ortho, Raritan, NJ) fluorescenceactivated cell sorter.
cells/ well
Generation of T-Lymphocyte Colonies from CSF
The expansion of T lymphocytes was performed using a
modification of the method described recently by Moretta
and colleagues [18]. Briefly, CSF cells were seeded at 0.3, 1,
3 , 10, 30, 60, and sometimes 180 cells/well in HEPESbuffered RPMI 1640 medium supplemented with 5% fetal
calf serum and 30% crude T-cell growth factor (prepared
from supernatants of 24-hour phytohemagglutinin-stimulated PBLs from young male blood donors) in the presence
of irradiated (5,000 rads), allogeneic feeder cells (PBLs adjusted to 1 x lo6 cellslml; HLA type: A2/A2, Bw44/
Bw62, Cw5/-, DRUDR9) in 96-well plates (Nunc, Roskilde, Denmark). Growing colonies were further expanded
in medium supplemented with 1% lectin-free interleukin 2
(IL 2; Lymphocult T-HP; Biotest, Dreieichenhain, FRG) until functional analysis became possible.
O O’i
Fig 1. Cloning efFciency of three representative cloning experiments using CSF T cells from Patient 1 (V),Patient 2 (a),and
Patient 3 (a)are shown.
Table 3 . Phenotypic Analysis of Mononuclear CSF Cells
Lymphproliferative Assay
In proliferative assays, TCLs or PBLs were screened for
antigen-specific proliferative activity as described previously
1191. Briefly, 1 to 2 x lo4 cells of the TCLs or 2 x lo5
PBLs were stimulated with specific antigen in the concentrations mentioned above in the presence of autologous or
allogeneic antigen-presenting cells (APCs) when the TCLs
were tested. Cultures were set up in complete medium
(RPMI 1640,200 pM L-glutamine, 10% FCS, and 50 pg/ml
gentamicin [Merck, Darmstadt, FRG]) supplemented with
10% human pool serum instead of fetal calf serum. After 72
(TCLs) or 144 (PBLs) hours of incubation (37°C; 5% CO?),
incorporation of 3H-thymidine (New England Nuclear,
Dreieich, FRG) was measured as described previously [19].
Immunofluorescence Staining
by Monoclonal Antibodies
(% positive cells)
Leu 4 (CD3)
Leu 3 (CD4)
Leu 2 (CD8)
OKIa (HLAclass 11)
OKM (monocytes)
Anti-IgM (B cells)
CSF = cerebrospinal fluid; HLA = human leukocyte antigen; Ig
immunoglobulin; NC = not counted.
Dot Blot Analysis
Helper function of TCLs was investigated with autologous B
cells (nonrosetting, or E-negative, cells, which were separated as described previously 120)) by dot blot analysis using
a modification of the technique described by Dorries and ter
Meulen [21). The intensity of staining was expressed semiquantitatively, ranging from zero to
The production of gamma-interferon by antigen-specific
TCLs was measured after 12 hours of incubation by the RXinterferon-gamma radioimmunoassay (Centocor, Malvern,
20 cells gave rise to a colony. A sample of representative experiments is given in Figure 1. Neither CSF
cells in the absence of irradiated feeder cells nor feeder
cells alone showed any outgrowth of colonies during a
period of 21 days. T-cell colonies from each cell concentration were further expanded in lectin-free IL 2
because only a minority of cells were expected to be
antigen specific in this chronic disease stage. Therefore, the term T-cell line instead of T-cell clone is used in
this report.
Efficiency o f the Limiting Dilution Procedure
As functional studies are impossible with the few cells
yielded by a diagnostic lumbar puncture, between 1
and 2 x lo4 CSF cells were seeded under limiting
dilution conditions using exogenous T-cell growth factor and allogeneic, irradiated (5,000 rads) feeder cells.
For Patient 1, several experiments were performed.
According to Poisson statistics, from 1 in 50 up to 1 in
Phenotypic Analysis of Mononuclear CSF Cells
and ofthe TCLJ
As shown in Table 3 , the majority of CSF mononuclear cells were T lymphocytes of the helpedinducer
phenotype (CD3
CD4 +). T lymphocytes of the
cytotoxic/suppressor subset (CD3 , CD8 + ) were
found to a minor extent. The control patient showed a
pleocytosis mainly composed of T cells equally distributed between both subsets. B cells (anti-IgM) and
Martin et al: Lyme Radiculomyelitis 511
Table 4. Proliferative Responses of Bb-Specific TCLr Isolated from the CSF of 3 Patients Suffering
from Chronic Lyme Radiculomyelitis
3H-Thymidine Incorporation (cpm) in the Presence of Autologous APs and:
Patient 1
Patient 2
Patient 3
Bb Antigen
IL 2
B lOi60
119,809 t 5,445
7,298 t 873
12,842 t 2,322
8,359 f 530
9,930 t 519
15,177 f 5,497
4,080 f 514
3,263 f 613
15,629 t 559
1,304 t 129
612 Ifr 49
1,110 f 445
1,348 f 210
369 74
888 f 23
862 t 137
247 f 86
344 t 40
816 2 186
438 t 14
255 t 8
721 2 46
746 Ifr 175
368 f 114
341 +. 359
1,152 f 351
556 2 65
984 f 125
886 f 298
462 t 32
281 t 41
1,081 t 275
657 t 243
438 r+_ 76
944 t 147
1,293 f 150
1,124 t 336
338 2 91
615 t 223
1,073 f 93
411 2 35
168 f 68
13,020 t 340
121,690 t 1,179
6,878 f 90
5,351 f 858
2,197 f 240
1,596 f 118
3,514 f 58
11,894 t 2,184
1,916 f 524
67,204 t 3,763
29,065 f 5,807
Bb = Bow& burgdaovfeovi;TCL = T-cell line; CSF = cerebrospinal fluid; APC = antigen-presenting cell; MBP
galactocerebroside; 1L 2 = interleukin 2; N D = not done; cpm = counts per minute.
monocytes (OKM) are rarely found during chronic
Lyme radiculomyelitis. The level of HLA class IIexpressing cells (OKIa) demonstrated that a considerable proportion of T cells were activated, as the antigen is normally found only on B cells, macrophages,
and dendritic cells, not on resting T cells. When the
CSF TCLs were stained for surface markers, it could
again be demonstrated that the majority of TCLs of all
3 patients and of the control patient were CD4 positive. Very few TCLs bore exclusively the CD8 antigen.
Half of the stained cell lines were mixtures of both
subsets, with a predominance of CD4-positive cells in
all of these TCLs (data not shown).
Proliferative Responses of CSF-Derived TCLr upon
Stimulation with Borrelia burgdorferi Antigen
Before testing the CSF-derived TCLs, the PBLs of the
3 patients and of the control patient were tested for
antigen-specific reactivity in proliferative assays. The
PBLs of all Lyme patients responded to Bb antigen.
The PBLs of Patient 3 also recognized GalCs (stimulation index 98.4; data not shown). Antigens closely related to B . burgdorferi, like those of Leptospira or T.
phagedenis, were not recognized (data not shown). The
control patient’s PBLs could not be stimulated by the
antigens used, although the responses to rnitogens
(phytohemagglutinin and pokeweed mitogen) were
normal (data not shown).
In six cloning experiments, a total of 505 CSFderived TCLs of the 3 patients were tested in proliferative assays upon stimulation with Bb antigen. In addition, 168 TCLs of the control patient were tested. In
the 3 patients suffering from chronic Lynie radiculomyelitis between 4.7 and 7.7% of tested TCLs
were specific for Bb antigen (stimulation index > 3).
Responses of sample Bb-specific cell lines for each patient are shown in Table 4. These TCLs did not react
to MBP, cardiolipin, and GalC (see Table 4 ) or to
512 Annals of Neurology Vol 24
No 4 October 1988
myelin basic protein; GalC =
bacterial control antigens (data not shown). Depending
on the state of activation of the cell line, 3H-thymidinc.
incorporation in the presence of IL 2 showed great
variability. Some of these TCLs could be kept in culture and were later analyzed in more detail (see below). Only 1 of 168 TCLs (0.6%) of the control patient weakly responded to Bb antigen, but could not
be kept in culture. However, 9.5% (16 of 168) of the
TCLs of the control patient responded to T .
Proliferutive Responses of CSF-Derived T C L upon
Stimulation with Autoantigens
All 505 TCLs were tested with MBP. Of these, 5.3%
(Patient I), 2.2% (Patient 2), 0% (Patient 3 ) , and 1%
(control patient) of tested TCLs responded only tci
stimulation with this white matter constituent (stimulation index > 3). The results of the proliferative testing
of some of these MBP-specific TCLs are given in
Table 5.
Reactivity to PM was tested only in Patient 2. Of
272 tested TCLs, 5.9% responded to stimulation with
PM (see Table 5).
In 673 TCLs of Patients 1 to 3 and the control
patient, only 3.1% (2 of 64) of Patient 3’s TCLs responded to GalC (data not shown). In this context it
should be noted that the PBLs of Patient 3 also proliferated following stimulation with GalC.
Of 169 TCLs from Patient 1 tested against cardiolipin, only 1 TCL (0.6%)gave a positive proliferative
response (data not shown).
It has already been stressed that most of the T-cell
populations were probably not clones, but oligoclonal
T-cell lines. We therefore expected to find a considerable number of TCLs to be reactive to more than one
antigen, but only a minority of all TCLs responded to
two, but not more, antigens. In Patient 1, 1.8% of 169
TCLs tested reacted to both Bb antigen and MBP. In
Table 5 . Proliferative Responses of a Sample of CSF-Derived TCLr upon Stimulation with Autoantigens
3H-Thymidine Incorporation (cprn) in t h e Presence of Autologous APCs and:
T C L No.
A 10130A
A 10/ I80B
D6/ 180A
2,546 ? 191
6,614 2 814
2,716 ? 445
4,777 2 301
858 2 205
IL 2
799 ? 232
3,113 220
2,378 2 236
3,187 2 64
5,865 413
798 +- 260
2,043 +- 10
816 k 269
1,516 +- 474
90 2 13
134 2 45
238 t 45
164 ? 21
357 k 28
160 2 23
193 19
21,985 ?
26,100 ?
8,543 ?
11,642 t
771 ?
2,604 ?
1,710 ?
3,161 ?
3,123 t
1,435 ?
CSF = cerebrospinal fluid; TCL = T-cell line; APC
interleukin 2; cpm = counts per minute.
antigen-presenting cell; MBP
* 14,295
rnyelin basic protein; PM
peripheral myelin; IL 2
Table 6. Proliferative Responses of TCLr Responding both upon Bb Antigen and MBP Stimulation (Part I )
and upon MBP and PM Stimulation (Part 11)
3H-Thvrnidine Incorporation (cprn) in the Presence of Autologous APCs and:
Patient No.
Bb Antigen
B 10/90
73,267 f 1,594
1,069 f 379
12,122 2 1,015
791 f 335
22,681 2 532
5,588 ? 413
1,069 ? 379
2,159 2 739
14,871 t 2,868
869 f 257
16,551 iz 3,060
2,903 f 550
13,432 i 5,127
1,862 f 515
869 ? 257
1,785 t 904
IL 2
54 i 38
128 t 65
5,405 t 1,084
175 2 168
708 i 208
101 f 37
128 ? 65
417 2 227
21,507 ? 5,757
4,411 t 138
16,209 ? 1,451
3,431 f 548
6,006 2 1,816
1,173 f 44
4,429 ? 138
1,522 f 384
576 ? 41
211 f 29
107 t 5
103 i 37
497 iz 33
2,207 ? 231
741 f 117
2,155 2 513
Patient 1
Patient 2
Patient 3
Patient 2
TCL = T-cell line; Bb = Bowelia burgdorferi; MBP
presenting cell; cprn = counts per minute.
1,934 ? 641
1,776 ? 775
1,784 f 629
1,835 t 580
4,153 ? 1,449
11,246 2 2,606
3,952 i 455
4,936 f 778
rnyelin basic protein; PM
Patient 2, 0.7% of 272 TCLs responded to Bb antigen
and MBP, and 2.2% to PM and MBP in parallel (see
Table 6).
Functional Analysis of a Bb-specific T C L
Using some of the previously tested TCLs, we asked
whether the proliferative response of specific TCLs
was restricted by HLA class I1 antigens. When the Bbspecific CD4-positive TCL D3/30A was tested using
autologous or allogeneic (HLA-DR type) APCs, it
could be demonstrated that the proliferative response
was restricted by HLA class I1 antigens (Fig 2). Because cells also responded to some extent to autologous APCs alone, possibly because of an autologous
peripheral myelin; IL 2
interleukin 2; APC = anrigen-
mixed lymphocyte reaction, the line was later recloned
with one cell per well. Ninety percent of the seeded
cells grew out. When these clones were tested again,
only 38% responded to Bb antigen. TCL D3/30A was
also tested for helper function. For that purpose, 2 X
lo4 cells of the line were cocultivated with or without
lo5 autologous E-negative cells, which were considered a source of B cells, in the presence or absence of
Bb antigen. After 3 days the supernatant was tested by
dot blot analysis. Thus, it could semiquantitatively be
shown that the cell line provided help to autologous B
cells in the production of specific IgG (Table 7). A
smaller amount of Bb-specific IgG was produced when
E-negative cells were stimulated by Bb antigen or
Martin et al: Lyme Radiculomyelitis 5 13
auto1 APC
Bb antigen
aufol APC
Table 8. Gamma-lnterfron Secretion of a Bb-Spec@
CD4-Positive T C L (D313OA) upon stimulation with Bb
Antigen in the Presence of Autologous or Akogeneic APCs
Bb antigen
cpm x 10-4
Fig 2. Human leukogte antigen ( H a )class I1 restriction of
the proliferative response upon stimuhtion with Borrelia burgdorferi (Bb) of a Bb-specific T-cell line (D3130A).HLA class I1
of autologous antigen-presenting cell (APC);DRwl 1lDRw6,
DQwllDQw3; HLA class 11 ofallogeneic APC: DR3IDR7,
DQw21DQw2. PHA = phytohemagglutinin.
Table 7. The Demonstration of Helper Function
to Autologous B Cells by a Bb-Spec$c TCL (D3130A)
B Cells
(1 x lo5
Bb Antigen
E-negative cells
E-negative cells
E-negative cells
- (PWM)
"Detected by dot blot analysis.
Bb = Borrefiu burgdovferi; Ig = immunoglobulin; TCL = 'T-cell line;
PBL = peripheral blood lymphocyte; PWM = pokeweed mitogen.
when the PBLs were cultivated in medium supplemented with pokeweed mitogen. In addition, TCL D3/
30A was tested for its ability to produce gammainterferon upon stimulation with Bb antigen in the
presence of autologous or allogeneic APCs. The results are shown in Table 8. TCL D3/30A secreted the
lymphokine only when Bb antigen was presented by
HLA class 11-compatible APCs, but not when allogeneic APCs or autologous APCs with T. phagedenis
antigen were used.
B. bzlrgdoyferi is able to spread to the CNS and can be
isolated from blood, skin, and CSF 122, 237. ,4 broad
spectrum of CNS disorders may be caused by the bacterium, ranging from meningitis to demyelinating encephalopathies 12-97. CSF findings include persistent
secretion of oligoclonal Bb-specific IgG [241 and lymphocytic pleocytosis. In the present investigation, we
addressed the question of whether and to what extent
antigen-specific T cells are found within the CSF of
patients with Lyme radiculomyelitis and whether B .
Annals of Neurology Vol 24
Autologous" Bb antigen
Autologous . . .
Allogeneicb Bb anrigen
Autologous Treponema phage-
. ..
denis antigen
IL 2
"HLA-DR type of autologous APCs: DRw6/DRw1 1.
bHLA-DR type of allogeneic APCs: DRlIDR9.
Bb = Borrefiu burgdorfer< TCL = T-cell line; APC
presenting cell; IL 2 = interleukin 2.
Szlrgdoyferz' is able to induce autoimmune phenomena.
(2 x lo4
No 4 October 1988
Pachner and colleagues 1137 and Sigal and colleagues
1141 have demonstrated that the bulk of mononuclear
cells isolated from CSF or joint fluid of Lyme patients
proliferate upon stimulation with B. bzlrgdorferi. However, more detailed analysis of the few cells isolated
from CSF becomes possible only when the recently
described cloning techniques are used 118, 19, 251.
We applied a modification of the limiting dilution technique originally described by Moretta and colleagues
[IS]. With 1 in 20 to 1 in 50 seeded cells giving rise to
a colony, the cloning efficiency in this group of patients
and in the control patient ranged in the same order of
magnitude as that seen in acute viral diseases of the
CNS 119, 251. T cells of the helperiinducer (CD4 + )
subset that are partly activated (HLA class I1 positive)
are predominant within the CSF and within a representative sample of CSF-derived TCLs. This was also the
case in a patient suffering from chronic rubella panencephalitis 1151, but was not the case in patients suffering from acute viral infections. B cells and monocytes
were only rarely found.
Before CSF-derived TCLs were investigated, PBLs
of all 3 patients were tested for proliferative responses
and were shown to respond upon stimulation with Bb
antigen but not upon exposure to leptospiral antigen
and T . phagedenis antigen. None of the autoantigens
were recognized except by Patient 3, whose PBLs
could be stimulated by cerebrosides. When the CSFderived TCLs were tested, a minority (4.7-7.7q1) of
505 TCLs were specific for B. burgdotjitrz. A similar
percentage of antigen-specific TCLs was also present in
a case of chronic rubella panencephalitis 1151, but in
acute mumps meningitis or measles encephalitis I19,
251, the majority of CSF T cells were antigen-specific
cytotoxic T cells. T&ng into account that only 1 in 50
to 1 in 20 seeded cells grew out, the real percentage of
antigen-specific T cells within the CSF is one to TWO
orders of magnitude lower than the percentage calculated from the proliferative assays.
As expected from. earlier reports [26), it could be
demonstrated that proliferative responses of CSFderived Bb-specific 'TCLs and also the response of a
TCL that had been established from the peripheral
blood (data not shown) are restricted by autologous
HLA class I1 molecules. In addition, the Bb-specific
TCL D3/30A provided helper functions to autologous
B cells producing Bb-specific JgG as shown by dot-blot
analysis. Cells with h.elper functions probably support
the persistent secretion of oligoclonal Bb-specific I&,
a phenomenon receritly described in the same group
of patients [24). It is not clear in chronic Lyme
radiculomyelitis and in other cases of chronic encephalomyelitis in animals and humans [27, 28) whether
continuing antibody secretion requires the persistence
of the antigen in the brain or is due to an ongoing
immune response.
The ability of these antigen-specific TCLs to secrete
gamma-interferon may be important for the induction
of HLA class I1 antigen expression on astrocytes that
can then serve as immunoregulatory cells within the
CNS 129-31).
Following a report noting that PBLs of patients suffering from measles, varicella, or rubella encephalitis
responded to stimulation with MBP [32), we addressed the question of whether B. bzlrgdorferi is able
to induce autoreactivity to brain or PNS antigens. In
contrast to the report .mentioned above [32), we tested
T cells directly expanded from the CSF in proliferative
assays using MBP, GtlCs, whole PM, and cardiolipin.
Although we did not find the same reactivity pattern in
all patients, between 0 and 5.3% of TCLs responded
to MBP, 5.9% of TCLs of Patient 2 reacted to PM,
and 3.1% of TCLs o i Patient 3 reacted to GalCs. A
minority of TCLs (0.4% of those of Patient 2 ) was
found to respond to ci-irdiolipin.
Comparable results have been shown in a case of
chronic rubella panencephalitis [l5), where reactivity
to MBP, proteolipid protein, cerebrosides, and actin
was seen. Similar investigations in multiple sclerosis
patients have provided conflicting results. Richert and
colleagues [33) found a mild degree of proliferation in
response to MBP and to measles virus. Their results
are difficult to compare with ours, because their CSF
cell lines were prestimulated by antigen, whereas the
TCLs described here were directly expanded from the
CSF and subsequently tested without prior antigenic
restimulation. Two cell lines isolated from CSF of multiple sclerosis patients by Bellamy and colleagues 134)
responded to purified gangliosides (GM1, G D l a ,
G D l b and GQlb). Johnson and colleagues [35] were
able to demonstrate reactivity to MBP and to myelinassociated glycoprotein, but not to proteolipid protein,
when PBLs of multiplt: sclerosis patients were tested.
When CSF cells and PBLs were analyzed by Hafler
and colleagues [36), responses to MBP or white matter were seen only after presensitization of PBLs. Fleischer and colleagues [37} applied the same technique as
described here and were not able to find any reactivity
to Epstein-Barr virus, parainfluenza type 3 virus,
mumps virus, measles virus, bovine MBP, or glycolipids. Regarding the autoreactive TCLs seen in Lyme
radiculomyelitis, we have not yet determined whether
these cells are of pathogenetic significance or represent
only an epiphenomenon. Sun and Wekerle [31) have
demonstrated that MBP-specific TCLs that are able to
mediate experimental autoimmune encephalomyelitis
exert cytotoxic function against astrocytes presenting
MBP in the context of appropriate HLA class I1 antigens. Whether the MBP- or Bb-specific TCLs described in this report also have the potential to lyse
cells presenting the antigen is currently being investigated. A small proportion of TCLs recognized either
Bb antigen and MBP (0.7--1.8%) or MBP and PM
(2.2%). The question of whether these responses are
due to cross reactions or are caused by TCLs containing two cell populations has to be clarified by using Tcell clones and well-defined B. burgdorferi antigens.
The finding of one Bb-specific TCL in the control patient can be interpreted only as a nonspecific activation
by lipopolysaccharides in the bacterial cell wall. The
positive response of two TCLs to MBP in this acute
meningitis a€unknown origin indicates that reactivity
to this white matter antigen is not a phenomenon
unique to certain diseases.
This report illustrates for the first time that, in good
correlation with clinical features, B . burgdorfez is able
to cause chronic nervous system involvement and to
induce autoimmune reactions directed against CNS
and PNS antigens. Compared to acute viral diseases,
antigen-specific T cells represent a minority of all T
cells but persist for long periods following an infecting
tick bite. During the disease course, autoreactive T
cells may become more prominent, although the
pathogenetic significance of these cells remains to be
determined. From these data, we conclude that cellular
immune reactions within the CNS play an important
role in chronic Lyme disease. Future studies will have
to clarify whether B. bzrrgdorfen' itself is able to cause
demyelination or whether myelin breakdown is mediated by the host's immune response to the pathogen.
This work was supported by grant Sti 6314-1 of the Deutsche
Leptospiral antigen was provided by D r Brem, Landesuntersuchungsamt fiir das Gesundheitswesen Sudbayern, Oberschleisslem, FRG.
D r U. Schwulera, Biotest Serum-Instimt, Offenbach, FRG, kindly
helped with the gamma-interferon determination.
We thank D. Drenkard and M. Rische for excellent technical assistance.
Martin et al: Lyme Radiculomyelitis
5 15
1. Burgdorfer W, Barbour AG, Hayes SF, et al. Lyme disease-a
tick-borne spirocherosis? Science 1982;216:1317-1319
2. Ackermann R, Gollmer E, Rehse-Kiipper B. Progressive Borrelien-Enzephalomyelitis. Dtsch Med Wochenschr 1985;110:
3. Broderick JP, Sandok BA, Mertz LE. Focal encephalitis in a
young woman 6 years after onset of Lyme disease: tertiary Lyme
disease? Mayo Clin Proc 1987;62:3 13-316
4. Klenk W, Heitmann R, Ackermann R. Rezidivierende
Erythema-Chronicum migrans Krankheit des Nervensystems:
Querschnittsmyelitis als Riickfall einer Meningopolyneuritis
Garin-Bujadoux-Bannwarth.Akt Neurol 1985;12:20--23
5. Pachner AR, Steere AC. The triad of neurologic manifestations
of Lyme disease: meningitis, cranial neuritis, and radiculoneuritis. Neurology 1985;35:47-53
6. Reik L, Smith L, Khan A, Nelson W. Demyelinating encephalopathy in Lymr disease. Neurology 1985;35:267-269
7 . Reik L, Steere AC, Bartenhagen N H , et al. Neurologic abnormalities of Lyme disease. Medicine 1979;58:281-294
8. Vallat JM, Hugon J, Lubeau M, et al. Tick-bite meningopolyneuritis: clinical, electrophysioiogical, and histologic findings in
10 cases. Neurology 1987;37:749-753
9. Weder B, Wiedersheim P, Matter L, et al. Chronic progressive
neurological involvement in Bow& burgu'urferz infection. J
Neurol 1987;234:40-43
10. Hardin JA, Walker LC, Steere AC, et al. Circulating immune
complexes in Lyme arthritis: detection by the "'I-C lq binding,
C l q solid phase and Raji cell assays. J Clin Invest 1979;63:468477
11. Moffat CM, Sigal LH, Steere AC, et al. Cellular immune findings in Lyme disease: correlation with serum IgM and disease
activity. Am J Med 1984;77:625-632
12. Steere AC, Hardin JA, Malawista SE. Erythema chronicum migrans and Lyme arthritis: cryoimmunoglobulins and clinical activity of skin and joints. Science 1977;196:1121-1122
13. Pachner AR, Steere AC, Sigal LH, Johnson CJ Antigenspecific proliferation of CSF lymphocytes in Lyme disease. Neurology 1985;35:1642- 1644
14. Sigal LH, Steere AC, Freeman DH, Dwyer JM. Proliferative
responses of mononuclear cells in Lyme disease. Arthritis
Rheum 1986;29:761-769
15. Martin R, Borkenstein M, Kreth HW. Virus-induced autoimmunity in progressive rubella panencephalitis. Eur J Pediatr
15a. Delpech B, Lichtblau E. Etude quantitative des immunoglobulines G et de l'albumine d u liquide CCphalorachidien. Clin
Chim Acta 1972;37:15-23
16. Huppenz H1. Stichr-Groh V, Schwan T. Borderline antibody
response in initial stages of lymphocytic meningitis docs not rule
out borreliosis. Lancet 1986;2:1468-1469
17. Banik NL, Davison AN. Isolation of purified basic protein from
human brain. J Neurochem 1973;21:489-494
18. Moretta A, Pantaleo G, Moretta L, et al. Direct demonstration
of the clonogenic potential of every human peripheral blood T
cell. J Exp Med 1983;157:743-754
19. Fleischer B, Kreth HW. Clonal expansion and functional analysis of virus-specific T lymphocytes from cerebrosplnal fluid in
measles encephalitis. Hum Immunol 1983;7:239-248
516 Annals of Neurology Vol 24
No 4
October 1988
20. Weiner MS, Bianco C, Nussenzweig V. Enhanced binding of
neuroaminidase-treated sheep erythrocytes to human T lymphocytes. Blood 1973;42:939-946
2 1. Dorries R, ter Meulen V. Detection and identification of virusspecific, oligoclonal IgG in unconcentrated cerebrospinal fluid
by immunoblot technique. J Neuroimmunol 1984;7:77-89
22. Benach JL, Bosler EM, Hanrahan JP,et al. Spirochetes isolated
from the blood of two patients with Lyme disease. N Engl J
Med 1983;308:740-742
23. Steere AC, Grodzicki RL, Kornblatt AN, et al. The spirochetal
etiology of Lyme disease. N Engl J Med 1983;308:733-740
24. Martin R, Martens U, Sticht-Groh V, et al. Persistent intrathecal
secretion of oligoclonal, Barrelid burgdurferi-specific IgG in
chronic meningoradiculomyelitis. J Neurol 1988;235:229-233
25. Fleischer B, Kreth HW. Clonal analysis of HLA-restricted cytotoxic T lymphocytes from cerebrospinal fluid in mumps meningitis. J Immunol 1983;130:2187-2190
26. Meuer SC, Schlossmann SF, Reinherz EL. Clonal analysis of
human cytotoxic T lymphocytes: 'T4 and T8 effector T cells
recognize products of different major histocompatibility regions. Proc Natl Acad Sci USA 1982;79:4395-4399
27. Dorries R, Watanabe R, Wege H , ter Meulen V. Analysis ofthe
intrathecal humoral immune response in Brown Norway (BN)
rats, infected with the murine corona-virus JHM. J Neuroirnmunol 1987;14:305-316
28. Vandvik B, Skddenberg B, Forsgren M, et al. Long-term pcrsistence of intrathecal virus-specific antibody responses h e r
herpes simplex encephalitis. J Neurol 1985;231:307-312
29. Wong GHW, Bartlett PF, Clark-Lewis I, et al. Interferongamma induces the expression of H-2 and Ia antigens on brain
cells. J Neuroimmunol 1985;7:2 5 5-2 7 8
30. Fontana A, Fierz W, Wekerle H . Astrocytes present myelin
basic protein to encephalitogenic T-cell lines. Nature 1984;
31. Sun D, Wekerle H. Ia-restricted encephalitogenic T lymphocytes mediating EAE lyse autoantigen-presenting astrocytes.
Nature 1986;320:70-72
32. Johnson RT, Griffin DE, Hirsch RL, et al. Measles encephalttis-clinical and immunological studies. N Engl J Med
33. Richert JR, McFarlin DE, Rose JW, et al. Expansion of antigen-.
specific T cells from cerebrospinal fluid of patients with multipl~
sclerosis. J Neuroimmunol 1983;5:317-324
34. Bellamy A, Davison AN, Feldmann M. Derivation of ganglio-side-specificT cell lines of suppressor or helper phenotype from
cerebrospinal fluid of multiple sclerosis patients. J Neuroimmunol 1986;12:107- 120
35. Johnson D, Hafler DA, Fallis RJ, et al. Cell-mediated immunity
to myelin-associated glycoprotein, proteolipid protein, and myelin basic protein in multiple sclerosis. J Neuroimmunol 1986;
1399- 108
36. Hafler DA, Buchsbaum M, Johnson D, Weiner HL. Phenotypic
and functional analysis of T cells cloned directly from the blood
and cerebrospinal fluid of patients with multiple sclerosis. Ann
Neurol 1985;18:451-458
37. Fleischer B, Marquardt P, Poser S, Kreth HW. Phenotypic
markers and functional characteristics of T IymphoLyre clones
from cerebrospinal fluid in mdtiple sclerosis.J Neuroimmunol
1984;7: 151-162
Без категории
Размер файла
798 Кб
autoreactive, burgdorferiчspecific, radiculomyelitis, lyme, line, borrelia, fluid, cells, cerebrospinal
Пожаловаться на содержимое документа