close

Вход

Забыли?

вход по аккаунту

?

Comparison of antigenic sources for acetylcholine receptor antibody assays in myasthenia gravis.

код для вставкиСкачать
Comparison of Antigenic Sources
for Acetylcholine Receptor Antibody
Assays in Myasthenia Gravis
Mildred W. McAdams, BS, and Allen D. Roses, M D
Antibody titers to the acetylcholine receptor (AChR) from patients with my asthenia gravis, in identical serum
samples, were directly compared using denervated rat, human, and baboon muscle as the source of AChR antigen
for radioimmunoassay (RIA). Calculations were standardized by using binding isotherms for each antigen source
and calculating the percentage of AChR sites labeled with ['~~jI]a-bungarotoxin
at the concentration used i n the
RIA. I n patients with high AChR antibody titers, t h e antibody concentration when human muscle was used as the
antigen source measured up to tenfold higher than that with denervated rat muscle. I n patients who had low
antibody titers with human muscle antigen, assays using rat denervated muscle AChR frequently failed to demonstrate diagnostically abnormal titers. T h e data explain differences among several reported series in the percentages of patients with myasthenia gravis who had elevated serum antibody concentrations. AChR antibody
concentrations with baboon muscle as t h e antigen source were comparable to those in which human muscle was
used.
MiAdams MW, Roses AD. Comparison of antigenic sources for acetylcholine receptor antibody assays in
myasthenia gravis. Ann Neurol 8 6 - 6 6 , 1980
Workers from several laboratories have discussed the
correlation between the concentration of serum antibodies to acetylcholine receptor (AChR) and the
clinical status of patients with myasthenia gravis [ 1,
la, 3,6, 7-9, 11, 1 2 , 2 0 , 2 3 , 2 4 ] . Inconsistent results
exist in the literature, reflecting the use of different
methods of calculating arid expressing the data [ 1, 2,
9, 10, 151. For example, Lindstrom et a1 [12] found
that 8?$& of patients with myasthenia gravis had eievated antibody concentrations (using human muscle
antigen), Appel and his associates [l] found 7Oci;.
(with denervated rat antigen), and Brenner and coworkers [2] found 3?% (with rat muscle antigen,
staphylococcal protein A precipitations rather than
antihumad antibody).
Accordingly, we have compared denervated rat
muscle and human muscle directly in an attempt to
develop a radioimmunoassay (RIA) with the least
variance for long-term patient follow-up. W e initiated a prospective analysis of sera from all known
and possible cases of myasthenia gravis in our
clinic, using both human muscle and denervated
rat muscle as antigen sources to measure AChR antibody concentrations in identical serum samples. In
the course of evaluating AChR antibody concentrations with each antigen, the variability obtained
with multiple human preparations and their inconstant relationship to the concentrations obtained with
denervated rat AChR assays became apparent. We
also initiated studies using baboon muscle antigen,
since this was a source of large amounts of antigen
that could be stored for use with sera from the same
patient followed over years.
LWaterials and Methods
m-Bungarotosin labeled with iodine 125 (specific activity,
I 0 to 20 pCiIpg: was obtained from New England Nuclear.
Benzoquinuniurrr, a gift from Sterling-Winthroy Pharmaceuticals. wa5 used to inhibit binding of [lz:IJabungarotoxin to AChR binding sites. Antihuman immunoglobulin G (1gG) serum was collected from rabbits immunized with hunian gamma globulin obtained from Miles
Laboratories. AChR was prepared from human leg muscle
obtained from therapeutic amputations, from rat leg muscle that had been denervated (sciatic nerve) ten days prior
to preparation, and from the extremity muscles of baboons
killed during pulmonary oxygen toxicity experiments by
Dr Walter Wolfe. All baboon muscle was obtained within
30 minutes of sacrifice.
AChR Antigen
AChR antigen was prepared by a slight modification of the
method of Lindstrom [lo]. Fresh muscle was processed
~
From the Departments of Medicine (Neurology) and Biochemistry, Howard Hughes Medical Institute, Duke University Medical
Center, Durham, NC.
~
Received June 7, 1979, and in revised form Sepr 20 and Nov 1.
Accepted for publication Nov 4, 1979.
Address reprint requests to Dr Roses, Box 2900, Duke University
Medical Center, Durham, NC 27710.
0364-5134180l070061-06$01.25 @ 1979 by Allen D. Roses
61
0.2
1
V
=0
m
0.15
u
n
u)
-m
0.I
0
I,
0.05
0
I
5
0
10
A
20
20
30
35
[Iz5I q-Eutx] X 10" M
40
45
50
.
,
.
,
40
45
50
15
0 2
0
c
0
=
m
ri
0.15
a
u)
-u
y1
z
0.1
0
0.05
0
L/
*I.
,
,
[IZ3I d-Butx]
B
,
,
X 10'" M
0 2
P
3
m
0 15
u
n
(0
VI
01
sa
0.05
0
I
0
C
within one hour after removal. Connective tissue and far
were dissected off before weighing, and the muscle was
homogenized in a Waring blender at 4°C for one minute
with four volumes (w/v) of buffer containing 0.1 M sodium
chloride, 0 01 M sodium phosphate, and 0.01 M sodium
a i d e at pH 7.0. The honiogenate was centrifuged for two
hours at 25,000 g. T h e pellet that was produced was
homogenized for 30 seconds in a Waring blender with two
volumes ( d v ) of 2% Triton X-100, 0.1 M sodium chloride, 0.01 M sodium phosphate, and 0.01 M sodium atide
at pH 7.0, and was stirred for one hour at 4°C Following
centrifugation at 2 5,000 g for two hours, the supernatant
contaming AChR was decanted. Protein content was measured by the method of Lowry et a1 [13]. The baboon muscle preparation was filtered through an Amicon TCFlO
system using an XM5O filter to concentrate the antigen.
Aliquots (0.1 to 0.5 ml) were incubated overnight at 4°C
with 5 nM ['2"Ila-bungarotoxin with and without additioh
of 1 m M benzoquinonium in a total volume of 1 ml. The
62 Annals of Neurology Vol 8 No 1 July 1980
5
10
15
20
25
30
35
[ 1 2 5 u~- ~ u t x ] x 1010 M
F i g 1 Saturatron of human (A),denerzated rat (B), and baboon (C) muscle AChR anttgens with ['Z9]a-bungaratoxzn.
Incubatzonr and Sephadex G-200 chromatography were performed at dacribed zn MethodJ (A) At 8 X lo-'' M ['2'lIlabungarotoxzn, 75 s/r of total human AChR ~rteswere labeled,
lBl at 7.1 x 10 M, 74% of tatuldenemated rut AChR
sztes were labeled, (C)at 8 x 1O - l 0 M , 35 % of total baboon
AChR sites were labeled.
AChR-['2iI]cr-bungarotoxin complex eluted with the void
volume o n 3 x 2 5 cm Sephadex G-200 columns using
0.5% Triton X-100, 0 1 M sodium chloride, 0.01 M
sodium phosphate, and 0.01 M sodium azide at pH 7.0.
The specific activity of AChR was calculated by subtracting
the ['251]a-bungarotoxin bound in the presence of benzoquinonium from that bound in its absence. Smaller volumes of denervated rat muscle preparation (20 to 40 pl)
were used dbe to the higher concentration of AChR in this
material
Following determination of the concentration of AChR
with ['ZSI]a-bungarotoxin,the percentage of specific binding sites occupied at a concentration of 2 X lo-'' M AChR
was measured using a saturation assay, with [12sI]abungarotoxin ranging from 0.1 to 5.0 nM. Each assay contained 2 x lo-'" M AChR preincubated for 30 minutes at
4°C with and without 1 mM benzoquinonium before addition of ['251]a-bungar~toxin.Assays were equilibrated
overnight at 4°C prior to chromatography with Sephadex
G-200 as just described. T h e concentration of [12sI]abungarotoxin that gave 75% saturation of the specific
['2511a-bungarotoxin binding sites was determined from the
plot shown in Figure 1 and was used routinely in the RIA
ill]; for the baboon muscle preparation the concentration
of ['ZTja-bungarotoxin was that which would give 35%
saturation (Fig 1). Concentrations below saturation were
used to decrease background and so that a relatively constant concentration of ['251]a-bungarotoxin was being
utilized.
Radioimmunoasmy
Serum antibody concentrations were determined using 2 x
lo-'" M AChR preincubated with and without benzoquinonium as described before ['2511a-bungarotoxin in a
total volume of 1 ml was added. The labeling reaction was
incubated for two hours at 4"C, then 5 p l of patient serum
was added and the mixture was incubated overnight at 4°C.
Antihuman I g G serum was then added and incubated for
four hours at 4"C, and the precipitate was centrifuged for
two minutes in a Beckman Microfuge. The pellet was
washed once with 1 ml of 0.5%' Triton-phosphate buffer
and radioactivity was counted in an Intertechnique gamma
counter. The specific amount of AChR-antibody complex
precipitated per 5 pl was obtained by subtracting the
counts per minute with benzoquinonium from the total
counts per minute and correcting for 7 5 % (35% for baboon) labeling. Division by the specific activity for ['251]abungarotoxin (cpm/mol) gives the moles of AChR precipitated per volume of serum tested.
The linear precipitin range for the AChR antibody in this
assay was determined to be between 1 x lo-'' M and 2 x
lo-@M, Any serum with an antibody concentration higher
M was assayed again using 1 p1 of it diluted
than 2 x
with 4 pI of normal serum. Determinations lower than 5 x
M were repeated with 10p1 of serum and precipitated
with twice the usual volume of rabbit antiserum; this increased the sensitivity for low concentrations of antibody.
The data from multiple serum antibody assays using different antigenic sources were subjected to two-way analysis of
variance [22].
An AChR concentration (2 x lo-'" M) was used in our
RIA system (Fig 2) that was found to be in excess for all but
the highest concentration of serum antibodies. This allowed accurate calculation of the number of AChR sites.
Using 2 x lo-'" M AChR, the assay was linear through a
wide range of antibody concentrations. Figure 3 illustrates
linearity using serum dilutions with a total volume of 5 p1
of serum in the assays; dilutions were made with normal
serum. The data in Figure 3 were representative of experiments performed with each antigen batch and with multiple
sera.
0.1
0
L
1
0
1.5
I
05
[AChR]
2.5
2
3
X 10" M
F i g 2. Maximum precipitation of 0.1 pmol of a myasthenia
gravis antibody (Ab) (5 pl of a serum with 20.8 nM antibody) with human AChR antigen (Ag). The ['25i]abungarotoxin Concentration (20 x
Mi was sufjcient to
saturate all AChR sites. Assay conditions were as described i n
Methods for radioimmunoassay.
0
0
02
04
$1
06
00
I
MG Standard Serum
F i g 3. Linear precipitin range of 2 x I0 " M human AChR
antigen (Ag). Similar data were obtained with multiple M G
sera and each batch of AChR antigen. Total volume of serum
at each point was S pl. Dilutions were made with normal
serum. (Ab = antibody.)
Results
Table 1 presents the data from AChR antibody assays
using 14 representative serum samples: A through G
are from patients with relatively high antibody concentrations, H through N are from patients with low
concentrations. Each line of data represents the same
serum sample assayed with the three antigen sources.
In every sample, the concentration obtained with
human muscle antigen was higher than that measured
McAdams and Roses: AChR Antibody Levels in MG 63
Table 1. Antibody Response i n Rudioimmunoassays of Identical
Sera Tested with Multiple Antigen Sourcesa
Serum
Sample
Human
Muscle
78-VIIa
Denervated
Rat
Muscle
Baboon
Muscle
78-1
HIGH ANTIBODY CONCENTRATION
A
B
C
D
E
F
G
11.1
9.7 1
6.35
7.59
1.47
6.09
4.54
2.16
0.284
0.835
2.40
1.07
3.32
0.272
12.5
6.86
4.24
7.88
1.99
4.47
2.92
LOW ANTIBODY CONCENTRATION
H
I
J
K
L
M.
N
0.121
0.235
0.180
0.660
0.776
0.82 1
0.294
0.008 ( - )
0.025 (-)
0
(-1
0.2 10
0.221
0.305
0.011 ( - )
0.180
0.126
0.183
0.97 1
1.02
1.37
0.280
“Assay conditions were as described in Methods, and each serum
sample was tested with each antigen source. Values are in
nanornoles per liter. Two-way analysis of variance for human versus denervated rat muscle: F = 9.616, p < 0.01; human versus
baboon muscle: F = 1.232, not significant. “High” is operationally
defined as >1 nM (1-422).
(-) = values indistinguishable from controls; normal ranges were
defined by testing a series of normal control sera with each type of
antigen. These four sera would not be useful for diagnosis using
denervated rat muscle antigen.
with denervated rat muscle. These results are representative of the complete series, but this group of
serum samples was also tested with multiple human
muscle preparations (Table 2). Four of the seven
samples with low titers would be indistinguishable
from controls using denervated rat muscle as the
antigen source (see Table 1). However, at both
higher and lower antibody concentrations, the assays
performed with baboon muscle antigen were as sensitive as those done with human muscle antigen.
Table 2 demonstrates the variability of measured
antibody concentrations in identical sera using four
different human muscle AChR preparations. The assays were performed three times on each serum sample using one of the human muscle preparations
(Table 3). The apparent variability lay between assays
performed with different muscle preparations, implying that AChR antigen obtained from different
human donors possesses variable antigenic properties.
The data in Table 1 are representative of the prospective series of patients examined in our clinic. In
every case, the clinical diagnosis was established be-
64
Annals of Neurology
Vol 8 No 1 July 1980
Table 2. Antibody Respon > < in Radioimmunoassays
Using Multiple Human hiuscle A C h R Preparations
with Identical Sera“
H u m a n Preparations
Serum
Sample
77-1
77-11
78-VI
78-VIIa
A
B
C
D
E
F
G
7.76
7.50
3.08
5.24
1.78
6.02
2.45
11.1
9.62
3.39
7.69
1.62
6.32
2.17
10.3
13.4
4.74
6.80
2.16
6.82
2.59
11.1
9.71
6.35
7.59
1.47
6.09
4.54
aAssay conditions were as described in Methods. Values are in
nanomoles per liter. Two-way analysis of variance: F = 4.0, p <
0.05.
Table 3 . Antibody Responre i n Rudioimmunoassays Using
the Identical Humun Muscle A C h R Preparation a
Serum
Sample
A
B
C
D
E
F
G
Assays
78-VIIa
78-VlIb
78-VlIC
11.1
9.7 1
6.35
7.59
1.47
6.09
4.54
9.57
10.4
6.36
7.62
1.57
6.22
3.90
10.4
10.4
6.12
6.39
1.46
5.82
4.17
“Human muscle AChR preparation 78-VII was used on three
separate days (a, b, and c) over a period of three months. Values
are in nanomoles per liter. Two-way analysis of variance: F =
0.766, not significant.
fore the antibody assay was made available to the attending physician. Clinical testing included history,
physical examination, response to edrophonium,
Jolly test, single fiber electromyography, computed
tomographic scan of the thorax, and other appropriate measures. With human muscle as the antigen
source, 29 of 33 patients (8896) clinically diagnosed
as having myasthenia gravis had measurable AChR
antibody concentrations; only 2 3 had measurable
concentrations with the denervated rat muscle
assay. In addition, all 28 patients who entered the
study as controls or as diagnostic problems but who
were not clinically diagnosed as having myasthenia
gravis showed no measurable AChR antibody in their
sera. Our data explain the differences between the
initial results reported by Lindstrom and associates
[12] and Appcl and his colleagues [l], for they employed, respectively, human muscle and denervated
rat muscle assays. The lower sensitivity of the denervated rat muscle assay, rather than selection of differ-
ent populations of patients, accounts for its decreased
effectiveness as a diagnostic tool.
Discussion
To understand the potential errors involved in using
the AChR antibody assay as a diagnostic test, the
method of the determination should be briefly outlined. Serum antibodies to AChR are allowed to bind
to AChR antigen that has been labeled by addition of
a noncompetitive agent such as [1251]a-bungarotoxin.
The complex is then precipitated by addition of an
antiserum, usually rabbit or goat antihuman IgG
serum. Two important variables exist: the serum
AChR antibodies that are being assayed and the
AChR antigen source. T o compare data o n AChR
antibodies accurately in patients with myasthenia
gravis, the second variable must be tightly controlled.
Furthermore, to compare the same sera with different antigen sources, the method of calculation must
be standardized.
Our data demonstrate that assays with denervated
rat muscle as the antigen source d o not directly correlate with assays utilizing human muscle antigen. All
AChR antibody levels were lower when denervated
rat muscle was used. Indeed, several patients with
high serum AChR antibody levels determined by
human muscle antigen had levels an order of magnitude lower with denervated rat antigen. Several
patients with low but diagnostic levels with human
muscle preparations were more sensitive for purposes
trols when rat muscle AChR preparations were used
(H, I, J, and N in Table 1). Therefore, the human
muscle preparations were more sensitive for purposes of diagnosis.
Variation between human muscle antigen preparations makes such preparations unsuited for evaluating the progress of individual patients [5, 9, 20, 211.
Substantial variation (F = 4.0, p < 0.05) in antibody
response was observed when multiple AChR preparations were used to test identical serum samples
(see Table 2). The consistent results (F = 0.766, not
significant) o b t i n e d with multiple determinations
using a single human muscle AChR preparation (see
Table 3 ) demonstrate that the large variation between multiple AChR preparations in Table 2 is due
to human antigenic differences rather than day-today laboratory variation.
Baboon muscle AChR antigen was found to be as
sensitive as human muscle antigen in assays done for
the purpose of diagnosis (see Table 1). Serum AChR
antibody levels were comparable to those obtained
with human antigen in every case. Since the baboons
were being killed infrequently (see Materials) but
predictably, large amounts of fresh, healthy muscle
could be processed and batches of antigen stored
for future use. This allows one to eliminate variability of antigen by using the identical batch of
AChR, thereby allowing more accurate assay of patients' AChR antibody levels over long periods
of follow-up.
Our clinical experience in following and treating a
substantial number of patients with myasthenia gravis
is that their antibody levels fail to correlate directly
with clinical status [ 191; others have reported similar
findings [2, 14, 211. I n our clinic, patients treated
at the time of diagnosis with plasmapheresis and
thymectomy (without other medications) have improved clinically and maintained their improvement
over several years while mctintaining high antibody
levels [221. Following thymectomy, antibody titers in
individual patients fall over a different time scale
(months to years) from the clinical response, suggesting that the pathogenesis of the disease may involve factors more directly linked to the thymus [4,
14, 16-19, 23, 251.
Supported in part by a Clinical Research Grant from the Muscular
Dystrophy Association of America and a Research Grant from the
National Multiple Sclerosis Society.
We wish to thank Ms Bobbie Williams for excellent secretarial
assistance, Dr Jeffery Vance for valuable suggestions and criticism,
and Dr Walter Wolfe, of the Division of Thoracic Surgery, for
coordinating and making available the baboon muscle.
References
1. Appel SH, Almon RR, Levy N : Acetylcholine receptor antibodies in myasthenia gravis. N Engl J Med 293:760-761,
1975
la. Bradley RJ, Dwyer D, Morley BJ, et al: Humoral immunity
in myasthenia gravis: relationship to disease severity and steroid treatment. Prog Brain Res 49:441-448, 1979
2. Brenner T, Ambramsky 0, Lisak RP, et al: Radioimmunoassay of antibodies to acetylcholine receptor in serum of myasthenia gravis patients. Isr J Med Sci 14:986-989, 1978
3. Christiansen FT,Houliston JB, Dawkins RL: HLA, anti-DNA
and complement in myasthenia gravis. Muscle Nerve 1:467470, 1978
4. Conti-Tronconi BM, Morgutti M, Sghirlanzoni A, et al: Cellular immune response against acetylcholine receptor in
myasthenia gravis: I. Relevance to clinical course and pathogenesis. Neurology 29:496-501, 1979
5. Dau PC, Lindstrom JM, Cassel CK, et al: Plasmapheresis and
immunosuppressive drug therapy in myasthenia gravis. N
Engl J Med 297:1134-1140, 1977
6 . Heilbronn E. Lefvert AK, Stalberg E: Acetylcholine receptor
antibodies in the diagnosis of human and experimental myasthenia gravis. Muscle Nerve 1:427-431, 1978
7 . Kalies I, Kalden JR, Heinz F, et al: [Demonstration of acetylcholine receptor antibodies in the serum of myasthenia gravis
patients applying affinity chromatographically purified human
acetylcholine receptor preparations] (author's translation).
Klin Wochenschr 57:875-881, 1979
8. Kay P, Dawkins B, Garlepp M, et al: Experience with the
anti-acetylcholine receptor antibody assay in myasthenia
gravis. Abstracts of the Fourth International Congress on
Neuromuscular Diseases, Montreal, September, 1978
McAdams and Roses: AChR Antibody Levels in M G 65
9. Lefvert AK, Bergstrom K, Matell G, et al: Determination of
acetylcholine receptor antibody in myasthenia gravis: clinical
usefulness and pathogenic implications. J Neuroi Neurosurg
Psychiatry 41:394-403, 1978
10. Lindstrom J: An assay for antibodies to human acetylcholine
receptor in serum from patients with myasthenia gravis. Clin
Immunol Immunopathol 7:36-43, 1977
11. Lindstrom JM, Lennon VA, Seybold ME, et al: Experimental
autoimmune myasthenia gravis: biochemical and immunochemical aspects. Ann N Y Acad Sci 274:251-274,1976
12. Lindstrom JM, Seybold ME, Lennon VA, et al: Antibody to
acetylcholine receptor in myasthenia gravis. Neurology (Minneap) 261054-1059, 1976
13. Lowry O H , Rosebrough NJ, Farr AL, et al: Protein measurements with the Folin phenol reagent. J Biol Chem
193:265-275, 1951
14. Massoud A, Lofti J, Imanian H, er al: Cell mediated immunity
(CMI) and lymphocyte subpopulation studies in thymectomized and non-rhymectomized patients with myasthenia
gravis. Abstracts of the Fourth International Congress on
Neuromuscular Diseases, Montreal, September, 1978
15. Mirtag T, Kornfeld P, Tormay A, et al: Detection of antiacetylcholine receptor factors in serum and thymus from patients with myasthenia gravis. N Engl J Med 294:691-694,
1976
16. Morgutti M, Conti-Tronconi BM, Sghirlanzoni A, et al: Cellular immune response to acetylcholine receptor in myasthenia gravis: 11. Thymectomy and corticosteroids. Neurology
29:734-738, 1979
17. Morrell RM: Acetylcholine receptors on peripheral lympho-
66 Annals of Neurology Vol 8 N o 1 July 1980
cytes and monocytes in myasthenia gravis. Abstracts of the
Fourth International Congress on Neuromuscular Diseases,
Montreal, September, 1978
18. Richman DP, Patrick J, Arnason BGW: Cellular immunity in
myasrhenia gravis. N Engl J Med 294:694-698, 1978
19. Roses AD, Olanow CW, McAdams MW, et al: There is no
direct correlation between serum anti-acetylcholine receptor
antibody levels and the clinical status of individual patients
with myasthenia gravis. Neurology (in press)
20. Scadding GK, Webster ADB, Ross M, et al: Humoral immunity before and after thymectomy in myasthenia gravis. Neurology 29:502-506, 1979
2 1. Seybold ME, Baergen RN, Nave B, et al: Anti-acetylcholine
receptor antibody concentrations after rhymectomy in patients with myasthenia gravis. Br Med J 2:1051-1053, 1978
22. Sokal RFL, Rohlf FJ: Two way analysis of variance. In Emerson
R, Kennedy D, Park RB, et al (eds): Biometry. San Francisco,
Freeman, 1969
23. Wechsler AS, Olanow CW: Myasthenia gravis. In Sabiston
DC (ed): The Davis Christopher Textbook of Surgery.
Philadelphia, Saunders (in press)
24. Yagi K, Beppu H, Uono M, et al: Antibody to acetylcholine
receptor in myasthenia gravis and its clinic significance.
Abstracts of the Fourth International Congress on
Neuromuscular Diseases, Montreal, September, 1978
25. Zilko PJ, Dawkins RL,Holmes K, et al: Genetic control of
suppressor lymphocyte function in myasthenia gravis: relationship of impaired suppressor function to HLA-B8I
DRW3 and cold reactive lymphocytotoxic antibodies. Clin
Immunol lmmunopathol 14:222-230, 1979
Документ
Категория
Без категории
Просмотров
0
Размер файла
500 Кб
Теги
acetylcholine, antigenic, myasthenia, source, antibody, receptov, gravis, comparison, assays
1/--страниц
Пожаловаться на содержимое документа