вход по аккаунту


Reactivity of serum antibodies to the keratin layer of rat esophagus in patients with rheumatoid arthritis.

код для вставкиСкачать
Serum antibodies reactive with the keratin layer
of rat esophagus (AKA) were found in 46 of 80 (57.5%)
rheumatoid arthritis (RA) patients. In contrast, A M
were present in only 7 of 82 (9.5%) patients with other
types of rheumatic disorders and in 2 of 47 (4.2%)
healthy subjects. AKA were not specific for RA, however, because in the former group, A M were present in 4
of 20 (20%) systemic sclerosis patients and in 3 of 12
(25%) ankylosing spondylitis patients. AKA belong predominantly to the IgG class and are complement fixing.
Although found in some RA joint fluids, AKA were not
selectively concentrated in the joint fluid. Absorption of
RA serum with type I human collagen or with human
epidermal keratin did not remove AKA activity. The
frequency of AKA in RA patients both negative and
positive for DR4 was equal. There was no relationship
between the frequency of AKA and the occurrence of
other serum autoantibodies such as antibodies to intermediate filaments, smooth muscle, and nuclear antigens. Serum antibody reactive with human stratum
corneum found in patients with psoriatic arthritis was
From the Clinical Immunology and Rheumatic Disease
Section, Department of Medicine, USC School of Medicine, Los
Angeles, California and the Rancho Los Amigos Hospital, Downey,
Supported in part by a grant from the Southern California
Chapter of the Arthritis Foundation to the Arthritis Service at
Rancho Los Amigos Hospital.
Francisco P. Quismorio, Jr., MD: Associate Professor of
Medicine; Ronald L. Kaufman, MD: Assistant Professor of Medicine; Thomas Beardmore, MD: Associate Professor of Clinical
Medicine; Edward S. Mongan, MD: Associate Professor of Medicine.
Address reprint requests to Francisco P. Quismorio, Jr.,
MD, USC School of Medicine, 2025 Zonal Avenue, HMR #715, Los
Angeles, CA 90033.
Submitted for publication August 30, 1982; accepted in
revised form November 9. 1982.
Arthritis and Rheumatism, Vol. 26, No. 4 (April 1983)
shown to be different from AKA. Rabbit antiserum to
human keratin did not inhibit the reaction of AKA
against the keratin layer of rat esophagus. Autoimmunity to structural proteins including collagen, vimentin
intermediate filaments, smooth muscle antigens, and
keratin is a characteristic feature of RA.
Rheumatoid arthritis (RA) is characterized by
the presence of a wide variety of circulating autoantibodies including rheumatoid factors, antinuclear antibodies, and antibodies to different types of structural
proteins such as collagen, intermediate filaments, and
smooth muscle antigens (1-3). In 1979, Young and
associates (4) described serum antibodies to the keratin layer of rat esophagus in RA. Keratins comprise a
heterogenous group of fibrous soluble proteins that are
found in the epidermis and epidermal appendages (5).
It has also been observed that keratin-related proteins
function as structural proteins forming a cytoskeleton
of a number of epithelial cells (6). Over half of RA
patients tested were found to have serum antikeratin
antibodies (AKA), and it was suggested that the antibody may be a useful marker for the disease (4).
The purpose of our research was 1) to examine
immunologic properties of AKA including complement fixation and cross-reactivity with human keratin,
2) to determine the presence of AKA in synovial
fluids, and 3) to study the relationship of AKA with the
leukocyte cell surface determinant HLA-DR4 as well
as the association of AKA with other types of serum
autoantibodies in RA.
Patients. Three groups of patients were studied:
group I consisted of 80 consecutive patients with classic or
definite RA (7). Group I1 was composed of 84 patients with
Figure 1. Cross-section of a rat esophagus showing a positive test
for antikeratin antibodies. There is a laminar staining pattern of the
keratin layer. There is also nuclear staining of the epithelial cells
indicative of the presence of antinuclear antibodies.
other types of rheumatic disorders such as systemic lupus
erythematosus (SLE), seronegative spondyarthritides, systemic sclerosis, and polymyositis. Group 111 consisted of 47
healthy subjects. All patients with SLE and systemic sclerosis (scleroderma) fulfllled the proposed criteria for the classification of these diseases (8,9).
Serologic tests. Antikeratin antibodies were tested by
an indirect immunofluorescence test described by Young et al
(4). Briefly, a 1: 10 dilution of the test serum was reacted
with 5p thick cryostat sections of rat esophagus. The middle
third of the esophagus was used as a substrate (10). After
washing the sections with phosphate buffered saline (PBS),
0.01M pH 7.3, fluorescein labeled anti-human Ig was added.
A positive test showed laminar pattern of fluorescent staining of the keratin layer (Figure 1). In other experiments,
monospecific sheep ahtibodies to human IgG, IgM, IgA, and
C3 were used. The complement fixing property of the
antibody was tested by a 3-step indirect immunofluorescence test described for antinuclear antibodies in a previous
report from our laboratory (1 1).
Antibodies to intermediate filaments were measured
by an indirect immunofluorescence test using acetone fixed
human fetal fibroblasts grown on glass slide as substrate (2).
Antinuclear antibodies (ANA) were tested by conventional
method using acetone fixed mouse kidney as substrate.
Typing for DR4 antigen was performed by a microcytotoxicity test in Dr. Paul Terasaki’s laboratory (UCLA)
Other methods. Human keratin was isolated using a
method described by Sun and Green ( 5 ) . Briefly, human
epidermal calluses clipped from the feet were minced, homogenized, and extracted serially with 20 mM Tris-hydrochloride pH 7.4 and then with the same buffer containing 8M
urea to remove the nonkeratin proteins. Finally, the pellet
was extracted with 8M urea and 0.02M dithiotreitol. The
dissolved keratin preparation showed multiple bands in
sodium dodecyl sulfate gel electrophoresis similar to those
described by Sun et a1 ( 5 ) .
A New Zealand white rabbit was immunized by
repeated subcutaneous injection of keratin emulsified with
complete Freund’s adjuvant. The gamma globulin fraction of
the rabbit serum was isolated by ammonium sulfate precipitation and labeled with fluorescein isothiocyanate. The molar fluoresceidprotein ratio of the conjugate was 3.60,
F(ab)i fraction of serum IgG was isolated by pepsin
digestion using a method described by Poulsen and Hjort
(13). Type I collagen was extracted from human infant skin
by limited pepsin digestion followed by neutral and acid salt
precipitation. The collagen t y b was verified by cyanogen
bromide peptide analysis (14).
Absorption and blocking experiments. RA and normal
human serum were inactivated at 56°C for 30 minutes. To
0.5-ml aliquots of 1 : 5 dilution of serum were added different
amounts of human type I collagen (1-3 mg) as well as
lyophilized human keratin (1-3 mg). The tubes were incubated at room temperature for 2 hours and then at 4°C for 16
hours. After centrifugation, the absorbed sera were tested
for AKA activity.
RA sera with high titer of antikeratin antibodies (1 : 80
or above) were used in all absorption and blocking experiments.
Sections of rat esophagus were incubated with normal (preimmune) rabbit serum and rabbit anti-human keratin
antiserum for 30 minutes. After washing with PBS, the
sections were incubated with AKA positive RA serum (1 : 10
dilution). After a PBS wash the sections were reacted with
fluorescein labeled rabbit anti-human Ig.
Sections of rat esophagus as well as notmal human
skin were incubated with 1 : 10 dilution of RA serum containing AKA and normal human serum. After washing with PBS,
the tissue sections were incubated with different dilutions
(1 : 5 and 1 : 10) of fluorescein labeled rabbit anti-human
keratin. The sections were examined for blocking of the
fluorescent reaction.
A fluorescent microscope equipped with a Ploem
incident light illuminator was used in the study.
Prevalence of AKA. AKA were found in 46 of 80
(57.5%) patients with RA. In contrast, only 8 of 84
(9.5%) patients with other types of rheumatic disorders (group 11) were positive. Two of 47 healthy
subjects (4.2%) were positive.
Among the patients in group 11, AKA were
found in 4 of 20 (20%) patients with systemic sclerosis
and in 3 of 12 (25%) patients with ankylosing spondylitis. In contrast, none of the patients with SLE, psoriatic arthritis, Reiter’s syndrome, or polymyositis were
positive (Table 1).
Charactristics of AKA. AKA belonged predominantly to the IgG class of immunoglobulins (Table 2).
Two of 5 RA patients, tested had IgM AKA and 1
patient had IgA AKA. The F(ab); fragment of 2 AKA
positive RA patients retained antibody activity. In
contrast, F(ab); fragments prepared from normal hu-
Tnbk 1.
Frequency of antikeratin antibodies
Rheumatoid arthritis
Systemic sclerosis
Ankylosing spondylitis
Systemic lupus erythemaiosus
Psoriatic arthritis
Reiter’s syndrome
Pol ymyositis
Healthy controls
46 (57.5%)
4 (20.0%)
3 (25.0%)
2 (4.2%)
man serum showed negative test for AKA by indirect
immunofluorescence test.
AKA fixed complement when tested with a 3stage indirect immunofluorescence test using human
serum as source of complement and fluorescein labeled anti-human C3 as conjugate. The fluorescent
reaction was abolished when heat inactivated (56°C for
30 minutes) normal human serum was substituted as
the complement source.
Figure 2 shows the distribution of antibody titer
of AKA in 17 patients with RA. The titer ranged from
1:lO to 1:320 with a mean of 1:72.
AKA in synovial fluids. Seven of 15 (47%)
synovial fluids obtained from RA patients contained
AKA. In contrast, none of 1 I non-RA joint fluids was
positive. These included specimens obtained from
patients with SLE, Reiter’s syndrome, gout, and osteoarthritis.
We measured the AKA titer in paired samples
of serum joint fluid from 5 RA patients. The concentration of IgG in the serum was adjusted by dilution to
equal the IgG content of the corresponding joint fluid.
Two of the 5 RA synovial fluids were AKA positive.
Both specimens were from RA patients with circulating AKA. Thc antibody titer of IgG AKA was equal in
the serum and joint fluids, indicating no preferential
concentration of AKA in the joint fluids. Sera and joint
fluids from the 3 remaining RA patients were negative
for AKA.
Association of AKA with DR4 antigen and other
autoantibodies. Thirty-four RA patients were typed for
the presence of the HLA antigen DR4. Nine of 19 DR4
positive RA patients (47.4%) and 10 of 15 DR4 negative RA patients (56.7%) had circulating AKA. The
prevalence of AKA did not differ in the two groups of
RA patients.
The frequency of other types of autoantibodies
was compared in AKA positive RA patients and AKA
negative RA patients. The results are shown in Table
3. Antibodies to intermediate filaments were found in 9
of 17 (52.9%) AKA positive patients and in I I of 18
(61.1%) AKA negative patients. The difference in the
prevalence was not significant. Similarly, there was no
difference in the frequency of antibodies to smooth
muscle antigens and to nuclear antigens in the two
groups of RA patients.
Immunoglobulin class and complement fixing property of
antikeratin antibodies (AKA) in rheumatoid arthritis
Table 3. Relationship of antikeratin antibodies (AKA) with other
autoantibodies in rheumatoid arthritis
Tnble 2.
Ig class
Figure 2. Distribution of serum antibody titer of antikeratin antibodies in 17 patients with rheumatoid arthritis.
positive RA*
9/17 (52.9%)
11/18 (61.1%)
4/15 (26.7%)
11118 (61.1%)
5/15 (33.3%)
12/21 (57.1%)
Antibody to
Intermediate filaments
Smooth muscle
Nuclear antigens
* Number of positive testshumber of sera tested.
Cross-reactivity of AKA with human stratum
corneum. We examined the cross-reactivity of AKA
with human stratum corneum by an indirect immunofluorescence test using cryostat sections of normal
human skin as substrate. Two groups of RA patients
were tested. Eight of 10 (80%) RA patients who were
AKA positive with rat esophagus reacted with human
stratum corneum. The sera displayed a laminar as well
as a diffusely granular staining of the keratin layer of
skin. However, 2 of 10 (20%) AKA negative RA
patients with rat esophagus also reacted with human
stratum corneum. In contrast, none of 5 AKA negative
normal sera were positive.
We studied the reaction of sera from patients
with psoriatic arthropathy which were AKA negative
with rat esophagus. Four of the 8 (50%) sera tested
reacted with human stratum corneum.
The antibody titer of AKA in 5 RA sera did not
change following absorption of the sera with either
type I human collagen or purified human stratum
corneum keratin.
Blocking with rabbit antiserum to human keratin. Rabbit antiserum to human keratin was shown by
indirect immunofluorescence using sheep anti-rabbit
Ig as conjugate to react with keratin layer of rat
esophagus. A diffusely granular staining pattern was
observed which was different from the sharp laminar
pattern displayed by RA serum containing naturally
occurring AKA. Against sections of normal human
skin, the rabbit antikeratin antiserum reacted with the
stratum corneum in a diffusely granular pattern, as
well as with the cytoplasm of the epidermal cells. The
antiserum did not react with the connective tissues of
the dermis.
The capacity of rabbit antikeratin antiserum to
block the reaction of AKA in RA serum was investigated. Preincubation of rat esophagus sections with
rabbit antikeratin antiserum did not block the reaction
of naturally occurring AKA in RA serum. This was
noted in 3 different RA sera tested. Furthermore, the
reaction of RA AKA against normal human skin was
not blocked when the skin sections were preincubated
with the rabbit antikeratin antiserum. Similarly, the
reaction of fluorescein labeled rabbit antikeratin antiserum against normal human skin was not blocked by
preincubation of the skin sections with either AKA
positive RA serum or normal human serum. These
observations indicate the wide heterogeneity of antibodies reacting with keratin layer of rat esophagus as
well as human skin.
Reaction of AKA with synovial membrane. Five
RA sera containing AKA did not react with normal
human and rat synovial membrane lining cells by
indirect immunofluorescence test. Rabbit antiserum to
human keratin also failed to react with synovial lining
The results of our study confirm the high frequency of AKA in the sera of RA patients (5,10,15). It
appears, however, that the antibody is not specific for
RA, thus its purported usefulness as a diagnostic
marker for RA is limited. Despite the fact that only
9.5% of 84 patients with other types of rheumatic
disorders were AKA positive, patients with systemic
sclerosis and ankylosing spondylitis have a high prevalence of AKA, 20% and 25% respectively. It is of
interest that both conditions have in common with RA
a high frequency of antibodies to collagen, another
structural protein (16,17).
Studies by other investigators have shown that
serum antibodies to stratum corneum of human skin
detected by the indirect immunofluorescence method
are frequently seen in patients with psoriasis (18). We
were especially interested to learn whether antibodies
to human stratum corneum were identical to AKA
found in RA sera. Our observations, however, indicate
that the two antibodies are different in several respects. Firstly, none of our patients with psoriatic
arthritis had AKA, yet half of them had circulating
antibodies to human stratum corneum. Secondly,
there were AKA negative RA patients whose sera
contained antibodies to human stratum corneum.
Thirdly, absorption of RA serum with human epidermal keratin did not block the fluorescent reaction of
RA AKA. Finally, preincubation of rat esophagus with
rabbit anti-human keratin antiserum did not inhibit the
reaction of AKA.
It has been pointed out that in certain situations
the fluorescent reaction obtained by the indirect immunofluorescence test for serum antibodies to human
stratum corneum was not due to true antigen-antibody
reaction but was brought about by the nonspecific
binding of IgG through its Fc portion (18). Thus our
findings showing that the F(ab); fragment of RA serum
which retained AKA activity was especially important
because it established the antibody nature of the
fluorescent reaction in AKA.
We investigated the relationship between the
presence of circulating AKA and the histocompatibility antigen DR4 because it has previously been reported that the immune reaction to collagen in RA patients
as well as normal subjects was associated with the
DR4 antigen (19). Our study showed no such association between AKA and DR4. Moreover, we found that
the presence of AKA was not related to the occurrence of other types of autoantibodies such as antibodies to intermediate filaments, smooth muscle, and
nuclear antigens. This observation suggests that factor(s) involved in the induction of AKA may be
different from those involved in the induction of autoimmunity to other antigens. On the other hand, a
strongly positive correlation between the presence of
rheumatoid factor and AKA has been described by
other investigators (4,10,15). That AKA were not a
type of cross-reacting rheumatoid factor was established by Young and associates (4) who demonstrated
that while the rheumatoid factor activity of RA serum
was abolished by absorption with aggregated gamma
globulin, the AKA activity was retained.
The significance of AKA in RA in not clear. It is
possible that the antibody cross-reacts with keratinlike proteins of intermediate filaments of epithelial
cells such as synovial lining cells. However, we observed that the naturally occurring AKA as well as the
rabbit anti-human keratin did not react with human or
rat synovium. This confirms the recent observation
that the intermediate filaments of the synovial lining
cells consist of vimentin (20).
The complement fixing property of AKA and its
presence in synovial fluids of some RA patients suggests that it might be of potential pathogenetic significance. It is conceivable that AKA may form soluble
immune complexes and participate in the pathogenesis
of synovial inflammation, much like the proposed role
of other autoantibodies in RA such as anticollagen
antibodies and rheumatoid factors (21). Although we
found no preferential concentration of AKA in RA
joint fluids, further studies that will examine the
presence of AKA in isolated immune complexes
aqd/or cryoproteins from joint fluids will have to be
The authors gratefully acknowledge the help of Mrs.
Min Huang, Mrs. Maria Hutchinson, Ms Nancy Jones, Dr.
Paul Benya, and Dr. Howard J. Marshall.
1 Andriopoulos NA, Mestecky J, Miller EJ, Bradley EL:
Antibodies to native and denatured collagens in sera of
patients with rheumatoid arthritis. Arthritis Rheum
19~613-616, 1976
2. Osung OA, Chandra M, Holborow EJ: Antibodies to
intermediate filaments of the cytoskeleton in rheumatoid
arthritis. Ann Rheum Dis 41:69-73, 1982
3. Andersen I, Andersen P, Graudal H: Smooth-muscle
antibodies in rheumatoid arthritis. Acta Pathol Microbiol Scand 88:131-135, 1980
4. Young BJJ, Mallya RK, Leslie RDG, Clark CJM,
Hamblin TJ: Antikeratin antibodies in rheumatoid arthritis. Br Med J 2:97-99, 1979
5 . Sun T, Green H: Keratin filaments of cultured human
epidermal cells. J Biol Chem 253:2053-2060, 1978
6. Sun ‘IT, Shih C, Green H: Keratin cytoskeletons in
epithelial cells of internal organs. Proc Natl Acad Sci
USA 76:2813-2817, 1979
7. Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA:
1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9: 175-176, 1958
8. Cohen AS, Reynolds WE, Franklin EC, Kulka JP,
Ropes MW, Shulman LE, Wallace SL: Preliminary
criteria for the classification of systemic lupus erythematosus. Bull Rheum Dis 21543-648, 1971
9. Masi AT, Rodnan GP, Medsger TA Jr, Altman RD,
D’Angelo WA, Fries JF, LeRoy EC, Kirsner AB, MacKenzie AH, McShane DJ, Myers AR, Sharp GC: Preliminary criteria for the classification of systemic sclerosis (scleroderma). Bull Rheum Dis 31:l-6, 1981
10. Johnson GD, Carvalho A, Holorow EJ, Goddard DH,
Russell G: Antiperinuclear factor and keratin antibodies
in rheumatoid arthritis. Ann Rheum Dis 40:263-266,
1 1 . Beaulieu AD, Quismorio FP Jr, Kitridou RC, Friou GJ:
Complement fixing antibodies to ds-DNA in systemic
lupus erythematosus. J Rheumatol 6:389-396, 1979
12. Terasaki PI, Bernoco D, Park MS, Ozturk G, Iwaki Y:
Microdroplet testing for HLA-A, -B, -C and -D antigens.
Am J Clin Patbol 69: 103-120, 1978
13. Poulsen F, Hjort T: A simple method for the production
of F(ab)z’ preparations by pepsin digestion of total
serum protein. Acta Pathol Microbiol Scand 88:241-245,
14. Benya P, Padilla SR, Nimni ME: The progeny of rabbit
articular chondrocytes synthesize collagen types I trimer, but not type 11: verifications by cyanogen bromide
peptide analysis. Biochemistry 16:865-872, 1977
15. Scott DL, Delamere JP, Jones LJ, Walton KW: Significance of laminar antikeratin antibodies to rat esophagus
in rheumatoid arthritis. Ann Rheum Dis 40:267-271,
16. Mackel AM, Delustro F, Harper FE, LeRoy EC: Antibodies to collagen in scleroderma. Arthritis Rheum
25522-53 1, 1982
17. Clague RE, Shaw MJ, Holt PJL: Incidence of serum
antibodies to native type I and type I1 collagens in
patients with inflammatory arthritis. Ann Rheum Dis
39~201-206, 1980
18. Beutner EH, Jablonska S, Jarzabek-Chorzelska M, Mar-
ciejowska E, Rzesa G, Chorzelski TP: Studies in immunodermatology. VI. IF studies of autoantibodies to
the stratum corneum and of in vivo fixed IgG in stratum
corneum of psoriatic lesions. Int Arch Allergy Appl
Immunol 48:301-323, 1975
19. Solinger AM, Bhatnagar R, Stobo JD: Cellular, molecular and genetic characteristics of T cell reactivity to
collagen in man. Proc Natl Acad Sci USA 78:3877-3881,
20. Osung OA, Chandra M, Holborow EJ: Intermediate
filaments in synovial lining cells in rheumatoid arthritis
and other arthritides are of vimentin type. Ann Rheum
Dis 41:74-77, 1982
21. Menzel J , Steffen C, Kolarz G, Eberl R, Frank 0,
Thumb N: Demonstration of antibodies to collagen and
of collagen-anticollagen immune complexes in rheumatoid arthritis synovial fluids. Ann Rheum Dis 35:446450, 1976
Без категории
Размер файла
486 Кб
patients, antibodies, keratin, arthritis, layer, serum, reactivity, rat, esophagus, rheumatoid
Пожаловаться на содержимое документа