вход по аккаунту


Studies of isolated synovial lining cells of rheumatoid and nonrheumatoid synovial membranes.

код для вставкиСкачать
Studies of Isolated Synovial Lining Cells of Rheumatoid
and Nonrheumatoid Synovial Membranes
T. Douglas Kinsella, John Baum and Morris Ziff
Synovial membranes flom patients with rheumatoid arthritis and other arthritides were digested with tryps.n to produce suspensions of synovial cells. The
cells were stained by the direct fluorescent antibody technique for IgG, IgM and
the ple (C3) component of complement, using fluorescein and rhodamine isothiocyanate-labeled antisera. A characteristic diffuse staining pattern for IgG and
pIcwas observed in the cytoplasm of the phagocytic lining cells of rheumatoid
synovium. In seropositive patients, inclusions containing IgG, IgM, and plCwere
also stained in these cells. The role of these immunoglobulins, localized in the
lining cells, is discussed in relation to rheumatoid inflammation.
Proliferation of the synovial lining cells
occurs in a number of arthritides, although
the underlying mechanism of this change is
not known. Electron microscopic examination of both normal and abnorma, synovial
tissue (1-5) has differentiated three types
of lining cells: a) type A cells, which are
structurally adapted for phagocytosis by
virtue of the presence of numerous lysosomes; b) type B cells, adapted for protein
synthesis by virtue of a well-developed
rough endoplasmic reticulum; and c) type
C cells (intermediate cells) which have the
From the Rheumatic Disease Unit, Department
of Internal Medicine, The University of Texas
Southwestern Medical School, Dallas, Texas.
Supported by Research Grant AM-09989 US Public Health Service and a grant from The Arthritis
KINSELLA,MD: Fellow of the Canadian Arthritis and Rheumatism Society. Present
address: Rheumatic Disease Unit, Queen’s Univeraity, Kingston, Ontario, Canada. JOHN BAUM,MD:
Present address: University of Rochester School of
Medicine, Rochester, New York. MORRIS
MD: Recipient of Research Career Award, National
Institute of Arthritis and Metabolic Diseases.
Submitted for publication March 6, 1970; accepted July 13, 1970.
capacity for both phagocytosis and proteiii
Much attention has recently been given
to the phagocytosis of inclusions by cells
present in synovial effusions (6-11). These
inclusions have been shown to contain IgG,
IgM,rheumatoid factor and athe ple or CY
component of complement. T h e evidence
from electron microscopic investigations,
which indicates that a large fraction of
synovial lining cells have phagocytic activity, has suggested the possibility that components present in the synovial fluid might
also be taken u p by these cells. PrevioL:s
studies by Kaplan and Vaughan (12), KapIan (13) Rodman et a1 (14), and Fish et
al (15) demonstrated the presence of various immunoglobulins and, in some instances (14, 15), complement components
in the superficial layers of rheumatoid synovial tissue. However, because of the limitations imposed by
use of frozen settions of synovial tissue, these authors could
not specifically identify the types
. _ of cells
involved in this process. in the present
investigation, an attempt has been made to
isolate the lining cells of rheumatoid and
Mhritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
other synovial tissues with a view toward
characterizing these cells further and establishing the identity of the synovial fluid
components that might be present wi4thin
Fraser and Catt (16) and Williamson et
a1 (17) demonstrated that synovial lining
layer cells could be recovered in suspension
after the trypsinization of cadaver joints. I n
the present studies, suspensions of synovial
lining cells have been obtained by trypsin
treatment of fresh human synovium obtained either at arthrotomy or by needle
biopsy. T h e individual types of cells were
identified by appropriate histx!?emiral
procedures. I n the phagocytic lining cel s
of patients with rheumitoid synovitis, immunoglobulins and compleme i t components have been identified.
Synovlal specimens. Synovial tissue was obtained from 3.1 patients at arthrotomy and by
needle biopsy using the Parker-Pearson needle
(18). Nine patients with adult rheumatoid arthritis
had either definite or classical disease by ARA
ciiteria (19). Four adult rheuwatoid patients and 1
Iuvenile rheumatoid pltient h x l scropositive disease (RA [RF+]) . Seven patients had seronegative
juvenile rheumatoid arthritis. Synovium from 5
patients with ruptured meniscus was obtained at
elrctite surgery and was grossly n o r n i ~ lin appearance in all cases. Three patients with gonococcal
arthritis had effusions which in all cases gave
positive cultures for N gonorrheae.
Synovicrl digestion technic. Stel ile precautions
were employed u p to the point at which smears
were prepared in the cytocentrifuge.* Within 2 hi
after surgery, specimens of synovium were washed
free of blood and debris in medium (normal saline,
10 ml; fetal calf serum,t 1.5 nil; NaHCO, 7.5y0, 1
ml; penicillin 200,000 tJ/nil, 1 ml; and EDTA in a
final concentration of 0.4y0) and trimmed of fat
and fibrous tissue. Specimens were then placed in
fresh medium in a petri dish and minced with
scissors. T h e minced t i m e was transferred by
*Shandon Scientific Co, Costa Mesa, Calif.
tHyland Laboratories, Sewickley, Pa.
forceps to sterile test tubes containing fresh medium and, when the desired amount of tissue was
added, a 2.5% solution of trypsint was added to
each tube in a final concentration of 0.25y0. Tubes
were gently inverted several times, placed in a
horizontal position in a 35" C incubator for 15 to 20
min, during which time they were again gently
in\ertetl several times. A t the end of this period,
they were placed in a 35" C water bath in a n
upright position and gross particles were sedimented by gravity. T h e tubes were then allowed to
stand in the bath for the remainder of the
procedure, with occasional inversion and further
sedimentation by gravity, to insure the adequate
dispersion of cells and more complete removal of
gross particl-s. Four to 8 drop-aliquots of the
supernatant cell suspension were transferred with
Pasteur pipettes to sample chambers of the cytocentrifuge and spun at 700 rpm for 7 min. A discrete
layer of cells, 1/4 inch in diameter, was obtained on
a microscope slide. Cytocentrifuge slide preparations
were air-dried for 5 min and then stained with
Riu's stain (20). Depending on the appearance of
these initial slides, the volume of cell suspension
transferred to the cytocentrifuge was varied in the
preparation of subsequent slides. To terminate the
procedure, a 50/, solution of soybean trypsin inhibitor: was added to each tube in a final concentration
of 0.5y0, usually 40 to 50 min after the initial
addition of the trypsin solution. T h e number of
slides derived from each digestion varied according
to the amount of synovium and volume of digest
medium employed. After synovectomy of rheumatoid joints, 2-3g wet weight of tissue were digested
in 15 ml of medium and yielded 35 to 50 slides.
When tissue was obtained by needle biopsy, approximately 100 mg of synovium was added to 3-5 ml of
digest medium and yielded 12 to 16 slides. Slides
were air-dried and stored without fixation at 4 O C.
Stored slides could be kept for a t least four weeks
with excellent preservation of morphology for both
light and fliiorescence microscopy.
Cell counrs. Total cell counts were performed
by supravital staining with 0.1% methylene blue in
saline. Cell viability was determined by the trypan
blue, dye-exclusion method. Differential synovial
cell counts were performed on cytocentrifuged
preparations after applying Wright's stain. T w o
hundred lining cells were counted under oil immersion and classified as type A, type B, or type C
(intermediate) cells.
fworthington Biochemicals Ltd, Freehold, NJ.
Mhritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
Idenilffcailon of lining cell iypes. Cytocentrifuge preparations were employed to measure cell
dimensions with a Leitz-Periplan micrometer
In two instances, differential counts of lining cells
in digest supernatant cell suspensions, made by
light microscopy, were compared with differential
counts obtained independently on the same suspensions under low magnification with an RCA EMU
3C electron microscope.* T h e technique used to
prepare the cells for electron microscopy has been
decribed previously (5).
Sera employed In immunoffuorescent studies.
Individual specimens of serum obtained from 12
children with rheumatoid arthritis (JRA) were
pooled after insuring that each specimen was free
of: a) rheumatoid factor (RF) by testing with both
the slide latex test? and a micro-adaptation of the
sensitized sheep cell agglutination test (SSCA)
(21), and b) antinuclear factor (ANF) by a
standard immunofluorescent technic employing rat
liver as substrate (22). T h e pooled sera were
brought to one-third saturation with ammonium
sulfate. T h e precipitated crude globulin fraction
was conjugated with fluorescein isothiocyanatef
(22), passed through Sephadex G-25, absorbed
twice with tissue powders, and the final protein
concentration adjusted to approximately 1.5gx. A
control pool of serum was similarly prepared from
14 sera obtained from nonrheumatoid pediatric
Human Cohn fraction 11". prepared in 0.5% and
1.x solutions in normal saline, was conjugated
with fluorescein isothiocyanate, passed through
Sephadex G-25, absorbed with tissue powders and
aggregated by heating to 63" C for 10 min (23). Of
five batches of aggregated FII thus prepared, only
LWO provided consistently reproducible results for
the immunofluorescent detection of rheumatoid
Commercial antisera were obtained from several
sources** T h e antisera employed in studies for the
direct immunofluorescent reaction included rabbit
*The electron microscopic counts, utilizing pellets
obtained from the cell suspensions, were performed
by Dr. K. Hirohata.
tRA-Test, Hyland Laboratories, Costa Mesa, Calif.
fBaltimore Biologicals, Ltd, Baltimore, Md.
SPharmacia Ltd, Uppsala, Sweden.
JIHyland Laboratories, Costa Mesa, Calif.
**Hyland Laboratories, Baltimore Biologicals,
Ltd, Pentex Laboratories.
and goat antihuman IgG, and goat antihuman IgA,
IgM, albumin, orosomucoid, Cohn fraction IV-4,
fibrin and the ple/pI. component of complement. I n
addition, antisera to human 1 6 , IgM and the
&./p,. component of complement, conjugated with
lissamine rhodamine B, were obtained. Unconjugated antisera with the same specificities were
obtained from the same sources for use in blocking
procedures and in the indirect immunofluorescent
reaction. A fluorescein-conjugated rabbit antiserum
to the human C l q component of complement was
obtained from Dr. P. Stastny. To reduce nonspecific
staining, all conjugated antisera were diluted with
normal saline from 1:2 to 1:4 and, in some
instances, passed through a column of Sephadex
G-25. T o stain protein polysaccharide (PP-H)
(25). a rabbit antiserum to cartilage PP-H, obtained from Dr. John Sandson, was absorbed twice
with whole human serum before use.
T o insure specificity of the various antisera, each
was permitted to react with its specific antigen,
with fetal calf serum, and with whole serum by
Ouchterloiiy double diffusion in agar (24).
Fluorescent antibody siaining technics. Airdried, cytocentrifuge preparations were used
throughout. No fixative was employed. Slides were
not washed before staining, since preliminary experiments failed to reveal differences in the staining
patterns between washed and unwashed preparations. Initial staining of individual preparations was
always carried out within 48 h r of preparation.
With minor exceptions, conventional methods were
employed in the immunofluorescent staining reactions ( 2 2 ) . Smears were exposed to antisera for 30
min and, after each exposure, were washed for 12
min in two changes of 0.01M phosphate-saline
buffer, p H 7.2.
Combined immunofluorescent staining was employed to demonstrate the concomitant cytoplasmic
localization of various combinations of IgG. IgM or
the p../p,, component of complement (hereinafter
referred to as &).
Slide preparations were first
incubated for 30 min with antiserum conjugated
with rhodamine, washed twice and then processed
in the same manner with antiserum conjugated
with fluorescein.
Specificity of the immunofluorescent reactions was
assured by standard blocking techniques (22), and
by the disappearance of reactivity after the pertinent
antiserum was absorbed with its specific antigen or
whole human serum. As a routine procedure, all
slides were coded, and known control slide preparations containing stainable IgG, IgM and p,. globu-
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 197Oj
lins were included for comparison.
Leitz Ortholux microscope with a dark-field
condenser was employed for fluorescent microscopy.
‘The light source was an Osram HBO-200 lamp; the
exciting filter was a UG-1; and the barrier filters,
depending upon the type of conjugate employed,
ranged from 460K to 510K. Photographs were
taken with a Leitz Orthomat camera, employing
high speed Ektachrome or Anscochrome 500 film
for colored transparencies and Kodak Tri-X film for
hlack and white reproductions.
Cell yields and viabilities. Total cell
counts and the percentage of viable cells
obtained after the synovial specimens were
digested are given in Table 1. Lining
cells comprised a variable fraction of the
total cells present in the digest supernatant
(Table 2), ranging from a mean of 29% in
seronegative rheumatoid arthritis (to SO’jr,
in synovium from patients with ruptured
Identification of individual types of
lining cells. I n addition to white blood
cells, three types of nonhematogenous,
morphologically distinct cells were readily
identified in all digest preparations. As
noted in Table 3, the type A cell is
elongate in shape, has moderate to marked
cytoplasmic basophilia and frequently contains a variety of ingested materials. T h e
nucleus of the type A cell is invariably
Table 2. Percentage of Lining Cells in
Digest Cell Suspensions*
Percentage of
lining cells
No. of
Digest cell suspension patients Mean
Ruptured meniscus
Degenerative joint
Juvenile RA (RF-)t
Adult RA (RF-)
RA (RF+)
Miscellaneous inflamrnatory arthritides#
* Remaining cells
principally WBC.
# Reiter’s syndrome, 1; gonococcal arthritis, 1;
pseudogout, 3.
t Rheumatoid factor
located off-center and contains very dense,
rope-like chroma tin material characteristically clumped at the nuclear membrane
(Fig 1). By contrast, the type B cell is
larger and displays homogeneous, less basophilic cytoplasm; its nuc eus is ro:ind, polar in location and possesses homogeneous,
acidophilic chromatin containing one or
two very prominent nucleoli (Fig 2). T h e
type C cell, intermediate between the tyFe
A and B cells in size and cytoplasmic
characteristics, has a distinctive kidney or
bean-shaped nucleus invariably locatcd at
the cell periphery (Fig 1). T h s relative
Table 1. Total Cell Counts and Percent Cell Viability in Digest Cell Suspension*
Cells/cu mrnt
Ruptured meniscus
Degenerative joint disease
Rheumatoid arthritis$
No. of
* All specimens obtained at arthrotomy.
t Cell counts not corrected for amount of synovium or volume of digest medium employed.
$ JRA (RF-),
5; ARA (RF-),
2; RA (RF+), 5.
Arthritis and Rheumatism, Vol. 13, No. 6 (November-kcember 1970)
numerical distribution of each of these
types of cells in various diagnostic groups
can be seen in Fig 3.
It was found that A cells appeared to
phagocytose not only particulate materials,
such as bacteria, cells, and crystals but by
fluorescent antibody studies, also engaged
in endocytosis of serum a n d synovial fluid
proteins. However, #theendocytosis of these
proteins was highly variable depending
upon the disease state, as will be shown
below. T h e B cells showed evidence of their
known synthetic activities by staining
strongly for cytoplasmic protein polysaccharide [PP-(H)] (Fig 4) with specific fluorescent antibody (25). Thus the C, or intermediate cell, showed evidence of both
phagocptic and synthetic activities.
As can be
Light microscopic studies.
seen in Fig 3, variations in the percentage
distribution of the 3 types of lining c:lh
were observed in the different diagnostic
groups. T h e highest percentage of type A
cells (average 63%) was found i!i the presumably norma! synovia obtained from p ~ tients with ruptured meniscus. By contrast,
in rheumatoid arthritis and the group of
miscellaneous inflammatory arthritides, the
percentage of typ: A ceds decreazed to
42Y0 and 2574, respectively. However, in
both of the latt-r conditions there appeared to be a compensatory increas: in
the percentages of type C cells, which were
3S70 and 557,, respectively, as compared
with the value of 14<,!$ found in the s)novia
of patients with ruptured menisci. T h e
percentage of type B cells remained fairly
constant, a t approximately 20y0 of the total, in all conditions studied, except in
degenerative joint disease where these cells
constituted an average of 38% of the total
lining-cell population.
Although the combined percentage of
potentially phagocytic lining cells reArthritis and Rheumatism, Vol. 13, No. 6 (November-December 1979)
iiiained unchanged in all groups studierl, spect to the observed amomt of in vivo
significant differences were readily seen by phagocytosis of particulate matter. In rheulight microscopy between inflammatory matoid arthritis, gonococcal arthritis, Reitand noninflammatory disease states in re- er's syndrome and pseudogout, 2 0 4 0 % of
Fag 1 (left). Synovial lining cells, DM, juvenile rheumatoid arthritis. Center (upper)-type A cell with
vacuolated, granular cytoplasm. Center (lower)-type C cell with white blood cell inclusion and peripheral
nucleus. (Wright stain,
950) Fig 2 (right). Type B synovial lining cell, AC, Reiter's syndrome. Note
the hmgyneous cytoplasm and polar location of the nucleus. (Wright stain, x 950)
Fig 3. Distribution of differential
counts in various diagnostic groups.
Two hundred lining cells counted for
each patient at magnification of
950 X.
... - -. .
I JRA (RF-I 7'
t ~ E R S
ARA (RF-) 5 I; &.
RA (RF+) 5 .
krthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
Fig 4. Binucleate synovial type 6
cell from patient (AC) with Reiter's
syndrome, showing positive diffuse
staining for the heavy protein-polysaccharide component of cartilage
(PP-H). (X 950)
type A and type C cells contained cells and
cellular debris, while in the synovia from
patients with ruptured meniscus and degenerative joint disease less than 5% of
type A and type C cells containe.1 such
phagocytosed materials.
Multinucleated giant cells (Fig 5) were
commonly observed in R A synovial digests,
but were rare in all other groups. These
cells, which accounted for u p to 5.0o/b of
the total lining cells counted, had a n average diameter of 38.1, f 3.9. T h e cytoplasm
was deeply basophilic, frequently contained phagocytosed cellular debris, and
exhibited from three to ten or more
peripherally-located nuclei, which resembled those of the type A cell. Morphologically and functionally, this cell appeared to
be a multinucleate type A cell.
To establish the relationship between
the classification of synovial cells by their
light microscopic appearance and that pre740
viously developed by Barland and coworkers (I), using the electron microscope, diflerential lining-cell counts were obtained in
two instances on the same cell suspension
by light and electron microscopy. There
was a close correspondence in the differential counts of type A, B and C cells by the
two technics (Table 4).
lmmunofluorescence studies. Synovial
cells of 17 patients with rheumatoid arthritis were tested with a fluoresceinconjugated seronegative adult R A serum
and also with a similarly conjugated pool
of seronegative juvenile RA serum, utilizing the direct immunofluorescent technique. N o staining was observed in the lining cells from any synovium using either
When synovial cell preparations were
stained directly with conjugated antisera to
immunoglobulin G (IgG), immunoglobu-
Arthritis and Rheumatism, Vol. 13.
No. 6 (IYmmber-December 1970)
lin M (IgM) and the C l q and PIC coinponents of complement, two patterns of
cytoplasmic fluorescence, restricted entirely
to the phagocytic type A and C lining cells,
were noted. T h e first of these was a diffuse
staining, which on occasion took on a finely
granular appearance (Fig 6). This will be
referred to as diffuse staining to distinguish
it from the second pattern which was characterized by the presence of comparatively
large and discrete cytoplasmic inclusions.
Applying the criterion that at least 50% of
the 'type A and C cells show 2+ or greater
specific cytoplasmic fluorescence, all of the
seropositive RA patients, all of the adult
seronegative rheumatoid patients, and 5 of
Fig 5. Cluster of A and C cells (RR, seropositive juvenile rheumatoid arthritis) with central multinucleate giant cell containing cellular debris. (Wright stain, x 950)
Table 4.
Comparison of Differential Lining Cell Counts Obtained by
Light and Electron Microscopy
Juvenile rheumatoid arthritis (RF-)
Adult rheumatoid arthritis
* EM = electron microscopy;
Differential cell count
Method of
counting* TypeA
Type B
LM = light microscopy.
t 136 cells counted by EM; 200 cells counted by LM.
3 132 cells counted
by EM; 200 cells counted by LM.
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
7 of .the seronegative juvenile rheumatoid
patients showed diffuse staining for IgG
(Table 5). Similar staining was observed in
the case of only l of the 13 patients with
other arthritides. Weak staining (< 2+), involving less than 10% of the A and C cells,
was observed in the synovia of the 3 patients with ruptured meniscus and 3 of the
Fig 6. Type A cell. Diffuse staining pattern of IgG (DM, seronegative juvenile rhematoid arthritis):
(top) showing diffuse granular staining for IgG; (b3ttm) clgster of A and C cdls with positive diffuse
staining for I&. Note the gradation from diffuse granular (A call, lower left) to diffuse homogeneous
staining. (X 950)
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
Table 5.
Diffuse Fluorescent Staining of Cytoplasm of Type A and Type C
Lining Cells in Various Diagnostic Groups
No. of patients positive*
No. of
Ruptured meniscus
Degenerative joint disease
Miscellaneous arthritides8
RA (RFi-)
R Ft
* Values represent number of patients showing greater than 2+ staining of over 50% of types A and C
lining cells.
t RF = rheumatoid factor.
# Represents sequential staining with rhodamine-conjugated anti-lgG and fluorescein-conjugated
8 Reiter's syndrome, 1; gonococcal arthritis, 3: pseudogout, 1; nail-patella syndrome, 1.
11 Only 6 patients tested, 4 positive.
# Only 4 patients tested, all positive.
4 patients with degenerative joint disease,
but not in the 6 patients with niiicel!an-.ous inflammatory arthritides.
Diffuse staining for IghI o r rheumatoitl
factor (RF) was not observed in any of the
groups studied, including the RF Fosi'ive
patients. Diffuse staining for the p,.. component of complement was not found in
any of the nonrheumatoid patients, but
was found in 15 oE the 17 RA pati-ntj
tested (Fig 7). T h e 2 negative patiznts,
both of whom had juvenile rheumatoid
arthritis, were tlie same 2 patients who had
shown only a weak (negative) st-ining
pattern for IgG. As with IgG, positive
staining for plCappeared not to be related
to the serologic status of the R4 pstients.
I n 5 RA patients who showed diffuse staining for &, diffuse granular cytoplasmic
staining for C l q was also noted in all cases.
When sequential staining with rhodamine-conjugated anti-IgG and fluoresceinconjugated anti-p,, was carried out, 13 of
15 patients with RA showed simultaneous
localization of both IgG and p,, within
individual pliagocytic lining cells (Table 5
and Plate 1A). No patient i n any of the
other disease groups showed this phenomenon. When positive combined staining
occurred, it was exhibited by the vast majority of the phagocytic lining cells, although an occasional cell was found to
contain only IgG.
I n addition to immunoglobulin and
complemmt components, diffuse staining
reactions in phagocytic lining cells were
observed for fibrin, albumin and orosomucoid. These reactions occurred in both
nonrheumatoid and rheumatoi 1 patients,
but no consistent pattern waj observed in
individual patients with respect to the
presence of tlie .e nonimmunoglobulin
serum proteins. With regard to the percent
of cells stained, positive stainirg for fibrin
was relatively infrequent, being obzerved in
less than 10% of the phagocytic cells of
occasional patients in both the rheumatoid
and nonrheumatoid g r o u p . Staining for
serum albumin was found in about 25y0
a n d for orosomucoid in about 50% of the
Mhritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
Fig 7. Diffuse staining of complement components (RR, seropositive juvenile rheumatoid arthritis): (left)
Cluster of A and C cells with positive diffuse staining for p... Note the brightly stained inclusion in the C
cell at lower center. (x 540); (right) Type A cell (DM, seronegative juvenile rheumatoid arthritis) showing
diffuse granular staining for Clq.
phagocytic cells of occasional patients in
both groups. I n contrast to these findings,
in which there appeared to be no preferential distribution in rheumatoid or nonrheuinatoid lining cells, substantial diffuse
staining for IgG and pie was limited to KA
synovial cells.
T h e second pattern of immunofluorescent staining, characterized by the presence
of discrete inclusions, was observed only in
the lining cells oE RF-positive rheumatoid
patients (Table 6). I n these cells, discrete,
and usually large, inclusions were observed
within the cytoplasm of type A and type C
lining cells and also within the cytoplasm
of polymorphonuclear cells obtained in the
digests. Lining cells from 4 of the 5 RFpositive patients examined contained such
inclusions. T h e single patient lacking inclusions did, however, display diffuse staining. In these patients a spectrum of stain744
ing reactions was observed i n individual
phagocytic cells, which ranged from either
diffuse or discrete staining (Plate 1B) to
both diffuse and discrete staining patterns
for IgG and PIC(Fig 7) in the same cells
T h e composition of these inclusions, as
determined by appropriate specific immunofluorescent staining, varied within individual patients (Table 6) and within indi.
vidual cel's. Inclusions containing either
IgG or IgM were present in 4 of the 5
rheumatoid factor-positive patients studied
(Fig 8). Inclusions containing both IgG
and IgM (Plate 1C) were found in all of 3
patients studied by the combined staining
method. Rheumatoid factor-containing inclusions were also identified in 3 of the 5
patients. T w o also demonstrated inclusions
which gave combined staining for IgG and
file (Plate 1D). Although all inclusions in
seropositive patients contained IgG, and
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
Plate 1A. Combined staining of synovial cells (RA, seropositive juvenile rheumatoid arthritis) with rhodamine anti-lgG and FlTC anti-p,.. A cluster of A and C
cells shows diffuse yellow-green stain suggesting presence of lgG-pl.. (x 540)
Plate 1B. Combined staining of synovial cells (RR, seropositive juvenile rheumatoid arthritis) using rhodamine anti-lgG and FlTC anti-,%.. On left, an A cell
shows diffuse yellow-green stain suggesting presence of IgG-pl.. On right is an
A or C cell with discrete orange (16-containing) inclusions. (X 540) Plate 1C.
Combined staining of a polymorphonuclear cell from synovial digest (SP, seropositive adult rheumatoid arthritis) with rhodamine anti-lgG and FlTC anti-lgM.
Yellow-green (IgG-lgM) inclusions are seen. (X 950) Plate 1D. Combined
staining of synovial lining cells (SP, seropositive adult rheumatoid arthritis) with
rhodamine anti-I& and FlTC antig.. At left, a C cell with solitary orange (IgG)
and yellow-green (IgG-p,,) inclusions. At right, a C cell with solitary orange,
yellow-green and green @,J inclusions. (X 950)
most contained IgM, occasional inclusions
were observed, which failed to stain for
IgM upon combined staining for IgG and
IgM. I n the 2 patients in whom ple-containing inclusions were found, the frequency of IgG positive, IgM negative inclusions
appeared to be equal to that of IgG positive, plCpositive inclusions, suggesting that
in these patients the PICcomponent was
attached to IgG, which was not complexed
with IgM. One synovium from a seropositive patient, which contained many inclusions of both the IgG-IgM and IgGp,,
varieties, showed rare inclusions staining
positively for both IgM and pic. In no
instance were the discrete inclusions found
to stain positively for seruu albumin, fibrin
or orosomucoid.
Trypsinization of surgical and biopsy
specimens of human synovium and concentration of the cell suspensions by a cytocentrifuge has provided a simple and reproducible method by which synovial lining
cells can be isolated for study by a variety
oE microscopic techniques. Williamson and
coworkers (17), employing the method of
Fraser and Catt (16), for the trypsinization
of synovial tissue from normal cadaver
joints, have delineated a lining-cell population similar to that described in the present
report, although t h e e authors did not specifically define the intermediate or type C
Light (17) and electron microscopic investigations (1,3) and fluorescent antibody
studies of the present report have demonstrated that three types of synovial lining
cells subserve two, clearly-distinct functions: phagocytosis, in the case of the A and
C cells (1, S), and synthesis of a protein
antigenically related to hyaluronateprotein in the case of the B and C cells
(25). Thus, the intermediate or type C
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
cell, which possesses both of these specialized functions, appears to be a welldifferentiated cell, which upon appropriate
stimulation can proliferate to provide increased numbers of cells for either phagocytosis or hyaluronate synthesis. Some support for this suggestion is provided by the
differential counts of cell suspensions from
patients with rheumatoid arthritis and the
miscellaneous group of inHammatory arthritides studied. I n these, a decrease in the
percentage of phagocytic type A cells was
invariably associated with an increase in
the percentage of type C cells actively engaged in phagocytosis.
When a fluorescein-conjugated pool of
Fig 8. Inclusions in lining cells of seropositive patient with adult rheumatoid arthritis (SP). (Top) Upper
left: Type A cell showing numerous discrete inclusions staining for IgG. Lower right: Shows polymorphonuclear cell with numerous discrete inclusions staining for IgG. (Bottom) Type A cell showing large discrete
inclusions staining positively for IgM. (x 950)
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
serum, obtained from a group of seronegative juvenile rheumatoid patients, and a
fluorescein-conjugated serum from a single
seronegative adult rheumatoid patient
were applied to rheumatoid synovial cell
preparations in an attempt to demonstrate
a specific "rheumatoid" antigen in isolated
lining cells, no specific staining was observed. It is possible, of course, that this
negative result is not conclusive, since free
antibody might not occur in the circulation
of patients with RA if an antigen were
present in the synovium in excess amount.
Conversely, if an excess of antibody were
present in the rheumatoid synovial
effusion, the antibody combining sites of
the hypothetical synovial antigen would be
saturated with antibody in vivo.
Two groups of investigators (14, 15)
have demonstrated, by immunofluorescent
technics, that immunoglobulins and complement components are deposited in rheumatoid, but not in nonrheumatoid, synovia1 membranes. However, the use of frozen
sections of synovium in both of the studies
cited has prevented definitive identification
of the cells that were the source of the
positive fluorescence. In the present investigation, the availability of cell suspensions
obtained by trypsin digestion of synovial
tissue has provided a unique opportunity
to study isolated lining layer cells. These
studies have shown that in RA, the phagocytic lining cells of the rheumatoid synovium engage in the in vivo uptake of immunoglobulins and complement components.
Specifically, it has been possible to delineate two morphologic patterns of phagocytosis of proteins, the first characterized by
a diffuse or diffusely granular pattern of
fluorescence for IgG and ple common to
virtually all patients with R A regardless of
serologic reaction for rheumatoid factor,
and the second, characterized by the
presence of large inclusions usually con-
taining IgM or RF, found only in seropositive rheumatoid patients.
Though diffuse cytoplasmic fluorescence
specific for IgG was observed in the cytoplasm of phagocytic lining cells in nonrheumatoid patients, it was of very low intensity and involved few cells. In the patients
with Reiter's syndrome and gonococcal arthritis, it was virtually absent. I n approximately two thirds of the group of patients
with ruptured meniscus and degenerative
joint disease, not more than 10% of the
phagocytic cells showed diffuse fluorescence
and the intensity of staining was weak,-ie,
1+ to 2+. In the rheumatoid patient-group,
on the other hand, almost all phagocytic
cells showed 3+ to 4+ staining. Furthermore, file staining almost uniformly accompanied the IgG staining in the rheumatoid
synovia, but was not observed in the patients with nonrheumatoid arthritides.
The immunofluorescent localization in
phagocytic lining cells of the Plc or C3
component of complement, and also the
Clq component of complement, clearly delineated the RA patients from all others
studied. T h e localization of complement
components in the cytoplasm was associated with the presence of active rheumatoid synovitis whether juvenile or adult
and whether seronegative or seropositive.
T h e common pattern of staining for the
Ple component was diffuse and/or diffusely
granular. Sequential staining for both IgG
and plC invariably revealed that both proteins had the same diffuse distribution
within individual lining cells and that in
no instance was plC found alone within
synovial lining cells. Thus, the synovitis of
RA appears to be characterized by a widespread and concurrent deposition of IgG
and complement in most of the phagocytic
lining cells of the synovial lining layer.
It seems unlikely that the simultaneous
uptake of IgG and complement components
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
represents independent phagocytosis of
these substances by rheumatoid lining cells.
I n the patients with miscellaneous infiammatory arthritides, phagocytosis of particulate material such as cells and crystals was
at least as prominent as in the RA group,
yet the lining cells of these patients showed
virtually no phagocytosis of IgG and none
of the plc component. Furthermore, although some of the patients in the r u p
tured meniscus and degenerative joint disease groups showed minimal phagocytosis
of IgG, none showed evidence of phagocytosis of .the plCcomponent. Thus, in the
patients -with rheumatoid arthritis, immunofluorescent staining for the Ple component appears to be specifically linked to
the phagocytosis of IgG and argues that
these are taken u p together in the form of
a complex.
A number of studies (15,26,27,28) have
demonstrated that the synovial fluid complement level of many patients with RA is
lower than in nonrheumatoid patients with
inflammatory arthritis, suggesting that complement is being fixed in the rheumatoid
joint. T h e demonstration in the present
studies of the simultaneous in vivo uptake
of complement components and IgG i n 15
of 17 rheumatoid synovia suggests that this
may be the mechanism responsible for the
lowering of synovial fluid complement
levels in rheumatoid arthritis.
IgM and R F were found only in large,
discrete IgG-containing inclusions and never in the diffuse pattern observed with IgG
and PIC. A similar observation has been
previously reported (14). Inclusions of this
type -were detected within the phagocytic
lining cells of seropositive patients exclusively. All patients who exhibited the characteristic IgM and RF containing inclusions also showed the diffuse staining produced by IgG in combination with Pie,
often within the same cell. Thus, the
present results show a more constant association of rheumatoid synovitis with the
presence of IgG and plc in the cytoplasm
than with IgG-RF inclusions, suggesting
that the rheumatoid factor complex is not
a n essential ingredient in the pathogenesis
of rheumatoid synovitis.
T h e nature of the IgG distributed diffusely with PICin the RA lining cells has
not been established by these studies, but
two possible alternatives exist. On the one
hand, the IgG may be the antibody component of a classical immune complex containing a specific, but unidentified antigen.
On the other hand, the IgG may be present
in an aggregated or altered form. This
form of IgG is capable of fixing complement (29) and substantial amounts of aggregated IgG possessing rheumatoid factor
activity have, in fact, been demonstrated in
rheumatoid synovial effusions in contrast to
nonrheumatoid effusions (30,31). Thus, it
is possible that the diffuse IgG-plc staining
pattern represents a complex of IgG rheumatoid factor with autologous IgG and
complement components.
T h e second pattern of staining was characterized by the presence of large, discrete
inclusions within the cytoplasm of p h a g e
cytic lining cells, made u p of IgG in combination with IgM but also showing in 2 of 4
cases combined staining for IgG and pie.
Although the combined staining data do
not prove this conclusively, it seems likely
that IgG-IgM-p,, complexes were present.
T h e observation that the large, discrete inclusions contained IgG in combination with
IgM and with PICalso suggests that the IgC
in these inclusions was in a n altered form
since altered IgG, in addition to possessing
the capacity to bind complement, also
reacts with IgM rheumatoid factor to produce an insoluble complex.
T h e extensive and widespread deposition of IgG with fllC in the phagocytic
Arthritis and Rheumatism, Vol. 13, No, 6 (November-Deember 1970)
lining cells of 15 of 17 patients with rheumatoid arthritis, whether seropositive or
seronegative, suggests that complexes containing these two components are present
in the synovial fluid of these patients and
that they are phagocytosed as such by these
cells. It is, in fact, possible that this process
may trigger rheumatoid synovial inflammation through labilization of the lysosomes
of the synovial A and C cells leading to
chemotaxis of polymorphonuclear cells
and, in turn, to further propagation of
synovial inflammation as a result of the
release of lysosomal enzymes and other
inflammatory constituents from the latter
cells (32).
T h e present observations are compatible
with the possibility that either a n immune
complex containing IgG and bound complement or a complex of IgG with IgG and
complement (33), may mediate rheumatoid synovitis by a direct effect on the
synovium. T h e recent evidence cited
above, which indicates that the rheumatoid
synovial effusion may be uniquely characterized by the presence of aggregated IgG,
lends support to the latter interpretation of
the results of the present studies. From the
present observations, it would also appear
that the presence of IgM rheumatoid factor
is not necessary for the propagation of this
Synovial lining cells were separated from
human synovial membrane by trypsinization in uitro. Aliquots of the resulting cell
suspensions were cytocentrifuged to produce discrete monolayers which have been
examined by light and fluorescence microscopy. Phagocytic Type A cells, nonphagocytic type B cells and intermediate type C
cells, have been identified and characterized. Rheumatoid arthritis and other
inflammatory arthritides were characterized by a fall in the percentage of A cells
and a marked increase in the percentage of
intermediate type C cells.
T h e rheumatoid synovial membrane
showed morphologic evidence of active in
vivo phagocytosis of particulate matter, immunoglobulins and complement components. Two patterns of intracytoplasmic
fluorescence were observed in phagocytic
lining cells. I n the more common pattern,
diffusely granular staining for IgG and PIC,
apparently i n the form of a complex, was
observed. This was found in 15 of 17 rheumatoid synovia from both seropositive and
seronegative R A patients. T h e second pattern, observed only in seropositive patients,
showed large, discrete inclusions, the majority of which contained IgM or Ighfrheumatoid factor in combination with
IgG. Inclusions also stained positively for
PICin some patients. Neither pattern was
detected in nonrheumatoid cells. Although
the nature of the IgG in the diffusely
distributed IgG -pic pattern was not tletermined, a complex of this type may trigger
rheumatoid synovitis by effecting lysis of
phagocytic lining cells.
After this manuscript was submitted, a
paper appeared by Tursi et al: Clin Exp
Immun 6:767, 1970, describing IgG-pl, and
IgG-rheumatoid factor complexes in the
synovial membrane of patients with rheumatoid arthritis.
The authors express their appreciation lor
the generous cooperation extended by Dr. Chester Fink, Dr. Louis Paradies and the Staff of
the Department of Orthopedic Surgery at Texas
Scottish Rite Hospital, Parkland Atelnorial
Hospital and St. Paul Hospital, Dallas.
1. Barland P, Novikoff AB, Hamerman D:
Fine structure and cytochemistry of the
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
rheumatoid synovial membrane, with special reference to lysosomes. Amer J Path
449353, 1964
2. Wyllie JC, Haust MD, More RH: T h e fine
structure of synovial lining cells in rheumatoid arthritis. Lab Invest 15:519, 1966
3. Norton WL, Ziff M: Electron microscopic
observations on the rheumatoid synovial
membrane. Arthritis Rheum 9:589, 1966
4. Ghadially FN, Roy S: Ultrastructure of
synovial membrane in rheumatoid arthritis.
Ann Rheum Dis 26:426, 1967
5. Hirohata K, Kobayashi I: Fine structures
of the synovial tissues in rheumatoid arthritis. Kobe J Med Sci 10:195, 1964
6. Parker RL, Schmid FR: Phagocytosis of
particulate complexes of -,,-globulin and
rheumatoid factor. J Immunol 88:519,
7. Hollander JL, McCarty DJ, Astorga G,
et al: Studies on the pathogenesis of rheumatoid joint inflammation. I. T h e RA cell
and a working hypothesis. Ann Int Med
62:271, 1965
8. Rawson AJ, Abelson MM, Hollander JL:
Studies on the pathogenesis of rheumatoid
joint inflammation. 11. Intracytoplasmic
particulate complexes in rheumatoid synovial fluids. Ann Intern Med 62281, 1965
9. Restifo RA, Lussier AJ, Rawson AJ, et al:
Studies on the pathogenesis of rheumatoid
joint inflammation. 111. T h e experimental
production of arthritis by the intraarticular
injection of purified 7s gamma globulin.
Ann Intern Med 62285, 1965
10. Vaughan JH, Barnett EV, Sobel MV, et al:
Intracytoplasmic inclusions of immunoglobulins in rheumatoid arthritis and other
diseases. Arthritis Rheum 11:125, 1968
11. Vaughan JH, Jacox RF, Noel1 P: Relation
of intracytoplasmic inclusions in joint fluid
leukocytes to antiyG globulins. Arthritis
Rheum 11:135, 1968
12. Kaplan MH, Vaughan J: Cbnference on
host response mechanisms in rheumatoid
arthritis, Arthritis and Rheufnatism Foundation Conference Series. Vol 6. 1962, p 57
13. Kaplan MH: Conference on the immunologic aspects of rheumatoid arthritis and
systemic lupus erythematosus.12
14. Rodman WS, Williams RC, Bilka PJ, et al:
Immunofluorescent localization of the third
and fourth component of complement in
synovial tissue from patients with rheumatoid arthritis. J Lab Clin Med 69:141, 1967
15. Fish AJ, Michael AF, Gewurz H , et al:
Immunopathologic changes in rheumatoid
arthritis synovium. Arthritis Rheum 9267,
16. Fraser JRE, Catt KJ: Human synovial cell
culture. Use of a new method in a study
of rheumatoid arthritis. Lancet 21437,
17. Williamson M, James K, Ling MR, et al:
A study of the morphology and a n examination of protein synthesis of synovial
cells. Ann Rheum Dis 25:534, 1966
18. Parker RH, Pearson CM: A simplified
synovial biopsy needle. Arthritis Rheum
6:172, 1963
19. Ropes MW, Bennet GA, Cobb S, et al:
Diagnostic criteria for rheumatoid arthritis.
Bull Rheum Dis 9:175, 1958
20. Yam LT: A rapid method of preparing
smears of effusions and solid mass aspirates
for cytologic diagnosis. Amer J Clin Path
47:797, 1967
21. Sever JL: Application of a microtechnique
to viral serological investigations. J Immunol 88:320, 1962
22. Holborow EJ, Johnson GD: Handbook of
Experimental Immunology. Edited by DM
Weir. Philadelphia, FA Davis Co, 1967
23. Mellors RC, Heimer R, Corcos J, et al:
Cellular origin of rheumatoid factor. J Exp
Med 110:875, 1959
24. Ouchterlony 0: Diffusion-in-gel methods
for immunological analysis. Prog Allergy
5:1, 1958
25. Blau S, Janis R, Hamerman D, et al: Cellular origin of hyaluronate protein in the
human synovial membrane. Science 150:
353, 1965
26. Hedberg H: Studies on the depressed
hemolytic complement activity of synovial
fluid in adult rheumatoid arthritis. Acta
Rheum Scand 9:165, 1963
27. Pekin TJ, Zvaifler NJ: Hemolytic comple-
Arthritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
ment in synovial fluid. J Clin Invest 43:
1372, 1964
28. Fostiropoulos G, Austen KF, Bloch KJ:
Total hemolytic complement and second
component of complement activity in serum
and synovial fluid. Arthritis Rheum 8:219,
29. Christian CL: Studies on aggregated gamma globulin. I. Sedimentation and electrophoretic properties. J Jmmun 84:112, 1960
30. Hannestad K: Presence of aggregated yGglobulin in certain rheumatoid synovial
effusions. Clin Exp Immun 2:511, 1967
31. Winchester R, Kunkel HG: High molecular weight complexes involving y-globulin
in joint fluids from rheumatoid arthritis
patients. Arthritis Rheum 11:849, 1968,
32. Weissmann G: Lysosomes and joint disease.
Arthritis Rheum 9:834, 1966
33. Winchester RJ, Agnello V, Kunkel HG:
T h e joint fluid yG-globulin complexes and
their relationship to intraarticular complement diminution. Ann NY Acad Med
(in press)
Mhritis and Rheumatism, Vol. 13, No. 6 (November-December 1970)
Без категории
Размер файла
3 521 Кб
nonrheumatoid, isolated, lining, membranes, synovial, studies, rheumatoid, cells
Пожаловаться на содержимое документа