вход по аккаунту


Light microscopic characterization of the fibroblast-like synovial intimal cell synoviocyte.

код для вставкиСкачать
Objective. To reassess synovial intimal cell populations by light microscopy.
Methods. Non-inflamed, rheumatoid and osteoarthritic synovia were analyzed as tissue sections and
cytospin preparations by a series of combined immunohistochemical and cytochemical staining techniques.
Results. Two populations of intimal cells were
identified. The first carried macrophage markers. The
second showed high uridine diphosphoglucose dehydrogenase (UDPGD) activity, minimal cytoplasmic CD68,
absent non-specific esterase (NSE) activity, and absent
leukocyte and endothelial antigens. The majority of
these cells showed a high content of prolyl hydroxylase.
Conclusion. Combined cytochemical staining for
NSE and UDPGD activity allows effective separation of
intimal cell populations. We suggest that the cells of high
UDPGD activity are the fibroblast-like or type B synovial intimal cells defined by electron microscopy. High
UDPGD activity probably reflects a preferential ability
to synthesize glycosaminoglycans, including hyaluronan.
The identity of synovial lining cells has long
been a matter of debate ( I ) . Following electron microFrom the Synovial Biology Group. Department of Rhetimatology Research. University College and Middlesex School of
Medicine, London, England.
Supported by the Arthritis and Rheumatism Council of
Great Britain.
Linda S. Wilkinson. MPhil: Senior Research Fellow; Andrew A. Pitsillides, PhD: Postdoctoral Fellow; Jennifer G . Worrall.
MRCP: Honorary Clinical Lecturer; Jo C. W. Edwards. MD.
FRCP: Senior Lecturer.
Address reprint requests to Jo C. W. Edwards. MD. FRCP.
Department of Rheumatology Research. Arthur Stanley House.
Tottenham Street, London WIP 9PG. UK.
Submitted for publication January 21, 1992; accepted in
revised form May 5 , 1992.
Arthritis and Rheumatism, Vol. 35, No. 10 (October 1992)
scopic studies by Barland et al ( 2 ) , evicdnce has
accumulated that the synovial intima contains two cell
populations with different origins and functions (3).
The type A cell appears to be a mature monocytederived macrophage (4-8). The type B cell resembles a
fibroblast (2), but its relationship to other fibroblasts is
not clear. Synthesis of synovial fluid hyaluronan has
been ascribed to the type B lining cell, and these cells
may be responsible for the relatively high proportion
of hyaluronan in the glycosaminoglycans produced by
synovial “fibroblasts” in culture (9,lO).
Analysis of the behavior of the type B cell has
been hindered by difficulty in distinguishing these cells
from subintimal fibroblasts in cultures of synovial
cells. Stevens and coworkers ( I I ) found that monoclonal antibody 67 bound to the surface of type B cells in
tissue sections, but use of this antibody to identify
type B cells in culture has not been reported.
The activity of a number of enzymes has been
documented in synovial lining cells by quantitative
cytochemistry (12). However, these studies do not
distinguish between type A and type B cells. Moreover, no formal comparison has been made between
activity in type B synovial lining cells and fibroblasts
in the subintima. Mehdizadeh et al (13) have recently
described an improved cytochemical method for assessing uridine diphosphoglucose dehydrogenase
(UDPGD) activity. UDPGD is responsible for the
synthesis of UDP-glucuronate, which is then copolymerized with UDP-N-acetylglucosamine to form hyaluronan (14). In initial studies we have reported the
findings of (a) high UDPGD activity in lining cells of
non-inflamed synovium, as compared with subintimal
cells, and (b) a relative reduction of UDPGD activity
in lining cells in disease, as compared with non-
inflamed tissue (15). T h e latter observation suggests
that UDPGD activity is not a marker of non-specific
activation, but an indication of a specialized resting
By a series of double-labeling experiments, we
sought t o establish whether high UDPGD activity in
synovium is restricted t o fibroblast-like cells. Macrophages were identified using (a) antibodies t o CD68,
which is present in all macrophages (16) but may occur
in other cells in diseased tissue, a n d (b) non-specific
esterase (NSE) activity, which marks a restricted
macrophage population in synovium ( 5 ) but gives clear
cytoplasmic staining in intimal macrophages. Prolyl
hydroxylase is involved in collagen synthesis, and t h e
antibody 5B5 (17), which is directed against t h e p
enzyme subunit, was used as t h e most likely marker t o
distinguish fibroblasts within t h e intima.
Tissue samples. Non-inflamed synovium was obtained from the limbs of 5 patients undergoing amputation
for sarcomata. Tissues were taken from clinically normal
areas at least 20 cm from the tumor site. The samples
showed no histologic evidence of disease. Inflamed synovium was obtained (within 20 minutes of removal) from
patients with osteoarthritis or rheumatoid arthritis (definite
or classic, according to the criteria of the American College
of Rheumatology [formerly, the American Rheumatism Association] [ 181) who were undergoing arthroplasty.
Synovia were trimmed of fat, dissected into 3-5-mm
pieces, snap-frozen in n-hexane at -7O"C, and stored at
-70°C until required for sectioning. Seven-micron sections
were cut at -35°C on a Slee cryostat. Adjacent samples of
diseased synovia were minced with scissors and incubated
for 1 hour at 37°C in RPMI (Gibco, Grand Island, NY)
containing collagenase (Worthington, Freehold, NJ) at 30
mg/gm of tissue. The digest was sieved, centrifuged, washed
twice in RPMI, and cytospun in a Cytospin 2 (Shandon,
Pittsburgh, PA) at 800 revolutions per minute for 5 minutes
at a cell density of 2 x lO'/ml. Cytospin preparations were
air dried, and tested immediately for UDPGD activity.
UDPGD activity. UDPGD activity was assessed using the method of Mehdizadeh et a1 (13). Freshly cut 7 p
sections or freshly prepared cytospin preparations were
incubated for 20 minutes at 37°C with a reaction medium
consisting of UDP and nicotinamide adenine dinucleotide in
glycyl glycine buffer at pH 7.8, containing 30% polyvinyl
alcohol and saturated with nitrogen, to which nitroblue
tetrazolium was finally added. Control preparations were
incubated with substrate-free medium. Sections were
washed in water, and mounted in Aquamount (BDH, Poole,
UK) or, for subsequent immunostaining, were washed in
phosphate buffered saline (PBS). Preparations for double
labeling were fixed in acetone for 5 minutes before reacting
for enzyme activity. Unfixed controls ensured that the
activity was not lost during fixation.
Immunohistochemistry. The following monoclonal
antibodies were used: EBMl 1 (anti-CD68) for macrophages
(16) at 1 :ZOO (Dako, Burlingame, CA); 5B5 (anti-prolyl
hydroxylase) for fibroblasts (17) at 1 5 0 (Dako); 4KB128
(anti-CD22) for B cells (19) at 1:200 (from Dr. P. Lydyard,
University College, London, UK); T3-4B5 (anti-CD3) for T
cells (20) at 1 : 10 (Dako); NP57 (anti-neutrophil elastase) for
neutrophils at 1: 100 (Dako); EN4 for endothelial cells at 1:5
(Sera Lab, Oxford, UK); W632 (anti-class I major histocompatibility complex [MHC]) for all nucleated cells used neat
(from Dr. P. Beverley, The Imperial Cancer Research Fund,
London, UK). Antibodies were used in native form or were
biotinylated for double labeling.
Binding of monoclonal antibodies was visualized
using either goat anti-mouse IgG-fluorescein isothiocyanate
(FITC) conjugate (Sera Lab) at 1:20 in PBS plus 20% normal
human serum (NHS) or, for biotinylated monoclonals,
streptavidin-tetramethylrhodamine isothiocyanate (TRITC)
conjugate (Sera Lab) at 1:20 in PBS plus 20% NHS or, for
permanent enzymatic preparations, rabbit anti-mouse Ig
(Dako) 120 in Tris buffered saline (TBS) plus 20% NHS and
alkaline phosphatase-anti-alkaline phosphatase (APAAP;
Dako) at 150 in TBS plus 20% NHS.
Sections were incubated with primary reagents for 1
hour at room temperature; all subsequent incubations were
for 30 minutes. For immunofluorescence, incubations were
in PBS at pH 7.6, and for the APAAP technique, in TBS at
pH 7.8. Specimens for immunofluorescence were mounted
in glycerol containing 1,4-diazabicyclo[2.2.2]octane(Sigma,
Poole, UK), and specimens stained with APAAP were
mounted in Aquamount (BDH).
NSE activity. NSE activity was assessed using
a-naphthyl acetate as described previously (5). When NSE
assessment was combined with UDPGD assessment, the
UDPGD reaction was performed first. When combined with
immunochemistry, the NSE assessment was performed first.
Double labeling. All double-labeling procedures were
controlled by comparison with single-labeled serial sections.
Assessment of UDPGD activity was combined with antiCD68, anti-prolyl hydroxylase, class I MHC, and a cocktail
of antibodies (all at their optimal dilutions) directed against
CD3 (T cells), CD22 (B cells), neutrophil elastase, and EN4
(endothelial cells). A fluoresceinated second antibody was
used. For double immunofluorescence, preparations were
incubated successively with one primary antibody, antimouse IgG-FITC conjugate, 20% normal mouse serum, the
second (biotinylated) primary antibody, and streptavidinTRITC conjugate.
Assessment of labeling. Stringent criteria were applied to the assessment of intimal staining. Negativity was
defined as no cytoplasmic staining, together with no staining
outlining the cell other than that explained by staining of the
adjacent cell membranes (in practice, no complete rim of
staining). Positivity required either cytoplasmic staining or a
complete rim of staining. Interpretation was occasionally
difficult, in that cells surrounded by other cells with positive
rim staining could only be judged positive, and this may have
inflated the number of cells showing unexpected staining
combinations. Cytospin preparations were used to assess
whether double labeling was maintained when cells were
physically separated. However, interpretation of disaggre-
Figure 1. Non-inflamed synovium. showing high uridine diphosphoglucose dehydrogendse dctivity confined to cells of the lining
ldyer (original magnification x M)O)
gated cell populations was made in the knowledge that both
intimal and subintimal cells were present. In sections, double labeling was assessed on at least 50 consecutive adjacent
intimal cells of high UDPGD activity. read from a randomly
chosen point in the intima. In cytospins. 100 cells of high
UDPGD activity were counted.
UDPGD activity. I n non-inflamed tissues, a majority of intimal cells showed high UDPGD activity
(Figure I). Subintimal cells showed minimal activity.
Very low activity was seen in some vascular cells. At
low magnification, UDPGD reaction product was seen
as a single line following the contour of the intima.
Figure 2. Rheumatoid \ynovium. 4howing an area retaining cells
with moderate uridine diphosphoglucose dehydrogenase activity in
the basal part of the lining layer. overlaid by cells of low activity
(original magnification x 2 0 0 ) .
Figure 3. Cytospin preparation of osteoarthritic synovium double
labeled for uridine diphosphoglucose dehydrogenase (UDPGD) activity (bluish purple) and CD68 (red). Two distinct populations are
\een. Many of the cells are still in sheets or clusters, with the
branching processes of UDPGD-rich cells encircling CD68-positive
cell5 (original magnification x 400).
In diseased tissue, a lower proportion of cells of
the thickened intima showed high UDPGD activity,
and the contrast between activity in these cells and
that in subintimal cells was less marked in rheumatoid
samples. In osteoarthritic samples, UDPGD activity
varied. with some areas within a sample showing
activity similar to that of non-inflamed tissue and other
areas showing only moderate activity. In diseased
tissue, cells with high UDPGD activity were typically
found beneath the most superficial intimal cells (Figure
Figure 4. Rheumatoid synovium double labeled for uridine diphosphoglucose dehydrogenase (UDPGD) activity (bluish purple) and
nonspecific esterdse (NSE) (brick red). As in Figure 2, UDPGD
activity i h seen in the basal part of the lining. whereas the most
superficial lining cells show high NSE activity, which is characteristic of mature macrophages (original magnification x 100).
Table 1. Percentage of cells with high uridine diphosphoglucose
dehydrogenase activity double labeling for CD68 and for prolyl
hydroxylase (PH)
Tissue sections
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
Sample 6
Sample 7
Sample 8
Sample 9
Sample 10
Sample I I
Sample 12
Sample 13
Sample 14
Sample 15
Sample 16
Sample 17
Sample 18
Sample 19
% of cells
R of cells
PH +
Cytospin preparations
% of cells
% of cells
PH +
* Staining largely confined to filopodia.
t Extremely variable staining intensity.
2), in a zone known to lack macrophage markers
In cytospin preparations, a proportion of cells
showed high UDPGD activity. These cells often separated in sheets, which were held together by a meshwork of processes (Figure 3). These processes enclosed rounded cells of low UDPGD activity carrying
CD68 (see below). They appeared, as would be expected, in an en face display of the intima, which
suggests that they corresponded to cells of similar
activity seen at the intima of tissue sections.
Findings of double-labeling studies. Evidence for
the identity of UDPGD-rich intimal cells was adduced
from a series of double-labeling experiments. All cells
stained for class I MHC molecules, whether or not
UDPGD assessment preceded this step, which confirmed that the UDPGD reaction product did not
significantly quench FITC.
UDPGD/NSE staining. We have previously
demonstrated that macrophages in the intima show
high NSE activity (5). NSE gives a cytoplasmic reaction product, overcoming the difficulties associated
with CD68 on cell peripheries. Double labeling for
UDPGD and NSE showed a clear segregation of cells
(Figure 4). No cells were convincingly positive with
both markers. Single-labeled serial sections and parallel cytospin preparations confirmed that the initial
UDPGD incubation did not alter NSE activity.
UDPGD/CD68 staining. In non-inflamed tissue,
cells with high UDPGD activity and cells positive for
CD68 formed distinct intimal subpopulations. In diseased tissue, the distinction was largely maintained,
but in some areas, there was overlap of UDPGD
activity and CD68. Table 1 shows the percentages of
cells with high UDPGD activity which double labeled
for CD68. In cytospin preparations from inflamed
tissue, cells could be separated consistently into 1 of
the following 3 groups: group 1, CD68-negative or dull
cells of high UDPGD activity with branching processes; group 2, CD68-negative or dull cells of low
UDPGD activity and less branching than group I ; or
group 3, rounded CD68-positive cells of minimal
UDPGD activity.
Although a minority of cells with high UDPGD
activity showed staining for CD68, this was qualitatively different from that seen on rounded, strongly
CD68-positive cells. CD68 staining of cells of high
UDPGD activity occurred chiefly on filopodia, with
only a few flecks of cytoplasmic staining. For all
labeled cells, the perinuclear cytoplasm was dominated by one or the other label, which suggests that
there are 2 discrete cell types showing minor convergence, rather than a spectrum with intermediate forms
(as previously suggested based on the ultrastructure
[ 2 2 ] ) .Group 2 cells showed staining similar to that of
subintimal fibroblasts seen in tissue sections.
In view of the cytospin findings, the overlap of
UDPGD activity and CD68 staining in some tissue
sections was judged consistent with the known problems of identifying individual cells in close apposition
within sections and the presence of CD68 on fibroblast
UDPGD/prolyl hydroxylase staining. Pro1yl
hydroxylase-positive cells were present throughout
tissue sections. In non-inflamed tissue, the strongest
staining was in intimal cells, but in inflamed tissue,
prominent staining was seen in all areas. In many
sections, double labeling of intimal cells was virtually
complete, in contrast to the subintima, where there
were many prolyl hydroxylase-positive cells with low
or absent UDPGD activity. In other sections, only
60-70% of cells of high UDPGD activity were strongly
positive for prolyl hydroxylase (Table I). Marked
variation in prolyl hydroxylase staining intensity was
seen at the intima of inflamed tissues, particularly in
the rheumatoid samples. Cells with the highest
UDPGD activity did not necessarily correspond to
those with the strongest prolyl hydroxylase staining.
Cytospin preparations confirmed that there was considerable independent variation of the two enzymes.
UDPGDIother markers. Cells with high UDPGD
activity did not stain with the cocktail of antibodies to
leukocyte and endothelial antigens, either in the tissue
sections or the cytospin preparations.
Prolyl hydroxylase/CD68 and NSE staining.
Double labeling was performed to check segregation of
prolyl hydroxylase and CD68 or NSE, in view of the
apparent overlap of CD68 and UDPGD staining. Overlap of the macrophage markers with prolyl hydroxylase was similar to that seen for UDPGD activity. In
tissue sections, there appeared to be some prolyl
hydroxylase-positive cells surrounded by CD68positive cells, but in cytospin preparations, cells were
clearly segregated, except for some prolyl hydroxylasepositive cells showing peripheral CD68 staining. NSE
activity and prolyl hydroxylase were clearly segregated in both preparations.
We conclude from these experiments that
synovial intimal cells can be segregated into macrophages or fibroblast-like cells by light microscopy.
High activity of UDPGD appears to be a useful
marker, within synovium, for fibroblast-like intimal
cells. UDPGD activity appears to be maintained in
fresh cytospin preparations, and may allow separation
of intimal and subintimal fibroblastic cells in vitro.
Anti-prolyl hydroxylase also labels intimal fibroblastlike cells but will not distinguish these from subintimal
fibroblasts. Moreover, prolyl hydroxylase content was
highly variable in fibroblastic intimal cells, even in
non-inflamed tissue.
A minority of cells appeared to carry both CD68
and either high UDPGD activity or prolyl hydroxylase. In our view this does not represent a spectrum of
cell behavior but is better explained by such factors as
technical problems in assessing closely apposed cells,
partial convergence of the behavior of 2 cell types in
disease, and exchange (by contamination or ingestion)
of material between cells. CD68 is associated with
lysosomes, and in inflamed tissue, fibroblasts can
contain small numbers of lysosomes. Mature macrophages can display prominent rough endoplasmic reticulum. Such convergence may explain reports of
cells of intermediate ultrastructure in diseased tissue
(23). Double labeling with NSE is the most practical
way of demonstrating both fibroblast-like and macrophage populations by light microscopy.
In diseased tissue, fibroblastic intimal cells
showed more variable UDPGD activity with a general
decrease in rheumatoid inflammation, suggesting that
the enzyme activity is down-regulated in inflammation. UDPGD activity may also be affected by antirheumatic drugs (24), but no such information has
been published for the drugs used by these patients,
which included several nonsteroidal agents, gold, sulfasalazine, and azathioprine. It remains unclear how
these cells relate, in terms of origins, to fibroblast-like
intimal cells in normal tissue and whether UDPGD
activity is controlled by short-term events in the cell
environment or whether fibroblast-like synovial lining
cells differ constitutively in this respect from other
fibroblasts. We intend to address these questions by
longer-term in vitro studies.
The nomenclature of intimal cells has been
confused, largely because of difficulties in relating
studies of tissue sections to studies of in vitro preparations and in correlating light and electron microscopic findings. When the ultrastructural classification
of type A (macrophage-like) and type B (fibroblastlike) was proposed (21, separation of cell types by light
microscopy was difficult. This study suggests that it is
now possible to identify by light microscopy a fibroblast subtype in the intima, which is likely to be
equivalent to the type B cell as defined ultrastructurally. Proof of this equivalence was beyond the scope of
this study, and the difficulty of such a proof illustrates
the potential advantages of moving to a light microscopic classification system for intimal cells.
At present, we see the most practical terms as
being “fibroblast-like intimal cell” and “intimal macrophage.” This classification emphasizes the similarity
between intimal and other macrophages. Although the
term “synoviocyte” has been used in several different
ways in the past, restricting its use to the fibroblastlike cell would have the advantage of both clarity and
brevity. Moreover, this is probably the only cell type
peculiar to synovial tissue.
I . Key JA: The synovial membrane of joints and bursae,
Special Cytology. Vol. 2. New York, PB Hoeber, 1932
2. Barland P, Novikoff AB, Hamerman D: Electron microscopy of the human synovial membrane. J Cell Biol
14~207-216. 1962
3. Henderson B, Edwards JCW: The Synovial Lining in
Health and Disease. London, Chapman & Hall, 1987
4. Henderson B, Revell P, Edwards JCW: Synovial lining
cell hyperplasia in rheumatoid arthritis: dogma and fact.
Ann Rheum Dis 47:34%349, 1988
5 . Broker B, Edwards JCW, Fanger M, Lydyard P: The
prevalence and distribution of macrophages bearing
FcRI, FcRII and FcRIII in synovium. Scand J Rheumato1 19:123-135, 1990
6. Wilkinson LS, Worrall JG, Sinclair HS, Edwards JCW:
Immunohistochemical reassessment of accessory cell
populations in normal and diseased human synovium. Br
J Rheumatol 29259-263, 1990
7. Edwards JCW, Willoughby DA: Demonstration of bone
marrow derived cells in synovial lining by means of giant
intracellular granules as genetic markers. Ann Rheum
Dis 41:177-182, 1982
8. Dreher R: Origin of synovial type A cells during inflammation. Immunobiology 161:232, 1982
9. Castor CW, Cabral AR: Connective tissue activating
peptides. Methods Enzymol 163:731-738, 1989
10. Clams BJ, Fraser JRE: Barrier around synovial cells in
vitro. Nature 214:1159, 1%7
11. Stevens CR, Mapp PI, Revell PA: A monoclonal antibody (Mab 67) marks type B synoviocytes. Rheumatol
Int 10:103-106, 1990
12. Chayen J , Bitensky L, Butcher RG: Practical Histochemistry. New York, John Wiley & Sons, 1973
13. Mehdizadeh S, Bitensky L, Chayen J: The assay of
uridine diphosphoglucose dehydrogenase activity: discrimination from xanthine oxidase activity. Cell Biochem Funct 9:103-110, 1991
14. Hadler N: The biology of the extracellular space. Clin
Rheum Dis 7:71-98, 1981
15. Pitsillides AA, Mehdizadeh S, Worrall JG, Wilkinson
LS, Edwards JCW: Demonstration of diphosphoglucose
dehydrogenase activity in normal and rheumatoid synovial lining cells (abstract). Br J Rheumatol (suppl) 1:6,
Franklin WA, Mason DY, Pulford K , Falini B, Bliss E,
Gatter KC, Stein H, Clarke LC, McGee JO: Immunohistological analysis of human mononuclear phagocytes
and dendritic cells by using monoclonal antibodies. Lab
Invest 54:3:322-326, 1986
Hoyhtya M, Myllyla R, Siuva J , Kivirkko KL, Tryggvason K: Monoclonal antibodies to human prolyl-4hydroxylase. Eur J Biochem 141:477-482, 1984
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
Nadler LM: B cell leukaemia panel workshop, Proceedings of the Second International Workshop on Human
Leucocyte Differentiation Antigens. Edited by E Reinhertz. Berlin, Springer-Verlag, 1985
Beverley PCL, Callard RE: Distinctive functional characteristics of human T lymphocytes defined by E rosetting or a monoclonal antibody. Eur J Immunol I1:329334, 1981
Burmester GR, Locher P. Koch B, Winchester RJ,
Dimitriu-Bona A, Kalden JR, Mohr W: The tissue
architecture of synovial membranes in inflammatory and
non-inflammatory joint diseases. Rheumatol Int 3: 173181, 1983
Hogg N , Palmer DG, Revell PA: Mononuclear phagocyte of normal and rheumatoid synovial membrane
identified by monoclonal antibodies. Immunology 56:
673-68 1, 1985
Ghadially FN. Roy S: Ultrastructure of synovial membrane in rheumatoid arthritis. Ann Rheum Dis 26:426
443, 1967
Raab W, Gmeiner B, Muckenhuber P: D-penicillamine
and enzymatic activities. Enzyme 23:84-88, 1978
Без категории
Размер файла
576 Кб
like, intimal, microscopy, synoviocyte, light, characterization, synovial, cells, fibroblasts
Пожаловаться на содержимое документа