British Journal of Rheumatology 1990;29:259-263
IMMUNOHISTOLOGICAL REASSESSMENT OF ACCESSORY CELL POPULATIONS IN NORMAL AND DISEASED HUMAN SYNOVIUM BY L. S. WILKINSON, J. G. WORRALL, H. D. SINCLAIR AND J. C. W. EDWARDS Department of Rheumatology Research, University College and Middlesex School of Medicine, Arthur Stanley House, Tottenham Street, London Wl
KEYWORDS: Normal synovium, Dendritic cell, Macrophage, MHC class 2.
The present study was designed to address the following points. First, we wished to establish whether evidence could be found for the presence of dendritic antigen presenting cells in normal adult human synovium. Second, we wished to reinvestigate antigen presenting cell populations in rheumatoid synovium to establish whether, with the advent of a broad spectrum antibody to macrophages, IDC and macrophage populations could be more clearly distinguished.
THE identification of antigen presenting cells in human tissue remains difficult. Interdigitating dendritic cells (IDC) in the T-cell rich zones of lymphoid tissue and veiled cells in peripheral lymph can be distinguished from classical macrophages by ultrastructure [1]. There is, however, no unambiguous way of distinguishing these cells in tissue sections examined by light microscopy. It has been suggested that the monoclonal antibody (mab) RFDl stains these cells preferentially, both in lymphoid and inflamed tissue [2-4]. Using the immunogold technique, Knight and co-workers [5] have shown that in rheumatoid synovial fluid, RFDl binds only to cells of dendritic ultrastructure. Duke et al. [6] found little or no evidence for the presence of such cells in normal fetal synovium using light microscopy and a combination of antibodies (RFDl, the macrophage marker RFD7 and an antibody to HLA-DR). In contrast, RFDl positive cells lacking RFD7 have been found in rheumatoid arthritis (RA) synovium [4], suggesting that cells with the features of IDC accumulate in diseased tissue. Recent observations have, however, indicated that RFD7 is not a pan macrophage marker. Mabs such as EBMll [7], directed against the CD68 epitope, show a wider tissue distribution than RFD7. Our own observations on synovial tissue [8] indicate that the distribution of EBMll includes cells stained with other markers such as RFD7, anti-CD14 and non-specific esterase activity. Other studies using the mab 3.9, which recognizes CDllc (the pl50/95 antigen), indicate that this marker is expressed on both cells with IDC features and tissue macrophage [9]. This has cast doubt on the distinction between the two cell types, a subject reviewed by Bofill and Janossy [10].
MATERIALS AND METHODS Tissues Normal tonsil (kindly supplied by the Royal National Throat, Nose and Ear Hospital) was obtained from routine tonsillectomy. Normal human synovial tissues were obtained from six individuals undergoing amputation for proximal sarcomata (Table I). Tissue was taken from joints showing no macroscopic abnormality and sited at least 15 cm distant from the tumour. Forty samples were analysed using RFDl and RFD7. A further 25 samples from the same six individuals were studied using RFDl and EBMll. Ten rheumatoid synovial tissues were obtained from patients undergoing diagnostic biopsy or joint replacement (6 knee, 2 hip, 2 wrist). Monoclonal antibodies The mouse anti-human mabs used in the study are shown in Table II. Reagents RFDl, RFD7 and RFBmix were kindly provided by Dr L. W. Poulter of the Royal Free Hospital Medical School. RFDl [2, 3] is believed to bind to a class II antigen with the restricted tissue distribution of HLA-DQ, but governed by genes outside and telomeric to the DQ region. It is expressed strongly on IDC. RFD7 binds to mature macrophages outside follicular structures in tonsil and in other tissues including synovium. UCHT1 was kindly provided by Dr P. Beverley of the Imperial Cancer Research Fund and
Submitted 6 June 1989; revised version accepted 5 January 1990. Correspondence to Dr J. Edwards. 259
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
SUMMARY Accessory cell populations in normal and diseased synovial tissue have been reanalysed following the development of the monoclonal antibody, EBMll, directed against the CD68 epitope which is expressed by macrophages in all locations so far examined. Previous studies used as a macrophage marker the monoclonal antibody RFD7 which binds only a proportion of (mature) macrophages. Using double indirect immunofluorescence, normal and rheumatoid synovial samples were examined for the presence of cells which bind the putative dendritic cell marker, RFDl, in conjunction with either RFD7 or EBMll. RFDl positive cells were found infiveof 40 normal synovia. Of these cells, 30-35% were negative for RFD7 or EBMll and, when closely apposed to T-lymphocytes, showed a typical interdigitating morphology. In contrast, all of ten rheumatoid synovia contained RFDl positive cells; the extent of double labelling with macrophage markers varied with the position of the cells in the tissue. As expected, EBMll stained a larger number of cells with macrophage morphology than RFD7. The combination of RFD 1 and EBMl 1 appears to be a useful method for identifying interdigitating dendritic cells in connective tissue, these cells being characterized by positive RFDl and negative EBMll binding. On this criterion, interdigitating dendritic cells were plentiful in rheumatoid synovium and present, albeit infrequently, in normal synovium.
BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXIX NO. 4
260
TABLE I SITES OF ORIGIN OF NORMAL SYNOVIAL TISSUE AND PRESENCE OF RFDl POSITIVE CELLS
Site
No. containing RFD1+ cells
8 2 1 6 6 3
2 1 1 1 0 0
7 7 9 8 4 3
1 1 2 0 0 0
These joints were obtained from the limbs of six separate individuals.
MID3 by Dr P. Lydyard of the Department of Immunology, UCMSM. All reagents were used at optimal dilution established by prior titration on positive tissue. Immunofluorescent staining All tissues were snap frozen in isopentane containing solid carbon dioxide and stored at — 70°C prior to use. Six-micron thickness frozen sections were cut on a SLEE TE cryostat, taken up on to glass slides, dried for 30 min under a cool fan and then stored wrapped in foil in hermetically sealed boxes at -70°C until required. Prior to staining, sections were wanned to room temperature before unwrapping, fixed for 10 min in cold acetone and then washed in PBS. Sections were incubated for 30 min with a 20 |xl aliquot of single or premixed mabs as appropriate, washed in PBS and then incubated with a 20 \i\ aliquot of single or premixed goat anti-mouse IgM rhodamine conjugate (SBA 1020-03 Sera Lab) and goat anti-mouse IgG fluorescein conjugate (SBA 1030-02 Sera Lab). Sections were finally washed in PBS and mounted in glycerol containing diazabicyclo-octane. Fluorescent staining was observed using a Nikon Optiphot fluorescence microscope. RESULTS Normal human tonsil Normal human tonsil showed staining with RFDl in all areas, but large cells with prominent branching processes were particularly evident in the T-cell areas and in the peripheral stroma. In the T-cell rich zone, brightly RFD1 staining cells occurred in groups, forming a dense net of processes (Fig. la). These cells were negative for the epitopes detected by EBM11, UCHT1, and R4/23. Distinct EBM11 positive cells with less extensive branching processes were present within the T-cell zone (Fig. lb). A minority of these cells showed slight to moderate RFDl staining. The double labelled cells represented only 17% of all RFDl positive cells in this area. Limited overlap of RFDl and EBM11 staining was confirmed on one specimen using optical sectioning
Normal human synovial tissue RFDl positive cells occurred infrequently in normal synovial membrane (Table I), but were occasionally found in both the lining layer and deep stroma. In tissues in which RFDl positive cells occurred, the frequencies of RFDl positive and EBM11 positive cells were in a ratio of 1:8 (300 cells counted). In double labelling studies with RFDl and EBM11,24 RFDl positive cells were seen, eight of which were EBM11 negative. These were isolated oval cells with short membrane extensions in the lining and stromal layers (Fig. 2a, b). Similarly, in tissues double labelled with RFDl and RFD7, eight RFDl positive cells were seen, three of which were RFD7 negative. EBM11 positive macrophages with dendritic morphology were not observed in normal tissues, probably because of the absence of T-cell clusters surrounding and indenting these cells. Rheumatoid human synovium Rheumatoid synovium showed plentiful RFDl postive cells, although the number varied markedly from sample to sample. RFDl showed a much more restricted distribution than MID3, staining many fewer cells in the lining and stroma, most of the bright staining being within lymphoid clusters. Anti-Leu-10 produced a pattern similar to RFDl but without the concentration of bright staining in lymphoid clusters. None of these cells stained with UCHT1. Double staining with EBM11 and RFD7 produced comparable results, but EBM11 always stained more cells than RFD7. The mean percentage of RFDl positive cells double labelling with EBM11 was estimated for 10 tissue samples: 89.5% of RFDl positive cells in the lining layer labelled with EBM11; 41.3% of cells in the stroma double-labelled, but in T-cell clusters only 6.3% of RFDl cells labelled with EBM11. Moreover, the latter RFDl positive cells were distinctly more branched and occurred in groups very similar to those seen in tonsil. Within T-cell clusters, quite distinct EBM11 positive/RFD 1 negative cells TABLE II THE MONOCLONAL ANTIBODIES USED IN THE STUDY
Mab
CD
Subclass Expression
RFDl RFD7
ND ND
IgM IgG
RFBmix CD20 CD22 EBM11 CD68
IgG IgM IgG
R4/23 ND UCHT1 CD3 MID3 ND ND LeulO
IgM IgG IgG IgG
ND, not designated.
Reference
HLA-DQ related 2,3 Mature 2,3 macrophages B-cells All macrophages/ monocytes FDRC T-cell receptor MHC class II HLA-D/DQw2
7 11 (DAKO) 12 13 Becton Dickinson
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
Upper limb Elbow Wrist Carpus Metacarpophalangeal Proximal interphalangeal Distal interphalangeal Lower limb Knee Ankle Tarsal Metatarsophalangeal Proximal interphalangeal Distal interphalangeal
No. of samples
performed on a BioRad Lasersharp MRC500 Confocal Imaging System. These findings support the idea that the clusters of large RFDl positive cells are IDC rather than macrophages.
WILKINSON ETAL.: ACCESSORY CELL POPULATIONS IN SYNOVIUM
261
were found. These cells showed several long processes similar to those seen with tingible body macrophages or macrophagesof tonsillarT-cell areas. DISCUSSION Accessory or antigen presenting cells are thought to play a major role in the pathogenesis of RA [14]. Such a role is supported by the close proximity of lymphocytes and non-lymphoid MHC class II positive cells in rheumatoid synovium and by the immunogenetic link between RA and the class II allotypes DR4 and DR1 [15]. Antigen presentation could be performed by macrophages, non-phagocytic monocyte derived cells
('dendritic cells'), or stromal fibroblastic cells subjected to activating stimuli. Evidence for the presence of the second type of cell has come from the work of Harding and Knight [16] which shows that cells with IDC-like ultrastructure are present in rheumatoid synovial fluid. In the human, the lineage of these cells is uncertain but an analogous type of cell in the mouse appears to belong to a lineage distinct from that of typical macrophages [17]. Such a lineage would in the human include IDC and the veiled cells of lymph, which have a similar ultrastructure. It would not include Langerhans cells, which are OKT6 positive [18], or follicular dendritic cells, which are MHC class II negative [19].
FIG. 2.—Surface frond of normal human synovium stained by double indirect immunofluorescence for RFDl (a) and EBMll (b). Both RFDl positive/EBMll negative (far right) and RFDl negative/EBMll positive (right of centre) cells are present. (x332).
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
FIG. 1.—Normal human tonsil stained by double indirect immunofluorescence for RFDl (a) and EBMll (b). RFDl positive cells are large with extensive interdigitating extensions. EBMll positive cells are smaller and more evenly distributed. A small proportion of cells double label. (x332).
262
BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXIX NO. 4
The findings of the present study suggest that both macrophages and typical IDC are present in the T-cell rich areas of normal tonsil and rheumatoid T-cell clusters. RFDl positive cells elsewhere in rheumatoid and normal synovium are more difficult to classify. The majority of RFDl positive cells outside T-cell clusters in rheumatoid tissue appear to be macrophages on the basis of EBMl 1 staining and morphology. They appear as large oval mononuclear cells with a small number of processes which are not part of a dense reticular net. Similar cells were found in some normal synovia. The nature of RFDl positive/EBMll negative cells seen outside T-cell clusters in rheumatoid synovia and, very occasionally, in the stroma and lining of normal synovia, must remain speculative. However, we suggest that they belong to the lineage of IDC for the following reasons. First, they are totally negative for EBMl 1. When EBMl 1 staining occurs in other cells it varies in intensity, but is never difficult to observe, the antibody providing a particularly high signal-to-noise ratio for immunohistochemistry. Second, the presence of IDC in synovia! fluid [16] indicates that IDC must pass through the stroma and lining cell layer on their way from blood vessel to fluid, at least in diseased tissue. The RFDl positive/EBMll negative cells in the lining layer do not have an interdigitating appearance, but this may indicate that they are relatively immature cells in transit and such an appearance would not in any case be expected in the absence of closely apposed T-cells. Taken with previous published evidence it seems that RFDl cannot be used alone as a marker of IDC lineage, but when used in combination with EBMU it may be the best indicator available for light microscopy. It has a more restricted distribution than other MHC class II markers and the combination of IgM (RFDl) and IgG (EBM11) mabs facilitates double
staining. Other macrophage markers cannot be relied on because of either too restricted (e.g. CD14) or too broad (e.g. 3.9) a distribution. In conclusion, it appears likely that true IDC can be demonstrated in synovial tissue, and that these are distinct from EBM11 positive cells. They are present in very small numbers in some areas of normal adult human synovial tissue but could be involved in the earliest events during the development of rheumatoid synovitis. It is beyond the scope of this paper to define the relative importance of the resident cell populations of normal synovial tissue. However, these putative IDC may represent a small garrison of immunocompetent cells ready to react immediately to a tissue insult. Together with the resident macrophage population they may serve to elaborate the immunological response in collaboration with any infiltrating cell population. REFERENCES
1. Balfour BM, Drexhage HA, Kamperdijk EWA, Hoefsmit ECM. Antigen presenting cells, including Langerhans' cells, veiled cells and interdigitating cells. In: Microenvironments in haemopoietic and lymphoid differentiation. Ciba Foundation Symposium. London: Pitman, 1984;281-301. 2. Poulter LW, Campbell DA, Munro C, Janossy G. Discrimination of human macrophages and dendritic cells by means of monoclonal antibodies. Scand J Immunol 1986;24:351-7. 3. Janossy G, Bofill M, Poulter LW, etal. Separate ontogeny of two macrophage-like accessory cell populations in the human fetus. J Immunol 1986; 136:4354-61. 4. Poulter LW, Janossy G. The involvement of dendritic cells in chronic inflammatory disease. Scand J Immunol 1985;21:401-7.
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
FIG. 3.—Rheumatoid synovium stained oy double indirect immunofluorescence for RFDl (a) and EBMl 1 (b). The two reagents give markedly different patterns with a minority ofdouble stained cells. RFDl positive/EBMll negative cells have a more ramifying appearance. (x332).
WILKINSON ETAL.:
ACCESSORY CELL POPULATIONS IN SYNOVIUM
12. Beverley PCL, Callard RE. Distinctive functional characteristics of human T lymphocytes denned by E rosetting or a monoclonal antibody. Eur J Immunol 1981;ll:329-34. 13. Lydyard PM, Banga JP, Guarnotta G, Walker P, Mackenzie L, Markandy S. Human lymphocyte antigens a mini-review. Biochem Soc Trans 1985;13:429-32. 14. Poulter LW, Duke O, Hobbs S, Janossy G, Panayi G, Seymour G. The pathogenesis of rheumatoid arthritis. Clin Exp Immunol 1983;51:247-51. 15. Lee SH, Matsuyama T, Logalbo P, Silver J, Winchester RJ. la antigens and susceptibility to rheumatoid arthritis. Clins Rheum Dis 1985; 11:645-64. 16. Harding B, Knight SC. The distribution of dendritic cells in the synovial fluid of patients with arthritis. Clin Exp Immunol 1986;63:594-602. 17. Steinman RM, Inaba K, Schula G, Witmer M. Stimulation of the immune response: contributions of dendritic cells. In: Steinman RM, North RJ, eds. Mechanisms of host resistance to infectious agents, tumors and allografts. New York: Rockefeller University Press, 1986; 71-97. 18. Murphy GF, Bhan AK, Harris TJ, Mihm MC Jnr. In situ identification of T6 positive cells in normal human dermis by immunoelectron-microscopy. BrJ Dermatol 1983;108:423-31. 19. Szakal AK, Gieringer RL, Kosco MH, Tew JG. Isolated follicular dendritic cells: cytochemical antigen localization, Nomarski, SEM and TEM morphology. J Immunol 1985;134:1349-59.
CLINICAL CONUNDRUM Is there a treatment dermatomvositis?
for the calcinosis of juvenile
is no drug therapy which appears to influence the course [1]. The initial thought that warfarin might be of value has not been substantiated, with probably only 50% of patients showing any improvement and the risk of haemorrhage including cerebral bleeding is very real. However, spontaneous regression does occur [2] and it would appear that this is most likely in patients where the disease is well controlled and who have a very active mobilization THERE
programme. B. ANSELL
The Consulting Rooms, 9 Beaumont Road, Windsor SL4 1HY 1. Ansell BM. Management of polymyositis and dermatomyositis. In: Clinics in Rheumatic Diseases. Inflammatory disorders of muscle. 1984; 10: 205-13. 2. Sewell JR, Liyanage B, Ansell BM. Calcinosis in juvenile dermatomyositis. Skeletal Radio! 1978; 3: 137-43.
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
Dixey J, Fryer P, Griffiths S, Harding B, Ansell B, Knight S. Class II antigens on dendritic cells (DCs) from synovial fluid of patients with inflammatory arthritis. Br J Rheumatol 1988;27(Abstr Suppl 2): 90. Duke O, Poulter LW, Sorg C, Panayi GS. Accessory cells in fetal joints. Br J Rheumatol 1987;26(Abstr Suppl 2): 123. Franklin WA, Mason DY, Pulford K, etal. Immunohistological analysis of human mononuclear phagocytes and dendritic cells by using monoclonal antibodies. Lab Invest 1986;54:322. Wilkinson LS, Edwards JCW. Analysis of macrophages in synovial tissue using monoclonal antibody EBM11. Br J Rheumatol 1989;28(Abstr Suppl 1):44. Hogg N, Takacs L, Palmer DG, Selvendran Y, Allen C. The pl50,95 molecule is a marker of human mononuclear phagocytes: comparison with expression of class II molecules. Eur J Immunol 1986;16:240-8. 10 Bofill M, Janossy G. The reactivity of macrophages and interdigitating cells with the monoclonal antibodies of the Vienna leucocyte workshop. In: Leucocyte typing IV. Oxford: Oxford University Press (In press). 11 Naiem M, Gerdes J, Abdulaziz Z, Stein H, Mason DY. Production of a monoclonal antibody reactive with human dendritic reticulum cells and its use in the immunohistological analysis of lymphoid tissue. J Clin Pathol 1983;36:167-75.
263
IMMUNOHISTOLOGICAL REASSESSMENT OF ACCESSORY CELL POPULATIONS IN NORMAL AND DISEASED HUMAN SYNOVIUM BY L. S. WILKINSON, J. G. WORRALL, H. D. SINCLAIR AND J. C. W. EDWARDS Department of Rheumatology Research, University College and Middlesex School of Medicine, Arthur Stanley House, Tottenham Street, London Wl
KEYWORDS: Normal synovium, Dendritic cell, Macrophage, MHC class 2.
The present study was designed to address the following points. First, we wished to establish whether evidence could be found for the presence of dendritic antigen presenting cells in normal adult human synovium. Second, we wished to reinvestigate antigen presenting cell populations in rheumatoid synovium to establish whether, with the advent of a broad spectrum antibody to macrophages, IDC and macrophage populations could be more clearly distinguished.
THE identification of antigen presenting cells in human tissue remains difficult. Interdigitating dendritic cells (IDC) in the T-cell rich zones of lymphoid tissue and veiled cells in peripheral lymph can be distinguished from classical macrophages by ultrastructure [1]. There is, however, no unambiguous way of distinguishing these cells in tissue sections examined by light microscopy. It has been suggested that the monoclonal antibody (mab) RFDl stains these cells preferentially, both in lymphoid and inflamed tissue [2-4]. Using the immunogold technique, Knight and co-workers [5] have shown that in rheumatoid synovial fluid, RFDl binds only to cells of dendritic ultrastructure. Duke et al. [6] found little or no evidence for the presence of such cells in normal fetal synovium using light microscopy and a combination of antibodies (RFDl, the macrophage marker RFD7 and an antibody to HLA-DR). In contrast, RFDl positive cells lacking RFD7 have been found in rheumatoid arthritis (RA) synovium [4], suggesting that cells with the features of IDC accumulate in diseased tissue. Recent observations have, however, indicated that RFD7 is not a pan macrophage marker. Mabs such as EBMll [7], directed against the CD68 epitope, show a wider tissue distribution than RFD7. Our own observations on synovial tissue [8] indicate that the distribution of EBMll includes cells stained with other markers such as RFD7, anti-CD14 and non-specific esterase activity. Other studies using the mab 3.9, which recognizes CDllc (the pl50/95 antigen), indicate that this marker is expressed on both cells with IDC features and tissue macrophage [9]. This has cast doubt on the distinction between the two cell types, a subject reviewed by Bofill and Janossy [10].
MATERIALS AND METHODS Tissues Normal tonsil (kindly supplied by the Royal National Throat, Nose and Ear Hospital) was obtained from routine tonsillectomy. Normal human synovial tissues were obtained from six individuals undergoing amputation for proximal sarcomata (Table I). Tissue was taken from joints showing no macroscopic abnormality and sited at least 15 cm distant from the tumour. Forty samples were analysed using RFDl and RFD7. A further 25 samples from the same six individuals were studied using RFDl and EBMll. Ten rheumatoid synovial tissues were obtained from patients undergoing diagnostic biopsy or joint replacement (6 knee, 2 hip, 2 wrist). Monoclonal antibodies The mouse anti-human mabs used in the study are shown in Table II. Reagents RFDl, RFD7 and RFBmix were kindly provided by Dr L. W. Poulter of the Royal Free Hospital Medical School. RFDl [2, 3] is believed to bind to a class II antigen with the restricted tissue distribution of HLA-DQ, but governed by genes outside and telomeric to the DQ region. It is expressed strongly on IDC. RFD7 binds to mature macrophages outside follicular structures in tonsil and in other tissues including synovium. UCHT1 was kindly provided by Dr P. Beverley of the Imperial Cancer Research Fund and
Submitted 6 June 1989; revised version accepted 5 January 1990. Correspondence to Dr J. Edwards. 259
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
SUMMARY Accessory cell populations in normal and diseased synovial tissue have been reanalysed following the development of the monoclonal antibody, EBMll, directed against the CD68 epitope which is expressed by macrophages in all locations so far examined. Previous studies used as a macrophage marker the monoclonal antibody RFD7 which binds only a proportion of (mature) macrophages. Using double indirect immunofluorescence, normal and rheumatoid synovial samples were examined for the presence of cells which bind the putative dendritic cell marker, RFDl, in conjunction with either RFD7 or EBMll. RFDl positive cells were found infiveof 40 normal synovia. Of these cells, 30-35% were negative for RFD7 or EBMll and, when closely apposed to T-lymphocytes, showed a typical interdigitating morphology. In contrast, all of ten rheumatoid synovia contained RFDl positive cells; the extent of double labelling with macrophage markers varied with the position of the cells in the tissue. As expected, EBMll stained a larger number of cells with macrophage morphology than RFD7. The combination of RFD 1 and EBMl 1 appears to be a useful method for identifying interdigitating dendritic cells in connective tissue, these cells being characterized by positive RFDl and negative EBMll binding. On this criterion, interdigitating dendritic cells were plentiful in rheumatoid synovium and present, albeit infrequently, in normal synovium.
BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXIX NO. 4
260
TABLE I SITES OF ORIGIN OF NORMAL SYNOVIAL TISSUE AND PRESENCE OF RFDl POSITIVE CELLS
Site
No. containing RFD1+ cells
8 2 1 6 6 3
2 1 1 1 0 0
7 7 9 8 4 3
1 1 2 0 0 0
These joints were obtained from the limbs of six separate individuals.
MID3 by Dr P. Lydyard of the Department of Immunology, UCMSM. All reagents were used at optimal dilution established by prior titration on positive tissue. Immunofluorescent staining All tissues were snap frozen in isopentane containing solid carbon dioxide and stored at — 70°C prior to use. Six-micron thickness frozen sections were cut on a SLEE TE cryostat, taken up on to glass slides, dried for 30 min under a cool fan and then stored wrapped in foil in hermetically sealed boxes at -70°C until required. Prior to staining, sections were wanned to room temperature before unwrapping, fixed for 10 min in cold acetone and then washed in PBS. Sections were incubated for 30 min with a 20 |xl aliquot of single or premixed mabs as appropriate, washed in PBS and then incubated with a 20 \i\ aliquot of single or premixed goat anti-mouse IgM rhodamine conjugate (SBA 1020-03 Sera Lab) and goat anti-mouse IgG fluorescein conjugate (SBA 1030-02 Sera Lab). Sections were finally washed in PBS and mounted in glycerol containing diazabicyclo-octane. Fluorescent staining was observed using a Nikon Optiphot fluorescence microscope. RESULTS Normal human tonsil Normal human tonsil showed staining with RFDl in all areas, but large cells with prominent branching processes were particularly evident in the T-cell areas and in the peripheral stroma. In the T-cell rich zone, brightly RFD1 staining cells occurred in groups, forming a dense net of processes (Fig. la). These cells were negative for the epitopes detected by EBM11, UCHT1, and R4/23. Distinct EBM11 positive cells with less extensive branching processes were present within the T-cell zone (Fig. lb). A minority of these cells showed slight to moderate RFDl staining. The double labelled cells represented only 17% of all RFDl positive cells in this area. Limited overlap of RFDl and EBM11 staining was confirmed on one specimen using optical sectioning
Normal human synovial tissue RFDl positive cells occurred infrequently in normal synovial membrane (Table I), but were occasionally found in both the lining layer and deep stroma. In tissues in which RFDl positive cells occurred, the frequencies of RFDl positive and EBM11 positive cells were in a ratio of 1:8 (300 cells counted). In double labelling studies with RFDl and EBM11,24 RFDl positive cells were seen, eight of which were EBM11 negative. These were isolated oval cells with short membrane extensions in the lining and stromal layers (Fig. 2a, b). Similarly, in tissues double labelled with RFDl and RFD7, eight RFDl positive cells were seen, three of which were RFD7 negative. EBM11 positive macrophages with dendritic morphology were not observed in normal tissues, probably because of the absence of T-cell clusters surrounding and indenting these cells. Rheumatoid human synovium Rheumatoid synovium showed plentiful RFDl postive cells, although the number varied markedly from sample to sample. RFDl showed a much more restricted distribution than MID3, staining many fewer cells in the lining and stroma, most of the bright staining being within lymphoid clusters. Anti-Leu-10 produced a pattern similar to RFDl but without the concentration of bright staining in lymphoid clusters. None of these cells stained with UCHT1. Double staining with EBM11 and RFD7 produced comparable results, but EBM11 always stained more cells than RFD7. The mean percentage of RFDl positive cells double labelling with EBM11 was estimated for 10 tissue samples: 89.5% of RFDl positive cells in the lining layer labelled with EBM11; 41.3% of cells in the stroma double-labelled, but in T-cell clusters only 6.3% of RFDl cells labelled with EBM11. Moreover, the latter RFDl positive cells were distinctly more branched and occurred in groups very similar to those seen in tonsil. Within T-cell clusters, quite distinct EBM11 positive/RFD 1 negative cells TABLE II THE MONOCLONAL ANTIBODIES USED IN THE STUDY
Mab
CD
Subclass Expression
RFDl RFD7
ND ND
IgM IgG
RFBmix CD20 CD22 EBM11 CD68
IgG IgM IgG
R4/23 ND UCHT1 CD3 MID3 ND ND LeulO
IgM IgG IgG IgG
ND, not designated.
Reference
HLA-DQ related 2,3 Mature 2,3 macrophages B-cells All macrophages/ monocytes FDRC T-cell receptor MHC class II HLA-D/DQw2
7 11 (DAKO) 12 13 Becton Dickinson
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
Upper limb Elbow Wrist Carpus Metacarpophalangeal Proximal interphalangeal Distal interphalangeal Lower limb Knee Ankle Tarsal Metatarsophalangeal Proximal interphalangeal Distal interphalangeal
No. of samples
performed on a BioRad Lasersharp MRC500 Confocal Imaging System. These findings support the idea that the clusters of large RFDl positive cells are IDC rather than macrophages.
WILKINSON ETAL.: ACCESSORY CELL POPULATIONS IN SYNOVIUM
261
were found. These cells showed several long processes similar to those seen with tingible body macrophages or macrophagesof tonsillarT-cell areas. DISCUSSION Accessory or antigen presenting cells are thought to play a major role in the pathogenesis of RA [14]. Such a role is supported by the close proximity of lymphocytes and non-lymphoid MHC class II positive cells in rheumatoid synovium and by the immunogenetic link between RA and the class II allotypes DR4 and DR1 [15]. Antigen presentation could be performed by macrophages, non-phagocytic monocyte derived cells
('dendritic cells'), or stromal fibroblastic cells subjected to activating stimuli. Evidence for the presence of the second type of cell has come from the work of Harding and Knight [16] which shows that cells with IDC-like ultrastructure are present in rheumatoid synovial fluid. In the human, the lineage of these cells is uncertain but an analogous type of cell in the mouse appears to belong to a lineage distinct from that of typical macrophages [17]. Such a lineage would in the human include IDC and the veiled cells of lymph, which have a similar ultrastructure. It would not include Langerhans cells, which are OKT6 positive [18], or follicular dendritic cells, which are MHC class II negative [19].
FIG. 2.—Surface frond of normal human synovium stained by double indirect immunofluorescence for RFDl (a) and EBMll (b). Both RFDl positive/EBMll negative (far right) and RFDl negative/EBMll positive (right of centre) cells are present. (x332).
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
FIG. 1.—Normal human tonsil stained by double indirect immunofluorescence for RFDl (a) and EBMll (b). RFDl positive cells are large with extensive interdigitating extensions. EBMll positive cells are smaller and more evenly distributed. A small proportion of cells double label. (x332).
262
BRITISH JOURNAL OF RHEUMATOLOGY VOL. XXIX NO. 4
The findings of the present study suggest that both macrophages and typical IDC are present in the T-cell rich areas of normal tonsil and rheumatoid T-cell clusters. RFDl positive cells elsewhere in rheumatoid and normal synovium are more difficult to classify. The majority of RFDl positive cells outside T-cell clusters in rheumatoid tissue appear to be macrophages on the basis of EBMl 1 staining and morphology. They appear as large oval mononuclear cells with a small number of processes which are not part of a dense reticular net. Similar cells were found in some normal synovia. The nature of RFDl positive/EBMll negative cells seen outside T-cell clusters in rheumatoid synovia and, very occasionally, in the stroma and lining of normal synovia, must remain speculative. However, we suggest that they belong to the lineage of IDC for the following reasons. First, they are totally negative for EBMl 1. When EBMl 1 staining occurs in other cells it varies in intensity, but is never difficult to observe, the antibody providing a particularly high signal-to-noise ratio for immunohistochemistry. Second, the presence of IDC in synovia! fluid [16] indicates that IDC must pass through the stroma and lining cell layer on their way from blood vessel to fluid, at least in diseased tissue. The RFDl positive/EBMll negative cells in the lining layer do not have an interdigitating appearance, but this may indicate that they are relatively immature cells in transit and such an appearance would not in any case be expected in the absence of closely apposed T-cells. Taken with previous published evidence it seems that RFDl cannot be used alone as a marker of IDC lineage, but when used in combination with EBMU it may be the best indicator available for light microscopy. It has a more restricted distribution than other MHC class II markers and the combination of IgM (RFDl) and IgG (EBM11) mabs facilitates double
staining. Other macrophage markers cannot be relied on because of either too restricted (e.g. CD14) or too broad (e.g. 3.9) a distribution. In conclusion, it appears likely that true IDC can be demonstrated in synovial tissue, and that these are distinct from EBM11 positive cells. They are present in very small numbers in some areas of normal adult human synovial tissue but could be involved in the earliest events during the development of rheumatoid synovitis. It is beyond the scope of this paper to define the relative importance of the resident cell populations of normal synovial tissue. However, these putative IDC may represent a small garrison of immunocompetent cells ready to react immediately to a tissue insult. Together with the resident macrophage population they may serve to elaborate the immunological response in collaboration with any infiltrating cell population. REFERENCES
1. Balfour BM, Drexhage HA, Kamperdijk EWA, Hoefsmit ECM. Antigen presenting cells, including Langerhans' cells, veiled cells and interdigitating cells. In: Microenvironments in haemopoietic and lymphoid differentiation. Ciba Foundation Symposium. London: Pitman, 1984;281-301. 2. Poulter LW, Campbell DA, Munro C, Janossy G. Discrimination of human macrophages and dendritic cells by means of monoclonal antibodies. Scand J Immunol 1986;24:351-7. 3. Janossy G, Bofill M, Poulter LW, etal. Separate ontogeny of two macrophage-like accessory cell populations in the human fetus. J Immunol 1986; 136:4354-61. 4. Poulter LW, Janossy G. The involvement of dendritic cells in chronic inflammatory disease. Scand J Immunol 1985;21:401-7.
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
FIG. 3.—Rheumatoid synovium stained oy double indirect immunofluorescence for RFDl (a) and EBMl 1 (b). The two reagents give markedly different patterns with a minority ofdouble stained cells. RFDl positive/EBMll negative cells have a more ramifying appearance. (x332).
WILKINSON ETAL.:
ACCESSORY CELL POPULATIONS IN SYNOVIUM
12. Beverley PCL, Callard RE. Distinctive functional characteristics of human T lymphocytes denned by E rosetting or a monoclonal antibody. Eur J Immunol 1981;ll:329-34. 13. Lydyard PM, Banga JP, Guarnotta G, Walker P, Mackenzie L, Markandy S. Human lymphocyte antigens a mini-review. Biochem Soc Trans 1985;13:429-32. 14. Poulter LW, Duke O, Hobbs S, Janossy G, Panayi G, Seymour G. The pathogenesis of rheumatoid arthritis. Clin Exp Immunol 1983;51:247-51. 15. Lee SH, Matsuyama T, Logalbo P, Silver J, Winchester RJ. la antigens and susceptibility to rheumatoid arthritis. Clins Rheum Dis 1985; 11:645-64. 16. Harding B, Knight SC. The distribution of dendritic cells in the synovial fluid of patients with arthritis. Clin Exp Immunol 1986;63:594-602. 17. Steinman RM, Inaba K, Schula G, Witmer M. Stimulation of the immune response: contributions of dendritic cells. In: Steinman RM, North RJ, eds. Mechanisms of host resistance to infectious agents, tumors and allografts. New York: Rockefeller University Press, 1986; 71-97. 18. Murphy GF, Bhan AK, Harris TJ, Mihm MC Jnr. In situ identification of T6 positive cells in normal human dermis by immunoelectron-microscopy. BrJ Dermatol 1983;108:423-31. 19. Szakal AK, Gieringer RL, Kosco MH, Tew JG. Isolated follicular dendritic cells: cytochemical antigen localization, Nomarski, SEM and TEM morphology. J Immunol 1985;134:1349-59.
CLINICAL CONUNDRUM Is there a treatment dermatomvositis?
for the calcinosis of juvenile
is no drug therapy which appears to influence the course [1]. The initial thought that warfarin might be of value has not been substantiated, with probably only 50% of patients showing any improvement and the risk of haemorrhage including cerebral bleeding is very real. However, spontaneous regression does occur [2] and it would appear that this is most likely in patients where the disease is well controlled and who have a very active mobilization THERE
programme. B. ANSELL
The Consulting Rooms, 9 Beaumont Road, Windsor SL4 1HY 1. Ansell BM. Management of polymyositis and dermatomyositis. In: Clinics in Rheumatic Diseases. Inflammatory disorders of muscle. 1984; 10: 205-13. 2. Sewell JR, Liyanage B, Ansell BM. Calcinosis in juvenile dermatomyositis. Skeletal Radio! 1978; 3: 137-43.
Downloaded from http://rheumatology.oxfordjournals.org/ at New York University on May 25, 2015
Dixey J, Fryer P, Griffiths S, Harding B, Ansell B, Knight S. Class II antigens on dendritic cells (DCs) from synovial fluid of patients with inflammatory arthritis. Br J Rheumatol 1988;27(Abstr Suppl 2): 90. Duke O, Poulter LW, Sorg C, Panayi GS. Accessory cells in fetal joints. Br J Rheumatol 1987;26(Abstr Suppl 2): 123. Franklin WA, Mason DY, Pulford K, etal. Immunohistological analysis of human mononuclear phagocytes and dendritic cells by using monoclonal antibodies. Lab Invest 1986;54:322. Wilkinson LS, Edwards JCW. Analysis of macrophages in synovial tissue using monoclonal antibody EBM11. Br J Rheumatol 1989;28(Abstr Suppl 1):44. Hogg N, Takacs L, Palmer DG, Selvendran Y, Allen C. The pl50,95 molecule is a marker of human mononuclear phagocytes: comparison with expression of class II molecules. Eur J Immunol 1986;16:240-8. 10 Bofill M, Janossy G. The reactivity of macrophages and interdigitating cells with the monoclonal antibodies of the Vienna leucocyte workshop. In: Leucocyte typing IV. Oxford: Oxford University Press (In press). 11 Naiem M, Gerdes J, Abdulaziz Z, Stein H, Mason DY. Production of a monoclonal antibody reactive with human dendritic reticulum cells and its use in the immunohistological analysis of lymphoid tissue. J Clin Pathol 1983;36:167-75.
263
