Scand. J. Immunol. 36, 681-688, 1992
T-Cell Receptor V-Gene Usage in Synovial Fluid and Synovial Tissue from RA Patients S. GUDMUNDSSON, J R O N N E L I D , A. KARLSSON-PARRA, J. LYSHOLM*, B. GUDBJORNSSONt. B. WIDENFALKJ, C. H. JANSON§ & L. KLARESKOG Department of Clinical Immunology. Uppsala Utiiversity Hospital, Sweden, 'Department of Rheumatology. Falu Hospital, Falun. Sweden, tDepartment of Rheumatology, Uppsala University Hospital, t Department of Hand Surgery, Uppsala University Hospital, and ^Department of Immunology, Karolinska lnstitutet, Stockholm, Sweden
Gudmundsson S. Ronnelid J, Karlsson-Parra A. Lysholm J, Gudbjornsson B, Widenfalk B. Janson CH. Klareskog L. T-Cell Receptor V-Gene Usage in Synovial Fluid and Synovial Tissue from RA Patients. Scand J Immunol 1992:36:681-8 The question of whether Ihcre is a preferential use of certain V genes in T cells entering an inflamed joint has hitherto been studied mainly using unfractionated eells from synovialfluidand tissue respectively, and no clear answer to the queslion has yet been provided. Concomitantly. evidence has been provided that the use of V genes may dilVer considerably between CD4 ' and CD8*^ T cells, and consequently thai detection of biased V-gene expression within an inflammatory lesion may require separate analysis of the Iwo T-cell subsets. In this paper we have therefore studied T-cell receplor V-gene expression in rheumatoid arlhritis by means of double stainings of synovial fluid and blood for available anti-TCR monoclonal antibodies and antibodies to CD4 and CD8. respectively. Double stainings were also performed with anti-TCR antibodies and antibodies to activation markers HLA-DR and 1L-2R. A certain bias towards the preferential use of certain V genes was seen particularly in the synovial fluid samples within both the CD4 ' and CDS * T-cell populations, but no uniform pattern was evident among the 35 patients investigated. Sveinn Gudtnundsson MD. Deparinieiil of Clinical Immunology, Uppsala Unirer.sitv Ho.spiial. S751 85 Upp.sala. Sweden
Activated T lymphocytes are thought to be involved in the development ofsynovitis as well as in other manifestations of rheumatoid arthritis (for a review see Ref. 1). It is likely that at least some of these cells are activated loeally in the synovial tissue, in contact with synovial MHC class ll-expressing cells and antigen [2]. It is also likely that T cells of particular importance for the pathogenesis of the disease are restricted by a particular MHC class II HLA-DR4 molecule which is characterized by a detined sequence in its ji chain [3]. Although these features suggest that the T cells driving the disease may be activated by defined antigen(s) (see discussion in Ref. 3). it has so far been difficult to define such an antigen. Another way to utilize the knowledge of the MHC class II restriction of this T-cell-dependent disease is by taking advantage of the fact that T cells that escape elimination in the thymus. but
are still capable of mediating an autoimmune response, tend to make use of a restricted number of variable T-cell receptor (TCR) genes [4, 5], If T cells restricted by the particular HLA-DR4 fi chain that confers susceptibility to rheumatoid arthritis (RA) would also tend to make use of one or a few variable TCR ji genes, this would open possibilities both for a further analysis of the specificity of T cells driving the disease and for specific therapy where a subset of potentially arthritogenic T cells can be affected (see Refs 68). Analysis of TCR usage among synovial fluid eells, and to some extent synovial tissue eells, has so far been carried out by a number of groups. They have used either a limited number of monoclonal antibodies for analysis of the total number of T cells in the synovial fluid [9], or V^ specific oligonucleotide probes to quantitate 681
682
S. Gudmundsson el al.
mRNA correspotiding to various V/( genes cither in the unfractionated synovial fluid or tissue by means of polymerase chain reaction (PCR) amplification [10 12], or in in vitro expanded T eells by means of direct blotting techniques [1316]. Quite variable results have been obtained, particularly from the analyses on in vitro expanded cell populations, which may be due to selection biases introduced by the various culturing techniques (for results and discussions, sec Refs 14 and 16). The few analyses published using PCR amplifieation techniques have also yielded uneonclusive results so far. In one study a certain relative aeeumulation of V/fM TCR among RA synovial fluid cells was reported [11], which was not eonfirmed in another study [10]; the interpretation ofthesc studies is. so far, also litnitcd by the difficulties encountered in tnaking the actual PCR-based techniques sufficiently quantitative. Thus, it appears that monoclonal antibodies that permit a direct quantitative analysis of synovial cell V-gcne expression remain important tools in the search for restricted V-gene usage among inflammatory cells, something that Is further ctnphasized by the demonstration of different Vgene usage among CD4-and CD8-positiveT cells respectively by means of double itnmtinofluorescence labelling techniques [17]. Against this baekground, we have in the present sludy analysed both synovial fluid and synovial tissue T cells in a number ol" synovial specimens from RA patients using a wider panel ofTCR-speeific monoclonal antibodies than previously published. By means of separate analysis both for CD4- and CD8-positive T-celt subsets, as well as T-cell subsets carrying dilTerent activation markers, we also wanted to see whether there was a bias regarding TCR rearrangement within relevant T-cell subsets.
routine surgery and the biopsies kindly provided by Dr Claes Juhlin, Department of Surgery, Uppsala University Hospital, The synovial biopsies were snap-frozen as earlier described in detail [9] and kept frozen at —70 C until cryosectioncd. Mononucloar cells from the synovial lluid and peripheral blood samples were obtained by Ficoll Hypaque centrifugation. Antihtiilics. Monoclonal antibodies reactive with the following TCR V-gene segments were obtained from TCell Sciences Inc. (Cambridge. MA. USA): V/y5.1 (LC4. Ref. IS); V/i5.2-1-5.3(ICI,Ref. 19); V/y5.3 (WI 12, Ref. 20); V/(6.7 (OTI45, Ref. 21); V/(8 (idGK. Ref. 20) and V^12 (S511, Ref. 22). The monoelonai antibody (MoAb) Fl (Va2.3) was produced in our laboratory [2.1]. Notably, the TCR MoAb listed above may also have reactivities with V-gene segments other than the indieated ones. Other antibodies used were antibodies to gamma-delta and to the ^(/(TCR (T-Cell Sciences), Leu4 (anti-CD.l), Leu3a (anti-CD4). Leu2 (anti-CD8), anti-HLA-DR. and anti-I L-2 receptor (anti-CD2.^) (all from Bctton-Dickinson. Sunnyvale. CA, USA). Imniuni'hisUnhcmiiol stdiuin^s. Immiinohistoehemical stainings were done on frozen sections of 9 biopsies from inflamed synovial tissue from RA patients and from normal lymph nodes and one normal spleen. Stainings were made as previously described [24] utilizing a goat anti-mouse secondary reagent and a monoelonai peroxidase-anti-peroxidase (PAP) reagenl (Dakopatts, Giostrup. Denmark) in tbe third step. In the estimation of the fraetion of T eells stained with the respective anti-TCR MoAbs. the Leu4-positive cells within a given area of the biopsy were first counted. Thereafter, cells stained with the respective anti-TCR MoAb.s were counted on serial sections encompassing the same area. Cylofluarinii'lric analysis. Cytofluorimetric analysis was made on suspended cells from 35 synovial fluid samples and 13 parallel peripheral hlood samples and from 15 normal individuals. Stainings were made with the various anti-TCR antibodies or anli-Leu4 as the primary reagents followed by an FITC-labelled goat anti-mouse IgG antibody, Kor double labellings phycoerythrin-conjugated anti-HLA-DR. anti-CD4. artiCDS and anti-CD25 antibodies (Beeton-Dickinson) were added after blocking the secondary anti-mouse antibody with l"'(i normal mouse serum. Labelled samples were analysed in a FACScan cytofluorimeter (Beeton-Diekinson).
MATERIALS AND METHODS Patients, biopsies and .synorial fluid specimens. Synovial tissue biopsies from patients sufTering from active RA were obtained in conjunction with routine surgery at the Department of Hand Surgery, Uppsala University Hospital. Synovial fluids (SF) from knee joints of active RA patients as well ;is heparinized peripheral blood from these patients were provided from the Department of Rheumatology. Falu Lasarctt, Sweden and from the Department of Rheumatology. Uppsala University Hospital. Peripheral blood (PBL) from nonnal healthy controls was obtained from blood donors. Biopsies from normal human lymph nodes and nonnal spleen were obtained in conjunction with
RESULTS Itwnutwhtstochemical stainings oti biopsies of ilieumatoid synovial tissue Stainings with the V/i5,2-F5.3, V/J5.3, V/i6.7 and anti-gamma-delta TCR were distinctly positive for lymphocyte subpopulations in both lymph nodes and synovial specimens, as was the anti-Lcu4 antibody. No stainings on the frozen seetions were, however, seen with the anti-a^ TCR antibody, and only weak stainings, difficult
TCR I-Gene U.sage in RA Paiienis
683
•fa)
f.
V%S
A
^
•A^-^^
• V
:V«»"^>'•
#•?
•-
riy^"-^^ X:..
•
•
^
^
*
'«.-;--> \*-"^- *!
F I G . I. Results of immunohisUichcniiciil slaiiiings oTseri;!! sections Irom a rheumaloid synovial tissue with a[Ui-I.,eu4 (anti-CD.1) antibodies (a) and ami V/y5.2 4-5.3 antibodies Ib).
684
S. Gudmundsson et al. TABLE I, Numbers of cells stained with certain anti-TCR antibodies in frozen sections from inflamed synovial tis.sucs* Antibody used Synovial tissues
\{fS.2 + i
V^5.3
\fi$_\
V^6.7
CD3
2
1 1 2 1
T-Cell Receptor V-Gene Usage in Synovial Fluid and Synovial Tissue from RA Patients S. GUDMUNDSSON, J R O N N E L I D , A. KARLSSON-PARRA, J. LYSHOLM*, B. GUDBJORNSSONt. B. WIDENFALKJ, C. H. JANSON§ & L. KLARESKOG Department of Clinical Immunology. Uppsala Utiiversity Hospital, Sweden, 'Department of Rheumatology. Falu Hospital, Falun. Sweden, tDepartment of Rheumatology, Uppsala University Hospital, t Department of Hand Surgery, Uppsala University Hospital, and ^Department of Immunology, Karolinska lnstitutet, Stockholm, Sweden
Gudmundsson S. Ronnelid J, Karlsson-Parra A. Lysholm J, Gudbjornsson B, Widenfalk B. Janson CH. Klareskog L. T-Cell Receptor V-Gene Usage in Synovial Fluid and Synovial Tissue from RA Patients. Scand J Immunol 1992:36:681-8 The question of whether Ihcre is a preferential use of certain V genes in T cells entering an inflamed joint has hitherto been studied mainly using unfractionated eells from synovialfluidand tissue respectively, and no clear answer to the queslion has yet been provided. Concomitantly. evidence has been provided that the use of V genes may dilVer considerably between CD4 ' and CD8*^ T cells, and consequently thai detection of biased V-gene expression within an inflammatory lesion may require separate analysis of the Iwo T-cell subsets. In this paper we have therefore studied T-cell receplor V-gene expression in rheumatoid arlhritis by means of double stainings of synovial fluid and blood for available anti-TCR monoclonal antibodies and antibodies to CD4 and CD8. respectively. Double stainings were also performed with anti-TCR antibodies and antibodies to activation markers HLA-DR and 1L-2R. A certain bias towards the preferential use of certain V genes was seen particularly in the synovial fluid samples within both the CD4 ' and CDS * T-cell populations, but no uniform pattern was evident among the 35 patients investigated. Sveinn Gudtnundsson MD. Deparinieiil of Clinical Immunology, Uppsala Unirer.sitv Ho.spiial. S751 85 Upp.sala. Sweden
Activated T lymphocytes are thought to be involved in the development ofsynovitis as well as in other manifestations of rheumatoid arthritis (for a review see Ref. 1). It is likely that at least some of these cells are activated loeally in the synovial tissue, in contact with synovial MHC class ll-expressing cells and antigen [2]. It is also likely that T cells of particular importance for the pathogenesis of the disease are restricted by a particular MHC class II HLA-DR4 molecule which is characterized by a detined sequence in its ji chain [3]. Although these features suggest that the T cells driving the disease may be activated by defined antigen(s) (see discussion in Ref. 3). it has so far been difficult to define such an antigen. Another way to utilize the knowledge of the MHC class II restriction of this T-cell-dependent disease is by taking advantage of the fact that T cells that escape elimination in the thymus. but
are still capable of mediating an autoimmune response, tend to make use of a restricted number of variable T-cell receptor (TCR) genes [4, 5], If T cells restricted by the particular HLA-DR4 fi chain that confers susceptibility to rheumatoid arthritis (RA) would also tend to make use of one or a few variable TCR ji genes, this would open possibilities both for a further analysis of the specificity of T cells driving the disease and for specific therapy where a subset of potentially arthritogenic T cells can be affected (see Refs 68). Analysis of TCR usage among synovial fluid eells, and to some extent synovial tissue eells, has so far been carried out by a number of groups. They have used either a limited number of monoclonal antibodies for analysis of the total number of T cells in the synovial fluid [9], or V^ specific oligonucleotide probes to quantitate 681
682
S. Gudmundsson el al.
mRNA correspotiding to various V/( genes cither in the unfractionated synovial fluid or tissue by means of polymerase chain reaction (PCR) amplification [10 12], or in in vitro expanded T eells by means of direct blotting techniques [1316]. Quite variable results have been obtained, particularly from the analyses on in vitro expanded cell populations, which may be due to selection biases introduced by the various culturing techniques (for results and discussions, sec Refs 14 and 16). The few analyses published using PCR amplifieation techniques have also yielded uneonclusive results so far. In one study a certain relative aeeumulation of V/fM TCR among RA synovial fluid cells was reported [11], which was not eonfirmed in another study [10]; the interpretation ofthesc studies is. so far, also litnitcd by the difficulties encountered in tnaking the actual PCR-based techniques sufficiently quantitative. Thus, it appears that monoclonal antibodies that permit a direct quantitative analysis of synovial cell V-gcne expression remain important tools in the search for restricted V-gene usage among inflammatory cells, something that Is further ctnphasized by the demonstration of different Vgene usage among CD4-and CD8-positiveT cells respectively by means of double itnmtinofluorescence labelling techniques [17]. Against this baekground, we have in the present sludy analysed both synovial fluid and synovial tissue T cells in a number ol" synovial specimens from RA patients using a wider panel ofTCR-speeific monoclonal antibodies than previously published. By means of separate analysis both for CD4- and CD8-positive T-celt subsets, as well as T-cell subsets carrying dilTerent activation markers, we also wanted to see whether there was a bias regarding TCR rearrangement within relevant T-cell subsets.
routine surgery and the biopsies kindly provided by Dr Claes Juhlin, Department of Surgery, Uppsala University Hospital, The synovial biopsies were snap-frozen as earlier described in detail [9] and kept frozen at —70 C until cryosectioncd. Mononucloar cells from the synovial lluid and peripheral blood samples were obtained by Ficoll Hypaque centrifugation. Antihtiilics. Monoclonal antibodies reactive with the following TCR V-gene segments were obtained from TCell Sciences Inc. (Cambridge. MA. USA): V/y5.1 (LC4. Ref. IS); V/i5.2-1-5.3(ICI,Ref. 19); V/y5.3 (WI 12, Ref. 20); V/(6.7 (OTI45, Ref. 21); V/(8 (idGK. Ref. 20) and V^12 (S511, Ref. 22). The monoelonai antibody (MoAb) Fl (Va2.3) was produced in our laboratory [2.1]. Notably, the TCR MoAb listed above may also have reactivities with V-gene segments other than the indieated ones. Other antibodies used were antibodies to gamma-delta and to the ^(/(TCR (T-Cell Sciences), Leu4 (anti-CD.l), Leu3a (anti-CD4). Leu2 (anti-CD8), anti-HLA-DR. and anti-I L-2 receptor (anti-CD2.^) (all from Bctton-Dickinson. Sunnyvale. CA, USA). Imniuni'hisUnhcmiiol stdiuin^s. Immiinohistoehemical stainings were done on frozen sections of 9 biopsies from inflamed synovial tissue from RA patients and from normal lymph nodes and one normal spleen. Stainings were made as previously described [24] utilizing a goat anti-mouse secondary reagent and a monoelonai peroxidase-anti-peroxidase (PAP) reagenl (Dakopatts, Giostrup. Denmark) in tbe third step. In the estimation of the fraetion of T eells stained with the respective anti-TCR MoAbs. the Leu4-positive cells within a given area of the biopsy were first counted. Thereafter, cells stained with the respective anti-TCR MoAb.s were counted on serial sections encompassing the same area. Cylofluarinii'lric analysis. Cytofluorimetric analysis was made on suspended cells from 35 synovial fluid samples and 13 parallel peripheral hlood samples and from 15 normal individuals. Stainings were made with the various anti-TCR antibodies or anli-Leu4 as the primary reagents followed by an FITC-labelled goat anti-mouse IgG antibody, Kor double labellings phycoerythrin-conjugated anti-HLA-DR. anti-CD4. artiCDS and anti-CD25 antibodies (Beeton-Dickinson) were added after blocking the secondary anti-mouse antibody with l"'(i normal mouse serum. Labelled samples were analysed in a FACScan cytofluorimeter (Beeton-Diekinson).
MATERIALS AND METHODS Patients, biopsies and .synorial fluid specimens. Synovial tissue biopsies from patients sufTering from active RA were obtained in conjunction with routine surgery at the Department of Hand Surgery, Uppsala University Hospital. Synovial fluids (SF) from knee joints of active RA patients as well ;is heparinized peripheral blood from these patients were provided from the Department of Rheumatology. Falu Lasarctt, Sweden and from the Department of Rheumatology. Uppsala University Hospital. Peripheral blood (PBL) from nonnal healthy controls was obtained from blood donors. Biopsies from normal human lymph nodes and nonnal spleen were obtained in conjunction with
RESULTS Itwnutwhtstochemical stainings oti biopsies of ilieumatoid synovial tissue Stainings with the V/i5,2-F5.3, V/J5.3, V/i6.7 and anti-gamma-delta TCR were distinctly positive for lymphocyte subpopulations in both lymph nodes and synovial specimens, as was the anti-Lcu4 antibody. No stainings on the frozen seetions were, however, seen with the anti-a^ TCR antibody, and only weak stainings, difficult
TCR I-Gene U.sage in RA Paiienis
683
•fa)
f.
V%S
A
^
•A^-^^
• V
:V«»"^>'•
#•?
•-
riy^"-^^ X:..
•
•
^
^
*
'«.-;--> \*-"^- *!
F I G . I. Results of immunohisUichcniiciil slaiiiings oTseri;!! sections Irom a rheumaloid synovial tissue with a[Ui-I.,eu4 (anti-CD.1) antibodies (a) and ami V/y5.2 4-5.3 antibodies Ib).
684
S. Gudmundsson et al. TABLE I, Numbers of cells stained with certain anti-TCR antibodies in frozen sections from inflamed synovial tis.sucs* Antibody used Synovial tissues
\{fS.2 + i
V^5.3
\fi$_\
V^6.7
CD3
2
1 1 2 1
