Highly Focused Clonal Composition of CD8+ CD28neg T Cellsin Aqueous Humor of Fuchs Heterochromic Cyclitis

Exp. Eye Res. (2002) 75, 317±325 doi:10.1006/exer.2002.2026, available online at http://www.idealibrary.com on

Highly Focused Clonal Composition of CD8‡ CD28neg T Cells in Aqueous Humor of Fuchs Heterochromic Cyclitis P I E R R E L A B A L E T T E * , D AV I D CA I L L A U , C L A U D I N E G R U T Z M AC H E R , J E A N - PA U L D E S S A I N T A N D M Y R I A M L A B A L E T T E a

Service d'Ophtalmologie and bService d'Immunologie, EA 2686, Centre Hospitalier et Universitaire de Lille, Lille 59045, France (Received 29 November 2001 and accepted in revised form 18 March 2002) Fuchs heterochromic cyclitis (FHC) is characterized by a predominant CD8 ‡ T cell subset in®ltrating the anterior chamber, but the clonal composition of these T cells is unknown. In the present study, T cell repertoire diversity of the accumulating T cells was analyzed to investigate if a high degree of restriction could indicate antigen-driven immune response. Aqueous humor (AH) and peripheral blood cells were collected in two patients with FHC. T cell repertoire diversity was screened by T cell receptor (TCR) BV family expression. In one patient, several BV gene segments were used by lymphocytes from the AH but with over-representation of BV16 that accounted for around half of the expressed intraocular repertoire. In the other patient, a more restricted TCRBV usage was found in AH, as only BV15 and BV18 were expressed in the ocular sample. In this patient, virtually all AH CD8 T-cells were CD28- and CD57-negative by three-color ¯ow cytometry, an immunophenotype suggestive of past antigenic stimulation. High resolution immunoscope analysis of TCRBV CDR3 pro®les and sequencing of subcloned TCRBV-BJ PCR products evidenced a highly restricted TCRBV-BJ usage, since virtually all the intraocular cells comprise only ®ve clonotypes in this patient. Unique peaks of CDR3 length were found in BV15 joined to BJ2S1, BJ2S3 and especially BJ2S5, in AH but did not predominate in blood. Conversely, identical clonotypes using rearranged BV18 and BJ2S7 gene segments were detected in both AH and peripheral blood. Maintenance of the TCRBV18-BJ2S7 clonotypes in aqueous humor was demonstrated over 6 months in this patient, with a switch in the predominance of two clonotypes. Our results show the presence of a ®nite number of CD8 ‡ CD28neg T cell clonotypes, which suggests an antigen-driven process and the involvement of these T cells in the pathogenesis of FHC. # 2002 Elsevier Science Ltd. Key words: uveitis; immunopathology; CD8 T lymphocytes; T cell receptor; aqueous humor.

1. Introduction Fuchs heterochromic cyclitis (FHC) (Jones, 1993), also called Fuchs uveitis syndrome (Liesegang, 1982), is a singular uveitis of unknown origin constantly affecting the anterior segment and characterized by a long-standing course of low grade in¯ammation with persistent in®ltrating cells in the anterior chamber. La Hey et al. (1994) proposed that FHC may result from reduced ef®cacy of anterior chamber-associated immune deviation, as the atrophic process of the iris would decrease the constitutive production of immunosuppressive cytokines such as transforming growth factor-b (de Boer et al., 1994; Muhaya et al., 1998; 1999). Several observations have suggested the implication of various infectious agents, ocular trauma and subsequent sensitization against various ocular autoantigens (Toledo de Abreu et al., 1982; Saraux et al., 1985; Jones, 1993; Yamamoto et al., 1996; Barequet et al., 2000). Consistent with the view that CD8 ‡ T cells respond to viruses and other * Address correspondence to: Pierre Labalette, Service d'Ophtalmologie, HoÃpital Huriez, Centre Hospitalier et Universitaire de Lille, 59037 Lille, France. E-mail: [email protected]

0014-4835/02/$35.00/0

intracellular pathogens, recent phenotypic studies have demonstrated that CD8 ‡ T cells dominate the in®ltrating cells into the anterior chamber of patients with FHC (Muhaya et al., 1998). However, it remains to be determined whether the infection occasionally detected initiates the disease or has a bystander role, in reinforcing a pre-existing mild in¯ammatory process or by causing the release of potent autoantigens. Besides, several signi®cant questions regarding the clonal composition, the speci®city, and the role of T cells in initiating and maintaining chronic in¯ammation within the anterior chamber remain unanswered. The accumulation of T cells into the anterior chamber could be driven by a single or several immune recognition events, by a superantigen, or simply re¯ect secondary non antigen-speci®c recruitment by chemoattractants released by local in¯ammation and unopposed by an altered blood±ocular barrier (Whitcup, 2000). Each of these possibilities implies ®nding a different pattern in the T cell repertoire within the aqueous humor (AH). The majority of T cells express a clonally variable antigen receptor, the ab T cell receptor (TCR), which is # 2002 Elsevier Science Ltd.

318

generated through somatic recombination of variable (V), diversity (D; in the b chain), and joining (J) gene elements during T cell ontogeny. Repertoire diversity is further increased by the imprecise joining of these segments and the quasi-random insertion or deletion of nucleotides. Therefore, the complementarity-determining region 3 (CDR3) which encompasses the V±(D)±J junctional region is highly variable both in length and in codon usage between functionally distinct TCR clonotypes, and the corresponding loop spans the major antigen binding site (Garboczi et al., 1996; Garcia et al., 1996; Garcia et al., 1999). Each of the possibilities advanced to explain the presence of T cells in the AH of FHC can thus be assessed by the distribution of TCR rearrangements. Non antigenspeci®c recruitment would give a polyclonal repertoire mirroring that of the blood, characterized by a similar breadth of V subfamilies usage and CDR3 lengths distributed in a near Gaussian manner. A superantigen in the anterior chamber would trigger polyclonal expansion of entire families of T cells bearing one or a limited set of variable segment(s) of the TCR b-chain (TCRBV) but heterogeneous CDR3s. Antigen-driven responses would result in oligoclonal T cell expansions characterized by biased usage of some TCRBV and TCRBJ, with CDR3 regions of ®nite length and de®ned nucleotide sequence (Pannetier et al., 1995; Maini et al., 1999). In this study, we analyzed the TCRBV repertoire of T cells in®ltrating the anterior chamber of two patients with FHC. Three other uveitis patients with mononuclear cells into the anterior chamber were used as control of intraocular in¯ammation. In one of the FHC patients, there was enough material to determine the clonality (TCRBJ usage and CDR3 length and sequence) and the immunophenotype of the intraocular lymphocytes.

P. L A B A L E T T E E T A L .

As expected in FHC, a mild to moderate blood±aqueous barrier breakdown was found in the two patients (23.5 + 2 and 9.9 + 2.3 photon counts msec 1, respectively). Systemic examination for associated disease was negative for both patients, and laboratory examinations were negative, especially for sarcoidosis (blood-ACE level, serum lysozyme, chest X-ray, accessory salivary gland biopsy), lupus and syphilis. As controls, three other uveitis patients were selected, one with bacterial endophthalmitis (postoperative staphylococcal infection), one with BehcËet disease, and one with active ocular toxoplasmosis. The aqueous ¯are was evaluated for each patient, reaching values of (mean + S.D.) 611 + 16.9, 96 + 4.4 and 32.9 + 4.2 photon counts msec 1, respectively. Aqueous Humor Collection AH (0.15±0.25 ml) was obtained at diagnosis by paracentesis under topical anesthesia at the operating room. AH mononuclear cells were isolated by centrifugation. Total numbers of 10 000±40 000 T cells were obtained in AH samples. Peripheral blood mononuclear cells were collected on the same day and isolated from EDTA-anti-coagulated blood by centrifugation over Ficoll-paque gradient. In one of the patients with FHC (FHC#1) a second aqueous tap was analysed, when he complained of worsened blurred vision in his right eye with a mild increase of the aqueous cells and typical keratic precipitates. This research followed the tenets of the declaration of Helsinki, and informed consent was obtained from the patient in accordance to protocols approved by the institutional ethics committee of the Lille University Hospital. Analysis of TCRBV Family Expression

2. Materials and Methods Patients Two young males presenting with unilateral FHC were included. Both patients described a right visual loss. Examination of their anterior segments showed numerous stellate keratic precipitates with a diffuse distribution, mild stromal iris atrophy without posterior synechiae, subcapsular cataract and moderate vitreous haze allowing the diagnosis of FHC in their right eye (Jones, 1993). Furthermore, several iris crystals and some Koeppe nodules on the pupillary margin were noted in the two patients, associated with multiple Bussaca nodules regularly disposed on the collarette of the brown iris of patient 1, as previously described in FHC (Rothova et al., 1994; Callear et al., 1999). Fundus examination disclosed no chorioretinal scar but rare zones of limited retinal vasculitis in this patient. Aqueous ¯are intensity was measured with a KOWA FC-500 laser ¯are meter (Kowa Co. Ltd, Tokyo, Japan).

As AH specimens were limited both in volume and in abundance of mononuclear cells, priority was given to the analysis of TCRBV repertoire. Semi-quantitative retrotranscription-PCR assays were carried out according to published protocols (Pannetier et al., 1993). Brie¯y, total cellular RNA was isolated from AH and peripheral blood lymphocytes isolated as mentioned in the previous section, respectively, using RNA-Plus (Appligene, Illkirch, France) with the addition of 2 ml yeast tRNA (BD Biosciences, Le Pont-de-Claix, France) as a carrier. First stand cDNA was synthesized using oligo-(dT)15 (Roche, Meylan, France) and murine Moloney murine leukemia virus reverse transcriptase (Omniscript, Quiagen, Courtaboeuf, France) in the presence of RNase inhibitor (Biotech, Orsay, France). BV-speci®c cDNA was then ampli®ed using one of a series of 50 sense BV-speci®c primers (TCRBV1-BV24) (Genevee et al., 1992), each paired with a single 30 TCRBC antisense primer. As an internal control, a pair of TCRAC region primers were

O L I G O C LO N A L I T Y O F C D 8 ‡ C D 2 8 n e g A Q U E O U S H U M O R T C E L L S I N F H C

319

included in the reaction tube. The PCR was performed using 5 U of rTth DNA polymerase (Perkin Elmer, Wellesley, MA, U.S.A.) for 30 cycles of ampli®cation at 908C melting, 608C annealing, and 728C extension for 1 min each. The ampli®ed products were sizeseparated by electrophoresis on a 2 % agarose gel, and transferred to a nylon membrane (Hybon-N ‡, Amersham, Les Ulis, France). Blots were pre-hybridized and hybridized with 32P-labelled 50 -AC and 50 -BC probes, and the autoradiograms were analysed by densitometry using Multi-analyst software (Biorad, Hercules, CA, U.S.A.). Representation of each TCRBV family was calculated as the percentage of the sum of all BV signals extracted of the autoradiograms.

vector (Invitrogen, Groningen, The Netherlands) and cloned in Escherichia coli (Invitrogen). Sequencing reactions were conducted on positive clones using standard protocols and nucleotide sequences of the CDR3 region were determined using the ABI-Prism 377 XL sequencing system (Applied Biosystems, Foster Citu, CA, U.S.A.). Diversity was measured by the frequency at which a particular cloned BV18-BJ2S7 insert occurred: n signi®es the number of identical TCR BV18-BJ2S7 junctional region sequences / total number of recombinant clones tested. CDR3 length was determined according to Rock et al. (1994).

High Resolution Immunoscope Analysis of TCRBV CDR3 Pro®le

Antibodies, Immuno¯uorescent Flow Cytometry, and Fluorescent-activated Cell Sorting

Five ml of the cDNA ampli®ed using BV15- and BV18-speci®c primers were further ampli®ed by a second round of 30 PCR cycles run in the same conditions. The PCR products were then subjected to a ®ve-round run-off linear ampli®cation using a series of 13 ¯uorophore-labelled BJ-speci®c primers (Toyonaga et al., 1985). The size and the ¯uorescence intensity of the run-off products were determined on a model 377 DNA sequencer (Perkin Elmer, Wellesley, MA, U.S.A.). Peak intensities were extracted using GENESCAN analysis 2.1 software (Applied Biosystems, Foster Citu, CA, U.S.A.). Percent relative peak intensity is the speci®c peak area/total area of all peaks. CDR3 sizes are calculated in aminoacids.

The following mouse monoclonal antibodies were used: anti-CD4 (T4, FITC-conjugated IgG1), anti-CD8 (CD8, phycoerythrin-cyanin 5.1(PeCy5)-conjugated IgG1), and anti-CD56 (NKH-1, phycoerythrin (PE)-conjugated IgG1) were from Beckman-Coulter (Hialeah, FL, U.S.A.); anti-TCR ab (PE-conjugated IgG1), anti-CD28 (Leu-28, PE-conjugated IgG1), and anti-CD57 (Leu-7, FITC-conjugated IgM) were from Becton Dickinson (Erembodegem, Belgium). Irrelevant mouse IgG1 and IgM directly conjugated with the appropriate ¯uorochrome were used as negative controls. Staining procedures and ¯ow cytometric analysis were carried out as described previously (Hazzan et al., 1997). For each combination of markers, at least 2500 cells were analysed in AH. CD8 ‡ CD28 ‡ and CD8 ‡ CD28neg lymphocyte subsets were 498 % puri®ed from blood lymphocytes by cell sorting on an Epics Elite ¯ow cytometer (Coulter, Miami, FL, U.S.A.) (Labalette et al., 1999).

Sequence Analysis of TCRBV CDR3 Region The cDNA ampli®ed using BV18- and BJ2S7speci®c primers were ligated into the Topo TA cloning

Other Uveitis

1 2 3 4 5S1 5S3 6 7 8 9 11 12 13 14 15 16 17 18 20 21 22 23 24

80 Patient #2 60 40 20 0

1 2 3 4 5S1 5S3 6 7 8 9 11 12 13 14 15 16 17 18 20 21 22 23 24

Percent of VB usage

Percent of VB usage

Fuchs Heterochromic Cyclitis 80 Patient #1 60 40 20 0

100 80 Ocular Toxoplasmosis 60 40 20 0 1 2 3 4 5S1 5S3 6 7 8 9 11 12 13

14 15 16 17 18 20 21 22 23 24

80 Beh et Disease 60 40 20 0

1 2 3 4 5S1 5S3 6 7 8 9 11 12 13 14 15 16 17 18 20 21 22 23 24

80 Endophthalmitis 60 40 20 0

1 2 3 4 5S1 5S3 6 7 8 9 11 12 13 14 15 16 17 18 20 21 22 23 24

F IG . 1. Broad vs restricted TCRBV usage by aqueous humor lymphocytes from patients with various types of uveitis. The representation of individual BV families is shown as a percentage of all the expressed families ampli®ed using BV-speci®c primers.

320

P. L A B A L E T T E E T A L .

3. Results TCRBV Usage in Aqueous Humor T Cells from Uveitis Patients As shown in Fig. 1 (left part), a highly restricted TCRBV usage was found in AH of the ®rst patient with FHC. Only BV15 and BV18 were expressed in the AH sample obtained at diagnosis, while almost all TCRBV families were expressed in the peripheral blood collected on the same day. The second patient with FHC expressed most BV gene segments, but with a marked dominance of BV16 that accounted for almost 50 % of the expressed repertoire. This pattern is similar to the one found in our patient with BehcËet disease (Fig. 1, right part). A more restrictive usage of two TCRBV gene segments was detected in AH of a patient presenting with active ocular toxoplasmosis. Conversely, unrestricted usage of BV families, all expressed at similar frequencies, was observed in AH T cells of a patient with endophthalmitis (Fig. 1, right part). In these different clinical situations associated with various levels of blood±ocular barrier permeability, the diversity of the TCRBV repertoire seemed to increase roughly with the intensity of the breakdown, as estimated by the laser ¯are meter. Clonal Composition of Aqueous Humor T Cells in a Patient with FHC

F IG . 2. Comparison of CD28 and CD57 expression by CD8 ‡ T cells in aqueous humor and in blood of a patient (FHC#1) at diagnosis of Fuchs heterochromic cyclitis. AH and peripheral blood mononuclear cells were stained with ¯uorescent anti-CD8, anti-CD28, and anti-CD57 and analysed by three-color ¯ow cytometry. Numbers indicate the proportions of in each quadrant.

A CD8 ‡ CD28neg T cell population in aqueous humor of FHC#1 patient. Most of the AH lymphocyte population expressed brightly the CD8 marker (90 % of total lymphocytes). Using three-color ¯uorescence, these CD8 ‡ lymphocytes were found to express T cell markers including CD3 and TCRab, but were CD56negative; around half expressed the CD45RO isoform. Virtually all AH CD8 ‡ T cells were lacking CD28 expression and were also CD57-negative (Fig. 2). In peripheral blood, however, CD8 ‡ CD28neg CD57neg, CD8 CD28neg CD57 ‡ and CD8 CD28neg CD57neg T cell subsets each represented around one third of the total CD8 ‡ T cell population.

In BV15 AH T cells, a restricted usage of the BJ2S1, BJ2S3 and especially BJ2S5 gene segments was detected, each with a single peak of CDR3 length (Fig. 3, top). In the corresponding BV15 blood T cells, most BJ gene segments were expressed, four being predominant. The BJ2S1, BJ2S3 and BJ2S5 run-off products displayed the expected Gaussian-like pro®le of CDR3 lengths, indicating polyclonality in blood lymphocytes, and the unique peaks of CDR3 length found in AH T cells did not predominate in blood. Conversely, with BJ2S7 a major peak of CDR3 length (8 aminoacids) was found in blood, but a counterpart of this gene segment could not be detected in AH, where in fact no BJ2S7 transcript was present. In BV18 AH T cells, a highly restricted usage of the BJ2S7 gene segment was recorded, contrary to the representation of most BJ segments in BV18 PBMC (Fig. 3, bottom). The BV18-BJ2S7 transcripts showed a similar oligoclonal pro®le in AH and in peripheral blood T cell, with only two peaks matching in size (8 and 10 aminoacid-long CDR3, respectively) (Fig. 3, bottom).

Highly restricted TCRBV usage in aqueous humor T cells of FHC#1 patient. As shown in Fig. 1, a highly restricted TCRBV usage was found in AH: only BV15 and BV18 were expressed in the AH sample obtained at diagnosis, while almost all TCRBV families were expressed in the peripheral blood collected on the same day. TCR repertoire was further analysed by using TCRBJ-speci®c primers (Fig. 3).

Persisting expansion of CD8 ‡ CD28neg T cells expressing TCRBV18-BJ2S7 with highly restricted CDR3 pro®les in aqueous humor and in peripheral blood of FHC#1 patient. To distinguish between unexpectedly large expansions of a few T cell clones in AH as opposed to the effects of accumulation of a peculiar (i.e. CD28neg) T cell subset in the anterior chamber, TCR repertoire was compared between fractionated

As AH specimens offered limitations both in volume and in abundance of mononuclear cells, the clonal composition of T cell in®ltrating the AH could only be determined in one of the patients with FHC (FHC#1). The immunophenotype of the in®ltrating cells was determined and TCR repertoire was further analyzed by using TCRBJ-speci®c primers.

O L I G O C LO N A L I T Y O F C D 8 ‡ C D 2 8 n e g A Q U E O U S H U M O R T C E L L S I N F H C

321

FHC#1 BJ1S2

BJ1S3

BJ1S4

BJ1S5

BJ1S6

BJ2S1

BJ2S2

BJ2S3

BJ2S4

BJ2S5

BJ2S6

BJ2S7

BJ2S2

BJ2S3

BJ2S4

B J2S5

BJ2S6

BJ2S7

BV15 BJ1S1

BJ1S2

BJ1S3

BJ1S4

BJ1S5

BJ1S6

BJ2S1

AH

Fluorescence intesity Blood AH

Blood

BJ1S1

8 10 aa

BV18

F IG . 3. Comparative BV BJ immunoscope analysis of aqueous humor and peripheral blood T cells from a patient (FHC#1) at diagnosis of FHC. Size distributions of TCR BV-BJ transcripts ampli®ed with either BV15 (top) or BV18 (bottom) primers, then run-off with BJ-speci®c primers. CDR3 size given in aminoacids.

CD8 ‡ CD28 ‡ and CD8 ‡ CD28neg peripheral blood T cells. Almost every BV family was expressed in both subsets with minor difference in the relative levels of some BV segments (data not shown). Comparing the TCRBV18-expressing cells, striking differences were found, however, in the representation of BJ gene segments: there was no distortion in the CD8 ‡ CD28 ‡ subset, which displayed a large spread of TCRBJ usage, whereas only BJ2S7 combined with BV18 was utilized in the CD8 ‡ CD28neg subset. The TCRBV18-BJ2S7 pro®le of CD8 ‡ CD28 ‡ blood T cells comprised successive peaks of CDR3 lengths, but only two CDR3 sizes (8 and 10 aminoacid-long) were detected in CD8 ‡ CD28neg blood T cells (Fig. 4(A)). This highly restricted pro®le in CD8 ‡ CD28neg blood T cells was similar to the BV18-BJ2S7 pro®le obtained in unfractionated blood cells analysed 6 months before, with only slight differences in relative intensities of the two size peaks (Fig. 4(B)). Two matching sizes of 8 and 10 aminoacid-long BV18-BJ2S7 CDR3 were expressed in the AH samples at diagnosis and 6 months later, but with greater differences in relative peak intensities: the 10 aminoacid peak was relatively more represented in the AH collected at 6 months from diagnosis (Fig. 4(B)). Identical junctional sequences in the VB18-JB2S7 transcripts in aqueous humor and in CD8 ‡ CD28neg peripheral blood T cells of FHC#1 patient. In complex mixtures of lymphocytes, many clones may share the same BV, BJ and even CDR3 size, but have distinct

CDR3 sequences. Junctional region sequences of TCRVB18-BJ2S7 transcripts from the AH and peripheral blood T lymphocytes were thus determined (Table I). The majority of TCRBV18-BJ2S7 rearrangements in peripheral blood T cells were identical at the nucleotide sequence of the CDR3 region, corresponding to a clonotypic transcript of 8 aminoacid-long CDR3 with 84 % dominance (16/19 clones) in blood at diagnosis, and 100 % dominance in CD8 ‡ CD28neg T cells tested 6 months later, presumably because polyclonal CD8 ‡ CD28 ‡ T cells have been removed. This CDR3 nucleotide sequence was also detected in AH, together with another clonotypic transcript of 10 aminoacid-long CDR3. Notably, the dominance (45 % vs 55 %, respectively; Table I) of the two clonotypes ranked as the relative peak size intensities (47 % vs 52 %, as shown in Fig. 4(B)).

4. Discussion Expansions of antigen-speci®c T cells occur frequently during in¯ammatory diseases. Global analysis of TCR repertoire using ¯ow cytometric or PCR-based methods has indeed shown that most pathological in®ltrates are either oligoclonal or display oligoclonal expansions over a polyclonal background (Pannetier et al., 1995). This was indeed the case in four of our uveitis patients. In the ®rst patient with FHC (FHC#1), ®ve (oligo)clonal CD8 ‡ CD28neg T cell populations comprise virtually all the cells that in®ltrate the AH, and the T cell clonotypes may persist for at least

322

P. L A B A L E T T E E T A L .

BV18- BJ2S7

A

May 2001

Fluorescence intensity

Blood CD8+ CD28+

Dec. 2000 PBMC

Blood CD8+ CD28neg

AH

AH

8 10 aa

8 10 aa

CDR3 Length

B

RELATIVE PEAK INTENSITIES

Dec. 2000 CDR3 SIZE (amino acids) PBMC AH 8 9 10 11

72% 8% 15% 5%

69% 31% -

May 2001 Blood CD8+ CD28neg

AH

95% 5% -

47% 52% -

F IG . 4. CDR3 length distribution analysis of TCR BV18BJ2S7 transcripts from aqueous humor and blood T cells collected in a patient (FHC#1) with FHC at presentation (Dec. 2000) and 6 months later (May 2001). (A) CDR3 pro®les are displayed in ¯uorescence intensity (arbitrary units) as a function of CDR3 size (aminoacids) of TCRBV18BJ2S7 transcripts in unfractionated peripheral blood lymphocytes, puri®ed CD8 ‡ CD28 ‡ and CD8 ‡ CD28neg blood T cells, and AH T cells tested on the indicated time. (B) Relative peak intensities of the CDR3 length pro®les presented above.

6 months in the anterior chamber of the affected eye of this patient. In the second patient (FHC#2), although more TCRBV gene segments were utilized, a marked dominance of one BV family was seen in intraocular lymphocytes. Distinct BV gene families were used prevalently in the two patients, however. The patient with active ocular toxoplasmosis presented with a highly restrictive pattern of only two BV families, one highly dominant, which argues for an expansion driven by antigen or superantigen stimulation (Purner et al., 1998) as expected in an intraocular infection where Toxoplasma gondii-speci®c

T cells have been reported in intraocular ¯uids (Feron et al., 2001). The AH repertoire of a patient with BehcËet disease, a relapsing multisystem in¯ammatory disorder, is made of many BV families, one being highly prevalent, a result in agreement with a previous report (Keino et al., 2000). In the patient with bacterial endophthalmitis, however, a highly diverse repertoire was characterized in the AH sample. This suggests a polyclonal response in a persisting massive intraocular infection, presumably favored by a leaky blood±ocular barrier, as indicated by a high photon count with laser ¯are meter. Laser ¯are meter studies in our two patients with FHC have shown on the contrary that the alteration of the blood±aqueous barrier is moderate, in agreement with comparative studies between FHC and other variants of chronic anterior uveitis (Norrsell et al., 1998; Kuchle and Nguyen, 2000). One patient with FHC (FHC#1) could be investigated more extensively. Our study shows that essentially all (99 %) the CD8 ‡ T cell population in®ltrating the AH of the FHC was lacking expression of CD28, with around half expressing the CD45RO isoform, an immunophenotype suggestive of past antigenic stimulation (Labalette et al., 1999; Nociari et al., 1999; Posnett et al., 1999; Weekes et al., 1999; Borthwick et al., 2000). Large or persistent clonal expansions are indeed more commonly detected in the CD8 ‡ T cell population (McMichael and O'Callaghan, 1998; Maini et al., 1999), particularly in the circulating CD28neg and CD45RO ‡ subsets of CD8 ‡ T lymphocytes (Hingorani et al., 1993; Batliwalla et al., 1996, 2000; Posnett et al., 1999; Wedderburn et al., 1999; Weekes et al., 1999). A much more restricted pattern of TCRBV and BJ gene usage was seen in AH than in blood of the FHC#1 patient, even when compared to sorted CD8 ‡ CD28neg blood T cells. By analysis of CDR3 length, of the ®ve predominant clonotypes characterized in the AH sample, two were also detected in the CD8 ‡ CD28neg subset of peripheral blood T cells of the patient. These ®ndings raise the question of how the clonotypes are recruited into, survive, and possibly expand within the anterior segment in FHC. On stimulation, T cells proliferate by clonal expansion and differentiate, which enable them to migrate through the blood to sites of in¯ammation (Lanzavecchia and Sallusto, 2000; von Andrian and Mackay, 2000). However, some single-sized CDR3 peaks found in circulating CD8 ‡ CD28neg T cells of the patient, as exempli®ed by the BV15-BJ2S7 transcript, were not represented in the AH. Likewise, CD8 ‡ CD28neg T cells using BV15 with the BJ2S1, BJ2S3 and BJ2S5 gene segments displayed several peaks of CDR3 length in blood, while a single peak size was found in AH T cells. These data are in agreement with other reports showing non-overlapping T cell oligoclonality between in¯amed sites and peripheral blood (Scotet et al., 1999), even when circulating CD8 ‡

O L I G O C LO N A L I T Y O F C D 8 ‡ C D 2 8 n e g A Q U E O U S H U M O R T C E L L S I N F H C

323

TABLE I Nucleotide and deduced amino acid sequences of TCRVB18-BJ2S7 junctional regions* in AH and peripheral T cells of a patient with Fuchs heterochromic cyclitis (FHC#1) CDR3{ Length

TCRBV18

-NDN-{

BJ2S7

Frequency}

PBMC (Dec. 2000)

8

- TCA S

GACAGCGGC DSG

GACGAGCAGTACDEQY

16/19

CD8 ‡ CD28neg blood T cells (May 2001)

8

- TCA S

GACAGCGGC DSG

GACGAGCAGTACDEQY

9/9

CD8 ‡ CD28neg AH T cells (May 2001)

8

- TCA S - TCA S

GACAGCGGC DSG CCGCCGACAGATCCC PPTDP

GACGAGCAGTACDEQY TACGAGCAGTACYEQY

4/9

10

5/9

* Transcripts were reverse-transcribed and ampli®ed using BV18 and BJ2S7 primers, cloned and sequenced as described in Materials and Methods. { CDR3 length was determined according to Rock et al. (1994). { CDR3 sequences are given as one-letter amino-acid code. } Frequency of bacterial clones displaying the indicated sequence.

CD28neg T cells were analyzed (Wedderburn et al., 1999). Our results make it clear that the T cell population accumulating in the AH is not simply a `concentrated' pool of peripheral T cells with features of past antigenic stimulation that would be collected through a leaky blood±aqueous barrier. Nevertheless, as these cells were detected in the peripheral blood, they could potentially migrate into various tissues; we cannot completely exclude that the eye is only one of the organs where they do accumulate, but the unilateral disease argues against the possibility that the clones are sequestrated in an unspeci®c manner. Furthermore, the BV gene segments found to be overexpressed in our patients are not the ones utilized in (oligoclonal) NKT cells (Godfrey et al., 2000). We cannot exclude, however, that the oligoclonal CD8 ‡ CD28neg T cells found in the FHC#1 patient may have gained expression of some NK cell receptors, as reported for currently activated, effector CD8 ‡ CD28neg T cells (McMahon and Raulet, 2001). Our data show that distinct BV gene families were used prevalently by intraocular lymphocytes between the two patients and the highly restricted usage of CDR3 length and sequence by AH T cells in the FHC#1 patient altogether argues against the involvement of a superantigen in FHC and point to antigendriven stimulation of the clones detected in the AH. The identi®cation of identical clonotypes using rearranged BV18 and BJ2S7 gene segments in blood and in AH of the ®rst patient argues for a circulation of these clonotypes. FHC is not a systemic disease, but experimental models indicate that effector or regulatory T cells speci®c for antigen from the eye can be activated in the peripheral lymphoid compartment before reaching the anterior chamber (Streilein et al., 1997; Niederkorn, 1999). Recruitment of speci®c T cell clones into (Jinquan et al., 1993) and survival within (Taga et al., 1994) the AH could be facilitated

by IL-10, as levels of this cytokine are reportedly higher in FHC than in other forms of uveitis (Muhaya et al., 1998). Besides, CD8 ‡ CD28neg T cells are less prone to apoptotic cell death (Posnett et al., 1999). Maintenance of the TCRBV18-BJ2S7 clonotypes in AH was demonstrated for 6 months after the ®rst visit of the patient, albeit with a switch in the predominance of two clonotypes expressed in the AH while the proportion of the corresponding two clonotypes in blood only varied a little. This observation suggests either that the recruitment into the AH of speci®c T cells with the 10 aminoacid-long CDR3 augmented, or that an in®ltrating clone had been stimulated to expand in situ by an antigen, whether foreign or self, present in the anterior segment. Whatever happens, the presence of a small number of CD8 ‡ CD28neg T cell clonotypes suggests the involvement of these T cells in the pathogenesis of FHC. The limited number of cells collected from the AH did not allow functional analysis, but it is known that CD8 ‡ CD28neg T lymphocytes may exert effector function (i.e. cytotoxicity, production of diverse cytokine pattern) (Azuma et al., 1993; Hamann et al., 1999; Labalette et al., 1999; Nociari et al., 1999; Borthwick et al., 2000) and therefore may represent pathogenic lymphocytes. CD8 ‡ CD28neg T lymphocytes may instead exert suppressor or regulatory functions, including the production of IL-10 (Hamann et al., 1999; Nociari et al., 1999), thereby possibly contributing to the benign clinical course of FHC. Contrary to interleukin-2 and interleukin-15 (Borthwick et al., 2000), interleukin-4 has been shown to maintain membrane expression of CD28 and to favor acquisition of the CD57 molecule by human CD8 ‡ T cells that differentiate upon activation (d'Angeac et al., 1994; Labalette et al., 1999). The CD8 ‡ CD28neg CD57neg T cells that accumulate in the AH could then be activated to CD8 ‡ T cells that

324

expand in a cytokine environment poor in interleukin-4. Accordingly, measurement of cytokine levels in ocular ¯uids constantly failed to detect interleukin-4 (Ongkosuwito et al., 1998; Muhaya et al., 1999). In conclusion, this report represents the ®rst characterization of the TCR repertoire of T cells in®ltrating the AH of two patients with FHC and demonstrates the persistence of clonally accumulated CD8 ‡ CD28neg CD57neg T cells in the AH of the patient that could be studied more extensively. Demonstration of the pivotal role of speci®c CD8 CD28neg CD57neg T cells in anterior uveitis now awaits the detection of similar clonal accumulations in the anterior chamber of other patients. Besides providing a better understanding of the pathogenesis of ocular in¯ammatory diseases, functional analysis of the clones will help searching the involved antigens, be they infectious organisms or autoantigens of the eye.

Acknowledgements This work was supported by a grant from the Projet Hospitalier de Recherche Clinique 1997. The authors thank Michel CreÂpin and Christine Mouton.

References Azuma, M., Phillips, J. H. and Lanier, L. L. (1993). CD28- T lymphocytes. Antigenic and functional properties. J. Immunol. 150, 1147±59. Barequet, I. S., Li, Q., Wang, Y., O'Brien, T. P., Hooks, J. J. and Stark, W. J. (2000). Herpes simplex virus DNA identi®cation from aqueous ¯uid in Fuchs heterochromic iridocyclitis. Am. J. Ophthalmol. 129, 672±3. Batliwalla, F., Monteiro, J., Serrano, D. and Gregersen, P. K. (1996). Oligoclonality of CD8 ‡ T cells in health and disease: aging, infection, or immune regulation?. Hum. Immunol. 48, 68±76. Batliwalla, F. M., Rufer, N., Lansdorp, P. M. and Gregersen, P. K. (2000). Oligoclonal expansions in the CD8(‡)CD28( ) T cells largely explain the shorter telomeres detected in this subset: analysis by ¯ow FISH. Hum. Immunol. 61, 951±8. Borthwick, N. J., Lowdell, M., Salmon, M. and Akbar, A. N. (2000). Loss of CD28 expression on CD8(‡) T cells is induced by IL-2 receptor gamma chain signalling cytokines and type I IFN, and increases susceptibility to activation-induced apoptosis. Int. Immunol. 12, 1005±13. Callear, A. B., Reynolds, A., Harry, J. and Murray, P. I. (1999). Iris crystals in chronic uveitis. Br. J. Ophthalmol. 83, 703±6. d'Angeac, A. D., Monier, S., Pilling, D., Travaglio-Encinoza, A., Reme, T. and Salmon, M. (1994). CD57‡ T lymphocytes are derived from CD57- precursors by differentiation occurring in late immune responses. Eur. J. Immunol. 24, 1503±11. de Boer, J. H., Limpens, J., Orengo-Nania, S., de Jong, P. T., La Heij, E. and Kijlstra, A. (1994). Low mature TGFbeta 2 levels in aqueous humor during uveitis. Invest. Ophthalmol. Vis. Sci. 35, 3702±10. Feron, E. J., Klaren, V. N., Wierenga, E. A., Verjans, G. M. and Kijlstra, A. (2001). Characterization of Toxoplasma gondii-speci®c T cells recovered from vitreous ¯uid of

P. L A B A L E T T E E T A L .

patients with ocular toxoplasmosis. Invest. Ophthalmol. Vis. Sci. 42, 3228±32. Garboczi, D. N., Ghosh, P., Utz, U., Fan, Q. R., Biddison, W. E. and Wiley, D. C. (1996). Structure of the complex between human T-cell receptor, viral peptide and HLAA2. Nature. 384, 134±41. Garcia, K. C., Degano, M., Stan®eld, R. L., Brunmark, A., Jackson, M. R. and Peterson, P. A et al. (1996). An alphabeta T cell receptor structure at 2.5 A and its orientation in the TCR-MHC complex. Science. 274, 209±19. Garcia, K. C., Teyton, L. and Wilson, I. A. (1999). Structural basis of T cell recognition. Ann. Rev. Immunol. 17, 369±97. Genevee, C., Diu, A., Nierat, J., Caignard, A., Dietrich, P. Y. and Ferradini, L. et al. (1992). An experimentally validated panel of subfamily-speci®c oligonucleotide primers (V alpha 1-w29/V beta 1-w24) for the study of human T cell receptor variable V gene segment usage by polymerase chain reaction. Eur. J. Immunol. 22, 1261±9. Godfrey, D. I., Hammond, K. J., Poulton, L. D., Smyth, M. J. and Baxter, A. G. (2000). NKT cells: facts, functions and fallacies. Immunol. Today. 21, 573±83. Hamann, D., Roos, M. T. and van Lier, R. A. (1999). Faces and phases of human CD8 T-cell development. Immunol. Today. 20, 177±80. Hazzan, M., Labalette, M., Noel, C., Lelievre, G. and Dessaint, J. P. (1997). Recall response to cytomegalovirus in allograft recipients: mobilization of CD57‡, CD28‡ cells before expansion of CD57‡, CD28 cells within the CD8‡ T lymphocyte compartment. Transplantation. 63, 693±8. Hingorani, R., Choi, I. H., Akolkar, P., Gulwani-Akolkar, B., Pergolizzi, R. and Silver, J. et al. (1993). Clonal predominance of T cell receptors within the CD8‡ CD45RO‡ subset in normal human subjects. J. Immunol. 151, 5762±9. Jinquan, T., Larsen, C. G., Gesser, B., Matsushima, K. and Thestrup-Pedersen, K. (1993). Human IL-10 is a chemoattractant for CD8‡ T lymphocytes and an inhibitor of IL-8-induced CD4‡ T lymphocyte migration. J. Immunol. 151, 4545±51. Jones, N. P. (1993). Fuchs' heterochromic uveitis: an update. Surv. Ophthalmol. 37, 253±72. Keino, H., Sakai, J., Nishioka, K., Sumida, T. and Usui, M. (2000). Clonally accumulating T cells in the anterior chamber of Behcet disease. Am. J. Ophthalmol. 130, 243±5. Kuchle, M. and Nguyen, N. X. (2000). Analysis of the blood aqueous barrier by measurement of aqueous ¯are in 31 eyes with Fuchs' heterochromic uveitis with and without secondary open-angle glaucoma. Klin. Monatsbl. Augenheilkd. 217, 159±62. Labalette, M., Leteurtre, E., Thumerelle, C., Grutzmacher, C., Tourvieille, B. and Dessaint, J. P. (1999). Peripheral human CD8(‡)CD28(‡)T lymphocytes give rise to CD28( )progeny, but IL-4 prevents loss of CD28 expression. Int. Immunol. 11, 1327±36. La Hey, E., de Jong, P. T. and Kijlstra, A. (1994). Fuchs' heterochromic cyclitis: review of the literature on the pathogenetic mechanisms. Br. J. Ophthalmol. 78, 307±12. Lanzavecchia, A. and Sallusto, F. (2000). Dynamics of T lymphocyte responses: intermediates, effectors and memory cells. Science. 290, 92±7. Liesegang, T. J. (1982). Clinical features and prognosis in Fuchs' uveitis syndrome. Arch. Ophthalmol. 100, 1622±6.

O L I G O C LO N A L I T Y O F C D 8 ‡ C D 2 8 n e g A Q U E O U S H U M O R T C E L L S I N F H C

Maini, M. K., Casorati, G., Dellabona, P., Wack, A. and Beverley, P. C. (1999). T-cell clonality in immune responses. Immunol. Today. 20, 262±6. McMahon, C. W. and Raulet, D. H. (2001). Expression and function of NK cell receptors in CD8‡ T cells. Curr. Opin. Immunol. 13, 465±70. McMichael, A. J. and O'Callaghan, C. A. (1998). A new look at T cells. J. Exp. Med. 187, 1367±71. Muhaya, M., Calder, V., Towler, H. M., Shaer, B., McLauchlan, M. and Lightman, S. (1998). Characterization of T cells and cytokines in the aqueous humour (AH) in patients with Fuchs' heterochromic cyclitis (FHC) and idiopathic anterior uveitis (IAU). Clin. Exp. Immunol. 111, 123±8. Muhaya, M., Calder, V. L., Towler, H. M., Jolly, G., McLauchlan, M. and Lightman, S. (1999). Characterization of phenotype and cytokine pro®les of T cell lines derived from vitreous humour in ocular in¯ammation in man. Clin. Exp. Immunol. 116, 410±4. Niederkorn, J. Y. (1999). Anterior chamber-associated immune deviation. Chem. Immunol. 73, 59±71. Nociari, M. M., Telford, W. and Russo, C. (1999). Postthymic development of CD28-CD8‡ T cell subset: age-associated expansion and shift from memory to naive phenotype. J. Immunol. 162, 3327±35. Norrsell, K., Holmer, A. K. and Jacobson, H. (1998). Aqueous ¯are in patients with monocular iris atrophy and uveitis. A laser ¯are and iris angiography study. Acta Ophthalmol. Scand. 76, 405±12. Ongkosuwito, J. V., Feron, E. J., van Doornik, C. E., Van der Lelij, A., Hoyng, C. B. and La Heij, E. C. et al. (1998). Analysis of immunoregulatory cytokines in ocular ¯uid samples from patients with uveitis. Invest. Ophthalmol. Vis. Sci. 39, 2659±65. Pannetier, C., Cochet, M., Darche, S., Casrouge, A., Zoller, M. and Kourilsky, P. (1993). The sizes of the CDR3 hypervariable regions of the murine T-cell receptor beta chains vary as a function of the recombined germ-line segments. Proc. Nat. Acad. Sci. U.S.A. 90, 4319±23. Pannetier, C., Even, J. and Kourilsky, P. (1995). T-cell repertoire diversity and clonal expansions in normal and clinical samples. Immunol. Today. 16, 176±81. Posnett, D. N., Edinger, J. W., Manavalan, J. S., Irwin, C. and Marodon, G. (1999). Differentiation of human CD8 T cells: implications for in vivo persistence of CD8 ‡ CD28 cytotoxic effector clones. Int. Immunol. 11, 229±41. Purner, M. B., Berens, R. L., Tomavo, S., Lecordier, L., Cesbron-Delauw, M. F. and Kotzin, B. L. et al. (1998). Stimulation of human T lymphocytes obtained from Toxoplasma gondii-seronegative persons by proteins derived from T. gondii. J. Infect. Dis. 177, 746±53.

325

Rock, E. P., Sibbald, P. R., Davis, M. M. and Chien, Y. H. (1994). CDR3 length in antigen-speci®c immune receptors. J. Exp. Med. 179, 323±8. Rothova, A., La Hey, E., Baarsma, G. S. and Breebaart, A. C. (1994). Iris nodules in Fuchs' heterochromic uveitis. Am. J. Ophthalmol. 118, 338±42. Saraux, H., Laroche, L. and Le Hoang, P. (1985). Secondary Fuchs's heterochromic cyclitis: a new approach to an old disease. Ophthalmologica. 190, 193±8. Scotet, E., Peyrat, M. A., Saulquin, X., Retiere, C., Couedel, C. and Davodeau, F. et al. (1999). Frequent enrichment for CD8 T cells reactive against common herpes viruses in chronic in¯ammatory lesions: towards a reassessment of the physiopathological signi®cance of T cell clonal expansions found in autoimmune in¯ammatory processes. Eur. J. Immunol. 29, 973±85. Streilein, J. W., Takeuchi, M. and Taylor, A. W. (1997). Immune privilege, T-cell tolerance, and tissue-restricted autoimmunity. Hum. Immunol. 52, 138±43. Taga, K., Chretien, J., Cherney, B., Diaz, L., Brown, M. and Tosato, G. (1994). Interleukin-10 inhibits apoptotic cell death in infectious mononucleosis T cells. J. Clin. Invest. 94, 251±60. Toledo de Abreu, M., Belfort, R., Jr. and Hirata, P. S. (1982). Fuchs' heterochromic cyclitis and ocular toxoplasmosis. Am. J. Ophthalmol. 93, 739±44. Toyonaga, B., Yoshikai, Y., Vadasz, V., Chin, B. and Mak, T. W. (1985). Organization and sequences of the diversity, joining, and constant region genes of the human T-cell receptor beta chain. Proc. Nat. Acad. Sci. U.S.A. 82, 8624±8. von Andrian, U. H. and Mackay, C. R. (2000). T-cell function and migration. Two sides of the same coin. N. Engl. J. Med. 343, 1020±34. Wedderburn, L. R., Maini, M. K., Patel, A., Beverley, P. C. and Woo, P. (1999). Molecular ®ngerprinting reveals non-overlapping T cell oligoclonality between an in¯amed site and peripheral blood. Int. Immunol. 11, 535±43. Weekes, M. P., Carmichael, A. J., Wills, M. R., Mynard, K. and Sissons, J. G. (1999). Human CD28-CD8‡ T cells contain greatly expanded functional virus-speci®c memory CTL clones. J. Immunol. 162, 7569±77. Whitcup, S. M. (2000). The double-edged ocular immune response: the Cogan lecture. Invest. Ophthalmol. Vis. Sci. 41, 3243±8. Yamamoto, S., Pavan-Langston, D., Kinoshita, S., Nishida, K., Shimomura, Y. and Tano, Y. (1996). Detecting herpesvirus DNA in uveitis using the polymerase chain reaction. Br. J. Ophthalmol. 80, 465±8.