Transaldolase (TAL) is expressed at selectively high levels in oligodendrocytes and targeted by autoreactive T cells of patients with multiple sclerosis (MS). Among 14 TAL peptides with predicted HLA-A2 binding, TAL 168–176 (LLFSFAQAV, TALpep) exhibited high affinity for HLA-A2. Prevalence of HLA-A2-restricted CD8+ T cells specific for TALpep was increased in PBMC of HLA-A2+ MS patients, as compared with HLA-A2− MS patients, HLA-A2+ other neurological disease patients, and HLA-A2+ healthy donors. HLA-A*0201/TALpep tetramers detected increased frequency of TAL-specific CD8+ T cells, and precursor frequency of TAL-specific IFN-γ-producing T cells was increased in each of seven HLA-A2+ MS patients tested. Stimulation by TALpep or rTAL of PBMC from HLA-A2+ MS patients elicited killing of TALpep-pulsed HLA-A2-transfected HmyA2.1 lymphoma cells, but not HLA-A3-transfected control HmyA3.1 targets. Without peptide pulsing of targets, HLA-A2-transfected, but not control MO3.13 oligodendroglial cells, expressing high levels of endogenous TAL, were also killed by CD8+ CTL of MS patients, indicating recognition of endogenously processed TAL. TCR Vβ repertoire analysis revealed use of the TCR Vβ14 gene by T cell lines (TCL) of MS patients generated via stimulation by TAL- or TALpep-pulsed APCs. All TAL-specific TCL-binding HLA-A*0201/TALpep tetramers expressed TCR Vβ14 on the cell surface. Moreover, Ab to TCR Vβ14 abrogated cytotoxicity by HLA-A2-restricted TAL-specific TCL. Therefore, TAL-specific CTL may serve as a novel target for therapeutic intervention in patients with MS.
Multiple sclerosis (MS)3 is considered to be an autoimmune disease leading to progressive loss of oligodendrocytes and demyelination in the CNS. In the acute stage of disease, lesions contain macrophages, CD4+ and CD8+ T cells, and Ig deposits, suggesting that the demyelination process is mediated by the immune system (1, 2, 3, 4). The inflammatory picture of early lesions, which is followed by a progressive gliosis, also suggested that the pathological process may be initiated by infectious agents and then self-perpetuated by a cross-reactive autoimmune process (5, 6, 7, 8, 9, 10). Although the Ag(s) driving this self-destructive process in MS has not been identified (11), the importance of myelin-derived Ags was demonstrated by their abilities to elicit an MS-like demyelinating disease, experimental allergic encephalomyelitis (EAE), in various animal models (12).
The major difficulties in applying the EAE model to MS stem from a lack of identification of relevant autoantigen(s). Studies on relapsing EAE have shown that different encephalitogenic molecules or epitopes within them are selected that are compatible with the heterogeneity of the immune response in MS, suggesting that disease initiation and relapse episodes are induced by different Ags (13, 14). Although cell-mediated mechanisms may have a primary role in EAE, augmentation of humoral immunity within the CNS is a well-recognized feature of MS (15). A breakdown of the blood-brain barrier would allow Abs to enter the CNS and cause demyelination by complement activation. In fact, complement may be directly involved in death of oligodendrocytes (16). Most efforts have been focused on myelin basic protein (MBP) and proteolipid protein that make up as much as 30 and 50% of CNS myelin, respectively (17, 18). T cell responses to MBP and proteolipid protein, or another oligodendrocyte-specific protein myelin oligodendrocyte protein (MOG), did not differ considerably between MS patients and control donors (11). Although oral vaccination with a predefined inducing Ag may successfully prevent and treat disease in animal models (19), a similar approach with MBP in 30 patients with MS led to no significant clinical improvement (20).
Molecular mimicry, i.e., cross-reactivity between self Ags and viral proteins, has been implicated in the initiation of autoimmunity and MS. Based on homology to retroviral sequences, a novel autoantigen, partially encoded by a retrotransposon and selectively expressed in oligodendrocytes at high levels (21), was identified as human transaldolase (TAL) (22). TAL is a key enzyme of the pentose phosphate pathway (PPP). Although glucose is largely metabolized through the glycolytic pathway and the tricarboxylic acid cycle, the significance of PPP in the brain has long been established. During brain development, PPP provides NADPH for the biosynthesis of lipids (23). The latter is particularly important at the period of active myelination (24, 25). The overall activity of PPP in the brain declines 5-fold from birth to maturity (26). Although under normal conditions only as little as 1% of the glucose enters the PPP (27), at times of rapid myelination up to 60% of the glucose is metabolized via the PPP (28). Involvement of PPP in myelination provided a physiological explanation for the high level of TAL in oligodendrocytes (21, 22, 23, 24, 25, 26, 27, 28, 29). PPP also plays an essential role in neutralization of oxygen radicals, and elevated levels of TAL expression increase susceptibility to apoptosis signals (30, 31, 32). Effector phase of the demyelination process in MS is thought to be mediated by reactive oxygen intermediates. Intralesional CTLs produce TNF-β, which, in turn, induces apoptosis, an oxidative stress-mediated programmed cell death, of oligodendrocytes (33). The Fas receptor/ligand system has also been implicated in death of oligodendrocytes (34, 35). Thus, oligodendrocyte-specific expression of TAL is possibly linked to production of large amounts of lipids, as a major component of myelin and vulnerability of the vast network of myelin sheaths to oxygen radicals. Immunohistochemical studies of postmortem brain sections revealed decreased staining by MBP- and TAL-specific Abs in MS plaques, indicating a concurrent loss of these Ags from sites of demyelination (29). Patients with MS have TAL Abs in their blood and cerebrospinal fluid (21, 22, 23, 24, 25, 26, 27, 28, 29). TAL autoantibodies recognize immunoblotted and three-dimensional epitopes and inhibit enzymatic activity of TAL (36). By constrast, TAL Abs were absent in normal individuals and patients with other autoimmune and neurological diseases (21) and, under identical conditions, no MBP Abs were found in serum and cerebrospinal fluid of MS patients (29). The fact that TAL Abs were absent in controls including patients with other neurological disease (OND) or systemic autoimmune diseases such as SLE and Sjögren syndrome indicated that the autoimmune process targeting TAL may be specific for MS. In addition, TAL elicit proliferation, aggregate formation, and skewing of the TCR Vβ repertoire of peripheral blood T lymphocytes from patients with MS with respect to MBP as control Ag. The results suggested that TAL may be a more significant target than MBP of myelin-reactive T cells and of humoral autoreactivity in patients with MS (29).
Most reports attribute oligodendrocyte cell death, at least in part, to the cytotoxic effect of CD8+ T cells (37). Adoptively transferred MBP-specific (38) or MOG-specific CD8+ T cells induce severe CNS demyelination in animal models (39). In MS brain lesions, infiltrating CD8+ CTL were reported to outnumber CD4+ T cells 10-fold (2). Also, actively demyelinating lesions of MS brains are enriched for clonally expanded CD8+ T cells as compared with CD4+ T cells (40, 41). These myelin-reactive CD8+ T cells may play a role in recruitment and retention of myelin-reactive CD4+ T cells by secreting proinflammatory soluble mediators to promote and mediate inflammatory response in MS brain. In the present study, we show that among 14 TAL peptides tested, TAL peptide 168–176 (TALpep) has high binding affinity for HLA-A2. Full-length human rTAL- or TALpep-stimulated CTL from HLA-A2+ MS patients kill TALpep-pulsed HLA-A2-transfected (Hmy A2.1), but not HLA-A3-transfected control target cells. By contrast, HLA-A2+ healthy or OND controls and HLA-A2− control donors do not show increased killing of peptide-pulsed Hmy A2.1 targets. HLA-A2-transfected, but not control MO3.13 oligodendroglial cells, expressing high levels of endogenous TAL, were also killed by CD8+ CTL of MS patients without peptide pulsing the targets, thus indicating that endogenously processed TAL was recognized by HLA-A2-restricted CTL. As detected by PE-conjugated HLA-A*0201/TALpep tetramer staining, the frequencies of TALpep-specific CD8+ T cells were increased in PBL of HLA-A2+ MS patients when compared with HLA-A2+ OND and healthy controls (HC), or HLA-A2− MS patients. In response to stimulation by rTAL or TALpep, precursor frequency of IFN-γ-producing T cells was increased in HLA-A2+ MS patients. T cell lines (TCL) of HLA-A2+ MS patients raised by stimulation with TAL- or TALpep-pulsed APC derived from autologous PBMC or EBV-transformed B cells expressed TCR Vβ14 on their cell surface. Pretreatment of TCL lines with TCR Vβ14 mAb abrogated their cytotoxicity against HLA-A2-transfected Hmy A2.1 or MO3.13 cells, suggesting that TAL-specific CTL may serve as a therapeutic target to prevent oligodendrocyte destruction in patients with MS.
Materials and Methods
Patient and control cells
Fourteen patients with MS, eight patients with OND, and eight HC donors were investigated (Table I). Patients with MS were diagnosed according to the criteria of Poser et al. (42). PBMC were isolated from heparinized venous blood on Ficoll-Hypaque gradient and resuspended in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.25 mg/ml amphotericin B, and cultured in a humidified atmosphere with 5% CO2 at 37°C. Cell culture products were obtained from Mediatech. PBMC were HLA class I Ag typed by the National Institutes of Health microcytotoxicity (43). Viability and functional capacity of all PBMC samples have been monitored by trypan blue exclusion (>95%) and PHA-induced proliferative responses by [3H]TdR incorporation, respectively (29).
|Donor .||HLA-A2 .||Age/Sex .||Dx Duration (years) .||Dx/Type .||Hmy .||MO3.13 .||Tetramer .||ELISPOT .|
|Donor .||HLA-A2 .||Age/Sex .||Dx Duration (years) .||Dx/Type .||Hmy .||MO3.13 .||Tetramer .||ELISPOT .|
Dx, Diagnosis; M, male; F, female; R/R, relapsing/remitting; PP, primary progressive; SP, secondary progressive; S, stable; OND, other neurological disease; PD, Parkinson’s disease; ALS, amyotrophic lateral sclerosis; HC, healthy control; Hmy, HLA-A2-restricted TALpep-specific cytotoxicity to Hmy cells; MO3.13, HLA-A2-restricted TALpep-specific cytotoxicity to MO3.13 cells; Tetramer, frequency of HLA-A*0201/TALpep tetramer-positive cells in PBMC stimulated with TALpep for 7 days; ELISPOT, frequency of IFN-γ-producing cells in PBMC stimulated with TALpep for 7 days; +, values above mean + 3 SD of all healthy and OND controls; −, values below mean + 3 SD of all healthy and OND controls; N/A, not applicable.
Ags and Abs
Full-length rTAL-H protein (clone 1425) was expressed as a fusion protein with GST by pGEX-2T plasmid vector (44), affinity purified through binding of GST to glutathione-coated agarose beads, cleaved from GST by 1 National Institutes of Health unit of thrombin (Sigma-Aldrich), and separated from the agarose bead-bound GST by centrifugation (21). Control GST Ag was prepared by elution from glutathione-coated agarose beads in 10 mM reduced glutathione and 50 mM Tris-HCl (pH 8.0). The purified full-length rTAL was quantified by the Bradford assay, analyzed by SDS-PAGE and Western blot, and tested for TAL enzyme activity, as earlier described (22). Synthetic peptides were produced and purified to >99% homogeneity by Genemed (Genemed Synthesis). BSA, Con A, and PHA were obtained from Sigma-Aldrich. Highly specific polyclonal rabbit Abs 169 and 170 directed to the 139-aa-long N-terminal segment of TAL were developed earlier (22).
Prediction and identification of HLA-A2-binding TAL peptide epitopes
The human TAL amino acid sequence was analyzed for the presence of nonamer peptides with HLA class I-binding motifs using a computer-based algorithm that predicts the stability of HLA-Ag/peptide complexes by quantitating positive and negative effects on binding of each amino acid within a nonamer peptide (National Institutes of Health BioInformatics & Molecular Analysis Section website server: 〈http://bimas.dcrt.nih.gov/molbio/hla_bind/〉). Binding of peptides to HLA-A2 was measured by incorporation of 125I-labeled β2-microglobulin (125I-β2m) into HLA-A2 H chain/β2m/peptide heterotrimeric complexes and analyzed by gel filtration. Stability of HLA-A2 H chain/β2m/peptide complexes was determined by the rate of dissociation (t1/2 in minutes) of 125I-β2m at 37°C, as earlier described (45). TAL peptides with measured t1/2 of >100 min were selected to stimulate CTL.
Establishment of TAL-specific TCL
PBMC were incubated in petri dishes precoated with autologous serum for 1 h at 37°C to remove monocytes (46). Nonadherent cells were enriched for CD8+ T cells by selective depletion of CD4+ cells using mAb-coated Dynabeads M-450 CD4, according to the manufacturer’s instructions (catalogue 111.08; Dynal Biotech). Usually, a CD4 T cell depletion of >99% was achieved with Dynabeads M-450 CD4. The resultant effector cells (107 cells per donor) were stimulated with 5 μg/ml TAL or TALpep. On day 8, and subsequently at regular 7- to 10-day intervals, TCL were restimulated with autologous APC derived from EBV-transformed B cells or PBMC. APC were generated by incubation with 5 μg/ml TAL or TALpep for 1 h, irradiated, and added to effector cells at a 1:1 ratio. In between restimulation with APC, 10 U/ml human rIL-2 was also added to the effector cells.
Generation of B cell lines by EBV infection
EBV was harvested from the supernatant of B95-8 cells (47), filtered through a 0.45-μm mesh, and stored at −80°C. A total of 107 PBMC was infected by EBV by incubation with 3 ml of B95-8 supernatant plus 2 ml of fresh complete RPMI 1640 medium at 37°C overnight. Next day, 4 ml of fresh complete RPMI 1640 medium and cyclosporin A (Sigma-Aldrich) was added to a final concentration of 0.5 μg/ml and cultures were left undisturbed for 10–14 days.
T cell-mediated cytotoxicity assays
To assess cytotoxic potential and the MHC class I restriction of the TAL peptide-reactive TCLs, HLA-A2-transfected (Hmy A2.1) and HLA-A3-transfected (Hmy A3.1) Hmy lymphoblastoma and HLA-A2-transfected (MO3.13/A2) and control MO3.13 oligodendroglial cells were used (48). Target cells were pulsed with 5 μg/ml TALpep for 2 h at 37°C. Then, 106 target cells were labeled with 200 μCi of Na251CrO4 for 1 h, washed three times, and seeded at a concentration of 5 × 103/well U-bottom 96-well plates. Depending on availability, effectors were added at 2.5:1, 5:1, 10:1, 25:1, and 50:1 E:T ratios. TALpep was added at 10 μg/ml during the cytotoxicity assay. Control cultures included peptide-pulsed and unpulsed target cells with peptide or medium alone. Maximal release was measured in the presence of 0.1 M HCl. After incubation for 4 h at 37°C, the supernatants were harvested and counted in a gamma counter. Percentage of cytotoxicity was calculated: 100 × ((test cpm − spontaneous cpm)/(maximal cpm − spontaneous cpm)).
Cytotoxic activity against adherent MO3.13 oligodendroglial cells (49) was measured by detachment of killed cells from the monolayer. A total of 2500 MO3.13 cells/well of flat-bottom 96-well plates was prelabeled with [3H]TdR (ICN Biomedicals) with or without 10 μg/ml TALpep and allowed to form nonconfluent monolayer for 24 h. Targets were washed three times, and effectors were added for 24 h. Subsequently, plates were washed six times to remove effectors and killed target cells, and cytotoxicity was determined based on [3H]TdR content of remaining viable cells, as earlier described (46).
PE-conjugated HLA-A*0201/LLFSFAQAV tetramer and control HIV peptide tetramers, HLA-A2/HIV pol 464 (HLA-A*0201/ILKEPVHGV), and HLA-A2/HIV gag 77 tetramers (HLA-A*0201/SLYNTVATL) were synthesized at the National Institutes of Health Tetramer Core Facility at Emory University. Freshly isolated PBMC from HLA-A2+ control donors, MS patients, and HLA-A2+ MS patients as well as TCL from MS patients were stained with PE-conjugated tetramers for 30 min at 4°C, washed three times, and resuspended in PBS before analysis on a BD Biosciences LSRII flow cytometer equipped with 20 mW argon (emission at 488 nm) and 16 mW helium-neon lasers (emission at 634 nm) (BD Biosciences). Dead cells and debris were excluded from the analysis by electronic gating of forward (FSC) and side scatter (SSC) measurements. Detection of staining with the tetramer (FL-2 fluorescence) was matched with emission spectra of FITC-conjugated CD8 (FL-1) and PE-Cy5-conjugated CD4 Abs (FL-3). Expression of TCR Vβ14 protein on the surface of TAL-specific TCL was assessed by staining with anti-human TCR Vβ14 mAb (mAb, clone CAS1.1.3; Serotec) and isotype control mouse IgG1 mAb ME.05 (Serotec clone ME.05), followed by detection with FITC-conjugated goat anti-mouse IgG.
IFN-γ ELISPOT assay
Millipore MultiScreen IP 0.45-μm Immobilon-P hydrophobic polyvinylidene difluoride membrane ELISPOT plates (catalogue MAIPN4510) were coated with 15 μg/ml anti-human IFN-γ capture mAb (Mabtech) overnight at 4°C, washed six times with 200 μl/well sterile PBS, and blocked with 100 μl/well 10% FBS in RPMI 1640 for 4 h at 37°C. After washing once with 10% FBS in RPMI 1640, 105 cells in 100 μl of RPMI 1640 with 10% FBS were incubated in each well in triplicates for 20 h at 37°C. Subsequently, the ELISPOT plates were washed six times with 200 μl of PBS per well. Next, 1 μg/ml biotinylated anti-IFN-γ detection mAb was added in 100 μl of PBS with 0.5% FBS for 20 h at 4°C. On day 4, plates were washed six times with 200 μl of PBS per well, and HRP-conjugated streptavidin was added in 100 μl of PBS with 0.5% FBS for 1 h at room temperature. Plates were washed six times with 200 μl of PBS per well and developed with 100 μl of 3-amino-9-ethyl-carbazole (AEC) substrate. AEC substrate included 20 mg of AEC dissolved in 2.5 ml of dimethylformamide and supplemented with 47.5 ml of acetate buffer (4.6 ml of 0.1 N acetic acid, 11.0 ml of 0.1 M sodium acetate, 46.9 ml of dH2O, and 25 μl of 30% H2O2) and was passed through a 0.4-μm filter to remove precipitates. After 4-min incubation with substrate, ELISPOT plates were washed six times with 200 μl of H2O per well. Spots were counted at 40-fold magnification using a dissection microscope.
Analysis of the TCR Vβ repertoire by RT-PCR
Total cellular RNA was isolated from cultured cells (Qiagen). A total of 3 μg of RNA was reverse transcribed into cDNA with oligo(dT) primer by 200 U of Superscript reverse transcriptase, according to the manufacturer’s protocol (Invitrogen Life Technologies). For analysis of TCR V gene usage, cDNA aliquots were subjected to PCR using 1 of 24 Vβ sense strand-specific oligonucleotides in combination with an antisense strand-specific Cβ primer (Cβ-REVF, 5-CGGGCTGCTCCTTGAGGGGCTGCG-3) (29). As an internal control, each PCR included Cα-specific sense (Cα-FWJ, 5-CCCTGACCCTGCCGTGTACCAGCT-3) and Cα antisense primers (Cα-REVE, 5-GTTGCTCCAGGCCGCGGCACTGTT-3) (29). Amplifications were conducted in 30 cycles at a denaturing temperature of 94°C (1 min), annealing at 60°C (1 min), and extension at 72°C (2 min). Amplified products were analyzed by electrophoresis in 1.5% agarose gel. Automated densitometry was used to quantify the relative levels of DNA band intensity using a Kodak Image Station 440CF with Kodak 1D Image Analysis software (Eastman Kodak).
Western blot analysis
A total of 500 ng of rTAL-H protein in 10 μl per well was separated by SDS-PAGE and electroblotted to nitrocellulose (50). Nitrocellulose blots were incubated in 100 mM Tris (pH 7.5), 0.9% NaCl, 0.1% Tween 20, and 5% skim milk with TAL Ab 169 (22). For detection of rabbit Abs, after washing, blots were incubated with HRP-conjugated goat anti-rabbit IgG (Boehringer Mannheim). In between the incubations, the strips were vigorously washed in 0.1% Tween 20, 100 mM Tris (pH 7.5), and 0.9% NaCl. The blots were developed with a substrate comprised of 1 mg/ml 4-chloronaphthol and 0.003% hydrogen peroxide.
TAL enzyme activity
TAL enzyme activity was measured by the transfer of the dihydroxyacetone three-carbon unit from the donor d-fructose-6-phosphate, to the acceptor d-erythrose-4-phosphate, as earlier described (22). Enzyme activity of 50 ng of rTAL was assayed in the presence of 3.2 mM d-fructose 6-phosphate, 0.2 mM d-erythrose-4-phosphate, 0.1 mM NADH, 10 μg of α-glycerophosphate dehydrogenase/triosephosphate isomerase at a 1:6 ratio in 1 ml of PBS (pH 7.4) at room temperature by continuous absorbance reading at 340 nm for 20 min.
Results were analyzed by Student’s t test or Mann-Whitney rank sum test for nonparametric data. Correlation was measured using Pearson’s correlation coefficient. Changes were considered significant at p < 0.05.
Detection of HLA-A2-restricted TALpep-specific CTLs in patients with MS
Demyelination in patients with MS is thought to be mediated by a selective destruction of oligodendrocytes (11). CD8+ CTL outnumber CD4+ T cells near MS lesions up to 10-fold (2). Cytosolic proteins, like highly soluble TAL, constitute a major source of peptides presented to CD8+ T cells by HLA class I molecules (51). To investigate the possible role of CD8+ T cell responses to TAL in the pathogenesis of MS, we used a computer-based algorithm that predicts the stability of HLA-Ag/peptide complexes by quantitating positive and negative effects on binding of each amino acid within a nonamer peptide (45). Incorporation of a typically 8- to 9-aa-long peptide is required for the stabilization of class I H chain and β2m complexes and their transport to the cell surface (51). Binding affinity to MHC class I molecules reliably predicts the capacity of a peptide epitope to elicit a CTL response (48, 49, 50, 51, 52). Therefore, the TAL amino acid sequence was analyzed for the presence of nonamer peptides with HLA class I-binding motifs (45). Ten peptides with high predicted HLA-A2-binding stabilities, i.e., t1/2 of greater than 100 min at 37°C, were identified in human TAL (Table II). Binding of TAL peptides to HLA-A2 was measured by incorporation of 125I-β2m into HLA-A2 H chain/β2m/peptide heterotrimeric complexes and analyzed by gel filtration. Stability of HLA-A2 H chain/β2m/peptide complexes was determined by the rate of dissociation (t1/2 in min) of 125I-β2m at 37°C, as earlier described (45). Among 14 peptides with predicted HLA-A2-binding stabilities of greater than 100 min (t1/2) at 37°C, TALpep has the highest binding affinity for HLA-A2. TAL peptides with measured t1/2 of >100 min (TALpep and TAL peptide 307–317) were tested for recognition by CTL in HLA-A2+ donors (Table II).
|Position .||HLA-A2 Binding Sequence .||Predicted t1/2 (min) .||Measured t1/2 (min) .|
|Position .||HLA-A2 Binding Sequence .||Predicted t1/2 (min) .||Measured t1/2 (min) .|
Binding stability was based on calculated half-life (predicted t1/2, in min at 37°C) of peptide-HLA-A2 complexes as earlier described (45 ). ∗, Values >100.
To test for the existence of HLA class I-restricted TAL-specific CTL activity, PBMC of MS and control donors were prestimulated with human rTAL for 1 wk. CTL from HLA-A2+ MS patients MS1, MS2, MS3, and MS4 showed specific killing of HLA-A2-transfected (HMY A2.1), but not HLA-A3-transfected control target cells (HMY A3.1) pulsed with TALpep (Fig. 1,A). Pretreament of effectors with rTAL was more efficient than synthetic peptides in inducing CTL activity against TALpep-pulsed targets (data not shown). Cytotoxic activity escalated with higher E:T ratios. By contrast, effector cells of three HLA-A2+ HC (HC1, HC2, HC3) and three HLA-A2− control donors (HC5; Fig. 1,A; HC6, data not shown) did not elicit killing of peptide-pulsed targets. Pulsing of targets with HLA-A2-binding TAL peptide 307–317 with an order of magnitude lower binding affinity (Table II) did not elicit killing.
MO3.13 oligodendroglial cells express high levels of endogenous TAL activity, ∼0.2 U/mg protein, as compared with −0.015 U/mg protein in PBMC (22). HLA-A2-transfected, but not control MO3.13 cells were efficiently killed by CD8 T cells prestimulated with TAL protein or TALpep (Fig. 1,B). CTL activity by HLA-A2+ (MS1, MS2, MS3, MS4, and MS5) and HLA-A2− MS patients (MS8 and MS9), and HLA-A2+ control donor (HC3) against MO3.13 oligodendroglioma target cells is shown in Fig. 1,B. In the absence of pulsing the target cells with peptides, killing by CTL of MO3.13/A2 with respect to control MO3.1 target cells was significantly increased in all HLA-A2+ MS patients (p < 0.0001). Killing by TAL or peptide-stimulated CTL of HLA-A2-transfected, but not control MO3.13 cells occurred without pulsing the targets with peptides. This indicated that endogenously processed TAL was recognized by HLA-A2-restricted CTL. Pulsing of the targets with TALpep further enhanced killing of MO3.13/A2 cells (p < 0.02; Fig. 1,B). Pretreatment of control MO3.13 target cells with TALpep did not elicit killing by HLA-A2+ MS CTL. Although PBL of MS patients MS1 and MS2 exhibited considerable cytotoxic activities, 24.8 ± 1.8% and 19.2 ± 2.0%, respectively, PBL from MS3 only exhibited 5.8 ± 1.0% cytotoxicity to HmyA2.1 cells (Fig. 1 A). HLA-A2-restricted TAL-specific CTL activity was enhanced to 23.6 ± 3.7% when using Dynabead M-450 affinity-purified CD8+ T cells of MS3, as effectors (data not shown). Overall, MO3.13 oligodendroglioma cells were more susceptible than Hmy lymphoblastoma cells for detection of TAL-specific HLA-A2-restricted cytotoxicity in each patient (p < 0.0001). HLA-A2− MS patients and HLA-A2+ control donors showed no significant CTL activity against TALpep-pulsed or unpulsed target cells. The results revealed the existence of TALpep-specific HLA-A2-restricted CTL in patients with MS.
Increased precursor frequency of TALpep-specific CD8+ T cells in peripheral blood of MS patients
Multimeric peptide-MHC complexes bind more than one TCR and thus have a relatively slow dissociation rate and allow staining and detection of epitope-specific T cells. Staining with PE-conjugated HLA-A*0201/LLFSFAQAV tetramers was measured by FL-2 fluorescence. Frequency of HLA-A2/TALpep/tetramer-binding cells was determined in seven HLA-A2+ MS patients, six HLA-A2− MS patients, three HLA-A2+ OND controls, three HLA-A2− OND controls, four HLA-A2+ HC, and three HLA-A2− HC. As shown in Fig. 2, increased numbers of cells staining with HLA-A*0201/LLFSFAQAV tetramer were detected in HLA-A2+ MS patients following rTAL stimulation (0.98 ± 0.07%; p < 0.01). Higher frequency of HLA-A*0201/LLFSFAQAV tetramer positivity was noted in TALpep-stimulated cells (1.4 ± 0.1%; p < 0.0001). In contrast, no significant increase in HLA-A*0201/TALpep/tetramer-PE staining was found in HLA-A2+ OND (0.34 ± 0.11%) and HC (0.36 ± 0.21%). HLA-A2/HIV pol 464 (HLA-A*0201/ILKEPVHGV) and HLA-A2/HIV gag 77 tetramers (HLA-A*0201/SLYNTVATL) did not show increased staining of TAL-stimulated PBMC from HLA-A2+ MS patients (Fig. 2).
Increased prevalence of TAL-reactive IFN-γ-producing T cells in MS patients
CTL produce IFN-γ that, in turn, enhances HLA class I expression on human oligodendrocytes (53). Therefore, the production of IFN-γ by TAL-reactive T cells of MS patients and controls was assessed. PBMC from seven HLA-A2+ MS patients, six HLA-A2− MS patients, three HLA-A2+ OND controls, three HLA-A2− OND controls, two HLA-A2+ HC, and one HLA-A2− HC were prestimulated for 1 wk with rTAL or TALpep, or PHA and the frequency of IFN-γ-producing T cells were determined by the ELISPOT assay seeding 105 cells/well (Fig. 3,A). There was a substantial increase of IFN-γ spots in all HLA-A2+ MS patients after both TAL (40.4 ± 2.7, p = 2.3 × 10−8) and TALpep stimulation (48.5 ± 6.2, p = 2.4 × 10−6) as compared with HLA-A2− MS patients or HLA-A2+ OND controls (Fig. 3,B). In HLA-A2− MS patients, stimulation by TAL resulted in enhanced spot production (7.0 ± 0.8) in comparison with stimulation by the TALpep (0.58 ± 0.28; p = 0.0001; Fig. 3 B). This indicated that TAL contains non-HLA-A2-restricted epitopes outside residues 168–176.
Generation of HLA-A2-restricted TALpep-specific CD8+ TCLs from MS patients
Freshly isolated PBMC from HLA-A2+ MS patients MS2 and MS7 were enriched for CD8+ T lymphocytes by depletion of CD4+ T cells with Ab-coated magnetic beads. TAL-specific CD8+ TCL were established by stimulation at regular 7- to 10-day intervals with rTAL- or TALpep-pulsed and irradiated APC derived from fresh autologous PBMC or EBV-transformed B cells. Functional capacity of TALpep tetramer-staining TCL was assessed by production of IFN-γ using the ELISPOT assay. Each of the TCL tested was maintained in vitro for >1 year by repetitive stimulation with TAL- or TALpep-pulsed autologous APC derived from PBMC (PBMC/APC) or EBV-transformed B cells (EBV/APC). As shown in Fig. 4, A and B, highest frequency of IFN-γ-producing cells was obtained when TCL were exposed to the same Ag used for repetitive stimulation, i.e., TAL-prestimulated cells produced more IFN-γ in the presence of TAL, while TALpep-prestimulated cells produced more IFN-γ in the presence of TALpep. MS2 TCL stimulated with TALpep PBMC/APC had higher frequencies of IFN-γ-producing cells than MS2 TCL stimulated with TAL PBMC/APC (p = 0.022; Fig. 4,B). Along the same line, TALpep was more effective than TAL in generating IFN-γ-producing T cells when using EBV APC in both patients MS2 (p = 0.029) and MS7 (p = 0.003; Fig. 4 B).
While maintaining these cell lines for >3 years, frequency of HLA-A2-restricted TALpep-specific T cells was repeatedly assessed by flow cytometry via staining with PE-conjugated HLA-A*0201/LLFSFAQAV tetramer (FL-2 channel), anti-FITC-conjugated CD8 (FL-1 channel), and PE-Cy5-conjugated CD4 mAb (FL-3 channel). As shown in Fig. 4,C, TAL PBMC/APC- and TALpep PBMC/APC-stimulated TCL from patient MS2 were double positive for HLA-A*0201/TALpep tetramer and CD8 at rates of 97.7 and 98.6%, respectively (Fig. 4,C). The same HLA-A*0201/TALpep tetramer-positive populations were CD4 negative. In contrast, HLA-A*0201/TALpep tetramer PE and CD8 FITC double-positive T cells comprised only 40.5 and 60.6% of TAL EBV/APC- and TALpep EBV/APC-stimulated TCL from patient MS2. Such TCL series were generated from patient MS2 on two occasions 1 year apart and gave similar results. The nontetramer-staining population in MS2 TCL stimulated with TALpep EBV/APC was shown to be all (99.1%) CD8+/CD4−. The HLA-A*0201/TALpep tetramer-negative population comprised two distinct populations, 40.1% CD4−/CD8+ and 14.5% CD4+/CD8− (Fig. 4 C).
IFN-γ production was tested on TCL cultured for 7 wk in culture (Fig. 4, A and B), while HLA-A*0201/TALpep tetramer staining was performed at a later stage, after 58 wk in culture (Fig. 4 C), indicating that repetitive stimulation with TAL led to a predominant recognition of TALpep.
TCL from HLA-A2+ MS patient MS7 were only stimulated with EBV-transformed B cell APC. TAL EBV/APC- and TALpep EBV/APC-stimulated MS7 TCL contained 32.4 and 43.5% TALpep tetramer and CD8 double-positive T cells, respectively (Fig. 4,C). Similar to the findings in patient MS2, the nontetramer-staining population in MS7 TALpep EBV/APC TCL was 99.4% CD8+/CD4−, while the MS7 TAL EBV/APC TCL contained 23.8% tetramer-negative CD4+/CD8− cells (Fig. 4 C). These findings indicated that stimulation with TAL-pulsed EBV/APC expanded tetramer-negative CD8+ and CD4+ T cells in both HLA-A2+ MS patients. None of the TAL- or TALpep-stimulated TCL of patients MS2 and MS7 showed significant staining with HLA-A2/HIV pol 464 and HLA-A2/HIV gag 77 control tetramers (<0.5%; data not shown).
TAL- and TALpep-specific TCL express the TCR Vβ14 gene
To determine clonal distribution of TAL- and TALpep-specific TCL, the repertoire of TCR Vβ gene usage was investigated by RT-PCR using 24 TCR Vβ primer sets (29). Twenty of 24 and 19 of 24 TCR Vβ genes were detected in unstimulated PBMC from MS patients MS2 and MS7, respectively (Fig. 5,A). TAL PBMC/APC- and TALpep PBMC/APC-stimulated MS2 TCL, both comprised of HLA-A*0201/TALpep tetramer/CD8 double-positive T cells, expressed the TCR Vβ14 gene exclusively (Fig. 5 A). Moreover, all TCL stimulated with TAL- or TALpep-pulsed EBV/APC expressed the TCR Vβ14 gene. Three of four EBV/APC-stimulated TCL also contained one additional TCR Vβ RT-PCR product. TCR Vβ19 was also detected in MS7 TCL stimulated with TAL- or TALpep-pulsed EBV/APC, which may represent T cells recognizing EBV Ags (54).
The TCR Vβ14 RT-PCR product from each TCL and PBMC of MS2 and MS7 was sequenced. Variations in the translated amino acid sequences were confined to residues 54–70, corresponding to the CDR3 of TCR Vβ14 (Fig. 5 B). Both MS2 TAL PBMC/APC TCL and MS2 TALpep PBMC/APC TCL had the identical TCR Vβ14 sequence through all regions, including CDR3. Although the TCR Vβ14 sequences of EBV/APC-fed TCL were not identical, they all shared the Jβ2.1 motif EQFF at aa positions 65–68, which corresponded to a similar Jβ2.7 motif EQYF in MS2 TCL stimulated by TAL- or TALpep-pulsed PBMC/APC.
Expression of TCR Vβ14 protein on the surface of TAL-specific TCL
Anti-human TCR Vβ14 mAb (Serotec clone CAS1.1.3) was used to test for the expression TCR Vβ14 on the surface of TAL-specific HLA-A2+ TCL (Fig. 6,A). IgG1 mAb ME.05 (Serotec clone ME.05) was used as isotype control. A total of 97.5 and 99.4% of TAL PBMC/APC- and the MS2 TALpep PBMC/APC-stimulated MS2 TCL expressed the TCR Vβ14 Ag, respectively. A total of 51.9 and 57.3% of TAL EBV/APC- and TALpep EBV/APC-fed MS2 TCL expressed TCR Vβ14, respectively. MS7 TAL EBV/APC TCL and MS7 TALpep EBV/APC TCL also exhibited significant TCR Vβ14+ populations, 29.2 and 38.5%, respectively. Small, but significant TCR Vβ14 staining was seen in PBL of HLA-A2+ MS patient MS2, as compared with HLA-A2+ HC PBL. No positive staining was noted with the ME.05 IgG1 isotype control Ab (Fig. 6,A). Double staining of TAL EBV/APC- and TALpep EBV/APC-fed MS2 and MS7 TCL revealed exclusive binding of HLA-A*0201/TALpep tetramer to cells expressing TCR Vβ14 (Fig. 6 B).
TCR Vβ14 mAb inhibits killing of oligodendroglioma and lymphoblastoma target cells by TAL-specific HLA-A2-restricted CTL
To assess the involvement of TCR Vβ14 in killing by TAL-specific CTL, effector cells were preincubated with or without mouse anti-human TCR Vβ14 mAb or mouse isotype control IgG1 mAb, and subsequently added to MO3.13 oligodendroglioma or Hmy lymphoblastoma target cells at 10:1 E:T ratio. The TCR Vβ14 mAb dramatically reduced the cytotoxicity by all TAL- and TALpep-specific TCL against both types of target cells (Fig. 7). MS2 TAL PBMC/APC TCL and MS2 TAL pep PBMC/APC TCL exhibited the highest cytotoxicity against TALpep-pulsed Hmy/A2 targets (91 and 97%, respectively) and TALpep-pulsed MO3.13/A2 targets (90 and 93%, respectively; Fig. 7). TCR Vβ14 mAb reduced cytotoxicity against TALpep-pulsed Hmy/A2 and MO3.13/A2 targets by an average of 59 and 72%, respectively (p < 0.0001). Additionally, TCR Vβ14 mAb inhibited the cytotoxicity against the nonpeptide-pulsed MO3.13/A2 target cells. A total of 66% killing by MS2 TAL PBMC/APC TCL and 75% killing by MS2 TALpep PBMC/APC TCL of MO3.13/A2 targets was reduced by 70% after addition of anti-human TCR Vβ14 mAb (p < 0.0001). Cytotoxic activities of EBV/APC-fed TCL were similarly reduced by TCR Vβ14 mAb. IgG1 isotype control mAb did not influence the cytotoxicity of TCL against either types of target cells (Fig. 7).
Previous studies from this laboratory, pursuing the goal to isolate autoantigens containing epitopes cross-reactive with viral proteins, demonstrated that an autoantigen, partially encoded by a retrotransposon and selectively expressed in oligodendrocytes at high levels (21), corresponds to TAL (22), a rate-limiting enzyme of the pentose phosphate pathway (PPP). The purified full-length rTAL was found to be functional in the TAL enzyme assay by showing a sp. act. of >16 U/mg protein (22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36). As references, normal human lymphocytes contain a TAL activity of 0.015 U/mg protein (22). By contrast, TAL activity in human oligodendroglioma cell lines M03.13 (49) and HOG (55) was >0.2 U/mg protein, indicating that high expression of TAL may be retained in oligodendrocytes undergoing malignant transformation.
In this study, we document the existence and increased prevalence of TAL-specific HLA-A2-restricted CD8+ CTL in patients with MS (Table I). In MS brain lesions, infiltrating CD8+ CTL outnumber CD4+ T cells (2); actively demyelinating lesions in MS brain are enriched for clonally expanded CD8+ T cells, as compared with CD4+ T cells (40, 41); and oligodendrocyte cell death is mostly attributed to the cytotoxic effect of CD8+ T cells (37). Adoptively transferred MBP-specific (38) or MOG-specific CD8+ T cells induce severe CNS demyelination in animal models (39). Although MBP-specific CD8 T cells were detected in human PBL, there was no difference in cytotoxicity of MBP 110–118 peptide-stimulated CD8 T cells between healthy donors and MS patients (56, 57). Recently, increased precursor frequency of MBP 111–119-reactive CD8 T cells was found in a minority of MS patients (58). In contrast, the present data show that the precursor frequency, cytotoxic activity, and IFN-γ production of TALpep-specific CD8 T cells are increased in each of seven HLA-A2+ MS patients as compared with seven HLA-A2− MS patients, four HLA-A2+ OND patients, four HLA-A2− OND patients, and four HLA-A2+ and two HLA-A2− healthy donors.
Patients with MS exhibit TAL-specific T cell proliferation and have Abs to TAL in their blood and cerebrospinal fluid, whereas TAL Abs are absent in normal individuals and patients with other autoimmune and neurological diseases (21, 29). Meanwhile, under similar conditions, Abs to MBP were not found in serum and cerebrospinal fluid of MS patients (29). These results suggested that TAL may be a more significant target than MBP of myelin-reactive T cells and of humoral autoreactivity in MS. Demyelination in patients with MS is thought to be mediated by a selective destruction of oligodendrocytes by cytotoxic effect of CD8+ T cells (2, 37, 41, 59). Cytosolic proteins, like highly soluble TAL, constitute a major source of peptides presented to CD8+ T cells by HLA class I molecules (51). Among 14 peptides with predicted HLA-A2-binding stabilities of greater than 100 min (t1/2) at 37°C, TALpep has the highest binding affinity for HLA-A2 and it is specifically recognized by CD8+ CTL in HLA-A2+ MS patients.
HLA-A2-transfected, but not control MO3.13 oligodendroglial cells, expressing high levels of endogenous TAL with respect to Hmy cells, were efficiently killed by TAL-specific CD8 T cells without pulsing the target cells with TALpep. This indicated that endogenously processed TAL was recognized by HLA-A2-restricted CTL. Pretreatment of the targets with TALpep further enhanced killing of MO3.13/A2, but not control MO3.13 cells. TAL-specific CTL activity was observed in each of seven HLA-A2+ MS patients, as compared with the mean of three HLA-A2+ OND patients, three HLA-A2+ healthy donors, or six HLA-A2− MS patients. The results clearly showed the presence of TALpep-specific HLA-A2-restricted CTL in patients with MS. Furthermore, staining with HLA-A*0201/TALpep (LLFSFAQAV) tetramer revealed a higher frequency of TAL-specific CTL in TAL- or TALpep-stimulated PBL of seven HLA-A2+ MS patients in comparison with HLA-A2+ OND and HC. The frequency of IFN-γ-producing T cells was also increased in all seven HLA-A2+ MS patients following stimulation with TAL or TALpep. In HLA-A2− MS patients, the frequency of IFN-γ-producing T cells was moderately increased following stimulation by TAL (7.0 ± 0.8/105 cells) in comparison with stimulation by the TALpep (0.58 ± 0.28/105 cells; p = 0.0001), suggesting that TAL contains non-HLA-A2-restricted epitopes outside residues 168–176. Repetitive stimulation of MS2 TCL with TAL or TALpep PBMC-derived APC elicited a single clonal population of TAL-specific CTL double positive for HLA-A*0201/TALpep tetramer PE and CD8 FITC. TAL-specific TCL exhibited markedly enhanced cytotoxic activity (Fig. 6) and robust IFN-γ production (Fig. 4). IFN-γ stimulates HLA class I expression on human oligodendrocytes (53); thus, increased IFN-γ production by TAL-specific T cells is likely to increase susceptibility of oligodendrocytes to killing by CTL.
Stimulation of CD8 effector cells with TAL elicited a clonal population of HLA-A*0201/TALpep tetramer-positive cells. All TAL and TALpep PBMC/APC-stimulated CTL lines used a unique TCR Vβ14 gene, thus suggesting that TALpep appears to be the dominant HLA-A2-restricted T cell epitope in human TAL. Along this line, all TAL or TALpep EBV/APC-stimulated TCL from patients MS2 and MS7 contained a dominant HLA-A*0201/TALpep tetramer/CD8+ population that expressed TCR Vβ14. In accordance with the presence of HLA-A*0201/TALpep tetramer-negative cells, EBV/APC-fed TCL were oligoclonal, using TCR Vβ9, Vβ7, Vβ4, and Vβ19. Of note, all of these Vβ were also expressed in CD8+ TCL obtained from EBV-seropositive HLA-A2+ healthy donors stimulated with irradiated autologous EBV-transformed B cells (54). Thus, HLA-A*0201/TALpep tetramer−/CD8+ cells in EBV/APC-fed TCL may respond to EBV Ags. TCL fed by EBV/APC pulsed with TAL, but not TALpep, also comprised CD4+ T cells, 14.5% in MS2 and 23.8% in MS7, respectively, suggesting that TAL also harbored HLA class II-restricted epitopes.
The TCR Vβ14 RT-PCR products of each TCL as well as PBMC of Mβ2 and Mβ7 were sequenced. Variations in the translated amino acid sequences were confined to residues 54–70, harboring the CDR3 region of TCR Vβ14 (Fig. 5 B). Both MS2 TAL PBMC/APC TCL and MS2 TALpep PBMC/APC TCL had the identical TCR Vβ14 sequence through all regions, including CDR3. Although the TCR Vβ14 sequences of EBV/APC-fed TCL were not identical, they all shared the EQFF motif at aa positions 65–68. The EQFF motif was preceded by neutral and polar amino acids GTSGY and LSGGY in MS2 TAL EBV/APC TCL and MS2 TALpep EBV/APC TCL, respectively. By contrast, the EQFF motif was preceded by highly charged RPRE and GRRD residues in MS7 TAL EBV/APC TCL and MS7 TALpep EBV/APC TCL, respectively. The conserved EQFF motif corresponded to the functionally similar EQYF motif in MS2 TCL stimulated by TAL- or TALpep-pulsed PBMC/APC. The EQFF and EQYF motifs are derived from the Jβ2.1 and Jβ2.7 regions, respectively. The findings indicate that the EQF/YF motif may contibute to recognition of TALpep by TCR Vβ14.
Use of the TCR Vβ14 gene by TAL-specific TCL was confirmed by demonstrating expression of TCR Vβ14 Ag on the surface of TAL-specific TCL. Flow cytometry analysis revealed exclusive binding of HLA-A*0201/TALpep tetramer to cells expressing TCR Vβ14 in all TAL-specific TCL, suggesting that TCR Vβ14 may be uniquely responsible for recognition of TAL in HLA-A2+ MS patients. Pretreatment of CTL lines with TCR Vβ14 mAb profoundly inhibited cytotoxicity against HLA-A2-transfected Hmy A2.1 or MO3.13 target cells, establishing the functional relevance of the TCR Vβ14/TALpep/HLA-A2 trimolecular interaction. Therefore, selective elimination of TAL-specific CTL may prevent oligodendrocyte destruction and serve as a novel target for therapeutic intervention in patients with MS.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by Grants RO1 DK 49221 from the National Institutes of Health and RG 2466 from the National Multiple Sclerosis Society.
Abbreviations used in this paper: MS, multiple sclerosis; AEC, 3-amino-9-ethyl-carbazole; β2m, β2-microglobulin; EAE, experimental allergic encephalomyelitis; FSC, forward scatter; HC, healthy control; 125I-β2m, 125I-labeled β2m; MBP, myelin basic protein; MOG, myelin oligodendrocyte protein; MS, multiple sclerosis; OND, other neurological disease; PPP, pentose phosphate pathway; SSC, side scatter; TAL, transaldolase; TALpep, TAL peptide 168–176; TCL, T cell line.