Characterization of CD4+ T cells specific for glutamic acid decarboxylase (GAD65) and proinsulin in a patient with stiff-person syndrome but without type 1 diabetes

The patient

The patient is a 33-year-old (as of 2010) Caucasian female, previously healthy other than having partial complex epilepsy. In March 2006, she was hospitalized because of lower back pain, walking difficulty and severe abdominal and lower lumbar muscle spasms. Serum creatinine kinase levels increased 10-fold during the first 5 days in the hospital. The clinical picture raised suspicion of SPS, and very high levels of GADA were serially measured from sera (Table 1). GADA was also measured from a CSF sample drawn concomitantly with the second serum sample, and the GADA titer was more than four times higher than in serum. Clonazepam was initiated and the patient was able to walk unassisted upon release from the hospital. Brain and cervical, thoracic and lumbar spinal cord MRIs were normal. CSF leukocyte count and CSF protein concentration and IgG index were normal, but total IgG concentration in CFS was elevated to 59 mg/L (normal value < 45 mg/L). Protein electrophoresis of CFS showed three oligoclonal gamma bands, indicating immunoglobulin production in the CNS, or a compromised blood–brain barrier. No viral, Mycoplasma or Borrelia burgdorferi-antibodies were present in the CSF. IgA deficiency was diagnosed. ENMG examination was normal, and serum autoantibody titers against tissue transglutaminase and gliadin, as well as parietal cell-, nuclear, acetyl choline, insulin and IA-2 autoantibodies were all negative. Anti-thyroid peroxidase antibody titre was intermediately positive 120 IU/mL (normal < 10 IU/mL). Islet cell antibodies were of a very high titer (3033 JDFU, normal value < 2.5 JDFU) which was likely due to the high GADA titer. Blood glucose levels, monitored every 3–6 months for 42 months and oral glucose tolerance test (performed twice) were normal. The clinical symptoms of SPS worsened in September 2006. Intravenous immunoglobulin treatment was contraindicated because of IgA deficiency, and she received oral prednisolone and tizanidine therapy for 3 months in addition to clonazepam. Thereafter, her SPS symptoms have been relatively stable until today (January 2010). The patient was HLA typed for class II alleles 31 and she carried the following genotype: HLA-(DR3)-DQA1*05-DQB1*02/HLA-(DR7)-DQA1*0201-DQB1*0303.

Peripheral blood samples for this study were obtained from the patient after informed consent, and when possible they were drawn simultaneously to blood samples collected for diagnostic purposes. The use of the blood samples for immunological studies was approved by the Ethical Committee at the Hospital District of Southwestern Finland (20.5.2003, 134).

GAD65 enzyme activity assay

GAD65 enzyme activity was measured by the ¹⁴CO₂-trapping method described previously 32. Recombinant human GAD65 (a kind donation by Amgen, Seattle, WA, USA) (100 ng) was incubated with the reaction buffer (50 mM K₂HPO₄, 0.03 mM PLP, 0.1 mM DTT, pH 6.8) for 1 h at room temperature, with or without the indicated amounts of isolated IgG. The enzymatic reaction was initiated by the addition of 0.56 mM L-glutamate and 0.018 µCi ¹⁴C-glutamate (Perkin–Elmer, Boston, MA, USA) and allowed to continue for 2 h at 37 °C with gentle agitation. During incubation, released ¹⁴CO₂ was captured on filter paper (Kontes, Vineland, NJ, USA) soaked in 50 µL 1 M NaOH. After the incubation, the absorbed radioactivity was determined in a Beckman scintillation counter. The results are presented as: [1-(residual activity = cpm in the presence of IgG/cpm in the absence of IgG)] × 100.

GADA assay

GADA were quantified with a radioligand assay in two different laboratories as previously described 33, 34. These two laboratories were the University of Washington laboratory (C.H.; University of Washington, WA, USA) and the University of Helsinki laboratory (M.K.; University of Helsinki, Finland). GADA titers (index) presented in Figure 1 were derived using the Washington laboratory assay, while GADA titers presented in Table 1 summarize results obtained in the Helsinki laboratory.

In the Washington assay 33, the cut-off limit for autoantibody positivity was 0.05 (GAD65Ab index). The sensitivity and specificity for T1D in this assay were 86% and 93%, respectively, in the 2007 Diabetes Autoantibody Standardization Program (DASP) Workshop. In the Helsinki assay 34, the cut-off limit for autoantibody positivity was 5.36 relative units (RU), representing the 99th percentile in 373 non-diabetic subjects. The sensitivity and specificity for T1D were 78% and 95%, respectively, in the 2009 DASP Workshop.

IA-2-Ab and IAA assay

IA-2 antibodies were analysed with a radioligand assay as described 35. The cut-off limit for autoantibody positivity was 0.77 RU, representing the 99th percentile in 354 non-diabetic subjects. The sensitivity and specificity for T1D were 64% and 99%, respectively, in the 2009 DASP workshop.

IAA were measured with a radiobinding microassay as described 36. The cut-off limit for IAA positivity was 2.80 RU, representing the 99th percentile in 354 non-diabetic subjects. The sensitivity and specificity for T1D were 42% and 99%, respectively, in the 2009 DASP workshop.

Recombinant Fab fragments

Monoclonal antibodies b96.11 and b78 were derived from a patient with Autoimmune Polyendocrine Syndrome—type 2 37, and recognize conformational epitopes formed by the 3D structure of amino acid residues 308–365 and 451–585, respectively 38. N-GAD65mAb was raised in mice and recognized linear epitopes representing amino acid residues 4–22 32. DPD was isolated from a patient with T1D 39 and recognized epitopes that mapped between amino acid residues 483–585 and 96–173, respectively. MICA-3 was isolated from a patient with T1D 40 and recognized epitopes of amino acid region 451–585 41. All mAbs recognized GAD65 in its native conformation and do not bind GAD67.

Competition studies using rFab

The capacity of the rFab to inhibit GAD65 binding by human serum GAD65Ab was tested in a competitive radioligand-binding assay using Protein A Sepharose (Invitrogen) as described 39. The rFab were added at the optimal concentration (0.7–1 µg/mL) as determined in competition assays using the intact mAb as a competitor. The background competition for each rFab was established in competition experiments with normal control sera. The background was subtracted prior to calculation of percent inhibition. The cut-off for specific competition was determined as > 15% by using as a negative control rFab D1.3 (a kind gift from Dr J. Foote, Arrowsmith Technologies, Seattle, WA, USA), specific to an irrelevant target, hen-egg lysozyme, at 5 µg/mL. Binding of GAD65Ab to GAD65 in the presence of rFab was expressed as follows: cpm in the presence of rFab/cpm in the absence of rFab × 100.

Peptides

Peptides used in the assay were GAD65 339–352 (TVYGAFDPLLAVAD), GAD65 247–266 (NMYAMMIARFKMFPEVKE) and ProIns B24–C36 (FFYTPMSRREVED).

Generation of tetramers

Generation of the MHC class II tetramers has been previously described 42. Briefly, a site-specific biotinylation sequence was added to the 3′ end of the DRB1*0301 leucine zipper cassette, and the chimeric cDNA was subcloned into a Cu-inducible Drosophila expression vector. DR-A and DR-B expression vectors were co-transfected into Schneider S-2 cells, the class II monomers then purified, concentrated and biotinylated. The desired peptide was loaded for 48–72 h, and tetramers were formed by incubating class II molecules with PE-labelled streptavidin.

Primary culture and stimulation

Peripheral mononuclear cells (PBMCs) were isolated from heparinized blood samples by Ficoll-Hypaque gradient separation (Lymphoprep, Nycomed, Oslo, Norway) and frozen. The frozen PBMC were shipped to Benaroya Research Institute in Seattle, USA on dry ice. PBMCs were thawed and cultured (2.0 × 10⁶ cells/well) in RPMI media with HEPES (25 mM), 15% pooled human serum, 1% L-glutamine, 1% Penicillin/streptomycin and 1% sodium pyruvate. 10 µg/mL of peptide was added to the culture and incubated for 14 days at 37 °C. Human recombinant IL-2 (10 U/mL; Chiron, Emeryville, CA, USA) was added on days 6 and 10.

Tetramer staining of PBMC

PBMCs were stained with a peptide loaded, phycoethrin-labelled DRB1*0301 tetramer for 1 h at 37 °C and subsequently with fluorochrome-labelled CD4 (Ebioscience, San Diego, CA, USA) antibody for 20 min on ice. The cells were then washed with PBS containing 1% BSA and 0.01% sodium azide and analysed using a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA, USA). The data were analysed using FlowJo Software (Tree Star Inc., Ashland, OR, USA).

Isolation of peptide-specific T-cell clones

The peptide stimulated cells were single cells sorted into 96-well plates using a FACSVantage cell sorter (Becton Dickinson). Sorted clones were expanded for 10 days by stimulation with irradiated unmatched PBMC (1.5 × 10⁵/well), 5 µg/mL PHA (Sigma, St Louis, MO, USA) and 10 U/mL IL-2 (Chiron) for two cycles, followed by stimulation with irradiated HLA-DR301 matched PBMC pulsed with 10 µg/mL of the peptide and 10 U/mL IL-2. On days 10–12, clones were selected based on the growth for further expansion. 5 × 10⁴ resting T cells were tested for the specificity by stimulation with irradiated HLA-DR301 matched PBMC (1 × 10⁵/well), with and without the specific peptide in the culture. Proliferation as measured by ³H-thymidine incorporation was tested at 72 h of culture. Supernatants were collected from culture supernatants at 48 h and cytokines were measured by using MSD's electrochemiluminescence MesoScale assay (Mesoscale Discovery, Gaithersburg, MD, USA) according to manufacturer's instructions. The assay was performed using a human Th1/Th2 cytokine kit to measure interferon (IFN)-γ, IL-10, IL12p70, TNF-α, IL-5, IL-4, IL-13, IL-2 and IL-6. T-cell clones were tested for tetramer binding by staining with 10 µg/mL specific or negative control tetramer for 1 h at 37 °C followed by fluorochrome conjugated CD4 Ab on ice for 20 min and analyed by flow cytometry.

TCR sequencing of single cell clones

TCR alpha and beta chain CDR3 sequences of the CD4+ T-cell clones were determined by RT-PCR and sequencing analysis. Briefly, total RNA was isolated from CD4+ T-cell clones using an RNeasy Mini Kit (Qiagen) and cDNA was synthesized with a Taqman Reverse Transcription Kit (Applied Biosystems). A set of five multiplex PCR reactions covering a majority of the human V-beta repertoire were performed as per Akatsuka et al. 43 and two series of amplifications, first, with pooled V-alpha primers and then, with specific V-alpha primers covering the CDR3 regions, were performed as per Seitz et al. 44. PCR products were visualized on an ethidium bromide stained 2% agarose gel, sequenced using a Big Dye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems) with either a TCR constant region 3′ primer or a specific variable region 5′ primer, and then run on an ABI3100 Genetic Analyzer. TCR alpha and beta CDR3 sequence data were analysed using the IMGT/V-QUEST (http://imgt.cines.fr) web-based program from the Université Montpellier, France 34.

Inhibition of GAD65 enzymatic activity by SPS sera

The manifestation of SPS symptoms was associated with high levels of GADA in both serum and CSF (up to 1.1 × 10⁶ RU in serum samples, Table 1). We tested the ability of the patient's sera to inhibit GAD65 enzyme activity by incubating recombinant GAD65 with longitudinally collected serum samples and subsequently measuring GAD65 enzyme activity (Figure 1). The enzymatic activity of GAD65 was inhibited significantly at all time points by at least 57%. A slight, but insignificant decrease of inhibition (65–57%) was observed between months 6 and 10.

Epitope mapping of GADA

We analysed the GADA epitope specificities present, using rFab derived from GAD65-specific monoclonal antibodies (Figure 2). In our epitope-specific RBA, we determined the binding of the serum sample to radiolabelled GAD65 in the presence of GAD65-specific monoclonal rFab. Competition of serum GAD65Ab with the respective rFab for a GAD65 epitope is reflected in reduced binding of the serum to radiolabelled GAD65. Significant competition (binding of serum sample to radiolabelled GAD65 is reduced by at least 15%) was observed only for rFab b96.11 and rFab b78. The longitudinal analysis showed that the competition with rFab b96.11 was stable throughout the analysis, while competition with rFab b78 increased from 3% at baseline to 20% in the final sample.

T-cell reactivity with GAD and proinsulin tetramers

PBMC from four serial blood draws spanning over 40 months were stimulated with two GAD65 peptides (247–266 and 339–352) and a proinsulin peptide B24–C36, and stained with DR301 tetramer containing the same peptide as used in the stimulation. At the first time point, the CD4+ T cells bound to all specific tetramers at frequencies from 2.09 to 6.56% (Figure 3) which was 10–32.8-fold higher than the background staining with the negative control tetramer (∼0.2%). At the subsequent time points, DR301-GAD339–352 tetramer-positive T cells were also identified but at much lower frequencies ranging from 1.06 to 2.00% whereas the T-cell response to proinsulin peptide B24–C36 was observed to be modestly positive (0.88%) and then disappeared. The response to the GAD339–352 peptide seemed to be the most persistent one as tetramer-binding cells were detected in all samples. All the samples were also stimulated with the DR301 binding NS1 peptide derived from flu virus as positive control and showed a strong tetramer binding (data not shown).

Avidity and cytokine profiles of the T-cell clones

Tetramer-positive cells from the cultured CD4+ lymphocytes were single cell sorted and T-cell clones were generated. T-cell clones specific for the GAD65 339–352 peptide were isolated from the first three samples whereas the attempt to raise clones from the fourth time point was not successful. T-cell clones specific for GAD247–266 and proinsulin B24–C36 peptides were isolated only from the first blood sample. Tetramer binding when used at non-saturating concentrations has been shown to correlate with the structural avidity of TCR 45, 46. In Figure 4, tetramer staining of the GAD65 339–352-specific T-cell clones shows that high-avidity T-cell clones with a strong tetramer-binding profile (77–100%) were detected at the first and third time point. However, some T-cell clones, even derived from the same sample, bound the tetramer at much lower level (34%, clone 339–1, sample #3) or, displayed lack of staining (clone 339–201, sample #2). All clones were capable of proliferating when stimulated with the specific peptide (data not shown) and produced cytokines (Figure 4) demonstrating specificity to the given T-cell epitope peptide. When the high-avidity T-cell clones derived from the first blood sample were stimulated with the GAD65 339–352 peptide robust IFN-γ and IL-13 production was observed and two of the clones (#309 and 332) also produced high levels of IL-5 and low amounts of IL-4. Three other clones isolated from the same sample displayed no tetramer binding but produced low levels of IFN-γ, IL-13 and IL-4 (50–80 pg/mL) demonstrating specificity but a low-avidity phenotype (data not shown). GAD339–352-specific T-cell clones derived from samples drawn at the later time points seemed to be biased more towards predominant IL-13 production. T cells specific for the other peptide epitopes, GAD65 247–266 and proinsulin B24–C36, also produced IL-13, and displayed low/intermediate avidity. T-cell clones specific for NS1 flu peptide were isolated from all time points (data not shown). These T cells displayed a strong tetramer binding (∼100%) and significant IFN-γ production upon stimulation with the specific peptide.

Overall, GAD65 339–352 is likely to be a predominant peptide epitope presented by DRB1*0301 in our patient as T-cell reactivity to this peptide was demonstrated in all four samples studied and, furthermore, T-cell clones with the given specificity, some with a very strong tetramer binding, were derived from three out of four samples. The T-cell clones generated at the first sampling produced high levels of IFN-γ but also IL-13 and IL-5, suggesting a Th1/Th0 profile which shifted more towards Th2 during the follow-up. T-cell responses to GAD247–266 and proinsulin B24–C36 peptides seemed to be more subdominant in nature. In addition, T-cell clones with these specificities displayed low avidity and Th2-biased cytokine profiles and were generated only from the first time point. Sequence analysis of the complementarity determining regions demonstrated that the T-cell clones used distinct TcR-Vβ chains and that the ones derived from different time points were not identical (data not shown).