Molecular tracking of antigen-specific T cell clones in neurological immune-mediated disorders

Abstract

T cells recognizing self or microbial antigens may trigger or reactivate immune-mediated diseases. Monitoring the frequency of specific T cell clonotypes to assess a possible link with the course of disease has been a difficult task with currently available technology. Our goal was to track individual candidate pathogenic T cell clones, selected on the basis of previous extensive studies from patients with immune-mediated disorders of the CNS, including multiple sclerosis, HTLV-I associated myelopathy/tropical spastic paraparesis (HAM/TSP) and chronic Lyme neuroborreliosis. We developed and applied a highly specific and sensitive technique to track single CD4(+) and CD8(+) T cell clones through the detection and quantification of T cell receptor (TCR) alpha or beta chain complementarity-determining region 3 transcripts by real-time reverse transcriptase (RT)-PCR. We examined the frequency of the candidate pathogenic T cell clones in the peripheral blood and CSF during the course of neurological disease. Using this approach, we detected variations of clonal frequencies that appeared to be related to clinical course, significant enrichment in the CSF, or both. By integrating clonotype tracking with direct visualization of antigen-specific staining, we showed that a single T cell clone contributed substantially to the overall recognition of the viral peptide/MHC complex in a patient with HAM/TSP. T cell clonotype tracking is a powerful new technology enabling further elucidation of the dynamics of expansion of autoreactive or pathogen-specific T cells that mediate pathological or protective immune responses in neurological disorders.

Fig. 1

Annealing sites of clonotype-specific oligonucleotide primers and probes. The nucleotide and deduced amino acid sequences of TCR β (T cell clones, P2–10, 1C2, CSF3) or α chain (T cell clone A6) CDR3 are shown. The annealing sites of oligonucleotide primers and probes for TaqMan quantification of clonotypic TCR transcripts from each T cell clone are indicated by boxes (single line, forward primer; dashed line, internal probe; double line, reverse primer).

Fig. 2

High sensitivity of TCCT. ‘Spiking’ experiments assessed the sensitivity of TCCT for the detection of clonal cells at low frequency. (A) Conventional PCR and quantification of clonotypic TCR transcripts of T cell clone 1C2 using P₃₃-radiolabelled dCTP allowed the detection of 10 target cells/10⁶ irrelevant cells. (B) As compared with conventional PCR, TCCT by TaqMan real-time quantitative PCR yielded a 10-fold increase in sensitivity, resulting in positive amplification of a 1 : 20 dilution of cDNA derived from of a mixture of 1 clonal cell/10⁶ control cells. (C) The outstanding efficiency of amplification of CSF3 TCR transcripts allowed detection of 1 clonal cell in 5 × 10⁶ control cells. High specificity of amplification was shown by the absence of any amplification of control cDNAs after up to 40 PCR cycles (Table 2).

Fig. 3

Detection of in vitro induced clonal expansion. Selective expansion of clone 1C2 in the unmanipulated ‘parent’ PBMC from donor HD4 was detected by TCCT as a result of stimulation with the specific antigen MBP_111–129 or with the superantigen SEA, which were known from previous experiments to stimulate proliferation of the isolated clone. There was no increase of 1C2 TCR transcripts upon PBMC stimulation with control stimuli FIU-HA_307-319 and SEE. Error bars indicate standard deviation.

Fig. 4

Expansion of MBP_83–99-specific T cell clone P2-10 before a multiple sclerosis exacerbation. (A) Representative post-contrast T₁-weighted brain MRI scans of patient MS502 at month 1 and at month 1.25 (exacerbation) of follow-up. (B) Cellular immune reactivity of PBMC from patient MS502 in response to the prevailing epitopes (MBP_83–99, black fills; and PLP_190–209, grey fills) being recognized at two different time-points (T1, left pair; T5, right pair), expressed as sum of stimulation indexes (SI) of all proliferating (SI > 2) wells in an IL-7-modified proliferation assay. P values were calculated using the Mann-Whitney rank sum test. A significant expansion of MBP_83–99-specific T cells was observed at month 1 of treatment with APL. In contrast, at month 5 the response to PLP was increased. The lower variance of SI for PLP_190–209 accounts for the higher significance. (C) Molecular quantification of P2–10 clonal frequency in PBMC (circles) over time (expressed in months; x-axis) is shown along with disease activity assessed by MRI (triangles). SD of P210 mRNA levels were: 1.5 (TO), 2.5 (Tl), 1.2 (T1.25), 1.4 (T2) and 0.7 (T5). A 5-fold increase of P2–10 frequency (filled circles) was observed 1 week (Tl) before the onset of a clinical exacerbation associated with the appearance of a large number of new gadolinium-enhancing lesions at MRI of the brain (open triangles). No increase of clone P2–10 was found at the time of a second relapse (T5).

Fig. 5

Persistent expansion of T cell clone A6 and relationship with CD8⁺ T cell recognition of HTLV-I Tax_11–19/A2 complex in a patient with HAM/TSP. (A) Stable frequency of A6 in the peripheral blood over several years of chronic progression of HAM/TSP. (B) Increased frequency of A6 in sorted CD8⁺ cells and in unmanipulated CSF cells (12-fold). (C) Staining of PBMC from a healthy control (left) and patient HAM3 (right) with A2 Ig loaded with HTLV-I Tax_11–19 (top) or with a control peptide, HIV gag_77–85 (bottom). Percentages in upper right quadrants indicate the proportion of Tax/A2 Ig positive CD8⁺ cells in the total CD8⁺ cells. (D) By staining with Tax/A2 Ig and TCCT, we determined that clone A6 alone represents 14.2% (or, one of seven) of all Tax_11–19-specific, A2-restricted CD8⁺ T cells in the peripheral blood. This figure can be viewed in colour as Supplementary material at Brain Online.

Fig. 6

Frequency of T cell clone CSF3 in PBMC and CSF cells from a patient with Lyme neuroborreliosis. CSF3 TCR transcripts were undetectable in PBMC samples collected at 6 different time points. Lower limit of detection was 1 cell/5 × 10⁶ (see Fig. 2C). In contrast, CSF3 was repeatedly found in CSF, with the highest frequency at the time of a severe exacerbation of neurological symptoms, an intermediate frequency during a milder relapse, and the lowest frequency at the time of clinical remission.