A minority of proliferating human CD4+ T cells in antigen-driven proliferation assays are antigen specific

Abstract

Antigen-driven T-cell proliferation is often measured using fluorescent dye dilution assays, such as the CFSE-based proliferation assay. Dye dilution assays have been powerful tools to detect human CD4⁺ T-cell responses, particularly against autoantigens. However, it is not known how many cells within the proliferating population are specific for the stimulating antigen. Here we determined the frequency of CD4⁺ T cells specific for the stimulating antigen within the antigen-responsive population of CFSE-based proliferation assays. We compared CD4⁺ T-cell responses to a type 1 diabetes autoantigen (proinsulin C-peptide) and to a vaccine antigen (tetanus toxoid). The TCRs expressed by antigen-responsive CD4⁺ T cells were sequenced, and their antigen specificity was tested functionally by expressing them in a reporter T-cell line. Responses to C-peptide were weak, but detectable, in PBMC from individuals with T1D, whereas responses to tetanus toxoid were much stronger. The frequency of antigen-specific CD4⁺ T cells correlated with the strength of the response to antigen in the proliferation assay. However, antigen-specific CD4⁺ T cells were rare among antigen-responsive CD4⁺ T cells. For C-peptide, an average frequency of 7.5% (1%-11%, n = 4) of antigen-responsive CD4⁺ T cells were confirmed to be antigen specific. In the tetanus-toxoid-stimulated cultures, on average, 45% (16%-78%, n = 5) of the antigen-responsive CD4⁺ T cells were tetanus toxoid specific. These data show that antigen-specific CD4⁺ T cells are a minority of the cells that proliferate in response to antigen and have important implications for in vitro CD4⁺ T-cell proliferation assays.

Keywords: C-peptide; CD4+ T cell; CFSE assay; antigen specific T cell; autoimmunity; clonal expansion; proliferation; tetanus toxoid.

Figure 1

Workflow to identify antigen-specific CD4⁺ T cells in the CFSE-based proliferation assay. The workflow started from isolating PBMC from each donor, which were stained with CFSE and cultured with tetanus toxoid or C-peptide for 7 days (A). Then, hash-tagging was used to mark cells of different antigen treatments. Using FACS, CFSE^dim CD4⁺ cells were sorted for different antigen treatments to isolate the antigen-responsive population (B). These CFSE^dim CD4⁺ cells derived from different antigens were pooled and processed for single-cell TCR sequencing using 10X Genomics’ 5′ platform (C). The first output from the TCR sequencing is represented by the TCR clonotype abundancy distribution plot (D). Then, the most abundant paired TCR clonotypes were screened to identify antigen-specific TCRs among the antigen-responding population (E).

Figure 2

Screening TCRs for response to tetanus toxoid and C-peptide. Frequency distribution of the 25 most abundant TCR clonotypes derived from Donor 282T2’s antigen-responsive CD4⁺ T cells for tetanus toxoid (A) and C-peptide (B). The TCR clonotypes are numbered according to their abundance in the antigen-responsive population (1–25, most to least abundant). TCR clonotypes that comprised two TRAVs are indicated as “A” and “B” after the clone number. JNL avatars expressing the most abundant ~25 TCR clonotypes were screened for antigen specificity. Responses to antigen tetanus toxoid (C) and C-peptide (D) are expressed as Δluciferase. The antigen-specific responses plotted for individual JNL avatars are shown as a percentage of their response to anti-CD3 for tetanus toxoid (E) and C-peptide (F). Anti-CD3 responses to the TCR negative, parental JNL cells plus 3× standard deviation of the “nil antigen” were used as a threshold for a positive response. Responses >30% of the anti-CD3 response, represented by a blue dotted line, are weak responding clones, and those >90% of the response to anti-CD3, represented by a red dotted line, were considered to be strongly antigen-specific TCRs.

Figure 3

Validating the screening assays using T-cell avatar. (A) Antigen-responsive TCRs specific for tetanus toxoid or C-peptide derived from each of the three categories in the screening assay were validated using purified CD4⁺ T-cell avatar against their respective antigens (A). The proportion of verified, antigen-specific TCRs is shown as the percentage of all TCRs tested for each category of response. Total TCRs tested for each category and for each antigen (C-peptide, shown as red bars, and tetanus toxoid, shown as a blue bar) were represented. Responses by JNL cells expressing tetanus-toxoid-specific TCRs (B) from Donor 312T4 and C-peptide (C), from Donor 282T2, are shown for each category of response. A Δluciferase reading of 2 × 10⁴ RLU was determined as a threshold response and is shown here with a dotted line.

Figure 4

Analyzing the frequency of antigen-specific CD4⁺ T cells and their distribution. The proportion of tetanus-toxoid-specific (blue bars) or C-peptide-specific (red bars) T cells (A) from each donor is shown as a percentage of the TCRs tested. The numbers are the frequency for each bar. (B) The strength of the antigen-specific response in the original CFSE assay measured by the cell division index (CDI log₁₀) for C-peptide (red) and TT (blue) was correlated with the proportion of antigen-specific TCR clone identified for each donor. R ² is the correlation coefficient. (C) The magnitude of all antigen-specific CD4⁺ T-cell responses in the screening assays, relative to anti-CD3, is shown plotted against their individual clonotype rank. Responses to tetanus toxoid are represented by black triangles, and those of C-peptide are represented by red triangles. The blue dotted line indicates the threshold for a weak (30%) antigen-specific response, and the dotted red line represents the threshold (100%) for high-responding antigen-specific TCRs. The green-shaded area indicates the top-10-ranked clonotypes. (D) The inverse of the minimum stimulatory concentration) for C-peptide was plotted for all avatars tested for each donor. The C-peptide concentration used was between 0.016 and 50.0 μM. Any JNL avatars that did not respond to the highest concentration (50 μM) of C-peptide were arbitrarily given a value of 50 μM or 0.02. These responses are highlighted with a gray box.