The development of tolerizing therapies aiming to inactivate autoreactive effector T-cells is a promising therapeutic approach to control undesired autoimmune responses in human diseases such as Type 1 Diabetes (T1D). A critical issue is a lack of sensitive and reproducible methods to analyze antigen-specific T-cell responses, despite various attempts. We refined a proliferation assay using the fluorescent dye 5,6-carboxylfluorescein diacetate succinimidyl ester (CFSE) to detect responding T-cells, highlighting the fundamental issues to be taken into consideration to monitor antigen-specific responses in patients with T1D. The critical elements that maximize detection of antigen-specific responses in T1D are reduction of blood storage time, standardization of gating parameters, titration of CFSE concentration, selecting the optimal CFSE staining duration and the duration of T-cell stimulation, and freezing in medium containing human serum. Optimization of these elements enables robust, reproducible application to longitudinal cohort studies or clinical trial samples in which antigen-specific T-cell responses are relevant, and adaptation to other autoimmune diseases.
Type 1 Diabetes (T1D) is a chronic, incurable autoimmune disorder in which insulin-producing β-cells are selectively destroyed by islet-infiltrating T-cells. The diagnosis of T1D is currently made at the onset of clinical symptoms. However, a long prodrome of autoimmune T and B cell activation leading to immune-mediated β-cell destruction precedes the onset of clinical diabetes. CD4+ T-cell responses are central to the pathogenesis of T1D and several pathogenic self-epitopes have been reported. Human islet-infiltrating CD4+ T-cell clones were shown to recognize epitopes from proinsulin-derived C-peptide, as well as neoantigens such as hybrid insulin peptides (HIPs). An important goal is the development of immunotherapies that prevent progression of β-cell destruction, thus intercepting T1D. Detecting and analyzing the function of islet-specific T-cells in humans has been challenging because of the low frequency of antigen-specific T-cells in the circulation and incomplete knowledge of their antigen specificity. Consequently, there is currently an unmet need for a robust assay capable of tracking islet antigen-specific autoreactive T-cells to monitor immune-mediated activity while patients are undergoing immunotherapy. The desirable requirements of such an assay include 1) reliable performance with small blood volumes, 2) simplicity, 3) high sensitivity and specificity for detecting antigen specific T-cells in patients with or at-risk of T1D, 4) reproducibility, and 5) applicability to frozen peripheral blood mononuclear cells (PBMCs).
Fluorescent dye-based proliferation assays represent a popular method to monitor antigen-specific T-cell responses in vitro. In these assays, PBMCs are labeled with a fluorescent dye, such as carboxyfluorescein diacetate succinimidyl ester (CFSE), and divide in response to antigenic stimuli. The resulting progeny retain half the number of CFSE molecules of its parent. The corresponding decrease in fluorescence intensity can be measured by flow cytometry and identifies proliferating cells. Dividing cells can also be phenotypically characterized using antibodies specific for surface markers and/or intracellular cytokines. CFSE is an ideal dye to measure cell division in view of its capacity to label lymphocyte populations with a high fluorescent intensity. It is also compatible with a broad range of other fluorochromes, thus allowing multi-color flow cytometry, and single-cell sorting to clone antigen-reactive T-cells. The primary limitations of flow cytometric dye-dilution assays relate to reproducibility, resulting from variable background proliferation in unstimulated wells, subjective gating, and inter-operator variability of up to 78%. Background proliferation decreases assay sensitivity, particularly with antigens that induce low levels of T cell proliferation, e.g., due to low precursor frequency in circulation or low affinity T cell receptors (TCRs). Autoreactive T-cells may also be more sensitive to apoptosis than other antigen-specific T cells. For example, low affinity interactions between a pre-proinsulin peptide and the Human Leucocyte antigen (HLA)-A2 molecule reduces binding affinity of the TCR-peptide-HLA complex, potentially enabling T-cells to escape thymic deletion and enter the periphery.
Here we optimized the sensitivity of a published CFSE-based T-cell proliferation assay that demonstrated proinsulin 33-63–specific CD4+ T cells in patients with recent-onset T1D. We used PBMCs from children with T1D for less than 3 months, to systematically optimize assay parameters that contribute to either death of the responding antigen-specific T-cells or that increase background proliferation in the absence of added antigen. The described protocol results in a reproducible assay that can be replicated in other laboratories and adapted for monitoring in other autoimmune diseases.
Fresh blood (5–15 mls) was obtained from children (aged 2–16 years) with T1D duration of ≤ 3 months after informed consent. Recruited individuals carried alleles associated with high risk of T1D (HLA DR3-DQ2, DR4-DQ8, or DR3-DQ2/DR4-DQ8) as previously described. The study was approved by the Children’s Health Queensland, Mater Hospital, and University of Queensland Human Research Ethics Committees. T1D was defined according to the criteria from the American Diabetes Association.
PBMCs were isolated by Ficoll Density (GE Healthcare, Sweden) gradient centrifugation then washed twice in phosphate buffered saline. Islet antigen-specific T-cell responses were measured using freshly isolated or thawed frozen PBMCs labeled with CFSE and cultured with or without islet peptides. Cells were cultured in complete medium [RPMI1640 medium, 1 U/ml penicillin/streptomycin/glutamate (Invitrogen, Thermo Fisher Scientific), 1 mM sodium pyruvate (Gibco, Thermo Fisher Scientific)] supplemented with 5% human serum (HS, Merck). Culture medium was filtered through a 0.2 µM filter (Sartorius-minisart) prior to culture or washing steps. PBMCs from single patients were analyzed separately.
PBMCs were frozen at 10–25 × 10 6 cells/vial in 1.0 ml of either 10% DMSO/fetal bovine serum (FBS, heat inactivated, Life Technologies), 10% DMSO/HS or a commercial serum-free freezing medium (CTL Cryo™ ABC Media Kit, Immunospot, USA). Counted cells were initially resuspended in 0.5 ml of either HS or CTL solution C and an equal volume of either 20% DMSO/FBS, 20% DMSO/HS or CTL solution AB was added. The cell suspension was aliquoted into cryovials and placed in a “Mr Frosty” (Nalgene, Thermo Scientific, Denmark) containing iso-propanol, at −80°C for 24–48 h before being transferred to liquid nitrogen.
Cells were thawed in a 37°C water-bath for 30 s and then added slowly to pre-warmed 5% HS/complete media containing 12.5 μg/ml DNase I (Sigma). The cell suspension was centrifuged, and the supernatant discarded. Cells were then resuspended in pre-warmed 5% HS/complete media supplemented with 6.25 μg/ml DNase I and rested for 30 min at 37°C 5% CO 2. The cell suspension was passed through a cell strainer to remove dead cell clumps, then centrifuged and resuspended in complete medium.
Tetanus toxoid (AJ Vaccines) was used at a concentration of 10 Lf/ml. Ultra-LEAF purified anti-human CD3 (Biolegend clone OKT3) was used at 0.1 µg/ml. Proinsulin 33-63 (PI 33-63, Sequence: EAEDLQVGQVELGGGPGAGSLQPLALEGSLQ, GL Biochem), was initially reconstituted to 16.55 M in DMSO (Sigma Aldrich) before making a working stock in PBS, of 2.5–5.0 mM for use at a final concentration of 10 µM as previously reported to reliably induce CD4+ T cell proliferation.
We followed methods previously described for a CFSE-based T-cell proliferation assay to measure antigen-specific CD4+ T-cell proliferation to PI 33-63. Briefly, a total of 10–20 × 10 6 PBMCs were re-suspended at 1 × 10 6 cells/ml in pre-warmed (37°C) 1XPBS before staining with CFSE dye (CellTrace™ CFSE Cell Proliferation Kit for flow cytometry, Invitrogen, Thermo Fisher Scientific). CFSE concentration and staining times are as indicated in the figure legends. The CFSE labeling reaction was terminated with pre-warmed, filtered complete media containing 2.5% HS. The cells were washed and then re-suspended at a concentration of 2 × 10 6 PBMCs/ml in pre-warmed, filtered 5% HS/RPMI. CFSE-stained cells (200 µl per well) were cultured in 96-well round bottom plates (Costar) for the number of days indicated in the figure legends, in 37°C 5% CO 2 with medium alone (negative control) or with PI 33-63, or with positive controls, tetanus toxoid and/or anti-CD3. In most experiments, cells were cultured in triplicate for each condition. Unstained cells were included in all cultures and used as single color stained controls to determine compensation settings on the flow cytometer.
After culture, triplicate wells were pooled, washed in saline and stained on ice with the following antibodies: anti-human CD3-APC/Cy7 (Mouse IgG1, κ, clone UCHT1), anti-human CD4-PE/Cy7 Antibody (Mouse IgG2b, κ, clone OKT4) and anti-human CD8a-Pacific Blue™ Antibody (Mouse IgG1, κ, clone HIT8a) (all from Biolegend, San Diego, CA). Gating of the major T cell populations was set using Fluorescence minus one (FMO) controls. LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit (Invitrogen, Thermo Fisher Scientific) was used to exclude dead cells. Harvested cells were acquired on a Gallios flow cytometer (Beckman Coulter) and analyzed with Kaluza software (Version 2.1, Beckman Coulter). Optimal voltage settings were determined for each experiment based on unstained and single stained samples. Viable CD3+CD4+ lymphocytes were gated. We acquired at least 2 × 10 5 lymphocytes in order to identify 5,000 CD4+ CFSE(undivided) cell events. The results are presented as a cell division index (CDI), the ratio of CD4+ T-cells that proliferated in response to antigen, relative to cells that proliferated in absence of antigen. The CDI was calculated based on a fixed number of 5,000 CD4+ CFSE(undivided) cells using the following formula:
C D I=n u m b e r o f d i v i d e d C D 4+T c e l l s p e r 5,000 C D 4+T c e l l s i n C F S E(u n d i v i d e d)f r o m"w i t h a n t i g e n"g r o u p n u m b e r o f d i v i d e d C D 4+T c e l l s p e r 5,000 C D 4+T c e l l s i n C F S E(u n d i v i d e d)f r o m"w i t h o u t a n t i g e n"g r o u p
A CDI of ≥3 was considered to represent the threshold for the positive control responses. Data from CFSE-based T-cell proliferation assays in which the CDI for both positive control antigen(s) did not exceed 3.0 were excluded from the analysis.
Freshly isolated PBMCs (10–20 × 10 6) were labeled with BD Horizon™ Violet Proliferation Dye 450 (VPD 450) at 10 × 10 6 cells/ml in pre-warmed PBS with the concentrations of VPD450 as indicated. After 15 min the reaction was terminated with pre warmed 5% HS/complete medium. After washing, PBMC were resuspended to 2 × 10 6/ml in pre-warmed, filtered 5% HS/complete medium.
To determine the precision and reproducibility of the CFSE-based T-cell proliferation assay, we produced a standard operating procedure (SOP) and measured CD4+ T-cell responses to PI 33-63 and tetanus toxoid from cryopreserved PBMC of a single patient diagnosed with T1D ≤ 3 months. We qualified the assay as fit-for purpose for research environments and non-regulated laboratories per recent guidelines for flow cytometry assays. Two analysts conducted independent experiments from the same sample. In each assay, six wells were set up for each condition. After 7 days of incubation, two wells were pooled together and analyzed as a single sample, resulting in three replicates. CDI values were calculated either from each
