Magnetic-activated cell sorting of TCR-engineered T cells, using tCD34 as a gene marker, but not peptide-MHC multimers, results in significant numbers of functional CD4+ and CD8+ T cells

Abstract

T cell-sorting technologies with peptide-MHC multimers or antibodies against gene markers enable enrichment of antigen-specific T cells and are expected to enhance the therapeutic efficacy of clinical T cell therapy. However, a direct comparison between sorting reagents for their ability to enrich T cells is lacking. Here, we compared the in vitro properties of primary human T cells gene-engineered with gp100(280-288)/HLA-A2-specific T cell receptor-αβ (TCRαβ) on magnetic-activated cell sorting (MACS) with various peptide-MHC multimers or an antibody against truncated CD34 (tCD34). With respect to peptide-MHC multimers, we observed that Streptamer(®), when compared with pentamers and tetramers, improved T cell yield as well as level and stability of enrichment, of TCR-engineered T cells (>65% of peptide-MHC-binding T cells, stable for at least 6 weeks). In agreement with these findings, Streptamer, the only detachable reagent, revealed significant T cell expansion in the first week after MACS. Sorting TCR and tCD34 gene-engineered T cells with CD34 monoclonal antibody (mAb) resulted in the most significant T cell yield and enrichment of T cells (>95% of tCD34 T cells, stable for at least 6 weeks). Notably, T cells sorted with CD34 mAb, when compared with Streptamer, bound about 2- to 3-fold less peptide-MHC but showed superior antigen-specific upregulated expression of CD107a and production of interferon (IFN)-γ. Multiparametric flow cytometry revealed that CD4(+) T cells, uniquely present in CD34 mAb-sorted T cells, contributed to enhanced IFN-γ production. Taken together, we postulate that CD34 mAb-based sorting of gene-marked T cells has benefits toward applications of T cell therapy, especially those that require CD4(+) T cells.

FIG. 1.

Gene constructs and sorting reagents to MACSort TCR-engineered human T cells. (A) Schematic representation of TCRα, TCRβ, and TCRα-2A-tCD34-2A-TCRβ transgenes used to gene-engineer primary human T cells. TCR specific for gp100/A2 comprised TRAV13-1*02/J52*01/CA and TRBV27*01/J2-7*01/D2*02/CB2 (Schaft et al., 2003). tCD34 represents a truncated and functionally inert variant of CD34. T cells were transduced either with pBullet:TCRα+pBullet:TCRβ and termed “TCR T cells” or with pBullet:TCRα-2A-tCD34-2A-TCRβ and termed “TCR-tCD34 T cells.” Abbreviations: V, TCRαβ-variable domain; C, TCRαβ-constant domain; D, TCRβ-diversity domain; J, TCRαβ-joining domain; 2A, 2A sequence encoding a self-cleaving peptide; tCD34, truncated CD34 molecule. (B) Schematic illustration of reagents used to MACSort TCR-engineered human T cells. From left to right: Tetramer (Altman et al., 1996), pentamer (Sebestyen et al., 2008), Streptamer (Knabel et al., ; Neudorfer et al., 2007), and CD34 mAb (Stull et al., 2000). Tetramers consist of 4 gp100 peptide (YLEPGPVTA)/HLA-A2 (gp100/A2) monomers that were biotinylated and multimerized with streptavidin–phycoerythrin; pentamers consist of 5 gp100/A2 monomers that were linked to phycoerythrin and multimerized by a self-assembling, coiled-coil structure; and Streptamers consist of 8–12 gp100/A2 monomers that were Strep-tagged and multimerized with Strep-Tactin–phycoerythrin (depicted with 5 monomers). Anti-CD34 mAbs consist of heavy and light chains providing two antigen-binding sites and are directly coupled to magnetic microbeads.

FIG. 2.

MACS with Streptamer or CD34 mAb improves T cell yield. Primary human T cells were transduced either with TCR or TCR-tCD34 transgenes as depicted in Fig. 1A. TCR T cells were MACSorted with tetramers, pentamers, and Streptamers, whereas TCR-tCD34 T cells were MACSorted with Streptamers and CD34 mAbs. After MACSorting, T cell numbers were counted microscopically and T cells were expanded with feeder cultures (as described in Materials and Methods). Mean T cell numbers and SEM (A) directly after MACS or (B) 1 week after MACS are from two to six repeat experiments with T cells from two to four healthy donors. (C) T cell expansion 1 week (solid columns) and 3–6 week(s) (open columns) after MACS is expressed as the fold increase in T cell numbers when comparing 1 week and directly after MACS and 3–6 weeks and 1 week after MACS, respectively. T cell expansions are from representative T cell cultures out of two to four repeat experiments with T cells from two to four healthy donors with similar results. Statistically significant differences were calculated with Student t tests: *p<0.05; **p<0.005; ***p<0.0005.

FIG. 3.

MACS of TCR T cells with Streptamers results in enhanced peptide–MHC binding that is stable over time. Primary human T cells were transduced and MACSorted as described in the legend to Fig. 2. Sorted T cells were analyzed by flow cytometry for (A) peptide–MHC binding (using gp100/A2 pentamer) and (B) surface expression of tCD34 (using CD34 mAb) at various time points: before MACS (solid columns), 1 week after MACS (open columns), and 3–6 weeks after MACS (shaded columns). Columns and error bars represent the mean percentage and SEM of two to four repeat experiments with T cells from two to four healthy donors. Binding of peptide–MHC or CD34 mAb before MACS was 20 and 70%, respectively. Enrichment of T cells that bind peptide–MHC or CD34 mAb is presented as the fold difference between T cells 1 week after and before MACS, and is indicated above the corresponding bars. Statistically significant differences were calculated with Student t tests: *p<0.05; **p<0.005; ***p<0.0005.

FIG. 4.

Human T cells sorted with CD34 mAb demonstrate increased antigen-specific responses. Primary human T cells were transduced with empty virus particles (mock T cells) or with TCR (TCR T cells) or TCR-tCD34 transgenes (TCR-tCD34 T cells) and were either not MACSorted (mock T cells, patterned columns) or MACSorted with Streptamers (TCR T cells, shaded columns; TCR-tCD34 T cells, open columns) or anti-CD34 mAbs (TCR-tCD34 T cells, solid columns). T cells sorted with CD34 mAb show upregulated (A) surface expression of T cell CD107a and (B) IFN-γ production on stimulation with antigen-positive melanoma cells. T cells were stimulated with the following target cells: BLM cells (gp100^–/A2⁺) loaded without or with 10 μM gp100 peptide, and FM3 cells (gp100⁺/A2⁺). CD107a expression (percent) was measured by flow cytometry, gating on viable and CD3-positive T cells, after a 2-hr stimulation, and IFN-γ production (ng/ml) was measured after overnight stimulation with conditioned supernatants by ELISA. Columns and error bars represent mean values and SEM of two independent measurements from two healthy donors. CD107a and IFN-γ data were corrected for differences in peptide–MHC binding, with peptide–MHC binding of CD34 mAb-sorted T cells set to 100% (see Supplementary Fig. S1A for details). Statistically significant differences were calculated by Student t test: *p<0.05; **p<0.005.

FIG. 5.

IFN-γ production of CD34 mAb-sorted T cells, but not peptide–MHC-sorted T cells, depends on both CD8⁺ and CD4⁺ T cells. Primary human T cells were transduced and MACSorted as described in the legend to Fig. 4. T cells were analyzed by flow cytometry after staining with APC-conjugated anti-CD8α mAb in combination with one of the following reagents: (A) PE-conjugated gp100/A2 tetramer (n=12); (B) PE-conjugated anti-TCR-V_β27 mAb (n=6); (C) FITC-conjugated anti-CD34 mAb (n=14); and (D) FITC-conjugated anti-IFN-γ mAb (n=2). In the case of IFN-γ stainings, T cells were stimulated for 6 hr with BLM (gp100^–/A2⁺) cells loaded with 10 μM gp100 peptide, permeabilized, and stained with anti-IFN-γ mAb, as described in more detail in Materials and Methods. Shown are representative dot plots, indicating percentages (all quadrants) and MFIs (upper and lower right quadrants when percent ≥5, in italic) of stained T cells.