Chimeric antigen receptor T (CAR-T)-cell therapy is revolutionizing the treatment of hematological malignancies1,2 with unprecedented response rates reaching 80%–90% observed in acute lymphocytic leukemia and diffuse large B-cell lymphoma. With recent Food and Drug Administration (FDA) approvals of Kymriah and Yescarta and hundreds of clinical trials3 underway for various indications, including treatment of solid tumors, demand for these therapies is expected to rise significantly. The production of CAR-T therapies requires the transduction of a chimeric antigen receptor for tumor recognition followed by large scale expansion of autologous T cells to meet the dose requirements for therapeutic response. The transduction of viral vectors is more efficient when the T cells are activated using a combination of activating and costimulating antibodies resulting in T-cell proliferation.
Two approaches have been broadly used for T-cell activation; one involving surface-immobilized4 anti-CD3 and anti-CD28 antibodies (flat surfaces or beads) and another using antibodies in solution. The most commonly used platform for immobilized antibodies is bead-based due to the enhanced available surface area. While bead-based approaches require a debeading step postexpansion leading to significant cell loss, the expansion rates with soluble antibodies are much lower requiring longer expansion times. This reduction in expansion rate is related to a lack of significant receptor clustering with untethered antibodies. Soluble anti-CD3 has also been used in conjunction with antigen-presenting cells (APCs) or artificial APCs with impressive results.5 However, time and cost involved with the inclusion of another engineered cell line in the expansion process and variability in results associated with artificial APCs is not ideal. A more recent approach using colloidal nanoparticles that does not require a separate debeading step provides similar levels of expansion as Dynabeads has also emerged though, a trend towards higher programmed cell death protein-1 receptor expression in cells activated with the nanomatrix has been reported.5 Thus, an alternate approach that maintains robust expansion while eliminating the need for bead removal or utilizing additional cell lines is highly desirable. This report describes a novel T-cell activation platform [DNA-based T-cell activator (DBTA)] comprised of a linear single-strand DNA (ssDNA) scaffold prepared by rolling circle amplification (RCA) and scaffold-complementary oligo conjugated anti-CD3 and anti-CD28 antibodies.
MATERIALS AND METHODS
DNA Scaffold Preparation
All deoxyoligonucleotides [oligo(s)] were synthesized by Integrated DNA Technologies (Coralville, IA) and purified by desalting. The ssDNA scaffold was prepared by a modification of the standard double-stranded RCA reaction conditions [supplementary information (SI)].
Preparation of Oligo-modified Antibody Conjugates
Antibodies were conjugated to scaffold-complementary oligonucleotides in a 2-step maleimide-thiol coupling strategy (SI). Oligonucleotide loading on each antibody was calculated by the method of Zhou et al6 using A260 and A280 absorbance measurements.
Activation studies were conducted in 6-well tissue culture plastic plates using commercially available Pan T cells (AllCells, Alameda, CA). Cultures were diluted as needed (1:3 or 1:4) to enable a 7-day expansion protocol (SI).
Clinical CAR-T vector transduction was carried out at day 2 at a multiplicity of infection of 40 using GMP-grade lentiviral particles harboring a chimeric antigen receptor construct specific for CD19 and CD227 (SI).
Flow Cytometry Studies
Antibodies used and their sources are listed in Supplementary Materials (Supplemental Digital Content 1, http://links.lww.com/JIT/A559, Table 1). Samples were analyzed using the CytoFLEX flow cytometer (Beckman Coulter, Brea, CA) for T-cell subsets, T-cell differentiation states, and homing and senescence markers.
TABLE 1 -
Antibodies Used and Their Sources
|Mouse anti-human CD3-PerCPCy5.5 clone UCHT1
|Mouse anti-human CD4-V500
|Mouse anti-human CD4-PE
|Mouse anti-human CD8-AF488 clone RPA-T8
|Mouse anti-human CD25-PE clone MA251
|Mouse anti-human CD57-APC clone NK-1
|Mouse anti-human CD45RO-PECy7
|Mouse anti-human CD62L-V450 clone DREG-56
|Mouse anti-human CCR7 clone 150503
|Mouse isotype control IgG2a-PE
All data were expressed as the mean±SD. Statistical comparisons were performed using Student t test or 1-way analysis of variance (P<0.05 were considered to be statistically significant).
Characterization of the Platform Components
DBTA is a DNA hybridization–based platform (Fig. 1) used to cluster antibodies and their target receptors on T cells to initiate cell activation and subsequent clonal expansion. It consists of the following components: (1) a concatenated ssDNA polymer prepared by ligating a linear oligo into a circle (Supplementary Fig. 1, Supplemental Digital Content 1, http://links.lww.com/JIT/A559) and using this circle as a template for RCA, and (2) activating and costimulating antibodies conjugated to an oligonucleotide with partial or complete complementarity to the ssDNA. Melting temperature analysis of the ssDNA demonstrated no hyperchromatic shift seen with double-stranded DNA (Supplementary Fig. 2, Supplemental Digital Content 1, http://links.lww.com/JIT/A559) as the hydrogen bonds between the bases are broken. The length of the RCA product was determined to be around 40,000 bases and a gel mobility shift assay was also accomplished that demonstrated hybridization of the oligo conjugated antibody to the ssDNA (data not shown).
DBTA Is a Flexible and Tunable Activation Reagent Platform
We compared the T-cell activation and expansion potential of the novel DBTA reagent against Dynabeads CD3/CD28 CTS (Invitrogen, Carlsbad, CA) utilizing antibodies to the CD3 subunit of the T-cell receptor complex and the CD28 costimulatory molecule. Initial studies were carried out with an early version of the DBTA reagent which employed the anti-CD3 antibody-oligo conjugate immobilized on the ssDNA and an unmodified anti-CD28 antibody. Early 24 hours activation kinetics (measured by CD25 expression) were significantly slower with the unmodified anti-CD28 in comparison to Dynabeads (Fig. 2A), with similar fold expansion by day 7 (Fig. 2B). When both the antibodies were modified with oligo and immobilized on the linear ssDNA, activation (Fig. 3A) and expansion (Fig. 3B) was higher than the unmodified anti-CD28 antibody version and comparable to Dynabeads.
To further characterize the ability of the DBTA reagent to tune activation kinetics, we evaluated the effect of varying levels of conjugate antibody—ssDNA hybridization on activation kinetics by varying the oligo:antibody (protein) (O/P) ratios of the conjugates. Figure 4 shows the performance of DBTA where both anti-CD3 and anti-CD28 antibodies are conjugated to the Mal-o20b(+)act oligo (a 20-base oligo with maleimide functionality for conjugation and a modified β-actin sequence with full complementarity to ssDNA scaffold) but with varying O/P ratios. Using antibody conjugates having O/P ratios of 2.8 or higher, the initial 24-hour expression of CD25 postactivation was found to be ≥90% (Fig. 4C). This improved version of the DBTA reagent was utilized for T-cell viral transduction studies. Preliminary studies with a commercially available GFP lentiviral vector (LentiBrite GFP Particles, MilliporeSigma), showed high (>90%) transduction with both DBTA and Dynabeads at day 5 (Supplementary Fig. 3, Supplemental Digital Content 1, http://links.lww.com/JIT/A559). Based on this, an activation-transduction-expansion study was carried out with a CD19/CD22 bispecific chimeric antigen receptor.7 At the end of expansion on day 8, much higher fold expansion (57- vs. 13-fold; Fig. 5A) and activation (65% vs. 43.6% CD25+; Fig. 5B) were observed with the DBTA in comparison to the Dynabeads, even with transient 24 hours activation with the DBTA reagent. The ratio of CD4/CD8 subsets within the final cell population was similar (Fig. 5C). In a single donor study, transduction efficiency (Fig. 5D) with DBTA was higher compared with Dynabeads. In our earlier studies with a commercial GFP vector (Lentibrite GFP) used at 20 multiplicity of infection, also in a single donor, >95% transduction efficiency was observed (data not shown) with both reagents. Despite having a higher fold expansion at day 8 in comparison to the Dynabeads, no major differences in the exhaustion (programmed cell death protein-1) marker expression (Fig. 5E) was apparent between the 2 groups. The expanded cell population in the DBTA group had a much higher percentage of central memory T cells (Tcm) in comparison to Dynabeads for both, the CD4+ and CD8+ subpopulation of cells (Fig. 5F, Supplementary Fig. 4, Supplemental Digital Content 1, http://links.lww.com/JIT/A559). Ex vivo T-cell expansion protocols generating a higher percentage of Tcm cells are critical to in vivo efficacy and persistence of the CAR-T product.8
In vitro expansion of cytotoxic T cells is an essential step in generating a CAR-T dose. Several recent studies have highlighted the effect of the choice of cytokines and costimulatory domains on the quality and quantity of the expanded CAR-T cells.9–12 Similarly, shortening the duration of T-cell activation by providing transient stimulation has been shown to enhance CD8+ T-cell expansion.13 We have developed a novel soluble, tunable, T-cell activator that is based on in situ DNA hybridization to cluster the T-cell surface receptors required for T-cell activation and provide costimulatory signals to increase proliferation. The noncovalent and reversible immobilization of anti-CD3 and anti-CD28 on the ssDNA-based scaffold provides higher activation than soluble antibodies and allows for unprecedented fine-tuning of T-cell response. Our results indicate that the activation and proliferation kinetics with the DBTA is robustly comparable to the CD3/CD28 Dynabeads and potentially scalable (Supplementary Fig. 5, Supplemental Digital Content 1, http://links.lww.com/JIT/A559). Reducing the in vitro expansion time for CAR-T cells is also being considered as a strategy to limit differentiation to effector T cells.14 Such protocols rely on the ability of the activation reagent to induce rapid activation and T-cell proliferation.
The in situ DNA hybridization strategy is particularly attractive offering many advantages. The scaffold used in this strategy provides a high degree of flexibility in controlling the strength of immobilization, amount of clustering, and multiple and flexible receptor/coreceptor targeting by changing length and composition of the complementary sequences and the types and loading of antibody conjugates. Finally, the ssDNA scaffold can be readily released or degraded, if required, by using DNases (Supplementary Fig. 6, Supplemental Digital Content 1, http://links.lww.com/JIT/A559) which are already FDA approved for use in clinical cord blood applications.15
The authors thank N. Shah (Pediatric Oncology Branch, NCI, NIH) for providing the lentiviral vector for the CD19/CD22 bispecific chimeric antigen receptor.
Conflicts of Interest/Financial Disclosures
Supported by funds from GE Healthcare. V.K., A.S., E.L., E.K., R.S.D., C.C., R.S., M.B. are current or past employees of GE Research. The remaining authors have declared that there are no financial conflicts of interest with regard to this work.
1. Park JH, Geyer MB, Brentjens RJ. CD19-targeted CAR T-cell therapeutics for hematologic malignancies: interpreting clinical outcomes to date. Blood. 2016;127:3312–3320.
2. Quintás-Cardama A. CD19 directed CAR T cell therapy in diffuse large B-cell lymphoma. Oncotarget. 2018;9:29843–29844.
3. ClinicalTrials.gov. Search of: CAR-T—list results. 2019. Available at: www.clinicaltrials.gov/ct2/results?cond=&term=CAR-T&cntry=&state=&city=&dist=
. Accessed February 7, 2019.
4. Hollyman D, Stefanski J, Przybylowski M, et al. Manufacturing validation of biologically functional T cells targeted to CD19 antigen for autologous adoptive cell therapy. J Immunother. 2009;32:169–180.
5. Mock U, Nickolay L, Philip B, et al. Automated manufacturing of chimeric antigen receptor T cells for adoptive immunotherapy using CliniMACS prodigy. Cytotherapy. 2016;18:1002–1011.
6. Zhou Z, Xiang Y, Tong A, et al. Simple and efficient method to purify DNA-protein conjugates and its sensing applications. Anal Chem. 2014;86:3869–3875.
7. Qin H, Ramakrishna S, Nguyen S, et al. Preclinical development of bivalent chimeric antigen receptors targeting both CD19 and CD22. Mol Ther Oncolytics. 2018;11:127–137.
8. Finney OC, Brakke H, Rawlings-Rhea S, et al. CD19 CAR T cell product and disease attributes predict leukemia remission durability. J Clin Invest. 2019;129:2123–2132.
9. Hinrichs CS, Spolski R, Paulos CM, et al. IL-2 and IL-21 confer opposing differentiation programs to CD8+ T cells for adoptive immunotherapy. Blood. 2008;111:5326–5333.
10. Zeng R, Spolski R, Finkelstein SE, et al. Synergy of IL-21 and IL-15 in regulating CD8+ T cell expansion and function. J Exp Med. 2005;201:139–148.
11. Klebanoff CA, Finkelstein SE, Surman DR, et al. IL-15 enhances the in vivo antitumor activity of tumor-reactive CD8+ T cells. Proc Natl Acad Sci. 2004;101:1969–1974.
12. Zhang H, Snyder KM, Suhoski MM, et al. 4-1BB is superior to CD28 costimulation for generating CD8+ cytotoxic lymphocytes for adoptive immunotherapy. J Immunol. 2007;179:4910–4918.
13. Kagoya Y, Nakatsugawa M, Ochi T, et al. Transient stimulation expands superior antitumor T cells for adoptive therapy. JCI Insight. 2017;2:e89580.
14. Ghassemi S, Nunez-Cruz S, O’Connor RS, et al. Reducing ex vivo culture improves the antileukemic activity of chimeric antigen receptor (CAR) T cells. Cancer Immunol Res. 2018;6:1100–1109.
15. Kaur I, Zulovich JM, Gonzalez M, et al. Comparison of two methodologies for the enrichment of mononuclear cells from thawed cord blood products: the automated Sepax system versus the manual Ficoll method. Cytotherapy. 2017;19:433–439.