Rapidly generated multivirus-specific cytotoxic T lymphocytes for the prophylaxis and treatment of viral infections

Abstract

Severe and fatal viral infections remain common after hematopoietic stem cell transplantation. Adoptive transfer of cytotoxic T lymphocytes (CTLs) specific for Epstein-Barr virus (EBV), cytomegalovirus (CMV), and adenoviral antigens can treat infections that are impervious to conventional therapies, but broader implementation and extension to additional viruses is limited by competition between virus-derived antigens and time-consuming and laborious manufacturing procedures. We now describe a system that rapidly generates a single preparation of polyclonal (CD4(+) and CD8(+)) CTLs that is consistently specific for 15 immunodominant and subdominant antigens derived from 7 viruses (EBV, CMV, Adenovirus (Adv), BK, human herpes virus (HHV)-6, respiratory syncytial virus (RSV), and Influenza) that commonly cause post-transplant morbidity and mortality. CTLs can be rapidly produced (10 days) by a single stimulation of donor peripheral blood mononuclear cells (PBMCs) with a peptide mixture spanning the target antigens in the presence of the potent prosurvival cytokines interleukin-4 (IL4) and IL7. This approach reduces the impact of antigenic competition with a consequent increase in the antigenic repertoire and frequency of virus-specific T cells. Our approach can be readily introduced into clinical practice and should be a cost-effective alternative to common antiviral prophylactic agents for allogeneic hematopoietic stem cell transplant (HSCT) recipients.

Figure 1

Growth-promoting cytokines enhance the activation and expansion of antigen-specific cytotoxic T lymphocytes (CTLs). Peripheral blood mononuclear cells (PBMC) were stimulated with pp65 pepmix in the presence of IL2, IL15, IL4+7 or without exogenous cytokines. Cell expansion was evaluated after 9–11 days of culture by cell counting using trypan blue exclusion (n = 5). Results are shown as mean cell numbers ± SEM. (a). Panel b CD3⁺ T cell proliferation in the different culture conditions as evaluated by CFSE dilution. M1 shows the percentage of cells that underwent at least seven cell doublings on day 10 after stimulation. Bulk cultures were analyzed for T and NK-cell marker expression on day 10 after activation. Mean expression ± SEM in CTL lines generated from five donors are shown in c. (d) cytokine production from CD3/CD4⁺ (helper) and CD3/CD8⁺ (cytotoxic) CTLs on day 9 after initiation in one representative donor (dot plots shown were gated on CD3⁺ cells). Summary intracellular cytokine production results from three donors (mean ± SD) are shown in (e). Finally the cytokine production profile of pp65-specific CTL initiated with or without cytokines was evaluated by multiplex assay using supernatant harvested 18 hours after antigenic restimulation (n = 4). Th1 cytokines are shown in the left panel while prototypic Th2 cytokines are shown in the right panel (f). Presence of regulatory T cells were evaluated by FoxP3 staining. Plots shown are gated on CD3⁺/CD4⁺ CTLs (g).

Figure 2

Peptide-stimulated and plasmid-activated cytotoxic T lymphocytes (CTLs) share similar phenotypic and functional characteristics. (a) CTLs were stimulated either directly with a pp65 pepmix or using DCs nucleofected with a DNA plasmid encoding the same antigen. Cell expansion was evaluated by counting using trypan blue exclusion (n = 4). (b) The expression of cell surface markers (average ± SD expression) on CTLs 11 days after stimulation (n = 4). The breadth of T cell reactivity in plasmid and pepmix-activated pp65-specific CTLs was evaluated by IFNγ ELISpot on day 9 using a total of 22 mini peptide pools representing all pp65 peptides. Data were normalized to 100% for maximum number of SFC per 1 × 10⁵ CTL. Data from 3 donors screened is shown in (c). (d) The T-cell receptor (TCR) avidity of plasmid versus pepmix-activated CTL generated from 2 representative donors. To assess avidity pp65-CTLs were stimulated bulk CTL cultures with serial dilutions of pp65 pepmix (pp65) or relevant (HLA-matched) epitope peptides (NLV, QAD). IFNγ release of stimulated CTLs was evaluated by ELISpot assay and maximum SFC/1 × 10⁵ cells was normalized to 100% for comparison purposes.

Figure 3

Peptide length does not affect breadth of reactivity. (a) A schematic of three peptide libraries spanning a portion of Adv-Hexon which were used for CTL initiation. Peptide libraries consisted of 15aa, 20aa, or 30aa peptides covering the immunogenic C-terminal 414aa of Adv-Hexon. (b) Phenotypic analysis of cytotoxic T lymphocytes (CTLs) performed on day 10 after stimulation (n = 6). Results are shown as mean ± SEM. Breadth of reactivity was tested using interferon γ (IFNγ) ELISpot as a readout, with the 15mer Hexon overlapping peptide library divided into minipools such that each pool contained 5–6 contiguous peptides, as a stimulus.

Figure 4

Pepmix-activated trivirus-specific CTL lines show similar specificity to plasmid-activated T cells. CTL lines were generated using DCs nucleofected with DNA plasmids encoding EBNA1, LMP2, BZLF1 (EBV), Hexon, Penton (adenovirus), IE-1, and pp65 [cytomegalovirus (CMV)] or direct peripheral blood mononuclear cells (PBMC) stimulation with the corresponding pepmixes. Specificity was determined 10 days after initiation with interferon γ (IFNγ) ELISpot as readout. Results are expressed as SFC/1 × 10⁵ input cells. Control was IFNγ release in response to stimulation with irrelevant pepmix.

Figure 5

Generation of multivirus-specific cytotoxic T lymphocytes (CTLs). (a) A schematic of antigen pooling strategy for CTL initiation. Peripheral blood mononuclear cells (PBMCs) were stimulated with pepmixes pooled by virus (condition A), divided into subpools—immunodominant and subdominant (condition B), divided into subpools encompassing antigens from latent or lytic viruses (condition C), and finally all antigens were pooled together in a mastermix (condition D). After activation PBMCs were pooled and transferred to the G-Rex10 (15 × 10⁶/G-Rex). After 10 days the specificity of the CTL lines generated using these 4 pooling strategies were analyzed using interferon γ (IFNγ) ELISpot assay as readout and individual pepmixes as a stimulus. Results from two representative donors are presented in (b) showing no difference in the specificity of lines. (c) Confirms that multivirus CTL can be reproducibly generated by pooling all pepmixes into one mastermix for activation (n = 8). Results are expressed as SFC/1 × 10⁵ input cells ± SEM. Control was IFNγ release in response to stimulation with an irrelevant pepmix. Antigen specificity of CD3/CD8⁺ (cytotoxic) and CD3⁺CD8- (helper) T cells was evaluated by intracellular IFNγ staining after overnight stimulation with the equivalent antigens. Results from one representative donor are shown in (d). (e) The lines are polyfunctional as assessed using intracellular cytokine staining (ICS) for IFNγ and TNFα in one representative donor.

Figure 6

Multivirus-specific cytotoxic T lymphocytes (CTLs) can be expanded in vitro. On day 9 after initial stimulation CTLs were restimulated using pepmix-pulsed PHA blasts. (a) shows the expansion of CTLs from initiation (day 0) to day 16, following a 2nd stimulation on day 9/10 (n = 4). CTL expansion was evaluated using trypan blue exclusion and results are shown as mean cell numbers ± SD. (b) Results from 1 representative donor illustrating the antigen specificity of CD3/CD8⁺ and CD3/CD8^– (CD4⁺) CTLs after the 2nd round of stimulation using interferon γ (IFNγ) intracellular cytokine staining (ICS). (c) Summary results from six donors after the 1st (day 9) and 2nd (day 16) stimulation, using IFNγ ELISpot as a readout. Results are expressed as SFC/1 × 10⁵ input cells ± SD and the control was IFNγ release in response to stimulation with irrelevant pepmix. The cytotoxic abilities of the generated CTLs were evaluated by standard 4–6 hours Cr⁵¹ release assay using pepmix-pulsed PHA blasts as targets. Specific lysis after the 1st and 2nd stimulation from two representative donors are shown in (d).