Viral infections caused by human adenovirus (HAdV) or CMV remain life-threatening complications in immunocompromised patients undergoing allogeneic hematopoietic stem cell transplantation. Adoptive immunotherapy with virus-specific T cells showed impressive clinical results without or with only mild GvHD. However, because of high costs and high regulatory barriers, these protocols are accessible to only a few centers. The infusion of unmanipulated donor lymphocytes (DLIs) that contain virus-specific T cells is not feasible because of the risk of GvHD. Reports about three patients treated with irradiated granulocytes or DLIs that potentially comprised virus-specific T cells discussed an active role of virus-specific lymphocytes despite irradiation, but real evidence could not be provided. Therefore, we tested the effect of irradiation on HAdV-specific T cells, which had been expanded in vitro, by stimulating PBMCs with HAdV-peptide pools and IL-15 for 12 days. Cells were then irradiated with 30 Gy, as performed for normal granulocyte concentrates. Cell viability and polyfunctional activity were determined by flow cytometry. Even 48 h after irradiation, 15.6% of expanded HAdV-specific T cells were apparently viable and cytolytically active. Although the in vivo antiviral activity was not tested, these data support earlier assumptions about the potential role of irradiated cells in patients.
Allogeneic hematopoietic stem cell transplantation (HSCT) is the only treatment option for several hematological diseases.1 Severe infections in the context of delayed immune reconstitution, due to ex-vivo or in-vivo T-cell depletion or to GvHD, do, however, pose life-threatening complications.2,3
For the treatment of bacterial and fungal infections of pediatric patients after HSCT, leukocyte/granulocyte transfusions from G-CSF-mobilized healthy donors (mostly relatives) have been successfully used after irradiation with 25–50 Gy,4,5 and they showed evidence of clinical benefit in adults6 and children.7 Irradiation is necessary to prevent GvHD, which could be induced by potentially alloreactive T lymphocytes present within the granulocyte concentrates (GCs).
In a report from 1998, the authors described the case of a CMV-seronegative patient who acquired CMV after HSCT.8 He was cured after treatment with seven transfusions of irradiated GCs donated by a CMV-seropositive relative. Flow cytometric analyses revealed that he had received a median of 3–5 × 108 CD3+ T cells/kg/day, and the authors discussed the potentially curative role of virus-specific donor T cells. In another case report from 1999, two patients who developed EBV-lymphoproliferative disease after T-cell-depleted mismatched HSCT were successfully treated with a combination of rituximab (a chimeric monoclonal antibody against CD20) and 1–3 transfusions of 25 Gy-irradiated donor-derived lymphocytes (1 × 10 E6 CD3+ T cells/kg/infusion).9 In the first patient who showed limited response to rituximab due to CD20-negative EBV-lymphoproliferative disease, treatment with irradiated DLIs resulted in a dramatic reduction of EBV-DNA titer further supporting the potential impact of irradiated DLIs.9 New analytical tools like tetramers or the IFN-γ cytokine secretion assay (CSA) were later used to describe a strong correlation between the CMV-IgG serostatus and the absence or presence of CMV-specific T cells analyzed in peripheral blood.10,11 From the above authors, however, appropriate detection methods to address CMV- or EBV-specific T cells had not yet been available. Therefore, definite evidence that irradiated virus-specific T cells preserve functional activity has been missing.
Novel very sensitive detection methods combined with our recently developed short-term expansion protocol12 prompted us to test the effector functions of very rare human adenovirus (HAdV)-specific T cells after irradiation with 30 Gy, as performed for normal GCs. We were able to show that a considerable number of in vitro short-term expanded HAdV-specific T cells (seHAdV-T cells) are able to retain viability and different effector functions for at least 2 days after irradiation. These data provide the first possible rationale for the reduced viral load of patients treated with irradiated blood products containing T lymphocytes. Further clinical trials will be necessary to prove this concept, which, if confirmed, might be a fast and easy therapeutic option to fight viral infections.
Materials and methods
Generation of short-term expanded HAdV-specific T cells (seHAdV-T cells)
Cells from nine healthy donors were obtained upon approval from the local ethics committee of Vienna (EK Nr.514/2011) and after informed consent. Ficoll-isolated PBMCs were used to (i) isolate monocytes (purity 75–95%) via plastic adherence as described13 or via CD14 MicroBead-separation according to the manufacturer’s instructions (Miltenyi Biotec, Bergisch Gladbach, Germay), (ii) generate PHA blasts as described12 and (iii) generate in vitro expanded HAdV-specific T cells. Briefly, 5–10 × 106/mL PBMCs were cultured in AIM-V supplemented with 2% Octaplas (OP, Octapharma, Vienna, Austria), 2 mM L-glutamine and 25 mM HEPES, also referred to as AIM-V+ as described,12 in 96-well plates (for viability assays), 12-well plates (for IFN-γ-CSA and CD107a staining) or 75 cm2 flasks (for the cytotoxicity assay), and stimulated with HAdV (AdV5) PepTivator (Miltenyi Biotec; final concentration 0.6 nmol per ml for each pepide). On day 6, cultured cells were added to adherent monocytes and re-stimulated with PepTivator. In addition, the cells were either cultured without cytokines or with IL-2 orIL-15 (5 ng/mL; R&D Systems, Minneapolis, MN, USA) on days 3 and 9. On day 12, seHAdV-T cells were harvested and used for the analyses described below.
Functional and cytolytic activity assays by flow cytometry
First, autologous monocytes obtained by CD14-positive selection (see above) were pulsed overnight using the HAdV-peptides. For the IFN-γ-CSA, intracellular staining and CD107a staining, 8 × 105 irradiated (30 Gy) and non-irradiated seHAdV-T cells were washed, resuspended in 100 μL AIM-V+, mixed with pulsed monocytes at a ratio of 5:1, and stimulated with PepTivator for another 4 h. The IFN-γ-CSA was performed according to the manufacturer’s instructions (Miltenyi). For the intracellular staining of IFN-γ, TNF-α and Bcl-2 (BD Biosciences, San Diego, CA, USA), cells were stimulated for 4 h with the HAdV-PepTivator and stained according to the manufacturer's instructions (eBioscience, San Diego, CA, USA). Cells were then labeled with LIVE/DEAD Fixable Aqua Dead Cell Stain (Invitrogen, Carlsbad, CA, USA) to distinguish between viable and non-viable cells, and surface staining was performed as described previously, but including CD107a.12 Samples without stimulation served as negative controls, and cells stimulated with 1 μg/mL of Staphylococcal enterotoxin B (Sigma-Aldrich, St Louis, MO, USA) served as positive controls (data not shown). All antibodies and HAdV-specific MHC class I streptamers applied for flow cytometry are described in Geyeregger et al.12
The cytolytic activity of seVirus-T cells was assessed by flow cytometry. Carboxyflourescein succinimidyl ester (CFSE)-labeled PHA targets were pulsed with the appropriate viral peptides for 2 h or overnight, respectively. Non-pulsed targets were used as negative controls. Autologous targets (1.25 × 104) were mixed with seVirus-T cells at a ratio 1:20. Four hours after incubation at 37 °C, the cell suspension was transferred to Trucount tubes and stained with DAPI (4′,6-diamidino-2-phenylindole, dilactate; 0.03 μg/mL). The absolute number of late apoptotic/necrotic targets (=CFSE+/DAPI+) was analyzed. For data evaluation, average values from three to six experiments were calculated based on average values of duplicates or triplicates.
Notably, only viable (>90% viability) non-irradiated or irradiated cells, sorted on a FACSAria, were used for the assays. The viability was only assessed on the basis of scatter parameters (forward scatter vs side scatter) as already recommended by Fritsch et al. and Vermes et al.14,15
HLA-typing of healthy blood donors was performed as described Geyeregger et al.12
Student's t-test analyses (for paired samples) were applied to determine the statistical significance values defined as NS: P>0.05; *P0.05; **P0.01; ***P0.001.
Short-term expansion of HAdV-specific T cells
To obtain sufficiently high numbers of seHAdV-T cells, PBMCs from nine healthy donors were expanded with the HAdV-PepTivator and IL-15 as described in the Materials and methods section. Phenotypic characterization revealed that the resulting seHAdV-T-cell fraction comprised mainly CD3+ T cells (mean: 78.6±4.7%), but also minor populations of NK cells (mean: 15.1±1.1%) and B cells (mean: 5.6±2.3%). The percentage values of CD4+ and CD8+ T cells among the leukocytes were 47.3±9.2% and 23.1±5.5%. The number of streptamer+ HAdV-specific T cells/μL was about 2 log increased, from a median of 0.2 cells/μL±0.03 within the starting PBMC population, to a median of 26 cells/μL±7.5 after the expansion period. Representative dot plots of different T-cell populations within seAdV-T cells (Figure 1a) and the donor-to-donor variability of streptamer+ T cells/μL (Figure 1b) are shown.
Impact of irradiation on the viability of HAdV-specific T cells
After expansion and subsequent irradiation, seHAdV-T cells were fluorescence-activated cell-sorted to obtain >90% viable cells. This was an important step to improve both the outcome of functional assays and the differentiation between non-irradiated control samples (containing low percentages of apoptotic cells) and irradiated samples (containing high percentages of apoptotic cells). These cells were then used as starting material to test the impact of irradiation on the viability of seHAdV-T cells. Untreated controls and cells irradiated with 30 Gy (a dosage generally used in the clinical routine and shown to be safe) were cultured for 24, 48, and 72 h. The viability of cells was assessed by their side scatter-A and forward scatter-A properties via flow cytometry, using Trucount tubes. This was important to avoid the potentially toxic influence of apoptotic dyes on the function of sorted cells. Non-irradiated and irradiated cells after culture were quantified as cells/μL based on the single-platform enumeration technique, and expressed as percentage of the non-irradiated starting cell population (Figure 1c). Generally, the viability of non-irradiated T cells after culture was slightly lower than that of the starting population, with 83.4%, 81.5% and 69% after 24 h, 48 h and 72 h, respectively (Figure 1c). In contrast, viable seHAdV-T-cell numbers were significantly reduced after irradiation, to median values of 39.6% after 24 h and 15.6% after 48 h. After 72 h, we still found 9.4% viable seAdV-T cells compared with the non-irradiated starting population (0 h). When compared with the respective time-matched controls, the recovery of viable cells in the irradiation groups was higher, with 47.6% after 24 h, 19.1% after 48 h and 13.6% after 72 h (Figure 1c). These data show for the first time that, even 3 days after irradiation, reasonable numbers of viable seHAdV-T cells are still present.
Effect of irradiation on the functionality of seHAdV-specific T cells
Using flow cytometry, we tested the capacity of irradiated seHAdV-T cells to express the activation marker IFN-γ and the cytotoxic marker CD107a.16 Again, non-irradiated and irradiated seHAdV-T cells (>90% viability) were cultured for 24 h before they were sorted and re-stimulated for 4 h with the HAdV-peptide pool-pulsed CD14+ monocytes obtained from magnetic selection. Non-irradiated cells cultured under identical conditions served as positive controls. After re-stimulation, high percentage values of IFN-γ-secreting CD8+ and CD4+ T cells were found in both the non-irradiated (CD8+: 14.7±4.8%, CD4+: 16.4±7%) and irradiated (CD8+: 12±3.2%, CD4+: 17.5±5.9%) cell cultures (Figures 2a and b). These results indicate that the IFN-γ secretion by CD4+ and CD8+ T cells was not significantly affected 24 h after irradiation. To evaluate the cytotoxic potential of irradiated cells, we analyzed the concomitant expression of CD107a by CD8+ T cells. We found a highly induced expression of CD107a in both the non-irradiated (7.1%±2.4) and the irradiated (3.9%±1.5) CD8+ HAdV-specific T cells (Figure 2c).
To determine whether the resistance to irradiation was a consequence of IL-15 supplementation (known to have anti-apoptotic effects17), we additionally expanded seHAdV-T cells in the absence of cytokines and with IL-2. Twenty-four hours after irradiation, the number of residual viable cells and the antigen-induced production of IFN-γ, TNF-α and CD107a of differentially cultured seHAdV-T cells were analyzed by intracellular cytokine staining, which was combined with a live/dead marker and with Bcl-2, a pro-survival protein. Although before irradiation the total cell number of streptamer+ HAdV-specific T cells/μL was highest in IL-15-supplemented cultures (data not shown), the median absolute number of viable cells per well was not significantly different between cultures with or without cytokines (no cytokine ( × 103): 13.3±1.8, IL-2 ( × 103): 14.3±4.1, IL-15 ( × 103): 23.1±10.2). Furthermore, we could show that survival was directly associated with the expression of Bcl-2 (Figure 3a). Our data show that the percentage of IFN-γ (no cytokine (-IL): 23.3±3.7, IL-15: 26.1±10, IL-2: 33±11.5), TNF-α (no cytokine (-IL): 18.4±4, IL-15: 19.6±8.1, IL-2: 27.1±11.1) and CD107a (no cytokine (-IL): 21.3±7.9, IL-15: 19.2±11.5, IL-2: 29.8±1.9) producing seHAdV-T cells did not differ significantly between the various cell culture conditions (Figure 3b). In addition, we could show that CD3+ seHAdV-T cells were polyfunctional, producing both IFN-γ and TNF-α, which is known to be associated with superior in vivo activity18 (Figure 3b). Polyfunctionality was also shown for CD4+ and CD8+ T-cell fractions, independent of the cytokines supplemented (data not shown).
Taken together, these results clearly indicate that seHAdV-T cells, even 24 h after irradiation, are still polyfunctional and cytolytically active, irrespective of whether or not they had been cultured in the presence of IL-15.
To examine the capability of irradiated seHAdV-T cells to lyseviral-peptide-pulsed target cells, a flow cytometry-based cytotoxicity assay was performed as described elsewhere.12 Irradiated seHAdV-T cells were cultured for 6, 24 and 48 h before they were viability-sorted (>90%), and incubated for 4 h, at a ratio of 20:1, with either non-pulsed (negative control) or appropriate HLA-dependent HAdV-peptide-pulsed autologous target cells. The number of late apoptotic/necrotic target cells or so called ‘dying’ cells/well was determined by flow cytometry. Lysis of peptide-pulsed target cells by non-irradated seHAdV-T cells was about 3.4-fold increased compared with that of non-pulsed control target cells (Figure 4). When using irradiated seHAdV-T cells, the results of target-cell lysis were very similar (3.4-fold), when the experiment was started 6 h after irradiation. When starting after 24 h, the specific lysis of HAdV-peptide-pulsed target cells was still 2.1-fold higher than in the controls. Even after 48 h, we found a median 1.9-fold difference in three experiments, although—because of high variability—the mean values were not significant (Figure 4). To our knowledge, this is the first report showing that T cells retain their capacity to lyse peptide-pulsed target cells even 24 h after irradiation.
Fatal HAdV disease (reported in 13–50% of infected patients) is associated with transplant-related mortality rates of 2–6% in high-risk patients.19,20 Despite impressive clinical results showing total or partial clearance of viral infections,21, 22, 23, 24, 25 the risk of GvHD can never be excluded. Furthermore, a cost-intensive good manufacturing practice facility including a lot of paper work is indispensable for such advanced therapeutic medicinal products26 even if rapidly generated virus-specific T cells are used.12,25,27 The infusion of unmanipulated donor lymphocytes to fight viral infections would be easy and inexpensive, but is of course unacceptable because of the high risk of GvHD induction in patients.28 The assumption that T cells within irradiated blood products from three G-CSF-stimulated allogeneic third party donors, such as GCs or DLIs, might have contributed to CMV and EBV clearance,8,9 prompted us to test the direct effect of gamma-irradiation on seHAdV-T cells.
We first demonstrated that a mean of 9% of in vitro generated HAdV-specific T cells were still viable 72 h after irradiation. This confirms data reported by Waller et al.29 who showed that, even 2–3 days after 20 Gy irradiation, the majority of T cells were still viable, although their proliferative capacity was completely blocked. It was shown that virus-specific memory T cells rapidly (within few hours) acquire cytotoxic activity against adoptively transferred target cells in vivo.30 In a very recent study, T cells, which were transiently redirected (for 2–3 days) with a TCR against hepatitis B virus by mRNA electroporation, showed efficient prevention of tumor seeding in a xenograft model of hepatocellular carcinoma.31 Brodie et al.32 showed that, within 3 days, HIV-specific T cells were sufficient enough to localize at sites of viral replication. These data support the assumption that there might be sufficient time—even for irradiated T cells—to fulfill their task before undergoing apoptosis. This is further supported by a study by Waller et al.,29 where at least 3 out of 12 patients showed objective responses against their leukemia or lymphoma within 2 weeks, after six infusions of allogeneic donor lymphocytes irradiated with 7.5–30 Gy.
A hallmark of T cells is their ability to secrete IFN-γ and TNF-α and to lyse virus-infected target cells. Strikingly, re-stimulation of HAdV-specific T cells 24 h after irradiation showed secretion characteristics of IFN-γ (via the IFN-γ CSA) similar to non-irradiated cells of both cytotoxic CD8+ T cells and CD4+ helper T cells. Also the expression of CD107a, representing the cytotoxic potential of seHAdV-T cells, was not significantly decreased in CD8+ T cells after irradiation.
Next, we analyzed whether IL-15 (known to support anti-apoptotic effects), which was supplemented to the cell culture, was responsible for the moderate effects of irradiation. Therefore, seHAdV-T cells were cultured without cytokines or with IL-2 and IL-15 followed by an intracellular cytokine staining. Interestingly, neither the number of viable cells nor their ability to express IFN-γ, TNF-α and CD107a differed significantly between differently cultured and irradiated seAdV-T cells. This could be partially explained by the rather short-term treatment of cell cultures with cytokines (6 days instead of 12 days). Of note, the irradiation-dependent reduction of absolute numbers of seHAdV-T cells compared with non-irradated cells could explain the observed higher percentage values of irradiated cells producing IFN-γ, TNF-α and CD107a as compared with non-irradiated cells. These results—together with the finding that irradiated seHAdV-T cells lyse autologous HAdV-peptide-pulsed target cells, which are only loaded with a single HLA-type-dependent epitope—indicate that, even 48 h after irradiation, some cells have retained their functional activity. This is in accordance with other data showing preserved effector function against, for example, Chlamydial pneumonia in a mouse model33 and cytolytic activity against ALL targets, after irradiation with 2.5–40 Gy.34 To our knowledge, this is the first report describing cytolytic activity 24 and 48 h after irradiation, instead of only 1 h as reported by others.34
In contrast to irradiated GCs and DLIs, which, if at all, only contain non-stimulated virus-specific T cells, we only used in vitro expanded and therefore stimulated T cells. Several studies showed that T cells stimulated with either IL-235 or PHA36 were less radio-sensitive than non-stimulated T cells. Whether irradiation has different effects on either freshly isolated or in vitro expanded HAdV-specific T cells was not tested in our study. Because irradiated cells cannot multiply and their life span is reduced to a few days only, multiple doses of irradiated T cells would be necessary to cure viral infections. This is supported by several other studies in mice33 and humans.8,9
In the clinical situation, the median time span between the detection of HAdV in the stools of patients after HSCT, and the onset of disseminated disease, is approximately 21 days.37 Although immediate antiviral treatment is generally preferred,38 it is still a matter of debate whether or not the amount of viral load is influencing the efficacy of immunotherapy. Another challenge is that matched third party donors are not always available. Therefore, preemptive or early treatment on demand, with HAdV-specific T cells, seems recommendable for patients with early viremia. Partially matched third party healthy relatives might be a source of HAdV-specific T cells infused after irradiation. In the present work and in a recent study, we showed that in vitro short-term expanded HAdV-specific T cells are able to kill mismatched HAdV-peptide-loaded target cells, even if they are only matched in a single HLA-type (see above and references Geyeregger et al.,12 and Sellar and Peggs25). This would highly increase the chance to find an appropriate third party donor for such adoptive immunotherapies. As recently shown by others, it is feasibile to adoptively transfer virus-specific T cells from third party donors.22,23
The number of circulating virus-specific T cells is very low, ranging from 0.01 to 1% of CD3+ T cells.12 However, this is compensated by the high total number of T cells collected in an apheresis product (approximately 5 × 109), which should contain at least 105 virus-specific T cells. In their communication,8 Witt et al. reported the infusion of 2.8 × 109 CD3+ T cells, which is in fact close to the above assumption.
Our findings support the suggestion that virus-specific T cells remain functional after irradiation. Nevertheless, we are aware that only a clinical trial has the potential to prove whether multiple doses of irradiated virus-specific T-cell-containing blood products could possibly bridge the immunosuppressive period after HSCT.
We thank M Zavadil for carefully proofreading and language editing the manuscript. The work was funded by the Children’s Cancer Research Institute (CCRI).