High-intensity interval training in allogeneic adoptive T-cell immunotherapy – a big HIT?

Background Adoptive transfer of virus-specific T cells (VSTs) represents a prophylactic and curative approach for opportunistic viral infections and reactivations after transplantation. However, inadequate frequencies of circulating memory VSTs in the T-cell donor’s peripheral blood often result in insufficient enrichment efficiency and purity of the final T-cell product, limiting the effectiveness of this approach. Methods This pilot study was designed as a cross-over trial and compared the effect of a single bout (30 min) of high-intensity interval training (HIT) with that of 30 min of continuous exercise (CONT) on the frequency and function of circulating donor VSTs. To this end, we used established immunoassays to examine the donors’ cellular immune status, in particular, with respect to the frequency and specific characteristics of VSTs restricted against Cytomegalovirus (CMV)-, Epstein–Barr-Virus (EBV)- and Adenovirus (AdV)-derived antigens. T-cell function, phenotype, activation and proliferation were examined at different time points before and after exercise to identify the most suitable time for T-cell donation. The clinical applicability was determined by small-scale T-cell enrichment using interferon- (IFN-) γ cytokine secretion assay and virus-derived overlapping peptide pools. Results HIT proved to be the most effective exercise program with up to fivefold higher VST response. In general, donors with a moderate fitness level had higher starting and post-exercise frequencies of VSTs than highly fit donors, who showed significantly lower post-exercise increases in VST frequencies. Both exercise programs boosted the number of VSTs against less immunodominant antigens, specifically CMV (IE-1), EBV (EBNA-1) and AdV (Hexon, Penton), compared to VSTs against immunodominant antigens with higher memory T-cell frequencies. Conclusion This study demonstrates that exercise before T-cell donation has a beneficial effect on the donor’s cellular immunity with respect to the proportion of circulating functionally active VSTs. We conclude that a single bout of HIT exercise 24 h before T-cell donation can significantly improve manufacturing of clinically applicable VSTs. This simple and economical adjuvant treatment proved to be especially efficient in enhancing virus-specific memory T cells with low precursor frequencies.


Background
Latent herpes viruses such as Cytomegalovirus (CMV) and Epstein-Barr virus (EBV) as well as lytic viruses such as Adenovirus (AdV) remain an important cause of morbidity and mortality after hematopoietic stem cell transplantation (HSCT) and solid organ transplantation (SOT) in immunocompromised patients [1][2][3].
With a median of 3 month after HSCT, 11% of all deaths are induced by viral infections with about one-third are caused mainly by CMV (up to 20%), EBV (up to 50%) and AdV (up to 75%) [4][5][6][7][8][9][10]. With an incidence of 6 to 28%, AdV infections are the most common infectious complications after HSCT, especially in children [2,5] Viral complications also influence the outcome of SOT [11] and in general, the incidence varies depending on the transplanted organ, the resulting immunosuppressive treatment, the serostatus of the donor and recipient, and the type of antiviral prophylaxis [12]. Due to high immunosuppression reactivation of EBV and/or outgrowth of latently EBV-infected cells can result in the development of malignant lymphoma (post-transplant lymphoproliferative disease; PTLD), which represents the most common EBV-associated malignancy after transplantation [13].
The pharmacotherapy of viral complications is often limited by both lack of efficacy and toxic side effects. Adoptive transfer of virus-specific T cells has become an attractive option for second line treatment [5,[14][15][16][17]. It is established that the delay in the reconstitution of virusspecific CD8 + and/or CD4 + T-cell responses is a critical factor in viral recrudescence and viral disease and that the permanent control and elimination of a virus depends on the presence of functional virus-specific T cells (VSTs) [18,19]. The adoptive transfer of donor-derived memory VSTs with defined specificity has developed into a prophylactic and/or curative approach for common post-HSCT and post-SOT viral infections without a significant risk of toxicity or GvHD [2,3,[20][21][22][23][24][25][26]. Recently, we demonstrated the successful use of EBV-specific T cells as a consolidating treatment for a patient suffering from EBVassociated PTLD of the central nervous system [27].
Methods to increase the number and function of VSTs in healthy donors are an attractive way to ensure the success of adoptive T-cell therapy. As lymphocytosis occurs during and directly after exercise [37], pre-donation exercise could be a time-and cost-efficient way to increase the starting frequencies of circulating peripheral memory VSTs in the donor before apheresis [38][39][40]. Especially single bouts of exercise evoke increased hemodynamics and the release of catecholamines and glucocorticoids, resulting in a striking leukocytosis [37]. Simultaneously, a redistribution of effector cells between the blood compartment and the lymphoid and peripheral tissues occurs [37]. On the other hand, longer periods of exhaustive exercise can have the adverse effect and weaken immunity [37]. Recent studies have shown that a single bout of continuous exercise (CONT) enhances the quantity of VSTs when they are stimulated and expanded over 8 days [39]. In the study of Kunz et al. they clearly demonstrated that single bouts of exercise can mobilize AdV-specific T cells and increase their function [41]. There is a growing body of evidence showing that high-intensity interval training (HIT) is an effective alternative to traditional endurance training. Alongside similar or even superior effects on many health-related markers (e.g. cardiovascular risk factors), lymphocyte numbers were shown to increase directly after HIT in sedentary men [38-40, 42, 43].
The present study aimed to investigate the benefit of CONT and HIT for adoptive antiviral T-cell therapy. We compared the impact of a 30 min single bout of HIT versus 30 min of CONT on the cellular immune status of healthy donors, in particular, regarding the total number, functionality and proliferative capacity of VSTs. These parameters were analysed in blood samples collected at different time points before and after exercise to identify the time point with the most effective induction level of functional active VSTs and therefore to define the optimal time point for T-cell donation. In summary, we found that both training protocols boost the number and function of VSTs in healthy donors and that the highest response was observed in VSTs against CMV, EBV and AdV with low precursor frequencies in the peripheral blood and in donors with a moderate fitness level who performed HIT 24 h prior to donation.

Study population
After approval by the Institutional Review Board of Hannover Medical School (MHH, approval number 3366-2016), 12 healthy adult volunteers (six females and six males with a mean age of 24 years; range: 20-31 years) were recruited into this study. Their demographic characteristics are presented in Table 1. To be eligible  [44] for classification of fitness level. All 12 donors had a moderate (n = 8) or a high (n = 4) fitness level.

Exercise protocols and blood sampling
Participants were asked to refrain from exercise 24 h prior to each laboratory visit. The individual exercise capacity as measured by maximum power output (P max ) was determined with a graded exercise test (GXT) on a cycle ergometer (Ergoline, Bitz, Germany) 4 weeks prior to testing. GXT started at 50 W, the workload increased by 16.7 W/min until physical exhaustion. Heart rate (HR) was measured continuously with a 12-channel ECG (GE Electric, Boston USA). Earlobes were hyperaemised at least for 3 min before the first sample was drawn. After 50 W, every 3 min afterwards and at exhaustion capillary blood drawn into 20 µl end-to-end capillaries was used to analyse glucose and lactate concentrations within the next 2 h. Analyses were performed with the Biosen S line analyzer (EKF Diagnostics, Barleben, Germany) including the glucose/lactate test set-up with haemolysing, Glucose/Lactate system, multi standard and Ready-Con test solutions (EKF Diagnostics). On two further occasions separated by a wash-out phase of 14 days, all 12 participants visited the Institute of Sports Medicine at MHH. They were assigned to two equally-sized groups based on the availability of the participants. Blinding was not deemed possible. Each donor executed the two exercise protocols in crossover fashion (Table 1). At the first visit (week 1), one group (n = 6, group 1) performed a single bout of continuous exercise (CONT) at 50% P max for 30 min, while the other group (n = 6, group 2) performed a single bout of high-intensity interval (HIT, more precisely High Intensity High Volume Training, HIHVT) training for 20 min of 30-s 100% P max intervals each followed by a 30-s pause 5 min before and after HIT exercises at 50% P max . At the second visit (week 3), each group switched over to the other exercise protocol. Heart rate was measured continuously throughout each visit with a 3-channel ECG (Ergoline, Bitz, Germany), and earlobe capillary blood samples for lactate and glucose testing were drawn before, during and after exercise. Intravenous blood samples were collected from a cubital or antecubital vein before and directly after exercise as well as 1 h and 24 h after exercise ( Fig. 1). Complete blood counts and cellular immune status tests were performed immediately on whole blood samples. Peripheral blood mononuclear cells (PBMCs) were isolated by discontinuous density gradient centrifugation and examined for the specified VST characteristics using appropriate in vitro immunoassays. Representative flow cytometric plots of the cellular immune status (A), VST enrichment efficiency (B) and T-cell activation using CD69 as a specific marker (C) are visualized in Additional file 1.

Serological testing by ELISA and Western blot
All donors were pretested for CMV, EBV and AdV serostatus as described previously using commercially available IgG Western blot and IgG ELISA kits [45,46]. Briefly, CMV and EBV serology was performed by confirmatory Western blot tests designed to quantitatively determine anti-CMV or anti-EBV IgG antibodies against major CMV and EBV proteins (recomLine CMV or EBV IgG, Mikrogen, Neuried, Germany). IgG antibodies

Table 1 Physical and fitness characteristics of the participants
Each donor executed the two exercise protocols in crossover fashion. At the first visit (week 1), one group (n = 6, group 1) performed a single bout of continuous exercise (CONT) at 50% P max for 30 min, while the other group (n = 6, group 2) performed a single bout of high-intensity interval training (20 min of 30-s 100% P max intervals each followed by a 30-s pause 5 min before and after HIT exercises at 50% P max , more precisely High Intensity High Volume Training, HIHVT). At the second visit (week 3), each group switched over to the other exercise protocol. In other words, group 1 did continuous exercise during the first visit now and did HIT exercise on the second (CONT/HIT), and group 2 vice versa (HIT/CONT)

Determination of the frequencies of antiviral memory T cells by enzyme-linked immunospot assay
Detection of virus-specific IFN-γ-producing T-lymphocytes was achieved by IFN-γ EliSpot assay as previously described. Briefly, PBMCs were isolated from whole blood samples by discontinuous density gradient centrifugation, resuspended in culture medium (CM) consisting of RPMI1640 (Lonza, Vervies, Belgium) with 10% human AB serum (C.C.pro, Oberdorla, Germany) at a concentration of 1 × 10 7 cells/ml, seeded on 24-well plates and rested overnight. Rested PBMCs were cocultured in anti-IFN-γ precoated EliSpot plates (Lophius Biosciences, Regensburg, Germany) for 16-18 h at a density of 2.5 × 10 5 cells/well with specific antigens of interest. For stimulation CMV-derived peptide pools CMV pp65 and CMV IE-1, EBV-derived peptide pools EBV EBNA1 and EBV Consensus and AdV-derived peptide pools AdV5 Hexon and AdV5 Penton were used at a final concentration of 1 µg of each peptide/ml peptide pool (Miltenyi Biotec, Bergisch Gladbach, Germany). Cells stimulated with staphylococcal enterotoxin B (1 µg/ ml, SEB, Merck, Taufkirchen, Germany) served as the positive control and PBMCs incubated in media alone as the negative control. IFN-γ secretion was detected using streptavidin-alkaline phosphatase (Mabtech Stockholm, Sweden) and revealed by 5-13 bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT Liquid Substrate, Merck). Spots were counted using AID EliSpot 8.0 on an AID iSpot spectrum reader system (both from AID, Strassberg, Germany). Findings are indicated as number of spots per well (spw), representing the number of spots in the antigen well after subtracting the respective negative control well. Results were furthermore indicates as spw/1000 CD3 + T cells.

IFN-γ Cytokine Secretion Assay
The IFN-γ Cytokine Secretion Assays (CSA, Miltenyi Biotec) was performed as previously described [47]. After overnight resting, 1 × 10 7 isolated PBMCs were stimulated with the CMV peptide pools CMV pp65 or CMV IE1, EBV peptide pools EBV EBNA1 or EBV Consensus, and AdV peptide pools AdV5 Hexon or AdV5 Penton, respectively, at a final concentration of 1 µg of each peptide/ml peptide pool. Unstimulated PBMCs served as the negative control. Activated IFN-γ-secreting T cells were specifically captured during the magnetic cell sorting enrichment process by using anti-IFN-γ-phycoerythrin (PE) antibodies and paramagnetic anti-PE microbeads.

Detection of the cellular immune status by flow cytometry
All flow cytometric analyses were performed using the FACSCanto 10c system (BD Biosciences) and BD FACS-Diva Software version 8.0.1. The cellular immune status was determined by using specific markers for monocytes, T cells, B cells and natural killer (NK) cells. Briefly, whole blood samples were analysed on a single-cell platform using TruCount tubes (BD Biosciences). Absolute counts of the investigated leukocyte subsets were calculated according to the manufacturer's instructions. Cells were stained for 20 min at room temperature using anti-CD45 APC-H7, anti-CD3 FITC, CD8 APC, anti-CD4 peridinin chlorophyll protein (PerCP), anti-CD19 PE-Cy7, anti-CD56 PE and anti-CD14 BV510 mAbs (all from BD Biosciences) before the lysis of erythrocytes using 1× BD FACS Lysing Solution according to the manufacturer's instructions (BD Biosciences). The number of spots per 1000 CD3 + T cells, as determined by IFN-γ EliSpot assay, was calculated based on the CD3 frequencies determined by staining PBMCs with anti-CD45 APC-H7, anti-CD3 FITC, anti-CD8 APC and anti-CD4 PerCP).

Real-time PCR detection of HSP70, Ki67, IFN-γ, and granzyme B expression measurement of telomere length
The mRNA levels of heat shock protein 70 (HSP70), Ki67, IFN-γ, and granzyme B were analysed as previously described [48]. Briefly, total cellular RNA was isolated (RNeasy Mini Kit; Qiagen, Hilden, Germany), and cDNA was amplified using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Darmstadt, Germany). Quantification of HSP70, Ki67, IFN-γ, and granzyme B mRNA levels was achieved by using inventoried mixes (Applied Biosystems). TaqMan Gene Expression Master Mix (Applied Biosystems) was used for amplification. The constitutively expressed glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene served as the reference gene.
As an indicator of telomere length, p16 expression was analysed. Quantitative PCR of p16INK4a was performed using the following intron-spanning primers and probes: p16INK4a forward 5′-GGG CAC TGC TGG AAGCC-3′, reverse 5′-AAC GTT GCC CAT CAT CAT C-3′, and probe 5′-CCG AAC TCT TTC GGT CGT A-3′. For analysis of telomere length, genomic DNA was first extracted from isolated PBMCs using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). DNA quantification was then performed on an Infinite M200 Pro microplate reader (Tecan, Maennedorf, Switzerland) using SYBR ® Green I Dye (Invitrogen, Carlsbad, CA, USA). Telomere length was calculated as abundance of telomeric template versus a single copy gene (36B4) by quantitative real-time PCR (qPCR) on a 7900HT Fast Real-Time PCR System (Applied Biosystems, Darmstadt, Germany) as previously described [49,50]. Results are expressed as the ratio of the telomere repeat copy number to the single gene copy number (T/S). All measurements were performed in quadruplicate, and all samples were measured on a single plate to reduce inter-assay variability.

Quantification of the total cellular cholesterol level of cytotoxic CD8 + T cells
CD8 + T cells were isolated from blood samples collected at different time points directly before and after exercise (Whole Blood CD8 MicroBeads, Miltenyi Biotec) for quantification of total cellular cholesterol. The cells were fixed with 0.1% glutaraldehyde (Merck, Taufkirchen, Germany) and treated with 2 U/ml cholesterol oxidase (Ther-moFisher) for 15 min to oxidize the plasma membrane cholesterol. After extracting intracellular cholesterol with methanol/chloroform (vol/vol, 1:2), total cellular cholesterol was quantified using the Amplex Red Cholesterol Assay Kit (ThermoFisher) according to the manufacturer's instructions. Plasma membrane cholesterol levels were calculated by subtracting intracellular cholesterol from total cellular cholesterol. The purity of isolated CD8 + T cells was assessed by flow cytometry after staining the cells with anti-CD8 APC mAb (BD Biosciences). At least 10,000 events were acquired in the CD8 + T-cell gate. The percentage of activated CD8 + T cells was determined by staining the cells with anti-CD3 PerCP, anti-CD8 APC, and anti-CD69 FITC mAbs (all from BD Biosciences). At least 20,000 events were acquired in the CD3 + T-cell gate for each analysis, while CD69 + T cells were gated on the scatter properties of CD3 + /CD8 + T lymphocytes.

Statistical analysis
This pilot study was carried out as a non-randomized two period two sequences cross-over trial. We did not adjust for multiplicity due to the exploratory nature of this trial and hence, our results are not to be interpreted in a confirmatory sense.
Statistical analysis was performed using the Prism v5.02 software (GraphPad, San Diego, California, USA) and SAS 9.4 (SAS Institute Inc., Cary, NC, USA). The results are displayed as mean ± standard deviation (SD). Data were analysed using the Wilcoxon test to compare each post training measurement with the pre-training measurement separately. Significance levels were calculated and expressed as p-values (*p < 0.05, **p < 0.01, ***p < 0.001). In addition, we carried out calculations to compare the two training methods with the two sample t approach adjusting for period as described in chapter 3 of [51].

Exercise induced lymphocytosis
All donors successfully completed the two 30 min exercise sessions. The effects of both protocols on the standard complete blood count (CBC) and cellular immune status were investigated with respect  18:148 to changes in heart rate, lactate and glucose levels ( Table 2) as well as to leukocyte subsets, erythrocytes and platelets (Table 3, Additional file 2). Heart rate was increased significantly by both programs during and directly after exercise, while glucose levels decreased. The measured lactate levels were higher during exercise and began to decrease again directly after. Both CONT (6.48 × 10 3 /µl ± 1.38 before exercise vs. 8.242 ± 1.50 1 h after exercise, p < 0.01) and HIT (6.72 × 10 3 /µl ± 1.19 before exercise vs. 8.36 × 10 3 /µl ± 2.32 1 h after exercise) increased the total count of leukocytes (white blood cells (WBCs)) in the standard CBC (Table 3, Panel A). All investigated immune cell subsets analysed in blood samples by using the haemocytometer, namely lymphocytes, granulocytes and monocytes (Table 3, Panel A, Additional file 2), indicated an increase of the absolute cell numbers 1 h after exercise, with recovery at least after 24 h. There was no significant difference between both training programs. Concerning the leukocytes, we found a higher increase of granulocytes compared to lymphocytes. This was confirmed by a slight decrease of the absolute lymphocyte count compared to an increase of the absolute leukocytes count ( Table 3, Panel A). Similar results were obtained by single-cell flow cytometric analysis, investigating the frequency and absolute number of CD45 + leukocyte and lymphocyte subsets ( Table 3, Panel B; Additional file 2 and Additional file 3, shown as normalized data). Interestingly, the absolute number of CD3 + T cells was increased continuously up to 24 h after exercise (Additional file 2E, shown as normalized data), while the CD4/CD8 T-cell ratio increased only slightly within 1 h after both training protocols and almost completely returned to pre-exercise levels within 24 h post training (Additional file 2F, shown as normalized data). No significant differences between both training programs were identified with regards to the frequencies of the investigated cellular immune subsets (CD45 + leukocytes, CD45 + lymphocytes, CD14 + monocytes, CD19 + B cells, CD3 + CD56 + NKT cells and CD3 − CD56 + NK dim/bright cells (Additional file 3, shown as normalized data). The frequency of CD14 + monocytes, was increased directly after both exercise programs, while the frequency of CD19 + B cells increased within 1 h after both exercise programs and recovered within 24 h (Additional file 3, shown as normalized data). NK cell subsets displayed an inversely proportional trend (Additional file 3). The frequency of CD3 − / CD56 dim NK cells increased within 1 h after both types of exercise, while the frequency of CD3 − /CD56 bright NK cells decreased. Regarding donor fitness level, highly fit individuals had slightly higher frequencies of lymphocytes, in the resting state (Additional file 4A), while absolute numbers of white blood cells (WBC), lymphocytes, granulocytes and monocytes did not differ between donors with high and moderate fitness levels (Additional file 4B).

Exercise increased HSP70 levels while downregulating proliferation markers and not affecting replicative senescence
Quantitative real-time PCR showed that mRNA expression levels of HSP70 increased more than twofold within 24 h after HIT (2.9-fold), but only slightly after CONT exercise (up to 1.8-fold); the latter increases were detected directly and 1 h post exercise (Additional file 5A). The mRNA expression levels of the proliferation marker Ki67 were lower than baseline at all time points after both types of exercise. The mRNA levels of IFN-γ remained unchanged after CONT and increased slightly within 24 h after HIT. Granzyme B mRNA levels showed the same trend for both exercise programs: they increased slightly directly after exercise and then dropped to levels lower than starting point within 1 h and 24 h after exercise. The latter trend was more noticeable for CONT exercise. Telomeres were studied by quantification of telomere length, telomere ratio and Table 2 Blood glucose, blood lactate, and heart rate in response to exercise Glucose and lactate levels in peripheral blood and heart rate before, during and after exercise (CONT: continuous exercise, HIT: high-intensity interval training). Glucose and Lactate measured periodically, heart rate measured continuously. Asterisks indicate statistically significant differences between before, during and after exercise (*p < 0.05, **p < 0.01 and ***p < 0.001). Data are mean ± SD  p16 expression (Additional file 5B). None of these three parameters changed significantly in terms of both exercise programs and sample time, even though the telomere ratio tended to be lower 24 h after HIT and CONT exercise.

HIT had the biggest impact on the number of functionally active VSTs
The exercise-induced change in the number of functionally active VSTs against the investigated peptide pools derived from CMV, EBV and AdV was measured by IFN-γ EliSpot assay at different time points before and after HIT and CONT (Fig. 2). The positive control showed a positive result for all analysed samples specified as TNTC (too numerous to count, TNTC ≥ 1000 spw). Overall, donors exhibited a higher increase 24 h after HIT than 24 h after CONT exercise, which induced almost no or minimal change. The change in the number of functionally active VSTs, expressed as the number of spots per 1000 CD3 + T cells from before to 24 h after exercise, was 1.2-fold CMV pp65, 2.6-fold CMV IE-1 (Fig. 2a), 1.3-fold EBV EBNA1, 1.2-fold EBV Consensus (Fig. 2b), 1.6-fold AdV5 Hexon, 2.2-fold AdV5 Penton (Fig. 2c) for HIT compared to 1.0-fold CMV pp65, 2.7-fold CMV IE-1, 0.7-fold EBV EBNA1, 0.7fold EBV Consensus 0.7-fold, 0.8-fold AdV5 Hexon and AdV5 Penton for CONT. Even though this trend was strong, it only reached statistical significance with AdV5 Hexon directly after HIT exercise, and a 1.6-fold increase occurred within 24 h (p < 0.05, Fig. 2c).
Comparing the effect of HIT and CONT with the two-sample t approach adjusting for period, there was no significant difference between the programs except for AdV5 Penton and EBV Consensus 24 h after exercise (p = 0.013 and p = 0.006), where HIT had the bigger effect (Additional file 6).
Concerning the effects of the individual donor fitness level on VST activation (Fig. 3), HIT most effectively boosted VSTs against AdV-and EBV-derived antigens in moderately fit donors (AdV: up to 2.5-fold increased; EBV: up to 2.8-fold increased) and was less efficient in augmenting the immunodominant CMV-derived antigen pp65 (2.5-fold increased) than the less immunodominant CMV IE1 (4.5-fold increased).

Proof-of concept: enrichment of antiviral T cells using small scale Cytokine Secretion Assay
Encouraged by the positive effect of HIT on VST activation identified by the IFN-γ EliSpot assay, we also examined its effect on IFN-γ-secreting T cells using the IFN-γ CSA, which revealed a similar outcome (Fig. 4). All antigens except the highly immunodominant antigen CMV pp65 [46] evoked fold increases in CD3 + T cells after short stimulation with the largest increased detected 24 h after exercise. The fold increases at 24 h after exercise compared to baseline were as follows: CD3 + /IFN-γ + : 0.73-fold CMV pp65, 1.8-fold CMV IE-1, 2.5-fold EBV EBNA1, 1.4-fold EBV Consensus, 2.3-fold AdV5 Hexon and 5.3-fold AdV5 Penton.

A single bout of exercise directly affected CD8 + T-cell cholesterol and activation levels
A single bout of CONT or HIT resulted in a steady increase in cholesterol in isolated CD8 + T cells (Fig. 5).
The overall change from before to 24 h after exercise was 1.46-fold (Fig. 5). This change correlated to expression levels of the activation marker CD69 on isolated CD8 + T cells; the total change was 1.66-fold before to 24 h after exercise (Fig. 5b)

Discussion
Research has shown that physical exercise augments the number of circulating VSTs in peripheral blood [41]. The current study demonstrates for the first time that (1) HIT is the preferable type of exercise for this purpose, (2) the optimal time for efficient augmentation is 24 h prior to leukapheresis, and (3) HIT most effectively boosts the proliferation of VSTs with low resting frequencies in moderately fit donors. These results may have important implications for VST production with regard to coaching donors prior to leukapheresis.
Jamurtas et al. [52] recently observed that WBC and lymphocyte counts increase immediately after both continuous exercise and high-intensity interval training, while monocyte counts remain unchanged. The fact that the increase after HIT was significantly greater suggests that the VST yield may also be augmented post interval training. In agreement with our results, the observed leuko-and lymphocytosis returned to baseline within 24 h after training. Walsh et al. reported that the same biphasic pattern of mobilization can be observed for T cells and is mainly due to the direct and indirect effects of epinephrine [40]. It has been shown that a single bout of various types of continuous exercise mobilizes VSTs specific for latent and non-latent herpesviruses (i.e. CMV, EBV and AdV) to the peripheral bloodstream and has a beneficial effect on their in vitro expansion [25,41].
Accordingly, we also observed an increase in leukocyte and lymphocyte counts directly after both exercise protocols, while the number and function of CD3 + T cells was highest 24 h after exercise. This response could be due to the sympathetically mediated elevation of stress levels directly after exercise resulting in higher glucocorticoid production and, therefore, temporary inhibition of cytokine production until recovery [53].
Strikingly, we observed the highest T-cell response after 30 min of HIT against AdV5 Hexon and AdV5 Penton in both IFN-γ EliSpot Assay and IFN-γ CSA, while CMV pp65 did not respond as much to exercise. HIT appears to augment IFN-γ secretion, especially in the case of circulating memory VSTs with low precursor frequencies, like AdV-specific T cells [46]. Conversely, T cells against the highly immunodominant antigen CMV pp65 [54], which have higher frequencies in the circulation of healthy donors, did not respond as much to exercise. This finding suggests that HIT is most beneficial for boosting antigen-specific T cells with low precursor frequencies.
Interestingly, these beneficial effects on mobilization and function were more distinct in moderately fit individuals. It is known that regular prolonged exercise promotes immunosenescence of T cells [55]. This could be (See figure on next page.) Fig. 3 Changes in frequencies of VSTs between moderately and highly fit donors. Donors were additionally classified into a moderately and a highly fit group according to the International Physical Activity Questionnaire (IPAQ). Frequencies of functional-active virus-specific T cells were determined by IFN-γ EliSpot assay. Negative and positive controls were carried out by using either medium or 1 μg/ml staphylococcal enterotoxins B (SEB). Results a in response to CMV, b EBV and c AdV from before and 24 h after exercise are shown as fold changes of spots per well per 1000 CD3 + T cells after subtracting those of the respective negative control. Positive control showed a positive result for all analysed samples detected as TNTC (too numerous to count, TNTC ≥ 1000 spw). Results are displayed as mean ± SD a reason why the highly fit individuals did not respond as much to a single session of exercise. Another reason could be that highly fit individuals are more used to exercise stimuli and therefore have a lower VST activation response to exercise. Among other co-stimulatory factors, a key component of this activation response is the clustering of T-cell receptors (TCR) at the centre of the T-cell/antigen-presenting-cell interface [56]. Recent studies have demonstrated that enhanced cholesterol levels of the whole cell and the plasma membrane are linked to increased functionality of activated CD8 + T cells resulting in enhanced TCR clustering and signalling as well as more efficient formation of the immunological synapse [57][58][59][60]. For this reason, we investigated whether a single bout of HIT or CONT results in higher cholesterol levels, which could lead to increased activation and cytotoxic function of CD8 + T cells by strengthening TCR signaling. Indeed, cholesterol levels in CD8 + T cells increased in response to exercise. Even though the difference between the two exercise protocols was not statistically significant, cholesterol levels and CD69 expression in CD8 + T cells increased significantly after continuous exercise, predominantly in moderately fit donors. As early marker for VSTs activation [61], the observed increases in CD69 and cholesterol correlated well with higher IFN-γ secretion. In a comprehensive future study, we will investigate the impact and modulatory role of cholesterol on VST activation and function.
Another interesting trend could be seen with HSP70. Different types of stress, one of which is exercise, induces HSP70 expression in order to adapt organisms appropriately [62,63]. In a rat model, it was shown that at a high-intensity workload, HSP70 increases while the proliferation response decreases [62]. This corresponds to our findings concerning HSP70 and the proliferation marker Ki67.
However, it is not surprising that a single bout of exercise did not alter measures of replicative senescence such as telomere length. Habitual exercise slows the age-related decrease in telomere length. However, this seems to be one of the long-term effects of regular exercise [64] and is not observed after a single exercise bout.
Although they are now used as second-line treatment in patients who do not respond to first-line treatments such as cidofovir for AdV infections, researchers are evaluating the usefulness of VSTs as an adjuvant to standard agents in the initial treatment [65].
In the present study, the peak VST frequencies differed from the peak levels of general lymphocyte counts: HIT did not evoke the highest VST activity when the CD4 + and CD8 + frequencies peaked. In fact, a higher antigen response was observed 24 h after HIT even though frequencies of the above-mentioned T-cell subsets were higher at other time points. Spielmann et al. found that although continuous exercise augments the ex vivo manufacture of VSTs, this effect does not appear to be due to numerical changes in the number of VSTs stimulated; hence, they deemed it unlikely that the beneficial effect of exercise achieved in their ex vivo expansion experiments could be replicated by simply drawing larger volumes of blood [39].
Thus, a single bout of HIT exercise 24 h before donation can serve as a simple and economical adjuvant to significantly improve adoptive T-cell therapy with respect to manufacturing functional-active antigen-specific T-cell products in sufficient cell numbers. This is especially effective in the case of memory VSTs with low precursor frequencies.

Conclusions
A single bout of exercise, preferably HIT 24 h prior to blood donation, can significantly improve T-cell enrichment in VST production for adoptive T-cell immunotherapy. Exercise has benefits, especially for viral antigens for which most healthy individuals possess less frequent T cells, particularly in donors who are moderately fit. As single bouts of exercise in general seem to be beneficial, they can be recommended to all donors. Fig. 4 Effects. of a single continuous and interval exercise on the activation capacity of functional IFN-γ + antigen-specific T cells. Peripheral blood samples of healthy donors (n = 6) were analysed before, 1 h after and 24 h after a single 30 min interval (HIT) using IFN-γ Cytokine Secretion Assay (CSA). Isolated PBMCs were stimulated overnight with a CMV-, b EBV-and c AdV-specific peptide pools (CMV pp65, CMV IE1, EBV EBNA1, EBV Consensus, AdV5 Hexon and AdV5 Penton). Negative control was carried out by using unstimulated PBMCs. The relevant cell fractions before and after enrichment were used for a detailed flow cytometric analysis of IFN-γ + T-cell subsets. Results are shown as fold changes of the frequency of CD3 + /IFN-γ + , CD8 + /IFN-γ + and CD4 + /IFN-γ + cells after subtracting those of the respective negative control. CMV and AdV results are displayed as mean ± SD (See figure on next page.)