SEARCH

SEARCH BY CITATION

Keywords:

  • interleukin-2;
  • interleukin-21;
  • regulatory T cells;
  • tumour immunotherapy

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Interleukin-2 (IL-2) is a mainstay for current immunotherapeutic protocols but its usefulness in patients is reduced by severe toxicities and because IL-2 facilitates regulatory T (Treg) cell development. IL-21 is a type I cytokine acting as a potent T-cell co-mitogen but less efficient than IL-2 in sustaining T-cell proliferation. Using various in vitro models for T-cell receptor (TCR)-dependent human T-cell proliferation, we found that IL-21 synergized with IL-2 to make CD4+ and CD8+ T cells attain a level of expansion that was impossible to obtain with IL-2 alone. Synergy was mostly evident in naive CD4+ cells. IL-2 and tumour-released transforming growth factor-β (TGF-β) are the main environmental cues that cooperate in Treg cell induction in tumour patients. Interleukin-21 hampered Treg cell expansion induced by IL-2/TGF-β combination in naive CD4+ cells by facilitating non-Treg over Treg cell proliferation from the early phases of cell activation. Conversely, IL-21 did not modulate the conversion of naive activated CD4+ cells into Treg cells in the absence of cell division. Treg cell reduction was related to persistent activation of Stat3, a negative regulator of Treg cells associated with down-modulation of IL-2/TGF-β-induced phosphorylation of Smad2/3, a positive regulator of Treg cells. In contrast to previous studies, IL-21 was completely ineffective in counteracting the suppressive activity of Treg cells on naive and memory, CD4+ and CD8+ T cells. Present data provide proof-of-concept for evaluating a combinatorial approach that would reduce the IL-2 needed to sustain T-cell proliferation efficiently, thereby reducing toxicity and controlling a tolerizing mechanism responsible for the contraction of the T-cell response.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

The ability to regulate the development of an immune response by cytokines provides the theoretical basis for the use of cytokines as adjuvants to potentiate the immune response to tumour vaccines. Interleukin-2 (IL-2) is a well-known cytokine that is implicated in the agonistic stimulation of immune responses for its ability to act as a powerful T-cell growth factor.[1, 2] In the clinical setting, IL-2 is currently approved for the treatment of renal cell carcinoma and melanoma and is also being tested in clinical trials for adoptive cell transfer immunotherapy, as it favours the ex vivo expansion of autologous tumour-reactive effector T cells as well as their long-lasting survival following re-infusion.[1, 2] However, IL-2 usage is limited by a severe toxicity, sometimes even requiring intensive care[3]. Moreover, experimental and clinical evidence shows that IL-2 contributes to maintaining peripheral tolerance by supporting the survival and function of CD4+ CD25+ Foxp3+ regulatory T (Treg) cells, a central component of tumour-mediated immunosuppression and capable of suppressing the development of protective anti-tumour effector T-cell responses.[4-7] Cancer cells and infiltrating normal cells at the tumour site secrete high levels of transforming growth factor β (TGF-β).[8] This cytokine has a profound inhibiting effect on the immune system; among other actions, it can directly convert conventional T cells into Treg cells and, most importantly, synergizes with IL-2 in facilitating Treg cell development.[9, 10] Hence, paradoxically, IL-2 administered to patients, would both amplify and temper the T-cell response against tumour at the same time, providing one possible explanation of why patients do not improve despite an increased frequency of cytotoxic T cells: the suppressive effect of Treg cells can overrule the activation, and so immune homeostasis is re-established upon treatment.[11]

These observations make it important to search for a cytokine endowed with the ability to tip the balance against tolerance by sustaining effector T-cell proliferation in the absence of Treg cell induction. Interleukin-21 is a recently discovered type I cytokine made by activated CD4+ cells and natural killer T cells, and endowed with pleiotropic effects that appear to depend on its concentration and the presence of other cytokines.[12-14] It has its own receptor and shares the common γ-chain receptor with IL-2. Favourable preclinical features of IL-21 in the context of tumour immunotherapy include facilitation of interferon-γ (IFN-γ) production and, in combination with IL-2 or IL-15, an additive effect on natural killer lytic function.[12, 15] Most importantly, IL-21 reportedly curbs Treg cell suppressive activity and survival in vitro, and decreases Treg cell accumulation in the tumour microenvironment, thereby probably contributing to subvert immune suppressive networks in the tumour microenvironment.[16-18] These favourable features make IL-21 superior to other cytokines belonging to the common γ-chain family, namely IL-7 and IL-15, which appear to drive Treg expansion and mobilization as IL-2.[19, 20] From a clinical standpoint, IL-21 seems to have a better safety profile than IL-2 and has been explored in phase I and II studies in various malignancies.[21-23] However, a major drawback of IL-21 is its scarce efficiency in sustaining T-cell proliferation compared with IL-2,[12, 24] although IL-21/IL-2 combination enhances CD8+ cell expansion in vitro and in animal models.[25, 26]

Against this background, it was deemed important to study whether IL-21 could be combined with IL-2 to best exploit the IL-2 pro-proliferative activity and IL-21 anti-Treg cell activity. Data presented here show that IL-21 synergizes with IL-2 in increasing T-cell receptor (TCR) -dependent T-cell proliferation to a level that is impossible to achieve with IL-2 alone, and concomitantly curtails Treg cell development. From a molecular standpoint, Treg cell blockage reflects the ability of IL-21 to bias intracellular signalling against Treg cell development. Contrary to early conclusions,[16, 17] IL-21 does not reverse the suppressive function of Treg cells.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
Peripheral blood mononuclear cell isolation and immunomagnetic cell sorting

Peripheral blood mononuclear cells (PBMC) were isolated from healthy adult volunteers via density gradient centrifugation using Ficoll–Hypaque (Sigma-Aldrich, Munich, Germany). Local Ethics Committee approval and informed consent were obtained from all donors. CD25+ cells to be used as Treg cells were isolated from PBMC using immunomagnetic anti-CD25 microbeads (Miltenyi Biotec, Bergisch-Gladbach, Germany). CD25-depleted PBMC to be used as responder cells were obtained by double negative selection using anti-CD25 microbeads (Miltenyi Biotec). This procedure usually leaves behind 1–2% CD4+ Foxp3+ cells.[27] For experiments aimed at separately studying the activity of IL-21 in naive and memory T cells, untouched T cells were first purified with the Pan T cell Isolation Kit (Miltenyi Biotec) to which anti-CD25 microbeads were added. Naive CD45RA+ and memory CD45RO+ cells were then purified by immunomagnetic CD45RA and CD45RO microbeads (Miltenyi Biotec). Autologous monocytes to be added to CD45RA+ and CD45RO+ cell cultures were purified from PBMC using anti-CD14 microbeads (Miltenyi Biotec). Experiments were performed only if the purity of sorted cells exceeded 90%.

Cell cultures and immunosuppression assay

Cells were plated in complete medium consisting of RPMI-1640 (Gibco, Grand Island, NY) supplemented with GlutaMAX™ (Invitrogen Life Technologies, San Diego, CA), 2% human AB serum (Sigma-Aldrich), 100 μg/ml streptomycin and 100 U/ml penicillin (Sigma-Aldrich). Cells were cultured with T Cell Activation/Expansion Kit (TCAE; Miltenyi Biotec) that allows the concomitant engagement of CD3, CD2 and CD28. Where indicated, culture media were supplemented with IL-21 (Prospec, Rehovot, Israel) and TGF-β (R&D Systems, Minneapolis, MN) at the beginning of culture, and with IL-2 (Roche Diagnostics GmbH, Mannheim, Germany) on day 3, according to the TCAE manufacturer's instructions, unless otherwise stated. To track cell proliferation, responder cells were stained with 0·2 μμ carboxyfluorescein succinimidyl ester (CFSE, Invitrogen), as described previously.[27] In some experiments, cells were stained with 0·8 μm Cell Proliferation Dye eFluor 670 (referred to here as CPD; eBioscience, San Diego, CA) and 2·5 μm CellTrace Violet (referred to here as CTV; Invitrogen), according to the manufacturers' instructions. Halving of the dye was measured by flow cytometry. The absolute number of growing cells in each population was determined by referring the total number of viable cells measured by an electronic cell counter (Z2, Beckman Coulter, Brea, CA) endowed with an electronic gate to exclude dead cells and debris on the basis of low electrical impedance, to the percentage of CD4+ or CD8+ cells. Because all cultures were performed in the presence of 2% human serum that may contains TGF-β, the latter was quantified in freshly prepared medium using the DRG ELISA kit for TGF-β (DRG Instruments GmbH, Marburg, Germany). TGF-β concentration was always below the lower detection limit (< 0·4 pg/ml), i.e. orders of magnitude lower than that used throughout the study (5 ng/ml). Assay for the Treg-cell-mediated suppression was carried out as described previously.[27] Briefly, CD4+ CD25+ cells and CFSE-loaded responder cells were co-cultured at a 1 : 1 suppressor/responder cell ratio in the presence of TCAE and the indicated cytokines for 4–5 days.

T-cell analysis by flow cytometry

Flow cytometry analysis of cell surface phenotype was performed using appropriate combinations of fluorescein isothiocyanate (FITC)-, phycoerythrin (PE)-, ECD-, PE/Cy5 (PC5)-, allophycocyanin (APC)-conjugated monoclonal antibodies (mAbs) to CD3, CD8, CD25, CD45RO and IL-21 receptor (IL-21R), clone 17A12 (all from BD Biosciences, Mountain View, CA), and CD4 and CD45RA (Beckman Coulter); mAb to Foxp3 (clone PCH 101 and clone 259/D) was from eBioscience and BD Biosciences, respectively. Glycoprotein-A repetitions predominant (GARP, Plato-1 clone) was from Enzo Life Sciences (Lausanne, Switzerland).[28] Proliferating CD4+ cells were sometimes assessed as CD8 cells, because direct recognition of CD4+ cells by anti-CD4 mAb in activated bulk T-cell cultures can be hampered by the down-regulation of the CD4 molecule. Appropriate isotype controls were included for each sample, with the exception of Foxp3 staining. Background value, which can be a problem when non-specific binding of mAb that results from the fixation and permeabilization steps adds to the noise, was objectively assessed by an isoclonic control obtained by pre-blocking Foxp3 binding sites with a 10-fold excess of unconjugated mAb to Foxp3. Sample was then incubated with the fluorescent-dye-conjugated anti-Foxp3 mAb. Flow cytometry analysis of signal transducer and activator of transcription 3 (Stat3), Stat5 and Smad2/3 phosphorylation status (pStat3, pStat5 and pSmad) was performed using phospho-specific FITC- or phycoerythrin-conjugated mAb to pStat3 and Stat5, respectively (BD Biosciences) and unconjugated mAb to pSmad2/3 (Cell Signaling Technology, Danvers, MA). Binding of the latter was revealed using an Alexa 488-conjugated F(ab')2 goat anti-rabbit IgG antiserum (Cell Signaling Technology). Appropriate combinations of cell surface antigen-specific fluorochrome-conjugated mAbs were used to correlate pStat3, pStat5 and pSmad2/3 expression to the cell subset of interest. PBMC were allowed to rest for 3 hr at 37° before being stimulated with the cytokines. Because Stat phosphorylation in peripheral T cells is relatively transient,[29] pilot time–course experiments were performed. The best time frame to detect changes in the phosphorylation status was 20 min for Stat3 and Stat5 and 60 min for Smad2/3. At the end of incubation, cells were immediately fixed and permeabilized using Cytofix fixation buffer and Phosflow Perm Buffer III (both from BD Biosciences). Staining was performed following the manufacturer's instructions and the original literature report.[30]

FACS analysis was performed using an EPICS-XL (Beckman Coulter) and a CyAn flow cytometer (Beckman Coulter). List mode data were analysed using Expo 32™ and Summit 4.3™ (both from Beckman Coulter) software. Forward and side scatter signals served to establish the lymphocyte gate. A minimum of 5000 cells of interest was acquired for each sample.

Statistics

Repeated measure anova followed by Tukey's test and paired Student's t-test were performed using the statistical software statistica (StatSoft Italia srl).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

IL-21/IL-2 combination acts synergistically on T-cell proliferation

The present experiments were purposely conducted to prove synergy between IL-21 and IL-2 in sustaining TCR-driven T-cell proliferation. Hence, it was mandatory to establish the amount of test substances that produces the highest effect. For this reason, the IL-21 and IL-2 concentrations that induced the maximal proliferative response in our system were assessed. On the grounds that IL-21 reportedly interferes with Treg-cell-mediated suppression,[16, 17] all experiments were conducted in PBMC cultures in which Treg cell frequency was made marginal by an extensive depletion of CD25+ cells. Proliferative responsiveness was assessed using the proliferation index (PI) as computed by the Proliferation Wizard contained in the ModFit LT program (Verity Software House, Inc., Topsham, ME). The PI takes into account the number of cells dividing and the number of divisions undergone, thereby providing an accurate measure of cell proliferation history and reducing the inherent subjectivity of visually gating on proliferating cells. IL-21 and IL-2 enhanced CD4+ and CD8+ cell proliferation in a dose-dependent manner, the maximal proliferative response being observed at 100 ng/ml for IL-21 and 40–80 U/ml for IL-2 (Fig. 1a). Cell proliferation calculated by CFSE dilution assay directly reflected corresponding differences in terms of viable cell number recovered at the end of culture (Fig. 1a, insets).

image

Figure 1. Effect of interleukin-21 (IL-21), IL-2 and different IL-21/IL-2 combinations on T-cell proliferation. CD25-depleted peripheral blood mononuclear cells (PBMC) were loaded with CFSE and stimulated with TCAE in the presence or absence of the indicated cytokines. Proliferative responsiveness was assessed by computing proliferation index (PI) in CD4+ and CD8+ gated cells. (a) Proliferation of CD4+ and CD8+ cells in the presence of increasing concentrations of IL-21 and IL-2. Horizontal dotted line is the maximal proliferative response. Insets are the number of cells recovered at the end of culture. Dashed columns are the proliferative response measured as CFSE halving or (inset) cell number to TCAE alone. Columns and data-points represent the mean value of duplicates of one experiment representative of four. (b) Proliferation of CD4+ and CD8+ cells in the presence of IL-21, IL-21 and different IL-21/IL-2 combinations. Results are shown as mean ± SD of four independent experiments run in duplicates.

Download figure to PowerPoint

Applying an asymptotic fractional response computation[31] showed that the relative efficacy was 71% for CD4+ and 84% for CD8+ cells. Based on these findings, to assess synergy TCAE-driven proliferation of CD4+ and CD8+ cells exposed to sub-maximal (20 U/ml) and supra-maximal (300 U/ml) IL-2 dosing in the presence of sub-maximal (25 ng/ml) and supra-maximal (100 ng/ml) IL-21 dosing was measured (Fig. 1b). The provision of IL-21 to IL-2-containing cultures significantly boosted CD4+ cell proliferation, even in the presence of supra-maximal IL-2 dosing (Fig. 1b), thereby indicating a positive synergistic effect on cell proliferation. Consequently, adding IL-21 to the lowest IL-2 amount (20 U/ml) induced a proliferative response that was considerably higher than that obtained with the highest IL-2 amount (Fig. 1b). CD8+ cell proliferation also increased but the variation fell short of statistical significance (Fig. 1b).

The most likely explanation for the comparatively lower sensitivity of CD8+ cells to IL-21 addition resides in the highest intrinsic responsiveness of this cell subset to the culture conditions, which did not allow for other than marginal increases, a view supported by the consistent observation that CD8+ cell proliferative responsiveness in unfractionated T-cell cultures was always higher than that of the correspondent CD4+ subset and confirmed by pilot experiments in which CD8+ cells were responsive to IL-21/IL-2 combination when cultured in the absence of the help provided by neighbouring CD4+ T cells (not shown).

Previous studies using purified T-cell subset cultures indicated that IL-21 preferentially favours naive over memory T-cell responses.[12, 24] We reasoned that although informative in the context of an in vivo situation, present experiments did not tease apart the contribution of memory and naive T cells. Hence, to scrutinize the effects on the two subsets and place our data in the context of current literature, CD25-depleted PBMC were immunomagnetically sorted into CD45RA+ (naive) and CD45RO+ (memory) T cells, loaded with CFSE, and cultured as described above. Interleukin-21 significantly enhanced IL-2-induced naive but not memory CD4+ cell proliferation (Fig. 2a). Naive CD8+ cell proliferation was somewhat increased by IL-21/IL-2 combination, which conversely did not modify memory CD8+ cell proliferation (Fig. 2b). The cytokine-induced modulation of cell proliferation reflected analogous changes in absolute cell counts (Fig. 2c,d).

image

Figure 2. Effect of interleukin-21 (IL-21), IL-2 and different IL-21/IL-2 combinations on naive and memory T-cell proliferation. Proliferative responsiveness was assessed by computing the proliferation index (PI) in CD4+ and CD8+ gated cells. CD25-depleted purified CD45RA+ cells and CD45RO+ cells were loaded with CFSE and stimulated with TCAE in the presence or absence of the indicated cytokines. Proliferation of (a) CD4+ and (b) CD8+ gated cells. Results are shown as mean ± SD of three independent experiments run in duplicates. (c) and (d) are the number of CD4+ and CD8+ cells recovered at the end of culture, respectively.

Download figure to PowerPoint

IL-21 inhibits Treg cell expansion following naive T-cell activation in the presence of IL-2 and TGF-β

Previous work has indicated that IL-21 can decrease Treg cell generation occurring during proliferation of tumour-specific T cells.[17] However, the activity of IL-21 on Treg cells has not yet been explored in the presence of context-specific signalling cues mimicking a highly immunosuppressive tumour microenvironment. To this end, cultures were supplemented with IL-2 and TGF-β to approximate a physiologically relevant condition, as IL-2 is commonly administered to sustain T-cell proliferation and TGF-β is often abundant in the tumour microenvironment being secreted by cancer cells and infiltrating normal cells.[8] Concentrations of IL-2 and TGF-β were those previously reported to provide robust induction of Treg cell development in vitro.[32] The study of Treg cells is complicated by a paucity of phenotypic markers to distinguish activated effector T cells from Treg cells. Because in addition to Foxp3, GARP has been recently added to the list of Treg cell markers,[33] here Foxp3 and GARP co-expression was used as a surrogate Treg cell indicator. Hence, CD4+ Foxp3+ GARP+ cells will be hereafter operationally referred to as Treg cells. Initially, we tested the ability of IL-21 to oppose Treg cell development in a TCAE-stimulated CD25-depleted unfractionated PBMC population. In these experiments, Treg cells were assessed at day 5. As expected,[32] the TGF-β/IL-2 combination was mostly efficient in inducing Treg cell development, an effect that was antagonized by IL-21 (Fig. 3a).

image

Figure 3. Interleukin-21 (IL-21) counteracts IL-2 and transforming growth factor-β (TGF-β) -driven regulatory T (Treg) cell induction on naive and memory CD4+ cells. CD25-depleted unfractionated peripheral blood mononuclear cells (PBMC) or purified naive and memory cells were stimulated with TCAE in the presence or absence of the indicated cytokines. Treg cell frequency was assessed 5 days later using FACS analysis gated on CD4+ cells. (a) Effect of IL-21 on Treg cell induction in unfractionated CD25-depleted PBMC. Results are shown as mean ± SD of three independent experiments run in duplicates. (b) Effect of IL-21 on Treg cell induction in naive and memory T cells. Results are shown as mean ± SD of two independent experiments run in duplicates. *Statistically different from all other conditions (P < 0·05).

Download figure to PowerPoint

Repeating the experiments using immunomagnetically purified CD25-depleted naive and memory CD4+ cells showed that both subsets contributed to the Treg cell induction seen in unfractionated CD25-depleted PBMC, but IL-21 significantly antagonized Treg cell development only in naive cells (Fig. 3b).

The observed reduction of Treg cell frequency could reflect impairment in either Treg cell de novo generation, proliferation or both. Moreover, it could just as well be better proliferation of other T cells that produced the result. To analyse the mechanisms behind IL-21 detrimental activity on Treg cell development, changes induced by IL-21 on Treg cell development from the early phases of T-cell activation were tracked. To this end, naive T cells were immunomagnetically purified from CD25-depleted T cells and stimulated by TCAE in the presence of cytokines. Parallel cultures were established using CFSE-loaded naive T cells. Flow cytometric analysis performed on electronically gated CD4+ cells, demonstrated an induction of Treg cells that began at 72 hr (Fig. 4a, upper panel). By that time, Treg cells were more evident in TGF-β/IL-2-supplemented cultures, although IL-21 perceptibly reduced their frequency (Fig. 4a, upper panel). The kinetics of changes of two indicators of early T-cell activation, namely CD25 expression and forward scatter (FSC), the latter as an estimate of cell size,[34] was not affected (Fig. 4a, middle and lower panel, respectively), thereby excluding the possibility that the observed differences in Treg cell induction merely reflected quantitative differences in cell activation. At the same time-point, cells had entered the first and, although to a lesser extent, second division stages, thereby generating one parental and two daughter cell populations, as exemplified in Fig. 4(b, upper panel), irrespective of the added cytokines. The strong Treg cell induction driven by the TGF-β/IL-2 combination was mostly evident in daughter cell populations (Fig. 4b, table). Interleukin-21, ineffective in modulating Treg cell formation in parental cell population, significantly counteracted Treg cell induction in the daughter cell populations (Fig. 4b, table).

image

Figure 4. Interleukin-21 (IL-21) counteracts IL-2 and transforming growth factor-β (TGF-β) -driven regulatory T (Treg) cell induction by hampering Treg cell proliferation. CD25-depleted naive T cells were stimulated with TCAE in the presence or absence of the indicated cytokines. Cytokine concentrations were as in Fig. 3. FACS analysis was performed after gating on CD4+ cells. (a) Time–course of the changes in Treg cell frequency, CD25 expression, and cell size (upper, middle and lower plot, respectively) in the presence of IL-2, IL-2/IL-21, IL-2/TGF-β, and IL-2/TGF-β/IL-21. Parameters were measured at the indicated time-points. Each data-point is the mean value of duplicates from one of four independent experiments with identical results. (b) Representative CFSE histogram at day 3 showing the strategy used to identify cells in the parental and generation 1 and 2 (P, G1 and G2, respectively) and assess Treg cell frequency in relation to proliferative status. Numbers in the table underneath the plot represent Treg cell frequency within each generation in the different culture conditions. Data are shown as range values from three independent experiments run in duplicate. (c) Following up with naive CD4+ cell cultures for 5 days. A representative experiment is shown. Upper and lower row, Treg and non-Treg cells, respectively. Numbers denote proliferation index (PI). (d) Results are shown as fold change value relative to TCAE alone for Treg and non-Treg cells, as detailed in the text (mean ± SD).

Download figure to PowerPoint

Following up with cell cultures, showed that most cells were in active proliferation by day 5. These cultures served to further analyse the influence of IL-21 on the proliferative status of Treg and non-Treg cells. To account for the differences in proliferative response induced by the diverse culture conditions, we first computed the PI of Treg and non-Treg cells generated in each culture condition, as shown in the example experiment in Fig. 4(c). Next, relative changes in PI of Treg and non-Treg cells resulting from cytokine supplementation were determined according to the following formula:

  • display math

By this method, it was demonstrated that TGF-β/IL-2 combination favoured Treg over non-Treg cell proliferation and the effect was significantly antagonized by IL-21 (Fig. 4d). In aggregate, these findings indicated that IL-21 counteracted Treg by affecting their proliferation but had marginal or no effect on the conversion of resting naive CD4+ cells into Treg cells.

Activated naive T cells express IL-21R at higher density than memory T cells

Expression of IL-21R was assessed in resting and activated T cells. The fluorescence minus one (FMO) strategy was used to assess background fluorescence in the PE (IL-21R) channel. To this end, PBMC were first stained with all reagents except for IL-21R mAb to allow recognition of the cell subsets of interest, i.e. naive and memory T cells, and B cells, and then divided into two aliquots. One was incubated with IL-21R mAb and the other one was incubated with isotypic control mAb. The IL-21R expression was clearly higher in B cells than non-B cells, the latter showing a generalized increase in the fluorescence signal (Fig. 5a), in accord with the technical data sheet provided by the manufacturer. Gating on naive and memory T cells showed that IL-21R expression level was identical in both subsets (Fig. 5b) and with no detectable difference between CD4+ and CD8+ T cells (not shown). Interleukin-21R expression reportedly increases following activation.[12, 16] Hence, CD25-depleted immunomagnetically purified naive and memory T cells were stained with CTV and CPD, respectively, mixed at a 1 : 1 ratio, and activated by TCAE. Cultures were harvested at day 3, stained with appropriate combinations of mAbs to allow recognition of IL-21R on the cell subsets of interest, i.e. CD4+ and CD8+ T cells following the FMO strategy, and receptor expression was visualized in relationship with successive rounds of cell replication. The IL-21R was preferentially expressed by proliferating T cells, and at a higher level in naive than memory CD4+ and CD8+ T cells (Fig. 5c).

image

Figure 5. Analysis of interleukin-21 receptor (IL-21R) expression in resting, and activated naive and memory T cells. (a) Peripheral blood mononuclear cells (PBMC) were stained with monoclonal antibodies (mAbs) to CD3, CD20, CD45RA, CD45RO and IL-21R to identify specific cell populations. (b) Naive and memory T cells were identified on the basis of CD45RA and CD45RO staining (not shown), and IL-21R expression was assessed in naive, vertical line histogram, and memory T cells, slashed line histogram. Isotypic staining is not shown to facilitate comparison between naive and memory T-cell staining. One flow cytometric analysis representative of four is shown. (c) CD25-depleted purified naive and memory T cells were loaded with CTV and CPD, respectively, and stimulated with TCAE for 72 hr. Cultures were stained with mAbs to CD3, CD4, CD8 and IL-21R to identify specific cell populations. Naive and memory cells were first identified among cultured cells by gating on CTV and CPD (not shown) and then electronically sorted into CD4+ and CD8+ T cells. The percentage of IL-21R-positive cells was measured from a cut-off set using an isotype-stained control. Numbers are % of cells in each quadrant. One of two identical experiments is shown.

Download figure to PowerPoint

IL-21 maintains Stat3 phosphorylation even in the presence of IL-2 and TGF-β, and down-modulates Smad2/3 phosphorylation evoked by IL-2/TGF-β combination

Having shown that IL-21 interfered with IL-2 and IL-2/TGF-β combination-induced Treg cells, we set out to delineate the molecular mechanism involved. There are reasons to believe that the Stat3/Stat5 signalling pathways are pivotal in Treg cell generation. IL-21 regulates Foxp3 negatively in a Stat3-dependent manner, whereas IL-2 promotes Treg cell differentiation via a direct binding of Stat5 to the foxp3 promoter.[35, 36] TGF-β cooperates with IL-2 in Treg cell induction via specific surface receptors that, in turn, phosphorylate and activate members of the Smad family of tumour suppressors, termed R-Smads, which includes Smad2 and Smad3.[37, 38] Against this background, CD25-depleted PBMC were subjected to cytokine stimulation or left unstimulated for 20 min to 1 hr and then stained with anti-pStat3, anti-pStat5 or anti-pSmad2/3, and target cell surface marker mAbs to separate effects on naive and memory CD4+ cells. Interleukin-21 induced a strong Stat3 phosphorylation that was not modified by IL-2/TGF-β combination in both naive and memory CD4+ cells, and significantly dampened Smad2/3 phosphorylation that was induced by IL-2/TGF-β combination (Fig. 6). However, IL-21 did not modify Stat5 phosphorylation induced by IL-2/TGF-β combination (Fig. 6).

image

Figure 6. Effect of interleukin-21 (IL-21), IL-2, transforming growth factor-β (TGF-β) and different cytokine combinations on signal transducer and activator of transcription (Stat3), Stat5 and Smad2/3 phosphorylation (pStat3, pStat5 and pSmad2/3, respectively) in naive and memory CD4+ T cells. CD25-depleted peripheral blood mononuclear cells (PBMC) were stimulated with the indicated cytokines as described in the Material and methods. Modulation of pStat3, pStat5 and Smad2/3 in response to various cytokine combinations was assessed by FACS analysis gated on CD45RA+ CD45RO or CD45RA CD45RO+ CD4+ cells. (a) Representative flow cytometry histograms from one experiment out of five that are summarized in (b) are shown. The vertical marker line in histograms referring to pStat3 and pStat5 was set on isotypic control stained unstimulated samples. Numbers denote percentage of cells whose fluorescent signal exceeded the marker. pSmad2/3 data are expressed as shift in median fluorescence intensity (MFI) over the entire population because of the uncertainties to determine positive/negative boundaries in most samples. (b) Results of five independent experiments are shown as mean ± SD. n.s., not significant; *P < 0·05 by paired Student's t-test.

Download figure to PowerPoint

IL-21 does not interfere with Treg-cell-mediated suppression

Previous works have indicated that IL-21 opposes Treg-cell-mediated suppression of the T-cell proliferative response.[16, 17] Immunomagnetically purified CD4+ CD25+ cells, i.e. Treg cells, were added back to the correspondent CD25-depleted unfractionated PBMC or purified naive and memory T cells at a 1 : 1 suppressor to responder cell ratio. Cells were cultured with TCAE in the presence of the indicated cytokines. Treg cells abrogated the proliferative responsiveness of responder cells in all culture conditions, irrespective of IL-21 presence, although in the absence of Treg cells, IL-2-containing cultures proliferated better than TCAE alone cultures and cultures containing an IL-21/IL-2 combination proliferated better than both IL-2 and TCAE alone cultures, as expected (Fig. 7).

image

Figure 7. Interleukin-21 (IL-21) does not subvert regulatory T (Treg) cell-mediated suppression. CD25-depleted unfractionated peripheral blood mononuclear cells (PBMC) or purified naive and memory T cells were loaded with CFSE and stimulated with TCAE in the presence or absence of autologous CD4+ CD25+ cells at a 1 : 1 suppressor/target ratio in the presence or absence of the indicated cytokines. CFSE dilution was assessed 5 days later using FACS analysis gated on CD4+ and CD8+ cells. Data are from one experiment run in duplicate. Similar results were seen in three to four independent assays. (a) CD25-depleted unfractionated PBMC. (b) CD45RA+ (naive) T cells. (c) CD45RO+ (memory) T cells.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Effective T-cell-based immunotherapy requires the successful and sustained mobilization of sufficient numbers of effector T cells that recognize tumour cells. High-dose IL-2 is currently used for effector T-cell expansion but its therapeutic usefulness appears limited by a concomitant Treg cell expansion.[4, 5] The discovery that IL-21 sustains T-cell proliferation without favouring Treg cell development and has a more favourable safety profile raised considerable interest for a possible use of this cytokine as an alternative to IL-2 in tumour immunotherapy.[21-23] Unfortunately, original findings in the field showed that IL-21 was less efficient than IL-2 in sustaining T-cell proliferation.[12, 24] Here we can restate the observation that IL-21 is not able to substitute IL-2 in full, which therefore remains unrivalled for its capability to enhance T-cell proliferation, and also add the observation that IL-21 synergistically up-regulates T-cell expansion in combination with IL-2. A positive interaction between the two cytokines in sustaining T-cell proliferation has been already described, but only an additive rather than synergistic activity could be demonstrated.[12] Data presented here show that IL-21/IL-2 combination takes T-cell proliferation to a level that is impossible to obtain with IL-2 alone and, consequently, that the maximum level of T-cell expansion obtainable with IL-2 alone can be reached by adding IL-21 to a considerably lower amount of IL-2. This may have clinical implications, as it might be anticipated that lowering the IL-2 amount needed to obtain a good T-cell response will result in a lower systemic toxicity.

At odds with the present data, the IL-21/IL-2 combination reportedly decreases IL-2-driven T-cell proliferation.[24] The basis for the differences between our and early results are not clear The discrepancy might be the result of differences in culture methods, e.g. the amount of cytokine used or the usage of polyclonal activators.

The observed T-cell expansion could reflect increased cell proliferation, enhanced survival, or both. Indeed, evidence from an animal model suggested that the effect on T-cell expansion seen in response to IL-21 primarily resulted from increased cell survival.[39] In humans also, IL-21 has been proposed to facilitate T-cell survival.[24] Although ad hoc experiments to formally exclude a contribution by cell survival were not performed, the strict relationship between the increase in cell number and the increase in rounds of cell divisions in all culture conditions tested rather suggests that IL-21 mostly cooperated with IL-2 in directly sustaining T-cell replication.

Proliferation experiments indicated naive CD4+ cells as the preferential target of IL-21. This prompted us to investigate whether, among the possible mechanisms involved, the phenomenon reflected differences in IL-21R expression. In freshly purified T cells IL-21R was homogeneously expressed by the various T-cell subsets and at quite a low level. Conversely, boosting receptor expression by in vitro activation demonstrated a higher expression in naive than memory CD4+ T cells. This observation fits well with the preferential synergistic activity of the IL-21/IL-2 combination on naive CD4+ T cells. Additionally, we documented that naive CD8+ T cells also expressed IL-21R at a higher level than their memory counterpart following activation, in line with the generalized greater susceptibility of naive T cells to IL-21. However, the ultimate effects of IL-21 cannot be justified solely on the basis of receptor expression, as CD8+ T cells, which were consistently less prone to be modulated by IL-21/IL-2 combination, expressed IL-21R at higher levels than CD4+ T cells.

An alternative explanation for the ability of IL-21 in facilitating T-cell proliferation is that IL-21 activity is more directed to Treg cells than to responder cells. We may infer that the increased T-cell proliferation seen in naive T cells reflected the ability of IL-21 to dampen Treg cell development only in this subset, with consequent release of responder cells from Treg blockage. On this point, a reduced Treg cell development has been invoked earlier to explain the enhanced cytotoxic T-cell proliferation.[17]

As alluded to above, here we show that IL-21 opposes Treg cell expansion and does so even in an experimental model mimicking one of the worst cytokine environments in tumour immunotherapy, i.e. TGF-β, which is abundant in the tumour microenvironment, and IL-2, which must be administered to foster T-cell proliferation but cooperates with TGF-β in Treg cell induction.[8, 10, 32] These findings confirm and complement those of a previous in vitro study in which IL-21 was found to counteract Treg cell development in the absence of additional cytokines and fits well with an animal study showing that achieving a high concentration of IL-21 in the tumour tissue reduces Treg cells at the tumour site.[40] Mechanistically, Treg cell reduction is mostly a reflection of the ability of IL-21 to favour non-Treg cell over Treg cell proliferation, rather than a direct activity of IL-21 on Treg cell conversion, a result that enlarges a previous report in which IL-21 was found not to sustain Treg cell proliferation.[16, 17] This view is also in accord with a mouse study in which the reduced expression of Foxp3 in the presence of IL-21 was deemed imputable to a preferential expansion of Foxp3 cells.[41]

Minimizing Treg cell development is a major objective in immunotherapeutic protocols, but attempts made by Treg cell depletion treatments are often frustrated by the rapid induction of newly developed Treg cells.[42] It can be envisaged that this unfavourable effect may be prevented by IL-21, which reduces Treg cells by limiting their proliferation rather than killing them.

The intracellular signalling pathways involved in the ability of IL-21 to interfere with Treg cell development were also investigated. It is well established that IL-2 and TGF-β cooperate at the molecular level to favour Treg cell induction, as IL-2-mediated Stat5 phosphorylation activates the Foxp3 promoter and TGF-β contributes to Treg cell development through Smad2 and Smad3 activation.[36, 43] It is also known that IL-21 regulates Foxp3 negatively in a Stat3-dependent manner.[35] Present findings show that IL-21 exerts a negative interaction with the intracellular signalling pathways devoted to Treg cell development, as it maintains Stat3 phosphorylation even in the presence of IL-2 and TGF-β, and down-modulates the Smad2/3 phosphorylation evoked by IL-2/TGF-β combination. In aggregate, it is conceivable that Stat3 phosphorylation prevails over the downstream consequences of Stat5 and Smad2/3 phosphorylation, with consequent Treg cell down-modulation. Regarding TGF-β signalling, evidence indicates that the TGF-β/Smad pathway can also directly suppress T helper type 1 cell development.[43] Hence, from an oncological perspective, a diminished Smad2/3 phosphorylation suggests an additional beneficial role of IL-21.

Stat3 phosphorylation experiments showed that IL-21 was equally effective in naive and memory CD4+ T cells. Although this may seem at odds with the preferential activity of IL-21/IL-2 combination on naive cell proliferation, it has to be underlined that Stat3 phosphorylation could be measured in resting cells only. We did not succeed so far in assessing differences in Stat3 phosphorylation in activated cells, because of the intrinsically high Stat3 phosphorylation.

Unfortunately, the potential usefulness of IL-21 in tumour immunotherapy does not seem to include a direct inhibition of Treg-cell-mediated suppression. This conclusion appears dissonant with the early assertions that IL-21 reverses the suppressive signal provided by Treg cells on T cells or that CD4+ and not CD8+ cells are made resistant to Treg-cell-mediated suppression.[16, 17] We can only conjecture as to why our results differ from those of other studies. Comparisons are difficult because of factors like differences in culture conditions and in the criteria that were used to measure responder cell proliferation. However, several populations of responder cells were tracked here namely, unfractionated CD25-depleted PBMC, and purified naive and memory CD4+ and CD8+ cells, and a virtually complete abrogation of cell proliferation was obtained in all instances.

In conclusion, the results of this study indicate the potential usefulness of IL-21/IL-2 combination to confer superior activation of the immune cells in the context of cancer immunotherapy. Additionally, as IL-21 has a more manageable profile of toxic effects than IL-2,[21-23] there would be an excellent chance to decrease IL-2 dosage while still inducing a sustained response characterized by a lower systemic toxicity.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

This work was partly supported by grant number 4210011 from MIUR, Italy.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References