SEARCH

SEARCH BY CITATION

Keywords:

  • Dendritic cells;
  • human;
  • NK cells;
  • rapamycin

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

Mammalian target of rapamycin kinase inhibitor (mTORi) rapamycin (RAPA) use in transplantation can lead to inflammatory complications in some patients. Our goal was to better understand how mTORi-exposed human monocyte-derived dendritic cells (DC) stimulated with pro-inflammatory cytokines shape T cell allo-immunity. RAPA-conditioned-DC (RAPA-DC) displayed a more immature phenotype than untreated, control (CTRL)-DC. However, subsequent exposure of RAPA-DC to an inflammatory cytokine cocktail (ICC) plus IFN-γ induced a mature Type-1 promoting phenotype, consisting of elevated HLA-DR and co-stimulatory molecules, augmented IL-12p70 and IL-27 production, but decreased IL-10 secretion compared to CTRL-DC. Co-culture of mature (m)RAPA-DC with allogeneic peripheral blood mononuclear cells resulted in significantly increased Type-1 (IFN-γ) responses by T cells. Moreover, NK cells acted as innate modulators that conveyed activating cell-to-cell contact signals in addition to helper (IFN-γ) and/or regulatory (IL-10) soluble cytokines. We conclude that production of IL12-p70, IL-27 and low IL-10 by RAPA-DC allowed us to elucidate how these cytokines as well as NK-DC interaction shapes T cell allo-immunity. Thus, lack of inhibitory NK cell function during allo-specific T cell activation by human ICC + IFN-γ-stimulated RAPA-DC may represent an unwanted effector mechanism that may underlie RAPA-induced inflammatory events in transplant patients undergoing microbial infection or allograft rejection.


Abbreviations
Ag

antigen

APC

antigen-presenting cells

CFSE

carboxyfluorescein diacetate succinimidyl ester

DC

dendritic cells

ELISA

enzyme-linked immunosorbent assay

FCS

fetal calf serum

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

GM-CSF

granulocyte-macrophage colony-stimulating factor

HSA

human serum albumin

ICC

inflammatory cytokine cocktail

IFN-γ

interferon gamma

MFI

mean fluorescence intensity

mTOR

mammalian target of rapamycin

NK cells

natural killer cells

PBMC

peripheral blood mononuclear cells

RAPA

rapamycin

rh

recombinant human

Treg

regulatory T cells

TLR

toll-like receptor.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

Dendritic cells (DC) are uniquely well-equipped, professional antigen (Ag)-presenting cells (APC) that can induce T cell immunity or tolerance [1, 2]. While immunogenic DC represent mature APC essential for stimulating naïve and memory T cell responses, tolerogenic DC include DC in the normal steady-state (immature DC), that can promote and maintain peripheral tolerance to auto- or allo-antigen. Mechanisms that underline DC tolerogenecity include: inducing effector T cell apoptosis, skewing of T cell polarization, promoting T cell anergy and generation of regulatory T cells (Treg) [1, 2]. Tolerogenic DC can be induced in vitro or in vivo by exposure to diverse immunosuppressive agents, that impact their phenotype and function, resulting in regulation of T cell immunity [3].

The mammalian target of rapamycin (mTOR) inhibitor, rapamycin (RAPA) is a macrocyclic triene with immunoregulatory properties [4-8]. While mTOR exists in two complexes, that is mTOR complex 1 (mTORC1) and mTORC2, RAPA mainly targets mTORC1, a highly conserved serine/threonine protein kinase, that controls cell responses to environmental cues [2, 9-11]. The ability of RAPA to inhibit myeloid DC differentiation, maturation and function has been studied extensively in vitro and in animal models [5, 9, 12, 13]. In murine systems, RAPA exerts a profound inhibitory effect on DC differentiation and function in vitro, impairing their phenotipic maturation in response to Toll-like receptor (TLR) or CD40 ligation, suppressing their T cell allostimulatory function and conferring the ability to induce allo-Ag-specific T cell hyporesponsiveness [5-7, 9, 12, 14]. In vivo, Taner et al. [5] first showed that recipient-derived RAPA-DC pulsed with donor allo-Ag and administered systemically before transplantation could prolong mouse heart allograft survival, while Turnquist et al. [9] found that a single infusion of similarly pulsed RAPA-DC, combined with a short course of low-dose RAPA, could promote transplant tolerance.

Human RAPA-DC generated from blood monocytes display a more immature phenotype than control (CTRL)-DC, suppress T cell proliferation and can induce Treg, depending on the protocol used and the timing of exposure to RAPA during DC differentiation [9, 15-17]. Paradoxically, however, following TLR ligation with LPS, both mouse and human RAPA-DC display enhanced IL-12p70 secretion, with concomitant decreased production of IL-10, thus acquiring a potential to promote activation of double positive Type-1/Type-2 T cell immunity, rather than immune regulation [14-16]. Moreover, when RAPA is administrated to transplant recipients, various adverse inflammatory reactions have been observed, including interstitial pneumonitis, increased incidence of acute cellular rejection, augmented pro-inflammatory cytokine production by myeloid cells and enhanced recall Ag-specific memory CD8+ T cell responses [18-20].

Notably, however, the impact of an inflammatory cytokine cocktail (ICC) IL-1β, IL-6, TNF-α plus IFN-γ on human RAPA-DC has not yet been examined. These two synergistic signals have been shown to direct human immature DC into Type-1 polarized DC (DC1) capable of a log increase in IL-12p70 production as compared to ICC alone, with subsequent strong downstream Type-1 T cell polarization [21]. This pathway of DC maturation has clinical relevance for transplant recipients on RAPA maintenance therapy, since both viral infections and acute cellular rejection after organ transplantation trigger inflammation and Type-1 interferon (IFN)-γ release.

Natural killer (NK) cells are critical components of the innate immune response and play an important role in early defense against viral infections and tumor growth [22-25]. In response to IL-12p70, they release large amounts of IFN-γ, a cytokine that mediates ‘helper’ activity during generation of Type-1 T cell immunity, including immunity to viruses, bacteria, oncogenic transformed cells and allo-immunity [26-28]. IFN-γ release by NK cells also promotes DC maturation and differentiation of Type-1-polarized DC with enhanced ability to secrete IL-12p70, and further elevate Type-1 immunity [29, 30]. Since NK cells constitutively express functional IL-12R β1 and β2, they can readily respond to IL-12p70 and thus may represent an important target of augmented IL-12p70 produced by RAPA-DC exposed to inflammatory signals.

Here, we show for the first time, that human monocyte-derived RAPA-DC exposed to ICC + IFN-γ display a mature phenotype and markedly enhanced secretion of both IL-12p70 and IL-27, associated with low production of IL-10. When co-cultured with allogeneic peripheral blood mononuclear cells (PBMC), ICC + IFN-γ matured (m)RAPA-DC deliver stimulatory cell-to-cell contact-dependent signals that target NK and T cells to augment downstream Type-1 immunity in allo-reactive CD8+ and bystander CD4+ and CD8+ T cells. In addition, mRAPA-DC cytokines IL-12p70 and IL-27 appear to instruct NK cells to exert either ‘regulatory’ or ‘stimulatory’ effects on IFN-γ production by allogeneic T cells. These findings have important implications for further understanding the impact of mTOR modulation on human allogeneic DC-NK-T cell interactions associated with inflammatory events, and for evaluation of RAPA as an immune regulatory drug in clinical transplantation.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

Media and reagents

RPMI-1640 supplemented with 2 mM L-glutamine, 10 mM HEPES, 100 IU/mL penicillin/streptomycin (all from Cellgro, Manassas, VA) and 5% normal human serum albumin (HSA) (Nabi, Boca Raton, FL) was used as complete medium. AIM-V medium (Gibco BRL) supplemented with 10% heat-inactivated fetal calf serum (FCS), 1000 U/mL recombinant human (rh) granulocyte–macrophage colony-stimulating factor (GM-CSF) (Berlex, Seattle, WA) and 1000 U/mL rhIL-4 (R&D Systems, Minneapolis, MN) is referred to as DC media. Recombinant IL-12p70, IL-27 and blocking antibodies (Ab) α-IL-12p70 (clone 24910) and polyclonal α-IL-27 were purchased from R&D Systems.

Human PBMC, monocyte, NK and T cell isolations

Leukocyte concentrates from 18 healthy adult volunteers were purchased from the Central Blood Bank of Pittsburgh. Twenty-five healthy adult control (HC) volunteers were recruited following informed consent under an IRB-approved protocol at the University of Pittsburgh (IRB#00608014). PBMC were isolated by Ficoll-Hypaque (GE Healthcare Bio-Science AB, Uppsala, Sweden) density gradient centrifugation. CD14+ monocytes, CD3+ T cells or CD56+ NK cells were isolated by positive selection using micro-magnetic beads (Miltenyi Biotec, Auburn, CA) according to the manufacturer's protocol. The purity of each sorted cell population was in all cases >95%.

DC generation and maturation

CD14+ monocytes were cultured for 7 days in DC media. Where specified, RAPA (10 ng/mL; Sigma, St. Louis, MO) was added to cultures on days 2 and 5 after plating, as previously described [9]. After 7 days, immature (i)DC were recovered (purity > 95%) and exposed for an additional 18–22 h to an ICC that comprised human IL-1β (10 ng/mL), IL-6 (10 ng/mL), TNF-α (10 ng/mL) (all from R&D Systems) plus IFN-γ (1000 U/mL) (PrepoTech, Rocky Hill, NJ). Following 18–22 h exposure to ICC + IFN-γ, allogeneic DC were recovered, washed with large volume of complete media, counted and used in co-cultures as indicated.

Flow cytometric analysis

DC cell surface phenotype was determined by flow cytometric analysis using different combinations of fluorochrome-conjugated mAbs, including CD14, HLA-DR, CD86, PD-L1 (programmed cell death ligand-1; B7-H1, CD274) (all from BD Biosciences, San Jose, CA) and CD40 (Ancell, Bayport, MN). Small aliquots of cells were stained for 30 min at 4°C, washed twice with FACS buffer (phosphate-buffered saline supplemented with 1% FCS and 0.05% sodium azide) and fixed with 2% paraformaldehyde-containing FACS buffer (Sigma).

When indicated, intracellular staining was performed to measure IFN-γ production. Briefly, cells were surface stained with CD3, CD4, CD8, CD19 and CD56 mAbs (all from BD Biosciences), fixed, permeabilized with 0.1% saponin (Sigma)-containing FACS buffer, stained with IFN-γ (BD Biosciences) for 30 min at 4°C, then washed twice with FACS buffer. Data acquisition was performed immediately using an LSR II (BD Biosciences) and analyzed using FlowJo software (Tree Star, Ashland, OR).

RNase protection assay (RPA)

RPA with 32P-uridine triphosphate-labeled anti-sense RNA probes to IL-12p35, p40 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was performed as described [31]. Briefly, RNA was isolated from cryopreserved DC using a total RNA Isolation Kit (BD Biosciences). RPA was performed using the RiboQuant Multi-Probe RPA System (BD Biosciences) and cDNAs encoding human IL-12p35, -p40 and the housekeeping gene GAPDH as templates. Quantification of bands was performed by densitometric assessment of scanned auto-radiographs using Scion Image v1.63 software (NIH). The signals from specific mRNAs were normalized to the signals from housekeeping genes run in each lane to adjust for loading differences.

Short-term cell cultures and functional assays

Bulk PBMC, CD3+T cells or CD56+NK cells were exposed to IL-12p70 (10 ng/mL) and/or IL-27 (100 ng/mL) in the absence or presence of α-IL-12p70 (4 μg/mL) and/or α-IL-27 (4 μg/mL) blocking Abs for 24 h. Golgi-Plug (BD Biosciences) was added for the last 4 h of culture. Cells were then recovered, surface stained and permeabilized for detection of intracellular IFN-γ, as described above. Cell culture supernatants were recovered from cultures without the addition of Golgi-Plug and analyzed for cytokine release by enzyme-linked immunosorbent assay (ELISA).

Long-term cell cultures and functional assays

PBMC or NK-depleted PBMC were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE) (Invitrogen, Eugene, OR) and used as responders in mixed leukocyte reactions (MLR) as previously described [32]. Washed allogeneic matured DC were incubated with CFSE-labeled cells (1:10 ratio) for 5 days at 37°C in 5% CO2 air. Activated responder cells were restimulated or not for 4 h with phorbol myristate acetate (PMA) (5 ng/mL; Sigma) and ionomycin (100 ng/mL; Sigma) in the presence of Golgi-Plug (BD Biosciences). In some experiments, trans-well plates (0.4 μm, Corning, Inc., Corning, NY) were used, or blocking α-IL-12p70 (4 μg/mL) and/or α-IL-27 (4 μg/mL) Abs were added at the initiation of co-cultures.

ELISA and Luminex

IL-12p70, IL-27, IL-10 and IFN-γ were quantified using ELISA kits from BD Biosciences following the manufacturer's protocol. Additional cytokines (IL-1β, IL-2, IL-4, IL-6, IL-17, IL-18, IL-21, IL-23p19, TNF-α and IFN-α) were measured by luminex at the University of Pittsburgh Cancer Institute Biomarkers Core Facility.

Statistical analyses

Results are expressed as means ± SD, as indicated. The differences between means were determined using the paired and unpaired Student's ‘t-test’, Mann–Whitney test or Wilcoxon matched-pairs signed rank test, as appropriate. The results were considered statistically significant if p ≤ 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

mTORC1 inhibition interferes with DC differentiation, but not with DC maturation induced by ICC + IFN-γ

To determine the influence of RAPA on monocyte differentiation into DC and on their ability to acquire a mature cell surface phenotype, peripheral blood monocytes were cultured in DC medium, with or without RAPA for 7 days. Cells were recovered as immature (i)DC, then exposed to an ICC + IFN-γ for an additional 24 h. Figure 1A shows that while iRAPA-DC and iCTRL-DC exhibited similar mean fluorescence intensity (MFI) levels of HLA-DR and lacked CD14, iRAPA-DC displayed significantly lower levels of the co-regulatory molecules CD40, CD86 and PD-L1 compared to iCTRL-DC. Interestingly, mRAPA-DC up-regulated CD40 and CD86, and maintained similar levels of HLR-DR to those expressed on mCTRL-DC. However, mRAPA-DC maintained lower cell surface expression of PD-L1, an important co-inhibitory ligand that engages its receptor, PD-1, on T cells to deliver inhibitory signals for TCR-mediated activation (Figure 1B). These results demonstrate that human RAPA-DC are not resistant to ICC + IFN-γ-induced maturation.

image

Figure 1. Human RAPA-DC exhibit a more immature surface phenotype than CTRL-DC, but do not resist maturation following exposure to pro-inflammatory cytokines. CD14+ blood monocytes were cultured in GM-CSF and IL-4, in the presence (RAPA-DC) or absence of RAPA (CTRL-DC) for 7 days, as described in the Materials and Methods Section. Immature (i)DC were recovered and stimulated with a inflammatory cytokine cocktail (ICC: IL-1β, IL-6, TNF-α) + IFN-γ for additional 24 h. Representative histograms and overall MFI of CD14, HLA-DR, CD40 and CD86 (n = 15) and of PD-L1 (n = 9) staining are shown from (A) iCTRL-DC (solid lines and black symbols) compared to iRAPA-DC (dotted lines and open symbols) or (B) matured (m)CTRL-DC compared to mRAPA-DC. Each symbol represents a single subject, while the horizontal line represents the mean value.

Download figure to PowerPoint

mTORC1 inhibition increases IL-12p70 and IL-27 and lower IL-10 production by human DC exposed to ICC + IFN-γ

The role of mTORC1 in regulating pro-inflammatory (IL-12p70) or anti-inflammatory (IL-10) cytokine production by human monocyte-derived DC upon their maturation in response to different TLR agonists has been described [9, 15, 33]. Here, we addressed for the first time, the role of mTORC1 inhibition in modulating cytokine production by RAPA-DC exposed to ICC + IFN-γ. Comparative analysis of IL-12-p35 and -p40 mRNA expression was conducted on CTRL and RAPA-DC exposed to media (iDC) and inflammatory cytokines with or without IFN-γ. Results were quantified by densitometry of bands as illustrated in Figure 2A. DC exposed to ICC alone showed up-regulation of mRNA for IL-12p40 as compared to media exposed DC (iDC), while maturation with ICC + IFN-γ significantly increased IL-12p35 and IL-12p40 mRNA expression (Figure 2A, right panel). Overall, RAPA-DC displayed significantly higher levels of mRNA for IL-12p40, IL-12p35 as well as of bioactive IL-12p70 protein secretion compared to mCTRL-DC (Figure 2B). In addition, mRAPA-DC displayed significant increases in IL-27 production, while IL-10 secretion was inhibited significantly when compared to mCTRL-DC (Figure 2C and D). Moreover, there were no quantifiable differences in IL-2, IL-15, IL-17, IL-18, IL-21, IL-23p19 or IFN-α levels in supernatants of mRAPA-DC or mCTRL-DC (data not shown). Taken together, these results indicate that human mRAPA-DC exhibit a signature of Type-1 promoting cytokines (high IL-12p70, high IL-27 and low IL-10) upon ICC + IFN-γ stimulation.

image

Figure 2. mRAPA-DC secretes elevated levels of IL-12p70 and IL-27, but reduced levels of IL-10. Immature CTRL- or RAPA-DC were exposed to inflammatory cytokines (ICC: IL-1β, IL-6, TNF-α) ± IFN-γ for 24 h. mRNA and culture supernatants were collected for RNA protection assay and ELISA, respectively. (A) Left panel: Northern blot analysis revealed that inflammatory cytokines ± IFN-γ enhance the expression of IL-12p35 and IL-12-p40 in RAPA-DC. Right panel: Northern blots were scanned as described in the Materials and Methods Section and arbitrary (Arb) units were calculated by dividing the values of IL-12p35 or IL-12p40 band intensity to correspondent values of GAPDH band intensity (n = 3) (B) IL-12p70 (n = 17), (C) IL-27 (n = 18) and (D) IL-10 (n = 18) secretion by mCTR-DC (black symbols) compared to mRAPA-DC (open symbols). Each symbol represents a single subject, while the horizontal line represents the mean value.

Download figure to PowerPoint

Exogenous IL-12p70 stimulates early (24 h) IFN-γ and IL-10 production, while IL-27 triggers exclusively IL-10 release by NK cells

It is well-accepted that IL-12p70 and IL-27 are critical for the generation and maintenance of Type-1 immunity [34]. In addition, IL-12p70 is a potent stimulator of NK cells, triggering large amounts of IFN-γ production and thus amplifying Type-1 immunity [28, 35]. Recent studies have reported that IL-27 can increase the production of IFN-γ by NK cells, but only if combined with other cytokines, such as IL-2 and IL-12 [36]. Since our results indicated that mRAPA-DC produced comparatively large amounts of both IL-12p70 and IL-27, we tested the early influence of soluble cytokines on NK and T cells. Bulk PBMC or highly purified T or NK cells were cultured in the presence or absence of exogenous IL-12p70, IL-27 or IL-12p70 + IL-27 for 24 h. The cells were recovered and analyzed for intracellular IFN-γ production by flow cytometry, while supernatants were collected and tested for IFN-γ and IL-10 by ELISA. As shown in Figure 3A, NK cells, but not T cells were the main sources of IFN-γ production within the PBMC population in response to short-term stimulation with IL-12p70. Neither CD3+ T nor NK cells produced IFN-γ in response to exogenous IL-27 alone, while IL-12p70 + IL-27 stimulation resulted in similar levels of IFN-γ release by NK cells to those observed with IL-12p70 alone (Figure 3A). These findings were confirmed by ELISA (Figure 3B). In addition, α-IL-12p70 mAb significantly decreased IFN-γ secretion by both PBMC and NK cells (Figure 3B). Thus, NK cells proved to be the major source of early IFN-γ secretion within PBMC upon exogenous IL-12p70 stimulation for 24 h, while soluble IL-27 had no early effect on IFN-γ release. In addition, both soluble IL-12p70 and IL-27 triggered IL-10 secretion by PBMC and NK cells (Figure 3C). Interestingly, blocking IL-12p70 increased IL-10 secretion, whereas blocking IL-27 significantly decreased IL-10 production (Figure 3C), suggesting that IL-12p70 and IL-27 differentially regulate IL-10 secretion by PBMC.

image

Figure 3. Soluble IL-12p70 but not soluble IL-27 stimulates NK cells to produce IFN-γ in short-term culture (24 h). (A) PMBC were exposed to IL-12p70 (10 ng/mL), IL-27 (100 ng/mL) or their combination for 24 h in the presence of Golgi-Plug for the last 4 h, and then analyzed by flow cytometry. Dot plots show the gating strategy and the detection of IFN-γ production by CD3+ T and CD56+ NK cells. One representative experiment of >5 experiments performed is shown. (B) PBMC, purified (CD3+) T cells or (CD56+) NK cells were exposed to IL-12p70, IL-27 or the combined treatments in the presence or absence of blocking Abs (α-IL-12p70 and α-IL-27; or both at 4 μg/mL) for 24 h. Supernatants were recovered and analyzed for IFN-γ (n = 14) or for IL-10 (n = 8) by ELISA. Each symbol represents a single subject, while the horizontal line represents the mean value. *One-tail Wilcoxon matched-pair signed rank test.

Download figure to PowerPoint

mRAPA-DC promote early IFN-γ production and low IL-10 production by NK cells and T cells

To further ascertain the combined impact of cell-to-cell contact and soluble cytokine delivery by these APC, we co-cultured mRAPA-DC or mCTRL-DC with allogeneic PBMC in short-term MLR co-culture. Figure 4A shows a significant augmentation of IFN-γ secretion (left panel) and a significant decrease of IL-10 release (right panel) detected in mRAPA-DC/PBMC co-culture supernatants at 24 h, compared to those from mCTRL-DC/PBMC co-cultures. Our results further demonstrate that, in addition to soluble factors, IFN-γ production was dependent on cell-to-cell contact, as assessed in trans-well experiments in which DC and allogeneic PBMC were incubated in separate chambers (Figure 4B). To further dissect the role of IL-12p70 and IL-27 in IFN-γ release, co-cultures were set-up with PBMC, T cells and NK cells in the presence or absence of α-IL-12p70, α-IL-27 blocking Abs or their combination (Figure 4C). Although our data indicate that soluble IL-12p70 but not IL-27 did stimulate PBMC to produce IFN-γ during 24 h co-culture (as shown in Figure 3C), blockade of IL-12p70 or of IL-27 or the combined treatment at the initiation of mRAPA-DC co-cultures decreased IFN-γ secretion by PBMC, suggesting that cell-to-cell contact provided by mRAPA-DC was required for IL-27 to contribute to enhanced IFN-γ production by PBMC together with IL-12p70 (Figure 4C). More importantly, our data further demonstrate that cell-to-cell contact, in addition to cytokine-mediated stimulation delivered by mRAPA-DC, also triggered early IFN-γ production by T cells, and that this was mostly dependent on IL-12p70 (Figure 4C). Since naïve T cells do not produce cytokines until 36–48 h poststimulation, these early IFN-γ producers are considered memory T cells [37]. Thus, in a pro-inflammatory milieu, mRAPA-DC deliver early potent signals to target NK and memory T cells, and thus are rendered more effective ‘IFN-γ helpers’ for the subsequent development of Type-1 immunity.

image

Figure 4. mRAPA-DC promote early IFN-γ production by NK and memory T cells. mCTRL- or mRAPA-DC were co-cultured for 24 h with allogeneic PBMC (1:10 ratio). (A) Left panel: overall IFN-γ concentrations in the supernatants (n = 32); right panel: IL-10 release by mCTRL- or mRAPA-DC co-cultured with allogeneic PBMC (n = 7). Each symbol represents a single subject, while the horizontal bar represents the mean value (B) IFN-γ concentration in co-cultures of mRAPA-DC separated from allogeneic PBMC by trans-well chambers or (C) mRAPA-DC co-cultured with allogeneic PBMC (left panel), CD3+ T cells (middle panel) or CD56+ NK cells (right panel) in the presence of blocking α-IL-12p70 (4 μg/mL), α-IL-27 (4 μg/mL) or combined treatments. Each symbol represents a single subject, while horizontal lines represent mean values.

Download figure to PowerPoint

mRAPA-DC promote late (5 days) Type-1 polarization in allogeneic T cells

The ability of mRAPA-DC to up-regulate co-stimulatory molecules and secrete enhanced levels of IL-12p70 and IL-27 upon exposure to ICC + IFN-γ suggests their augmented potential to promote Type-1 polarization of T cells. To test this, we set up co-cultures of mDC and CFSE-labeled allogeneic PBMC for 5 days. Our results reveal increased overall IFN-γ production by allogeneic CD4+ and CD8+ T cells in response to mRAPA-DC compared to mCTRL-DC (Figure 5A and B). This enhanced IFN-γ production was dependent on cell-to-cell contact, as was evident when mCTRL-DC or mRAPA-DC were separated from responder allogeneic PBMC by trans-well chambers (Figure 5C) and not on soluble factors (Figure 5D). Thus, blocking IL-12p70 and/or IL-27 at the initiation of co-cultures had no significant effect on IFN-γ production by either allogeneic CD4+ or CD8+ T cells (Figure 5D). Moreover, using the CFSE dilution assay to distinguish proliferating from nonproliferating T cells (Figure 5E), we found that mRAPA-DC could effectively instruct Type-1 polarization and thus significantly augment IFN-γ production by allo-reactive CD8+ T cells, as well as by nonproliferating, bystander allogeneic CD4+ and CD8+ T cells, compared with mCTRL-DC (Figure 5F).

image

Figure 5. mRAPA-DC promote late IFN-γ production by allo-reactive and memory T cells. mCTRL- or mRAPA-DC were co-cultured for 5 days with CFSE-labeled allogeneic PBMC (1:10 ratio). On day 5, PBMC were restimulated with PMA (5 ng) and ionomycin (100 ng) in the presence of Golgi-Plug for 4 h then analyzed by flow cytometry for IFN-γ production by CD4+ and CD8+ T cells. (A) One representative experiment is shown to depict the gating strategy and the detection of IFN-γ production by allogeneic CD4+ and CD8+ T cells. (B) Overall % of IFN-γ producing allogeneic CD4+ and CD8+ T cells co-cultured with mCTRL-DC or mRAPA-DC (n = 23). Each symbol represents a single subject, while horizontal bars represent mean values. (C) The % of IFN-γ producing allogeneic CD4+ and CD8+ T cells when mCTRL-DC or mRAPA-DC were separated from allogeneic PBMC by trans-wells at the initiation of co-cultures (n = 7). (D) Top panels: Addition of neutralizing antibody α-IL-12p70 to the initiation of co-cultures (n = 18). Bottom panels: Addition of neutralizing Abs α-IL-27 or α-IL-12p70 + α-IL-27 (n = 3) at the initiation of co-cultures. (E) One representative experiment is shown to depict the gating strategy and the detection of IFN-γ production in CFSEdim proliferating (allo-reactive) and CFSEhigh nonproliferating (allogeneic) CD4+ and CD8+ T cells stimulated by mRAPA-DC for 5 days. (F) Overall % of IFN-γ producing proliferating and nonproliferating CD4+ and CD8+ T cells in response to mCTRL-DC or mRAPA-DC stimulation (n = 25). Each symbol represents a single subject, while horizontal bars represent mean values. Dark symbols represent experiments with CTRL-DC, whereas open symbols represent those with RAPA-DC. *One-tail Wilcoxon matched-pair signed rank test.

Download figure to PowerPoint

NK cells modulate Type-1 T cell polarization promoted by mRAPA-DC

To further determine the role of NK cells during mRAPA-DC/allogeneic PBMC ‘cross-talk’, we co-cultured bulk PBMC and/or NK-depleted PBMC with mDC for 5 days (Figures 6 and 7). Depletion of NK cells at the initiation of co-cultures resulted in significantly decreased IFN-γ production by both CD4+ and CD8+ T cells in ∼60% of the co-cultures tested, suggesting a ‘stimulatory’ NK effect during Type-1 polarization (Figures 6A and B and 7A). Conversely, for the remaining 40% of co-cultures, NK cell depletion resulted in augmented IFN-γ secretion by CD4+ and CD8+ T cells, suggesting a ‘regulatory’ rather than a ‘stimulatory’ role for NK cells (Figures 6C and 7B). Interestingly, adding back NK cells to trans-well chambers (to prevent contact between NK cells and mRAPA-DC/allogeneic T cells) at the initiation of co-cultures where NK cells were ‘regulatory’ in corresponding experiments resulted in a complete IFN-γ inhibition. This indicated that the regulatory NK effect was dependent mainly on soluble factor(s), and that NK-DC-T cell contact, when present, can override the influence of soluble inhibitory cytokine(s) (Figures 6E and 7). On the other hand, adding back NK cells in trans-well chambers within the co-cultures in which they exerted a ‘stimulatory’ effect in corresponding experiments, did not further decrease IFN-γ secretion by CD4+ and CD8+ T cells, suggesting that NK do not release inhibitory soluble factor(s) and that cell-to-cell contact together with helper cytokines were the main factors promoting IFN-γ secretion by T cells, (Figures 6D and 7). Taken together, these data suggest that integration of cell-to-cell contact and a balance between inhibitory soluble factor(s) (IL-10) and helper cytokine (IFN-γ) are critical early signals delivered by NK cells that modulate mRAPA-DC/allogeneic T cell cross-talk and thus downstream T cell polarization (Figure 7).

image

Figure 6. NK cells modulate IFN-γ production by T cells exposed to mRAPA-DC. Allogeneic bulk PBMC or CD56+ NK-depleted PBMC were exposed to mCTRL-DC or mRAPA-DC for 5 days, then recovered and restimulated with PMA and ionomycin, in the presence of Golgi-Plug for 4 h. CD4+ and CD8+ T cells were analyzed by flow cytometry for IFN-γ production. (A) Overall comparison of % IFN-γ positive allo-reactive CD4+ and CD8+ T cells in co-cultures depleted or not of CD56+ NK cells (n = 12). (B–C) Representative experiments in which CD56+ NK cell depletion resulted in (B) decreased % IFN-γ (stimulatory) (subjects 1–3) or (C) enhanced % IFN-γ (regulatory) (subjects 4, 5) effects. The IFN-γ fold change was calculated for these representative co-culture experiments in which mRAPA-DC/allogeneic PBMC depleted of CD56+ NK cells (gray columns) were supplemented back with CD56+ NK but separated in a trans-well chamber (open columns) in examples when CD56+ NK cells (D) stimulatory or (E) ‘regulatory’ effects are shown. The horizontal lines in D and E highlight the baseline IFN-γ values from bulk co-cultures. Each symbol represents a single subject, while horizontal bars represent mean values. Dark symbols represent experiments with CTRL-DC, whereas open symbols represent those with RAPA-DC. *One-tail Wilcoxon matched-pair signed rank test.

Download figure to PowerPoint

image

Figure 7. Proposed model for the interaction between mRAPA-DC, Type-1 T cells and NK cells. (A) mRAPA-DC up-regulate MHC Class I and II, co-stimulatory molecules as well as IL-12p70 and IL-27 production. mRAPA-DC cross-talk with allogeneic NK cells via cytokines and various surface molecules and render NK cells to produce IFN-γhigh/IL-10low, thus increasing Type-1 (IFN-γ) polarization of T cells after 5 days in co-culture. (B) Regulatory NK cells receive from mRAPA-DC IL-12p70 and IL-27 instructions in the absence of cell-to-cell contact and release soluble factor(s) (high IL-10) that profoundly inhibit Type-1 T cell polarization. These suggest that DC-NK contact override inhibitory cytokine(s) signal.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

A goal of our laboratory is to define the molecular and functional pathways in DC that are impacted by mTOR inhibition. Here we tested for the first time the responsiveness of human RAPA-DC generated from a cohort of healthy individuals to an ICC that mimic factors encountered by immature DC in peripheral tissues during viral infections or episodes of transplant rejection in patients on RAPA therapy. We incorporated IFN-γ in the inflammatory cocktail as a second signal for DC maturation because of its role in these inflammatory disorders, and since it has been shown that IFN-γ has the unique capacity to prime human DC into DC1 to produce high levels of IL-12p70 [38, 39]. Using this cytokine cocktail, our phenotypic analyses showed that RAPA-DC could up-regulate HLA-DR, CD40 and CD86 to the same levels as those expressed on mature CTRL-DC. In addition, when compared to CTRL-DC, immature or mature RAPA-DC exhibited low expression of PD-L1 (B7-H1), a PD-1 ligand that negatively regulates T cell activation and can promote peripheral tolerance [40]. Other reports have demonstrated a similar capacity of mTORi to modulate the phenotype of monocyte-derived DC. Although the timing of exposure of these APC to RAPA has varied between studies, the overall conclusion is that human RAPA-DC is not resistant to maturation.

Recent rodent and human studies have shown that RAPA-DC have a unique cytokine production profile upon stimulation with TLR agonists (e.g. lipopolysaccharide), that is increased secretion of IL-12 due to inhibited GSK-3, while IL-10 is reduced [9, 41]. Other reports have shown that mTORi controls IL-12p40 production by human monocyte-derived DC after TLR-dependent and -independent stimulation, due to enhanced NF-κB activity [14, 15]. Here we demonstrate that, when RAPA-DC are exposed to ICC + IFN-γ, not only IL-12p40 but also IL-12p35 is up-regulated, resulting in increased bioactive IL-12p70 secretion, paralleled by decreased IL-10. In addition, for the first time we provide direct evidence that exposure to inflammatory cytokines including IFN-γ elicits significantly increased IL-27 secretion by mRAPA-DC compared to mCTRL-DC. These findings may have important clinical relevance, since DC-derived IL-12p70 is regarded as the key cytokine for induction of Type-1 immunity (e.g. IFN-γ) in both NK cells and CD8+ effector T cells [37]. In addition, IL-27 is also a member of the IL-12 cytokine family, which acts synergistically with IL-12p70 to promote Type-1 responses [34, 42]. IL-27 also has a nonredundant role in selectively increasing T cell proliferation and cytotoxicity [34, 42]. However, IL-27 can also exert an inhibitory effect on Th1, Th2 and Th17 cells in pathogen-induced disease models, due to its ability to induce IL-10 production by T cells [43, 44]. Our data show that soluble IL-27 renders NK cells to produce IL-10, whereas cell-to-cell contact plus IL-27 may trigger NK to also produce IFN-γ and decrease IL-10. Thus, the role of IL-27 is more complex, due to its concomitant stimulatory and/or regulatory effects making such observations worthy of future investigation.

Based on the complex helper/regulatory cytokine and co-stimulatory/co-regulatory molecule expression profile, we postulated that mRAPA-DC would have unique abilities to integrate and deliver heterogeneous cell-to-cell contact (signal 1 and 2) and IL-12/IL-27 (signal 3) signals that are critically required for instructing allogeneic NK to modulate T cells to mount IFN-γ responses. In addition, the lower IL-10 secretion by RAPA-DC may favor Type-1 T cell polarization since IL-10 cross-regulates IL-12p70 and vice versa [45]. Similar to early IFN-γ ‘helper’ responses that facilitate the initiation and maintenance of Type-1 anti-viral or anti-cancer T cell immunity [28, 37, 46] those mediated by NK and memory T cells stimulated by mRAPA-DC appear to also modulate Type-1 allo-immunity. Although these responses were always higher than those induced by mCTRL-DC, they varied in strength between healthy individuals.

Increased IFN-γ production by allogeneic CD4+ and CD8+ T cells was observed in 5-day MLR co-cultures with mRAPA-DC compared to mCTRL-DC. This observation contrasts with results reported by others, who have found either a significant decrease or no difference in IFN-γ release by T cells when stimulated by allogeneic mRAPA-DC [16, 17]. These discrepancies may result from differences in RAPA-DC generation and maturation protocols implemented by different investigators. Interestingly, none of these studies measured levels of IL-12p70 production by mRAPA-DC. Our data demonstrate augmented Type-1 responses by allogeneic CD4+ and CD8+ T cells. Notably, we have found that mRAPA-DC generated ex vivo from kidney transplant recipients on RAPA monotherapy significantly augment IFN-γ secretion by allogeneic CD4+ and CD8+ T cells (Macedo et al., unpublished observations), consistent with our in vitro finding. Furthermore, the Type-1 polarization of mRAPA-DC-allogeneic PBMC 5 days co-cultures was highly dependent on cell-to-cell contact while soluble factors only minimally interfered with IFN-γ secretion by CD4+ or CD8+ T cells. These data suggest that DC/T/NK cell interactions may stabilize the immunologic synapse during late T cell activation [46, 47], thus, making it difficult to interfere with the action of soluble factors. Moreover, among the proliferating (allo-reactive) CFSEdim T cells, only CD8+ T cells exhibited significantly higher IFN-γ secretion following mRAPA-DC stimulation compared to mCTRL-DC stimulation. This difference may be due to the preferential role of IL-27 in priming naïve CD8+ T cells rather than CD4+T cells into Type-1 effectors in this setting. Conversely, mRAPA-DC may only be able to reactivate bystander memory CD8+ T cells specific to recall Ag (pathogen-specific) rather than to prime allo-reactive CD4+ and CD8+ T cell responses, as reported [20, 47]. Thus, the increased IFN-γ production observed within CFSEdim proliferating CD8+ T cells may represent reactivation of heterologous, anti-viral memory CD8+ T cells that cross-react with human MHC class I allo-Ags rather than de novo priming of allo-reactive CD8+ T cells [32, 48].

Our data show, for the first time, that human NK cells are targets of mRAPA-DC stimulation in vitro. This finding is consistent with that of Brouard et al. [19], who showed that the peripheral blood transcriptional profile induced by RAPA monotherapy in stable kidney transplant patients was dominated by pro-inflammatory features of innate immune cells, including NK cells. The present observations further reveal that mRAPA-DC instruct allogeneic NK cells to convey either stimulatory or regulatory signals to allogeneic T depending on the responder/stimulator combination pairs. In literature, the precise role of human NK cells in organ transplantation is unclear. NK cells were shown to integrate complex stimulatory (NKp46, NKp30, NKG2D) and inhibitory (KIRs, CD94/NKG2A) signals combined with the release of diverse cytokines [49, 50]. In general, NK cells are considered rapid initiators of a pro-inflammatory milieu that promotes the licensing of DC and T cells into Type-1-polarized effectors, able to mediate acute or/and chronic allograft injury [50]. Recent findings, however, have indicated that NK cells can also promote allograft tolerance, with DC and T cells serving as targets of NK cell killing as a result of ‘missing self’ [49-51]. While here we report the ability of human mRAPA-DC to instruct NK cells to produce increased IFN-γ, our findings add to the list of potential roles for NK (i) triggering receptors which further promote Type-1 help to allogeneic T cells in certain individuals and (ii) release of regulatory cytokine(s) that regulates allogeneic T cells by lowering their IFN-γ production in others. This effect may be mediated by IL-10, since our results and recently published data have shown IL-27 to induce IL-10 production by NK cells [52]. Therefore future studies are required to identify which NK regulatory cytokine(s) and/or activating receptors are relevant during the complex DC/NK/T cell interactions in the setting of allo-recognition.

In conclusion, we have characterized the phenotype and cytokine secretion profile of human mRAPA-DC generated as the result of exposure to inflammatory cytokines including IFN-γ. We demonstrated their marked influence on both allo-reactive NK and T cell functions. These data reveal both immune stimulatory and regulatory properties of NK cells stimulated by mRAPA-DC, similar to the dichotomous functions proposed for NK cells in relation to transplant outcome. In total, out findings indicate that clinical mTORi use in transplantation may be complicated by conferring DC a novel capacity to drive NK cell support of Type-1 immunity, especially in patients where NK cells serve a stimulatory function, or lack regulatory properties.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

We thank Beth Elinoff and Sheila Fedorek for their contributions as research coordinators. This work was supported by the National Institutes of Health (NIH) grants R01AI67541S1 (A.W.T. and D.M.), R01 HL094603 (D.M.), P0181678 (A.W.T.), R00HL097155 (H.R.T.) and diversity supplement (R01 AI67541 to M.C.-R.).

Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References