Effect of Oxygen Levels on the Physiology of Dendritic Cells: Implications for Adoptive Cell Therapy
Dendritic cell (DC)-based adoptive tumor immunotherapy approaches have shown promising results, but the incidence of tumor regression is low and there is an evident call for identifying culture conditions that produce DCs with a more potent Th1 potential. Routinely, DCs are differentiated in CO2 incubators under atmospheric oxygen conditions (21% O2), which differ from physiological oxygen levels of only 3–5% in tissue, where most DCs reside. We investigated whether differentiation and maturation of DCs under physiological oxygen levels could produce more potent T-cell stimulatory DCs for use in adoptive immunotherapy. We found that immature DCs differentiated under physiological oxygen levels showed a small but significant reduction in their endocytic capacity. The different oxygen levels did not influence their stimuli-induced upregulation of cluster of differentiation 54 (CD54), CD40, CD83, CD86, C-C chemokine receptor type 7 (CCR7), C-X-C chemokine receptor type 4 (CXCR4) and human leukocyte antigen (HLA)-DR or the secretion of interleukin (IL)-6, tumor necrosis factor (TNF)-α and IL-10 in response to lipopolysaccharide (LPS) or a cytokine cocktail. However, DCs differentiated under physiological oxygen level secreted higher levels of IL-12(p70) after exposure to LPS or CD40 ligand. Immature DCs differentiated at physiological oxygen levels caused increased T-cell proliferation, but no differences were observed for mature DCs with regard to T-cell activation. In conclusion, we show that although DCs generated under atmospheric or physiological oxygen conditions are mostly similar in function and phenotype, DCs differentiated under physiological oxygen secrete larger amounts of IL-12(p70). This result could have implications for the use of ex vivo-generated DCs for clinical studies, since DCs differentiated at physiological oxygen could induce increased Th1 responses in vivo.
Dendritic cells (DCs) are the most potent antigen-presenting cells (1,2) and are critical for the induction of immune responses to pathogens and cancer (1,3). Because of these properties, DCs are being widely used for vaccines and immunotherapeutic strategies (3, 4, 5, 6, 7, 8, 9, 10). The most common approaches for tumor immunotherapy involve the use of DCs generated from the progenitors CD34+ (11, 12, 13) or CD14+ in the presence of granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin (IL)-4 ex vivo (14). Tumor antigens are delivered to DCs using many different systems including whole tumor cells or lysates (15, 16, 17), RNA (18, 19, 20, 21), peptides or viral vectors (22). Exposure to antigen is followed by the addition of a maturation stimulus in vitro, since mature DCs induce more potent immune responses than immature DCs, which can induce T-cell tolerance (23,24). One of the originally used maturation stimuli was monocyte-conditioned media (14), which was subsequently refined to a cocktail containing four components including prostaglandin E2 (PGE2), tumor necrosis factor (TNF)-α, IL-1β and IL-6 (8),25, 26, 27).
Although immune responses are generated to DC-based vaccine approaches in patients, only in a few instances has tumor regression been observed in phase I clinical trials (28,29). Clearly, this result could have many explanations including the DC subtypes used, the maturation stimuli, the suppressive tumor microenvironment, treatment starting too late in the disease, and other explanations. An alternative possibility is that the ex vivo-generated DCs are not optimal for in vivo function because of the applied culture conditions. Optimization of the in vitro culture conditions should allow for the generation of DCs that will give the desired and maximal Th1-type immune response in vivo. In studies using ex vivo differentiated DCs, the DCs are generally differentiated in incubators that maintain atmospheric oxygen levels (21% O2 and 5% CO2). However, in vivo, most cells including DCs do not encounter such high oxygen levels. DCs exist both in blood and tissue, and most of them reside in the tissue where the oxygen levels are 3–5% (30, 31, 32). The effect of physiological and atmospheric oxygen levels has recently been compared with regard to its effect on T cells (32, 33, 34). At physiological oxygen levels, primary T cells proliferated less in response to CD3/CD28 stimulation, which correlated with higher intracellular nitric oxide levels (33). Furthermore, cytotoxic T cells developed under 2.5% O2 were more lytic but secreted lower amounts of IL-2 and interferon (IFN)-γ (32). Although the effect of hypoxia on DCs has been investigated (35,36), since this is of interest to understand the effect of hypooxygeneation on, for example, DCs in tumor tissue, no data exist on the effect of physiological oxygen levels on the differentiation of human DCs from progenitors and their maturation. Thus, we compared functionally and phenotypically monocytederived DCs that have been differentiated under physiological oxygen with those differentiated under atmospheric oxygen conditions with a goal to evaluate if we can generate more potent DCs for tumor immunotherapy.
Materials and Methods
HeLa cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). HeLa cells that express human CD154 (HeLa-CD154) were generated by electroporation of an expression vector containing the full-length human CD154 into HeLa cells, followed by selection and cloning as described (37).
Generation of Human Monocyte-Derived DCs
Peripheral blood mononuclear cells (PBMCs) were isolated from the blood of normal volunteers (San Diego Blood Bank) over a Ficoll-Hypaque (Amersham Biosciences, Uppsala, Sweden) density gradient. To generate DCs, PBMCs were allowed to adhere to culture plates for 1 h. The nonadherent cells were washed off, and the adherent cells were cultured in RPMI 1640 medium supplemented with 2 mmol/L L-glutamine (GIBCO-BRL Life Technologies, Grand Island, NY, USA), 50 µmol/L 2-mercaptoethanol (Sigma, St. Louis, MO, USA), 10 mmol/L HEPES (GIBCO-BRL), penicillin (100 U/mL)-streptomycin (100 µg/mL) (GIBCO-BRL) and 5% human serum (HS; Human AB serum, Gemini Bio Products, West Sacramento, CA, USA), supplemented with 1,000 units GM-CSF/mL (Bayer HealthCare Pharmaceuticals, Wayne, NJ, USA) and 200 units IL-4/mL (R&D Systems, Minneapolis, MN, USA) at days 0, 2 and 4. Immature DCs were harvested on days 5–7. If not otherwise, the DCs were differentiated starting from day 0 under two different oxygen levels. Physiological oxygen tensions, or 5% O2, were generated in a Sanyo MCO-18M O2/CO2 incubator (Sanyo Scientific, Bensenville, IL, USA). Gas phase O2 levels were controlled by continuous injection of medical grade N2 to reach the target oxygen level. DCs cultured in atmospheric oxygen levels (20% O2) were incubated in a standard incubator without the addition of N2. All cells were exposed to 5% CO2.
Immature DCs were collected on day 7 and incubated with 1 mg/mL Dextran-FITC (Molecular Probes, Eugene, OR, USA) for 30 min at 4°C or 37°C. The caps of the tubes were left opened to ensure that the cells were exposed to the respective oxygen levels. DCs were washed twice with 5% HS/RPMI, fixed in 3.7% formaldehyde in phosphate-buffered saline (PBS) (pH 7.2–7.4) and analyzed by flow cytometry using a FACSCalibur (Beckon Dickinson, San Jose, CA, USA). Data were analyzed using the FlowJo 7.2.2 software (Tree Star, Ashland, OR, USA).
Stimulation of DCs
At day 5 of culture, immature DCs were either left untreated (immature [IM]), stimulated with indicated doses of lipopolysaccharide (LPS) (E. coli serotype 026:B6; Sigma) or a cocktail of cytokines (CyC) consisting of TNF-α at 10 ng/mL, IL-1β at 10 ng/mL (all from R&D Systems) and PGE2 at 1 µg/mL (Sigma). At 48 h after stimulation, cell-free culture supernatants were collected, and cytokines were measured by enzyme-linked immunosorbent assay (ELISA) (eBioscience, San Diego, CA, USA).
Analysis of DC Phenotype
The 1 × 104 DCs were incubated for at least 20 min at 4°C in 100 µL PBS/5% fetal calf serum/0.1% sodium azide (staining buffer) with phycoerythrin (PE)-conjugated IgG specific for cluster of differentiation 54 (CD54), human leukocyte antigen (HLA)-DR (all from Becton Dickinson Immunostaining Systems, San Jose, CA, USA), CD83 (Immunotech-Beckman-Coulter, Marseille, France) and CD184 (C-X-C chemokine receptor type 4 [CXCR4]) (BD Biosciences, Franklin Lakes, NJ, USA) or fluorescein isothiocyanate (FITC)-conjugated IgG monoclonal antibody (mAb) specific for CD40 and CD58 (all from Becton Dickinson Immunostaining Systems). Cells were washed 4× with staining buffer, fixed in 3.7% formaldehyde in PBS and examined by flow cytometry using a FACScan(Calibur) (BD Biosciences). In all experiments, isotype controls were included using PE- or FITC-conjugated irrelevant mAb of the same Ig class.
T cells were isolated by negative selection using the RosetteSep antibody cocktail from StemCell Technologies (Vancouver, CA, USA) according to the manufacturer’s instructions. The purity of the isolated T cells was routinely ~99%.
Mixed Leukocyte Reaction
To assess levels of cellular activation and proliferation, cells were plated at 2 × 105 cells per well in a flat-bottomed 96-well tray at DC:T-cell ratios of 1:10 for 5 d in medium described above. T-cell proliferation was measured using the CellTrace CFSE Cell Proliferation Kit from Invitrogen (Eugene, Oregon, USA). Cells were stained in 0.1% bovine serum albumin/PBS for at least 10 min in a 37°C water bath, washed 3× with culture media and then plated. Cells were harvested on days 2 and 5, fixed in 10% formaldehyde in PBS and analyzed by flow cytometry.
Activation of DCs Using CD154-Expressing Hela Cells
Immature DCs were exposed to media or LPS at day 5 for 24 h. On day 6, immature and LPS matured DCs were collected and cocultured with HeLa or HeLa-CD154 cells. HeLa cells were plated in a flat-bottomed 96-well plate at 1.5 × 104 cells per well 24 h before addition of 2.5 × 105 DCs. After 12 h, cell culture supernatants were collected and IL-12(p70) was measured by ELISA. HeLa-only controls did not contain measurable levels of IL-12.
Ninety-six-well polyvinylidene fluoride (PVDF) plates (Millipore, Billerica, MA, USA) were coated overnight at 4°C with 5 µg/mL of the primary anti-human IFN-γ mAb (Mabtech, Cincinnati, OH, USA). The antibody-coated plates were washed 5× with PBS and blocked with RPMI 1640 containing 5% human serum for 1 h at 37°C. Immature and mature DCs were pulsed with 100 ng/mL MART-1 peptide (ELAGIGILTV) for 1 h at 37°C. Unpulsed and peptide-pulsed DCs were cocultured with MART-1 specific T cells (generated in our laboratory from normal donor PBMCs) at a ratio of 103 DCs:105 T cells and incubated overnight (approximately 16–18 h) at 37°C. Enzyme-linked immunosorbent spot (ELISPOT) plates were washed 5× with PBS containing 0.05% Tween-20 followed by a 2-h incubation at room temperature with 1 µg/mL biotinylated anti-IFN-γ mAB (Mabtech). Plates were washed 5× in PBS with 0.1% Tween-20. Streptavidin-horseradish peroxidase (1:500) was added to wells and incubated for 1 h at room temperature. The plates were washed 5× in PBS with 0.1% Tween-20 and then 2× in PBS only, followed by a 2- to 3-min incubation in tetramethylkbenzidine (TMB [Mabtech]) to develop the reaction. Plates were washed with tap water to stop the reaction. IFN-γ-secreting T cells were counted using an automated image analysis ELISpot reader (ImmunoSpot Series 1 Analyzer; Cellular Technology, Cleveland, OH, USA).
Data are represented as mean ± SD. Data were analyzed for statistical significance using a Student t test. P values <0.05 were considered statistically significant.
Immature DCs Differentiated Under Physiological Oxygen Levels Show Decreased Endocytic Activity
The number of DCs recovered per well at atmospheric and physiological oxygen levels.
3.3 × 105± 2.2 × 105
4.1 × 105± 2.7 × 105
5.5 × 105± 3.6 × 105
4.5 × 105± 1.9 × 105
4.4 × 105± 2.4 × 105
5.4 × 105± 2.7 × 105
DCs Differentiated and Matured Under Atmospheric and Physiological Oxygen Levels Show Similar Expression of Maturation Markers
Mature DCs Differentiated Under Physiological Oxygen Secrete Larger Amounts of IL-12
DCs Differentiated Under Physiological or Atmospheric Oxygen Levels Induce Similar Levels of T-Cell Activation
To examine whether antigen-specific CD8+ T-cell activation was affected by DCs differentiated under different oxygen conditions, we used HLA-A*0201-positive donors. Immature DCs, LPS-DCs and CyC-DCs generated at the different oxygen conditions were pulsed with MART-1 peptide and cocultured with a MART-1-specific CD8+ T-cell clone for 18 h, and secretion of IFN-γ was measured by ELISPOT (see Figure 4C). The result shows that oxygen levels had no influence on the assay. No difference was observed between immature physO2 or atmosO2 DCs or mature physO2 or atmosO2 DCs with regard to their ability to induce MART1-specific T-cell responses.
Because DCs are the most potent initiators of antigen-specific T-cell responses, they have been applied toward immunotherapy of cancer and chronic infectious diseases. For DC-based adoptive immunotherapy, DCs are differentiated ex vivo from CD14+ or CD34+ progenitors, pulsed with tumor antigen, exposed to a maturation stimulus and reinjected back into patients (7,40). For these studies, DCs were differentiated in CO2 incubators containing atmospheric oxygen levels of 20–21%. However, these levels are two to four times higher than physiological oxygen levels, which are approximately 12% in arterial blood and 3–5% in tissue (30, 31, 32), where most DCs reside.
In light of a recent finding showing altered T-cell responses at atmosO2 compared with physO2 (33), we sought to determine if differentiation of DCs from CD14+ progenitors under physiological oxygen levels would alter their physiology and antigen-presenting capacity, with the hope to identify improved conditions for DC-based studies. Surprisingly, no difference in expression of the surface molecules CD54, CD40, CD83, CD86, HLA-DR, CXCR4 and CCR7 was observed in immature DCs or mature DCs. In addition, the secretion of TNF-α, IL-6 and IL-10 from either immature or mature DCs did not differ between the different oxygen culture conditions.
We found that LPS-matured physO2 DCs secreted higher levels of IL-12(p70) than LPS-matured atmosO2 DCs. IL-12 is an important cytokine mediating Th1 polarization of CD4+ T cells, which provides help for the activation of cytotoxic CD8+ T cells. Although physO2 DCs did not elicit increased CD8+ T-cell responses in vitro, as measured by IFN-γ secretion, this could be due to the use of a CD8+ T-cell clone, which is easier to activate than, for instance, naive T cells. Another possibility is that the levels of IL-12 secreted in the two different oxygen conditions were sufficiently high to support T-cell activation; thus, no difference could be observed. The increased capacity of immature physO2 DCs to activate allogeneic T cells could be due to their increased secretion of IL-12 after ligation with CD40L on the T cells, as seen when DCs were stimulated by CD40L.
We found that the cytokine cocktail-matured DCs, when generated under physiological or atmospheric oxygen conditions, did not differ with regard to their ability to induce proliferation of allogeneic T cells or activation of antigen-specific CD8+ T cells. It was recently shown that cytokine cocktail-matured DCs can also induce immunosuppressive regulatory T cells (Tregs) (41), and it remains to be tested what effect oxygen levels will have on the induction of Tregs by DCs.
Overall, we show that DCs differentiated and stimulated with LPS or CD40L under physO2 conditions secreted higher amounts of IL-12(p70), suggesting that it may be more beneficial to culture and activate DCs at physO2 levels because higher IL-12 secretion may cause a more pronounced T-cell response in vivo. Overall, only a few differences were observed owing to oxygen levels and other parameters, such as the DC subtype, the maturation stimulus, or the presence of the immunosuppressive microenvironment of the tumor are more likely influencing the poor clinical outcome of the currently used adoptive DC immunotherapy protocols.
The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
This work was supported by the U.S. Army Medical Research and Materiel Command under agreement number W81XWH-07-1-0412 to D Messmer, the Swedish Research Council AI52731 and the Swedish International Development Cooperation Agency (SIDA and VINNMER [Vinnova]) to M Larsson. The authors thank Jessie F Fecteau and Dan Seible for the critical reading of the manuscript.
- 19.Van Tendeloo VF, et al. (2001) Highly efficient gene delivery by mRNA electroporation in human hematopoietic cells: superiority to lipofection and passive pulsing of mRNA and to electroporation of plasmid cDNA for tumor antigen loading of dendritic cells. Blood. 98:49–56.CrossRefPubMedGoogle Scholar