SEARCH

SEARCH BY CITATION

Keywords:

  • antigen presentation;
  • chemokines;
  • dendritic cell;
  • program;
  • T cell;
  • vaccine efficiency

Summary

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

In a companion article to this study,1 the successful programming of a JAWSII dendritic cell (DC) line's antigen uptake and processing was demonstrated based on pre-treatment of DCs with a specific ‘cocktail’ of select chemokines. Chemokine pre-treatment modulated cytokine production before and after DC maturation [by lipopolysaccharide (LPS)]. After DC maturation, it induced an antigen uptake and processing capacity at levels 36% and 82% higher than in immature DCs, respectively. Such programming proffers a potential new approach to enhance vaccine efficiency. Unfortunately, simply enhancing antigen uptake does not guarantee the desired activation and proliferation of lymphocytes, e.g. CD4+ T cells. In this study, phenotype changes and antigen presentation capacity of chemokine pre-treated murine bone marrow-derived DCs were examined in long-term co-culture with antigen-specific CD4+ T cells to quantify how chemokine pre-treatment may impact the adaptive immune response. When a model antigen, ovalbumin (OVA), was added after intentional LPS maturation of chemokine-treated DCs, OVA-biased CD4+T-cell proliferation was initiated from ~ 100% more undivided naive T cells as compared to DCs treated only with LPS. Secretion of the cytokines interferon-γ, interleukin-1β, interleukin-2 and interleukin-10 in the CD4+ T cell : DC co-culture (with or without chemokine pre-treatment) were essentially the same. Chemokine programming of DCs with a 7 : 3 ratio of CCL3 : CCL19 followed by LPS treatment maintained partial immature phenotypes of DCs, as indicated by surface marker (CD80 and CD86) expression over time. Results here and in our companion paper suggest that chemokine programming of DCs may provide a novel immunotherapy strategy to obviate the natural endocytosis limit of DC antigen uptake, thus potentially increasing DC-based vaccine efficiency.


Introduction

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

Among antigen-presenting cells, which stimulate adaptive immunity in the host immune response, dendritic cells (DCs) are recognized as the most potent tool for emerging immunotherapy vaccines.[2, 3] In vaccines, certain tumor-derived or infected cell-derived antigens can be loaded ex vivo into DCs in combination with adjuvants that induce DC maturation, and then mature DCs (mDCs) are injected back into the host to stimulate T cells in vivo, thereby inducing adaptive immunity through T-cell activation.[4-6] More recently, intense research efforts have focused on the delivery to DCs of genes (as mRNA or DNA) that encode for specific antigens.[7-9] Research on effective antigen loading or non-viral gene delivery systems to DCs has been achieved by targeting specific receptors on DCs (mediating internalization), such as: by mannose-decorated pDNA polyplexes;[10] direct antigen fusion with a single chain Fv antibody against the DC phagocytic receptor, DEC-205;[11] and DEC-205 monoclonal-antibody-targeted nanoparticles.[12]

However, relatively less attention has been paid to improvement of the intrinsic endocytosis (antigen internalization) process of DCs, even though enhancing endocytosis offers a novel alternative to effectively deliver a myriad of emerging therapeutic agents (antigens or genes) by in vitro, ex vivo or in vivo methods.[13-17]

Immature DCs (iDCs) actively sample antigens from the environment using their full endocytic capacity. Once DCs are matured, however, their endocytic capacity is naturally down-regulated, so that they can migrate to the lymph nodes wherein DCs present antigenic peptides to lymphocytes, hence inducing a successive adaptive immune response. The natural down-regulation of DC endocytosis upon maturation terminates the continued uptake of most immunotherapeutics targeting DCs.[18] Also, adjuvants within vaccines cocktails can inadvertently activate iDCs before antigen uptake,[14] thereby reducing vaccine efficiency.

Only a very few studies have considered artificially programming DCs to modulate endocytosis during or past DC maturation. For instance, when iDCs are pre-treated with dexamethasone and subsequently stimulated with tumor necrosis factor-α (TNF-α) they show an endocytic capacity four times higher than iDCs treated only TNF-α.[19] In addition, Yanagawa and Onoe[20] report that CCL3 or CCL19 stimulated an enhanced endocytic ability for a short time period (≤ 1 hr) in iDCs or mDCs, respectively, when dextran and chemokines were added simultaneously to the cell culture.

In the companion article to this work,[1] we report a study with the JAWII DC cell line where iDCs were pre-treated with various combinations of the chemokines CCL3 and CCL19, followed by intentional maturation using lipopolysaccharide (LPS). The DCs pre-treated with a chemokine cocktail of CCL3 : CCL19 at a 7 : 3 mass ratio then matured with LPS, retained their antigen uptake capacity at levels 36% higher than iDCs and 96% higher than iDCs treated only with LPS. This chemokine programming of iDCs also modulated the expression of MHC molecules and various cytokine responses of DCs. These initial cell line results suggest a new DC programming tool for enhancing ex vivo and in vivo immunotherapy vaccine strategies by overcoming the natural cessation of antigen uptake upon DC maturation. For example, even though iDCs are accidently pre-matured by an adjuvant before efficient internalization of antigens,[14] they would still retain their endocytic capacity at a certain level, which would increase the overall vaccine efficiency.

However, enhancing antigen uptake capacity of DCs does not guarantee that exogenous antigens captured by DCs will be processed and presented to T cells accordingly. Hence, the functional consequences of modifying antigen uptake on antigen presentation and T-cell activation must be quantified. In this second companion study, we pre-treat primary murine bone marrow-derived DCs (BDDCs) with the chemokine mixture of CCL3 + CCL19 at a ratio of 7 : 3, then cells were matured with LPS. Treated BMDCs were then cultured with antigen-specific CD4+ T cells to examine whether DCs successfully present an exogenous antigen, ovalbumin (OVA), to T cells. Again, we determined cytokine [interferon-γ ((IFN-γ), interleukin-1β (IL-1β), IL-2, IL-10] secretion responses in the DC–T-cell co-cultures and quantified the kinetics of DC maturation.

Results demonstrate that iDCs pre-treated with the chemokine mixture of CCL3 + CCL19 at a ratio of 7 : 3 then subsequently treated with LPS, induced OVA-specific CD4+ T-cell proliferation that was initiated from ~ 100% more undivided naive T cells as compared to iDCs treated only with LPS, with OVA antigen added after LPS treatment. In addition, this programming of DCs and subsequent LPS treatment induced live OVA-specific CD4+ T-cell proliferation at levels ~ 117% higher than iDCs treated with LPS only after 89 hr of DC–T-cell co-culture. Cytokine (IFN-γ, IL-1β, IL-2, IL-10) secretions into the DC–T-cell co-culture medium were approximately the same between iDCs treated with and without the addition of chemokine cocktail, regardless of LPS addition. Even after subsequent LPS treatment, iDCs pre-treated with the chemokine mixture exhibited maturation marker expressions (double-positive CD80 and CD86) at levels less than iDCs treated only with LPS; at least for the first 24 hr after LPS treatment. Both our previous paper with cell lines and the BMDC results here substantiate the effect of DC programming with a specific chemokine combination on the final antigen presentation to T cells, proffering a novel immunotherapy strategy to control adaptive immunity for DC-based vaccine applications.

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
  10. Supporting Information
Dendritic cell differentiation

Bone marrow-derived dendritic cells (BMDCs) were generated as described previously.[21] Briefly, bone marrow was collected from femurs and tibias of hind limbs of 6∼10 week old female C57BL/6 mice (Jackson Laboratory, Bar Harbor, ME). After red blood cells were lysed, cells were washed twice using PBS (without Ca2+ and Mg2+) (Sigma, St Louis, MO). Then cells were plated into Petri dishes (100-mm diameter) at 5 × 105 cells/ml of the culture medium (RPMI-1640; Cellgro, Manassas, VA), supplemented with 2 mm l-glutamine (Cellgro), 10 mm HEPES, 1 × non-essential amino acid, 1 mm sodium pyruvate, 0·1% 2-mercaptoethanol, 100 U/ml penicillin, 100 μg/ml streptomycin, 10% heat-inactivated fetal bovine serum (FBS) (all from Invitrogen, Carlsbad, CA), and 20 ng/ml granulocyte–macrophage colony-stimulating factor (GM-CSF; Peprotech, Rocky Hill, NJ) (Day 0). Then, on Day 3, 10 ml fresh medium (at 20 ng/ml GM-CSF) was added per sample, incubated until Day 5 when the existing medium was removed and replaced with 10 ml fresh medium (at 20 ng/ml GM-CSF). Two days later (Day 7), cells were harvested and CD11c+ BMDCs were isolated using CD11c-specific microbeads as per the manufacturer's protocol (Miltenyi Biotec, Auburn, CA). Isolated BMDCs were further cultured in the medium (at 10 ng/ml GM-CSF) for treatment with chemokines or LPS as described below.

Dendritic cell culture conditions

Dendritic cell controls and various treatments in this study are summarized in Fig. 1. Based on the results of our companion study,[1] a chemokine cocktail of CCL3 + CCL19 at ratio of 7 : 3 (70:30 ng/ml, respectively) (Peprotech) was selected for the chemokine treatment followed by intentional maturation using LPS (at 100 ng/ml) (Sigma). To examine the effects of DC maturation by LPS on antigen uptake and presentation of DCs to T cells, a model antigen, endotoxin-free OVA (EndoGrade, Hyglos GmbH, Germany), was either added to DCs both before and after LPS treatment (Ag 2×) or only after LPS treatment (Ag 1×) as shown in Fig. 1. Further, OVA was added at 0, 1, 10 or 100 μg/ml per DC control or treatment and was present in the medium throughout the duration of the DC–T cell co-culture to examine the full capacity of antigen uptake of DCs at various antigen concentrations for long-term antigen-specific T-cell stimulation.

image

Figure 1. Schematic representation of overall study protocol for immature dendritic cell (iDC) treatment with chemokine cocktail, maturation with subsequent lipopolysaccharide (LPS) treatment, and co-culture of treated and control DCs with ovalbumin (OVA)-specific CD4+ T cells. OVA antigen was added in various concentrations (0, 1, 10, and 100 μg/ml), and LPS added at 100 ng/ml. The chemokine cocktail of CCL3 and CCL19 was fixed as 7 : 3 (by 70 and 30 ng/ml, respectively) in this study. A total of four different combinations were examined based on duration of chemokine exposure (CKE = 24 hr or 137 hr) and number of OVA applications (twice = Ag 2× or once = Ag 1×).

Download figure to PowerPoint

When DCs are pre-treated with dexamethasone for 24 hr, the endocytic capacity of DCs is lower than that of DCs pre-treated for more than 24 hr,[19] and endocytic capacity of DCs is sensitive to the duration of DC treatment with chemokines.[20, 22] In our companion study using a JAWSII DC line, LPS was added to DCs in the presence of chemokines that were previously added for 24 hr. In this study, to examine the effect of DC treatment time with the chemokine cocktail and the presence/absence of the chemokine cocktail in the medium on antigen uptake and presentation of DCs to T cells, DCs were exposed to the chemokine cocktail for either 24 hr [chemokine exposure (CKE) for only 24 hr; i.e. CKE = 24 hr] or 137 hr (CKE = 137 hr); DCs of ‘CKE = 24 hr’ were exposed to the chemokine cocktail for the first 24 hr and then washed twice using PBS followed by being resuspended in fresh medium without chemokines, whereas the other series of DCs at ‘CKE = 137 hr’ were exposed to the chemokine cocktail throughout 137 hr without medium change (Fig. 1). In this way, a total of four combinations of DC treatments were examined in this study: CKE = 137 hr:Ag 2×, CKE = 137 hr:Ag 1×, CKE = 24 hr:Ag 2×, CKE = 24 hr:Ag 1×.

Collectively, the four different cell groups were: (i) untreated iDCs, (ii) iDCs treated only with LPS, (iii) iDCs treated only with the chemokine cocktail, and (iv) iDCs pre-treated with the chemokine cocktail and subsequent LPS. At time zero, all DC groups were plated into a 24-well plate at 5 × 105 cells/ml (1 ml/well) and Groups 3 and 4, the chemokine-treated iDCs, were exposed to the chemokine cocktail. At time = 24 hr, one portion of Groups 3 and 4 chemokine-treated iDCs had their medium removed and replaced with fresh medium without any chemokines (CKE = 24 hr). At time = 24 hr, one portion from each of the four Groups (1, 2, 3 and 4) was given the antigen OVA (this is the first OVA addition for ‘Ag 2×’ before LPS treatment), and then 1·5 hr later, Groups 2 and 4 received LPS. At time = 48 hr, all DCs received OVA (this is the second OVA for both ‘Ag 2×’ and ‘Ag 1×’ after LPS treatment) and then were mixed with CD4+ T cells for assays of antigen presentation to T cells as described below.

Antigen-specific T-cell proliferation assay

To examine the antigen-presenting capacity of the variously treated DCs, the DCs were treated as described above, then co-cultured with antigen-specific T cells. Briefly, 24 hr after LPS treatment (elapsed time of 48 hr) of DCs, OVA-specific CD4+ T cells were isolated from the lymph nodes and spleen of 6∼10 week old female OTII transgenic mice (Jackson Laboratory) using a CD4+ T-cell isolation kit (Miltenyi Biotech) as per the manufacturer's protocol. Isolated CD4+ T cells were labelled with carboxyfluorescein succinimidyl ester (CFSE) (Invitrogen) as per the manufacturer's protocol. Then, all groups of DCs were collected and mixed with CFSE-labelled T cells in fresh medium at a ratio of 1 : 10 (DC : T cell) (2 × 104 DCs : 2 × 105 T cells in 200 μl/well) and placed into a 96-well tissue culture plate and incubated at 37°. Eighty-nine hours after DC–T-cell co-culture (elapsed time of 137 hr), supernatants were saved for future cytokine analysis and all cells (DCs and T cells) were stained with anti-mouse monoclonal antibodies against CD4 (clone RM4-5; IgG2aκ) (BD Pharmingen, San Jose, CA). Then, cells were extensively washed using PBS, treated with the Live/Dead™ fixable staining kit (Invitrogen) to exclude dead cells in the flow cytometry assay as per the manufacturer's protocol, followed by fixation of cells using 2% paraformaldehyde. Cells were washed twice using PBS and then analysed immediately with 10 000 events per sample using FACS Canto (BD Biosciences, San Jose, CA). Proliferation of T cells (CFSE fluorescence dilution) was analysed for live CD4+ T cells using Flowjo Software (Tree Star, Ashland, OR). T-cell viability within the DC–T-cell co-culture was determined as the number of viable cells (Live Stain) that also showed CD4 antibody binding.

Cytometric bead array for cytokine release

After 89 hr of the co-culture of DCs and T cells as described above, the culture supernatant for all samples was saved and the accumulated cytokines (IL-1β, IL-2, IL-10 and IFN-γ) were analysed using a cytometric bead array (CBA) (BD Pharmingen) as per the manufacturer's protocol.

Effect of antigen and chemokine treatment of DC maturation

Effects of the antigen OVA and chemokine pre-treatment on DC maturation were examined as a function of exposure time using only DCs, under culture conditions identical to those for the T-cell proliferation assay above. To examine the effects of OVA presence on DCs, now OVA was added only once, simultaneously with the LPS. However, the DC exposure to the chemokine cocktail was identical to before; exposure for either 24 hr (CKE = 24 hr) or 137 hr (CKE = 137 hr). Briefly, DCs were exposed to the chemokines (CKE = 24 or 137 hr), then OVA was added once (at 0, 1, 10, or 100 μg/ml) at an elapsed time of 24 hr when the LPS was added. The DC surface marker expressions (CD80/86) were measured at elapsed times of 48, 96 and 137 hr. Collected DC samples were blocked with anti-mouse Fcγ III/II receptor monoclonal antibody (clone 2.4G2; IgG2bκ) and stained with CD80 (clone 16-10A1; IgG2κ), CD86 (clone GL1; IgG2aκ), and their isotypes (all from BD Pharmingen) for 30 min at 4° in the dark. After staining, cells were extensively washed three times using ice-cold FACS buffer (2% FBS/1× PBS) then analysed immediately with 10 000 events per sample using FACS Canto (BD Biosciences). An isotype per antibody against each surface marker was used as the negative control and the mean fluorescence intensity (MFI) of the isotype control cells was subtracted from that of cells stained with each antibody per surface marker. Data were analysed using Flowjo software (Tree Star).

Statistical analysis

For pairwise comparison between two treatments or between a treatment and a control, a one-sided Mann–Whitney U-test was used. To observe a significant difference between all DC controls and treatments in pairs or between groups, each of which includes multiple treatments of DCs, a general linear two-way analysis of variance (AVOVA) (Tukey test) model was applied pairwise for a mixed model with repeated measures. For all statistical methods, graphpad prism (Version 5·04, La Jolla, CA) or Minitab software (Version 14, State Collage, PA) was used. If not indicated, a P-value ≤ 0·05 was considered to be significant.

Results

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

Effects of chemokine pre-treatment of DCs on T-cell stimulation

Antigen-specific CD4+ T cells were co-cultured with DCs programmed with or without the specific CCL3:19 (7 : 3) chemokine cocktail to measure the antigen-presenting capacity of DCs. In the absence of any applied OVA antigen, DCs stimulated no T-cell proliferation as seen in Fig. 2 (Column 1). However, when DCs were exposed to OVA antigen, they stimulated T-cell proliferation, with an increasing T-cell proliferation with increasing applied OVA concentration. Even after LPS treatment, DCs programmed with the chemokine cocktail (Fig. 2, Row D) stimulated antigen-specific T-cell proliferation patterns closer in resemblance to untreated iDCs (Fig. 2, Row A) compared with that of iDCs treated only with LPS (Fig. 2, Row B).

image

Figure 2. Representative histograms of antigen-specific CD4+ T-cell proliferation upon co-culture with the four various dendritic cell (DC) groups. Column 1 indicates No ovalbumin (OVA) addition. Columns 2–4 indicate three different OVA concentrations applied twice to all DC groups (Ag 2×). Columns 5–7 indicate three different OVA concentrations applied only once to all DC groups (Ag 1×). Row A = untreated immature DCs (iDCs), Row B = iDCs treated only with lipopolysaccharide (LPS), Row C = iDCs treated only with the chemokine cocktail, and Row D = iDCs pre-treated with the chemokine cocktail and subsequent LPS. CFSE fluorescence dilution indicates T-cell proliferation measured once after 89 hr of DC–T-cell co-culture. Numbers shown in each histogram are the percentage of proliferated T-cell population out of total CD4+ T cells at the flow cytometry after the 89-hr duration. Only the result of CKE = 137 hr is shown here. Patterns of the histogram changes of T-cell proliferations from the CKE = 24 hr (data not shown) are similar to this CKE = 137 hr.

Download figure to PowerPoint

T-cell proliferation flow cytometry histograms were essentially identical for all conditions regardless of the duration of chemokine exposure (CKE = 24 or 137 hr) (statistically analysed data available). Regardless of the number of antigen exposures and antigen concentrations, DCs exposed to the chemokine cocktail for 24 or 137 hr stimulated the same T-cell proliferation levels (Fig. 3a,c or b,d). Exposing the DCs to Ag 2× versus Ag 1× had only a mild effect of increasing T-cell proliferation at OVA concentrations of 10 and 100 μg/ml, but not at 1 μg/ml (Fig. 3a versus b or c versus d). Increasing OVA concentration (1, 10, 100 μg/ml), as expected, increased T-cell proliferation, under all conditions. The DCs treated with a chemokine cocktail then LPS maturation exhibited greater antigen-specific T-cell proliferation versus DCs only treated with LPS, regardless of whether DCs were exposed to OVA once or twice, or OVA concentration, indicating that the specific CCL3:19 (7 : 3) pre-treatment was able to enhance antigen presentation and T-cell stimulation over that of LPS matured cells.

image

Figure 3. Per cent cell proliferation of antigen-specific CD4+ T cells upon co-culture with the four various dendritic cell (DC) groups. Per cent cell increase values are obtained from CFSE fluorescence dilutions shown in Fig. 2 and are identical to percentages shown in each histogram. Results are given for both number of antigen applications (Ag 2×: a, c; and Ag 1×: b, d) and for chemokine exposure time (CKE=137 hr: a, b; CKE=24 hr: c, d). Per cent cell increases are shown with mean ± SEM, n = 4 (four independent trials) for CKE = 137 hr or n = 3 (three independent trials) for CKE = 24 hr. Star indicates values significantly different ( 0·05) relative to % cell increase values for immature DCs (iDCs) treated only with lipopolysaccharide (LPS) at the same ovalbumin (OVA) concentration.

Download figure to PowerPoint

Viable CD4+ T cells at the end of the co-culture (after 89 hr) are reported in Fig. S1 (see Supplementary material). Percentages of all live CD4+ T cells after 89 hr of DC–T-cell co-culture decreased with addition of LPS and with increasing applied OVA concentration at Ag 1×, regardless of the chemokine exposure durations.

Effects of chemokine pre-treatment of DCs on cytokine expression in DC–T-cell co-culture

After 89 hr of the co-culture of DCs and T cells, the supernatants from the co-culture wells were saved and examined for various accumulated cytokines. Whereas IFN-γ secretions into the DC–T-cell co-culture wells increased with both addition of LPS and with increasing OVA concentrations, IL-2 secretions increased only with increasing OVA concentrations, regardless of the duration of chemokine exposure as shown in Fig. 4 (CKE = 137 hr) and the Supplementary material, Fig. S2 (CKE = 24 hr). Interleukin-1β secretion also increased upon addition of LPS but did not change with the various OVA concentrations, regardless of the duration of chemokine exposure. Contrary to all the other cytokines, IL-10 secretion significantly decreased when DCs were exposed to LPS at an OVA concentration of 100 μg/ml, regardless of the duration of chemokine exposure. Whereas each of the other cytokines exhibited very similar secretion levels between the numbers of antigen exposure (Ag 2× versus Ag 1×) per DC control or treatment, IL-2 secretion significantly decreased with (Ag 1 × ) versus (Ag 2 × ) at higher OVA concentrations of 10 and 100 μg/ml, regardless of the duration of chemokine exposure. Interestingly, when DCs were exposed to only chemokine cocktail for 24 hr followed by medium change with fresh medium (see Supplementary material, Fig. S2), secretions of IL-1β, IL-2 and IL-10 significantly increased versus CKE = 137 hr (Fig. 4) at an OVA concentration of 1 μg/ml, regardless of the number of antigen exposures. In addition, CKE = 24 hr (see Supplementary material, Fig. S2) induced decreases in cytokine levels of IL-1β, IL-2 or IL-10 versus CKE = 137 hr (Fig. 4) regardless of addition of LPS and chemokine cocktail, at the higher OVA concentrations of 10 and 100 μg/ml. However, for all cytokines, secretion levels between DCs treated only with LPS and pre-treated with chemokine cocktail/subsequent LPS were essentially the same, regardless of the number of antigen exposures, the antigen concentration, and the duration of chemokine exposure.

image

Figure 4. Cytokines secreted during 89 hr of dendritic cell (DC) and CD4+ T-cell co-culture for the CKE = 137 hr exposure condition. Cytokine concentrations (pg/ml) are shown for increasing ovalbumin (OVA) concentrations applied twice (Ag 2×, dark grey bars) or once (Ag 1×, light grey bars). White bars indicate no OVA addition. iDC = untreated immature DCs, LPS = iDCs treated only with lipopolysaccharide, CCL = iDCs treated only with the chemokine cocktail, and CCL + LPS = iDCs pre-treated with the chemokine cocktail and subsequent LPS. Cytokine levels are shown with mean ± SEM, n = 3 (three independent trials). Star indicates values significantly different ( 0·05) relative to cytokine values for iDCs treated only with LPS at the same OVA concentration. Bracket indicates cytokine values significantly different ( 0·05) between samples receiving antigen twice (Ag 2×) versus once (Ag 1×) at any antigen concentration.

Download figure to PowerPoint

DC expression of maturation surface markers is delayed by chemokine pre-treatment

To examine the effects of OVA antigens and chemokine cocktail pre-treatment on DC maturation, a separate series of experiments were carried out identical to those above where iDCs were treated with the chemokine cocktail (CKE = 24 hr and 137 hr), matured with LPS as before, then cultured in the presence of OVA antigen (no CD4+ T cells present), then maturation surface marker (CD80/86) expressions on DCs were measured. Representative flow cytometry quadrant density plots for the four DC groups are shown in the Supplementary material, Fig. S3 at an elapsed time of 48 hr (24 hr after LPS/OVA addition). Quadrant density plots for the four DC groups were identical for conditions of CKE = 24 hr and 137 hr; consequently only the results of the CKE = 137 hr are provided in Fig. S3. Note: at an elapsed time of 48 hr iDCs treated only with LPS exhibited black density cores (in the middle of the contour lines) shifting to CD80+ CD86+ quadrant more than all the other DCs.

As seen in Fig. 5, the presence of OVA at different concentrations did not affect DC maturation per DC group for both CKE = 24 hr and 137 hr through all time-points. When each DC group that includes four OVA concentrations is compared pairwise between four DC groups, iDCs treated only with the chemokine cocktail did not express the double-positive CD80/86 at levels different from untreated iDCs, regardless of the duration of chemokine exposure through all time-points. Interestingly, iDCs programmed with chemokine cocktail, even after LPS treatment, expressed the double-positive CD80/86 at levels significantly lower than iDCs treated only with LPS or higher than the other two DC groups at an elapsed time of 48 hr, regardless of the duration of chemokine exposure. However, from an elapsed time of 96 hr through to 137 hr, the maturation of DCs programmed with chemokine cocktail and then treated with LPS became the same as that of iDCs treated only with LPS, but these two LPS-treated DC groups still expressed the double-positive CD80/86 at levels significantly higher than the other two DC groups, regardless of the duration of chemokine exposure. Interestingly, including large standard deviations in Fig. 5(f), two LPS-treated DC groups of CKE = 24 hr exhibited the double-positive CD80/86 levels slightly higher than those of CKE = 137 hr through all time-points.

image

Figure 5. Effects of antigen (ovalbumin; OVA) presence with various concentrations and duration of exposure to chemokine cocktail (CKE = 24 hr and 137 hr) on dendritic cell (DC) maturation as a function of time. Various DC groups were cultured as stated, with only a single time of antigen addition in the absence of CD4+ T cells, then CD80/86 expressions per DC were measured as a function of time using the flow cytometer. iDC = untreated immature DCs, LPS = iDCs treated only with lipopolysaccharide, CCL = iDCs treated only with the chemokine cocktail, and CCL + LPS = iDCs pre-treated with the chemokine cocktail and subsequent LPS. The fraction of double-positive (CD80+ CD86+) cell populations (from total cells in a quadrant dot plot) for the various OVA concentrations are normalized to the fraction number of CD80+ CD86+ cells generated from untreated iDCs receiving no OVA. (a–c) Results for the CKE = 137 hr group; (d–f) results of the CKE = 24 hr group. (a, d) Results at elapsed time of 48 hr (24 hr after OVA/LPS addition); (b, e) results at elapsed time of 96 hr, and (c, f) results at elapsed time of 137 hr. Normalized fractions of the double positive are shown with mean ± SD, n = 3 (three independent trials). Bracket indicates statistical difference ( 0·05) between DC groups, each of which includes addition of four different OVA concentrations.

Download figure to PowerPoint

Dendritic cell viability was also measured at the same time when the maturation marker expression was examined. As a result, viabilities appeared to decrease over time for all DC groups and the lowest viability was above 80% at an elapsed time of 137 hr. For all DC groups, viabilities were not significantly different from each other through all time-points (data not shown).

Discussion

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

This is the first reported in vitro study showing how the programming of antigen uptake/processing capacity of DCs, against subsequent TLR stimulation, can control antigen-specific T-cell responses over a long-term (89 hr) co-culture of DC–T cells.

CD4+ T cells activate and proliferate only when they recognize a specific peptide (here derived from OVA) in the context of MHC class II complexes expressed on DCs.[23] When DCs present antigenic peptides to T cells, peptide–MHC complexes (signal 1) and co-stimulatory molecules such as CD80 or CD86 on DCs (signal 2) are essential for effective T-cell proliferation.[24-26] Among cytokines secreted from DC–T-cell interaction, IFN-γ is known as a pro-inflammatory cytokine playing an important role in T helper type 1 (Th1) polarization, whereas IL-2 is known as a growth factor of T cells. Interestingly, these two cytokines are also known for inhibiting T-cell expansion or promoting apoptosis of antigen-specific CD4+ T cells.[27-29] Most importantly, IL-2 has been recognized as a potent growth factor as well as a regulator of homeostasis by modulating the equilibrium between CD4+ T-cell proliferation and apoptotic death.[27] In detail, upon contact with antigenic peptides on DC surface, the naive T cells initiate proliferation, producing IL-2, and then this IL-2 stimulates the T cells for further proliferation.[30] However, to maintain the homeostasis, the immune system has a negative feedback loop to limit the immune responses,[31]; IL-2 paradoxically sensitizes T cells to apoptosis when simultaneously proliferating T cells to secrete IL-2 so that this sensitization helps terminate the immune response and prevent autoimmunity.[27]

Interleukin-2 is produced primarily by activated CD4+ T cells,[32] even though we cannot rule out secretion by DCs[33]; when DCs are activated by bacteria (Escherichia coli at MOI of 10)[34] or LPS (10 μg/ml) and CD40 ligand,[35] they secrete IL-2. However, possibly because of a different experimental protocol in this study wherein 100 ng/ml of LPS was used to stimulate iDCs, all four DC groups cultured for 89 hr (time duration for DC–T-cell co-culture in this study) in the absence of OVA antigens (Fig. 4 and Fig. S2) induced IL-2 secretion accumulated less than 10 pg/ml. As seen in Fig. 2, T cells did not proliferate for all DC controls and treatments in the absence of OVA antigen during 89 hr of the co-culture (Fig. 2, Column 1) but once OVA was added, T cells proliferated. Therefore, we suggest that accumulated IL-2 secretions shown in Fig. 4 depend mainly on T-cell proliferation proportionally stimulated by increasing OVA concentrations.

As one of the most important mechanisms for maintaining the homeostasis, activation-induced cell death (AICD) plays an effective role in T-cell deletions,[36, 37] which is a form of apoptosis of activated T cells that occurs upon repeated T-cell receptor triggering.[38] T cells co-cultured with untreated iDCs or iDCs treated only with the chemokine cocktail for all OVA concentrations (see Supplementary material, Fig. S1A–D) exhibited T-cell viability after 89 hr of DC–T-cell co-culture at levels higher or similar to those (slightly < 80%) of CD4+ T cells co-cultured with only DCs (no antigen) up to 72 hr.[38] These data all indicate that untreated iDCs and iDCs treated only with the chemokine cocktail maintained T-cell viabilities at normal levels for 89 hr of DC–T-cell co-culture irrespective of OVA concentrations (see Supplementary material, Fig. S1), and AICD can be considered a main factor inducing antigen-specific T-cell apoptosis for the cell culture protocol used in this study. Compared with these two DC groups, the other two DC groups that were exposed to LPS should have higher expression levels of the co-stimulatory molecules (signal 2) for T-cell activation. Hence, LPS-treated DCs can activate T cells to a greater extent than the other DCs (no LPS) so that LPS-treated DCs can induce AICD of T cells when T cells are highly activated/proliferated and repeatedly exposed to excessive antigenic peptides (maximizing AICD of T cells). Interestingly, the two DC groups without LPS exhibited live T-cell proliferation (measured after 89 hr of DC–T-cell co-culture) at higher or similar levels compared with the other two DC groups that were treated with LPS (Figs 2 and 3). Possibly as a result of the minimal maturation of iDCs in the absence of LPS but in the presence of GM-CSF,[21, 40, 41] as discussed in our companion study,[1] untreated iDCs or iDCs treated only with the chemokine cocktail induced activation of T cells in association with more antigen uptake/presentation to T cells for long-term duration of 89 hr (minimizing AICD of T cells).

As seen in Fig. 4, two DC groups that were exposed to LPS, induced IL-2 secretions higher or similar to the other two DC groups for both antigen ‘Ag 2×’ and ‘Ag 1×’ groups. These accumulative IL-2 levels indicate that these two DC groups that were exposed to LPS induced an accumulative total T-cell proliferation (live + dead) no less than the other two DC groups (no LPS). In addition, because AICD is proportional to the number of T-cell divisions,[42] more T-cell deletions shown in Fig. S1 (see Supplementary material) indicate that more T cells proliferated earlier, followed by more AICDs in line with the accumulated IL-2 results above. For these reasons, we can certainly postulate that live T-cell proliferation or T-cell viabilities measured only once after 89 hr of DC–T-cell co-culture (Figs 2 and 3, and Supplementary material, Fig. S1) were controlled mainly by AICD derived from DCs activated by LPS, which supports highly activated conditions of DCs and T cells by co-stimulatory molecules on DCs and death cytokines such as TNF [induced by LPS treatment in our companion study[1] present in the DC–T-cell co-culture.[30]

As loss of CFSE intensities directly indicates proliferation or deletion of CFSE-labelled cells,[43] higher intensities of the rightmost CFSE peak per histogram (Fig. 2) of T cells, which were co-cultured with iDCs treated only with LPS for ‘Ag 1×’, suggest that more T cells remained alive without proliferation[42] compared with all other DC groups. Because DC maturation by LPS induces loss of endocytic capacity, when OVA antigen was added only after LPS treatment (Ag 1×), these DCs would uptake/present less antigen, and then initiate less T-cell proliferation of undivided T cells compared with ‘Ag 2×’. This is supported by the observation that all DC groups exhibited significant increases of CFSE dilutions (Fig. 3) and accumulated IL-2 levels (Fig. 4) with ‘Ag 2×’ versus ‘Ag 1×’, along with increasing OVA concentrations for both CKE = 137 hr and 24 hr.

Clearly when OVA antigen is added only after LPS treatment (Ag 1×), iDCs pre-treated with the chemokines/subsequent LPS initiate T-cell proliferation with about ~ 100% more undivided T cells (more homogeneous T-cell population at the initial proliferation) as compared to iDCs treated only with LPS (Fig. 2), which is solely due to a higher capacity of endocytosis of these former DCs versus the latter DCs as demonstrated in our companion study.[1] Furthermore, iDCs pre-treated with chemokine cocktail/subsequent LPS maintained accumulated IL-2, IFN-γ and IL-1β at levels comparable to iDCs treated only with LPS. Hence, this DC programming with the chemokine cocktail can initiate a more homogeneous adaptive immune response[14] by preventing the endocytosis cessation of pre-matured DCs while DCs and T cells are still activated by adjuvants.

Taub et al.,[44] found that when splenocytes are incubated with CCL3 for 18 hr, they express CD80 at levels higher than with only LPS treatment. They concluded this CD80 expression induced by CCL3 induced antigen specific T cell proliferations they observed. By contrast, in our study (wherein BMDCs were examined), the chemokine cocktail/subsequent LPS treatment delayed CD80/86 expressions on DCs in time course as compared to only LPS treatment of DCs (Fig. 5 and S3). In our companion study[1] (wherein JAWSII DC line was examined), iDCs pre-treated with the chemokine cocktail (7 : 3) showed a semi-maturation status, and even after subsequent LPS treatment, the MHC II expression was not up-regulated higher than untreated iDCs, in line with typical semi-maturation phenotypes that are non-responsive to subsequent TLR stimulation[45] or are resistant to LPS-induced maturation.[46] Therefore, these maintained immature phenotypes of DCs may at least partially contribute to the retained endocytic capacity even after subsequent LPS treatment, as observed in our companion study.

Compared with all other cytokines in this study, two DC groups that were treated with LPS induced secretion of IL-10 at levels lower than the other two DC groups (no LPS) only at 100 μg/ml of OVA (Fig. 4). As an important immunoregulatory cytokine for Th2 polarization,[47] IL-10 is secreted from DCs upon LPS stimulation[48] or from activated T cells[49] and inhibits IL-2 production[50] as well as T-cell apoptosis.[51] Results suggest that when DCs are treated with LPS, addition of OVA (at 100 μg/ml) to DCs might suppress IL-10 secretions and more IFN-γ or IL-2 was secreted in the DC–T-cell co-culture wells in association with more AICDs, as seen in the Supplementary material, Fig. S1.

When DC medium was refreshed to wash out the initially added chemokines from the cell culture wells (CKE = 24 hr), the resultant levels observed were different from those of CKE = 137 hr for cytokine secretions (Fig. 4 and Supplementary material, Fig. S2) and CD80/86 expressions on DCs (Fig. 5). However, antigen-specific T-cell proliferation (Fig. 3) resulted in no statistical difference between CKE = 137 hr and 24 hr (data not shown), and statistical difference patterns of CD80/86 expression levels for all DC groups per time-point were the same between CKE = 137 hr and 24 hr (Fig. 5). These results suggest that the duration of chemokine exposure does not control interactions of DCs and antigen-specific T cells. Also, this observation emphasizes that once iDCs are pre-treated with the chemokines for 24 hr, this programming effect can be retained by DCs in the absence of the originally added chemokines in the medium. However, when medium was refreshed (CKE = 24 hr), IL-2 (at 0, 1 and 10 μg/ml of OVA) or IL-10 (at 1 μg/ml OVA) secretions by iDCs treated only with the chemokine cocktail were modulated at levels significantly higher versus iDCs treated only with LPS (see Supplementary material, Fig. S2). In addition, statistical differences of CFSE dilutions at 1 μg/ml of OVA for all four DC groups were different between CKE = 137 hr and 24 hr for ‘Ag 2×’ (Fig. 3). It seems that a ‘conditioned medium’ in CKE = 137 hr included specific soluble proteins that directly affect cell phenotype changes versus the ‘refreshed medium’ of CKE = 24 hr.

It is not possible to clearly understand how this medium change works based only on the data shown here. Hence, a more detailed study of those soluble proteins should be performed in subsequent studies. For example, CCL3 or CCL19 is generally secreted upon DC–T-cell interaction or DC maturation, respectively.[52, 53] Then, irrespective of the presence of the original chemokines (used for programming) in the cell culture medium, overall chemokine (CCL3 and CCL19) concentrations and their ratio would change over time. This change of the chemokine concentrations during the culture should be further examined to better understand the cell culture condition post-programming. Furthermore, most importantly, another DC–T-cell co-culture study should be examined, sampling cells and medium periodically with time so as to quantify the effect of LPS treatment on DC-induced T-cell proliferation/apoptosis, as a function of antigen addition schedules.

Following the first demonstration of DC programming in our companion study,[1] this programming is confirmed to enhance antigen presentation to T cells in this study. New immunotherapeutics have emerged using DC-based vaccine strategies[2, 3, 54] and DC programming strategies have been shown with their effects on successive adaptive immunity by injection of ex vivo programmed DCs[18] or signal molecules released in vivo from implant biomaterials.[16] In vivo pre-injection of cytokines has also been shown to impact T-cell trafficking and priming through regulation of DC migration.[55] Hence, ex vivo or in vivo animal trials would be critical to confirm the multifunctional impacts of DC programming using the chemokine combination on enhancing vaccine efficiency.

We demonstrate here that DCs programmed with a specific chemokine combination induced markedly robust and durable T-cell activation and proliferation (for 89 hr) even after DCs were stimulated by adjuvant such as LPS, because of their retained endocytic capacity against subsequent LPS treatment of DCs. Moreover, once iDCs were pre-treated with the chemokines for 24 hr, this programming effect was retained through DC–T-cell co-culture in the absence of the original chemokines in the cell culture well. This underscores that the chemokine programming can be employed for ex vivo immunotherapeutics wherein DCs are trained using specific protocols at the bench and then injected into the host as one of the DC-based vaccine strategies.

These results suggest that DC-based vaccine strategies could be modified, overcoming the natural limit (cessation of antigen uptake upon DC maturation) of the host immune response. For instance, in vivo programming of DCs could be possible using implanted biomaterials releasing chemokines and antigen sequentially or chemokine/antigen targeting iDCs residing in lymphoid organs.[56] In this way, this DC programming with the chemokine combination can initiate more homogeneous adaptive immune response[14] by preventing endocytosis cessation of pre-matured DCs while DCs and T cells are still activated enough by an appropriate adjuvant.

Acknowledgements

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

This work was financially supported by the National Institutes of Health: NIAID R01AI074661 and NIDCR R01DE018701.

References

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

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
imm12059-sup-0001-FigS1.tifimage/tif165KFigure S1. CD4+ T-cell viability after 89 hrs of T cell co-culture with all DC groups.
imm12059-sup-0002-FigS2.tifimage/tif241KFigure S2. Cytokines secreted during 89 hrs of DC and CD4+ T cell co-culture for the CKE=24 hrs exposure condition.
imm12059-sup-0003-FigS3a.tifimage/tif439K 
imm12059-sup-0004-FigS3b.tifimage/tif308KFigure S3. Representative quadrant dot plots (Zebra plots) of CD80 and CD86 expressions selected from results shown in Figure 5. (a) Column 1, 2, 3, 4, 5 indicate isotype (no OVA), and OVA concentrations = 0, 1, 10, 100 g/mL, respectively. (b) Representative result from CKE = 137 hrs at elapsed time of 96 hrs or 137 hrs is only shown.
imm12059-sup-0005-SupplementText.docxWord document49K 

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.