Introduction

Condyloma acuminatum (CA), one of the most common sexually transmitted diseases caused by human papillomavirus (HPV), is characterized by high recurrence. The persistent infection of HPV is considered to be a major epidemiological marker of individual risk for anal or cervical cancers [12, 17, 21, 26]. CA-related emotional distress and heavy economic burdens have been noticed. Therefore, a strategy to effectively manage CA would have far-reaching global health and economic impact.

Evidences have shown that the local cell-mediated immunity was suppressed in CA lesions [30], especially in the patients with recurrent CA [5, 14, 36, 38, 40]. Langerhans cells (LCs) or dermal dendritic cells (DCs) are the most important antigen-presenting cells in skin [11]. However, the number and the function of LCs are significantly reduced in CA lesions [27, 29]. This might attribute to the insufficient induction of a more threatening T help type 1 (Th1) response which would favor the development of cytotoxic T lymphocytes (CTLs). Additionally, the down-regulation of TNF-α and the up-regulation of the immunosuppressive cytokine IL-10 in CA lesions may subsequently suppress the anti-viral activities of the Th1 cells [2, 29].

Antigen-based therapeutic HPV vaccines are proved to potentially eliminate pre-existing lesions and malignant tumors by generating cellular immunity against HPV-infected cells [22]. The key point of these vaccine strategies is to elicit the necessary processing and presenting of peptide, and subsequently the enhanced activation of T cells [16, 18, 41]. In this regard, the delivery of ex vivo pretreated DCs may be an alternative [1, 32]. There are several studies on DC vaccine against high-risk HPV related cervical carcinoma [32]. In clinical trials, autologous DCs pulsed with HPV 16/18 E7 protein can induce T cell responses in a portion of both early and late stage cervical cancer patients, and finally lead to a slow tumor progression [15, 33]. Therefore, therapeutic DC vaccine such as HPV 11 E7 protein-pulsed DCs might be a potential approach to manage the low-risk HPV infection such as CA. On this basis, boosting of immune responses by effective adjuvants will benefit the vaccination.

Toll like receptors (TLRs) recognize motifs of microbial pathogens for early recognition of microbial invasion. Their ligation by microbial elements is critical in DC activation and maturation, and therefore important for the adaptive pathogen-specific T cell response [7, 23, 25 ]. To analyze the substantial effect of TLR agonists on adjuvants of a HPV epitope peptide vaccine regimen, we studied the ability of various TLR agonists to modulate the maturation of HPV 11 E7 CTL epitope peptide-loaded mdDCs, and further to induce HPV-specific Th1 and CD8+ T cell responses. MdDCs may represent immature dendritic cells which home at sites of inflammation [4]. Moreover, the availability of large numbers differentiated mdDCs facilitated the study to compare several TLR agonists side by side.

Materials and methods

Isolation of monocytes, T cells, and naïve T cells

Peripheral blood mononuclear cells (PBMC) were isolated from HLA-A*0201 healthy donors by Lympholyte-H (Cedarlane Laboratories, Canada) density gradient centrifugation as recommended by the manufacturer. CD14 positive monocytes were isolated with paramagnetic beads (MiltenyiBiotec) as previously described [39]. The purity of the cell separation tested by FACS was 95.3% (±4.1%).

T cells were isolated with pan T cell isolation kit (MiltenyiBiotec) from the remaining monocyte-depleted cells [39]. CD4+ T cells were then collected after depletion of CD8+ T cells with anti-CD8 paramagnetic beads (MiltenyiBiotec). Afterwards, CD4+CD45RA+ naïve T cells were labeled with anti-CD45RA paramagnetic beads (MiltenyiBiotec) and enriched on separation columns in magnetic field (MidiMACSTM, MiltenyiBiotec). All the cells were incubated in the standard medium [RPMI 1640 (Gibco, USA) supplemented with 10% heat-inactivated fetal bovine serum, 2 mmol/l l-glutamine, 100 IU/ml penicillin, 100 μg/ml streptomycin].

Generation and stimulation of mdDCs

Immature dendritic cells (imDC) were generated from CD14+ cells as previously described [39]. On day 6, the CD1a expression of the cultured cells was 61.2% (±11.36%). 2 × 106 differentiated DCs were then co-cultured with immunodominant HLA-A*0201-restricted HPV 11 E7 CTL epitope peptide (TLKDIVLDL) screened previously [39] in six-well plate (20 μg/ml) for 1–12 h. Cultures were then supplemented with different TLR agonists for 48 h at the following concentrations: LPS from Escherichia coli 0111:B4 strain (TLR4 agonist, Sigma, USA) 0.4 μg/ml, imiquimod (TLR7 agonist, Mingxin Pharmaceutical Ltd, China. Imiquimod was dissolved in 0.01 M HCl solution to obtain a concentration of 0.50 mg/ml) 4 μg/ml, polyinosinic acid: polycytidylic acid (PIC, TLR3 agonist, InvivoGen, USA) 20 μg/ml, CpG oligonucleotide 1826 (5′-TCCATCACGTTCCTGACGTT-3′, TLR9 agonist, Invitrogen, USA) 10 μg/ml, or a combination of PIC, imiquimod and CpG. The CD1a expression of the stimulated cells was 97.1% (±5.23%). The concentration of all TLR agonists were determined in preliminary experiments (data not shown), at which the agonists induced significant DC phenotypic differentiation without cell death. Additionally, exogenous human IL-10 (50 ng/ml medium, Peprotech Inc., USA) was added to cultures together with various TLR agonists during the DC maturation in an independent experiment.

Phenotyping of mdDCs

MdDCs were characterized on day 6 (immature mdDCs) and day 8 (mature mdDCs) using following fluorochrome-labeled monoclonal antibodies (anti-CD1a, anti-CD40, anti-CD80, anti-CD83, anti-CD86, anti-HLA-DR, as well as their corresponding PE- or FITC-labeled isotype control antibodies; eBioscience, USA) and analyzed on an EPICS-XL flow cytometer (Beckman Coulter, USA).

Cytokine production of mdDCs by stimulation with E7 peptide and TLR agonists

Supernatants were collected from cultures of immature, unstimulated, or stimulated mdDCs. Secretion of IL-12p70, IL-10, and IFN-α was measured with corresponding ELISA kits according to the manufacturer’s instructions (Biosourse, USA).

Th1/Th2 polarization

Isolated naïve T cells (2.5 × 105 cells/500 μl medium) were co-cultured with 5 × 104 HPV 11 E7-loaded mdDCs or various TLR agonist-pretreated mdDCs in 24-well plates for 6 days. Supernatants were collected, and cytokine (IFN-γ and IL-4) concentration was measured by ELISA according to the manufacturer’s instructions (Biosourse, USA).

Cytokine production of T cells cultured with mdDCs

The HPV 11 E7 CTL epitope peptide-loaded or various TLR agonist-pretreated mdDCs (2 × 105 cells/2 ml medium) as described were co-cultured with 2 × 106 T cells in six-well plates at 1-week interval for two times. T cell culture supernatants were analyzed for IFN-γ, TNF-α, and IL-2 with the corresponding ELISA Kits (Biosourse, USA). All culture supernatants were kept at −20°C until use.

Enzyme-linked immunospot (ELISPOT) assay for IFN-γ and IL-2 production

E7 peptide-loaded mdDCs, various TLR agonist-stimulated mdDCs, and E7 peptide plus various TLR agonist-stimulated mdDCs were used as stimulator cells. T cells (1 × 105) and stimulator cells (2 × 104) were seeded into 96-well polyvinylidene fluoride (PVDF)-backed microplates coated with monoclonal antibody specific for human IFN-γ or IL-2 (R&D Systems, USA). After incubation at 37°C for 24 h, the cells were removed and the plates were processed according to the manufacturer’s instructions. Resulting spots in each well were counted with an ELISPOT Reader (Cellular technology Ltd, USA) and analyzed with the ImmunoSpot 4.0 software. The values were expressed as spot-forming cells (SFCs) per 106 T cells.

Specific CTL assay

Purified CD8+ T cells primed with HPV 11 E7 epitope peptide-loaded, TLR agonist-pretreated mdDCs served as effector cells. The HPV11E7-expressing human embryonic kidney (HEK) 293 cells established previously [39] were used as target cells. Briefly, 5 × 103 target cells were cocultured with effector cells in effector-to-target cell (E/T) ratios of 100:1, 50:1 and 20:1. CD8+ T cells primed by mdDCs without peptide loading or by E7-mdDCs unstimulated with TLR agonists were served as controls. After 6 h of incubation at 37°C, the supernatant was collected to assess the lactate dehydrogenase concentration using CytoTox 96 nonradioactive cytotoxicity assay kits (Promega Corp, Madison, USA). The cytotoxicity activity of T cells was assessed based on the following formula with the mean values from quadruple wells: Percent cytotoxicity = (experimental value − effector spontaneous value − target spontaneous value)/(target maximum − target spontaneous value) × 100 [9].

Statistical analysis

All data were expressed as mean ± SD. One-way ANOVA was used to evaluate the significance of group differences. Values of P < 0.05 were considered to be statistically significant.

Results

Phenotypic maturation of mdDCs by TLR agonists’ stimulation

LPS and PIC induced a significant up-regulation of CD40, CD80, CD83, CD86, and HLA-DR expression on mdDCs as compared with medium alone (P < 0.05) (Fig. 1a). LPS and PIC induced a significantly higher expression of CD40, CD83, and CD86 on mdDCs as compared with imiquimod (P < 0.05) or CpG ODN (all P < 0.01) (Fig. 1a). Imiquimod induced higher expression of CD80, CD83, and CD86 on mdDCs than that of control mdDCs, except for the expression of CD40 and HLA-DR (P ≥ 0.05). No increased expression of CD40, CD80, CD86, and HLA-DR on CpG ODN-stimulated mdDCs was observed (P ≥ 0.05).

Fig. 1
figure 1

MdDCs differentiate in response to TLR ligation and/or IL-10 after culture for 48 h. Surface expression of various markers (CD40, CD80, CD86, CD83, and HLA-DR) of mdDCs after exposure to indicated TLR agonists (a) or to TLR agonists plus IL-10 (b) were measured by flow cytometry. Data presented are mean ± SD values of from five donors. *Indicates statistical significance (P < 0.05) when compared with the control group

Full size image

The CD40, CD83, and CD86 expression of the mdDCs stimulated with IL-10 was lower than that of the non-stimulated mdDCs, while the CD80 and HLA-DR expression was comparatively stable (Fig. 1b). The expression of CD83, CD86, and HLA-DR on mdDCs stimulated with IL-10 plus LPS, PIC or imiquimod stimulation was higher than that of mdDCs stimulated with IL-10 alone (Fig. 1b).

Cytokine production of mdDCs by HPV11E7 plus TLR agonists’ stimulation

We investigated the secretion of IL-12 p70 and IFN-α, two potential Th1-inducing cytokines, and IL-10 by HPV 11 E7 epitope peptide-loaded mdDCs in response to different TLR ligation. High level of IL-12 was produced by mdDCs stimulated with either LPS or PIC (2155.7 ± 133.5 and 1971.7 ± 144.8 pg/ml, respectively) (P < 0.01, compared with control mdDCs or mdDCs loading of HPV 11 E7 epitope peptide only) (Fig. 2). Imiquimod and CpG induced higher level of IL-12 by mdDCs (495.5 ± 132.7 and 484.3 ± 140.0 pg/ml, respectively) than that of the control group (P < 0.05) but much lower than that of LPS or PIC group (P < 0.01) (Fig. 2). IFN-α was poorly produced by mdDCs exposed to the studied TLR agonists except the mdDCs stimulated with LPS (2045.6 ± 389.2 pg/ml). IL-10 was undetectable by all the TLR agonist-stimulated mdDCs except LPS-stimulated mdDCs (1089.7 ± 264.0 pg/ml).

Fig. 2
figure 2

IL-12 p70 production by mdDCs stimulated with various TLR agonists for 48 h. Supernatants were assessed for IL-12 production by ELISA. Results shown represent the mean ± SD from five donors. *Indicates statistical significance (P < 0.05) when compared with the control group. **Indicates statistical significance (P < 0.05) when compared with the LPS group

Full size image

TLR agonist pre-treated mdDCs enhanced CD4+ naïve T cell differentiation

Th1 and Th2 development was assessed by quantifying IFN-γ and IL-4 secreted by CD4+ naïve T cell. As shown in Fig. 3, no IFN-γ secretion was detected from the unstimulated naïve T cells. And the IFN-γ secretion of naïve T cells was increased by HPV 11 E7 epitope peptide-loaded mdDCs stimulated with all TLR agonists. Among these agonists, PIC-treated mdDCs stimulated naïve T cells to produce IFN-γ nearly twofold higher than that of LPS, imiquimod or CpG groups. On the contrary, IL-4 was not detectable in each group (the test range of ELISA kit is from 0 to 500 pg/ml).

Fig. 3
figure 3

Effect of TLR agonists on cytokine production by CD4+ Th cells. Allogenic CD4+CD45RA+ T cells were co-cultured for 6 days with effector mdDCs pretreated with the indicated TLR agonists. The release of IFN-γ and IL-4 (not shown in the figure) was measured by ELISA. Results are representative of three independent experiments

Full size image

Cytokine production of T cells induced by mdDCs pretreated with various TLR agonists

T cells stimulated with HPV 11 E7 epitope peptide-loaded mdDCs which were subsequently stimulated with LPS or PIC, but not imiquimod or CpG ODN, secreted strikingly higher levels of IFN-γ (Fig. 4a), TNF-α (Fig. 4b), and IL-2 (Fig. 4c), compared with mdDCs loaded with E7 only. IFN-γ, TNF-α, and IL-2 production of the T cells stimulated with mdDCs which were pretreated with a combination of PIC, imiquimod, and CpG ODN was not higher than that of the T cells stimulated with the PIC-pretreated mdDCs (Fig. 4).

Fig. 4
figure 4

Cytokine production of IFN-γ, TNF-α and IL-2 by T cells in response to TLR agonists and HPV 11 E7 epitope peptide-pretreated mdDCs. Allogenic T cells were co-cultured with 2 × 105 indicated mature mdDCs. Cytokine production from each culture was measured after 14 days by ELISA. Results shown are means ± SD of three independent experiments

Full size image

The frequencies of IFN-γ and IL-2-producing T cells after stimulation with various TLR agonists pretreated mdDCs

As shown in Fig. 5, mdDCs pretreated with HPV 11 E7 epitope peptide plus LPS or PIC could induce more IFN-γ- and IL-2-secreting cells than irrelevant peptide pulsing mdDCs or PBS (P < 0.05). MdDCs stimulated with E7 epitope peptide plus LPS or PIC, but not plus imiquimod or CpG ODN, had significantly higher frequencies of IFN-γ- (Fig. 5a) and IL-2- (Fig. 5b) producing cells when compared to mdDCs stimulated with E7 epitope peptide alone or E7 plus IL-10 (all P < 0.01). There was no difference between the frequencies of IFN-γ- (Fig. 5a) and IL-2- (Fig. 5b) producing cells elicited by mdDCs pretreated with TLR agonists plus E7 and mdDCs pretreated with various TLR agonists alone (all P > 0.10).

Fig. 5
figure 5

IFN-γ and IL-2 ELISPOT assays. 1 × 105 purified allogenic T cells were cocultured with 2 × 104 the indicated effector mdDCs in vitro, and the frequency of IFN-γ or IL-2 secreting cells was assayed with the human IFN-γ/IL-2 ELISPOT kit according to the manufacturer’s instructions. The values are expressed as spot-forming cells (SFCs) per 106 T cells. The data are the mean ± SD of triplicate values

Full size image

Cytotoxicity of the E7 epitope peptide-specific CD8+T cells towards E7-expressing cells

The E7 epitope peptide-specific CD8+ T cells primed by all TLR agonist-pretreated mdDCs could induce lysis of E7-expressing cells. LPS- or PIC-pretreated mdDCs induced a strong CTL activity (83.1 and 81.3%, respectively) in an effector/target cell (E/T) ratio of 100:1, that is much more efficient than that of mdDCs stimulated with E7 epitope peptide only (56.5% cytotoxicity), mdDCs stimulated with non-sense peptide (19.7% cytotoxicity), and mdDCs unstimulated (16.1% cytotoxicity) (Fig. 6). The CD8+ T cells primed by imiquimod or CpG-pretreated mdDCs exhibited 62.2 and 59.7% cytotoxicity, respectively, in an E/T ratio of 100:1. No difference was observed when compared with mdDCs stimulated with E7 only. The CTL activities in the LPS, PIC, imiquimod, and CPG ODN group were above 50% in the E/T ratio of 50:1.

Fig. 6
figure 6

Cytotoxicity of the HPV 11 E7 epitope peptide specific CD8+ T cells against HPV 11 E7/B16 cells. CD8+ T cells were pretreated with peptide (HPV 11 E7 or non-sense pepitide) pulsed mdDCs stimulated or unstimulated with TLR agonists. The cytotoxicity against target cells was determined by the release of LDH after 6 h of co-culture at different effector/target cell ratio. Percentage of cytotoxicity was calculated as described. The result is a representation of three independent experiments

Full size image

Discussion

In order to evaluate the immunostimulatory effects of various TLR agonists upon mdDCs loading immunodominant CTL epitope peptide of HPV 11 E7, we assessed the phenotypic maturation and the cytokine production of mdDCs, as well as their ability to augment T cell responses including cytokine production and specific CTL activity.

MdDCs maturation is an important step in their functional capacity to activate T cells. The phenotypic maturation of mdDCs after exposure to various TLR agonists was characterized in the current experiment. MdDCs’ maturation in response to PIC and LPS was demonstrated by significant up-regulation of the costimulatory molecules CD40, CD80, CD86, maturation-associated marker CD83, and MHC class II (HLA-DR), while imiquimod and CpG induced less, but still notable phenotypic changes. It indicated that TLR agonists, especially PIC and LPS, could induce the maturation of mdDCs and might be severed as alternative maturation stimulus of mdDCs apart from the classical approaches.

A serious disadvantage of current ex vivo generated human mdDC is the poor production of IL-12p70 which is important for the type 1 polarization of T cell immunity [3, 35]. Prior studies showed that immature mdDCs produced high levels of IL-12 by TLR3, TLR4, TLR7/8, or TLR9 ligand stimulation [13, 24]. MdDCs matured with LPS, PIC, and/or R848 (a much more potent ligand of TLR7/8 than imiquimod) [37] were able to produce vast amounts of IL-12p70 [6]. TLRs thereby could be effective targets for enhancing IL-12 production of mdDCs. In accordance with these studies, we found that mdDCs were much more susceptible to PIC or LPS stimulation than imiquimod or CpG ODN stimulation in respect of IL-12 secretion. However, IFN-α, which can orient DC functions towards the priming and expansion of protective antiviral immune responses [34], was not detectable in the TLR-activated mdDCs except for LPS stimulation. This might be explained by the low level of IFN regulatory factor (IRF)-7 expression in immature mdDCs [10, 19], whereas the IFN-α production induced by LPS might be largely dependent on its ligation with the remaining CD14 (another receptor of LPS) expressing on the mdDCs rather than with TLR4. Therefore, of the cytokines measured, IL-12 p70 is the predominant cytokine up-regulated by TLR agonists in activated mdDCs.

IL-10, which increased in CA lesion [2, 29], was reported to impair DC function by holding the DCs in immaturity [20, 28]. The apparent inhibition of IL-10 on mdDCs’ maturation in this study is consistent with some previous studies [8]. We also found that LPS- or PIC-induced maturation of mdDCs was less suppressed by IL-10 as compared to imiquimod and CpG ODN. It suggests that the mdDCs’ maturation in the circumstance of increased IL-10, such as in CA lesions, could be partially restored by LPS or PIC.

The increased production of IL-12 and the decreased production of IL-10 would be predicted to affect Th effector cell development. By co-culturing various TLR agonist-pretreated effector DCs with naive Th cells, we delineated the ability of these agonist-treated DCs to regulate specific subset differentiation of CD4+ T cells. Evaluation of cytokine production profiles revealed that after stimulation with TLR agonists, E7-loaded mdDCs stimulated naïve CD4 ± T cells to produce high level of IFN-γ. PIC-pretreated mdDCs were much more effective in stimulating naïve T cells to produce IFN-γ than LPS-, imiquimod- or CpG ODN- pretreated mdDCs. However, the Th2 type cytokine, IL-4, was undetectable. It indicates that PIC-pretreated mdDCs, could strikingly induce the Th1 response and will favor the production of CTL effectors which are important in clearing virally infected cells.

We also found that TLR ligand-induced activation of mdDCs significantly enhanced production of cytokines (IFN-γ, TNF-α and IL-2) and accordingly increased frequencies of activated T cells. The PIC was the most effective stimuli for mdDCs as compared to the positive control LPS. Importantly, CD8+ T cells primed by LPS- and PIC- pretreated mdDCs loading of HPV 11 E7 epitope peptide strikingly increased specific CTL activity against HPV 11 E7-expressing cells as compared to T cells primed by mdDCs stimulated with imiquimod or CpG ODN, mdDCs loading of HPV 11 E7 epitope peptide, or mdDCs loading of non-sense epitope. This suggests that E7-specific anti-viral T cell immunity can be mostly augmented through LPS or PIC-modulated mdDCs. Imiquimod and CpG ODN were relatively poor adjuvants for eliciting T cell responses against HPV 11 E7-expressing cells.

Renn et al. [31 ] analyzed the TLR 1-10 expression in mdDCs by real-time PCR. They found that mdDCs expressed mRNAs for TLR 1-10 with high expression of TLR2 mRNA, intermediate expression of TLR8, TLR4, TLR3, and TLR10 mRNAs, and low expression of TLR1, TLR5, TLR6, TLR7, and TLR9 mRNAs. It was reasonable that in the current study, the activation of mdDCs by the TLR3 agonist PIC and the TLR4 agonist LPS was much higher than the activation induced by TLR7 agonist imiquimod, and TLR9 agonist CpG ODN. Therefore, LPS and PIC might be more promising adjuvants, than imiquimod and CpG ODN, in mdDC-based vaccine strategies against CA.

In summary, our study provides a reference of using predominant CTL epitope of HPV 11 E7 and TLR agonist immunization for eliciting specific cellular immune responses. These data might favor the rational design of TLR-based therapeutic vaccines or immune-modulating therapy against CA. Further investigation in mouse model expressing HPV 11 E7 is under research now.