Prothymosin α and a prothymosin α-derived peptide enhance TH1-type immune responses against defined HER-2/neu epitopes

Background Active cancer immunotherapies are beginning to yield clinical benefit, especially those using peptide-pulsed dendritic cells (DCs). Different adjuvants, including Toll-like receptor (TLR) agonists, commonly co-administered to cancer patients as part of a DC-based vaccine, are being widely tested in the clinical setting. However, endogenous DCs in tumor-bearing individuals are often dysfunctional, suggesting that ex vivo educated DCs might be superior inducers of anti-tumor immune responses. We have previously shown that prothymosin alpha (proTα) and its immunoreactive decapeptide proTα(100–109) induce the maturation of human DCs in vitro. The aim of this study was to investigate whether proTα- or proTα(100–109)-matured DCs are functionally competent and to provide preliminary evidence for the mode of action of these agents. Results Monocyte-derived DCs matured in vitro with proTα or proTα(100–109) express co-stimulatory molecules and secrete pro-inflammatory cytokines. ProTα- and proTα(100–109)-matured DCs pulsed with HER-2/neu peptides induce TH1-type immune responses, prime autologous naïve CD8-positive (+) T cells to lyse targets expressing the HER-2/neu epitopes and to express a polyfunctional profile, and stimulate CD4+ T cell proliferation in an HER-2/neu peptide-dependent manner. DC maturation induced by proTα and proTα(100–109) is likely mediated via TLR-4, as shown by assessing TLR-4 surface expression and the levels of the intracellular adaptor molecules TIRAP, MyD88 and TRIF. Conclusions Our results suggest that proTα and proTα(100–109) induce both the maturation and the T cell stimulatory capacity of DCs. Although further studies are needed, evidence for a possible proTα and proTα(100–109) interaction with TLR-4 is provided. The initial hypothesis that proTα and the proTα-derived immunoactive decapeptide act as “alarmins”, provides a rationale for their eventual use as adjuvants in DC-based anti-cancer immunotherapy.


Background
Anti-cancer vaccines are designed to break tolerance to self and stimulate strong and durable anti-tumor immunity. Administering defined tumor-derived epitopes to cancer patients for the activation of helper and cytotoxic T cells has been shown to enhance anti-cancer immune responses in vivo and in some cases to lead to objective clinical responses [1][2][3]. To optimize the efficacy of peptide-based anti-cancer vaccines, combinatorial approaches stimulating both innate and adaptive immunity are now being clinically evaluated [4,5]. Mature dendritic cells (DCs) are key players for eliciting such responses, as they present antigens to T cells and provide the necessary co-stimulatory signals and cytokines favoring the efficient activation of tumor-reactive immune cells [6,7]. DC maturation can be induced in vivo upon admixing and co-administering immunogenic peptides with adjuvants, but to date this strategy has been proven successful only when vaccinating against common pathogens [8]. In cancer patients, the presence of tumor-associated suppressive factors impairs endogenous DC functions [9], a condition that can be bypassed only by the adoptive transfer of ex vivo matured immunocompetent DCs [10,11].
In normal living cells, proTα is localized in the nucleus where it controls the cell cycle and promotes cell proliferation. Released from dead cells, extracellular proTα acquires multi-functional immunomodulatory properties [18]. We and others have previously shown that proTα upregulates the expression of IRAK-4 in human monocytes [19], ligates TLR-4 on murine macrophages and signals through MyD88-dependent and independent pathways [20]. Similar to its immunoreactive decapeptide proTα(100-109) [21], it upregulates the expression of HLA-DR [22], CD80, CD83 and CD86 and promotes maturation of human DCs in vitro [23].
Here, we show that DCs matured ex vivo in the presence of proTα or proTα(100-109) are not only phenotypically but also functionally competent, secrete pro-inflammatory cytokines and induce T H 1-type immune responses in the presence of tumor-associated immunogenic epitopes of the oncoprotein HER-2/neu. DCs matured with proTα or proTα(100-109) prime naïve CD8-positive (+) T cells to exert HER-2/neu peptide-specific cytotoxicity and CD4+ T cells to proliferate in a peptide-dependent manner. Finally, we provide preliminary evidence suggesting that both proTα and its decapeptide proTα(100-109) likely signal via TLR-4 in human DCs.

Results
Phenotype of and cytokine production by proTα-or proTα(100-109)-matured DCs We have previously shown that proTα and proTα(100-109) efficiently mature human DCs in vitro, as indicated by the induction of surface expression of established DC-markers to levels comparable to those induced by lipopolysacharide (LPS) [23] or tumor necrosis factor (TNF)-α (this report). As shown in Figure 1, LPSmatured DCs significantly upregulated the expression of HLA-DR, CD11b, CD80, CD83, CD86 and CD40, to levels comparable to TNF-α-matured DCs (p < 0.001 compared to iDCs for all values). Similarly, both agents caused a reduction of CD14 expression. In agreement with our previous study [23], iDCs matured with either proTα or its decapeptide presented a similar phenotype to LPS-or TNF-α-matured DCs, upregulating the expression of all co-stimulatory molecules (p < 0.05-0.001, compared to iDCs; Figure 1) and downregulating CD14 (p < 0.01, compared to iDCs).
In conjunction with their phenotype, functionally competent mature DCs secrete pro-and anti-inflammatory cytokines [6]. Therefore, we assessed the production of TNF-α, interleukin (IL)-12 and IL-10 from iDCs and mature DCs and determined the IL-12:IL-10 ratio. We present these data because high TNF-α levels, as well as a balance between IL-12:IL-10 in favor of IL-12 have been proposed to promote T H 1-polarization [24]. As shown in Figure 2, iDCs produced low amounts of all three cytokines, whereas mature DC supernatants collected 48 h after addition of LPS or TNF-α were rich in TNF-α and IL-12. Compared to iDCs, higher TNF-α and IL-12 levels were also found in supernatants of cultures of DCs matured with proTα and proTα(100-109). Although the absolute concentrations of cytokines varied among the differentially matured DCs, their overall cytokine-production patterns were comparable. Most importantly, the mean IL-12:IL-10 ratios were similar (6.61, 6.45, 7.89 and 5.18, for LPS-, TNF-α-, proTα-and proTα(100-109)-matured DCs, respectively). These data suggest that the peptides bias immunoreactivity towards a pro-inflammatory T H 1-type of response.

ProTα-and proTα(100-109)-matured DCs stimulate tumor peptide-specific CD8+ T cell responses
Cell-mediated immunity requires initial collaboration between T H 1 CD4+ and CD8+ T cells [26]. Thus, we next investigated whether proTα-and proTα(100-109)-matured DCs can elicit tumor peptide-specific cytotoxic T cell immune responses. CD8+ T cells recovered from the same stimulation cultures as aforementioned were assessed for the intracellular production of TNF-α. As shown in Figure 4A,  Where indicated, mAb to MHC class I molecules was added throughout the culture period at a final concentration of 5 μg/mL. In all assays the E: T ratio was 10:1. Data represent mean % cytotoxicity ± SD from 2-5 donors.
they also exhibited a similar pattern of enhanced cytokine production in the presence of HER-2/neu peptides as did CD4+ T cells. The percentage of TNF-α+ cells was increased from 0.35% (unpulsed) to 47.52% (pulsed) when T cells were stimulated with TNF-α-matured DCs, and from 0.12% to 45.38% for proTα-and from 0.13% to 42.88% for proTα(100-109)-matured DCs ( Figure 4A). In addition and in accordance with the results recorded for CD4+ T cells, IL-2-and IFN-γ-producing CD8+ T cells were also increased in the presence of peptide-pulsed DCs in the cultures, whereas differences in the percentages of IL-10-producing CD8+ T cells were only marginal (Additional file 1: Table S1A). The same cells were assessed for the expression of CD107a, as a surrogate marker for cytotoxicity [27]. In the absence of HER-2(9 369 ), a low percentage of CD8+ T cells stimulated with TNF-α-matured DCs expressed CD107a (3.70%; Figure 4A), which increased when cells were stimulated with HER-2(9 369 )-pulsed DCs (54.75%). Similar CD107a upregulation was observed in CD8+ T cells stimulated with proTα-and proTα(100-109)matured HER-2(9 369 )-pulsed DCs (36.86% and 41.99%, respectively, compared to 2.80% and 2.17% of the unpulsed groups; Figure 4A). Since TNF-α mediates target cell damage and CD107a-expressing CD8+ T cells are cytotoxic [27], our results suggest that proTα-and proTα(100-109)-matured DCs efficiently activate CD8+ cytotoxic T cells, which were able to kill targets presenting the immunogenic epitope versus which they were primed.
ProTα and proTα(100-109) induce the maturation of DCs via triggering TLR-4 We have previously reported that stimulation of human monocytes with proTα upregulated IRAK-4, a protein kinase involved in TLR downstream signaling [19], whereas Mosoian et al. [20] showed that proTα ligates TLR-4 and signals through both TRIF-and MyD88dependent pathways. To determine whether TLR-4 is triggered by our peptides, we studied the kinetics of TLR-4 surface expression on proTα and proTα(100-109)stimulated DCs. Immature DCs (iDCs) and DCs matured with LPS (a known TLR-4 ligand; [30]), proTα or proTα(100-109) for 15 min, 30 min, 1 h, 18 h and 36 h were analyzed by flow cytometry. The percentage of surface TLR-4 expression over time is presented in Figure 6. Maturation of DCs with LPS led to an early (15 and 30 min) decrease of TLR-4 expression (by~15%) due to internalization [31], and a subsequent increase from 1 to 18 h [32]. To extend these findings, we next investigated the intracellular expression levels of three adaptor molecules that participate in signaling pathways downstream of TLR-4, namely TRIF, an adaptor molecule common to TLR-3 and -4 signaling; TIRAP, a signaling adaptor common to TLR-2, and -4; and MyD88, a molecule upregulated upon ligation of all TLRs except TLR-3 [33]. We specifically selected these three adaptors because this constellation is unique to TLR-4 activation. Total cell extracts from iDCs and DCs matured with LPS, proTα or proTα(100-109) for 1 h and 18 h were immunoblotted ( Figure 7A). Upon densitometric quantification of each protein band detected, expression relative to GAPDH was calculated. As shown in Figure 7B, addition of LPS led to a significant~2-3 fold increase of the expression of all three adaptors within 1 h (3.05 for TRIF, 2.88 for TIRAP and 1.81 for MyD88) relative to iDCs (1.38 for TRIF, 1.00 for TIRAP and 0.74 for MyD88). At 18 h post-addition of , although in the latter case, the detected protein levels were lower. As with LPS, 18 h after proTα or proTα(100-109) DC-stimulation, the expression of TRIF, TIRAP and MyD88 was reduced and was similar to iDCs. These data, in conjunction with the cytokine profile shown in Figure 2, suggest that LPS, proTα, and possibly also proTα(100-109), activate DCs at least partly through one common TLR-4-dependent intracellular signaling pathway.

Discussion
We have previously shown that human monocytederived iDCs activated in vitro with proTα or its immunoreactive decapeptide, proTα(100-109), acquire a mature DC phenotype [23]. Here, we show that DC maturation induced by proTα or proTα(100-109) promotes the secretion of IL-12, rather than IL-10, from these cells. Thus, both proTα-and proTα(100-109)-matured DCs possess immunostimulatory properties appropriate for the efficient activation of T cells, through their enhanced antigen-presenting capacity (HLA-DR; signal 1), the increased expression of co-stimulatory molecules (CD80/CD86; signal 2) and the secretion of inflammatory mediators (IL-12), recently proposed to act as signal 3 for optimizing effector T cell functions [34,35]. We assessed whether these ex vivo generated DCs can present tumor-associated immunogenic peptides to autologous T cells, along with the appropriate signals for their activation. We pulsed DCs with one MHC class Iand one class II-restricted immunodominant epitope from the oncoprotein HER-2/neu, HER-2(9 369 ) and HER-2/neu(15 776 ), respectively [36,37]. Our results show that proTα-or proTα(100-109)-matured HER-2/neu peptide-pulsed DCs favor the generation of T H 1-type immune responses in vitro, by polarizing CD4+ T cells to produce pro-inflammatory cytokines. This cytokine milieu, characterized by high levels of IFN-γ and IL-2, results in the generation of strong CD8+ T cell responses [26,38], as we also observed. Indeed, CD8+ effectors recovered from the same stimulation cultures exhibited a pro-inflammatory cytokine profile similar to the CD4+ T cells (Additional file 1: Tables S1A and B) and enhanced HER-2(9 369 )-specific MHC class I-restricted cytotoxicity. Of interest, a high percentage of the peptide- specific CD8+ T cells generated in our stimulation cultures were polyfunctional, a quality reportedly associated with superior T cell performance [28,29,39]. These findings, in conjunction with the observed enhancement of HER-2(15 776 )-specific T cell proliferation, suggest that in the presence of tumor antigenic peptides, proTα-and proTα(100-109)-matured DCs efficiently promote the expansion of peptide-specific T cells.
Different DC-stimulating agents, including TLR ligands, have long been and still are being explored to optimize the immunostimulatory properties of DCs [10,11,40,41]. Although it was initially proposed that TLRs recognized only PAMPs, accumulating evidence to date suggests that TLRs also bind and respond to endogenous ligands released during tissue injury and inflammation, termed DAMPs or "alarmins" [42]. Most prominent among the alarmins are HMGB1, members of the HSP family and granulysin [43], all of which mature and activate DCs in vitro and bias immune responses towards a T H 1-type, when used as vaccine adjuvants in vivo [44][45][46][47][48]. We and others have previously shown that proTα promotes antigen-specific adaptive immune responses [20,[49][50][51][52] and based on the data presented herein, we now identify proTα as an alarmin. Moreover, in line with data on immunoreactive peptidefragments derived from either HMGB1 (Hp91; [53]) or HSP70 (HSP70 359-610 ; [46]), we show that the immunologically active site of proTα, the decapeptide proTα(100-109) [23], also favors T H 1-polarization and induces HER-2/neu peptide-specific immune responses.
Ex vivo education of DCs by proTα or proTα(100-109) results in their polarization to type-1 DCs, with increased capacity to stimulate tumor peptide-specific T cell responses and to render cytotoxic T cells polyfunctional. If this holds true also in vivo, then these molecules could be promising components of DC-based anti-cancer vaccines.
For T cell stimulation, 48 h matured DCs (1×10 6 /mL) were pulsed with 50 μg/mL HER-2(9 369 ) and HER-2 (15 776 ) for 6 h at 37°C, in a humidified 5% CO 2 incubator in X-VIVO 15. DCs were washed twice, resuspended in X-VIVO 15 and added to autologous lymphocytes (non-adherent fraction) at a DC:lymphocyte ratio of 1:10. T cells were stimulated thrice at weekly intervals and on days 3 and 5 after each stimulation, 40 IU/mL IL-2 (Proleukin; Novartis Pharmaceuticals Ltd, UK) were added to the cultures. At the third stimulation, Golgi-Plug (1 μL/mL; Becton-Dickinson (BD) Biosciences, Erembodegem, Belgium) was added in the cultures, and 12 h later, T cells were harvested and analyzed for cytokine production by flow cytometry.

Flow cytometry analysis
For DC phenotype analysis, iDCs and mature DCs were stained for the surface molecules HLA-DR, CD80, CD83, CD86, CD11b, CD40 and CD14. Triple staining was performed using appropriate combinations of FITC-, PE-or PE-Cy5-labelled mouse anti-human IgG1 and IgG2 mAbs (BD Biosciences) at saturating concentrations for 30 min on ice. DCs were also stained with irrelevant antihuman IgG1 and IgG2 mAbs (BD Biosciences), as isotype controls. Samples were measured using a FACSCalibur flow cytometer (BD Biosciences) and data were analyzed using CellQuest software. MFI was evaluated for each marker.
To determine maximal and spontaneous isotope release, targets were incubated with 3 N HCl and in plain medium, respectively. All cultures were set in triplicate. Percentage of specific cytotoxicity was calculated according to the formula: [(cpm experimental-cpm spontaneous)/ (cpm maximal-cpm spontaneous)] ×100.

Proliferation assay
Stimulated T cells were seeded in 96-well U-bottom plates (1 × 10 6 /mL; 100 μL). Autologous matured DCs pulsed with 50 μg/mL HER-2(15 776 ) or tyr (15 448 ) for 6 h, were added (1 × 10 5 /mL; 100 μL/well) and cocultured for 5 days. T cells incubated with unpulsed matured DCs or in the presence of IL-2 (500 IU/mL) were used as controls. Where indicated, mAb to MHC class II molecules (L243, kindly donated by Prof. S. Stevanovic) was added to the cultures at a concentration of 5 μg/mL for the entire culture period [66]. For the last 18 h of culture, 1 μCi 3 H-thymidine (Amersham Pharmacia Biotech, Amersham, Bucks, UK) was added per well and cells were harvested in a semi-automatic cell harvester (Skatron Inc., Tranby, Norway). The amount of incorporated radioactivity, proportional to DNA synthesis, was measured in a liquid scintillation counter (Wallac, Turku, Finland) and expressed as cpm. The S.I. of each experimental group was calculated using the formula: (average cpm of sample in the presence of peptidepulsed DCs)/(average cpm of sample in the presence of unpulsed DCs).

Statistical analysis
Data were analyzed by the Student's t-test and statistical significance was presumed at significance level of 5% (p < 0.05).
Additional file 2: Figure S1. Kinetics of CD14 and TLR-4 surface expression on monocytes, macrophages and iDCs/DCs upon stimulation with LPS, proTα or proTα(100-109). Monocytes, macrophages and iDCs (0 h) were stimulated with LPS (A), proTα (B), or proTα(100-109) (C) for 15 min, 30 min, 1 h and 18 h and assessed for the surface expression of CD14 and TLR-4 using flow cytometry. MFI values in the presence of neutralizing anti-TLR-4 Ab (+ a-TLR-4) are shown below each histogram. Histograms are from one representative donor of 3 tested. Using the loss of cell surface expression as a readout for TLR-4 and CD14 endocytosis from 0-36 h [31], data from all three donors are shown as mean values ± SDs for TLR-4 (D, E, F) and CD14 (G, H, I).
Additional file 3: Figure S2. CD14, TLR-4 and CD206 expression on monocytes, monocyte-derived macrophages and monocyte-derived iDCs. Macrophages were generated from human monocytes upon incubation with 100 ng/mL GM-CSF for 5 days. Human monocytes were isolated and iDCs were generated as described in Methods. Monocytes, macrophages and iDCs were assessed for the surface expression of CD14, TLR-4 and CD206 (as a specific marker for macrophages and DCs), using flow cytometry. Histograms are from one representative donor of 3 tested and numbers indicate MFIs.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions KI: performed the experiments, analyzed data, carried out statistical analyses and wrote the manuscript. ED: designed, analysed and interpreted flow cytometry data and helped to write the manuscript. ET: participated in immunoblotting data acquisition and analyses. PS: performed sample collection and helped to draft the manuscript. HK: carried out peptide synthesis and purification and helped to draft the manuscript. WV: helped in HLA-typing and to draft the manuscript. IPT: participated in the design of the study and reviewed the manuscript. GP: participated in the design and coordination of the study, helped to draft, reviewed and edited the manuscript. OET: conceived, designed and coordinated the study, drafted and reviewed the manuscript. All authors read and approved the final manuscript.