Recruitment of bone marrow CD11b+Gr-1+ cells by polymeric nanoparticles for antigen cross-presentation

The objective of this study was to investigate the function of poly(lactic-co-glycolic acid) (PLGA) nanoparticles (NPs) on the activation of antigen-specific CD8+ T cell responses via the CD11b+Gr−1+ myeloid subpopulations in murine bone marrow (BM). PLGA NPs containing ovalbumin (OVA) were fabricated by the double-emulsion method. The CD11b+Gr-1lowLy-6Chigh and CD11b+Gr-1highLy-6Clow subsets from mice bone marrow were sorted and treated with the PLGA/OVA NPs, followed by co-culture with the carboxyfluorescein succinimidyl ester (CFSE)-labelled OT-I CD8+ cells. Co-culture of OT-I CD8+ T cells with PLGA/OVA NPs-primed CD11b+Gr-1+ subsets upregulated the expression of IL-2, TNF-α, INF-γ, granzyme B, and perforin, resulting in proliferation of CD8+ T cells and differentiation into effector cytotoxic T lymphocytes (CTLs). In vivo proliferation of CFSE-labelled OT-I CD8+ cells in response to OVA was also obtained in the animals immunized with PLGA/OVA NPs. The results presented in this study demonstrate the ability of polymeric NPs to recruit two CD11b+Gr−1+ myeloid subsets for effective presentation of exogenous antigen to OT-I CD8+ T cells in the context of major histocompatibility complex (MHC) class I, leading to an induction of antigen-specific cell proliferation and differentiation into effector cells.

Biodegradable polymeric micro-or nano-particles (NPs) are of great interest in the field of drug delivery and have been extensively studied in vaccine delivery for the enhancement of presentation of exogenous antigens 1-6 , a process referred to as cross-presentation or cross-priming, in which the antigenic fragment derived from exogenous proteins is bound to the major histocompatibility complex (MHC) class I molecules of the antigen presenting cells (APCs) to stimulate the CD8 + T immune response [7][8][9] . The induction of cytotoxic CD8 + T cell-mediated immunity plays a pivotal role in the development of immunotherapeutic strategies against infection and cancer. Dendritic cells (DCs), the professional APCs in the processing and presentation of exogenous antigens, have served as the major target cells for antigen delivery to enhance vaccine efficacy [10][11][12][13][14] . Although it was reported in earlier studies that particulate antigens can promote presentation of the associated antigens to T cells via both macrophage and non-macrophage APCs that phagocytose the particles 15 , the delivery of antigens by nanoparticles (NPs) to other APCs for the elicitation of MHC class I immunity unfortunately has been largely ignored. The ability of neutrophils to process the phagocytosed bacteria via the MHC Class I pathway to trigger the CD8 + T cell responses and their ability to stimulate in vitro cross presentation of exogenous antigens employing the B3Z model have been previously reported 16,17 . Our recent study also demonstrated the activation of CD8 + T cells by the nanoparticles-primed Gr-1 high cells 18 . These results prompted us to further evaluate the potential of granulocytes from murine bone marrow to induce activation of cytotoxic T lymphocyte (CTL) effectors in nanoparticle (NPs)-based vaccination.
Immature myeloid cells in the bone marrow (BM) are a heterogeneous population of cells that differentiate into protective cell types such as granulocytes and macrophages 19 . BM granulocytes can be phenotypically characterized by the expression of the surface proteins CD11b and Gr-1, including the two isoforms Ly6C and Ly6G 19,20 . The CD11b + Gr-1 + subset is a heterogeneous myeloid population comprising at least two subsets: polymorphonuclear (PMN) and monocytic cells 21 . The polymorphonuclear granulocytes are the most abundant leukocytes continuously released from bone marrow (BM) into the blood circulation, and they play a critical role in innate immunity. Despite the established phagocytic activity of granulocytes, the role of BM CD11b + Gr-1 + cells in MHC class I antigen processing and presentation via polymeric nanoparticles (NPs) has been ignored.
In this study, we employed the anti-Gr-1 monoclonal antibody (RB6-8C5), previously used to detect the granulocyte-differentiation antigen on more differentiated granulocytes 22 , to characterize the two subsets of BM myeloid subsets, including the CD11b + Gr-1 high Ly-6C low (abbreviated as Gr-1 high ) subset that exhibits a polymorphonuclear or band-shaped nuclear morphology and the CD11b + Gr-1 low Ly-6C high (abbreviated as Gr-1 low ) subset, with a mononuclear morphology. We attempted to elucidate the role of CD11b + Gr-1 + polymorphonuclear (PMN) granulocytes in antigen cross presentation after treatment with the nanoparticle-based antigens. The CD8 + T cells from OT-I mice, expressing the transgenic T cell receptor (TCR) specific for OVA peptide residues 257-264 in the context of H2K b , were used to assess the effects of PLGA/OVA NPs on the activation of the OVA-specific CD8 + T cell response and the induction of the cytotoxic lymphocyte (CTL) effect. It was assumed that upon activation by the polymeric NPs-primed CD11b + Gr-1 + granulocytes, the antigen-specific CD8 + T cells undergo proliferation and differentiation into effectors (clonal expansion) that recognize specific peptides on MHC class I complexes and express type 1 cytokines, such as IFN-γ , TNF-α , and IL-2, for the elicitation of cytotoxicity (target elimination) 23,24 . The cytotoxic T lymphocytes (CTLs) are effector lymphocytes that play important roles in defence immunity against infectious diseases and cancers, in which perforin and granzyme B are involved in the induction of cell death, contributing to an efficient generation of immune effectors in the antigen specific immune response 25 .
The results of this study illustrated that priming the Gr-1 high and Gr-1 low subsets of BM CD11b + Gr-1 + cells with the PLGA/OVA NPs induced the expansion and proliferation of OVA-specific OT-I CD8 + T cells, resulting in an antigen-specific immune response in the context of MHC class I complexes.

Results
We investigated the potential recruitment of polymorphonuclear (PMN) granulocytes from the mouse bone marrow (BM) by polymeric nanoparticles (NPs) for cross-presentation of exogenous antigens and stimulation of cytotoxic T lymphocyte (CTL) effector functions.
Scanning electron microscopy (SEM) of PLGA/OVA NPs. The electron micrograph of the PLGA/OVA NPs, prepared by the double emulsion and solvent evaporation method, is shown in Fig. 1a, illustrating the spherical shape with a mean particle size of 690 ± 7.1 nm. Physicochemical characterization of the PLGA/OVA NPs and the release profile of antigen from PLGA/OVA NPs have been previously reported 18 . The mean particle size of the PLGA/OVA NPs was 690 ± 7.1 nm, and the size distribution is shown in Fig. 1b. The zeta potential of the PLGA/ OVA NPs, determined by a Malvern Zetasizer Nano ZS, was −30.5 ± 0.5 mV, and the encapsulation efficiency (the ratio of the amount of ovalbumin in the NPs to the initial amount loaded) was approximately 52.7% 18 .
Sorting of Lineage2 − CD11b + Gr-1 high and Lineage2 − CD11b + Gr-1 low subsets by fluorescenceactivated cell sorting (FACS) and microscopic examination. Bone marrow (BM) cells were flushed and isolated from the femurs and tibias of mice aged 6-8 weeks. The Lin2 − CD11b + Gr-1 high Ly-6C low (Gr-1 high ) and Lin2 − CD11b + Gr-1 low Ly-6C high (Gr-1 low ) subsets were sorted using a FACSAria III (BD Biosciences, San Jose, CA) (Lin2: CD3/CD19/B220/Ter119/NK1.1) 18,26 , and the morphology of the sorted populations was assessed by Wright-Giemsa staining. The gating scheme for sorting of these two subpopulations is presented in Fig. 1c. The phenotypic analysis by flow cytometry showed that Gr-1 high cells are Ly-6C low and polymorphonuclear in morphology, whereas most Gr-1 low cells are Ly-6C high and mononuclear with a heterogeneous nuclear shape. In contrast to the expression of CD115 in 35.6% of the Gr-1 low subset, most Gr-1 high cells do not express CD115 ( Fig. 1c-g).
Proliferation and expression of the activation markers CD25 and CD69 in OT-I CD8 + T cells after co-culture with PLGA/OVA NPs-primed Gr-1 + cells. To examine the effect of Gr-1 + cells after pre-treatment with the polymeric nanoparticles on the proliferation of OT-I CD8 + T cells in vitro, CD8 + TCRVβ 5 + T cells were enriched from the spleens of OT-I mice by negative selection. The purity of the sorted fraction was approximately 96.5% (Fig. 2a). These cells were then labelled with carboxyfluorescein succinimidyl ester (CFSE) and co-cultured for 72 hrs with the sorted Gr-1 + cells, either the Gr-1 high or Gr-1 low subset, that were pre-treated for 16 hrs with PBS, OVA, PLGA NPs, or PLGA/OVA NPs. The OT-I CD8 + T cell response, assessed by the CFSE-dilution assay, showed an extensive cell division and proliferation in response to the treatment with PLGA/ OVA NPs compared to the control groups treated with PBS, OVA, or PLGA NPs alone ( Fig. 2b-f).
CD25, the α chain of the IL-2 receptor, and CD69 are T cell activation markers. To determine if the Gr-1 + cells primed with PLGA/OVA NPs induced CD 25 and CD69 expression on T cells, the cells were surface stained with fluorochrome-conjugated antibodies against CD25 and CD69. Figure 2b-f illustrate a significant proliferation and upregulation of the expression of CD25 and CD69 in the TCRVβ 5 + CD8 + T cells after co-culture with the PLGA/ OVA NPs-treated Gr-1 high or Gr-1 low subsets compared to the control groups treated with PBS, OVA, or PLGA NPs without OVA, which did not induce significant CD25/CD69 expression or cell proliferation ( Fig. 2b-f). Statistical analysis revealed a significant induction of cell proliferation and more pronounced activation phenotypes in OT-I CD8 + T cells when co-cultured with the PLGA/OVA NPs-primed Gr-1 low subset compared to the Gr-1 high group (Fig. 2d-f). These data demonstrated the antigen presentation capabilities of the Gr-1 + cells, including both the Gr-1 high or Gr-1 low subsets, after acquiring antigens from the PLGA/OVA NPs.   Figures (f,g) illustrate the microscopic morphology of the CD11b + Gr-1 high (f) and CD11b + Gr-1 low (g) subsets after staining with the Wright-Giemsa stain. Scale bar = 10 μ m.
the Gr-1 high or Gr-1 low subsets were placed in 96-well plates at 1 × 10 6 cells/ml in complete RPMI 1640 medium containing 5 ng/ml GM-CSF, followed by treatment with PBS, 100 μ g/ml PLGA NPs, or PLGA/OVA NPs for 0, 8, 16, or 24 hrs. The cells were then harvested, washed with PBS, and analysed by flow cytometry. The data in Fig. 3 show the induction of the expression of CD115, a receptor for macrophage colony stimulating factor (M-CSF), in the Gr-1 high population. Treatment of Gr-1 high cells with the PLGA/OVA NPs reduced CD115 expression compared to the PBS control (Fig. 3a).
Approximately 25-35% of Gr-1 low cells expressed high levels of CD11b but were negative in CD11c and F4/80, the dendritic cells and macrophage markers, started expressing both markers after 16 hrs incubation in the culture medium containing GM-CSF (Fig. 3b). Approximately 34% of cells in the F4/80 + CD11c + double-positive (DP) subset also express MHC II, and 13% of cells in the F4/80 − CD11c − double-negative (DN) subset express CD115 (Fig. 3c). Plasticity has long been recognized as a hallmark in the differentiation pathway of myelomonocytic cells. These data demonstrate that less than 35% of Gr-1 low cells, but not the Gr-1 high subset, differentiated into F4/80 + CD11c + mononuclear phagocytes 18,27 , exhibiting characteristics of macrophages and dendritic cells, and that in vitro cell transformation induced by the polymeric NPs was relatively insignificant.
Microscopic examination of Gr-1 low cells illustrated that in comparison to the PBS treated control cells, treatment with PLGA/OVA NPs for 8 hrs induced the formation of spindle-shaped cells that are attached to the bottom of the plates, whereas the PLGA/OVA NPs are still present (Fig. 3d). These nanoparticles began to be cleared after incubation for 16 hrs when Gr-1 low cells start expressing CD11c and F4/80 ( Fig. 3b and d), and were extensively ingested when the spindle-shaped cells became more abundant, compared to the PBS control, at 24 hrs post-treatment (Fig. 3d).

Treatment of Gr-1 + cells with PLGA/OVA NPs enhanced intracellular cytokine expression.
To assess the cytokine secretion induced by the treatment of PLGA/OVA NPs, the Gr-1 high or Gr-1 low cells were pre-treated for 16 hrs with PLGA NPs, with or without the OVA antigen, followed by co-culture for 72 hrs with CD8 + T cells isolated and negatively enriched from the spleens of OT-I mice. The cells were then stained with fluorochrome-conjugated mAbs against CD8α , TCRVβ 5.1, Gr-1, Ly-6C, and CD11b. Co-culture of OT-I CD8 + cells with Gr-1 + cells pre-treated with PLGA/OVA NPs induced the expression of Th1-like cytokines, such as IL-2, TNF-α , and IFN-γ , at significantly higher levels than the control groups treated with PBS, OVA, or PLGA NPs alone (Fig. 4a-c). The percentages of cytokine-secreting cells were analysed and compared in the gated CD8 + TCRVβ 5 + T cells. Selected data are presented as delta mean fluorescence intensity (Δ MFI); i.e., the MFI of cytokine expression in the CD8 + TCRVβ 5 + T cells in each group minus the corresponding cytokine levels of CD8 + TCRVβ 5 + T cells in the matched isotypes. The data in the bar graphs ( Fig. 4d-e) depicting the percentages and the Δ MFI of cytokine-expressing cells demonstrate that treatment of CD8 + TCRVβ 5 + T cells with the Gr-1 + subsets pre-treated with the PLGA/OVA NPs significantly upregulated the expression of Th1-like cytokines in the OT-I CD8 + T cells.

Detection of perforin and granzyme B by flow cytometry and fluorescence microscopy.
Acquisition of the effector properties, such as upregulation of perforin and granzymes, is critical in the generation of cytotoxic T lymphocytes (CTLs). To detect the expression of granzyme B and perforin in the secretary granules of OT-I CD8 + cytotoxic T lymphocytes (CTLs) after co-culture with the PLGA/OVA NPs primed Gr-1 + populations, cells were stained with mAbs against granzyme B and perforin and analysed with a BD FACSVerse cytometer. The expression of these cytolytic effector molecules was significantly upregulated in the PLGA/OVA NPs-treated group compared to the control groups treated with PBS, OVA, or PLGA NPs alone (Fig. 5a-c). This finding is substantiated with further examination by fluorescence microscopy, confirming that treatment of Gr-1 + cells, either the Gr-1 high or Gr-1 low subset, with the PLGA/OVA NPs induced significant upregulation of perforin and granzyme B in the OT-I CD8 + T cells, and that the expression was relatively higher when co-cultured with the Gr-1 low cells than those with the Gr-1 high subset (Fig. 5d-e).

Treatment of Gr-1 + cells with PLGA/OVA NPs induced the cytolytic activity of OT-I CD8 + T cells.
To determine the effect of PLGA/OVA NPs on cytolytic activity, the enriched OT-I CD8 + T cells were co-cultured  for 72 hrs with the Gr-1 high or Gr-1 low subsets that were pre-treated for 16 hrs with PBS, OVA, PLGA NPs, or PLGA/OVA NPs. The cells were then co-cultured for 4 hrs with EL4 cells pulsed with or without 1 μ M SIINFEKL peptide followed by cytolytic assay using the CytoTox96 non-radioactive cytotoxicity assay kit (Promega). The results in Fig. 5f demonstrate that co-culture of OT-I CD8 + cells with PLGA/OVA NPs-primed Gr-1 + cells, either the Gr-1 high or Gr-1 low subset, stimulated antigen-specific cytotoxic activity in the SIINFEKL-pulsed target EL4 cells, in contrast to the untreated EL4 control groups (Fig. 5f).
Enhanced production of OVA-specific IgG in the blood of animals after immunization with PLGA/OVA NPs. In an attempt to evaluate the impact of PLGA/OVA NPs on the induction of the humoral immune response in vivo, C57BL/6J mice were i.v. injected with PBS, OVA, PLGA NPs, or PLGA/OVA NPs on days 0, 7, 14, 21, and 28, followed by blood collection on days 3, 10, 17, 24, and 31. The serum antibody titres were quantified by enzyme-linked immunosorbent assay (ELISA) with peroxidase-conjugated rat anti-mouse IgG1, IgG2a, or IgG2b and the plates were read with a spectrophotometer. The OVA-specific antibody response in each immunization group is shown in Fig. 6a-c. The OVA-specific IgG levels in the control groups treated with PBS and PLGA NPs were not detectable (data not shown), whereas mice immunized with PLGA/OVA generated significantly higher titres of IgG isotypes, including IgG1 and IgG2a, than the group immunized with OVA alone (Fig. 6a-c).

In vivo proliferation of antigen-specific OT-I CD8 + T cells induced by the PLGA/OVA NPs.
To monitor the antigen-specific T cell proliferation, the CD8 + T cells were isolated and enriched from spleens of OT-I mice, labelled with CFSE and i.v. injected into the tail veins of irradiated C57BL/6J recipient mice. Twenty-four hrs later, the mice were i.v. injected with PBS, OVA, PLGA NPs, or PLGA/OVA NPs and euthanized at 4 days post-treatment. Cells from the spleen and lymph nodes were analysed by flow cytometry. The proliferation profiles of the OT-I CD8 + T cells, gated on the 7-AAD¯CD3 + CD8 + TCRVβ 5 + population, are shown in  Fig. 6d, demonstrating a significant proliferation of the OT-I CD8 + T cells in the recipient mice immunized with PLGA/OVA NPs, compared to the control groups treated with PBS, OVA, or PLGA NPs alone (Fig. 6d).

Discussion
In view of the emergence of nanotechnology and the potential application of polymeric nanoparticles in cancer vaccines, we investigated in this study the combined potential of polymeric nanoparticles and the phagocytic capability of bone marrow Gr-1 + granulocytes, including the Gr-1 high and Gr-1 low subsets, for cross presentation of exogenous antigens to boost the MHC class I immune response. Treatment of BM Gr-1 + cells, either the Gr-1 high or Gr-1 low subsets, with antigen-encapsulated polymeric nanoparticles (NPs) significantly stimulated the in vitro proliferation of antigen-specific CD8 + T cells and enhanced the secretion of Th1 cytokines, such as IL-2, TNF-α , and IFN-γ , compared to the control groups treated with free antigen or the empty PLGA NPs alone (Fig. 4). It is noteworthy that these PLGA/OVA NPs do not stimulate the maturation of the costimulatory molecules in the bone marrow-derived dendritic cells (BMDCs) (data not shown). In contrast to the Gr-1 low group, Gr-1 high cells do not simulate significant expression of these cytokines in OVA-treated OT-I CD8 + cells (Fig. 4). Cytometric analysis of Gr-1 high cells with a monoclonal antibody 25-D1.16 (eBioscience) specific for the OVA (257-264) peptide SIINFEKL bound to H-2K b of MHC class I showed the capability of cross-presentation of the H-2K b bound OVA-peptide in the context of MHC I when treated with the PLGA/OVA NPs compared to the control groups 18 . It is established that monocytes can differentiate into DCs when receiving a phagocytic stimulus 27 . Upon in vitro culture in the presence of GM-CSF, the expression of CD115 in CD11b + Gr-1 low cells was upregulated, and approximately 30-35% of CD11b + Gr-1 low cells became CD11c + F4/80 + double-positive (DP), whereas approximately over 65% of the population remained CD11c − F4/80 − double-negative (DN) (Fig. 3b). In the double-positive (DP) population, 90% of cells express CD115 (CSFR), indicating the trans-differentiation potential of these cells into dendritic cells or macrophage-like monocytic phagocytes in the presence of GM-CSF, regardless of the treatment with the PLGA/OVA NPs. The CD11c + F4/80 + double-positive (DP) population can be further subdivided into MHC II + and MHC II¯ granulocytic fractions. The CD11c¯F4/80¯ double-negative (DN) population, however, is both CSFR¯ and MHC II¯ (Fig. 3c). Phagocytosis of the PLGA NPs by these two subsets, DP and DN, did not induce significant phenotypic changes in the cells. Gr-1 low cells exhibited superior capabilities of cross-presentation compared to the Gr-1 high subset (Fig. 2b-f), presumably caused by the trans-differentiation of 25-35% of Gr-1 low population into CD11c + dendritic cells, the most potent antigen presenting cells, in the culture medium containing GM-CSF.
The expression of CD11b and Gr-1 has been associated with immunosuppression in cancer or infection in the form of myeloid-derived suppressor cells (MDSCs). In contrast to the previous report by Peranzoni et al. 21 , employing the anti-CD11b and anti-Gr-1 mAb (RB6-8C5) mAbs to distinguish the two murine myeloid subsets of myeloid derived suppressor cells (MDSCs) with immunosuppressive properties 21 , including the CD11b + Gr-1 high Ly6G + Ly6C low/int (polymorphonuclear) and CD11b + Gr-1 int Ly6G − Ly6C high (mononuclear) cells, the results obtained in this study demonstrated that treatment of these cells with the antigen-containing polymeric NPs resulted in immunostimulation rather than immunosuppression. The experimental evidence from this study also demonstrated the potential capability of the CD11b + Gr-1 + cells, including the Gr-1 high and the Gr-1 low subsets, to present exogenous antigens in the context of MHC class I to the CD8 + T cells when primed with the PLGA/OVA NPs, leading to the upregulation of type 1 cytokines (Fig. 4) and the acquisition of cytolytic molecules such as perforin and granzyme B (Fig. 5), a signature of CTL activation and effector function, to eliminate the pathogen-infected or oncoprotein-specific transformed cells, one of the most desired effect in the design of vaccine-based cancer therapy. To the best of our knowledge, results reported herein showed the first experimental evidence demonstrating the acquisition of antigen-specific cytolytic T effector functions via the polymeric NPs-primed Gr-1 + subsets. However, the data in Fig. 5 reveals that the Gr-1 low subset express higher levels of perforin and granzyme B and hence exhibit higher antigen presentation capability than the Gr-1 high subset, presumably due to the trans-differentiation of Gr-1 low cells into CD11c + dendritic cells in the presence of GM-CSF. This speculation, on the other hand, was not explicitly justified in the results shown in Fig. 5f, where both the Gr-1 high and Gr-1 low subsets exhibit similar antigen-presentation capabilities on the stimulation of the CTL effect.
Our recent studies on cellular biodistribution of polymeric nanoparticles have illustrated the kinetics and localization of polymeric nanoparticles in the immune system after i.v. injection 28 . The i.v. administration route prompts the uptake of nanoparticles by the APCs in the circulation faster than any other routes, resulting in accumulation of nanoparticles in the lymphoid organs such as spleen and bone marrow 29 . The nanoparticles can then be taken up by the phagocytic APCs, such as dendritic cells and macrophages, followed by antigen cross-presentation and the ensuing antigen-specific immune response. The in vivo expansion of adoptively transferred CFSE-labelled OT-I CD8 + T cells was confirmed in this study in the recipient mice immunized with the PLGA-NPs (Fig. 6d), suggesting that the PLGA/OVA NPs are able to stimulate the antigen-specific CD8 + T cell proliferation in the PLGA/OVA NPs immunized animals. The function of polymeric nanoparticle-mediated antigen delivery in the cross-presentation of exogenous antigens in vivo is thus confirmed.

Conclusions
We have demonstrated in this study the efficient recruitment of Gr-1 + subpopulations, including both the Gr-1 high and Gr-1 low subsets from murine bone marrow, by PLGA/OVA nanoparticles (NPs) for cross presentation of exogenous antigens in the context of MHC class I complexes, resulting in antigen-specific OT-I CD8 + T cell proliferation and upregulation of the expression of Th1-type cytokines, including IL-2, TNF-α , IFN-γ , and effector molecules such as perforin and granzymes B, leading to the effector T cell response and antigen-specific cytotoxic T-lymphocyte (CTL) effects. The experimental results obtained in this study on nanoparticulate antigens have a crucial implication for the future design of vaccine delivery.

Materials and Methods
Materials. PURASORB PDLG (DL-lactide and glycolide) 5004A (50/50 DL-lactide/glycolide copolymer, MW44000), a biodegradable poly(lactic-co -glycolic acid) (PLGA) copolymer, was kindly provided by Purac Biomaterials (Gorinchem, The Netherlands). Polyvinyl alcohol (MW15,000) was obtained from Fluka and the OVA peptide 257-264 (SIINFEKL) was obtained from PolyPeptide Group (Strasbourg, France). Fetal bovine serum (FBS) and RPMI 1640 cell culture medium were obtained from HyClone Laboratory and Gibco Inc. respectively. 5-(and-6)-carboxyfluorescein diacetate succinimidyl ester (CFSE) was obtained from Molecular Probes. 4′ ,6-Diamidino-2-phenylindole dihydrochloride (DAPI) was purchased from Sigma-Aldrich (St. Louis, MO). The CytoTox 96 Non-Radioactive Cytotoxicity Assay was purchased from Promega (Madison, WI). Preparation of PLGA and PLGA//OVA nanoparticles. PLGA and PLGA/OVA NPs, the PLGA NPs containing the OVA antigen, were prepared by the water-in-oil-in-water (w/o/w) double emulsion and solvent evaporation method, as previously described 18 . A small amount of lyophilized NPs was placed on a double-sided tape attached on a sample stand and coated with a thin layer of gold using a Hitachi E-1010 ion sputter, and the surface morphology of the PLGA/OVA nanoparticles was examined by scanning electron microscopy (SEM) with a Hitachi S-4800 scanning electron microscope.

Purification of bone marrow granulocytes by fluorescence-activated cell sorting (FACS).
Bone marrow cells were prepared from the femurs and tibias of C57BL/6J mice. Briefly, the mice were anesthetized with isoflurane and euthanized. Bone marrow cells were harvested by flushing cells from the tibias and femurs of adult C57BL/6J mice, followed by lysis of red blood cells with ACK (ammonium-chloride-potassium) lysing buffer containing 154.95 mM ammonium chloride, 9.99 mM potassium carbonate, and 0.0995 mM EDTA. For phenotyping, sorted cells were stained with 7-AAD and fluorochrome-conjugated rat anti-mouse mAbs against Lineage 2 (Lin2), CD11b, Gr-1, and Ly-6C, where Lin2 includes the antibodies against CD3/CD19/NK1.1/B220/ Ter119 26 . The Lin2¯CD11b + Gr-1 high and Lin2¯CD11b + Gr-1 low subsets were sorted with a BD FACSAria III high-speed cell sorter (BD Biosciences) by the Cell Sorting Core Facility of our university. The purity of the final DC subpopulations was routinely greater than 98.5%.
Phenotypic analysis by flow cytometry and examination of cell morphology by Wright-Giemsa staining and microscopy. Gr-1 high and Gr-1 low subsets, isolated from bone marrow of mice by FACS, were stained with fluorochrome-conjugated mAbs against using the antibodies against Lin2, CD11b, Gr-1, and Ly-6, and analysed by flow cytometry with a LSRFortessa flow cytometer (BD Biosciences).
To examine the morphology of these cells, the two subpopulations were suspended in the cell culture medium at 1 × 10 5 in 200 μ l PBS, and cytospun at 1,000 rpm for 3 min onto glass slides with a Cytospin 4 cytocentrifuge (Thermo Scientific). The cells were fixed with methanol for 10 min at room temperature, washed with PBS, and stained with Wright-Giemsa stain (Sigma-Aldrich) for 1 min at room temperature 31 . The slides were then washed with double distilled water and dried, followed by mounting with DPX mountant (Sigma). The cell morphology was examined with a Zeiss Axio Imager A1 microscope and photographed.

Fluorescence microscopy.
To examine the expression of perforin and granzyme B by fluorescence microscopy, the Gr-1 + subsets were plated in complete RPMI 1640 medium containing GM-CSF (5 ng/ml), and treated with PBS or 100 μ g/ml PLGA/OVA NPs for 16 hrs, followed by co-culture with the OT-I TCRVα 2 + CD8 + T cells for 48 hrs. The cells were then treated with 10 μ g/ml brefeldin A (BFA) for 24 hrs and cytospun at 1,000 rpm for 3 min onto the silane-coated micro glass slides with a Shandon Cytospin4 (Thermo Scientific), followed by fixation with methanol at − 20 °C for 5 min and washes with PBS. The non-specific binding was blocked for 5 min with 10% goat serum/PBS. The cells were then incubated at 4 °C for 16 hrs with rat purified anti-mouse perforin (eBioOMAK-D) (1:50) or anti-mouse granzyme B (16G6) (1:100), washed with PBS again, and incubated for 1 hr with goat anti-rat biotin (1:500). After several washes with PBS, the cells were stained for 1 hr with tetramethyl-rhodamineisothiocyanate-conjugated extravidin (extravidin-TRITC) (1:100) and CD8-FITC, washed with PBS, and stained for 5 min with DAPI (1 μ g/ml). The cells were washed with PBS again, and the slides were mounted with a fluorescent mounting medium (DakoCytomation) and observed with a Zeiss Axio Imager A1 fluorescence microscope.
In vitro antigen-specific cytotoxicity assays. Cytotoxicity assays were performed using the EL4 target cells, with or without pre-coating with 1 μ M SIINFEKL peptide for 1 hr [35][36][37] . Briefly, the Gr-1 high (Lin2¯CD11b + Gr-1 high ) and Gr-1 low (Lin2¯CD11b + Gr-1 low ) subsets were isolated from the bone marrow of naïve mice and sorted by FACS with a BD FACSAria III. The cells were treated with or without the PLGA/OVA NPs for 16 hrs. The cells were then washed with PBS and co-cultured for 72 hrs in complete RPMI 1640 medium containing 10 U/ml IL-2 at 1:1 ratio with the CD8 + TCRVβ 5 + T cells isolated and enriched from the spleen of OT-I mice. The effector T cells were then assayed at various effector: target cell (E/T) ratios for cytotoxic activity on the target cells. Briefly, the effectors were co-cultured for 4 hrs at 37 °C at various effector: target cell ratios (E/T = 0.75:1, 1.5:1, 3:1, 6:1, and 12:1) with EL4 cells, with or without being pulsed with 1 μ M SIINFEKL in the 96-well plates. The cells were used for cytolytic analysis of lactic dehydrogenase (LDH) with the CytoTox96 non-radioactive cytotoxicity assay kit (Promega) according to the manufacturer's instruction. After the reaction was stopped, the microplates were centrifuged and fifty μ l of the supernatant was collected for the cytotoxicity assay. The amount of LDH release was determined by taking the absorbance at 490 nm, and background absorbance values of phenol-red free culture medium was subtracted from all absorbance values. The percent specific lysis was calculated from triplicate samples as follows: where Effector spontaneous and Target spontaneous are the spontaneous LDH release by the control OT-I CD8 + effector T cells and control target cells (EL4 or SIINFEKL-loaded EL4), respectively. Target maximum, on the other hand, was obtained by treatment of EL4 cells, without T cells, for 45 min with 10 μ l lysis buffer [9% (v/v) Triton X-100].