Ex vivo development, expansion and in vivoanalysis of a novel lineage of dendritic cells from hematopoietic stem cells

Cells and mice

CD34⁺ cells used in this study were purified from BM, mobilized peripheral blood (MPB) or cord blood (CB) using magnetic beads (Miltenyi Biotec) following the manufacturer's instructions or purchased from AllCell Inc. (San Mateo, CA), Cambrex (Baltimore, MD) and National Disease Research Interchange (Philadelphia, PA). Buffy coats of peripheral blood of healthy donors were purchased from Civitan Blood Center (Gainesville, FL). PBMCs were isolated from buffy coats of healthy donors or from blood of cancer patients with approval of the Institutional Review Board of University of Florida. B lymphoblastoid cell lines (BLCLs) were generated by transforming peripheral blood B lymphocytes with EBV as described previously [31]. The BLCLs were propagated in complete RPMI-1640 medium (Gibco, Grand Island, NY) supplemented with 2 mM L-glutamine, 100 ug/ml streptomycin, 100 IU/ml penicillin and 10% heat inactivated fetal bovine serum (FBS) at 37°C with 5% CO₂. The mouse fetal stromal cells were cultured in Minimal Essential Medium Alpha (Gibco, Grand Island, NY) supplemented with 2 mM L-glutamine, 100 ug/ml streptomycin, 100 IU/ml penicillin and 20% heat inactivated FBS (Gibco) at 37°C with 5% CO₂. CT26 mouse colon carcinoma cell line was purchased from ATCC (catalogue no. CRL-2638) and cultured in Dulbecco Modified Eagle's Medium (Gibco, Grand Island, NY) supplemented with 2 mM L-glutamine, 100 ug/ml streptomycin, 100 IU/ml penicillin and 10% heat inactivated FBS (Gibco) at 37°C with 5% CO₂. BALB/c mice were obtained from Jackson Laboratory (Bar Harbor, ME) with approval from the Institutional Animal Care and Use Committee of University of Florida.

Antibodies and reagents

Fluorescein isothiocyanate (FITC)-conjugated Abs to CD4, CD8, CD11c, CD33, CD34, CD38, CD86, IFN-γ, and HLA-DR, phycoerythrin (PE)-conjugated Abs to CD4, CD8, CD11c, CD34, CD83, CD90, CD123, HLA-DR, IFN-γ, and TNFα, PerCP Cy5.5-conjugated Ab to CD8, CD33, PE-Cy7-conjugated antibodies to CD8, CD11b, CD11c, CD34, CD62L, CD40 and CD80, and allophycocyanin (APC)-conjugated Abs to CD1a, CD3, CD11c, CD14, CD33, CD69, CD83, CD90, CD133, IFN-γ and TNF-α were purchased from BD Pharmingen (San Diego, CA), eBioscience (San Diego, CA), Miltenyi Biotec (Auburn, CA), and Caltag Laboratories (Invitrogen, Carlsbad, CA) as listed in Additional file 1, Table S1. Isotype-matched antibodies were included as controls. HLA-A2 restricted, EBV BMLF1 GLC-peptide (amino acid 280-288, GLCTLVAML) pentamer was purchased from Proimmune (Springfield, VA).

Lentivector preparation and gene transfer

Lentivectors (LVs) were constructed as described previously [32–34]. The growth factor cDNAs were amplified by PCR using primers designed to contain an optimized translation initiation sequence (-CCACC-5' to the initiation codon). The primers used in this study are listed in Additional file 2, Table S2. The amplified cDNAs were cloned into the self-inactivating pTYF plasmid behind the EF1α promoter. To generate feeder cells, mouse fetal stromal cells were multiply transduced with LVs at 10-50 infectious unit/cell in 12-well plates in a minimal volume of 0.3 ml per well. After 2 h, 0.5 ml of fresh media was added and cells were incubated at 37°C overnight. The infected cells were continuously propagated for more than 50 passages and stable lentiviral transgene expression was confirmed. The mouse CT26 tumor cells and DCs were transduced with LVs encoding a codon-optimized human HPV E6-E7 fusion protein and the chaperone protein calnexin as previously described [35].

RNA extraction, RT-PCR and microarray analysis

RNA was extracted using Tri-reagent (MRC Inc., Cincinnati, OH) and oligo(dT)₁₅-primed cDNA was made with MMLV reverse transcriptase (Promega Inc., Madison, WI). For semi-quantitative PCR, all reactions used the same serially diluted cDNA normalized to the mouse GAPDH (mGAPDH). The PCR amplification conditions were as follows: denaturing temperature, 95°C; annealing temperature, 55-62°C; extension temperature, 72°C; the amplification cycles were 25-35 cycles. Products were resolved by agarose gel electrophoresis and visualized by ethidium bromide staining. The PCR primers used in this study are listed in Additional file 2, Table S2.

For gene expression microarray analysis, RNA samples were harvested from purified CD34⁺HPCs, ex vivo cultured DCPs, and adherent PBMC-derived IL-4 and IL-15 DCs. These RNA samples were analyzed using Illumina Human RefSeq-8 Expression BeadChips. RNA quantity was determined with the Agilent RNA 6000 Nano Kit and Bioanalyzer. All samples displayed 28 S and 5.8 S peaks indicating intact full length RNA. Synthesis of double-stranded cDNA and in vitro transcription were performed with the Ambion Illumina TotalPrep kit according to manufacturers' instructions. For each sample, input quantity for the first strand synthesis was normalized to 200 ng. After in vitro transcription reaction, yield of purified cRNA was assessed with the RiboGreen assay and quality was assessed with the Agilent Bioanalyzer. BeadChip hybridization, staining and scanning were performed according to Illumina whole genome expression for BeadStation. For each sample, input of cRNA was normalized to 1500 ng. As control, Stratagene Universal Human Reference (SUHR) RNA was labeled with the Ambion TotalPrep kit. The labeled cRNA was used as interchip hybridization replicates and showed strong correlation. Biological replicate pairs were analyzed and for unnormalized data, the linear r² was greater than 0.94 for all replicates.

Generation of mature DCs and antigen-specific immune cells

PBMCs were isolated after Ficoll-Hypaque density centrifugation (Sigma Aldrich, St Louis, MO). After plastic adherence, the adherent cells were cultured in 50 ng/ml GM-CSF and 25 ng/ml IL-4 (eBiosource International, Inc. Camarillo, CA) in serum-free AIM-V medium (Invitrogen, San Diego, CA) to generate immature DCs. Immature DCs were transduced with LVs, and treated with LPS (1 ug/ml) and TNFα (20 u/ml) for 24 hr to induce maturation. The mature DCs were loaded with 5 ug/ml of specific peptides. The non-adherent PBMCs were cocultured with irradiated (10 Gy) mature DCs, at a 20:1 ratio, in AIM-V supplemented with IL-2 (12.5 U/ml) and IL-7 (10 ng/ml) in 24-well plates. At day 12 of coculture, the T cells were restimulated or harvested for analysis as previously described [36, 37]. The DCs of BALB/c mice were generated from BM of tumor-bearing mice and transduced with LVs as previously described [38].

Quantitative cytokine and chemokine multiplexed enzyme-linked immunosorbent assay (ELISA) arrays

DCs were washed twice with PBS and cultured in AIM-V without growth factors and other supplements at a density of 10⁶ cells/ml for 24 hr. The supernatants were collected and delivered to Quansys Biosciences (Logan, UT) for custom multiplexed sample testing as previously described [35]. Each sample was tested in triplicate. The list of cytokines and chemokines tested included: IL-1a, IL-1b, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-13, IL-17, IL-23, IFN-γ, TNF-α, TNF-β, Eotoxin, growth-related oncogene-alpha (GRO-α), monocyte chemotactic protein 1 (MCP-1), MCP-2, regulated upon activation, normal T-cell expressed and secreted cytokine (RANTES), I-309 and thymus and activation-regulated chemokine (TARC).

Antibody and pentamer staining and flow cytometry

For antibody (Ab) staining, single-cells were suspended in PBS containing 2% FBS and 0.05% sodium azide and pre-incubated with anti-CD16/CD32 Ab for 10 min to block FcRs. Expression of cell surface markers was analyzed by standard four-color staining using FITC-, PE-, PE-Cy7 or PerCP and APC-conjugated primary Abs. To evaluate the expression of intracellular molecules, cells were washed and restimulated for 5 hr in the presence of brefeldin A (1 ug/ml) during the last 2.5 hr of culture. The stimulated cells were stained with anti-surface marker Abs, washed and permeabilized with the CytoFix-Cytoperm kit (BD Pharmingen), according to the manufacturer's instruction, then stained with anti-intracellular marker Abs, and analyzed by flow cytometry. For multimer staining, the resting T cells were stained with PE-labelled pentamer (Proimmune) for 12 min at room temperature, followed by FITC-labeled or PE Cy7-labelled anti-CD8 antibody for 30 min on ice and analyzed by flow cytometry. Data acquisition and analysis were done on a FACSCalibur and FACSAria using CellQuest and FACSDiva software, respectively (BD Biosciences, San Jose, CA), or Flowjo software (Tree Star, Inc. Ashland, OR).

Analysis of antigen uptake

Immature DCs were harvested and washed with AIM-V twice and re-suspended in AIM-V at a concentration of 5x10⁵ cells per ml. DCs were incubated with Dextran-FITC or OVA-FITC (Molecular Probes, Inc., Eugene, OR) at 37°C for 1 h; a parallel control was incubated at 4°C for 1 h. Cells were washed three times with cold FACS buffer, resuspended in 100 ul of cold FACS buffer, stained with APC-conjugated anti-CD11c Ab (BD Biosciences, CA) and analyzed by flow cytometry.

In vivoDC vaccine tumor model

BALB/c CT26 colon cancer cells were transduced with LV-optE6E7 encoding a fusion protein of HPV16 E6/E7 to generate the CT26-E6E7 cell line. The BALB/c mice were inoculated with 1x10⁵ CT26-E6E7 tumor cells subcutaneously. Seven days later the mice were vaccinated with 2-5 x10⁵ immature DC/LV-optE6E7 or DC/LV-optE6E7/LV-calnexin derived from tumor-bearing mice, weekly for 2 weeks (n = 5 per group). Tumor size was measured over time using calipers and mean tumor volume (in mm³) was determined.

Statistical analysis

The statistical analysis was performed using Student's t-test and GRAPHPAD PRISM 4 software.

Expansion of CD34⁺HPCs and development of DC progenitors (DCPs)

HPCs can expand in culture but have limited potential of maintaining hematopoietic stem cell phenotype and function [39]. We established a series of lentivector-modified stromal cell (LSC) lines to provide cell-free and cell-associated signals that can support continuous expansion of HPCs and differentiation of DCs (Fig. 1). The LSC lines include LSC-KFT (KL, Flt3L, TPO), LSC-KFTb (KL, Flt3L, TPO and bFGF), LSC-KFT63 (KL, Flt3L, TPO, IL-6, and IL-3) and LSC-KFT63b (KT, Flt3L, TPO, bFGF, IL-6, IL-3 and bFGF) for the expansion of HPCs, and LSC-KFT-GM15 (KL, Flt3L, TPO, GM-CSF and IL-15) and LSC-KFT-mGM15 (KL, Flt3L, TPO, moues GM-CSF, and IL-15) for the differentiation and expansion of human and mouse DCs, respectively. LSC-KFT, LSC-KFTb, LSC-KFT63 and LSC-KFT63b supported HPC expansion to similar extents, and the total expansion fold varied with individual donors. Under this culture condition, HPCs consistently expanded twenty to one hundred-fold in twenty days, followed by more than one thousand-fold expansion and differentiation into DCPs in thirty days (Fig. 1A). This dual culture system supports expansion and development of DCs from both human and mouse HPCs.

To verify expression of the various growth factors in LSCs, RNAs harvested from LSCs were analyzed by semi-quantitative RT-PCR (Figure 1B and 1C). We confirmed that lentivector expression was stable even after 50 passages in these cell lines (data not shown). Highly enriched human CD34⁺ HPCs derived from adult mobilized peripheral blood and BM expressed high level of hematopoietic progenitor marker CD133 and low level of CD33 (Figure 2A). This culture system supported HPC expansion for both healthy donors and cancer patients; for example, adult peripheral blood (PB) HPCs expanded in LSC-KFT63b to about one hundred-fold in two to three weeks (Figure 2A). Surface phenotype analysis indicated that the ex vivo expanded HPCs gradually lost progenitor markers (CD34, CD90, and CD133), which was accompanied by increased expression of myeloid differentiation markers CD38 and CD33 (Figure 2B). Similar results have been obtained with mouse BM Sca1⁺Lin^- (lineage-minus) HPCs (data not shown).

Differentiation and expansion of DCPs toward meyloid DC-like phenotype

To see if the LSC culture system can generate functional DCs, we first expanded CD34⁺ HPCs in LSC-KFT63b. After an initial 20-40-fold expansion, the cells were transferred to LSC-KFT-GM15. The DCPs continued to expand several orders of magnitude in 30 days; they were then transferred to feeder-free culture supplemented with GM-CSF and IL-15 to generate functional immature DCs as illustrated in Figure 3A.

Analysis of myeloid and DC lineage differentiation markers including PU.1, Langerin, Id2, hIL7R-a, CCL17, hCCR6, and E-cadherin (E-CAD) of the DCPs from day 0, 4, 9, 13, 23 and 39 by semi-quantitative RT-PCR revealed a gradual increase in myeloid (PU.1) and DC differentiation markers (Id2, hIL7R-a, CCL17, and hCCR6), and a stochastic expression of differentiating Langerhans cell markers (Langerin or CD207 and E-CAD) as compared with monocyte-derived immature DCs (imDC, Figure 3B). After 35 days, the DCPs displayed a differentiation profile similar to that of monocyte-derived imDCs, except for E-CAD, which was down-regulated. Kinetic analysis of monocyte and DC markers including CD14, CD11c, CD1a, CD11b, HLA-DR, CD83, CD40, CD123, and CD86 by flow cytometry showed that the ex vivo-expanded DCPs gradually differentiated toward mature DCs with increased expression kinetics of costimulatory molecules CD40, CD86, and DC maturation marker CD83 (Figure 3C). A representative flow cytometry analysis of DC markers of day 39 DCPs is shown in Figure 3D; at this time point, DCPs displayed increased activation and maturation markers resembling conventional mature DCs. Similar results were observed with mouse DCPs expanded in the LSC-KFT-mGM15 culture (not shown).

We next examined the gene expression profile of the DCPs at different time points after differentiation from HPCs using Illumina BeadChip Human RefSeq-8 arrays. RNA samples were harvested from an early time point (day 4) and a late time point (day 23), and compared to RNAs harvested from CD34⁺ HPCs and adherent PBMC-derived IL-4 and IL-15 DCs. Cluster analysis of unnormalized sample data showed that all biological replicates (two DCPs, two IL-4 DCs and three CD34⁺ HPCs) sorted into the same groups (Figure 4). Sample dendogram revealed that the day 4 differentiated DCPs displayed gene expression profile resembled CD34⁺HPCs, whereas the day 23 differentiated DCPs displayed expression profile resembled IL-4 DCs (Figure 4).

Cytokine and chemokine secretion profiles of the ex vivogenerated DCs

Antigen capture and T cell activation functions of the ex vivogenerated DCs

Professional antigen presenting cells can uptake and process antigens and stimulate T cells. To examine these functions, we compared the ex vivo generated DCs with immature adherent PBMC-derived DCs (PBMC-DCs) for their antigen uptake function by feeding them with fluorescent DQ-OVA (OVA-FITC) and dextran-FITC particles. The PBMC-DCs captured fluorescent particles at 37°C but not at 4°C, as demonstrated with flow cytometry. The day 37 DCP-DCs, which contained a large number of CD11c-positive cells, captured antigens as efficiently as did the PBMC-DCs (Figure 5A).

To see if the DCP-derived DCs were capable of activating antigen-specific T cells, we set up a DC/T cell coculture system as previously described [31, 35, 37]. The DCs were transduced with LVs encoding a viral antigen, EBV BMLF (LV-BMLF), or a truncated self-antigen tNGFR (LV-tNGFR). After maturation, the DCs were cocultured with autologous monocyte-depleted PBMCs for 12 days to generate antigen-specific T cells. The activated T cells were restimulated with the corresponding DCs and incubated with a BMLF peptide-specific (GLCTLVAML, HLA-A2*-restricted) MHC pentamer to detect antigen-specific response. Both the DCP-DCs and the PBMC-DCs induced antigen-specific T cell response when transduced with LV-BMLF, but not LV-tNGFR (Figure 5B, left panel). Intracellular staining for IFN-γ expression in CD4 and CD8 T cells confirmed that the DCP-DCs activated BMLF-specific T cells as effectively as did the PBMC-DCs (Figure 5B, right panel). In contrast, the self-antigen tNGFR did not register a substantial response.

In vivotumor suppression mediated by DCP-DCs derived from tumor-bearing mice

The above assays demonstrated that the HPC-derived DCs displayed antigen presentation and T cell activation functions similar to that of monocyte-derived DCs. To examine their therapeutic potential, we designed an experiment using a previously established syngeneic mouse tumor model. BALB/c mice implanted with CT26 colon cancer cells expressing a codon-optimized Human Papilloma Virus 16 (HPV 16) E6 and E7 fusion protein (optE6E7) are protected through immunization with DCs transduced with LVs encoding optE6E7 and calnexin, a chaperone protein [31, 35]. To mimic situation in a cancer patient, we derived DCPs from BM of tumor-bearing mice to see if a protective anti-cancer immune response can be induced. Figure 6A illustrates the strategy of generating DCPs from BM of tumor-bearing mice. CT26/optE6E7 tumors were first established in BALB/c mice, after confirming tumor growth, Sca1⁺Lin^- HPCs were purified from BM of the tumor-bearing mice. The HPCs were expanded in the LSC culture system as described; a representative mouse DCP expansion curve is illustrated (Figure 6A). The mouse DCPs expanded more than 6 orders of magnitude within 30 days. To generate DC vaccines, the tumor mouse-derived DCPs were differentiated for two days in IL-15 DC medium as described in Materials and Methods and transduced with LV-optE6E7 or LV-optE6E7 plus LV-calnexin. BALB/c mice bearing CT26/optE6E7 tumors were divided into two groups of five mice each group, and received two DC vaccinations at a one-week interval. We observed that mice injected with DCs modified by LV-optE6E7 plus LV-calnexin displayed increased survival than those modified by LV-optE6E7 alone (Figure 6B), a result consistent with previous findings using BM-derived IL-4 DCs [35].