Exploring dendritic cell based vaccines targeting survivin for the treatment of head and neck cancer patients

Peripheral blood from five HLA-A2 positive HNSCC patients was used to monitor the presence of survivin specific T cells. The HNSCC patients were treated at the VU University Medical Center in Amsterdam, the Netherlands with surgery, chemotherapy, radiotherapy or a combination of these. Blood was drawn at least six weeks after the last treatment via a vena puncture. All patients signed an informed consent form, approved by the Institutional Review Board (METc-VUmc registrationnumber:2009/205). Lymph node derived T cells from a patient suffering from locally advanced breast cancer was used to determine TAA specific T cells by ELIspot. The patient took part in a IRB-approved clinical trial where she received 6 neoadjuvant chemotherapy cycles and signed a written informed consent (METc-VUmc IRB00002991 IORG number 0002436).

Multiple survivin constructs were designed, the survivin inserts were codon modified and generated by Geneart (Regensburg, Germany) and cloned into pGEM4Z vectors (kindly provided by dr. Viggo van Tendeloo, Antwerp, Belgium). The mini-survivin sequence (including the immunodominant HLA-A2 binding T cell epitope) was cloned into pGEM as an EcoRI-NotI fragment, resulting in pGEM-mini-survivin (mini-survivin in short for the mRNA). The resulting open reading frame encoded the following amino-acid-sequence; MQFTELTLGEFLKL DREEEREAEFKSAFTELTLGEFLKL DREEREEERRNKQFTELTLGEFLKL DREEREEERR FNKKQFTELTLGEFLKL DRE*.

The Ubi-mini-survivin sequence was cloned into pGEM as an XhoI-NotI fragment, resulting in pGEM-Ubi-mini-survivin (Ubi-mini-survivin in short for the mRNA). The open reading frame encoded the following amino-acid-sequence; MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGVVN SEKFQFTELTLGEFLKL DREEEREAEFKSAFTELTLGEFLKL DREEREEERRNKQFTELTLGEFLKL DREEREEERR FNKKQFTELTLGEFLKL DRE*.

The immunodominant HLA-A2 binding T cell epitope survivin(96–104) is indicated in bold in both sequences. The survivin(96–104 epitope is located in the 3rd exon of survivin and is therefore present in wild type survivin and the splice variants 2B and 3B, but absent in survivin splice variants 2alpha and Dex3.

Two full length survivin wild type constructs were used (pGEM-survivin and pGEM-Ubi-survivin along the lines described above. A codon-modified minigene (Ubi-mini-MART for the mRNA) containing four repeats of the altered peptide ligand ELAGIGILTV (was used as described before [38]). In vitro transcription was carried out according to manufacturer’s protocol of the T7 mMessage-mMachine kit (Ambion, Huntingdon, Cambridgeshire, UK) to generate m7G(50)pppGcapped IVT mRNA (CAP). mRNA quality was checked by agarose gel electrophoresis. RNA concentration was assayed by spectrophotometrical analysis at OD_260.

Healthy donor derived peripheral blood mononuclear cells (PBMC) were isolated from HLA-A2 positive buffycoats (Sanquin, Amsterdam, The Netherlands) by density gradient centrifugation using Lymphoprep (Nycomed, Oslo, Norway). Monocytes were isolated by positive selection using a MACS column (MACS; Miltenyi Biotec, Bergisch Gladbach, Germany). PBMCs were labeled with anti-CD14 beads (Miltenyi Biotec) followed by MACS sorting according to the manufacturer’s protocols. Subsequently, the monocytes were cultured with 100 ng/mL recombinant human granulocyte colony stimulating factor (rhGM-CSF; Sagramostim, Berlex) and 10 ng/mL interleukin 4 (IL4; R&D systems) for 6 to 7 days in IMDM medium supplemented with 8% fetal calf serum (FCS; HyClone), 2 mM L-glutamine and antibiotics (100 IE/ml penicillin and 100 μg/mL streptomycin, life technologies). The immature DCs were then matured by culturing them with monocyte conditioned medium (MCM) 25% v/v (generated as previously described [39]) and 50 ng/mL tumor necrosis factor (TNF)α (Miltenyi Biotec). DCs were analyzed phenotypically to determine the maturation status using flow cytometry. The following antibodies were used for phenotypic analysis: IgG1, CD14, CD80, CD83, HLA-DR (BD Biosciences, Heidelberg, Germany), CD1a, CD86, CD40 (Pharmingen, San Diego, CA, USA) labeled with Phycoerythrin (PE). Mean fluorescence index was calculated as follows: MFI-Index = (mean fluorescence intensity marker)/(fluorescence intensity isotype).

Mature DC were electroporated [40] with eGFP, Ubi-survivin or survivin mRNA. Cells were electroporated with 10 μg mRNA in 200 μL electrobuffer (Cell Projects Ltd) in 4 mm cuvets (Cell Projects Ltd) at 300 Volt and 150 μF (Easyject plus, Equibio, Kent, UK). One hour after electroporation cells were washed and plated out in 1*10⁶/mL IMDM medium supplemented with 8% fetal calf serum (FCS; HyClone), 2 mM L-glutamine and antibiotics (100 IE/ml penicillin and 100 μg/mL streptomycin, life technologies).

For Western blot, 4 hours after electroporation with mRNA encoding either eGFP, survivin or Ubi-survivin 1*10⁶ cells were harvested and washed with phosphate-buffered saline (PBS). DC were harvested for cytospin preparations, at time points 2, 4, 6, and 24 hours after electroporation with RNA encoding either eGFP, survivin or Ubi-survivin.

Isolation of resting CD8β positive T cells from HLA-A2 positive, healthy donor PBMC (Sanquin) was performed by positive selection using a MACS column (Miltenyi Biotec). For this purpose, total PBL were incubated with anti-CD8β mAb (clone: 2ST8.5H7, Beckman Coulter, Inc.) and subsequently with microbead-conjugated anti-mouse IgG Abs (Miltenyi Biotec), followed by MACS sorting according to the manufacturer’s protocol. Per donor 6*10⁶ CD8β + T cells were cultured in Yssels medium supplemented with 1% hAB in order to obtain resting T cells. Another aliquot of CD8β + T cells were cultured in the presence of 100 U/mL IL2 and 500 ng/mL PHA in order to obtain activated T cells. After a 48 hour incubation period, the T cells were harvested for FACS and Western blot analysis. T cell activation levels were determined by incubating the T cells with anti-CD25 PE and CD69 FITc (BD Biosciences). Cells were subsequently measured by flow cytometry.

Cell pellets were resuspended in lysis buffer (50 mmol/l Tris/HCl, pH 8.0, 0.5% NP-40, 5 mmol/l EDTA) containing protease inhibitors. Protein levels in the samples were quantified according to manufacturer’s protocol using BCA Protein Assay Kit (Pierce Biotechnology, Rockford, USA). Expression of survivin was detected using a 1:1000 dilution of rabbit-anti-survivin (clone 71G4B7, Cell Signaling Technology, Inc, Boston, USA), followed by incubation with a 1:1000 dilution of polyclonal swine-anti-rabbit immunoglobulin/HRP (DakoCytomation, Glostrup, Denmark). Proteins were visualized with the enhanced chemo-luminescence technique (Amersham Pharmacia Biotech, Piscataway, NJ, USA).

Cytospin samples of dendritic cells were prepared as followed; per condition 20.000 cells were taken up in 100 uL PBS and spun down for 3 minutes at 600 rpm on a Superfrost plus slide (Thermo Scientific, Braunschweig, Germany). Cytospins were next air dried overnight, fixed in acetone for 10 minutes and stained with anti-survivin(clone 71G4B7, Cell Signalling Technology, Inc.). According to the manufacturer this rabbit mAb was produced by immunizing animals with a synthetic peptide corresponding to residues surrounding cysteine 60 of human survivin and detects endogenous levels of total survivin protein.

Isolation of resting CD8β positive T cells from healthy donor PBMC (Sanquin) was performed by positive selection using a MACS column (Miltenyi Biotec). For this purpose, total PBL were incubated with anti-CD8β mAb (clone: 2ST8.5H7, Beckman Coulter, Inc., Marseille, France) and subsequently with microbead-conjugated anti-mouse IgG Abs (Miltenyi Biotec), followed by MACS sorting according to the manufacturer’s protocol. Mature DC were electroporated with survivin mRNA in combination with IL12 mRNA or IL21 mRNA as described above. One hour after electroporation DCs were washed and multiple mini-cultures containing 0.5–1 ×10⁶ CD8β T cells, 0.5–1 ×10⁵ mRNA-transfected DCs and 0.25–0.5 ×10⁶ irradiated autologous CD4⁺ T cells were set up in Yssel’s medium supplemented with 1% hAB and 10 U/mL IL7. After 10 days, T cells were analyzed for specificity using PE- and/or APC-labeled HLA-A^*0201 tetramers (Tm) presenting the survivin 5, 95, 96 or 96 M epitope (survivin 5–14: TLPPAWQPF, survivin 95–104: ELTLGEFLKL, survivin 96–104: LTLGEFLKL, survivin 96 M: LM LGEFLKL). We have used in-house prepared tetramers, which at that moment in time could not be tested for specific binding since bona fide survivin specific T cell clones were unavailable to us. In a similar case in the past we have validated HPV specific tetramers by using two different fluorescent dyes [41], thus excluding non-specific binding of one of the tetramers. Staining was performed in PBS supplemented with 0.1% BSA and 0.01% Sodium azide for 15 min at 37°C. On day 10 the bulk cultures were re-stimulated with mRNA electroporated DCs and the next day 20 U/mL IL2 (R&D systems, Oxon, UK) was added. Tetramer staining was performed after 7 days and the bulk cultures were re-stimulated weekly as described above.

Tumor-draining lymph node material, from patients suffering from locally advanced breast cancer, was used to determine the presence of survivin specific T cells. Antigen-induced interferon gamma (IFN-γ) secretion by survivin induced T cells was measured by an EnzymeLinked Immuno-spot (ELIspot) assay using a modification of a described technique [42]. After wells of nitrocellulose-bottom 96-well plates (MultiScreen-HA; Millipore, Billerica, MA) were coated overnight with 50 μL monoclonal anti-IFN-γ antibody (10 μg/mL; R&D Systems, Minneapolis, MN), the plates were washed, 1% bovine serum albumin was added to each well to block nonspecific binding, and the plates were washed again. The survivin long peptides shown in Additional file 1: Figure S1C were used to load target cells (autologous moDCs). DCs alone or with the survivin long peptides (5 μM) were added in a total volume of 100 μL of complete RPMI medium to the wells and incubated at room temperature for 1 h. The CD4+ or CD8+ T cells were added (1 × 10⁴ cells/well) in 100 μL of complete RPMI medium with added IL2 (20 U/mL) and incubated for 48 h at 37°C with the target cells (1 × 10³ cells per well). After the plates were washed, captured IFN-γ was detected by incubation with 50 μL of biotin-conjugated goat anti-human IFN-γ (R&D systems) for 2 h at 37°C in the dark, and the plates were rewashed, incubated for 1 h at 37°C with streptavidin-alkaline phosphatase (Zymed, San Francisco, CA) diluted 1:10.000 in conjugate buffer (1% bovine serum albumin in PBS plus 0.05% Tween 20), washed, and incubated with 50 μL of FAST 5-bromo-4-chloro-3-indolylphosphate nitroblue tetrazolium (BCIP/NBT) substrate (Sigma) for 30 min at 37°C in the dark. After drying the plates overnight in the dark, spot-forming cells (SFC) corresponding to individual IFN-γ-secreting cells were counted using an AID ELISPOT reader (Autoimmun Diagnostika GmbH, Strasbourg, Germany). The background number of SFC was determined by incubating the CD8⁺ cells with autologous moDCs alone or autologous moDCs loaded with irrelevant peptide (HPV long peptide), and wells containing CD8⁺ T cells were included as a negative control. Each determination was performed in duplicate, and the results were expressed as SFC/number of plated CD8⁺ T cells.

We measured the percentage of survivin specific T cells in PBMC derived from the peripheral blood of five patients with head and neck cancer. In three of these patients we found T cells binding HLA-A2 tetramers containing the survivin derived epitope TLPPAWQPFL (Survivin 5–14) (Figure 1A). For one of the patients T cell numbers enabled us to FACS sort the tetramer positive population and clone them by limiting dilution cloning (0.2 cell/well). In 3 out of 20 wells with growing T cells we detected 20-27% survivin specific T cells. A representative example of one of these cultures is given in Figure 1B. However, these three cultures fared poorly (scatter plot shown in Figure 1) and could not be maintained for prolonged periods of time, thus precluding functional analysis, e.g. cytokine production or tumor cell recognition and killing.

In a tumor-draining lymph node obtained from another patient, in this case suffering from locally advanced breast cancer and treated with six cycles of neo-adjuvant chemotherapy, we detected tumor antigen associated specific T cells. The spontaneously arising T cells were tested in an interferon-gamma ELIspot assay against a number of different nine-mer HLA-A2-binding peptides representing tumor associated CTL epitopes; Survivin-95, CEA-571, Her-2/Neu-369 and Her-2/Neu-654. As a negative control we used the virally derived HPV16E7(11–20) peptide. Indicated are the number of spots per 100.000 T cells. The results given in Figure 1C clearly shows survivin specific reactivity amongst reactivity against other TAA.

Having established that spontaneously arising survivin specific T cells can be detected in cancer patients we embarked on in vitro induction experiments using healthy donor derived mature DC, electroporated with in vitro transcribed mRNA, as stimulator cells and isolated CD8+ T cells as responders, the purpose being to optimize T cell stimulation protocols, employing mature DC, survivin and cytokine encoding mRNA, for in vivo vaccination studies in HNSCC cancer patients.

For the production of in vitro transcribed mRNA we designed a number of pGEM constructs encoding either full length survivin or mini-gene constructs containing T cell epitopes derived thereof. A survivin mini-gene construct containing four codon modified repeats of survivin95-107 was prepared as described previously for a different TAA (Mart-1/MelanA) [38]. DC need to process the immunodominant T cell epitope (survivin96-104) from this repeat and present it in the context of HLA-A2. In a number of constructs survivin or the mini-gene repeats were preceded by Ubiquitin sequences to enhance degradation through the proteasome resulting in enhanced MHC class I presentation. An overview of the different constructs, is shown in Figure 2.

Next we tested for the presence of survivin protein in transfected cells. Monocyte derived dendritic cells were electroporated with mRNA encoding either GFP as a control, or full length survivin either or not preceded by Ubiquitin sequences. Lysates of cells were prepared four hours after electroporation of mRNA. Western blot analysis showed a clear band for survivin protein, of the expected molecular weight of 16.3 kD, in mature DC electroporated with full length survivin mRNA (Figure 3A, lane 2), but not in control GFP electroporated DC (Figure 3A, lane 1). As expected no specific protein band for survivin was found in the lysate of DC after electroporation with Ubi-survivin mRNA (Figure 3A, lane 3). Staining of high molecular weight protein bands was observed in all three lanes.

Next we determined survivin expression in mature DC by immunohistochemical staining of cytospin preparations (Figure 3B). Cytospins were prepared at 2, 4, 6, 8 and 24 hours after electroporation of mature DC. Low level survivin staining was seen in GFP and in Ubi-survivin electroporated DC at all time points included in the analysis. Staining of survivin protein was most prominent in cytospins taken at 4 h and 6 h after electroporation of mature DC with full length survivin encoding mRNA. Specific staining was reduced to background level in these cells taken at time point 24 h. From these experiments we concluded that the full length survivin sequence, used in two of the DNA constructs shown in Figure 2, is capable of driving the production of survivin protein.

Next to the induction of survivin specific T cells we applied our protocols to the induction of Influenza specific (memory) T cells using peptide loaded mature DC, and Mart-1 specific (naive) T cells using mRNA electroporated mature DC [38, 41]. CD8+ T cells and monocyte derived mature DC from an HLA-A2 positive healthy donor were used.

After in vitro stimulation of CD8+ T cells with either peptide loaded (Influenza) or mRNA electroporated (Mart-1) mature DC we tested T cell bulk cultures for the presence of specific T cells employing tetramers. High percentages of specific T cells for Influenza (7% tetramer positive) and Mart-1 (19% tetramer positive) were detected after as little as two in vitro stimulations (Figure 4A). In this particular donor the percentage of survivin specific T cells remained below the detection limit. Influenza or Mart-1 specific T cells can be enriched easily by FACS sorting or cloned by limiting dilution cloning. Importantly such T cell clones or cultures can be maintained in culture, using feeder mix stimulations, at near hundred percent purity for prolonged periods of time without loss of specificity or reactivity (data not shown and data published by us previously [38, 41, 43‐45]).

Next we stimulated T cells with mature DC co-electroporated with full length survivin mRNA and IL21 mRNA. After as many as five in vitro stimulations we detected low levels of survivin specific T cells staining with tetramers containing either the wild type (0.2% positive for survivin(96–104) or M-modified survivin(96–104) epitope (0.5% positive)). Results are shown in Figure 4B. We noted interferon-γ production as measured in an ELIspot assay upon incubation of T cells with peptide loaded DC for this donor but not for a second donor (see Additional file 1: Figure S1 and Additional file 2: Table S1). After limiting dilution cloning, outgrowth of T cells was found in 69 of 920 wells seeded with survivin tetramer binding T cells obtained after FACS sorting. None of these wells,however, contained viable survivin(96–104) specific T cells.

We next sought to improve on the in vitro stimulation of specific T cells by using mRNA encoding Ubi-survivin, mini-survivin of Ubi-mini-survvin in combination with mRNA encoding IL12. Figure 5 shows the results after in vitro stimulation of T cells for three times with survivin/IL12 DC, followed by one IVS with Ubi-survivin/IL12 DC. The percentages of tetramer positive T cells are indicated for each well for survivin(5–14) (Figure 5A) and survivin(96–104) (Figure 5B) epitope (range 3-10% positive). Tetramer positive T cells were next MACS-isolated and put in culture on feeder mix. In contrast to our expectations based on prior experience with other T cell specificities, MACS enrichment of survivin specific T cells did not result in a T cell bulk culture with a high percentage of survivin specific T cells (Figure 5A and B right hand panel).

These results prompted us to perform more T cell inductions and monitor the appearance and persistence of survivin specific T cells more closely. We set up separate mini-cultures of T cells from one donor and stimulated them once with either mini-survivin/IL12 DC or Ubi-mini-survivin/IL12 DC. Subsequent restimulations were performed using B cells (JY), electroporated with mRNA encoding either mini-survivin or Ubi-mini-survivin in combination with mRNA encoding IL12. We set up 20 wells per condition, each indicated with a separate symbol in Figure 6. Prior to in vitro stimulation, survivin specific T cells were undetectable using HLA-A2-survivin(96–104) tetramers (data not shown). From the data presented it is clear that the number of tetramer binding T cells went up after the 3rd and 4th in vitro stimulation and reached a maximum after the 5^th stimulation (range 1-5% positive T cells) in conditions where we used IL12 mRNA in conjunction with either mini-survivin (Figure 6A) or Ubi-mini-survivin (Figure 6B). After the 6th in vitro stimulation the numbers went down, never to reach these levels again.

Fratricide by survivin specific T cells has been postulated as a possible explanation for the disappearance of survivin expressing T cells in culture. Endogenous expression of survivin by T cells would be a prerequisite for fratricide to take place in the experimental wells. In order to investigate this, we isolated resting and prepared activated T cells from three different healthy donors. Figure 7 shows the expression levels of the T cell activation marker CD25. As expected CD25 expression was low or absent on resting T cells (Figure 7A), whereas approximately 90% of the activated T cells showed high expression levels of CD25 (Figure 7B). Similar data were obtained for another T cell activation marker CD69 (data not shown). Resting and activated T cells were next used to determine endogenous survivin protein expression. From the Western blot data shown in Figure 7C it is clear that activated T cells express appreciable levels of survivin protein, whereas resting T cells do not.