ABSTRACT
Invariant chain (Ii) is traditionally known as the dedicated MHCII chaperone. Recent reports have broadened our understanding about various tasks that Ii plays including its physiological role in MHCI cross-presentation. Ii bound MHCI via the MHCII scaffolding CLIP peptide may facilitate MHCI trafficking to the endosomal pathway. The sorting function of Ii depends on two leucine-based sorting signals present in the cytoplasmic tail that acts as binding sites for the adaptor proteins AP-1/AP-2. Here we increased the Ii cross-presentation potency by replacing these with an AP3 motif resulting an efficient transport of Ii from TGN to late endosomes. We also replaced the CLIP region of li with a therapeutically relevant peptide, MART-1. We found the Ii AP3mutant-MART1 construct was capable of loading MHCI and stimulate specific T-cell response more efficiently than the wild type counterpart. The results show that Ii with an AP3 binding sorting motif carrying peptide epitope(s) can promote efficient antigen presentation to cytotoxic T cells (CTLs) independent of the ER located classical MHCI peptide loading machinery.
Introduction
Vaccination demands successful immune response that in turn depends on activation of both CD8+ and CD4+ T cells in the context of major histocompatibility complex I and II (MHCI and antigen presentation. The processing of the antigen and its subsequent presentation on MHCI or MHC II at the surface of antigen presenting cells (APCs) requires the coordinated action of different accessory molecules and chaperones. Many strategies have been described to obtain a robust immune response, such as the use of carrier proteins to improve peptide loading to MHC molecules, and targeting of antigens to “favorable” intracellular pathways where MHC reside 1-7. MHCI and MHCII are not loaded by the same cellular machinery, and they are dependent on different trafficking signals. A successful MHC II antigen presentation largely depends on Ii and peptide loading in the endosomal pathway, while MHCI peptide loading is independent of Ii and occurs primarily in the endoplasmic reticulum 8.
The targeting to the endosomal pathway of MHCII relies on the two-leucine-based sorting signals Leu7/Ile8 and Met16/Leu17 of Ii 9-11. These signals sequences are present in the cytoplasmic tail of Ii and bind the adaptor proteins AP-1 and AP-2 12, 13. These APs are generally found at the trans-Golgi network (TGN) (AP1) and plasma membrane (AP2) and act as coat proteins that bind the donor membrane in order to assemble a scaffold for vesicle budding 14,15. The APs thus mediate sorting from the TGN to endosomes directly or via the cell surface. Upon entry into the endosomal pathway, Ii is sequentially degraded leaving the class II-associated Ii peptide (CLIP) bound to the MHCII groove. CLIP is subsequently exchanged for specific antigenic peptides in the later parts of the endosomal pathway prior to transport to the cell surface for presentation to CD4+ T cells. Several vaccination studies have shown that replacement of the CLIP region with an antigenic peptide can lead to efficient MHCII loading and specific presentation to CD4+ T cells 1, 2, 16.
The classical view is that MHCI encounters its (endogenous) antigenic peptides in the ER and this complex is transported to the PM and presents the antigen to the cytotoxic T cell 17. However, it was demonstrated a few years ago that MHCI like MHCII may be loaded in the endolysosomal pathway guided by Ii 18, but it was not shown where in the endosomal pathway this took place. An Ii-MHCI interaction was also demonstrated by van Luijn and coworkers who showed that CLIP efficiently binds to several MHCI molecules in leukemic cells 19. Furthermore, our recent study showed an Ii with CLIP substituted by a MHCI specific tumor antigen is efficient enough to load MHCI and activate T cell specific response in a proteasome/TAP/tapasin independent manner 20. This strategy was found to be as efficient as exogenous loading of synthetic peptide in vitro, and thus identified a novel loading pathway for MHCI which may lead to novel vaccine strategies. In one of our earlier study, we have shown that it is possible to redirect a fusion protein with the li tail to late endosomes by introducing this AP-3 binding motif to the cytoplasmic tail of the li 21. Here we show that the introduction of AP-3 binding motif in Ii itself re-routed this molecule to proteolytic late endosomal compartments skipping the conventional trafficking route via the PM. As a consequence, this AP3 containing Ii protein had a dramatically shorter half-life than its wt counterpart. We have further investigated the potency of Ii-MHCI mediated antigen presentation with the Ii trafficking mutant designed to bind AP-3 21, 22 where we replaced the CLIP region with the cancer relevant peptide MART-1 20, 23, 24. This mutant with CLIP/Mart-1 replacement was found to have an improved potency to activate CD8+ T-cells compared to the Iiwt and was able to increase the amount of peptide-MHCI on the plasma membrane (PM). Taken together, we find that an improved Ii-based antigen loading/presentation of this peptide may be achieved by routing the Ii and most likely also MHCI to a late stage of the endosomal pathway.
RESULTS
Biochemical characterization of IiR4RP6/L17A
The Iiwt sorting signal (Q4RD6)L7I, found to bind the adaptors AP-1 and AP-2 13, was replaced as shown in Figure 1A by an (R4RP6)L7I motif, which is a strong AP-3 binder 22 and found to mediate direct TGN to endosomal sorting 21. We first tested whether the invariant chain with the AP3 motif, IiR4RP6/L17A was a trimer. In addition to Ii wild type we included the double leucine mutant IiL7A/L17A known to accumulate at the cell surface due to inactive sorting signals 25. As shown, all constructs were able to form trimers suggesting that the cytosolic tail mutations did not affect the trimer assembly (Fig. 1B). The abundance of mutant li trimers even in the reduced sample shows the efficiency of the mutant li in making stable trimers that did not compromise on its structural stability (Fig. 1B). However, the protein amount of the AP3 mutant was significantly less than the two others.
Arginine motifs may mediate ER retention and the Ii mutant (MDDRRPL7I) could therefore affect Ii release from the ER thus reducing the total protein level of Ii 26, 27. To test for this, we performed an Endoglycosidase H (Endo H) treatment, where passage through the Golgi apparatus prior to endosomal sorting is monitored by acquisition of Endo H resistance 28. Three Ii fractions were detected for the wild type and all the mutants of Ii (Fig. 1C). Thus, all constructs gained Endo H resistance indicating that despite the presence of the RRP amino acid sequence, the Ii mutant can egress the ER. Further to investigate why the level of IiR4RP6/L17A is differed from Iiwt, we performed a pulse-chase experiment to monitor the half-life (t1/2) of this protein. Transfected cells were pulsed with 35Met/35Cys containing media, and chased for various time points, followed by an immunoprecipitation of total Ii. As shown in Figure 1D, the t1/2 of Iiwt was approximately three hours, whereas IiR4RP6/L17A had a half-life closer to one hour, which suggests a faster kinetic to late endosomal compartments. IiL7A/L17A shown to accumulate at the plasma membrane 11 was not degraded after four hours and served as a control. As shown earlier, a TFR fusion protein with the IiR4RP6/L17A cytosolic tail can be targeted directly to the late endosomes/lysosomes 21. The short half-life indicates that Ii with the IiR4RP6/L17A tail also followed this pathway. To test for such a late endosomal proteolytic localization of IiR4RP6/L17A, we added either the Cathepsin S inhibitor or the broad protease inhibitor Leupeptin, both taken up by endocytosis to the transfected cells. A combination of both cathepsin S and leupeptin were able to protect IiR4RP6/L17A from degradation while the li L7A/L17A remained protected with either of the protease inhibitors (Fig. 1E). As expected, the Ii wild type, trafficking to the endosomal pathway via the PM 29, 30 was also protected by both Leupeptin and Cathepsin S. Interestingly, the IiR4RP6/L17A mutant needed both the leupeptin and the cathepsin A inhibitor for maximal inhibition, most likely as this construct is targeted to late endosomes which are more difficult to reach by the endocytosed inhibitors than the wild type Ii which traffics via the PM.
AP-1 is located at the TGN and AP-2 is a plasma membrane adaptor 14, AP-3 is involved in binding and sorting of proteins to late endosomes and detected both at the TGN and between early and late endosomes and is therefore believed to be involved both in sorting from TGN to late endosomes and endosomal maturation 31. To further study the pathway of our AP3 binding construct, we performed RNAi depletion of AP-3 and investigated the effect on the protein level. As shown in Figure 1F, AP-3 depletion resulted in a dramatic accumulation of both IiR4RP6/L17A and Iiwt was also protected from degradation, but less. Together with the protective effect of the protease inhibitors in the endosomal pathway, such a strong protective influence of AP-3 depletion on IiR4RP6/L17A is in line with a hypothesis that IiR4RP6/L17A is sorted directly from TGN to the late proteolytic pathway. The control Iiwt is less affected as it is sorted via the PM and only affected by the endosomal maturation inhibition caused by the AP3 inhibition.
Subcellular distribution of IiR4RP6/L17A
The subcellular distribution and trafficking of IiR4RP6/L17A was further investigated using live cell confocal imaging approach. Madine Derby Canine kidney (MDCK) cells were chosen for their high tolerance for laser exposure. The cells were transfected with either Iiwt or IiR4RP6/L17A N-terminally fused with red fluorescent protein mCherry, together with early and late endosomal markers, GFP-Rab5 or GFP-Rab7a respectively. To measure trafficking via the PM the cells were incubated with anti Ii-antibody, M-B741-Alexa647, one hour prior to imaging. It is furthermore known that Iiwt imposes enlargement of endosomes and causes a delay in endosomal maturation 29, 32-34. Due to this delay, at early time points the antibody reach primarily early endosomes, but gradually throughout the next 2-4 hours, the antibody was also seen in late endosomes 20, 29. We also observed enlarged endosomes in the cells transfected with Iiwt (Fig 2A and B) and colocalization with GFP-Rab5 and GFP-Rab7a, distributing almost equally (55-60%) within early and late compartments. In addition, more than 60% of Iiwt found to colocalize with M-B741-Alexa 647 (Fig. 2C), confirming trafficking via PM.
In contrast, mCherry-IiR4RP6/L17A showed a 75% colocalization with the late endosomal marker GFP-Rab7a and less than 10% with GFP-Rab5 (Figure 2A and B). This cellular distribution indicated that the Ii mutant followed a direct sorting route to late endosomes circumventing the cell surface. In further support, we demonstrate that only 12% of mCherry-IiRRP/L17A colocalized with M-B741-Alexa 647 (Fig. 2C), which corroborates direct sorting to late endosomes, and is in line with our biochemical characterization of IiR4RP6/L17A. Because of its rapid turnover (Fig. 1D), and the accumulation in late endosomal compartments, IiR4R6P/L17A, did not delay endosomal maturation. The residual uptake of M-B741 was most likely due high expression level of the mutant Ii in some of the cells being missorted to the PM. We confirmed our observations with Ii transduced SupT1 cells. In this assay, the Ii membrane expression was monitored by staining cells without prior permeabilization and later analyzed by flow cytometry (Fig. 2D). This demonstrated that the Ii mutant did not appear on the cell surface. Finally, we used live confocal imaging and confirmed that the IiR4R6P/L17A mutation did not affect the previously described MHCI-Ii association 20(Sup Fig. 1).
Soluble T cell receptors detect peptide loaded HLA-A2 from IiR4RP6/L17A
Since the IiR4RP6/L17A mutant is re-routed to a degradation trafficking pathway, we tested whether such a CLIP replaced Ii construct combined with this tail mutation could still load MHCI. To this end, we first detected that the peptide-MHC complex of cell expressing different constructs with a soluble T cell receptor (sTCR) 35 specific for MART1p. sTCRs have a low affinity for their target, however, we were able to detect Ii-MART1p, inserted in the CLIP region of IiR4RP6/L17A (Fig. 2E). Although very low, this result suggests that the peptide was well loaded on the MHCI molecule. As a control, cells expressing HLA-A2 single-chain trimer (SCT) combined with MART-1 peptide (SCT-M1) were used and showed an expected saturating signal. Taken together, our data support an improved antigen loading ability of IiR4RP6/L17A over the Ii wild type construct. In addition, the trafficking of Ii to the plasma membrane does not seem to be required to get an efficient loading.
Cells expressing Ii carrying tumor-associated epitopes efficiently load HLA-A2 and specifically activate CD8+ T cells
We have previously shown that Ii interacts with MHCI, and that the human MHCI allele, HLA-A*02:01 (HLA-A2) colocalized with Ii throughout the endosomal pathway. We have also demonstrated that CLIP-replaced Ii efficiently activated antigen specific CTLs when expressed in HLA-compatible APC 20. We therefore compared the ability of the IiR4RP6/L17A mutant to load HLA-A2 peptides with the CLIP-replaced Ii construct (Fig. 3A). J76 cells stably expressing MART-1 specific TCR (DMF5) 35 were incubated with HLA-A2 positive cells expressing different Ii constructs. IL-2 secretion was used as a read-out for specific TCR stimulation; SCT-MART1 and SCT with an irrelevant peptide (SCT-irr) expressing cells were used as controls. When MART-1 peptide was loaded on HLA-A2 utilizing IiR4R6P/L17A as carrier, the intensity of the stimulation was almost equal to the stimulation observed with SCT-M1, confirming an increase in peptide loading compared to Iiwt (Fig. 3B). In order to support these data, we performed a DC priming study using autologous donor cells. To this end, we assessed the priming ability of DC transfected with Ii mutant (Ii17R4RP6/L17A MART-1) compared to Iiwt MART-1 or the MART-1 peptide. We found that the mutant li (Ii17R4RP6/L17AMART-1) was significantly more efficient and superior to peptide at priming primary CD8+ T cells, whereas IiwtMART-1 seemed to be improved but did not reach significance (p=0,052, Fig. 3C), hence at this stage can be considered equal to peptide loading. In addition the Ii17R4RP6/L17A MART-1 was not significantly superior to IiwtMART-1 (p=0.53, not shown). Taken together we can at this stage only conclude that the new construct is functional in DCs, but we might require to test more peptides before we can reach the same conclusions as in the Jurkat system (Fig. 3B). This is in agreement with our previous data where we show that Iiwt construct performed as efficiently as peptide loaded cells 20. A possible explanation could be that a mutant Ii and wild type bind differently to MHCI. However, by co-immunoprecipitation experiments we found no difference in MHCI binding to the mutated li (Ii17R4RP6) as compared to binding to wild type Ii (Sup Fig. 2). Together these data support the proposition that IiR4R6P/L17A improved the loading of peptide placed in CLIP region for MHCI mediated antigen presentation.
DISCUSSION
Previous studies have revealed that neither MHCI nor MHCII are AP3 dependent for trafficking and antigen presentation 36. The kinetics of Ii transport and degradation are also unaffected in cells lacking AP-3 37. Through introduction of the AP-3, instead of the AP1/AP2 binding motif, we successfully re-routed the Ii towards late endosomal compartments. In addition to the re-routing of Ii, the insertion of AP3 binding motif increased the kinetics of trafficking of Ii to the late endosomal compartment. Overall, our effort of bringing mutations in the Ii created a strict and direct sorting pathway with improved MHCI peptide loading efficacy. Here, we characterized the sub-cellular distribution of IiR4RP6/L17A and compared with the Iiwt. Equal distribution of Iiwt to the early and late endosomes was observed whereas IiR4RP6/L17A was found to be mainly colocalizing with late endosomal markers. The direct sorting to endosomes also overcomes the property of delayed endosomal maturation 29 substantiating a faster antigen processing.
Recent studies have shown the role of Ii in trafficking of MHCI to the endosomal pathway and its implication in cross presentation 18, 20. The description of Ii as a vehicle to perform antigen presentation has brought this molecule to the doorstep of the clinic not only as a target for immunotherapy 38 but also as a vector for increasing immune reactions towards specific oncogenic antigens 20. Genetic exchange of the CLIP region with a peptide antigen substantially loads MHCI and presents the antigen on the cell surface. Our results showed that the antigenic peptide carried by IiR4RP6/L17A can be detected by T cell carrying a specific TCR. Additionally, the antigen detection by the sTCR supports the specificity and the robustness of the IiR4RP6/L17A mutant loading. Taken together, our data show that the modification in trafficking induced by the IiR4RP6/L17A mutation can improve peptide loading and thus MHCI-peptide levels at the cell surface. In this study, IiR4RP6/L17A carrying MART-1 antigen was shown to be equally efficient in activating specific TCR carrying cells in comparison to liwt. In addition, we found that IiR4RP6/L17A mutant was also competent to load enough peptide onto DC in order to prime primary naive T cells. These studies confirm the capacity of MHCI loading in the endosomal pathway and establish that the loading may take place in the proteolytic later parts of this pathway. Additional studies will be necessary to determine if such and AP3 binding mutation will be advantageous in vivo, for instance to improve immunotherapy using modified Ii as an immunization vector.
METHODS
Recombinant cDNA constructs
cDNA encoding human Iip33 wt 9, was subcloned into the pcDNA3 expression vector at KpnI-BamHI. Human Iip33 mutants, Ii L17A and Ii L7A L17A, in the PSV51L expression vector have also been described 25. KpnI and BamHI restriction sites were introduced up and downstream of the Ii sequences respectively by PCR and the following primers were used: Ii-KpnI forward 5’ AGAGA GGGTACCGTCATGGATGACCAGCGCGAC 3’. Ii-BamHI reverse - 5’ AGAGAGGGATCCTCACATGGGGACTGGGCCCAG 3’. The Ii mutants were thereafter subcloned into pcDNA3 at KpnI-BamHI, behind the T7-RNA polymerase promoter, and Ii L17A was subsequently used as template for PCR quick change mutagenesis (all reagents used were included in the kit; QuickChange® Site-Directed Mutagenesis (Stratagene, La Jolla, CA, USA)) in order to generate the AP-3 binding motif RRP 21. Primers used: L17A RRP sense 5’ CCGTCATGGATGACCGTCGTCCCCTTATCTCCAACAATG 3’ and L17A RRP anti-sense 5’ CATTGTTGGAGATAAGGGGACGACGGTCATCCATGACGG 3’. All primers were purchased by Eurofins MWG Operon (Ebersberg, Germany). mCherry-Ii was made by cloning Iiwt in frame to the C terminal end of mCherry without the stop codon in pcDNA3 (a kind gift from Terje Espevik, NTNU, Trondheim, Norway). mCherry-IiR4R6P/L17A was purchased from GenScript (Piscataway, NJ, USA). HLA-A2-GFP has been described and transfections were carried out as already described 20. GFP-Rab5 and GFP-Rab7 were supplied by Cecilia Bucci 39. All CLIP-antigenic peptide constructs (IiMART1) were cloned by site direct mutagenesis of the Iiwt and IiR4RP6/L17A construct subcloned in pENTR vector (Invitrogen, Oslo, Norway). The mutagenesis to change the CLIP peptide (MRMATPLLM) into antigenic peptides (MART1: ELAGIGILTV) was performed as described 20. After sequence verification, these constructs were recombined into a Gateway-converted pCI-pA102 40
Cell culture, transfections and RNA interference
HEK293 cells, human epithelial HeLa-Oslo and Madin Darby Canine Kidney (MDCK) cells were grown in Dulbecco’s Modified Eagle Medium (DMEM, Bio Witthaker, Walkersville, MD, USA). All media were supplemented with heat-inactivated 10% fetal calf serum (FCS, HyClone, Logan, UT, USA). J76 were a kind gift from Miriam Hemskerk, (Leiden University Medical Center, Leiden, The Nederland) SupT1 from Martin Pule (UCL, London, Great Britain), both cell lines were grown in RPMI+10% fetal calf serum. All cells were grown in a 5% CO2 incubator at 37°C. Transient transfections were performed with either lipofectamine 2000 reagent from Invitrogen (Hek293 cells, MDCK) or with FuGENE 6 (ProMega) (HeLa), both according to manufacturer’s protocols. For siRNA interference (RNAi) we used the following oligonucleotides; the sense µ3A, 5′-GGAGAACAGUUCUUGCGGC-3′ and the antisense 5′-GCCGCAAGAACUGUUCUCC-3′ oligos, for negative control a scrambled sequence was used, sense: 5′ ACUUCGAGCGUGCAUGGCUTT 3′ and antisense scrambled control 5′ AGCCAUGCACGCUCGAAGUTT 3′. All of the oligos were from Eurofins MWG Operon (Ebersberg, Germany) and are previously described 36, 41. Transfection of HeLa with siRNA was performed as previously described 42.
Antibodies and reagents
M-B741 was purchased from BD Biosciences (Franklin Lakes, NJ, USA). Labeling of antibody was with Alexa-647 performed according to manufacturer’s protocol (Invitrogen/Molecular Probes, Carlsbad, CA, USA). Anti-actin was purchased from AbCam, (Cambridge, UK). The anti AP-3 antibody is affinity-purified rabbit antiserum directed at its µ subunit and was a kind gift from Professor Margaret S. Robinson (Cambridge, UK). The secondary antibodies: sheep anti-mouse- and sheep-anti rabbit-HRP were acquired from Invitrogen/Bio-Rad (Hercules, CA, USA). Anti-FLAGM2 monoclonal antibody was purchased at Sigma-Aldrich (Oslo, Norway). Soluble DMF5 TcR was prepared as described by Walseng et al. 35.
Biochemical analyses
Metabolic labeling was performed using S35-labeled Cysteine/Methionine (Perkin Elmer, Waltham, MA, USA). Cells were seeded to 60%-70% confluence in 6-well plates; washed three times in Cys/Met-free DMEM; incubated in Cys/Met-free DMEM for 45 min followed by a 30 minutes pulse with Cys/Met-free DMEM supplemented with 50µCi S35. For the pulse chase assay, the cells were washed three times in DMEM containing 2mM L-glutamine, primocin, and 30% FCS and chased for indicated time periods. Immunoprecipitations were done at 4 °C over night with 1-2µg ml-1 antibody in lysis buffer (50 mM Tris-HCl, pH 7,5, 150 mM NaCl, 1% Tx100) supplemented with the protease inhibitor cocktail Protease Arrest (G-Biosciences, St. Louis, MO, USA). Antigen–antibody complexes were captured with Protein G-coupled Dynabeads (Invitrogen) and re-suspended in gel loading buffer containing 2% SDS, 125mM TrisHCL, 20% glycerol and 5% β-mercaptoethanol, or its non-reducing β-mercaptoethanol free counterpart. The samples were boiled for 5 min at 95° C, loaded onto 4-20% Tris-HEPES-SDS gels (Pierce, Rockford, IL, USA), and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). Antibody incubation was done in 5% skim milk (BioRad, Hercules, CA, USA) at room temperature and immunoprecipitated protein was detected using Amersham® ECL Plus Western Blot Detection System (GE Healthcare, Buckinghamshire, UK). Radioactivity, however, was detected directly on ECL films (GE Healthcare, Buckinghamshire, UK). For the experiments including protease inhibitors, the same procedure was followed as for metabolic labeling. During the 30 min S35-Cys/Met pulse, 20 nM Cathepsin S Inhibitor (Merck Chemicals Ltd., Nottingham, UK) and/or 100 μM Leupeptin (SIGMA ALDRICH) were added. The procedure was then continued as described above. For Endo H digestion, the beads were resuspended in 0.1 M sodium phosphate buffer (pH5.5) containing protease inhibitor as described above. The samples were divided into two and incubated for 15 minutes at room temperature with, or without 0,5 mU of Endo H (SIGMA). After the Endo H treatment, the samples were boiled at 95°C, and loaded onto gels as described above.
Flow cytometry
Indicated samples were acquired using a BD LSR II flow cytometer and the data were analysed using FlowJo software (Treestar Inc., Tilburg, The Netherlands).
Spinning disk – and confocal laser scanning microscopy
MDCK cells were grown to 70% confluence in 35 mm microwell dishes (MarTek, Ashland, MA, USA). The cells were then transfected with; GFP-Rab5, GFP-Rab7a, Ii constructs and HLA-A2-GFP/β2m. 1h prior to imaging, the cell medium was exchanged with complete DMEM without phenol red, and cells were incubated with M-B741-Alexa 647 to a final concentration of 1 µg/mL. Live imaging was performed to eliminate fixation artifacts. Confocal images were acquired on an Olympus FluoView 1000 inverted microscope equipped with Plan/Apo 60/1.10 NA oil objective (Olympus, Hamburg, Germany). Constant temperature was set to 37 °C and CO2 to 6% by an incubator enclosing the microscope stage. Fluorochromes were exited with 488nm, 543nm and 647nm lasers. All image acquisition was done by sequential line scanning to eliminate bleed-through. Live films were acquired using an Andor Revolution XD Spinning Disc microscope with PlanApo 60×1.42 NA oil immersion objective, as this microscope provides an ideal platform for high speed, high signal to noise imaging, with low bleach rate and low photo-toxicity. Three lasers were used; 488nm, 561nm, and 640 nm, and 4 frames per minute were acquired for the total of 25 min. Images was processed with ImageJ (NIH, USA) and Illustrator (Adobe systems Inc., San Jose, CA, USA).
In vitro generation of Dendritic cells for antigen presentation and T-cell priming assay
Immature dendritic cells (DCs) were generated essentially as described in Subklewe, et al. 43 Briefly, monocytes obtained from leukapheresis product (REC Project no: 2013/624-15) were cultured for 2 days in CellGro DC medium (CellGenix, Freiburg, Germany) supplemented with GM-CSF and Interleukin-4 (IL-4) in Ultra-low attachment cell culture flasks (Corning). The immature DCs were electroporated with either mRNA encoding for mutant Ii17R4RP6/L17A MART-1 or wild type Ii (liwt) carrying MART-1 peptide. Cytokines facilitating maturation were used (IL-1β, IL-6, TNF-α, IFN-γ (all from PeproTech, Rocky Hill, NJ), prostaglandin E2 (PGE2), and TLR7/8 agonist R848 (MedChem Express, Sweden)) 44 and cultured for 24h. Mature DCs were used in T cell priming experiment. DCs electroporated with wild type li mRNA (no CLIP replacement) and DCs loaded with MART-1 peptide (10 µM) were included as negative and positive control, respectively, in the priming assay. Briefly, the distinct DC populations were cultured with autologous PBMCs at 1:10 DC:T cell ratio. On day 3, T cell cultures were supplemented with IL2 and IL7.On day 8, T cell cultures were re-stimulated with DCs and 10 days later T cells were stained with MART-1 dextramer (Immudex, Copenhagen, Denmark) to assess the presence of MART-1 antigen-specific T cells in the cultures.
CONFLICT OF INTERESTS
The authors declare that they have no conflict of interests.
ACKNOWLEDGEMENTS
We thank the NorMIC Oslo imaging platform at the Department of Biosciences, University of Oslo for use of imaging facilities. The financial support of the Norwegian Cancer Society (grants 4604944 to O. B.), the Research Council of Norway (grant 230779 to O. B., grants 244388 and 254817 to E.M.I and M.R.M. and N.M, respectively and through its Centre of Excellence funding scheme to O.B., project number 179573) is gratefully acknowledged and South-Eastern Norway Regional Health Authority to S.W. (Innovation grant 13/00367-88) and to E.M. (grant 2010021).