The murine dendritic cell line DC2.4 (H-2^b) (25) and the murine macrophage cell line BMC-2 (H-2^b) (26) were originally obtained from K. Rock (University of Massachusetts Medical School, Worcester, MA, USA) and were cultured in RPMI 1640 (Lonza). The C57BL/6-derived methylcholanthrene-induced fibrosarcoma cell line MC57G (H-2^b) (27) was grown in MEM (Gibco) and HEK293T (ATCC, CRL-3216) were cultured in DMEM (Lonza). All media were supplemented with 10% fetal calf serum (Gibco) and 100 U/mL penicillin/streptomycin (Lonza).

To generate the expression construct HA-GFP-S8L-FKBP12 (GFP-S8L-F12) (Figure 1A, top), we amplified eGFP from the vector pCMV-GFP (addgene #11153) using the primer pair 5′-TATGATGGTACCTATGGTGAGCAAGGGCGAG-3′ and 5′-ATCATAGGATCCCTTGTACAGCTCGTCCAT-3′ introducing a 5′-KpnI and a 3′-BamHI restriction site. The amplified fragment was then digested with KpnI and BamHI and ligated to an annealed double-stranded SIINFEKL (Ova_257–264, S8L) oligo that additionally contained the 15 up- and downstream base pairs of the original sequence in ovalbumin, via a 3′-BamHI overhang. Next, GFP-S8L was ligated to annealed double-stranded HA oligos via the previously introduced KpnI restriction site at the 5′ end of GFP to generate the fragment HA-GFP-S8L. This fragment was again amplified by PCR using the primer pair 5′-CACCATGGCCTACCCCTACGAC-3′ and 5′-ATCATACTCGAGACTGGTCCATTCAGTCAG-3′ to generate a 3′-XhoI restriction site and a 5′-CACC-overhang required for Gateway cloning. Subsequently, FKBP12 (F36V) was amplified from pDONR-221-FKBP12^F36V (9) using the primer pair 5′-TATGATCTCGAGGGAGGCATGGGAGTGCAGGTGGAA-3′ and 5′-TCATTCCAGTTTTAGAAGCTCCACATC-3′ to introduce a 5′-XhoI restriction site. Finally, FKBP12 was ligated to HA-GFP-S8L via the 3′-XhoI site and cloned into the vector pENTR using the pENTR/D-TOPO-D-cloning kit (Invitrogen). The expression vectors based on pLex305 (constitutive, PGK promoter) and pCW57.1 (doxycycline-inducible) (addgene #41390, #41393) were generated using the Gateway LR clonase II enzyme mix (Invitrogen). The construct HA-GFP-S8L lacking the FKBP12 domain (ΔFKBP12) was amplified using the primer pair 5′-CACCATGGCCTACCCCTACGAC-3′ and 5′-TCAACTGGTCCATTCAGTCAGTT-3′ introducing a stop codon 3′ after the S8L sequence and the 5′CACC site for cloning into the vector pENTR and the expression vector pLex305 as described above.

HEK293T cells were seeded in 6-wells and transfected with 500 ng and 50 ng of the packaging plasmids psPAX2 and pCMV-VSV-G (Addgene #12260, #8454) and 500 ng of the expression vectors using Lipofectamine 2000 (Invitrogen). Supernatant was collected 24 and 48 h after transfection, pooled, sterile-filtered, and frozen at −80°C until use. Subsequently, cells were infected in full medium containing 8 µg/mL polybrene (Sigma) and centrifuged for 1.5 h at 30°C. 48 h after infection, cells were selected using 4 µg/mL puromycin (Sigma) for DC2.4 and BMC-2 cells and 6 µg/mL for MC57G cells. For all cell lines, single cell clones were generated by limiting dilution and analyzed for their GFP expression by flow cytometry.

For testing degradation of the antigen, 4 × 10⁵ BMC-2 cells were seeded into 6-well plates, incubated overnight and treated with indicated concentrations of dTAG-7 [synthesized as described before (7)] or 0.1% DMSO as a control for 18 h. Then, samples were collected in cold PBS before further analysis. For western blot analysis, samples were lysed in RIPA buffer [50 mM Tris-HCl (pH = 8), 150 mM NaCl, 1% NP-40, 0.5% Na-deoxycholate, 0.1% SDS] supplemented with Halt Protease and Phosphatase Inhibitor Cocktail (Thermo Fisher) for 30 min on ice. After centrifugation, the supernatant was harvested and protein concentration was quantified using Bradford reagent (Sigma) and a VersaMax ELISA plate reader (Molecular Devices). Subsequently, samples were mixed with reducing sample buffer (80 mM Tris-HCl, 10% glycerol, 2% SDS, 5% β-mercaptoethanol, bromphenolblue), heated at 95°C for 10 min, loaded onto SDS-PAGE Gels (4–15% Mini-Protean TGX Gels, Bio-Rad) and run at 120 V. Then samples were transferred onto a PVDF-membrane (Millipore) using a Bio-Rad blotting system, and membrane was blocked in 5% milk powder in PBS containing 0.05% Tween-20 (PBS-T) for 1 h. Membranes were stained with a mouse anti-HA (clone HA-7, Sigma) and a polyclonal rabbit anti-β-actin antibody (Sigma) in 2.5% milk powder in PBS-T overnight, washed with PBS-T, incubated with goat anti-mouse IRDye800CW and goat anti-rabbit IRDye680-labeled secondary antibodies (Li-Cor) for 1 h, and imaged with an Odyssey Infrared Imager (Li-Cor). Quantification of the blots was performed using ImageJ version 1.5.1. Uncropped images of western blots are displayed in Figure S1 in Supplementary Material.

In order to assess antigen stability, 4 × 10⁵ BMC-2 cells were seeded in 6-well plates, incubated overnight and then treated with 100 µg/mL of the translation inhibitor cycloheximide (Sigma), 1 µM dTAG-7 or 0.1% DMSO as a control. Samples were collected in cold PBS at indicated time points and protein levels were analyzed by western blot.

To investigate proteasome dependency of dTAG-7-mediated degradation, 4 × 10⁵ BMC-2 cells were seeded in 6-well plates, incubated overnight, and treated for 6 h in the presence or absence of 1 µM dTAG-7 and/or 10 µM of the proteasome inhibitor MG132 (Merck Chemicals) as indicated. Subsequently, cells were collected in cold PBS and protein levels were measured by western blot analysis.

To determine the toxicity of dTAG-7, 5 × 10³ BMC-2 cells were seeded in a 96-well plate and incubated with indicated concentrations of dTAG-7 (1 mM stock in DMSO) for 24 h or 10% DMSO as positive control for induction of cell death. Subsequently, the supernatant was removed by centrifugation and cells were resuspended in Cell Titer Glo reagent (Promega), according to the manufactures protocol. Luminescence was quantified using a GloMax Explorer (Promega), and values were normalized to a control treated with 0.1% DMSO, representing the highest concentration of dTAG-7.

To quantify H-2K^b/S8L presentation, 2 × 10⁵ cells were treated as indicated, resuspended in 50 µL FACS buffer (0.1% NaN₃, 0.5% BSA in PBS) containing a monoclonal anti-H-2K^b/S8L antibody, which specifically recognizes S8L (OVA_257–264) bound to the MHC class I molecule H-2K^b (25D1.16, APC-labeled, 1:160, eBioscience) and incubated on ice for 30 minutes. Cells were washed twice with 150 µL FACS buffer. Then, samples were resuspended in 200 µL FACS buffer and measured on a FACS Canto II (BD Biosciences). GFP-specific fluorescence was measured in parallel. Total MHC class I presentation was detected using anti-H-2K^b and anti-H-2D^b antibodies (1:200, PE-labeled, E-Bioscience), and the staining was performed for 20 min at 4°C. Using the FlowJo software (Version 7.6 (Treestar)), samples were gated on all living cells via forward and side scatter and background mean fluorescence intensities (MFI) (non-transduced or non-doxycycline treated cells, respectively, stained with the 25D1.16 antibody) were subtracted from all values. The impact of dTAG-7-mediated antigen degradation on GFP-specific fluorescence and H-2K^b/S8L presentation was tested by seeding 2 × 10⁵ cells in triplicates in 96-well plates. 1 µM dTAG-7 was added 30, 24, 18, 8, 6, 4, 2, 1.5, 1, and 0.5 h before collection to induce degradation of the fusion antigen. To exclude any influence of DMSO alone, cells were incubated with 0.1% DMSO, corresponding to the amount of DMSO in samples treated with 1 µM dTAG-7, for 30 h. After treatment with dTAG-7, cells were centrifuged and stained for flow cytometry as described above. To investigate the impact of newly synthesized antigen on H-2K^b/S8L presentation, 2 × 10⁵ cells were seeded in 96-well plates, treated with 1 µg/mL doxycycline (Sigma) and indicated concentrations of dTAG-7 for 3 and 6 h. Then cells were stained as mentioned above. The contribution of stable antigens to the pool of MHC class I substrates was calculated as % dTAG-7 dependent H-2K^b/S8L presentation = (1 − ((H-2K^b/S8L)_{no dTAG-7}/(H-2K^b/S8L)_{+1µM dTAG-7})) × 100.

4 × 10⁵ BMC-2 cells were treated with 1 µg/mL doxycycline to induce antigen expression. After 24 h, cells were washed twice with PBS, cultured again in full medium, and collected in QIAzol Lysis Reagent (Qiagen) at indicated time points. RNA was isolated using the recommended phenol-chloroform extraction protocol and quantified using a NanoDrop One Microvolume UV-Vis spectrophotometer (VWR). 1 µg of RNA was transcribed to cDNA using the RevertAid First Strand cDNA Kit (Thermo Fisher). Real-time PCR was performed using TaqMan Universal PCR Master Mix (Thermo Fisher) and the following primers: GFP: 5′-CTGCTGCCCGACAACCAC-3′, 5′-TGTGATCGCGCTTCTCGTT-3′, probe: 5′-FAM-ACCTGAGCACCCAGTCCGCCCT-TAMRA-3′; Eef1a1: 5′-GCAAAAACGACCCACCAATG-3′, 5′-GGCCTGGATGGTTCAGGATA-3′, probe: 5′-FAM-CACCTGAGCAGTGAAGCCAG-TAMRA-3′. The assay was run on a Quant Studio 5 Real-Time PCR System (Applied Biosystems) in triplicates and expression was normalized to the housekeeping gene Eef1a1 (eukaryotic translation elongation factor 1 alpha 1). To quantify the expression of inflammatory genes in cells treated with toll-like receptor (TLR) stimulants, 4 × 10⁵ cells were seeded in triplicates in 6-well plates. Treatment was performed by adding 20 ng/mL lipopolysaccharide (LPS) or 6 µg/mL poly(I:C) (both Invivogen) the next day. Samples were collected after 8 h in QIAzol Lysis Reagent (Qiagen), RNA was isolated and real-time PCR performed as mentioned above, using the TaqMan gene expression assays IL-6 (Mm00446190) and Mx1 (Mm00487796) (Thermo Fisher).

All results are shown as the mean ± SEM from experimental triplicates, unless indicated otherwise. Significance was calculated by student’s t-tests using GraphPad Prism 5 (ns, not significant; *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001).

In order to investigate the impact of dTAG-7-mediated antigen degradation on its MHC class I presentation, we generated a model system that would allow us (a) to monitor antigen expression and presentation and (b) to control antigen degradation using the small molecule dTAG-7. Therefore, we fused the well-studied H-2K^b-restricted T-cell epitope SIINFEKL (S8L) derived from chicken ovalbumin to the C-terminus of enhanced GFP that served as fluorescent reporter for flow cytometry. In addition, we added a mutant version of FKBP12 (F36V) to the C-terminus of GFP-S8L to act as a dTAG-7-dependent dTAG (7, 9) (Figure 1A, top, Figure 1B). Finally, an N-terminal hemagglutinin (HA) tag was added to allow detection by western blot. This construct was transduced into the C57BL/6-derived macrophage cell line BMC-2 (H-2^b) using a lentiviral delivery system to generate stable clones with constitutive expression. To demonstrate dTAG-7-dependent degradation of HA-GFP-S8L-FKBP12 (from here on called “GFP-S8L-F12”) in BMC-2 cells, we treated cells with increasing concentrations of dTAG-7 and measured GFP-S8L-F12 expression by western blot (Figure 1C). dTAG-7 triggered the degradation of GFP-S8L-F12 in a concentration-dependent manner, with minimal residual protein detected at 5 µM. A similar range of dTAG-7 activity has previously been observed in human cell lines (9) and importantly, dTAG-7 treatment had no influence on the expression of actin that served as control. Next, we performed a cycloheximide chase to demonstrate the decrease in antigen stability after dTAG-7 treatment (Figure 1D). From this experiment, we could conclude that GFP-S8L-F12 was intrinsically stable (~90% after 6 h) and that addition of dTAG-7 led to a reduction of its half-life (~20% after 6 h). Since western blot analysis is a measure of total protein (properly folded as well as misfolded), we further employed GFP-specific fluorescence as readout for the degradation of properly folded, mature GFP-S8L-F12 (Figures 1E,F). dTAG-7 triggered a reduction of the GFP-specific fluorescence in a concentration dependent manner, and samples treated with 1 µM dTAG-7 already reached values close to baseline. To further demonstrate the target specificity of dTAG-7, we additionally generated stable BMC-2 cells expressing a construct lacking the FKBP12 domain of GFP-S8L-F12 (ΔFKBP12) (Figure 1A, bottom). GFP-specific fluorescence of cells expressing ΔFKBP12 was not altered by increasing concentrations of dTAG-7, indicating that GFP-S8L-F12 degradation in this system is indeed FKBP12 dependent (Figures 1E,F). To further characterize our system, we aimed to confirm dTAG-7-specific degradation to be proteasome dependent. Therefore, we induced GFP-S8L-F12 degradation in the presence or absence of the proteasome inhibitor MG132 (Figure 1G). While treatment of cells with dTAG-7 alone lead to degradation of GFP-S8L-F12, inhibition of the proteasome reversed this effect, indicating that GFP-S8L-F12 degradation in this system was indeed proteasome dependent. To demonstrate that dTAG-7 treatment did not irreversibly affect GFP-S8L-F12 expression or the degradation machinery, we performed a wash-out experiment, in which cells were pre-treated with dTAG-7 for 18 h before the compound was removed and GFP-specific fluorescence was followed over time (Figure 1H). This experiment indicated that dTAG-7-induced degradation is completely reversible and already 8 h after wash-out of the compound GFP-S8L-F12 expression is restored to pretreatment levels. In line with this, dTAG-7 did not show cytotoxicity at the concentrations used in this study (Figure 1I). Together, these experiments demonstrate that we generated an expression system that could be used to study the impact of dTAG-7-mediated antigen degradation on its MHC class I presentation.

After having established the experimental setup, we next set out to test the impact of dTAG-7-mediated GFP-S8L-F12 degradation on its direct MHC class I presentation. We therefore treated GFP-S8L-F12-expressing BMC-2 cells with dTAG-7 and followed GFP-specific fluorescence, as well as S8L presentation using an H-2K^b/S8L-specific antibody, over time by flow cytometry (Figure 2A). In line with the experiments shown above, we could observe a strong decrease in GFP-specific fluorescence already during the first hours after dTAG-7 treatment. At the same time, a prominent increase in H-2K^b/S8L surface exposure could be detected with a peak at around 2 h after dTAG-7 treatment. To exclude any non-specific effects on the MHC class I presentation machinery, we performed the same experiment using cells expressing ΔFKBP12 (Figure 2B). Here, GFP-specific fluorescence as well as presentation of H-2K^b/S8L was comparable before and 2 h after dTAG-7 treatment, the time when maximal increase in H-2K^b/S8L presentation was observed using complete GFP-S8L-F12. In addition, we treated wild-type BMC-2 cells with dTAG-7 and measured the overall MHC class I surface expression (Figure S2 in Supplementary Material, left part). As expected, we did not observe any alterations in total H-2K^b and H-2D^b presentation after dTAG-7 treatment. These experiments indicated that the dTAG-7-dependent effect on MHC class I presentation was antigen-specific and FKBP12-dependent. Further, we also measured total MHC class I presentation of GFP-S8L-F12-expressing BMC-2 cells. While dTAG-7 treatment did not alter the total surface expression of H-2D^b, we observed a significant increase in the overall H-2K^b presentation on these cells (Figure S2 in Supplementary Material, right part). This demonstrates that epitope generation seems to be rate-limiting for MHC class I presentation under steady-state conditions and that the amount of H-2K^b/S8L presentation after dTAG-7-mediated degradation of GFP-S8L-F12 is strong enough to increase the total levels of H-2K^b. Next, we aimed to further confirm these findings in GFP-S8L-F12-expressing DC2.4 cells of dendritic cell origin and in the fibrosarcoma cell line MC57G (both H-2^b). In both, DC2.4 cells (Figure 2C) and MC57G cells (Figure 2D) addition of dTAG-7 induced a strong reduction of GFP-specific fluorescence within the first 2 h of treatment, indicating efficient dTAG-7-specific GFP-S8L-F12 degradation. In agreement with the results seen with BMC-2 cells, the simultaneous measurement of GFP-S8L-F12-derived surface expression of H-2K^b/S8L revealed that the dTAG-7-induced antigen degradation in DC2.4 and MC57G cells strongly increased the MHC class I presentation of S8L. Overall, we concluded from these experiments that dTAG-7-mediated antigen degradation enhanced MHC class I presentation of its target.

We next aimed to test the impact of dTAG-7-mediated antigen degradation in the absence of DRiPs and therefore chose to perform additional experiments in a doxycycline-inducible setting. Using the same lentiviral strategy, we generated stable cell lines expressing our model antigen GFP-S8L-F12 in a doxycycline-dependent manner. As DRiPs by definition have a very short half-life (<10 min) (18), doxycycline removal from GFP-S8L-F12-expressing cells should, therefore, depending on the residual stability of GFP-S8L-F12 mRNA, strongly reduce their presence. As a consequence, the previously accumulated pool of stable antigen would remain the only source to fuel the MHC class I presentation pathway. We, therefore, first induced GFP-S8L-F12 expression in BMC-2 cells to generate a pool of stable antigen and then removed doxycycline to stop its transcription. In order to determine the stability of GFP-S8L-F12 mRNA, we took samples at indicated time points after doxycycline withdrawal and performed a quantitative real-time PCR analysis (Figure 3A). According to these data, already 4 h after doxycycline withdrawal the amount of GFP-S8L-F12 mRNA was strongly reduced. Next, we induced expression of GFP-S8L-F12 in BMC-2 cells for 48 h, removed doxycycline for 4 h to reduce DRiP production, and then induced antigen degradation by dTAG-7 for 3 and 6 h (Figure 3B). Similar to the results observed before, dTAG-7-mediated antigen degradation was associated with a strong increase in H-2K^b/S8L presentation at both time points measured (Figure 3C). As before, these results could also be confirmed when using doxycycline-inducible MC57G cells as an independent validation (Figure 3D). We next quantified the percentage of H-2K^b/S8L presentation that was derived from dTAG-7-mediated GFP-S8L-F12 degradation. This quantification revealed that after DRiP-depletion, dTAG-7-mediated GFP-S8L-F12 degradation contributed around 70–80% to the overall H-2K^b/S8L presentation (Figure 3E). These results demonstrate the importance of stable antigens in serving as a substrate for MHC class I presentation in this system.

The previously observed increase in MHC class I-restricted GFP-S8L-F12 presentation upon dTAG-7 treatment originated from rapid degradation of a pool of stable antigen that accumulated in the cells over time. In order to further estimate the potential of DRiPs in our system, we next chose a setting that allowed DRiPs and mature antigens to contribute equally to the overall MHC class I presentation with regard of the time-of-synthesis ratio (23). We induced GFP-S8L-F12 expression in BMC-2 cells by addition of doxycycline and simultaneously started antigen degradation by adding increasing concentrations of dTAG-7 (Figure 4A). In this setting, it is important to highlight that degraded dTAG-7 substrates are by definition not converted into DRiPs, as they first need to acquire a mature conformation in order to being targeted. We quantified GFP-specific fluorescence and surface presentation of H-2K^b/S8L after 3 and 6 h (Figure 4B). Similar to the experiments with constitutive GFP-S8L-F12 expression and DRiP-depletion, we observed a strong dTAG-7-dependent increase in H-2K^b/S8L presentation, indicating that also without previous generation of an antigen pool, degradation of the mature, stable antigen contributed a major part to the overall MHC class I presentation of this antigen. These results could also be confirmed when using doxycycline-inducible MC57G cells (Figure 4C). To quantify the amount of H-2K^b/S8L presentation derived from dTAG-7-mediated antigen degradation, we again calculated the percentage of dTAG-7-dependent H-2K^b/S8L presentation, as above (Figure 4D). Interestingly, using this experimental setting, we consistently observed slightly lower percentages for the contribution of dTAG-7-mediated degradation of the mature, stable GFP-S8L-F12 of around 60–70% of the overall presented peptides. This number was around 10% lower than compared to the previous quantification after DRiPs depletion and we conclude from these experiments that this difference in H-2K^b/S8L presentation might reflect the contribution of DRiPs in our system.

Therapies based on bifunctional degraders aim to interfere with critical mechanisms of diseases, including cancer. Often, such mechanisms develop in a microenvironment with pronounced inflammation and immune cell activation. To test whether the enhancing effect of dTAG-7 treatment on MHC class I presentation can also be seen under inflammatory conditions, we treated GFP-S8L-F12-expressing cells with different TLR agonists before inducing dTAG-7-mediated antigen degradation. Since the contribution of DRiPs to direct MHC class I presentation has been thought to be of special importance during cellular infections (18, 28), these experiments also tested the role of DRiPs under inflammatory conditions in our model. First, we investigated whether our antigen-expressing cell lines could be induced by TLR stimulation. Therefore, we measured the expression of inflammatory genes in BMC-2 and MC57G cells in response to the TLR4 agonist LPS, and the TLR3 agonist and viral RNA mimic poly(I:C) (Figures 5A,B). Indeed, we could show that TLR-stimulated cells significantly induced the expression of the NFκB target gene Il-6, as well as the interferon-stimulated gene Mx1. We next pretreated BMC-2 cells with TLR agonists for 24 h to induce the expression of inflammatory genes and then cultured cells in the presence or absence of dTAG-7 (Figure 5C). Similar to the experiments performed before, we could observe pronounced dTAG-7-mediated GFP-S8L-F12 degradation as measured by reduction in GFP-specific fluorescence. Interestingly, TLR stimulation did not reduce the potency of dTAG-7 to increase H-2K^b/S8L presentation, as expected if a DRiP-associated pathway would show a more dominant role under these conditions. In fact, TLR stimulation rather increased H-2K^b/S8L presentation after dTAG-7 treatment, probably due to general upregulation of the MHC class I presentation machinery. The same results could also be observed using antigen-expressing MC57G cells (Figure 5D). To mimic the role of DRiPs shortly after infection, we performed experiments in which TLR stimulation and induction of GFP-S8L-F12 expression were performed simultaneously. Also in this setting, for BMC-2 cells (Figure 5E), as well as for MC57G cells (Figure 5F), we did not observe a reduction in dTAG-7-induced H-2K^b/S8L presentation 6 h after TLR stimulation. From these experiments, we conclude that the observed effect of dTAG-7 in boosting MHC class I presentation of its target works independently of the inflammatory status of the target cell that we induced in this study. At the same time, our results demonstrate that stimulation of antigen-expressing cells with LPS or poly(I:C) did not increase the contribution of DRiPs to the H-2K^b/S8L presentation.