Development of NanoLuc-targeting protein degraders and a universal reporter system to benchmark tag-targeted degradation platforms

Development of NanoTACs; a NanoLuc-targeting degrader system

We developed heterobifunctional degraders that bind NanoLuc (NanoTACs) to trigger protein degradation of NanoLuc-tagged substrates building upon the existing tTPD systems that direct degraders to FKBP^F36V or Halo-tagged substrates (Fig. 1a). While each tTPD system has specific advantages, NanoLuc is the only tag that offers luminescence properties. Additionally, NanoTACs are catalytic and can act as degradation catalysts, unlike HaloPROTACs (Fig. 1b). Our most potent NanoTAC recruits the CRL4^CRBN E3-ligase complex to trigger the degradation of NanoLuc-tagged substrate proteins (Fig. 1c). To generate this NanoTAC (NC4) (Fig. 1d), we synthesised two series of NanoTAC analogues. NC1 and NC2 were generated by linking NanoLuc inhibitor 1, a small-molecule inhibitor of NanoLuc with a low nanomolar IC₅₀ (31 nM) against the enzymatic activity of NanoLuc²⁷ (Supplementary Fig. 1a), with the CRBN ligand thalidomide using varied alkyl chain linker lengths (Supplementary Fig. 1b, c). NC3, NC4, NC5 and the structurally related control analogue unable to recruit CRBN, NC*, were prepared by coupling a derivative of the more potent NanoLuc inhibitor 2, with a single-digit nanomolar IC₅₀ (4.2 nM) against the enzymatic activity of NanoLuc²⁷ (Supplementary Fig. 1d) with the CRBN ligand thalidomide using varied linker lengths (Supplementary Fig. 1e-g). Lastly, NV1 and NV2 were prepared by coupling NanoLuc inhibitor 2 to the widely used VHL ligand derivatives: VH298 and VH032^18,19 (Supplementary Fig. 1h, i).

Characterisation of NanoTACs, and selection of a potent NanoTAC: NC4

To test the degradation capacity of the first series of analogues we cloned and stably expressed through lentiviral transduction a versatile and universal tTPD reporter protein that enabled comparative studies in HEK293T cells. This reporter contains all three protein tags: Halo, FKBP^F36V and NanoLuc in addition to the fluorescent protein EGFP to create the artificial recombinant fusion protein Halo-EGFP-NanoLuc-FKBP^F36V (H-E-N-F) (Supplementary Fig. 2a). All graphs in this study have been colour-coded corresponding to the targeted tag to ease interpretation. The F36V-CRBN degrader FC1 almost completely degraded the fusion protein at a concentration of 100 nM as judged by Western blot while the NanoLuc-CRBN targeting NanoTACs NC1 and NC2 led to partial degradation of the fusion protein at 500 nM and 1 µM (Supplementary Fig. 2b).

We next tested the second series of NanoTAC analogues. To explore whether NanoTAC degraders could be improved by recruiting a different E3 ligase, we employed NV1 and NV2 that hijack the CRL2^VHL complex (Supplementary Fig. 2c). NV1 and NV2 were unable to trigger degradation of the H-E-N-F fusion protein over the wide range of tested concentrations up to 16 μM (Supplementary Fig. 2d, e). The lack of degradation was not due to an absence of VHL activity as the F36V-VHL targeting degrader (FV1) triggered complete degradation of the fusion protein by Western blot (Supplementary Fig. 2d, e).

Luminescence is a higher throughput readout than Western blotting, however, NanoTAC degraders compete with NanoLuc substrates as they interfere with the conversion of furimazine to furimamide (Supplementary Fig. 2f). To address this, we generated a fusion protein that incorporates Firefly: Halo-Firefly-NanoLuc-FKBP^F36V (H-FF-N-F) (Fig. 2a). This construct allows for the direct side-by-side comparison of all tTPD systems including the NanoLuc tTPD system using a Firefly luciferase, which emits a luminescence signal that can be assayed independently from NanoLuc through the use of distinct substrates. Using the H-FF-N-F reporter protein we tested the second series of NanoLuc-CRBN targeting NanoTAC analogues NC3, NC4, NC5 and NC* (inactive control). We observed a greater reduction in Firefly luminescence, and protein levels by Western blot, upon NanoTAC treatment of cells expressing the H-FF-N-F fusion protein (Fig. 2b, c). Specifically, NC3 and NC4 showed a marked improvement in the extent of substrate degradation compared to the first series of analogues NC1 and NC2, particularly at lower concentrations (Fig. 2b, c). At higher concentrations, NC4 demonstrated the hook effect (Fig. 2c), which is a characteristic feature of heterobifunctional degraders whereby high concentrations block degradation due to ligand saturation and inability to form a ternary complex²⁸. Interestingly, NC5 which harbours a linker that is just one carbon length shorter than NC4 was incapable of triggering degradation of the fusion protein, possibly due to unfavourable steric interactions as a result of the decreased linker length (Fig. 2b, c). As expected, the control compound NC* did not trigger degradation of the fusion protein (Fig. 2b, c). NanoTAC degraders were equally capable of triggering degradation of the H-E-N-F substrate protein with NC3 and NC4 consistently outperforming the first analogue NC1 (Fig. 2d). Overall, NC4 displayed the most effective degradation profile and was therefore characterised further.

a Schematic depicting the Halo-Firefly-NanoLuc-FKBP^F36V (H-FF-N-F) fusion protein. b Western blot analysis of lysates from HEK293T cells expressing H-FF-N-F, induced with 20 ng/mL doxycycline overnight, stimulated for 5 h (ns non-specific band). c Firefly luminescence from cells in b, induced with 20 ng/mL doxycycline overnight, stimulated for 5 h. d Western blot analysis of lysates from HEK293T cells expressing H-E-N-F, induced with 20 ng/mL doxycycline overnight, stimulated for 5 h. e Western blot analysis of lysates from cells in d, induced with 20 ng/mL doxycycline overnight, stimulated with 125 nM of NC4 for the indicated times. f Western blot analysis of lysates from immortalized MDFs (iMDFs) constitutively expressing H-E-N-F or wild-type iMDFs (ctrl), cells were left untreated (Unt), stimulated with DMSO or 125 nM of NC4 for 5, 24, 48 or 72 h. NC4 was left in the culture medium for the duration of the time-course or removed from the culture medium at 5 h, and cells incubated post degrader removal for 24, 48 or 72 h. g Western blot analysis of lysates from cells in d, induced with 20 ng/mL doxycycline overnight, stimulated with NC4 for 5 h. h Volcano plot quantifying proteins significantly downregulated in 293 T cells expressing the H-FF-N-F reporter after treatment with DMSO or 125 nM of NC4 for 5 h. Proteins significantly downregulated (using the q-value) with NC4 treatment when compared to DMSO treatment are labelled and indicated with an arrow. i NanoLuc luminescence from cells in b, induced with 20 ng/mL doxycycline overnight, stimulated with 10 μM MG132 or 1 μM MLN4924 for 1 h prior to treatment with 15.6 nM NC4 or NC* for 4 h. Substrate (furimazine) was diluted 1:100, 1:70 and 1:50 prior to detecting luminescence. j Western blot analysis from cells in B, induced with 20 ng/mL doxycycline overnight, stimulated with 10 μM MG132, or 1 or 5 μM MLN4924 for 1 h prior to stimulation with DMSO or 125 nM of NC4 or NC* for 5 h. c, i Error bars (EB) represent mean ± SD from N = 4 technical repeats. All experiments were repeated independently three times and a representative Fig. is shown. Source data are provided as a Source Data file.

NC4 exhibited approximately a 30-fold reduced inhibition of NanoLuc enzymatic activity compared to the parent inhibitor (NanoLuc inhibitor 2) indicating that conjugation of the CRBN ligand altered the binding kinetics of NC4 (Supplementary Fig. 2g). Strikingly, even though NC4 exhibited reduced inhibition of NanoLuc enzymatic activity NC4 was still capable of rapidly reducing protein levels within 2 h and degradation was sustained over a period of 10 h (Fig. 2e). Further to this, degradation by NC4 was sustained over 24 h when stimulation was maintained in the culture medium, and reversible by 24 h when stimulation was removed after 5 h of initial treatment (Fig. 2f). NC4 was also capable of triggering degradation of the fusion protein at low concentrations, ~30 nM, with a slight hook effect observed at higher concentrations, similar to that observed in Firefly luminescence-based readouts (Fig. 2c). Importantly, the global proteomic analysis demonstrated that NC4 triggers specific degradation of the target substrate with no significant off-target degradation observed when assayed against 5591 proteins (Fig. 2h). Notably, our proteomics analysis did not detect the Thalidomide-induced CRBN neosubstrates, IKZF1 or IKZF3, and therefore we cannot comment on whether these neosubstrates are altered upon NC4 treatment.

To explore whether NanoLuc luminescence could be used in conjunction with NanoTAC treatment we titrated NC4 against NC* from 2–10 h. We reasoned that the catalytic nature of NC4 would allow for substrate degradation at concentrations below that required to inhibit NanoLuc’s enzymatic luminescence activity. To this extent we compared the ability of NC4 to reduce luminescence with the control compound, NC*, which lacks the degradative capability. As predicted, we observed clear differences in NanoLuc luminescence at 31.2 nM and 15.6 nM between NC4 and NC* (Supplementary Fig. 2h). Importantly, the reduction in NanoLuc luminescence and the observed depletion of the reporter by Western blot was indicative of cullin-based proteasome-mediated degradation, as pre-treatment with the proteasome inhibitor MG132 or the NEDD8-activating enzyme inhibitor MLN4924 could rescue both the reduction in luminescence and protein levels (Fig. 2i, j). Notably, a slightly enhanced rescue was observed when the substrate furimazine concentration was increased for NanoLuc detection, consistent with the competitive nature of NanoLuc inhibition by NC4 (Fig. 2i).

FKBP^F36V-targeting degraders outperform other tTPD systems

To benchmark existing tTPD systems against NanoTACs and assess the relative degradative performance of each system, we synthesised degraders that hijack VHL and CRBN to trigger proteasomal degradation of FKBP^F36V and Halo-tagged substrates^9,10,18,19, and generated a series of Halo targeting CBRN recruiting degraders; HC1, HC2 and HC3 (Supplementary Fig. 3a). Halo-VHL degraders (HV1 and HV2) target Halo-tagged substrates¹⁹, while both F36V-CRBN (FC1 and FC2) degraders and the F36V-VHL (FV1) degrader target FKBP^F36V-tagged substrates^9,10 (Fig. 3a). Notably, similar Halo-CRBN molecules have been tested for cellular penetration using the Chloroalkane Penetration Assay (CAPA)²⁹; however, degradation studies have not been performed. The Halo-CRBN molecules we generated were unable to trigger efficient substrate degradation over a wide concentration range (Supplementary Fig. 3b) possibly due to reduced cellular accumulation of these compounds or the inability to promote ternary complex formation.

a Schematic representation of the Halo-VHL, F36V-CRBN, F36V-VHL and NanoLuc-CRBN tag-targeting degrader systems. b Firefly luminescence from 293 T cells stably expressing doxycycline-inducible H-FF-N-F, treated with 20 ng/mL doxycycline overnight (induced), then stimulated with the indicated concentrations FC1 (F36V-CRBN), FC2 (F36V-CRBN), FV1 (F36V-VHL), HV1 (Halo-VHL), HV2 (Halo-VHL) or NC4 (NanoLuc-CRBN) for 5 h. c Firefly luminescence from cells used in b, treated with 20 ng/mL doxycycline overnight (induced), then stimulated with the indicated concentrations FC1, FC2, FV1, HV1, HV2 or NC4 for the indicated times. d Western blot analysis of total cell lysates from cells expressing H-FF-N-F, treated with 20 ng/mL doxycycline overnight (induced), then stimulated with the indicated concentrations of degraders for 5 h. e Firefly luminescence from cells in b, treated with 20 ng/mL doxycycline overnight (induced), following treatment with 20 μM MG132 for 1 h prior to stimulation with the indicated concentrations FC1, FC2, FV1, HV1, HV2 or NC4 for 5 h. b, c, e EB represent mean ± SD from N = 4 technical repeats. All experiments were repeated independently three times, and a representative Fig. is shown. Source data are provided as a Source Data file.

To directly compare the NanoTAC system to the FKBP^F36V and Halo systems (Fig. 3a) we employed cells expressing either the H-FF-N-F or the H-E-N-F fusion proteins (Fig. 2a and Supplementary Fig. 2a). Using Firefly luminescence, we were able to directly compare the degradative capacity of the F36V-CRBN, F36V-VHL, Halo-VHL and NanoLuc-CRBN systems against the same substrate. Overall, F36V-CRBN degraders (FC1 and FC2) outperformed the Halo-VHL degraders (HV1 and HV2), while the F36V-VHL degrader (FV1) triggered the most efficient substrate depletion over a wide concentration range after 5 h (Fig. 3b). A similar degradation profile was observed from the corresponding NanoLuc luminescence detected from the H-FF-N-F and H-E-N-F substrates in the same assay (Supplementary Fig. 3c, d). Halo-VHL degraders, although less efficient at triggering substrate depletion compared to the F36V-CRBN and F36V-VHL degraders, particularly at lower concentrations, displayed slightly improved substrate degradation over 8 and 24 h at relatively low concentrations (Fig. 3c, Supplementary Fig. 3e). The NanoLuc-CRBN targeting NanoTAC, NC4, was capable of triggering degradation of the fusion protein over a wide concentration range and to levels comparable to that of the Halo and FKBP^F36V tTPD systems displaying a calculated D_max of 77% and a DC₅₀ of 21.6 nM against our fusion (Fig. 3b, c). NC4, displayed a similar degradation profile to that of the HaloPROTACs (Fig. 3b), but reproducibly appeared to lose efficacy after 24 h (Fig. 3c). Consistent with our luminescence results, we observed reduced levels of the fusion proteins with all tested tTPD degrader systems by Western blot (Fig. 3d, Supplementary Fig. 3f), and importantly the observed reduction in luminescence could be rescued by co-treatment with the proteasome inhibitor MG132 (Fig. 3e, Supplementary Fig. 3g). Next we compared the durability of degradation across all tTPD systems by performing wash-out recovery experiments. We transitioned to a constitutive system where the H-E-N-F protein was expressed under the control of a ubiquitin promoter to avoid foreseeable complications with wash-out of the doxycycline in our inducible systems. Interestingly FC1 and FV1 maintained near complete degradation up to 48 h (Fig. 4a), and were capable of maintaining clear degradation for up to 48 h when the degrader was removed from the culture medium at 5 h post-treatment (Fig. 4b). FC2 and NC4 displayed very similar durability profiles with clear degradation seen up to 24 h (Fig. 4a), however, this was not maintained when the degrader was removed from the culture medium after 24 h (Fig. 4b). HaloPROTACs HV1 and HV2 were less durable in their response with depletion not maintained upon removal of the degrader from the culture medium and overall slower degradation kinetics (Fig. 4a, b).

a Western blot analysis of total cell lysates from immortalized MDFs (iMDFs) constitutively expressing H-E-N-F. Cells were stimulated with DMSO or stimulated with 500 nM of FC1, FC2, FV1, HV1, HV2 or NC4 degrader compounds for 5, 24 or 48 h. Asterisk (*) indicates possible cleavage product. b Western blot analysis of total cell lysates from immortalized MDFs (iMDFs) constitutively expressing H-E-N-F. Cells were stimulated with DMSO or stimulated with 500 nM of FC1, FC2, FV1, HV1, HV2 or NC4 degrader compounds for the first 5 h. Degraders were then removed from the culture medium at 5 h and cells were incubated for 24 or 48 h. Minus sign (−) indicates that sample was harvested at the time the treatment was removed (5 h). Asterisk (*) indicates possible cleavage product. c Schematic depicting the comparison of selected tTPD systems; F36V-VHL (FV1) and F36V-CRBN (FC1). d Western blot analysis of total cell lysates from MLKL^−/− HT29 cells stably expressing the doxycycline-inducible C-terminal fusion protein MLKL-FKBP^F36V. Cells were treated with 40 ng/mL doxycycline overnight to induce the constructs, then stimulated with the indicated concentrations of FV1 (F36V-VHL) or FC1 (F36V-CRBN) for 5 h. e Schematic depicting the comparison of selected tTPD systems; F36V-VHL (FV1) and Halo-VHL (HV2). f Western blot analysis of total cell lysates from MLKL^−/− HT29 cells stably expressing the doxycycline-inducible C-terminal fusion protein MLKL-FKBP^F36V or MLKL-Halo treated with 40 ng/mL doxycycline overnight to induce the construct, then stimulated with the indicated concentrations of FV1 (F36V-VHL) or HV2 (Halo-VHL). All experiments were repeated independently three times, and a representative Fig. is shown. Source data are provided as a Source Data file.

To functionally assess the impact of the different tTPDs on a model substrate, we selected the well-characterised pro-necroptotic pseudokinase MLKL. MLKL is the terminal effector protein for necroptotic cell death³⁰ which can be induced by the combined treatment of TNF, Smac mimetic (SM); compound A³¹ and z-VAD-fmk (TSZ)³². C-terminally tagged MLKL has previously been shown to be fully functional³³ and MLKL knockout cells cannot undergo necroptosis unless reconstituted with an active MLKL construct³². We therefore generated C-terminal fusion proteins of MLKL containing either FKBP^F36V, Halo or NanoLuc tags and stably expressed doxycycline-inducible versions of these proteins in MLKL^−/− HT29 cells that were generated by CRISPR/Cas9 targeting. The HT29 cell line is widely used and can undergo rapid necroptosis when stimulated with TSZ³². MLKL-NanoLuc expression was independently confirmed by NanoLuc luminescence (Supplementary Fig. 4a). The tagged versions of MLKL displayed reduced expression compared to untagged MLKL in MLKL^−/− HT29 cells, with the NanoLuc tag leading to a strong reduction in expression compared to adding a Halo or FKBP^F36V tag (Supplementary Fig. 4b). From this data it is unclear to what extent this will limit the application of the NanoLuc tag to other targets, and what causes the reduced expression. Nevertheless, the expression level of MLKL-NanoLuc was sufficient as MLKL-Halo, MLKL-FKBP^F36V and MLKL-NanoLuc could confer comparable sensitivity to necroptosis (Supplementary Fig. 4c).

The tagged MLKL substrates afforded us the opportunity to directly compare the efficiency of the CRBN and VHL degrader systems against a biologically-relevant substrate with the tags positioned at the same terminus (Fig. 4c). Interestingly the FV1 degrader that recruits F36V-VHL was faster at triggering degradation of MLKL compared to the F36V-CRBN degrader FC1, acting within 1 h (Supplementary Fig. 4d). FV1 was also more effective at triggering degradation of MLKL-FKBP^F36V compared to the F36V-CRBN degrader FC1 resulting in no detectable protein at 5 h at low nM concentrations (Fig. 4d). To directly compare whether the FKBP^F36V or Halo epitope tags could influence the degradation of tagged MLKL with the same CRL system we tested the F36V-VHL and Halo-VHL systems side-by-side (Fig. 4e). Consistent with our previous results we observed that the F36V-VHL-recruiting degrader FV1 was the most effective compound (Fig. 4f). Interestingly, we observed comparable degradation of murine MLKL-FKBP^F36V reconstituted into Mlkl^−/− mouse dermal fibroblasts (MDFs) with F36V-CRBN degraders (FC1, FC2) and the F36V-VHL degrader (FV1) (Supplementary Fig. 4e), suggesting that even slight changes in the substrate or cell type might influence the effectivity of degrader compounds. To explore this further, we inserted each tag into the genomic locus of the FBL gene that encodes for the protein Fibrillarin, which is exclusively expressed in the nucleolus. We performed concentration titrations and time courses on Fibrillarin-NanoLuc, Fibrillarin-Halo and Fibrillarin-FKBP^F36V expressing cell lines and calculated D_max values based on densitometry measurements (Supplementary Fig. 4f–m). Consistent with our results for MLKL and our fusion protein, we observed that FV1 was superior in degradation capacity compared to the other tag degraders displaying a D_max of ~95.9% after 10 h. FC1, HV1, HV2 and NC4 all demonstrated clear degradation of FBL over 10 h with calculated D_max values of ~70%, ~93.2%, ~94.8% and ~52.9%, respectively. Fibrillarin-NanoLuc predominantly localises to the nucleus (Supplementary Fig. 5a) suggesting that NC4 was capable of degrading substrates that primarily reside in the nucleus. To determine whether substrates in other cellular compartments could also be targeted by NC4 we generated a fusion construct containing NanoLuc and the plasma membrane transporter protein SLC38A2. We confirmed the location of NanoLuc-SLC38A2 in the membrane compartment (Supplementary Fig. 5b), and consistent with our previous results we observed a reduction in Nanoluc-SLC38A2 upon treatment with NC4 (Supplementary Fig. 5c).

tTPD systems trigger physiologically relevant degradation

To compare the biological activity of tTPD-mediated degradation across all tTPD platforms, we assessed whether our most potent NanoTAC could degrade substrates to a level that prevents necroptosis to achieve a response similar to protein knockout. Consistent with our previous results, NanoTAC degraders triggered efficient degradation of MLKL-NanoLuc, with NC4 displaying the most effective degradation profile (Supplementary Fig. 5d). Similar to the degradation of the Halo-EGFP-NanoLuc-FKBP^F36V fusion protein, NC4 triggered degradation of MLKL-NanoLuc within 2 h with sustained degradation over 10 h (Supplementary Fig. 5e). Impressively, NC4-induced degradation of MLKL-NanoLuc was sufficient to reduce MLKL-NanoLuc-driven necroptosis in HT29 cells in a dose-dependent manner (Fig. 5a), and importantly, NanoTAC degraders displayed no anti-necroptotic activity in MLKL^−/− cells expressing untagged MLKL (Fig. 5a) or wild-type HT29 cells (Supplementary Fig. 5f). F36V-CRBN targeting degraders (FC1, FC2) were capable of reducing necroptosis driven by MLKL-FKBP^F36V in MLKL^−/− HT29 cells in a dose-dependent manner, sustaining cell survival over 24 h at 62.5 and 15.6 nM, respectively (Fig. 5b, c). The F36V-VHL targeting degrader FV1 was extremely efficient at preventing necroptosis driven by MLKL-FKBP^F36V demonstrating a sustained and prolonged reduction in cell death at 3.9 nM over 24 h (Fig. 5d). HaloPROTACs (HV1 and HV2), although capable of reducing necroptosis driven by Halo-MLKL in MLKL^−/− HT29 cells, required much higher concentrations to maintain cell survival; 250 and 125 nM, respectively (Fig. 5e, f). Interestingly, HaloPROTACs also exhibited steep reductions in cell survival at lower concentrations, which may be attributed to the non-catalytic nature of these degrader compounds (Fig. 5e, f). Like NanoTAC degraders, FKBP^F36V degraders and HaloPROTACs displayed no anti-necroptotic activity in MLKL^−/− cells expressing untagged MLKL (Figs 5b–f) or wild-type HT29 cells (Supplementary Fig. 5g, h). A representative flow cytometry gating strategy that was used to analyse the induction of necroptosis is provided (Supplementary Fig. 5i). To test whether NanoTACs could be combined with other tTPD systems to modulate the levels of multiple proteins, we deleted endogenous caspase-8 via CRISPR/Cas9 targeting in HT29 cells expressing MLKL-NLuc, and re-expressed caspase-8-FKBP^F36V. Both caspase-8-FKBP^F36V and MLKL-NLuc could be depleted by FV1 and NC4 treatment in these cells, respectively (Fig. 5g). TNF and Smac mimetic (TNF + SM) treatment triggers caspase-8 dependent apoptotic cell death that can be switched to necroptotic cell death when caspase-8 is inhibited or deleted from cells. Induction of MLKL-NLuc with doxycycline and constitutive expression of caspase-8-FKBP^F36V rendered this cell line susceptible to necroptotic cell death and apoptotic cell death upon treatment with TNF + SM (Fig. 5h). As expected, removal of MLKL-NLuc with NC4 did not impact cell death induced by TNF + SM treatment (TNF + SM + NC4), owing to active apoptotic caspase-8 signalling. Impressively, removal of caspase-8-FKBP^F36V with FV1 (TNF + SM + FV1) caused a clear reduction in TNF + SM induced cell death that could be further reduced with MLKL-NLuc removal with NC4, but not NC*. These data indicate that TNF + SM induced apoptosis and necroptosis driven by caspase-8-FKBP^F36V and MLKL-NLuc can be dynamically fine-tuned in the same cell system by combining two different tTPD degraders (Fig. 5h).

a MLKL^−/− HT29 cells stably expressing the doxycycline-inducible untagged MLKL or the C-terminal fusion protein MLKL- NanoLuc were induced with 40 ng/mL doxycycline overnight, then stimulated with NC4 (NanoLuc-CRBN) for 5 h before the addition of TNF (100 ng/mL) + Smac mimetic (compound A; 500 nM) + caspase inhibitor; z-VAD-fmk (10 μM), for 24 h. Cell death was assessed by flow cytometric analysis of PI exclusion. N = 3 independent experiments (symbols), EB represent mean ± SD. b–d MLKL^−/− HT29 cells expressing doxycycline-inducible untagged MLKL or the C-terminal fusion protein MLKL-FKBP^F36V were induced with 40 ng/mL doxycycline overnight, then stimulated with FC1, FC2 or FV1 for 5 h before the addition of TNF (100 ng/mL) + Smac mimetic (compound A; 500 nM) + caspase inhibitor; z-VAD-fmk (10 μM), for 24 h. Cell death was assessed by flow cytometric analysis of PI exclusion. N = 3 independent experiments (symbols), EB represent mean ± SD. e, f MLKL^−/− HT29 cells expressing doxycycline-inducible untagged MLKL or the C-terminal fusion protein MLKL-Halo were treated with 40 ng/mL doxycycline overnight to induce the constructs, then stimulated with the indicated concentrations HV1 or HV2 for 5 h before the addition of TNF (100 ng/mL) + Smac mimetic (compound A; 500 nM) + caspase inhibitor; z-VAD-fmk (10 μM), for 24 h. Cell death was assessed by flow cytometric analysis of PI exclusion. e N = 4, f N = 3 independent experiments (symbols), EB represent mean ± SD. g Western from MLKL^−/−CASP8^−/− HT29 cells stably expressing the doxycycline-inducible C-terminal fusion protein, MLKL-NanoLuc (MLKL-NLuc) and constitutively expressing the C-terminal fusion protein caspase-8-FKBP^F36V. Cells were induced with 40 ng/mL doxycycline overnight, then stimulated with FV1 and NC4 for 5 h. h Cells from g were treated with 40 ng/mL doxycycline and stimulated with 125 nM of FV1, NC4, or NC* for 5 h before the addition of TNF (100 ng/mL) + Smac mimetic (compound A; 500 nM) for 19.5 h. Cell death was assessed using an IncuCyte and Cytotox-Red uptake. N = 3 independent experiments, one representative experiment shown, EB represent mean ± SD. Source data are provided as a Source Data file.

Collectively, these results suggest that efficient degradation of substrates with tag-targeting degraders is system dependent. Importantly, these results highlight that it is critical to compare each tTPD system against each substrate to identify the optimal degrader/tag pair to achieve prolonged and sustained degradation with minimal effective degrader concentration.