Targeted degradation of zDHHC-PATs decreases substrate S-palmitoylation

Abstract

Reversible S-palmitoylation of protein cysteines, catalysed by a family of integral membrane zDHHC-motif containing palmitoyl acyl transferases (zDHHC-PATs), controls the localisation, activity, and interactions of numerous integral and peripheral membrane proteins. There are compelling reasons to want to inhibit the activity of individual zDHHC-PATs in both the laboratory and the clinic, but the specificity of existing tools is poor. Given the extensive conservation of the zDHHC-PAT active site, development of isoform-specific competitive inhibitors is highly challenging. We therefore hypothesised that proteolysis-targeting chimaeras (PROTACs) may offer greater specificity to target this class of enzymes. In proof-of-principle experiments we engineered cell lines expressing tetracycline-inducible Halo-tagged zDHHC5 or zDHHC20, and evaluated the impact of Halo-PROTACs on zDHHC-PAT expression and substrate palmitoylation. In HEK-derived FT-293 cells, Halo-zDHHC5 degradation significantly decreased palmitoylation of its substrate phospholemman, and Halo-zDHHC20 degradation significantly diminished palmitoylation of its substrate IFITM3, but not of the SARS-CoV-2 spike protein. In contrast, in a second kidney derived cell line, Vero E6, Halo-zDHHC20 degradation did not alter palmitoylation of either IFITM3 or SARS-CoV-2 spike. We conclude from these experiments that PROTAC-mediated targeting of zDHHC-PATs to decrease substrate palmitoylation is feasible. However, given the well-established degeneracy in the zDHHC-PAT family, in some settings the activity of non-targeted zDHHC-PATs may substitute and preserve substrate palmitoylation.

Fig 1. Subcellular location and expression of Halo-zDHHC20 and Halo-zDHHC5 in Flp-In 293 T-REx cells.

A–Subcellular location of Halo-zDHHC5 visualised by staining cells with TAMRA-chloroalkane either without (uninduced) or with (induced) gene induction using tetracycline. B–Subcellular location of Halo-zDHHC20 visualised by staining cells with TAMRA-chloroalkane either without (uninduced) or with (induced) gene induction using tetracycline. Scale bar: 10μm. C–Western blot analysis confirming successful induction of Halo-zDHHC5 (left) and Halo-zDHHC20 (right).

Fig 2. Halo-PROTAC mediated zDHHC-PAT degradation in Flp-In 293 T-REx cells.

A–Halo-PROTAC structures B–Dose response relationship for Halo-PROTACs in Flp-In 293 T-REx cells expressing Halo-zDHHC5. Cells were treated with the indicated concentration of Halo-PROTAC for 18–24 hours and lysates immunoblotted as shown. V: vehicle (DMSO) control. The bar chart (right) shows the expression of Halo-zDHHC5 relative to GAPDH. *: P<0.05, **: P<0.01, one-way ANOVA followed by Dunnett’s multiple comparisons test. C–Dose response relationship for Halo-PROTACs in Flp-In 293 T-REx cells expressing Halo-zDHHC20. Cells were treated with the indicated concentration of Halo-PROTAC for 18–24 hours and lysates immunoblotted as shown. V: vehicle (DMSO) control. The bar chart (right) shows the expression of Halo-zDHHC20 relative to GAPDH. *: P<0.05, one-way ANOVA followed by Dunnett’s multiple comparisons test. D–Impact of compound 1 on abundance of endogenous zDHHC5 in Flp-In 293 T-REx cells.

Fig 3. Halo-PROTAC mediated zDHHC-PAT degradation requires recruitment of VHL, elongation of ubiquitin chains, and proteasome activity.

A–An epimer of compound 1 which does not recruit VHL does not induce degradation of Halo-zDHHC5 (left) or Halo-zDHHC20 (right). The bar charts show the expression of each Halo-zDHHC-PAT relative to GAPDH. **: P<0.01, one-way ANOVA followed by Tukey’s multiple comparisons test. B–Inhibition of the protease using Mg-132 (5μM) or inhibition of NEDD8 activating enzyme using MLN4924 (10μM) prevent degradation of Halo-zDHHC5 (left) or Halo-zDHHC20 (right) by compound 1. **: P<0.01, one-way ANOVA followed by Tukey’s multiple comparisons test. C–Immunoprecipitation experiments confirm the incorporation of ubiquitin chains into Halo-zDHHC20 induced by treatment with compound 1. Flp-In 293 T-REx cells in which zDHHC20 expression was induced or not with tetracycline (± Tet) were transfected with HA-ubiquitin and treated with compound (1μM) alone or in combination with Mg-132 (5μM). Whole cell lysates (unfractionated), immunoprecipitation fractions that did not bind the anti-Halo beads (unbound) and proteins immunoprecipitated by the anti-Halo beads were immunoblotted as shown.

Fig 4. Halo-PROTAC mediated zDHHC-PAT degradation decreases substrate palmitoylation in Flp-In 293 T-REx cells.

A–Palmitoylated proteins were purified from Flp-In 293 T-REx cells engineered to express tetracycline (Tet) inducible Halo-zDHHC5 and phospholemman (PLM) after expression was induced or not with tetracycline (± Tet). The Western blots show palmitoylated proteins (Palm) immunoblotted alongside corresponding unfractionated cell lysates (UF). The bar chart (right) shows the relative palmitoylation of PLM. **: P<0.01, one-way ANOVA followed by Tukey’s multiple comparisons test. B–Palmitoylated proteins were purified from Flp-In 293 T-REx cells engineered to express tetracycline (Tet) inducible Halo-zDHHC20 after expression was induced or not with tetracycline (± Tet). The Western blots show palmitoylated proteins (Palm) immunoblotted alongside corresponding unfractionated cell lysates (UF). The bar chart (right) shows the relative palmitoylation of IFITM3. **: P<0.01, one-way ANOVA followed by Tukey’s multiple comparisons test. C–Palmitoylated proteins were purified from Flp-In 293 T-REx cells engineered to express tetracycline (Tet) inducible Halo-zDHHC20 and transfected with HA-tagged SARS-CoV2 spike. The Western blots show palmitoylated proteins (Palm) immunoblotted alongside corresponding unfractionated cell lysates (UF). The bar chart (right) shows the palmitoylation of the SARS-CoV2 spike 80kDa cleavage product relative to its abundance in the corresponding unfractionated cell lysate.

Fig 5. Targeting Halo-zDHHC20 in Vero E6 zDHHC20 knockout cells.

A–Subcellular location of Halo-zDHHC20 visualised by staining cells with TAMRA-chloroalkane either without (uninduced) or with (induced) gene induction using tetracycline. Scale bar: 10μm. B–Western blotting confirms successful induction of Halo-zDHHC20 expression (left). Dose response relationship for Halo-PROTACs in Vero E6 zDHHC20 knockout cells expressing Halo-zDHHC20 (right). Cells were treated with the indicated concentration of Halo-PROTAC for 18–24 hours and lysates immunoblotted as shown. V: vehicle (DMSO) control. The bar chart (right) shows the expression of Halo-zDHHC20 relative to GAPDH. C–Impact of co-applying the multi-drug resistance transporter inhibitor verapamil (40μM) with compound 1 (1μM) for either 24h or 48h on Halo-zDHHC20 abundance in Vero E6 zDHHC20 knockout cells. The bar chart (right) shows the expression of Halo-zDHHC20 relative to GAPDH. *: P<0.05, **: P<0.01, one-way ANOVA followed by Tukey’s multiple comparisons test. D–Palmitoylated proteins were purified from Vero E6 cells engineered to express tetracycline (Tet) inducible Halo-zDHHC20, transfected with HA-tagged SARS-CoV2 spike. The Western blots show palmitoylated proteins (Palm) immunoblotted alongside corresponding unfractionated cell lysates (UF). The bar charts show the relative palmitoylation of the SARS-CoV2 spike 80kDa cleavage product and IFITM3.

Fig 6. Comparison of nanobody and PROTAC mediated Halo-zDHHC20-YFP degradation in Flp-In 293 T-REx cells.

Flp-In 293 T-REx cells engineered to express Halo-zDHHC20-YFP were treated with compound 1 (1μM) and / or transfected with the YFP-directed nanobody LAG-16 with the HECT Domain of NEDD4L fused at either its amino or carboxyl terminus. The bar chart (right) shows the expression of Halo-zDHHC20-YFP relative to GAPDH. *: P<0.05, **: P<0.01, ***: P<0.001, ****: P<0.0001, one-way ANOVA followed by Tukey’s multiple comparisons test.