Temporal Profiling Establishes a Dynamic S-Palmitoylation Cycle

Abstract

S-palmitoylation is required for membrane anchoring, proper trafficking, and the normal function of hundreds of integral and peripheral membrane proteins. Previous bioorthogonal pulse-chase proteomics analyses identified Ras family GTPases, polarity proteins, and G proteins as rapidly cycling S-palmitoylated proteins sensitive to depalmitoylase inhibition, yet the breadth of enzyme regulated dynamic S-palmitoylation largely remains a mystery. Here, we present a pulsed bioorthogonal S-palmitoylation assay for temporal analysis of S-palmitoylation dynamics. Low concentration hexadecylfluorophosphonate (HDFP) inactivates the APT and ABHD17 families of depalmitoylases, which dramatically increases alkynyl-fatty acid labeling and stratifies S-palmitoylated proteins into kinetically distinct subgroups. Most surprisingly, HDFP treatment does not affect steady-state S-palmitoylation levels, despite inhibiting all validated depalmitoylating enzymes. S-palmitoylation profiling of APT1^-/-/APT2^-/- mouse brains similarly show no change in S-palmitoylation levels. In comparison with hydroxylamine-switch methods, bioorthogonal alkynyl fatty acids are only incorporated into a small fraction of dynamic S-palmitoylated proteins, raising the possibility that S-palmitoylation is more stable than generally characterized. Overall, disrupting depalmitoylase activity enhances alkynyl fatty acid incorporation, but does not greatly affect steady state S-palmitoylation across the proteome.

Figure 1.. Depalmitoylase inhibition enhances 17-ODYA labeling.

(a) HDFP enhances 17-ODYA incorporation at all time points during labeling. Neutralized 0.5 M hydroxylamine (NH₂OH) hydrolyzes 17-ODYA labeling, demonstrating the presence of a thioester linkage. (b) Different concentrations of 17-ODYA do not affect HDFP-dependent enhanced probe incorporation. (c) Fluorophosphonate-TAMRA (FP-TAMRA) competitive ABPP demonstrates APT enzymes are inactivated by sub-micromolar HDFP concentrations. (d) Dose-dependent profile of HDFP-dependent enhancement of 17-ODYA incorporation. (e) APT inhibition does not affect 17-ODYA incorporation levels in 293T cells. 1 µM HDFP, 50 µM 2BP, 10 µM ML348, ML349, and PalmB, 1 µM ABC44 and ML211 were used. Data represents the average intensity of the ~20 kD band (encompassing Ras small GTPases) in biological replicates (N = 3, standard error) of 17-ODYA gels, shown in Figure S2b. (f) Competitive ABPP mass spectrometry analysis of HDFP inhibition across the serine hydrolase family. Error bars represent error propagated from standard errors at differing HDFP concentrations, quantified from multiplexed TMT isobaric reporter ions. N = 4. Serine hydrolases with >2 PSMs are shown. All experiments were performed in human 293T cells.

Figure 2.. Time-dependent profiling of 17-ODYA incorporation identifies kinetic clusters of dynamic S-palmitoylated proteins.

(a) Schematic of bioorthogonal temporal S-palmitoylation profiling by multiplexed quantitative mass spectrometry. (b) Hierarchical clustering of 17-ODYA enriched S-palmitoylated proteins across both control and 1 µM HDFP-treated cells reveals distinct kinetic clusters in 293T cells. Proteins with >2-fold enrichment and at least 3 PSMs are shown. The heat map is normalized to the highest point along the labeling time course. Branches were simplified to highlight higher level grouping. (c) Representative labeling time course of select proteins from each subclass. Unique peptides were used to confirm analysis of RhoA and not RhoB, as both are frequently annotated S-palmitoylated proteins. (d) Conserved kinetic profiles of temporal S-palmitoylation between human 293T and HAP1 cells (Figure S3).

Figure 3.. Comparison of 17-ODYA and acyl-RAC enrichment of S-palmitoylated proteins.

(A) Gel-based analysis of alkynyl fatty acid metabolic labeling and cli9+ck chemistry detection of S-palmitoylation compared to hydroxylamine-switch analysis by iodoacetamide-TAMRA detection. (B) Limited overlap of click chemistry and acyl-RAC analysis of specifically enriched S-palmitoylated proteins from 293T cells. Mass spectrometry data is reported in Tables S2 and S4.

Figure 4.. Depalmitoylase inhibition does not influence steady-state S-palmitoylation.

(a) In-gel analysis of S-palmitoylation using the hydroxylamine-switch analysis reports equivalent labeling profiles after 17-ODYA and HDFP treatment. Iodoacetamide-rhodamine (IAM-TAMRA) labeling reports the presence of hydroxylamine sensitive thiols. (b) Equivalent enrichment of S-palmitoylated proteins by quantitative proteomic analysis of acyl-RAC enriched proteins from control (DMSO) and HDFP (1 µM) or HDFP (1 µM) +17-ODYA (20 µM) treated cells, N=4. Proteins with >2 PSMs, multiple annotations in SwissPalm are shown. (c) Confirmatory acyl-RAC analysis of pan-RAS and SCRIB S-palmitoylation. (d) Quantitation of acyl-RAC enrichment relative to control (DMSO) across replicates after normalizing to input (Ras, N = 5 for all conditions except 17-ODYA, N = 3; Scrib, N = 3 for all conditions except 17-ODYA, N = 2, ±SEM) demonstrates no significant changes following HDFP treatment, but shows major reductions after 2BP treatment (two-tailed Student’s t-test with unequal variance).

Figure 5.. Steady-state S-palmitoylation is unaffected in APT1^−/−/APT2^−/− mouse brains.

(a) FP-TAMRA labeling of active serine hydrolases in whole mouse brain homogenates confirms selective deletion of APT1 and APT2 with no observable compensatory changes across other hydrolases. Knockout was also confirmed by western blot. Labeling is shown with 2 µM FP-TAMRA. (b) Hydroxylamine-switch analysis of S-palmitoylation in normal and knockout samples does not reveal major changes in S-acylation profiles. Detection with 1 µM IAM-TAMRA. (c) Volcano plot analysis of acyl-RAC enrichment and multiplexed TMT mass spectrometry analysis of APT1^−/−/APT2^−/− whole mouse brain homogenates (N=3). Red lines represent p < 0.05 from two-tailed Student’s t-test with unequal variance (horizontal) and 2-fold change (vertical).