Protocol for a semi-quantitative approach to identify protein S-palmitoylation in cultured cells by acyl biotin exchange assay

Summary Palmitoylation is a post-translational lipid modification in which palmitic acid is conjugated predominantly to cysteine residues of target proteins, allowing them to tether to cell membranes. Here, we describe a protocol to perform a stepwise acyl biotin exchange assay to identify protein S-palmitoylation. We describe steps for initial blocking of free thiols in protein lysates, subsequent replacement of thioester-linked palmitate groups with a biotin tag for affinity enrichment, and identification of palmitoylated proteins by SDS-PAGE. For complete details on the use and execution of this protocol, please refer to Leishman et al.1

Palmitoylation, or S-acylation, is a process whereby palmitic acid (a 16-carbon saturated fatty acid) is covalently added to cysteine residues via a labile thioester linkage. 2 Although rare, palmitoylation can also occur on serine, threonine, or lysine residues of a protein.The attachment of palmitate increases the hydrophobicity of proteins allowing them to tether to cellular membranes. 3Consequently, palmitoylation is pivotal for regulating the localization, function, and stability of proteins. 4ome of the first proteins identified to undergo palmitoylation were the heterotrimeric G proteins and the H-Ras proteins. 5,6In the case of heterotrimeric G proteins, mutating the palmitoylated cysteine residues in the Gqa subunit prevented membrane attachment and activation of the downstream phospholipase C signaling pathway. 5 our research on NLRs and inflammasomes, we recently identified that the sensor protein NLRP3 undergoes palmitoylation. 1 In agreement with its functions, we discovered that palmitoylation of NLRP3 promotes recruitment of the sensor protein to the trans Golgi network (TGN) vesicles.This translocation is important for the assembly and consequently the activation of the NLRP3 inflammasome.7 In cells exposed to NLRP3 stimuli such as ATP or nigericin, this membrane association triggers inflammasome signaling, resulting in the cleavage of protease caspase-1, secretion of pro-inflammatory cytokine IL-1b, and the induction of inflammatory cell death, pyroptosis.8,9 By contrast, inhibition of palmitoylation by a pharmacological inhibitor, 2-bromopalmitate (2-BP), results in reduced co-localization of NLRP3 with TGN vesicles and diminished caspase-1 cleavage and IL-1b secretion.1 Here, we describe a detailed protocol to identify palmitoylation in HEK293T cells. W strongly recommend using one of the established palmitoylated proteins as a positive control in this assay.In our protocol below, we employed Caveolin-1 as a positive control.Caveolin-1 is a major component of caveolae, membrane subdomains which are involved in organizing signaling molecules at the plasma membrane.10 In pulse-chase experiments conducted on endothelial cells with labeled palmitate, Caveolin-1 palmitoylation was found rather stable, making it a suitable marker for a positive control in this assay.10 Notably, this protocol is restricted to examining S-acylation in cellular lysates; it is not suitable for identifying O-palmitoylation or N-palmitoylation of proteins.11 In silico analysis of palmitoylation sites This initial in silico analysis allows predicting putative palmitoylated sites in proteins which may assure whether the protein of interest is palmitoylated before undertaking this protocol.However, it is important to note that these predictions are not always accurate.
1. Identify the protein of interest that is to be studied.For the remainder of this protocol, NLRP3 is used as protein of interest and Caveolin-1 is employed as a control.2. Retrieve the amino acid sequence for the protein of interest from publicly available tools such as PubMed or UniProt.3. Input the sequence into GPS-Palm (https://gpspalm.biocuckoo.cn/) or SwissPalm (https:// swisspalm.org/)which may help in the identification of putative palmitoylation residues in the protein of interest 12,13 We performed this protocol in HEK 293T cells.HEK 293T cells do not express any of the NLR proteins.Consequently, we transfected and overexpressed Myc-tagged mouse NLRP3 in these cells.However, we have also used this protocol successfully in primary bone marrow-derived and in immortalized mouse macrophages to identify if endogenous NLRP3 undergoes palmitoylation.As such, the protocol can be used to examine palmitoylation in proteins that are sufficiently expressed endogenously in a variety of mammalian cells.

MATERIALS AND EQUIPMENT
Buffers can be prepared in advance with the 'core buffer' acting as a base for lysis and flex buffers.CRITICAL: Sodium dodecyl sulfate (SDS) is an anionic detergent.It is harmful if swallowed and can cause skin and serious eye damage.Wear protective gloves and eye protection when handling.
Note: pH of the buffers can fluctuate when stored for a long time.Make fresh buffers or check and readjust the pH, if required.

STEP-BY-STEP METHOD DETAILS
Transfection and stimulation of HEK 293T cells

Timing: 2 days
In this step, cells are transfected with Myc-NLRP3 plasmid to over-express the NLRP3 protein.
2. Each confluent 6-well yields $200 mg of protein which should be sufficient for generating HAM positive, and HAM negative control samples later in the protocol.a.We recommend using a minimum of 100 mg protein per sample.b.The plating density may need to be optimized for other cell types so that the wells are approximately 90% confluent.3. Allow cells to grow in a 37 C incubator for 18 h-24 h.

Day 1
4. Change media to Opti-MEM to get cells ready for transfection.
a. Transfection can be omitted if the protein of interest is expressed well in HEK 293T cells or in your chosen cell type. 5. Add 1.5 mL Lipofectamine 2000 in a final volume of 25 mL OptiMEM in a 1.5 mL tube.Vortex gently for $10 s.
6.In a separate 1.5 mL tube, add plasmid DNA in a final volume of 25 mL OptiMEM.Vortex gently for $10 s. a.In this example, Myc-NLRP3 plasmid was transfected at 0.5 mg/well in a 6 well plate.b.If several wells are to be transfected, the above reagents can be prepared as a master mix. 7. Let the two tubes sit for 5 min at 22 C-24 C. 8. Mix the contents of the two tubes into a single tube and gently vortex for $10 s.Incubate the tube containing both Lipofectamine 2000 and plasmid DNA for 25 min at 22 C-24 C. 9. Add 50 mL complex to the desired wells dropwise.10.Incubate the cells for 18 h-24 h in a 37 C incubator.
a. Depending on the plasmid and the gene of interest, the cells may take between 24 h and 72 h to express the gene of interest.This may need to be optimized in a pilot experiment examining the protein expression at different time points.

Day 2
11. On the day of experiment, wash the cells once with PBS and change growth media to fresh 1 mL DMEM.12. Activate the inflammasome with a final concentration of 5 mM ATP for 30 min.
a.As HEK 293T cells do not express the full suite of inflammasome components, they do not undergo pyroptosis.b.Primary and immortalized mouse macrophages will first need to be stimulated with LPS (500 ng/mL) for 4 h in a 37 C incubator to allow cells to express inflammasome components.
Note that these cells rapidly undergo pyroptosis upon inflammasome activation which may affect protein yield between samples.c.It is important to note that palmitoylation may or may not require cell stimulation.This may depend on the expression of specific palmitoyl acyltransferases.13.Using a cell scraper, collect cells in a sterile 1.5 mL Eppendorf tube.14.Spin down the cells in a microcentrifuge for 5 min at 1,000 g at RT. 15.Wash the cell pellet in 1 mL PBS.

Timing: 1 day
In this step, N-ethylmaleimide (NEM), an alkene derived from maleic acid, is added to irreversibly block free (unbound) thiol group of cysteine residues.This step ensures that only cysteines that were originally palmitoylated are examined in the later steps of this protocol.
16. Resuspend the cell pellet from the previous step in 200 mL lysis buffer containing 20 mM NEM. 17. Vortex 3 times at 3 s intervals.
CRITICAL: pH of the lysis buffer is crucial for NEM to function and should be readjusted if the buffer was stored for a long time.The pH should be between 7.0 and 8.0 for optimal NEM function.
18. Incubate samples at 4 C for approximately 16 h-18 h with end-over-end rotation.19.Precipitate the protein with a mixture of chloroform, methanol, and water at a ratio of 1:0.4:1 (methanol: chloroform: H 2 O).
a.For the initial 2 3 106 cells used, the volumes we add are 500 mL methanol, 200 mL chloroform, and 300 mL H 2 O. 200 mL of lysis buffer contributes to the aqueous part.20.Vortex the samples for $10 s, and centrifuge at 15,000 g for 5 min.21.The centrifugation step should result in a phase separation resulting in three distinct layers (Figure 1A).With a p1000 pipette, remove the topmost aqueous layer making sure not to disturb the middle interphase consisting of proteins.a.If the aqueous layer cannot be completely removed without disturbing the protein interphase, it can be left for removal in the subsequent wash steps.22. Add 500 mL of methanol and mix by inverting the tubes 3-5 times.23.Spin at 15,000 g for 5 min in a microcentrifuge.This should result in a relatively stable pellet at the bottom of the tube.24.Carefully decant or remove the supernatant with a pipette tip.25.Repeat protein precipitation steps 19-24 for an additional 2x times.26.Leave the pellets to air dry for 10 min.
CRITICAL: Complete removal of residual NEM by repeating protein precipitation steps is important to prevent further recapping of the free thiol groups made available by cleavage of the thioester bond that linked palmitic acid to cysteine residues.Any leftover NEM in the samples will lead to a reduction or complete loss of palmitoylation signal.
Pause point: Samples can be stored at À20 C for up to 24 h.Leaving the samples for longterm at this step is not recommended.

Timing: 3-4 h
Addition of HAM in this step cleaves the thioester bond between the sulfhydryl group of cysteine residue and the palmitic acid.This newly available unbound (free) cysteine sulfhydryl group is used for biotin tagging.Hydroxylamine is not able to cleave the NEM bound sites, and these remain unavailable to the biotin tag.
27. Resuspend the protein pellet in 400 mL of flex buffer.Vortex to dissolve the protein pellet.Brief sonication up to $30 s may be required.28.Once the protein pellet is fully dissolved in the cleavage buffer, divide the samples equally (200 mL each) into pre-labeled HAM (À) and HAM (+) tubes.29.Add 4.5 mL HAM (0.7 M final concentration) to tubes labeled HAM (+).30.Incubate with end-over-end rotation (Figure 1B) at 22 C-24 C for 1 h.31.Finally, perform chloroform: methanol protein precipitation one time as described in steps 19-24.a. Allow pellets to dry fully before proceeding further.

Timing: 1 day
In this step, biotin tag is added to the newly available free thiol sites which were previously palmitoylated for affinity enrichment.The palmitoylated proteins can be detected using antibody against the protein of interest or if the protein was co-expressed with a tag on it.
32. Dissolve the protein pellet in 200 mL of flex buffer supplemented with 20 mM Biotin-HPDP.33.Incubate the samples at 22 C-24 C with end-over-end rotation for 1 h.34.Perform protein precipitation one final time as described in steps 19-24.35.Dissolve the protein in 100 mL of flex buffer.36.Estimate protein concentration by mini-BCA assay (Figure 1C) using a 10 mL protein sample.37. Normalize protein concentration for any variability across samples.a. Remove a 20 mL aliquot from each sample to load as an input control.Mix the sample with 20 mL of 2x SDS-PAGE Laemmli buffer and store at À20 C until ready to load.38.Dilute the remaining 70 mL sample 10 times by adding 630 mL of core buffer.This dilutes the SDS concentration to 0.1%.39.Add 30 mL of pre-washed streptavidin agarose beads directly to each sample and vortex for $10 s to mix the contents.a.To first wash and prepare the streptavidin-agarose beads, add 30 mL agarose beads per sample to 500 mL of core buffer.b.Spin at 10,000 g for 2 min at room temperature.c.Wash the agarose beads for a total of 3 times taking 500 mL core buffer each time.After the final wash step, resuspend the beads in 30 mL of core buffer per sample.40.Incubate the samples with end-over-end rotation at 4 C for 18 h-24 h.41.The following day, wash the samples with 500 mL of core buffer supplemented with 0.1% SDS and 0.1% NP-40.a. Repeat washes for a total of 3 times, spinning the samples at 12,000 g for 5 min between each of the washes.42.To the pellet (Figure 1D), add 75 mL of 1x Laemmli buffer and store at À20 C until ready for SDS-PAGE.

Protocol EXPECTED OUTCOMES
When this protocol is followed for a protein which has active palmitoylation sites at the time of sample collection, it can be expected to observe a band only in the HAM (+) set of samples after immunoblotting.A band should not be observed in the corresponding HAM (À) samples.Representative blots for our protein of interest NLRP3 and Caveolin-1 are shown in Figure 2A.The corresponding input control samples should also be run for immunoblotting.This allows for normalization of data and more accurate comparison between different streptavidin pull-down samples or different experimental conditions.When combined with standard densitometric analysis of immunoblots such as with ImageJ (NIH) or Image Lab (Bio-Rad), it is possible to obtain semi-quantitative data from these blots, as shown in Figure 2B.If the assay has proceeded expectedly as evaluated from the data obtained with controls in the experiment, but no bands have been observed in HAM (+) samples, it can be concluded that the protein is not S-palmitoylated.

LIMITATIONS
Whilst robust, there are limitations to this assay.The method can only detect cysteine palmitoylation (S-palmitoylation).Palmitoylation at serine, threonine or lysine residues cannot be detected with this assay.The assay involves multiple steps, and proteins which are rather unstable or insoluble, may be degraded or lost during the protocol giving misleading results.The assay is also less sensitive to detect proteins with limited palmitoylation or those which are expressed at low abundance in samples.Finally, the assay has quantitative limitations.It cannot quantify the number of active palmitoylated sites.But useful semi-quantitative information on the levels of palmitoylation between different experimental conditions can be obtained by using this protocol.

TROUBLESHOOTING Problem 1
Smearing or fragmentation of protein bands (related to steps 2 and 41).

Potential solution
Smearing on the gel can either be due to the presence of agarose beads in the sample when loading the samples, or higher protein concentration of the samples.This can be minimized by ensuring that samples are centrifuged to remove the agarose beads after all the bound protein has been recovered following boiling of the samples.Protein loading may need to be optimized.

Potential solution
This can be due to the protein having no active palmitoylation.It can also be due to protein having low palmitoylation levels which are below the sensitivity levels of this assay.However, if palmitoylation of the protein is known, then this could be due to the residual levels of NEM in the sample.Make sure that NEM is completely removed before proceeding to the addition of HAM to the samples.This can be done by increasing the number of protein precipitation steps after overnight incubation with NEM.If the problem still remains, NEM concentration in the buffer can be lowered to 15 mM but this may affect the specificity of the assay.

Problem 3
Presence of non-specific bands in HAM (À) samples (related to step 34 and Figure 3).

Potential solution
If the total amount of starting protein is too high, NEM may not efficiently block free thiol sites leading to non-specific biotinylation and consequently pull-down by streptavidin agarose.This may result in bands in the HAM (À) samples.An increased abundance of biotin in the samples may also lead to the above effect.Keep the starting protein concentration around the recommended range.Add an extra protein precipitation step to remove excess unbound biotin.Figure 3A shows a bad outcome of this protocol in which comparable bands are observed in HAM (+) and HAM (À) samples.Figure 3B shows a good outcome of the protocol illustrating no non-specific pull-down of NLRP3 in HAM (À) sample after the above troubleshooting steps have been taken.

Problem 4
Difficulties in dissolving the protein pellet after protein precipitation steps (related to step 27).

Potential solution
If the pellet will not dissolve, then it indicates the presence of methanol in the samples.This is a common occurrence and may require vigorous vortexing or physically breaking the pellet with a micro spatula.A few seconds in bath sonication may also help.Try and keep loss of protein minimal during the resuspension.

Problem 5
Complete absence of protein on gels.

Potential solution
Do a rigorous check of all molecular biology reagents to ensure that plasmid transfection of HEK 293T cells worked.Also, use a fresh aliquot of the antibody used in immunoblotting.

Note:with ddH 2 Owith ddH 2 O
Stock concentrations and storage conditions of the major components in the assay are below: N-ethylmaleimide Stock Concentration: 100 mM.Final concentration: 20 mM.Store at À20 C and use within 2 weeks of preparation.Core buffer, pH 7.4 (Tris-EDTA-NaCl) with ddH 2 O Note: Lysis buffer can be stored at 4 C for up to 1 month.Flex buffer, pH range 7.0 to 8Note: Flex buffer can be stored at 4 C for up to three months.Note: Laemmli buffer can be stored at À20 C for over a year.Hydroxylamine Stock Concentration: 32 mM.Final concentration: 0.7 M. Store at room temperature.Biotin-HPDP Stock Concentration: 20 mM.Final concentration: 1.0 M. Store at À20 C. CRITICAL: N-ethylmaleimide (NEM) is toxic upon contact and fatal if swallowed.Handle under a chemical fume hood.CRITICAL: NP-40 is a non-ionic surfactant.It is harmful if swallowed or inhaled and can cause skin and eye irritation.Handle under a chemical fume hood.CRITICAL: 2-mercaptoethanol (2-ME) is toxic and should be handled with care.

Figure 1 .
Figure 1.Images depicting some of the steps of the acyl-biotin exchange (ABE) assay (A) Protein precipitation: clear layers can be observed, with the upper phase (metabolites, salts, and amino acids), middle protein interphase shown with an arrow, and lower phase consisting of lipids.(B) Example of end-over-end rotation of samples to ensure proper reagent mixing.(C) Color development in BCA assay for protein quantification.(D) Affinity enrichment of palmitoylated proteins by streptavidin agarose.Note the protein pellet at the bottom of the tube shown with an arrow.

Figure 2 .
Figure 2. Expected outcomes from an acyl-biotin exchange (ABE) assay (A) HEK-293T cells were transfected with Myc-NLRP3 and an ABE assay was carried out.The samples and input controls were immunoblotted with either NLRP3 or Caveolin-1 antibody.Note the absence (or very low expression) of the two proteins in HAM (À) samples compared to HAM (+) samples.(B) Densitometric analysis of the blot shown in (A) to illustrate the semi-quantitative nature of this protocol.Data shown are mean with error bars showing standard deviation, and are representative of experiments done at least three times.

Figure 3 .
Figure 3. Examples of immunoblots depicting a bad vs. good outcome in an acyl-biotin exchange (ABE) assay Immunoblots were carried out on immortalized BMDMs exposed or not to LPS (500 ng/mL).Upper panels show palmitoylated NLRP3 while lower panels show NLRP3 immunoblot on input samples.The protein shows only low constitutive expression which is upregulated upon TLR4 ligation by LPS.Panel (A) shows a bad outcome in which HAM (À) sample exhibits high non-specific pull-down of NLRP3.Panel (B) depicts a good outcome with minimal NLRP3 pulldown in HAM (À) sample compared to a strong band observed in HAM (+) sample.Please refer to recommendations under 'troubleshooting' to achieve a good outcome of this protocol.