Garcinol Is an HDAC11 Inhibitor

Abstract

Garcinol is a natural product from the Garcinia Indica fruit and is well-known as an anti-oxidant, anti-inflammation, and anti-cancer agent. However, the understanding of its mechanism of action is still incomplete. It has been reported to be a histone acetyltransferase (HAT) inhibitor Here, we surprisingly found that Garcinol is a potent HDAC11 inhibitor (HDAC11 IC₅₀ ~5 μM in vitro HPLC assay and IC₅₀ ~10 μM in cellular SHMT2 fatty acylation assay) which is comparable to previously reported HDAC11 inhibitors. Additionally, among all the HDACs tested, Garcinol specifically inhibits HDAC11 over other HDACs. HDAC11 is the only class IV HDACs and there are very few inhibitors available for it. Therefore, this study provides a new HDAC11 inhibitor lead from natural products and may help explain the various biological activities of Garcinol.

Graphical Abstract

Natural product, Garcinol, is a poly-isoprenylated benzophenone compound extracted from the Garcinia indica fruit.^1–2 It has been reported to have various interesting biological properties, such as anti-cancer, anti-inflammation, and anti-oxidant activities (Figure 1).^1–5 Its mechanism of action, however, remains largely unknown. It has been reported to decrease the expression of COX2, iNOS, and NF-kB.⁵ In the cell-free extract, it could inhibit 5-lipoxygenase and microsomal prostaglandin E2 synthase 1 (mPGES-1).³ It could also inhibit STAT3 activation and suppresses tumorigenesis.⁴ However, the exact protein targets underlying its various biological functions is rather unclear. The only reported direct target is histone acetyltransferases (HATs), p300 and PCAF.⁶ However, many of the biological activities of Garcinol, such as protection against obesity^24–25 and multiple sclerosis²³ may not result from HAT inhibition.^7–8

Figure 1.

Structure and biological activities of Garcinol.

Surprisingly, here we found that Garcinol is a potent HDAC11 inhibitor. HDAC11 is the only class IV HDAC member. Recently, several reports showed that it does not have deacetylation activity but instead can hydrolyze long-chain fatty acyl groups from protein lysine residues.^9–11 The first reported physiological substrate of HDAC11 is SHMT2.⁹ HDAC11 regulates type I interferon signaling by defatty-acylating SHMT2 and promoting type I interferon receptor ubiquitination and degradation.⁹ Several recent reports showed that HDAC11 knockout in mice displays beneficial effects, including increased thermogenesis in adipose tissue, increased resistance to high-fat diet-induced obesity, as well as protection from symptoms of multiple sclerosis.^12–14 The interesting phenotypes of HDAC11 knockout warrants the development of HDAC11 inhibitors. Currently, only a few HDAC11 inhibitors are known.^15–16 The finding that Garcinol is an HDAC11 inhibitor will not only provide an opportunity to understand the mechanism underlying Garcinol’s various biological activities but also provide a new scaffold to develop new HDAC11 inhibitors to explore the therapeutic potential of targeting HDAC11.

Results and discussion

A new strategy to discover HDAC11 inhibitor.

Since Class I, II, and IV HDACs have a Zinc dependent catalytic mechanism, all previous HDAC inhibitors targeting Class I and II HDACs possess a Zinc-binding group which is essential for HDAC inhibition.^17–19 Previously, known HDAC inhibitors were examined as potential HDAC11 inhibitors. However, we and others soon discovered that this strategy does not work for developing HDAC11 inhibitors.^{11, 15–16} Most known HDAC inhibitors turned out to be unable to inhibit HDAC11, while a few (trapoxin A and its analogs) can inhibit HDAC11, but they do not have the desired selectivity for HDAC11.¹¹ Hence, a new strategy to discover HDAC11 inhibitors is needed.

We previously used an activity-based design approach and discovered a potent and selective HDAC11 inhibitor, SIS17,¹⁵ which is similar to FT895,¹⁶ an HDAC11 inhibitor developed by Forma Therapeutics. This approach focuses on compounds that contain zinc-chelating groups and long hydrophobic groups to occupy the acyl pocket of HDAC11 that accommodate long-chain fatty acyl groups. Guided by similar principles, we turned our attention to natural products with similar properties (zinc-chelating groups and large hydrophobic groups). We noticed that several natural products (curcumin, Garcinol, and anacardic acid) with interesting biological activities contain potential Zinc binding groups (Figure 2A), which is indispensable for HDAC11 inhibition. Furthermore, Garcinol and anarcardic acid also contain large hydrophobic moieties, which could potentially bind the active site pocket of HDAC11 that accommodate long-chain fatty acyl groups.

Figure 2.

Garcinol is an HDAC11-specific inhibitor in vitro. (A) Structures of metal-chelating groups-containing natural products tested for HDAC11 inhibition. The percent of inhibition at 10 μM concentration is indicated. (B) HDAC11 inhibition curves of Garcinol on two different myristoyl peptides. (C) Lineweaver-Burk plot for Garcinol. (D) IC₅₀ values of Garcinol on different HDACs. The following substrates were used in the assays: a, myristoyl H3K9; b, acetyl H3K9; c, trifluoroacetyl H3K9; and d, acetyl p53.

We tested whether they could inhibit HDAC11 with an in vitro HPLC assay using myristoyl-H3K9 peptide as the substrate. Anacardic acid showed weak inhibition at 10 μM (20% inhibition against HDAC11) while Curcumin showed no inhibition at 10 μM. Garcinol, on the other hand, exhibited 60% inhibition at 10 μM (Figure 2A).

We further examined its IC₅₀ values of Garcinol against HDAC11 with myristoyl-SHMT2 and myristoyl-H3K9 peptides. Garcinol showed 4.4 μM (Myristoyl-SHMT2) and 6.1 μM (Myristoyl-H3K9) IC₅₀ values based on our HPLC assay (Figure 2B).

We next explored whether Garcinol is a competitive inhibitor or not using a Lineweaver-Burk plot. We fixed HDAC11 concentration at 20 nM and measured the initial rates with different concentrations of myristoyl-H3K9 and Garcinol at 37 °C. Plotting the initial rates versus the concentrations of myristoyl-H3K9 at different Garcinol concentrations yielded lines intersecting on the y-axis. This result showed that Garcinol is a competitive inhibitor of HDAC11 (Figure 2C).

Garcinol is selective for HDAC11 among all the HDACs tested.

Given that Garcinol has been reported to inhibit HATs, we initially expected that Garcinol might be a pan-assay interfering compound and may inhibit many HDACs. So next, we tested its effects other HDACs using the same HPLC assay. Surprisingly, to our delight, Garcinol showed a remarkable selectivity among the HDACs tested. It only showed 10% inhibition against SIRT3 demyristoylation activity at 50 μM and 43% inhibition against HDAC8 demyristoylation activity at 25 μM. For other HDACs tested, we could not detect any inhibition at 50 μM (Figure 2D).

Given the surprising selectivity of Garcinol towards HDAC11 among all the HDACs tested, we then proceeded to test its ability to inhibit HDAC11 in cells. We measured the fatty acylation of the known HDAC11 physiological substrate, SHMT2,⁹ in cells. The acylation of SHMT2 was detected using an alkyne tagged palmitic acid analog (Alk14, 50 μM). HEK293T cells were incubated with Alk14 and after cell lysis, the Alk14-labeled proteins were conjugated with biotin-azide using click chemistry. Biotinylated proteins were immunoprecipitated with streptavidin beads and immunoblotted against SHMT2. Higher western blot signal would indicate a higher fatty acylation level on SHMT2. At 10, 20, and 40 μM of Garcinol, we could see a clear increase of fatty acylation on SHMT2 (Figure 3A). Thus, Garcinol can inhibit HDAC11 in cells.

Figure 3.

Garcinol inhibits HDAC11 in cells. (A) Garcinol increases the fatty acylation level of SHMT2, a substrate of HDAC11. (B) Garcinol does not affect the acetylation level of Histone H3, with or without co-incubation with trapoxin A (TpxA) in HEK293T. (C) Garcinol does not significantly affect the acetylation level of Histone H4, with or without co-incubation with trichostatin A (TSA) in HeLa cells. (D) HEK293T cell viability with Garcinol, Isogarcinol and SIS17. Signal intensities (quantified by ImageJ) shown in A-C were normalized to input SHMT2, actin, histone H4 levels, respectively.

Garcinol has been reported to inhibit HATs with IC₅₀ values around 5 μM.^{6, 20} To test whether Garcinol indeed inhibits HATs in cells, we also measured the histone acetylation levels in HEK 293T cells in the presence of Garcinol. DMSO and a pan-HDAC inhibitor, trapoxin A (TpxA), were used as controls. Garcinol did not change histone acetylation levels at 10, 20, and 40 μM, while TpxA increased the acetylation level of histone H3 drastically even at 5 μM (Figure 3B). Since the base level of acetyl-histone H3 was low, we treated cells with TpxA and Garcinol to see whether decrease in histone acetylation caused by Garcinol could be observed. However, Garcinol did not change the acetylation level of H3 even in the presence of TpxA (Figure 3B). Thus, our data showed that Garcinol inhibits HDAC11 in cells but it does not seem to inhibit HATs on H3 acetylation under the conditions we used.

Because previous HATs inhibition by Garcinol was carried out in HeLa cells by measuring H4 acetylation, we also measured the histone H4 acetylation levels in HeLa cells in the presence of Garcinol. We found that trichostatin A (TSA) at 0.1 μM did not kill HeLa cells but increased H4 acetylation level. We then examined whether garcinol could decrease the H4 acetylation in the presence of 0.1 μM TSA. A reported HAT inhibitor, CPTH6, was used as a positive control. CPTH6 at 70 μM (literature examples typically uses 50-100 μM) could decrease the acetylation level under the TSA treatment. Garcinol at 20 and 40 μM may slightly decreased the H4 acetylation level, but the decrease was not obvious (Figure 3C). Thus, our data suggest that Garcinol is much better at inhibiting HDAC11 than HATs. We do not know exactly why Garcinol does not exhibit strong HAT inhibition as previously reported. It could be due to different preparations of Garcinol. The Garcinol we used was purchased from Cayman. We confirmed its purity and structure using LC-MS and ¹H and ¹³C NMR (Supporting Information).

Pan HDACs inhibition and HATs inhibition is well-known to be toxic to mammalian cells, which is a major reason that they are considered anticancer therapeutic candidates.^{7, 21–22} Hence, if Garcinol is selective for HDAC11 inhibition, we expect that it should not be cytotoxic to mammalian cells. We examined the cytotoxicity of Garcinol along with SIS17 and isogarcinol in HEK293T cells. Garcinol did not cause cell death in HEK293T cells at concentrations tested. This result is consistent with the observation that under the conditions we used, Garcinol selectively inhibits HDAC11, but not other HDACs or HATs (Figure 3C).

The enol form of the β-diketone moiety of Garcinol contributes to zinc-chelation.

There are two potential zinc-chelating groups in Garcinol (Figure 4A), the enol form of the β-diketone moiety, or the catechol moiety. One commercially available Garcinol-derived compound is isogarcinol. Isogarcinol has the catechol but the β-diketone group is alkylated. Thus, we used isogarcinol to test whether the β-diketone moiety of Garcinol is important for HDAC11 inhibition. Using the same HPLC assay, we found that isogarcinol has much weaker HDAC11 inhibition activity (IC₅₀ = 40.7 ± 7.3 μM). This result support that the β-diketone part of Garcinol is important for HDAC11 inhibition and is likely the zinc-binding group of Garcinol (Figure 4B). This observation could help to modify the Garcinol structure to obtain even better HDAC11 inhibitors.

Figure 4.

The enol form of the β-diketone moiety is likely the zinc binding group of Garcinol. (A) Speculation of possible zinc-binding groups in Garcinol. (B) The most probable zinc-binding group of Garcinol based on IC₅₀ of isogarcinol.

In summary, by examining three natural products with potential zinc-chelating groups and large hydrophobic groups, we surprisingly found that Garcinol efficiently inhibits HDAC11. Furthermore, Garcinol showed selectivity toward HDAC11 among all the HDACs tested. In cells, it did not affect the acetylation of histone H3, but efficiently increased the fatty acylation level of SHMT2, a previously established HDAC11 substrate. These results are comparable to our best HDAC11 inhibitor previously developed, SIS17. Thus, our data established Garcinol as a potent and selective HDAC11 inhibitor.

Garcinol has been reported to have many beneficial biological activities, including anticancer, anti-oxidant, and anti-inflammation activities. Even though the biological effects of Garcianol are likely complex and may be through multiple targets, our study here suggests that some of its biological effects is likely through HDAC11 inhibition. Interestingly, several in vivo effects of Garcinol showed similarity with HDAC11 knockout phenotypes. Garcinol treatment in the experimental autoimmune encephalomyelitis (EAE) mice model exhibited a decreased level of demyelination in the spinal cord.²³ Loss of HDAC11 also significantly inhibited demyelination of the spinal cord in the EAE mice model.¹⁴ Additionally, Garcinol has been reported to reduce obesity and diabetes in the high-fat-diet mouse and rat models,^24–25 correlating with the phenotypes of HDAC11 knockout in mice on a high-fat diet.^12–13 These striking phenotypic similarities provide further support that Garcinol is an HDAC11 inhibitor. Our study thus not only provides a novel HDAC11 inhibitor to further investigate the biological function of HDAC11 and therapeutic potential of targeting HDAC11, but also an opportunity to understand the mechanism of action of a complex natural product with an interesting history and biological activities.

Methods

Reagents.

Commercially available Curcumin (#81025.1), Anacardic acid (#13144), Garcinol (#10566) and Isogarcinol (#21164) were purchased from Cayman. Anti-Flag affinity gel (#A2220) and anti-Flag antibody conjugated with horseradish peroxidase (#A8592) were purchased from Sigma-Aldrich. Acetyl-Histone H3 Antibody (#9675) and acetyl histone H4 antibody (#13534) were purchased from Cell Signaling Technology. SHMT2 (mSHMT Antibody (F-11): sc-390641- Non-conjugated), ß-actin antibody (SC-47778 HRP) and goat anti-mouse IgG-HRP (sc-2005) were purchased from Santa Cruz. And histone H4 antibody (#ab197517) was purchased from Abcam. Triple Flag peptide for eluting Flag-tagged HDAC11 enzyme and protease inhibitor cocktail were purchased from Sigma-Aldrich. ECL plus western blotting detection reagent and universal nuclease for cell lysis were purchased from Thermo Scientific Pierce. Acyl peptides (myristoyl-H3K9) and Alk14 were synthesized.²⁶ The peptide sequence for the H3K9 peptide is KQTARK(myristoyl)STGGWW with uncapped N-terminal and SHMT2 peptide is SDEVK(myristoyl)AHLLAWW with capped acetyl N-terminal.

Instrumentation.

Analytical HPLC analysis was carried out on a SHIMADZU HPLC with Kinetex 5u EVO C18 100A column (100 mm × 4.60 mm, 5 μm, Phenomenex) monitoring at 215 nm and 280 nm. Solvents for analytical HPLC were water with 0.1% (v/v) trifluoroacetic acid (TFA) as solvent A and acetonitrile with 0.1% (v/v) TFA as solvent B. Compounds were analyzed at a flow rate of 0.5 mL/min.

Cell culture.

HEK293T cells were cultured in designated DMEM media with 10% (v/v) FBS heat-inactivated at 56 °C. The cell lines used for experiments had been passaged no more than 20 times and all cell lines were tested for and showed no mycoplasma contamination.

Expression and purification of HDACs and sirtuins from HEK293T Cells.

Full-length human HDACs and sirtuins were inserted into the pCMV-Tag 4a vector with a C-terminal Flag tag. All plasmids were transfected into HEK293T cells using FuGene 6 Transfection Reagent according to the manufacturer’s protocol (from Promega). Eight 10-cm plates of cells were collected by centrifugation at 500 × g for 5 min and then lysed in 2 ml of Nonidet P-40 lysis buffer (25 mM Tris-HCl, pH 7.4, 150 mM NaCl, 10% (v/v) glycerol and 1% (v/v) Nonidet P-40) with protease inhibitor cocktail (1:100 dilution) at 4 °C for 30 min. After centrifugation at 15,000 × g for 15 min, the supernatant was collected and incubated with 20 μL anti-Flag affinity gel at 4 °C for 2 h. The affinity gel was washed three times with 1 ml of washing buffer (25 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.2% (v/v) Nonidet P-40) and then eluted with 100 μl of 300 μM 3× FLAG peptide (dissolved in 25 mM Tris-HCl, pH 7.4, 150 mM NaCl and 10% (v/v) glycerol) twice (total 200 μL) for 1 hr each time. The eluted proteins were checked by 12% SDS-PAGE to be at least 80% pure.

HDACs and sirtuins deacylase activity assay.

All inhibitors were dissolved in DMSO. For the deacylation assay using the zinc-dependent HDACs, HDAC1, HDAC4, HDAC6 and HDAC8 was incubated in 40 μL reaction mixtures (25 mM Tris-HCl, pH 8.0, 50 mM NaCl, 25 μM H3K9 acetyl for HDAC1 (32 nM), trifluoacetyl for HDAC4 (64 nM), myristoyl for HDAC8 (1 μM), and 25 μM p53 acetyl peptide for HDAC6 (128 nM)) at 37 °C for 20 min with 64 nM of HDAC11. To quench the reactions, 40 μL cold acetonitrile was mixed with the reaction mixture. After centrifuging at 15,000 × g for 15 min, the supernatant was collected and analyzed by HPLC using Kinetex 5u EVO C18 100A column (150 mm × 4.6 mm, Phenomenex). Solvents used for HPLC were water with 0.1% (v/v) trifluoroacetic acid (solvent A) and acetonitrile with 0.1% (v/v) trifluoroacetic acid (solvent B). The gradient for HPLC condition was 0% B for 2 min, 0-20% B in 2 min, 20-40% B in 13 min, 40-100% B in 2 min, and then 100% B for 5 min. The flow rate was 0.5 mL/min. Procedure for HPLC assay with SHMT2 myristoyl peptide is the same as H3K9 myristoyl procedure described above except 120 min incubation time used instead of 20 min.

For the deacylation assay using sirtuins, SIRT2, SIRT3, or SIRT6 (2 μM) were incubated in 40 μL reaction mixtures (25 mM Tris-HCl, pH 8.0, 50 mM NaCl, 1 mM DTT, 1 mM NAD⁺, 25 μM H3K9 myristoyl peptide) at 37 °C for 5 min (SIRT2 and SIRT3) or 30 min (SIRT6). To quench the reactions, 40 μL of cold acetonitrile was mixed with the reaction mixture. After centrifuging at 15,000 g for 15 min, the supernatant was collected and analyzed by HPLC following the same conditions described above.

Lineweaver-Burk Plot with Garcinol.

A 80 μL of reaction mixture (25 mM Tris-HCl ,50 mM NaCl, pH 8.0) containing 20 nM of HDAC11, different concentration of Garcinol (0, 3, 7, 12, 20 μM of) and myristoyl-H3K9 peptide (4, 8, 13.5 μM) was incubated for 40 mins at 37 °C. Then the reactions were quenched with 80 μL of Acetonitrile. A total volume of 150 μL of each reaction was placed in a borosilicate glass vial (Thermo Scientific/MSCERT5000-39TR). The reaction was analyzed by HPLC (monitoring at 280 nm) and the Lineweaver-Burk plot was drawn using KaleidaGraph.

Detection of histone H3 acetylation in HEK 293T cells.

HEK293T cells were cultured in 10-cm dishes. 80~90% confluent HEK293T cells were treated with the corresponding inhibitors (DMSO, 5 μM of TpxA, 5 μM of TpxA + 10 μM of Garcinol, 5 μM of TpxA + 20 μM of Garcinol, 5 μM of TpxA + 40 μM of Garcinol, 10 μM of Garcinol, 20 μM of Garcinol, and 40 μM of Garcinol) for 6 h. After 6 h, cell culture media were removed and 5 ml of PBS was added into each dish. The cells were scraped off the plates and collected by centrifugation at 500 g for 5 min. Collected cells were washed twice with 1 ml of PBS. After removing PBS gently removed by pipetting, cells were then lysed in 50 μL of 4% SDS lysis buffer (50 mM triethanolamine at pH 7.4, 150 mM NaCl, 4% (w/v) SDS) with protease inhibitor cocktail (1:100 dilution) and nuclease (1:1000 dilution). After vortexing, the samples were incubated at room temperature for 15 min. The protein concentration in the total cell lysate was determined using a bicinchoninic acid (BCA) assay and an equal amount of protein was loaded onto a 15% SDS-PAGE gel and resolved. The gel was used for western blot analysis with acetyl histone H3 antibody (Cell Signaling, #9675) and β-actin (SC-47778 HRP) as a loading check.

Detection of histone H4 acetylation in HeLa cells.

HeLa cells were cultured in 10-cm dishes. 80~90% confluent HeLa cells were treated with the corresponding inhibitors (DMSO, 0.1 μM of TSA, 0.1 μM of TSA + 70 μM of CPTH6, 0.1 μM of TSA + 20 μM of Garcinol, and 0.1 μM of TSA + 40 μM of Garcinol) for 6 h. After 6 h, cell culture media were removed and 5 ml of PBS was added into each dish. The cells were scraped off the plates and collected by centrifugation at 500 g for 5 min. Collected cells were washed twice with 1 ml of PBS. After removing PBS gently removed by pipetting, cells were then lysed in 50 μL of 4% SDS lysis buffer (50 mM triethanolamine at pH 7.4, 150 mM NaCl, 4% (w/v) SDS) with protease inhibitor cocktail (1:100 dilution) and nuclease (1:1000 dilution). After vortexing, the samples were incubated at room temperature for 15 min. The protein concentration in the total cell lysate was determined using a BCA assay and an equal amount of protein was loaded onto a 15% SDS-PAGE gel and resolved. The gel was used for western blot analysis with acetyl histone H4 antibody (Cell Signaling, #13534) and histone H4 antibody (Abcam, #ab197517).

Western Blot.

Proteins were resolved by 12% or 15% SDS-PAGE and transferred to PVDF membranes. The membrane was blocked with 5% (w/v) BSA in PBS with 0.1% (v/v) Tween-20 (TPBS) at room temperature for 60 min. The primary antibody was diluted with fresh 5% (w/v) BSA in TPBS (1:5,000 dilution for antibodies to Flag, Ac-histone H3, Ac-ɑ-tubulin, and SHMT2) and incubated with the membrane at room temperature for 1 h or at 4 °C for 12 h. After washing the membrane three times with TPBS, the secondary antibody (1:3,000 dilution with 5% (w/v) BSA in TPBS) was added and then incubated at room temperature for 1 h. The chemiluminescence signal in the membrane was recorded after developing in ECL Plus Western Blotting detection reagents using a Typhoon 9400 Variable Mode Imager (GE Healthcare Life Sciences).

Detection of Lysine Fatty Acylation on Endogenous SHMT2 by Western Blot.

HEK293T cells were treated with 50 μM of Alk14 and inhibitors or DMSO for 6 h. One 80-90% confluent 15-cm plate of cells (approximately 2 × 10⁶ cells per 15-cm dish) were collected by centrifugation at 1000 g for 5 min and lysed in 4% SDS lysis buffer (50 mM triethanolamine at pH 7.4, 150 mM NaCl, 4% (w/v) SDS) with protease inhibitor cocktail (1:100 dilution) and nuclease (1:1000 dilution) at room temperature for 15 min. The proteins were precipitated with cold methanol (200 μL per sample), cold chloroform (75 μL per sample), and cold water (150 μL per sample). After vortexing to mix well, samples were spun down at 17,000 g for 15 min. Proteins should be located between the upper layer and bottom layer. The upper layer of the solvents was gently removed. Then 1 mL of cold methanol was added to each sample to wash the proteins by vortexing. Samples were centrifuged at 17,000g for 5 min and the solvent was removed. The washing was repeated once more. After methanol washing, methanol was removed from tubes and protein pellets in tubes were air-dried for 5-10 min. Proteins were dissolved in 100 μL of click chemistry buffer (25 mM HEPES, pH 7.4, 150 mM NaCl, and 4% (w/v) SDS). The proteins were sonicated for 30 min at room temperature to be resolubilized. The concentration of the resolubilized proteins was determined using the BCA assay. For each sample, 800 μg of proteins were used for click chemistry (for samples that have higher protein concentration, the reaction volume was adjusted with click chemistry buffer). Biotin-N₃ (5 μL of 5 mM solution in DMF), Tris[(1-benzyl-1H-1,2,3-triazole-4-yl)methyl]amine (5 μL of 2 mM solution in DMF), CuSO4 (5 μL of 50 mM solution in H₂O) and Tris(2-carboxyethyl)phosphine (5 μL of 50 mM solution in H₂O) were added into the reaction mixture. The click chemistry reaction was allowed to proceed at 30°C for 60 min. Proteins were then precipitated using cold methanol, chloroform, and water as described above. After washing two times with cold methanol, the proteins were resolubilized in 120 μL of 4% SDS lysis buffer. Again, the concentrations of proteins were measured by BCA, and 10 μg of proteins for each sample were set aside as input. Then, 400 μg of proteins for each sample was added into 10 mL of IP washing buffer (25 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.2% (v/v) NP-40) (making sure that SDS is lower than 0.1% in IP washing buffer for efficient affinity purification with streptavidin beads). Streptavidin agarose beads (Thermo Fisher, 20 μl) were added into each sample. The mixture was agitated for 1 h at room temperature. After washing the beads three times with 1 mL of IP washing buffer, 24 μL of 4% (w/v) SDS lysis buffer and 6 μL of 6× loading dye were added into each sample and boiled for 10 min. The samples and inputs were resolved using 12% SDS-PAGE gel, transferred to nitrocellulose membrane, and analyzed by western blot for SHMT2. The signal was recorded using a Typhoon 9400 Variable Mode Imager (GE Healthcare Life Sciences).