Acetylation stabilizes ATP-citrate lyase to promote lipid biosynthesis and tumor growth

Abstract

Increased fatty acid synthesis is required to meet the demand for membrane expansion of rapidly growing cells. ATP-citrate lyase (ACLY) is upregulated or activated in several types of cancer, and inhibition of ACLY arrests proliferation of cancer cells. Here we show that ACLY is acetylated at lysine residues 540, 546, and 554 (3K). Acetylation at these three lysine residues is stimulated by P300/calcium-binding protein (CBP)-associated factor (PCAF) acetyltransferase under high glucose and increases ACLY stability by blocking its ubiquitylation and degradation. Conversely, the protein deacetylase sirtuin 2 (SIRT2) deacetylates and destabilizes ACLY. Substitution of 3K abolishes ACLY ubiquitylation and promotes de novo lipid synthesis, cell proliferation, and tumor growth. Importantly, 3K acetylation of ACLY is increased in human lung cancers. Our study reveals a crosstalk between acetylation and ubiquitylation by competing for the same lysine residues in the regulation of fatty acid synthesis and cell growth in response to glucose.

Figure 1. ACLY Is Acetylated at Lysines 540, 546, and 554

(A) Exogenous ACLY is acetylated. Ectopically expressed ACLY protein and acetylation levels were determined by western blot. Relative acetylation/protein ratios were quantified. Error bars represents ±SD for experiments performed in triplicate. (B) Endogenous ACLY is acetylated. Endogenous ACLY protein was purified from HEK 293T cells after NAM and TSA treatment as indicated. Acetylation levels were analyzed by western blot. (C) Mutations of 3K decrease ACLY acetylation. Acetylation of ectopically expressed ACLY^WT, ACLY^3KQ, and ACLY^3KR was analyzed, and relative acetylation was quantified. (D) NAM and TSA increases ACLY acetylation at 3K. A549 cells were treated with NAM and TSA. Acetylation was detected by α-Ac(3K) antibody. (E) Glucose increases both ACLY protein and 3K acetylation levels in A549 cells. A549 cells were maintained under various glucose concentrations, and endogenous ACLY protein and acetylation levels were determined against b-actin and ACLY protein. In order to measure ACLY protein levels, the loading was normalized to total protein as indicated by the actin western blot (upper left panels). In order to measure relative ACLY acetylation, the loading was normalized to ACLY protein levels as indicated by the ACLY western blot (lower left panels). (F) Glucose injection increases endogenous ACLY acetylation and protein levels in mice liver. BALB/c male mice were injected with PBS or 200 mg/g glucose into enterocoelia. Western blots for ACLY protein and acetylation were performed similar to that in (E). Endogenous ACLY protein levels and ACLY acetylation in mice livers at different time points were determined by western blot. ACLY mRNA levels were quantified by qPCR. Blood glucose levels were determined by ACCUCHEK Active (Roche). (G) Endogenous ACLY is accumulated by treatment of proteasome inhibitor MG132. HeLa cells were treated with or without 10 mm MG132 for the indicated times under low glucose. Endogenous ACLY level was determined by western blot. (H) High glucose stabilizes endogenous ACLY. A549 cells were treated with CHX maintained in 2.5 mM (upper panels) or 25 mM (lower panels) glucose for various time points. Endogenous ACLY protein level was analyzed by western blot. (I) ACLY^3KR is more stable than ACLY^WT under low glucose. A549 cells with stable ACLY knockdown and re-expression of the Flag-tagged shRNA-resistant wild-type and 3KR mutant were treated with CHX for the indicated time points. Stabilities of ACLY^WT and ACLY^3KR protein were determined by western blot and quantified. Error bars represents ±SD for experiments performed in triplicate.

Figure 2. Acetylation Protects ACLY from Proteasome Degradation by Inhibiting Ubiquitylation

(A) ACLY is ubiquitylated in vivo. Flag-tagged ACLY was cotransfected with HA-tagged ubiquitin into HEK 293T cells. Ubiquitylation of immunoprecipitated ACLY was determined. (B) Mutations of the acetylation lysine residues decrease ACLY ubiquitylation. Flag-tagged wild-type and 3KR mutant ACLY were cotransfected with HA-tagged ubiquitin into HEK 293T cells, and ubiquitylation of purified proteins was determined. (C) Deacetylase inhibitors NAM and TSA decrease wild-type ACLY ubiquitylation, but not 3KQ or 3KR mutants. Flag-tagged WT, 3KQ, or 3KR of ACLY was cotransfected with HA-tagged ubiquitin into HEK 293T cells with or without the treatment of NAM and TSA for indicated times. Ubiquitylation of purified proteins was analyzed by western blot. (D) 3K residues are the potential competition sites of ubiquitylation and acetylation. K540/546R, K546/554R, and K546/554R left with one lysine were cotransfected with Flag-tagged ubiquitin. The anti-Flag-UB immunoprecipitates (input, supposed to be all ubiquitylated substrate mixture) and the remaining supernatants (output, supposed to be nonubiquitylated substrate mixture) were blotted to compare the ACLY 3K acetylation level. Acetylation at 3K was detected in the output (nonubiquitylated ACLY), but not in the input (ubiquitylated ACLY). (E) NAM and TSA reduce ACLY ubiquitylation. Flag-tagged ACLY was cotransfected with HA-tagged ubiquitin into HEK 293T cells maintained under 2.5 mM or 25 mM glucose with or without NAM and TSA as indicated. Ubiquitylation of purified proteins was analyzed. (F) High glucose decreases acetylation of wild-type ACLY, but not 3KQ or 3KR mutants. Flag-tagged WT, 3KQ, or 3KR of ACLY was cotransfected with HA-tagged ubiquitin into HEK 293T cells maintained under 2.5 mM or 25 mM glucose concentrations. Ubiquitylation of purified proteins and acetylation were determined by western blot.

Figure 3. UBR4 Is the E3 Ligase of ACLY

(A) The interaction between E3 ligase domain of UBR4, noted as UBR4(D), and ACLY is enhanced by the treatment of MG132. Flag-tagged ACLY and Myc-tagged UBR4(D) was coexpressed into HEK 293T cells maintained in the presence or absence of MG132. The interaction between ACLY and UBR4(D) was determined by immunoprecipitation and western blotting (IP-western). (B) The interaction between UBR4(D) and ACLY is decreased under high glucose. Flag-tagged ACLY and Myc-tagged UBR4(D) were coexpressed into HEK 293T cells maintained in 2.5 mM or 25 mM glucose for indicated times. The interaction between ACLY and UBR4(D) was determined by IP-western. (C) UBR4 knockdown increases steady-state ACLY level. UBR4 in A549 cells was knocked down by siRNA, and endogenous ACLY level was determined and quantified. (D) Knocking down UBR4 stabilizes ACLY. A549 cells with or without siUBR4 were treated with CHX for indicated time points. Stability of endogenous ACLY protein level was determined by western blot. The knockdown efficiency of UBR4 was measured by qPCR with ±SD. (E) Knockdown of UBR4 decreases ACLY ubiquitylation. HEK 293T cells with or without siUBR4 were transfected with Flag-tagged ACLY and HA-tagged ubiquitn. The ubiquitylation of ACLY was determined by western blot. UBR4 knockdown efficiency was measured by qPCR with ±SD.

Figure 4. PCAF Is the Acetylase of ACLY

(A) Overexpression of PCAF, but not other acetyltransferases (HATs), increases endogenous ACLY acetylation at 3K. Flag-tagged PCAF, P300, or CBP or Myc-tagged GCN5 was tranfected individually into HEK 293T cells. Endogenous acetylation levels of ACLY were determined by anti-3K (Ac) antibody. (B) ACLY-PCAF interaction is enhanced under high glucose. Myc-tagged ACLY and Flag-tagged PCAF were coexpressed into HEK 293T cells maintained in 2.5 mM or 25 mM glucose for indicated times. The interaction between ACLY and PCAF was determined by IP-western. (C) PCAF knockdown decreases endogenous ACLY acetylation at 3K and its protein level. A549 cells were transfected with siPCAF or control as previously described, and endogenous ACLY acetylation at 3K was measured by western blot against ACLY protein level (upper panel). ACLY transcription level was measured by qPCR (lower panel). Endogenous ACLY levels with or without PCAF knockdown were determined by anti-ACLY antibody against b-actin. (D) PCAF knockdown decreases ACLY protein stability. A549 cells were transfected with siPCAF or control as previously described and were treated with CHX for the indicated times. The endogenous ACLY protein was determined and quantified by western blot against b-actin. (E) Overexpression of PCAF decreases ACLY ubiquitylation. Flag-tagged PCAF was cotransfected with Flag-tagged ACLY and HA-tagged ubiquitin. The ubiquitylation of ACLY was determined by western blot. (F) PCAF inhibition increases the interaction between UBR4(D) and wild-type ACLY, but not 3KR. Myc-tagged UBR4(D) and Flag-tagged wild-type and 3KR of ACLY were cotransfected into HEK 293T cells treated with high (25 mM) glucose. The interaction between ACLY and UBR4(D) was determined by western blot. PCAF knockdown efficiency is shown on the right.

Figure 5. SIRT2 Decreases ACLY Acetylation and Increases Its Protein Levels In Vivo

(A) NAM, but not TSA, increases ACLY acetylation. Flag-tagged ACLY was transfected into HEK 293T cells with or without sirtuin deacetylase inhibitor NAM and HDAC inhibitor TSA. Acetylation of ACLY was measured by western blot. (B) SIRT2 overexpression decreases ACLY acetylation. Flag-tagged ACLY was cotransfected with either HA-tagged SIRT1 or SIRT2 into HEK 293T cells. Acetylation of purified proteins was determined by western blot. (C) Catalytic activity of SIRT2 is required for the deacetylation of ACLY. Flag-tagged ACLY was cotransfected with HA-tagged SIRT2 WT or catalytically inactive mutant H187Y into HEK 293T cells. Acetylation was determined by western blot. (D) SIRT2, but not its catalytic mutant, promotes ACLY ubiquitylation. Flag-tagged ACLY and HA-tagged ubiquitin were cotransfected with Myc-tagged SIRT2 or SIRT2^H187Y into HEK 293T cells. Ubiquitylation of purified proteins was detected. (E) Inhibition of SIRT2 with inhibitor increases endogenous ACLY protein level. HEK 293T cells were treated with or without SIRT2 inhibitor salermide. Endogenous ACLY protein levels were determined by western blot, and relative ACLY mRNA levels were quantified by qPCR. (F) Sirt2 knockdown increases endogenous ACLY protein level in mice liver. Livers from BALB/c mice with tail vein injection of control and Sirt2 shRNA were prepared. The Sirt2 knockdown efficiency was determined by qPCR. Endogenous ACLY protein levels were determined by western blot. (G) High glucose decreases the interaction between ACLY and SIRT2. Flag-tagged ACLY were cotransfected with HA-tagged SIRT2 into HEK 293T cells. Protein interactions were determined. (H) SIRT2 is required for glucose-regulated ACLY acetylation. HEK 293T cells transfected with or without siSIRT2 were cultured in medium containing 2.5 mM or 25 mM glucose. Endogenous ACLY acetylation was determined by western blot. SIRT2 knockdown efficiency was determined by qPCR. (I) Restoration of SIRT2 destabilizes ACLY. MEF cells with Sirt2 ^−/− or Sirt2 ^−/− MEF cells stably re-expressing Sirt2 were treated with CHX at the indicated time points. The stability of endogenous ACLY protein level was determined by western blot.

Figure 6. Acetylation of ACLY at 3K Promotes Lipogenesis and Tumor Cell Proliferation

(A) ACLY^3KR mutant promotes cell proliferation. ACLY was stably knocked down with shRNA in A549 cells. shRNA-resistant wild-type and 3KR mutant of ACLY were stably re-expressed in the knockdown cells to a level that is compatible with endogenous ACLY. ACLY knockdown efficiency and re-expression level were determined by western blot. ACLY^WT or ACLY^3KR cells were seeded at the same number in each well. Cell numbers were counted every 48 hr. Error bars represent cell numbers ±SD for experiments performed in triplicate. (B) ACLY^3KR increases cellular acetyl-CoA level under low glucose. Flag-tagged ACLY^WT and ACLY^3KR were transfected into Chang shACLY stable cell line maintained in 2.5 mM glucose for the indicated times. Total cellular acetyl-CoA was measured using PicoProbe Acetyl-CoA Assay Kit (Abcam). Error bars represent ±SD for experiments performed in triplicate. (C) The ACLY^3KR mutant-expressing cells show reduced uptake of extra phospholipids. Intracellular accumulation of phospholipids was measured using phospholipids conjugated to fluorescent dyes. (D) Intracellular accumulation of phospholipids is quantified. The fluorescence of intracellular phospholipids was measured at optical density 594 (OD₅₉₄) on a plate reader. Error bars represent ±SD. (E) ACLY^3KR mutant promotes xenograft tumor growth. A549 cell lines characterized in (A) were injected subcutaneously into the flanks of nude mice. At 7 weeks after injection, tumors from 8 mice were extracted and photographed. (F) Tumor growth curves in nude mice. Tumor diameters were measured at the indicated time points, and tumor volumes were calculated. Significant difference curves were evaluated by t test. (G) ACLY^3KR expression promotes tumor cell proliferation in vivo. Tumor sections from xenografts were prepared for IHC. All eight pairs of hematoxylin and eosin (H&E)- and Ki-67-stained ACLY^WT and ACLY^3KR were analyzed.

Figure 7. Acetylation of ACLY at 3K Is Upregulated in Human Lung Carcinoma

(A) Lung cancer clinical cases with an increase in ACLY protein. Human lung carcinoma samples paired with carcinoma tissue (shown as T) and adjacent normal tissue (shown as A) were lysed. The ACLY protein levels were compared against b-actin under western blot. (B) Clinical cases with increased ACLY acetylation at 3K in ACLY-upregulated lung cancer. ACLY-upregulated human lung carcinoma samples paired with carcinoma tissue (shown as T) and adjacent normal tissue (shown as A) were lysed. The acetylation level of ACLY at 3K was compared against ACLY protein by western blot. (C) Shown is a working model depicting how acetylation at 3K protects ACLY from ubiquitin-mediated proteasome degradation in response to high glucose. This upregulation of ACLY protein level contributes to de novo lipogenesis and promotes tumor cell proliferation.