Cytoskeletal association of ATP citrate lyase controls the mechanodynamics of macropinocytosis

Significance

The scavenging of extracellular macromolecules via macropinocytosis is a key adaptive mechanism that supports the metabolic fitness of cancer cells. Although the importance of macropinocytosis for sustaining tumor cell growth under nutrient-limiting conditions is well documented, less is known about the molecular mechanisms by which macropinocytosis is regulated. This study describes a previously uncharacterized dependence of macropinocytosis on the compartmentalized generation of acetyl-CoA through the association of ACLY with the actin cytoskeleton. This metabolic channeling process establishes a mechanistic framework for understanding actin remodeling events that drive macropinocytosis.

Abstract

Macropinocytosis is an actin-dependent mode of nonselective endocytosis that mediates the uptake of extracellular fluid-phase cargoes. It is now well recognized that tumor cells exploit macropinocytosis to internalize macromolecules that can be catabolized and used to support cell growth and proliferation under nutrient-limiting conditions. Therefore, the identification of molecular mechanisms that control macropinocytosis is fundamental to the understanding of the metabolic adaptive landscape of tumor cells. Here, we report that the acetyl-CoA–producing enzyme, ATP citrate lyase (ACLY), is a key regulator of macropinocytosis and describes a heretofore-unappreciated association of ACLY with the actin cytoskeleton. The cytoskeletal tethering of ACLY is required for the spatially defined acetylation of heterodimeric actin capping protein, which we identify as an essential mediator of the actin remodeling events that drive membrane ruffling and macropinocytosis. Furthermore, we identify a requirement for mitochondrial-derived citrate, an ACLY substrate, for macropinocytosis, and show that mitochondria traffic to cell periphery regions juxtaposed to plasma membrane ruffles. Collectively, these findings establish a mode of metabolite compartmentalization that supports the spatiotemporal modulation of membrane–cytoskeletal interactions required for macropinocytosis by coupling regional acetyl-CoA availability with dynamic protein acetylation.

Engagement of nutrient scavenging pathways by oncogenic signaling is a key adaptive strategy that confers metabolic flexibility to cancer cells (1). This has been well described in the context of oncogenic Rat Sarcoma virus (RAS)-expressing cancer cells, which up-regulate macropinocytosis, a mechanism of nonselective, fluid-phase endocytosis (2–5). Macropinocytosis is initiated by the formation of membrane protrusions called membrane ruffles, a process dependent on the coordinated polymerization, cross-linking, and branching of cortical filamentous actin (F-actin) (6). As these membrane protrusions extend outward and collapse back onto the plasma membrane, they envelop extracellular fluid-phase cargoes into endocytic vesicles called macropinosomes (7). Cancer cells use macropinocytosis to scavenge extracellular macromolecules and necrotic cell debris, that are subsequently catabolized in the lysosome to yield anabolic substrates that fuel cell growth and proliferation (7). The importance of this nutrient scavenging pathway for the metabolic fitness of cancer cells is evidenced by studies demonstrating that genetic or pharmacological blockade of macropinocytosis in vivo attenuates tumorigenesis (3–5, 8). As such, there is a growing interest in defining and characterizing the key regulatory steps that control this process. Although progress has been made toward the identification of effectors and regulators of macropinocytosis (4, 5, 8–12), the interdependence between cellular bioenergetics and macropinocytosis remains poorly understood.

In this study, we demonstrate that macropinocytosis is dependent on ATP citrate lyase (ACLY), an enzyme that catalyzes the conversion of mitochondrial-derived citrate and coenzyme A (CoA) to oxaloacetate and acetyl-CoA (13). We provide evidence that ACLY is an F-actin-associated protein and describe a link between ACLY-mediated production of acetyl-CoA and the regulation of actin cytoskeleton dynamics that are required for membrane ruffle formation. The association between F-actin and ACLY defines a mechanism for the compartmentalized production of acetyl-CoA and establishes a functional interdependence between specific intracellular metabolic intermediates and nutrient supply via macropinocytosis.

Results

ACLY Activity Is Required for Macropinocytosis.

To explore the relationship between macropinocytosis and cellular bioenergetics, we have assessed the effects on macropinocytosis of acute pharmacological modulation of biosynthetic pathways that are critical for cancer cell fitness. This analysis revealed that treatment of cells with the ACLY inhibitor BMS-303141 (14, 15) caused a marked reduction in the uptake of dextran (a selective marker of macropinocytosis) in a panel of macropinocytic RAS mutant cell lines (Fig. 1 A and B). A similar inhibitory effect was displayed by a structurally unrelated, allosteric inhibitor of ACLY, NDI-091143 (16), indicating that ACLY activity might be required for macropinocytosis (Fig. 1 C and D). While the role of ACLY-dependent acetyl-CoA production in promoting lipogenesis and maintaining global histone acetylation is well established (17–20), a link between ACLY activity and the regulation of endocytic processes, such as macropinocytosis, has not been described. We, therefore, sought to further validate this link using a genetic approach involving the generation of ACLY knockout (KO) clones of T24 cells by CRISPR/Cas9 genome editing (SI Appendix, Fig. S1 A–C). Previous reports have demonstrated that acyl-CoA synthetase short-chain family member 2 (ACSS2) is able to compensate for genetic deletion of ACLY by synthesizing acetyl-CoA from acetate (15, 21) (Fig. 1E). Therefore, we assessed the macropinocytic capacity of ACLY WT and KO clones in the presence and absence of exogenous acetate. In contrast to ACLY WT clones, which displayed comparable levels of dextran uptake independently of acetate, ACLY KO clones showed impaired dextran uptake following acetate withdrawal (Fig. 1 F–H). Additionally, reexpression of WT ACLY, but not catalytically inactive ACLY (H760A) (15), was sufficient to restore dextran uptake in ACLY KO cells, rendering them insensitive to acetate withdrawal (Fig. 1 I–K). Taken together, these results uncover a previously uncharacterized dependence of macropinocytosis on ACLY activity and acetyl-CoA production.

Fig. 1.

ACLY activity regulates macropinocytosis. (A and B) Fluorescence micrographs (T24) and quantification of tetramethylrhodamine (TMR) dextran uptake in cell lines treated with ACLY inhibitors (1 h) BMS-303141 and (C and D) NDI-091143. (E) Schematic showing the major intracellular sources of acetyl-CoA. ACLY uses mitochondrial-derived citrate and CoA to synthesize acetyl-CoA and oxaloacetate (OAA). ACSS2 utilizes acetate and CoA to synthesize acetyl-CoA. (F) Fluorescence micrographs and quantification of TMR dextran uptake in (G) ACLY WT and (H) ACLY KO T24 clones in the presence or absence of sodium acetate. (I) Fluorescence micrographs and (J) quantification of TMR dextran uptake in ACLY KO T24 cells reconstituted with empty vector, WT ACLY, or H760A (catalytically inactive mutant) ACLY in the presence or absence of sodium acetate. (K) Immunoblot showing ACLY expression in ACLY KO T24 cells reconstituted with empty vector, WT ACLY, or H760A ACLY. α-tubulin was used as a loading control. All images and immunoblots are representative. (Scale bar for all images, 10 µm.) For TMR dextran uptake assays, at least 500 cells were counted per biological replicate (n = 3 to 4). Data are represented as mean ± SEM. **P < 0.01, ***P < 0.001, ns = not significant (unpaired, two-tailed, Student’s t test).

ACLY Associates with F-actin.

Many of the known regulators of macropinocytosis (e.g., V-ATPase, RAC1, RAS, SDC1, SLC4A7, and PI3K) have been shown to localize to the plasma membrane (7). Therefore, as an initial step toward defining the mechanistic basis for the regulatory role of ACLY in macropinocytosis, we examined its subcellular distribution using immunofluorescence microscopy. Consistent with prior reports, we found that ACLY displayed a prominent nucleocytosolic localization (22) (Fig. 2A). In addition, the staining pattern revealed a previously uncharacterized spatial distribution of ACLY, namely its localization to the cell periphery. The specificity of the staining was validated using ACLY WT and KO cells (Fig. 2B). To further characterize the nature of the observed peripheral ACLY staining pattern, membrane ruffles were labeled with phalloidin, which selectively binds F-actin (an obligatory constituent of membrane ruffles) (6). Confocal microscopy analysis revealed the presence of ACLY in phalloidin-labeled membrane ruffles in multiple cell lines (Fig. 2C). The lack of apparent membrane targeting sequences on ACLY and the high degree of spatial overlap between ACLY and cortical F-actin (Fig. 2C) raised the possibility that ACLY is associated with the actin cytoskeleton. To test this idea, detergent fractionation was used to extract soluble, cytosolic proteins and lipid-based organelles while retaining the F-actin cytoskeleton in the insoluble fraction. Immunoblot analysis of lysates from detergent-extracted cells demonstrated the presence of ACLY in the detergent-insoluble fraction in multiple cell lines (Fig. 2D and SI Appendix, Fig. S2A). Loss of α-tubulin from the detergent-insoluble fraction was used to assess extraction efficiency and fraction purity. Confocal immunofluorescence analysis of ACLY in detergent-extracted cells confirmed the presence of a detergent-resistant pool of ACLY that colocalized with F-actin (Fig. 2E and SI Appendix, Fig. S2 B and C). Importantly, this staining pattern was lost in detergent-extracted cells following treatment with the F-actin depolymerizing agent latrunculin A (Lat A), indicating that the retention of ACLY in the detergent-insoluble fraction is mediated by its interaction with F-actin (Fig. 2E and SI Appendix, Fig. S2 B and C).

Fig. 2.

ACLY associates with F-actin. (A) Fluorescence micrographs (epifluorescence microscopy) of cells immunostained for ACLY showing nucleocytosolic and peripheral labeling of ACLY. (B) Fluorescence micrographs (epifluorescence microscopy) of ACLY WT and KO T24 cells immunostained for ACLY showing antibody specificity. (C) Fluorescence micrographs (confocal microscopy) of cells immunostained for ACLY showing colocalization between ACLY and F-actin (phalloidin). (D) Immunoblot analysis of ACLY in detergent-soluble (S) and -insoluble (I) fractions in detergent-extracted cells. α-tubulin was used to assess fraction purity and β-actin was used as a loading control. (E) Fluorescence micrographs (confocal microscopy) of ACLY immunostaining and colabeled with phalloidin in detergent-extracted T24 cells treated with Lat A (1 h). (F) Immunoblots of biotinylated phalloidin F-actin pulldown in HeLa-KRAS(G12V) and T24. ARP3 and ACTN1 (known actin-binding proteins) and α-tubulin and proliferating cell nuclear antigen (PCNA) (nonactin-binding proteins) were used as positive and negative controls, respectively. (G) Immunoblot analysis of supernatant and pellet fractions of actin cosedimentation assay using purified F-actin incubated with either recombinant ACLY tetramer, recombinant alpha-actinin (ACTN) as a positive control, or purified ALB (albumin) as a negative control. Reactions without F-actin serve as background sedimentation controls. (Scale bar for all images, 10 µm.) All images and immunoblots are representative.

To test this idea, we first utilized a biotinylated phalloidin pulldown assay in which F-actin was isolated from cell lysates, and the associated proteins were detected by immunoblotting. Using this approach, we found that ACLY, as well as other known F-actin–binding proteins (ARP3 and ACTN1), copurified with F-actin (Fig. 2F). Non-F-actin–binding proteins (α-tubulin and PCNA) served as negative controls (Fig. 2F). In order to test whether the association of ACLY with F-actin is mediated through a direct interaction, we performed an actin cosedimentation assay using purified, polymerized F-actin and recombinant ACLY protein. From this analysis, we were able to detect ACLY in the pellet fraction of the binding reaction, only in the presence of F-actin, providing strong evidence for a direct interaction between ACLY and F-actin (Fig. 2G). Alpha-actinin and albumin served as positive and negative controls for F-actin binding, respectively (Fig. 2G). While the molecular details of the interactions between ACLY with F-actin remain to be determined, we have found that ectopically expressed, catalytically inactive ACLY (H760A) retained the ability to colocalize with F-actin, indicating that this association was not dependent on the enzymatic activity of ACLY (SI Appendix, Fig. S2 D and E). In addition, the colocalization of ACLY with F-actin was maintained in mutant KRAS-expressing cells following Kirsten RAS (KRAS) knockdown (SI Appendix, Fig. S3 A and B) and in wild type (WT) RAS-expressing cells (SI Appendix, Fig. S3C). Of note, protein levels of ACLY were also unaffected by oncogenic KRAS expression (SI Appendix, Fig. S3 B and D). Furthermore, F-actin–associated ACLY was also observed in cells treated with a pan-PI3K inhibitor (BMK120) (SI Appendix, Fig. S3 E and F). Collectively, these findings identify a previously unappreciated subcellular location for ACLY that is mediated by its direct interaction with F-actin and is independent of RAS-mediated signaling mechanisms that control macropinocytosis.

Capping Protein (CP)-Dependent Actin Remodeling Regulates Macropinocytosis.

Next, we sought to investigate the mechanistic link between the cytoskeletal association of ACLY and the regulation of macropinocytosis. To this end, we employed a proximity-dependent biotin identification (BioID) proteomic screen. To capture both short- and long-range proximal proteins, we generated BirA-ACLY fusion proteins with different sized linkers (16 or 46 amino acids) (SI Appendix, Fig. S4A). Importantly, we confirmed that the localization of these fusion proteins mirrored that of endogenous ACLY, in terms of F-actin association, and production of comparable total levels of biotinylation, relative to BirA control (SI Appendix, Fig. S4 B–D). In order to enrich for cytoskeletal-associated proteins, the screen was performed by affinity-purifying biotinylated proteins from detergent-extracted cells expressing BirA or BirA-ACLY. In addition to previously reported ACLY interaction partners such as GSK3β, PKA, NDPK, and SEC16A (20, 23), several heretofore-undocumented F-actin–binding proteins were among the screen hits (Fig. 3A and Dataset S1). Furthermore, gene ontology analysis revealed a specific enrichment of actin-dependent cellular components, consistent with actin cytoskeleton association (SI Appendix, Fig. S4E).

Fig. 3.

Actin CP controls actin dynamics and regulates macropinocytosis. (A) Gene ontology terms associated with BioID screen showing top actin-associated proteins among BirA-ACLY hits grouped by cellular component (Enrichr). Blue boxes indicate assignment to corresponding cellular component. (B) Coimmunoprecipitation (anti-FLAG) of endogenous ACLY with CP in HeLa-KRAS(G12V) cells. (C) Fluorescence micrographs of TMR dextran uptake in T24 cells and (D) quantification in CPβ-depleted cell lines (72 h knockdown). (E) Immunoblots showing CPβ knockdown in HeLa-KRAS(G12V) and T24 cells. α-tubulin was used as a loading control. (F and G) Fluorescence micrographs of F-actin (phalloidin) staining and quantification of membrane projections in T24 cells treated with siCPβ (72 h) or (H and I) ACLY inhibitors (1 h). (J) Live phase-contrast imaging of T24 cells treated with ACLY inhibitors (30 min), BMS-303141, and NDI-091143. Yellow arrows indicate regions of active membrane ruffling. The same cells were imaged pretreatment (vehicle) and posttreatment (inhibitor). (K) Quantification of membrane projections in ACLY KO T24 cells reconstituted with empty vector or WT ACLY in the presence or absence of sodium acetate. (L) Fluorescence micrographs of TMR dextran uptake and (M) quantification in CPβ-depleted (72 h knockdown) ACLY KO T24 cells reconstituted with empty vector or WT ACLY in the presence or absence of sodium acetate. (Scale bar for all images, 10 µm.) All images and immunoblots are representative. For TMR-dextran uptake assays, at least 500 cells were counted per biological replicate (n = 3). For membrane projection quantification, projections were counted from 50 cells per condition (each dot represents one cell) over three independent experiments and normalized to plasma membrane perimeter (per 10 µm). Data are represented as mean ± SEM **P < 0.01, ***P < 0.001, ns = not significant (unpaired, two-tailed, Student’s t test).

In order to prioritize candidate proteins, we cross-referenced the BioID screen hits with genes from a genome-wide siRNA screen for macropinocytosis regulators previously performed by our group (24). Heterodimeric actin CP subunit beta (CPβ, encoded by CAPZB) was among the top hits from the siRNA screen (positive regulator of macropinocytosis) and was also detected in the BioID screen. Therefore, CPβ was selected as a candidate for further investigation. In line with BioID identification, we found that immunoprecipitates of ectopically expressed heterodimeric CP (achieved by endogenous coexpression of the CPα and CPβ subunits) contained endogenous ACLY, providing further evidence for their spatial proximity (Fig. 3B). CP, which binds to the ends of actin filaments and terminates polymerization, controls the architecture of the actin cytoskeleton by fine-tuning filament length and branching and regulating monomeric G-actin availability (25). Dysregulated F-actin capping, either by unrestrained capping (which causes attenuated polymerization) or insufficient capping (which leads to unrestricted polymerization and G-actin depletion), results in impairment in the cell’s ability to generate F-actin–driven protrusive membrane forces, a critical cytoskeletal remodeling determinant of membrane ruffles (25). In accordance with this postulated role, knockdown of CPβ potently inhibited dextran uptake (Fig. 3 C–E and SI Appendix, Fig. S5 A and B). This was accompanied by loss of the lamellar architecture of membrane ruffles and the pronounced accumulation of thin projections of plasma membrane, a morphological feature previously shown to be induced by loss of CP function (26) (Fig. 3 F and G). Significantly, similar phenotypic changes in membrane architecture were observed following pharmacological inhibition of ACLY (Fig. 3 H–J). These effects were also recapitulated in ACLY KO cells following acetate withdrawal and could be rescued by reexpression of WT ACLY (Fig. 3K). Together, these findings indicate that ACLY activity may regulate macropinocytosis through modulation of F-actin capping dynamics that are required for membrane ruffle formation. Consistent with this idea, we found that neither reconstitution of ACLY nor acetate supplementation was sufficient to restore dextran uptake in ACLY KO cells in the setting of CPβ depletion (Fig. 3 L and M and SI Appendix, Fig. S5C). These complementation experiments place CPβ as an essential downstream mediator of ACLY- and acetyl-CoA–dependent regulation of macropinocytosis.

ACLY Controls CP Acetylation.

We next set out to determine the mechanistic basis for the functional interdependence between acetyl-CoA and CP in the context of the regulation of macropinocytosis. In agreement with data from proteome-wide acetylation screens demonstrating that CPβ is acetylated at multiple lysine residues (27–29), we detected acetylated CPβ using an antibody that recognizes acetylated lysine (AcK) residues (Fig. 4A). The abundance of acetylated CPβ was significantly increased following treatment with pan-histone deacetylase inhibitors [trichostatin A (TSA) and nicotinamide (NAM)], confirming specificity of the anti-AcK antibody (Fig. 4A). The requirement of ACLY for the maintenance of CPβ acetylation was established by demonstrating that treatment of cells with ACLY inhibitors led to a significant reduction in the levels of acetylated CPβ (Fig. 4B). Furthermore, disruption of the F-actin cytoskeleton network using Lat A, which results in the preferential loss of F-actin–associated ACLY (Fig. 2E), caused a significant reduction in CPβ acetylation, but had no effect on histone acetylation (Fig. 4C). The dependence of CPβ acetylation on cytoskeletal-associated ACLY suggests that the anchoring of ACLY to cortical F-actin could serve as a mechanism for the tight spatiotemporal regulation of actin cytoskeletal dynamics through the reversible modulation of protein acetylation. Using mass spectrometry analysis to verify CPβ acetylation status and identify acetylation sites, we detected several tryptic peptides with lysine acetylation corresponding to positions K78, K199, K235, K237, and K254 (Fig. 4D). Importantly, the latter four residues are located at the C terminus of CPβ (Fig. 4E), which has been shown to be essential for its binding to F-actin (30–32).

Fig. 4.

ACLY modulates actin CP acetylation. (A) Immunoblot analysis of acetylated CPβ-FLAG immunoprecipitated with anti-acetylated lysine antibody (anti-AcK) from HeLa-KRAS(G12V) cells treated with pan-histone deacetylase inhibitors TSA and NAM (5 h), (B) ACLY inhibitors BMS-303141 and NDI-091143 (2 h) or (C) Lat A (1 h). Ac-H3 (K9) is acetylated histone H3 (lysine 9). (D) Tryptic peptide sequences from mass spectrometry PTM analysis showing the positions of acetylated lysine residues (red) in CPβ-FLAG immunoprecipitated (anti-FLAG) from HeLa-KRAS(G12V) cells. (E) Structure of human CPβ protein (PyMOL) showing positions of acetylated lysine residues (red) identified from mass spectrometry PTM analysis in D. (F) Images of live-phase contrast/fluorescence microscopy of HeLa-KRAS(G12V) and T24 expressing Mito-DsRed (mitochondrial probe) showing mitochondria in proximity to membrane ruffles. (G) Fluorescence micrographs (confocal microscopy) of HeLa-KRAS(G12V) and T24 cells expressing Mito-DsRed immunostained for ACLY and colabeled with phalloidin (F-actin), showing mitochondria in proximity to membrane ruffle-associated ACLY. (H) Fluorescence micrographs of TMR dextran uptake in T24 cells and (I) quantification in cell lines treated with a CTP inhibitor, CTPI-2 (1.5 h). (J) Schematic of model showing mitochondria infiltration into actin-rich membrane ruffles. CTP-dependent export of citrate from membrane ruffle-localized mitochondria is used by actin-associated ACLY to locally generate acetyl-CoA that is required for the spatially defined acetylation of CP. Acetylation regulates the F-actin localization of CP, which controls actin polymerization and membrane ruffle formation. (Scale bar for all images, 10 µm.) All images and immunoblots are representative. For TMR dextran uptake assays, at least 500 cells were counted per biological replicate (n = 3). Data are represented as mean ± SEM. **P < 0.01, ***P < 0.001 (unpaired, two-tailed, Student’s t test).

A prerequisite for the proposed role of F-actin–associated ACLY in the localized production of acetyl-CoA is the channeling of its substrates (mitochondrial-derived ATP and citrate) into the vicinity of cortical ACLY. In fact, mitochondrial positioning at the leading edge of the cell has been shown to play an important role in supporting the ATP-dependent cytoskeleton rearrangements that drive cell migration (33, 34). In support of this idea, we observed by live cell imaging and immunofluorescence analysis that mitochondria traffic to the cell periphery toward areas of membrane ruffling (Fig. 4F and Videos S1 and S2) where they are found in close proximity to membrane ruffle-associated ACLY (Fig. 4G). Given that the utilization of mitochondrial-derived citrate by cytosolic ACLY requires its export via the mitochondrial citrate transporter protein (CTP, encoded by SLC25A1), we tested whether blocking mitochondrial citrate export affects macropinocytosis. In accordance with a requirement of macropinocytosis for mitochondrial-derived citrate, pharmacological inhibition of CTP significantly impaired dextran uptake (Fig. 4 H and I). These findings support a model wherein submembranous channeling of mitochondrial-derived citrate into cytoskeleton-localized ACLY might serve as a mechanism by which the acetylation-dependent actin dynamics that are required for membrane ruffling and macropinocytosis can be spatially fine-tuned (Fig. 4J).

Discussion

Macropinocytosis is now recognized as a central metabolic adaptive process that is utilized by tumor cells to scavenge extracellular macromolecules under nutrient-limiting conditions. In this study, we identify a previously unidentified step in the regulatory circuitry that controls macropinocytosis involving the localized production of acetyl-CoA by actin-associated ACLY. Our findings provide insights into the molecular underpinning of macropinocytosis and underscore the importance of understanding spatial determinants of the cellular machinery that drive this process.

Dynamic acetylation, which engenders a means of rapid and reversible modulation of protein activity and localization, is required for the regulation and execution of many fundamental biological processes including cell signaling, transcription, cell cycle regulation, and cytoskeleton organization (35). Consistent with the essential role of acetyl-CoA as the obligatory acetyl donor for acetylation reactions (36), the intracellular abundance of acetyl-CoA is a major determinant of global acetylation levels (18, 22). However, there is mounting evidence that local acetyl-CoA availability within defined cellular compartments contributes to the spatiotemporal control of acetylation (37–41). These observations suggest that regulation of the subcellular localization of acetyl-CoA–producing enzymes can be essential for specifying the spatiotemporal dynamics of protein acetylation.

In this study, we identify CPβ (the β subunit of heterodimeric CP) as a target for location-specific ACLY-dependent acetylation. CP has been shown to be required for the formation of many actin-driven cellular structures, including membrane ruffles, a function that is dependent on its ability to bind to the ends of actin filaments (25). There is evidence to suggest that several basic residues that lie within the F-actin–binding interface, which spans the C-terminal portion of the protein, promote the interaction of CP with F-actin (30–32). Of relevance, several of the CPβ acetylation sites identified in this study are located within this interface. Given that acetylation results in lysine charge neutralization, it is plausible that the acetylation of one or more of the lysine residues within the F-actin–binding interface might affect the electrostatic interactions between CPβ and F-actin. Indeed, acetylation of CP has been suggested to stimulate actin dynamics in phenylephrine-stimulated hypertrophying myocytes by promoting the dissociation of CP from F-actin (42). The identity of the lysine acetyltransferases (KATs) and lysine deacetylases (KDACs) that mediate the reversible acetylation of CP remains to be determined. Of note, several KDACs and KATs have been shown to localize to the actin cytoskeleton (42–44), suggesting that the control of actin assembly by dynamic acetylation can be spatially regulated at multiple levels.

While the association of ACLY and the actin cytoskeleton identified in the present study has been investigated specifically in the context of the F-actin–remodeling machinery that regulates membrane ruffling and macropinocytosis, its functional relevance might extend to other ACLY-mediated cellular processes. For example, nuclear F-actin has been shown to play an essential role in promoting the repair of DNA double-strand breaks (DSBs) (45, 46). In fact, many DNA repair factors have been shown to interact with F-actin, which has been suggested to contribute to their stabilization at sites of damage (47). Notably, nuclear ACLY has been implicated in the regulation of DNA repair through the localized production of acetyl-CoA, which promotes the acetylation of histones at DSBs and subsequently stimulates the recruitment of DNA repair factors (41). These findings suggest that the capacity of ACLY to interact with the actin cytoskeleton could be also critically important for the spatial orchestration of the DNA damage response.

Genetic and pharmacological blockade of ACLY has been shown to inhibit tumor growth in various in vivo cancer models (17, 20, 48). Thus far, the protumorigenic function of ACLY has largely been attributed to acetyl-CoA–dependent lipid biosynthesis and histone acetylation (13). By implicating ACLY in the regulation of macropinocytosis, our study suggests a modality by which ACLY might contribute to the metabolic fitness of tumor cells, namely by sustaining their capacity to support anaplerotic pathways via the macropinocytic uptake of extracellular proteins. Additionally, by virtue of playing a role in submembranous F-actin assembly, a process required for the formation of membrane ruffles that are known to generate the driving force for cell movement (49), ACLY likely serves as an essential component of the regulatory network that controls cancer cell migration and invasion. Future studies examining the mechanisms underlying the association of ACLY with F-actin, and their impact on spatially defined F-actin dynamics, will provide insights into the links between metabolic reprogramming and actin-dependent processes that drive cell transformation.

Methods

Reagents and DNA Constructs.

All reagents used were of analytical grade.tetramethylrhodamine (TMR)–Dextran (70 kDa) was purchased from Fina Biosolutions. BMS-303141 (SML0784), CTPI-2 (ENA018104423), puromycin (P8833), doxycycline (D9891), biotin (B4639), TSA (T8552), and NAM (N0636) were purchased from Sigma. NDI-091143 was purchased from Aobious (AOB17806). Lat A was purchased from Cayman Chemical (10010630). BKM120 was purchased from Selleckchem (S2247). Protease inhibitor cocktail was purchased from Roche (11697498001). The following primary antibodies were used in this study: anti-bovine albumin (A10-127, Bethyl Laboratories); anti–α-tubulin (clone B-5-1-2 – Sigma, T5168); anti-acetylated histone H3 (K9) (clone C5B11 – Cell Signaling, 9649); anti-acetylated lysine (Ac-K2-100) MultiMab (Cell Signaling, 9814); anti-ACLY (Sigma, HPA022434); anti-ACTN1 (Abclonal, A1160); anti-AKT (clone 40D4, Cell Signaling, 2920); anti-ARP3 (clone FMS338 – Abcam, ab49671); anti–β-actin (clone 13E5 – Cell Signaling, 4970); anti-CPβ (Bethyl Laboratories, A304-734A); anti-histone H3 (clone D1H2 – Cell Signaling, 4499); anti-KRAS (clone F234, Santa Cruz, sc-30); anti-MYC (clone 9B11, Cell Signaling, 2276); anti-proliferating cell nuclear antigen (PCNA) (clone PC10 – Cell Signaling, 2586); anti–phospho-AKT (Ser473) (clone 193H12, Cell Signaling, 4058); and anti-T7 (Novagen, 69522). For immunoblotting, the following secondary antibodies were used: Alexa Fluor 680 goat anti-mouse (Invitrogen, A21058); IRDye 800CW goat anti-rabbit (Li-Cor, 926-32211), horseradish peroxidase (HRP)-conjugated goat anti-rabbit (Cell Signaling, 7074); and HRP-conjugated goat anti-mouse (Cell Signaling, 7076). For immunofluorescence, Alexa Flour 488-conjugated goat anti-rabbit (Invitrogen, A11029), Alexa Flour 555-conjugated goat anti-rabbit (Invitrogen, A21429), or Alexa Flour 555-conjugated goat anti-mouse (Invitrogen, A21424) secondary antibodies were used. The following siRNAs were purchased from Dharmacon (Horizon):

nontargeting siRNA Pool #2 (D-001206-14): UAAGGCUAUGAAGAGAUAC; AUGUAUUGGCCUGUAUUAG; AUGAACGUGAAUUGCUCAA; UGGUUUACAUGUC​GACUAA.

siCAPZB (human) SMARTPool (M-011990-01): GAAGUACGCUGAACGAGAU; GGAGUGAUCCUCAUAAAGA; GAGACAAGGUGGUGGGAAA; CACCAUGGAGUAAC​AAGUA.

siKRAS (human) SMARTPool (M-005069-0010): CGAAUAUGAUCCAACAAUA; UAAGGACUCUGAAGAUGUA; GACAAAGUGUGUAAUUAUG; GCUCAGGACUUAGC​AAGAA.

pTRIPZ-KRAS(G12V) was constructed by cloning human, codon-optimized KRAS(G12V) with an N-terminal T7 tag into pTRIPZ (Dharmacon) using AgeI and MluI to simultaneously remove RFP. Human ACLY(H760A) was generated by site-directed mutagenesis using the pCMV6-ACLY(MYC-FLAG) vector (Origene, RC200508). Human CPα (CAPZA1, cloneID 3047937) and CPβ (CAPZB, clone ID 6715709) were purchased from Dharmacon and inserted into pcDNA3 (Invitrogen) with N-terminal MYC or FLAG tags. BirA-ACLY constructs were generated by inserting human ACLY into pcDNA3.1 mycBioID vector (N-terminal MYC-BirA, Addgene, plasmid 35700) with a 16 or 46 amino acid linker between ACLY and BirA. ACLY lentivirus constructs were generated by inserting ACLY variants with an N-terminal MYC tag into pLVX-IRES-Puro (Clontech, 632183). All DNA constructs were verified by Sanger sequencing. All primers used for cloning are shown in Table 1.

Table 1.

Cloning primer sequences

Construct (cDNA → vector)	Primer Sequences (5′-3′)
T7-KRAS(G12V) → pTRIPZ	F:AGCAACCGGTATGGCATCGATGACAGGTGGCCAACAGATGGGTACGGAATATAAGCTTGT R:TGCTGGACGCGTTTACATAATTACACACTTTG
MYC-ACLY → pLVX	F:ATTATAGAATTCATGGAGCAGAAACTCATCTCAGAAGAGGATCTGTCGGCCAAGGCAATTTCAGAGC R:AATATAGCGGCCGCTTACATGCTCATGTGTTCCGG
ACLY(H760A) mutagenesis (pCMV6)	F:GAGGTCCAGTTTGGCGCTGCTGGAGCTTGTGC R:GCACAAGCTCCAGCAGCGCCAAACTGGACCTC
MYC-BirA-ACLY(16) → pcDNA3	F:AATATAGAATTCGGTGGAGGCTCGGCCAAGGCAATTTCAGR:AATATAAAGCTTTTACATGCTCATGTGTTCCGG
MYC-BirA-ACLY(46) → pcDNA3	F:AATATAGAATTCGGTGGAGGCGGGTCTGGAGGCGGGGGTAGTGGCGGGGGTGGAAGCGGGGGTGGAGGCGGGTCGGGTGGCGGAGGTAGCGGAGGCGGTGGAAGTGGAGGCTCGGCCAAGGCAATTTCAG R:AATATAAAGCTTTTACATGCTCATGTGTTCCGG
MYC-CPα → pcDNA3	F:ATATAGGATCCATGGAGCAGAAACTCATCTCAGAAGAGGATCTGGCCGACTTCGATGATC R:AATATAGAATTCTTAAGCATTCTGCATTTCTTTGCC
FLAG-CPα → pcDNA3	F:ATATAGGATCCATGGATTACAAGGATGACGACGATAAGGCCGACTTCGATGATCGTGTG R:AATATAGAATTCTTAAGCATTCTGCATTTCTTTGCC
MYC-CPβ → pcDNA3	F:ATATAGGATCCATGGAGCAGAAACTCATCTCAGAAGAGGATCTGAGTGATCAGCAGCTGG R:AATATAGAATTCTTAGCATTGCTGCTTTCTCTTCAAAGC
FLAG-CPβ → pcDNA3	F:ATATAGGATCCATGGATTACAAGGATGACGACGATAAGAGTGATCAGCAGCTGGACTG R:AATATAGAATTCTTAGCATTGCTGCTTTCTCTTCAAAGC

Cell Culture, Treatment Conditions, and Generation of Stable Cell Lines.

All cell lines (HeLa, A549, Calu-6, DLD-1, HCT-116, MIA PaCa-2, T24, and 293T) were purchased from the American Type Culture Collection and routinely tested negative for Mycoplasma. The cells were maintained at 37 °C with 5% CO₂ in high-glucose Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, 11965-092) supplemented with 1% penicillin/streptomycin (Gibco, 15140-122) and 10% FBS (Gibco, 10438-026). Inducible HeLa-KRAS(G12V) cells were generated by transducing HeLa cells with pTRIPZ-T7-KRAS(G12V) lentiviral particles followed by selection with puromycin (2 µg/mL) for 72 h. HeLa-KRAS(G12V) cells were maintained in 10% tetracycline-free FBS (Takara Bio, 631101), and KRAS(G12V) expression was induced using doxycycline (250 ng/mL for 48 h). Reconstitution of ACLY in ACLY KO T24 cells was performed by pooling all four KO clones in equal ratios and infecting with pLVX-IRES-Puro-(MYC-ACLY) lentiviral particles; 48 h after transduction, the cells were selected with puromycin (2 µg/mL) for 72 h. Cells stably expressing the Mito-DsRed mitochondrial probe were generated by infecting cells with pLV-mitoDsRed (Addgene, plasmid 44386) lentivirus particles with puromycin (2 µg/mL) selection for 72 h. The lentivirus particles were packaged in 293T cells by cotransfecting carrier plasmids with psPAX (Addgene, plasmid 12260) and psPAX2 (Addgene, plasmid 12260). Virus-containing supernatant was concentrated using Amicon Ultra-15 100kDa MWCO centrifugal filters (Millipore, UFC910024). For siRNA knockdown experiments, siRNA (20 nM) was complexed with Lipofectamine RNAiMax Transfection Reagent (Thermo Fisher Scientific, 13778075) in OptiMEM (Thermo Fisher Scientific, 31985062) according to the manufacturer’s instructions. Transfection complexes were added to cells in complete media for 6 h before replacing media. For DNA transfections, DNA was complexed with X-tremeGENE 9 DNA Transfection Reagent (Sigma, 6365779001) in OptiMEM according to the manufacturer’s instructions. Unless otherwise indicated, drug treatment conditions were maintained the same throughout experimental assays. For BMS-303141 (5 µM) and NDI-091143 (10 µM), cells were treated with inhibitor for 1 h in serum-free DMEM after 4 h of serum starvation. For BKM120 (5 µM), cells were treated with inhibitor for 3 h in serum-free DMEM after 4 h of serum starvation. For CTPI-2 (50 µM), cells were treated for 90 min in serum-free DMEM after 4 h of serum starvation. For anti-AcK immunoprecipitation, cells were treated with TSA (2.5 µM) and NAM (5 mM) for 5 h in serum-free DMEM. For acetate withdrawal experiments, ACLY WT and KO cells were cultured in complete DMEM containing 10% dialyzed FBS with or without sodium acetate (1 mM) for 5 h followed by 3 h serum starvation with or without sodium acetate (1 mM). Lat A (200 nM) was used for 1 h in serum-free media.

CRISPR/Cas9 Genome Editing.

An sgRNA construct targeting exon 2 of the human ACLY gene was made by inserting annealed, phosphorylated oligonucleotides (5′-CACCGGAATCGGTTCAAGTATGCTC-3′, 5′-AAACGAGCATACTTGAACCGATTCC-3′) into pSpCas9(BB)-2A-Puro (PX459, Addgene, plasmid 48139). The sgRNA construct was then transfected (2 µg) into T24 cells using Lipofectamine 3000 (Invitrogen, 100022050) according to the manufacturer’s instructions. Cells were selected with puromycin (2 µg/mL) for 48 h and then single cell sorted by FACS into 96-well plates. Nonedited clones (i.e., without mutations in ACLY) were derived from the puromycin-selected population and used as ACLY WT controls. ACLY WT and KO cells were maintained in 1 mM acetate. Clones were verified by Sanger sequencing of PCR products spanning the sgRNA target site generated using the following primers:

5′-CCTTCTGACCAGCTTCTCTCTCC-3′ (F) and 5′-GGCATCACCAACAAACCAATGGC-3′ (R).

Macropinosome Visualization and Quantification.

Macropinocytic index was quantified using dextran uptake as previously described (3). Briefly, cells were seeded onto glass coverslips and allowed to adhere for 48-72 h. The cells were serum starved for 4 h and treated as indicated before incubation with TMR-dextran (70 kDa) diluted in serum-free media (1 mg/mL) containing indicated treatments for 30 min at 37 °C. The cells were then fixed with 3.7% formaldehyde solution for 15 min at room temperature and stained with DAPI (1 µg/mL). Coverslips were mounted using antifade mounting media ( DAKO, S3023). Images were captured using a Ti2 Eclipse Epifluorescence Microscope (Nikon). Image analysis and quantification was performed in ImageJ (v1.53q) using raw data files as previously described (3).

Live Cell Imaging.

Cells were seeded into glass-bottom 6-well plates in FluoroBrite DMEM (Gibco, A1896701) and allowed to adhere for 48 h. The cells were serum starved for 4 h in a preequilibrated, humidified chamber (37 °C, 5% CO₂) of a Ti2 Eclipse Epifluorescence Microscope (Nikon). Phase-contrast and fluorescence images were captured every 10 s for 30 to 60 min using 60× or 100× oil objectives. For inhibitor experiments, drugs were added directly to the media (i.e., without replacing media) to minimize temperature and pH changes. Image brightness and contrast adjustments were performed in ImageJ (v1.53q) using raw data files.

Immunofluorescence.

Cells were seeded onto glass coverslips in 24-well plates. Then, 48 to 72 h after seeding, the cells were fixed with 3.7% formaldehyde solution for 15 min at room temperature followed by two washes in PBS. The cells were then incubated in blocking solution (5% normal goat serum in 0.1% Triton X-100/PBS) for 30 min at room temperature and incubated with primary antibodies (1:250, anti-ACLY and 1:500, anti-MYC) diluted in blocking solution overnight at 4 °C. The cells were then washed in PBS, incubated with secondary antibodies diluted 1:1,000 in blocking solution, washed in PBS, and stained with Alexa Fluor 488-conjugated phalloidin (Invitrogen, A12379) or Alexa Fluor 647-conjugated phalloidin (Abcam, ab176759) for 1 h in 1% BSA/PBS. Coverslips were mounted using antifade mounting media ( DAKO, S3023). For detergent extraction, prior to fixation, the cells were incubated on ice for 5 min in freshly prepared, ice-cold extraction buffer: 10 mM HEPES (pH 7.4), 50 mM NaCl, 2.5 mM MgCl, 300 mM sucrose, 0.5% Triton X-100, 1 mM SOV, 1 mM NaF, and protease inhibitor cocktail. The cells were then washed in extraction buffer (without Triton X-100) and fixed as above. Loss of α-tubulin from the detergent-soluble fraction, which occurs through cold-induced depolymerization of microtubules, was used as a control to assess fraction purity. Images were captured using an LSM510 META Confocal Microscope (Zeiss) or a Ti2 Eclipse Epifluorescence Microscope (Nikon). Brightness and contrast adjustments were performed in ImageJ (v1.53q) using raw data files. See “Reagents and DNA Constructs” for the list of primary and secondary antibodies used.

Immunoprecipitation and Immunoblotting.

For immunoprecipitation experiments, cells were seeded into 15 cm plates and transfected with indicated constructs using X-tremeGENE 9 (Sigma, 6365779001) according to the manufacturer’s instructions. The cells were washed with ice-cold PBS before being scrapped, pelleted, and lysed in lysis buffer: 50 mM Tris (pH 7.4), 250 mM NaCl, 1 mM EDTA, 1mM MgCl, 0.5% NP40, 1 mM DTT, 5 mM NAM, 2.5 µM TSA, 1 mM SOV, 1 mM NaF, and protease inhibitor cocktail. Lysates were then sonicated, cleared by centrifugation, and incubated overnight with anti-acetylated lysine antibody (1:100) or for 1 h with 50 µL prewashed α-FLAG M2 affinity gel agarose slurry (Sigma, A2220) on a rotator at 4 °C. For acetylated lysine IP, antibody capture was performed with Protein A Dynabeads (Invitrogen, 10001D) for 2 h on a rotator at 4 °C. The beads were washed five times in lysis buffer and then proteins eluted by boiling in sample buffer (40 mM Tris pH 6.8, 1% SDS, 5% β-mercaptoethanol, 7.5% glycerol) for 10 min. For whole cell lysates, cells were washed in PBS, scraped in sample buffer, and then lysates were sonicated and boiled for 10 min. Cell lysates were resolved by SDS-PAGE and electrotransferred onto nitrocellulose or polyvinylidene difluoride (PVDF) membranes, which were then blocked using 5% skim milk in TBS-T (0.1% Tween-20) for 1 h at room temperature. Membranes were then incubated with primary antibodies diluted in 5% BSA/TBS-T for either 1 h at room temperature or overnight at 4 °C with shaking. Following primary antibody incubations, the membranes were washed in TBS-T and then incubated for 1 h at room temperature with secondary antibodies diluted at 1:10,000 in blocking buffer. For detection of biotinylated proteins, HRP-conjugated streptavidin (Thermo Fisher Scientific, SA10001) diluted at 1:5,000 in 5% BSA/TBS-T was used. For HRP detection, blots were developed with SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Scientific, 34579) or SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific, 34095). Imaging of membranes was performed using an ImageQuant800 Analyzer (GE Healthcare) or an Odyssey Scanner (Li-Cor). For quantification of protein levels from immunoblots, Odyssey Software (Li-Cor) was used to measure fluorescence intensity of individual bands. See “Reagents and DNA Constructs” for the list of primary and secondary antibodies used.

Proteomics Sample Preparation and Mass Spectrometry.

For BioID screening, HeLa-KRAS(G12V) cells were grown in 15 cm dishes and transfected with pcDNA3.1-BirA-Myc (2 µg + 8 µg pcDNA3 as a carrier plasmid) or pcDNA3.1-BirA-Myc-ACLY (10 µg) using X-tremeGENE 9. At the time of transfection, cells were treated with doxycycline (250 ng/mL) and biotin (50 µM). Then, 30 h after transfection, the cells were washed with PBS and then incubated with ice-cold extraction buffer (freshly prepared) for 5 min on ice: 10 mM HEPES (pH 7.4), 50 mM NaCl, 2.5 mM MgCl, 300 mM sucrose, 0.5% Triton X-100, 1 mM SOV, 1 mM NaF, and protease inhibitor cocktail. The extracted cells were washed with extraction buffer (without Trixon X-100) and then scraped and lysed in Radio-Immunoprecipitation Assay (RIPA) buffer: 50 mM Tris-HCl (pH 7.4), 250 mM NaCl, 5 mM EDTA, 0.5% DOC, 0.1% SDS, 1% NP40, 1 mM SOV, 1 mM NaF, and protease inhibitor cocktail. Lysates were incubated for 30 min at 4 °C on a rotator and then cleared by centrifugation (10,000 × g). Cleared lysates were incubated with 300 µL prewashed NeutrAvidin Agarose slurry (Thermo Scientific, 29200) for 1 h at 4 °C on a rotator. Beads were washed twice in 5 mL of high-salt RIPA containing 500 mM NaCl (5 min at 4 °C on a rotator) and then twice with RIPA (5 min at 4 °C on a rotator). The beads were then resuspended in sample buffer (40 mM Tris pH 6.8, 1% SDS, 5% β-mercaptoethanol, 7.5% glycerol), passed through a column (Chromotek, sct-50) to remove beads, and submitted for mass spectrometry analysis. For CPβ-FLAG posttranslation modification (PTM) analysis, HeLa-KRAS(G12V) cells were grown in 15 cm dishes and transfected with pcDNA3-FLAG (15 µg) or pcDNA3-CPβ-FLAG (7.5 µg) and pcDNA3-CPα-MYC (7.5 µg) using X-tremeGENE 9. At the time of transfection, the cells were treated with doxycycline (250 ng/mL). Subsequently, 30 h after transfection, the cells were washed in PBS, collected, pelleted (2000 x g), and then resuspended in lysis buffer: 50 mM Tris (pH 7.4), 250 mM NaCl, 1 mM EDTA, 1 mM MgCl, 0.5% NP40, 5% glycerol, 1 mM DTT, 2 µM TSA, 5 mM NAM, 1 mM SOV, 1 mM NaF, and protease inhibitor cocktail. Lysates were incubated for 30 min on ice, sonicated, and cleared by centrifugation (10,000 × g). Cleared lysates were incubated with 300 µL prewashed α-FLAG M2 affinity gel agarose slurry and incubated for 1 h at 4 °C on a rotator, and then beads were washed three times in lysis buffer. Immunoprecipitated proteins were eluted with 0.5 mg/mL (diluted in lysis buffer) 3× FLAG peptide (Sigma, F4799) for 30 min on a rotator. Eluates were then centrifuged and supernatants passed through a column (Chromotek, sct-50) to remove residual beads before concentrating using 3 kDa MWCO Amicon Ultra-0.5 centrifugal filters (Millipore, UFC5003). Concentrated eluates were submitted for mass spectrometry analysis. Samples were prepared for mass spectrometry analysis by reducing with DTT (2 µL of 0.2 M) for 1 h at 57 °C and alkylation with iodoacetamide (2 µL of 0.5 M) for 45 min at room temperature in the dark. Proteins were resolved on a NuPAGE 4 to 12% Bis-Tris gel (Life Technologies) then excised and gel slices dehydrated with methanol 1:1 (v/v) and ammonium bicarbonate (100 mM) mixture before digestion with 200 ng trypsin (Promega) overnight at room temperature. Peptides were extracted by incubating gel slices with R2 POROS beads (Thermo Scientific) followed by acidification with 0.1% trifluoroacetic acid (TFA) and desalted using Sep-Pak C18 solid-phase extraction (Waters). Peptides were eluted with 40% acetonitrile in 0.5% acetic acid and then 80% acetonitrile in 0.5% acetic acid and concentrated using a SpeedVac and stored at −80 °C. The peptides were gradient eluted from the column directly into the Q Exactive mass spectrometer (Thermo Scientific) using a 1-h gradient with solvent A (2% acetonitrile, 0.5% acetic acid) and solvent B (80% acetonitrile, 0.5% acetic acid). High-resolution full MS spectra were acquired with a resolution of 45,000, an AGC target of 3 × 10⁶, a maximum ion time of 45 ms, and scan range of 400 to 1,500 m/z. Following each full MS, 20 data-dependent high-resolution higher-energy C-trap dissociation (HCD) MS/MS spectra were acquired. All MS/MS spectra were collected using the following instrument parameters: resolution of 15,000, AGC target of 1 × 10⁵, maximum ion time of 120 ms, 1 microscan, 2 m/z isolation window, fixed first mass of 150 m/z, and normalized collision energy (NCE) of 27. MS/MS spectra were searched against a Uniprot Human database using Proteome Discoverer 1.4 (for BioID screen) or against human CPβ protein using Byonic (for PTM analysis).

Biotinylated Phalloidin Pulldown.

72 h after seeding into 15 cm dishes, cells were washed twice with ice-cold PBS, scraped, and pelleted. Cell pellets were suspended in lysis buffer: 50 mM Tris (pH 7.4), 50 mM KCl, 150 mM NaCl, 1 mM EDTA, 1 mM MgCl2, 0.5% Triton X-100, 1 mM ATP, 1 mM DTT, 1 mM SOV, 1 mM NaF, and protease inhibitor cocktail. Lysates were incubated for 15 min on ice and then cleared by centrifugation (3,000 × g). Supernatant was collected and precleared with protein G agarose beads (Pierce, 20397) for 15 min at 4 °C on a rotator. Beads were removed by centrifugation (3,000 × g for 1 min) and lysates split equally and incubated with either methanol (vehicle) or 1 µM biotinylated phalloidin (Biotium, 00028) for 2 h at 4 °C on a rotator. The lysates were then incubated with 30 µL prewashed NeutrAvidin agarose resin (Thermo Scientific, 29200) for 1 h at 4 °C on a rotator. The beads were washed 5× in lysis buffer and then resuspended in sample buffer (40 mM Tris pH 6.8, 1% SDS, 5% β-mercaptoethanol, 7.5% glycerol), boiled for 10 min, and then subjected to SDS-PAGE and immunoblotting.

Actin Cosedimentation Assay.

Purified nonmuscle actin (Cytoskeleton Inc., APHL99) was polymerized according to the manufacturer’s instructions. Binding reactions (100 µL) were set up using purified, polymerized F-actin (3 µM) with either recombinant ACLY (0.25 µM), recombinant alpha-actinin (0.125 µM, Cytoskeleton Inc., AT01), or purified bovine albumin (2 µM, Thermo Scientific, 23209) in reaction buffer containing 20 mM Tris pH 7.5, 300 mM NaCl, and 2 mM DTT. Duplicate reactions were set up without F-actin to control for background protein sedimentation. Reactions were incubated at room temperature for 1 h and then centrifuged (100,000 × g at 23 °C) for 30 min in a TLA-100 Fixed-Angle Rotor Package (Beckman Coulter). Supernatant (10%) and pellet (25%) fractions were subjected to SDS-PAGE and immunoblotting. Recombinant ACLY tetramer was purified as previously described (16).

Statistical Analysis.

Unless otherwise stated, data in all graphs are represented as mean ± SEM of n = 3-4 biological replicates. Two-tailed, unpaired Student’s t tests (equal variance) were used for statistical comparisons between groups (GraphPad Prism v9.2.0). P-values and number of samples analyzed for each experiment are indicated in figure legends.

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Movie S1.

Live phase contrast/fluorescence microscopy of HeLa-KRAS(G12V) cells expressing Mito-DsRed (mitochondrial probe) showing mitochondria trafficking to membrane ruffles. Scale bar, 10 μm

Movie S2.

Live phase contrast/fluorescence microscopy of T24 cells expressing Mito-DsRed (mitochondrial probe) showing mitochondria trafficking to membrane ruffles. Scale, 10 μm

Data, Materials, and Software Availability

All unique reagents generated in this study are available from the corresponding author with a completed Materials Transfer Agreement. Mass spectrometry proteomic data from the BioID screen are shown in Dataset S1. All study data are included in the article and/or SI Appendix.