Covalent capture and genetic code expansion enables chemoproteomic profiling and functional characterization of lysine acetoacetylation

Abstract

Lysine acetoacetylation (Kacac) driven by metabolite acetoacetic acid represents a molecular mechanism by which ketone bodies regulate cellular functions beyond energy provision. However, comprehensive characterization of Kacac has been hindered by technical limitations in detection and functional validation. Here, we report an integrated platform for systematic Kacac investigation. Exploiting the unique reactive ketone carbonyl moiety, we developed Aca-Bio, a hydroxylamine-based probe enabling specific enrichment of Kacac peptides through ketone-targeted covalent labeling and pH-controlled reversible enrichment. Application to mouse liver identified 260 Kacac sites across 125 proteins, revealing notable enrichment in metabolic pathways. Concurrently, we established a genetic code expansion system enabling site-specific Kacac incorporation. Using this approach, we demonstrated that K310acac in HMGCS2 substantially attenuates catalytic activity through impaired substrate binding. This dual-platform approach establishes a comprehensive framework for global profiling and site-specific functional characterization of Kacac, thereby facilitating systematic exploration of its physiological roles and pathological implications.

A chemical biology toolkit combining chemoproteomics and genetic code expansion deciphers the function of lysine acetoacetylation.

INTRODUCTION

Posttranslational modifications (PTMs) driven by metabolites play crucial roles in physiological and pathological processes by mediating dynamic protein regulation in response to cellular metabolic states. Ketone bodies, encompassing acetone, acetoacetic acid, and β-hydroxybutyrate, are hepatic metabolites produced via fatty acid β-oxidation, with biosynthesis significantly up-regulated during glucose-limited conditions such as fasting, prolonged exercise, and insulin-deficient pathologies (1). Beyond their canonical role as alternative energy supply, emerging evidence indicates that metabolites can function as signaling molecules by serving as precursors for various PTMs (2). Several short-chain metabolites, including lactate (3), succinate (4) and β-hydroxybutyrate (5), have been demonstrated to drive lysine acylations, thereby directly modulating protein function and cellular signaling networks.

Our prior investigations identified acetoacetic acid as a biosynthetic precursor for lysine acetoacetylation (Kacac) (6), a structurally distinctive PTM characterized by a reactive ketone moiety (Fig. 1A). This discovery expanded our understanding of ketone body regulatory mechanisms beyond their canonical roles as energy substrates. However, comprehensive elucidation of Kacac’s physiological and pathological importance necessitates addressing two fundamental challenges: first, achieving efficient, deep, and site-resolved identification of Kacac modifications; second, determining the functional impact of specific Kacac sites on substrate proteins.

Fig. 1. Chemical basis and experimental strategy for Kacac investigation.

(A) Chemical structure of Kacac. (B) Designing of the probe for the detection of Kacac. (C) Integrated technical platform combining chemoproteomic profiling and GCE for comprehensive characterization of Kacac. m/z, mass-to-charge ratio. WB, Western Blot.

Now, substantial technical limitations impede progress in both areas. Traditional approaches for identifying lysine acylations rely predominantly on antibody-based affinity enrichment. While we previously developed Kacac antibodies using antigen mimetics that demonstrated efficacy in Western blot applications, these antibodies exhibited suboptimal performance in Kacac peptide enrichment, necessitating development of alternative methodologies for specific Kacac enrichment and identification. Furthermore, conventional mutational analyses provide only indirect functional inferences that may not accurately reflect the authentic biological effects of the modification.

Unlike other known PTMs, Kacac has a distinctive active ketone carbonyl group capable of participating in specific chemical reactions without interference from canonical amino acids. In this study, exploiting this structural characteristic, we designed the Aca-Bio probe, comprising a hydroxylamine group that forms covalent oximes with ketone moieties, a polyethylene glycol (PEG) linker ensuring steric accessibility, and a biotin tag facilitating streptavidin-mediated enrichment. This probe achieved specific enrichment of Kacac-modified peptides through ketone-specific chemical targeting and reversible affinity purification, enabling subsequent mass spectrometric identification following controlled dissociation. Using this methodology, we identified 260 Kacac sites across 125 proteins in liver tissues from fasted C57BL/6 mice, substantially expanding the Kacac substrate landscape and revealing significant enrichment in metabolism-associated pathways.

Concurrently, we established a genetic code expansion (GCE) system enabling site-specific Kacac incorporation (Fig. 1, B and C). Using this system, we characterized the functional implications of K310acac in hydroxymethylglutaryl–coenzyme A (CoA) synthase (HMGCS2), a key regulatory enzyme in ketone body metabolism. Kinetic analyses revealed that the K310acac variant exhibited significantly attenuated catalytic activity compared to wild-type (WT) protein, indicating that Kacac substantially inhibits enzymatic function through impaired substrate binding capacity.

This integrated technical platform, combining chemoproteomic probe with GCE, establishes a comprehensive framework for both global Kacac substrate profiling and site-specific Kacac functional characterization. These methodological advances circumvent some limitations of antibody-dependent approaches and provide direct mechanistic insights into Kacac-mediated regulation, thereby facilitating in-depth exploration of Kacac’s roles in diverse cellular processes.

RESULTS

Development of pH-responsive probe for selective enrichment of Kacac

Unlike previously characterized PTMs, Kacac has a distinctive reactive ketone carbonyl structural feature. Notably, this chemical moiety is not naturally present among the 20 canonical amino acids comprising mammalian proteomes. This unique characteristic provides the molecular foundation for developing a chemoselective labeling strategy, wherein the ketone carbonyl group exhibits specific reactivity with hydroxylamine derivatives under physiological conditions, forming stable oxime conjugates. Crucially, these oxime bonds exhibit pH-dependent reversibility, maintaining high stability under neutral conditions while undergoing reversible cleavage in strongly acidic environments (Fig. 1B) (7, 8). On the basis of this dynamic covalent chemistry, we could construct a bifunctional system that simultaneously achieves specific covalent labeling of Kacac and allows acid-controlled targeted release of captured substrates, thereby establishing a technical platform for specific enrichment and controllable dissociation of Kacac-modified peptides.

Guided by these principles, we designed a probe Aca-Bio, with a molecular architecture comprising three critical functional modules: a hydroxylamine group that facilitates covalent labeling through ketone carbonyl-specific reactions; a PEG linker ensuring steric accessibility of the probe; and a biotin tag mediating streptavidin-based enrichment (Fig. 2A). This design integrates targeted covalent modification through ketone-specific chemistry with reversible affinity purification, constituting an acid-responsive labeling system specifically engineered for Kacac detection.

Fig. 2. Characterization of Aca-Bio probe reactivity and optimization of labeling parameters.

(A) Structure of the Aca-Bio probe. (B) Immunoblot analysis of unmodified BSA and Kacac-BSA using Kacac antibody. (C) Since Kacac proteins remain biotinylated after incubation with Aca-bio, they can be detected using horseradish peroxidase-conjugated streptavidin (HRP-SA). (D) Immunoblot analysis of Aca-Bio–labeled lysates from dimethyl sulfoxide (DMSO)–treated versus EAA-treated HepG2 cells. (E and F) Quantitative assessment of probe concentration-dependent (E) and reaction time-dependent (F) labeling efficiency of Aca-Bio in HepG2 cell lysates. (G) Evaluation of oxime bond reversibility under various delabeling conditions.

To validate probe performance, we initially constructed acetoacetylated bovine serum albumin (Kacac-BSA, BSA was chemically modified with acetoacetate) as a model protein (fig. S1). Western blot analysis demonstrated that Kacac-specific antibodies exclusively recognized chemically modified Kacac-BSA, while unmodified native BSA generated no detectable signal (Fig. 2B). Further examination using a horseradish peroxidase–conjugated streptavidin (HRP-SA) detection system revealed that biotin signal was specifically present in Aca-Bio-BSA samples (Kacac-BSA was labeled with Aca-Bio), confirming that the bifunctional design of the probe effectively recognized Kacac (Fig. 2C).

Given that our previous work had demonstrated that cellular Kacac levels could be modulated through ethyl acetoacetate (EAA) (6), we subsequently evaluate the probe’s capacity for detecting dynamic changes in global Kacac patterns. The results indicated that Aca-Bio sensitively reflected the significant elevation of Kacac levels in HepG2 cell lysates following EAA treatment (Fig. 2D). Through systematic optimization of experimental conditions, we determined that the probe achieved maximum labeling efficiency at 5 mM concentration (Fig. 2E), with the labeling reaction reaching kinetic equilibrium within 3 to 4 hours (Fig. 2F).

Building on the established optimal labeling system, we next sought to explore the conditions for the delabeling of Aca-Bio. Considering the reversible characteristics of the oxime bond in acidic environments, we hypothesized that labeled Aca-Bio could be displaced by using excess methoxyamine. As anticipated, the signal of Aca-Bio–labeled proteins rapidly decreased in a time-dependent manner following methoxyamine treatment, reaching a stable state within a timeframe of 6 to 12 hours (Fig. 2G).

Identification of Kacac sites in mammalian cells

On the basis of our optimized labeling and delabeling conditions, we established a systematic workflow for detection of Kacac substrates (Fig. 3A). The procedure involves incubating cell lysate with the Aca-Bio probe, followed by methanol/chloroform precipitation. This step effectively separates proteins from the vast majority of small molecules, including potential probe-labeled metabolites, which remain soluble in the supernatant and are subsequently discarded. After tryptic digestion of protein samples, the probe’s covalent labeling properties are leveraged to achieve selective enrichment. During this process, unmodified peptides lacking the requisite ketone carbonyl group are readily eluted because of their inability to form covalent bonds with the hydroxylamine moiety, whereas Kacac-modified peptides, having formed stable oxime linkages with the Aca-Bio probe, are specifically captured and retained by streptavidin agarose. Next, target Kacac peptides were eluted with 50 mM NH₂OMe buffer containing 20% AcOH, and the resulting samples underwent desalting to further remove any residual small molecules or salts. Last, Kacac site assignment was strictly based on high-quality tandem mass spectrometry (MS/MS) spectra matched to specific peptide sequences, providing unambiguous discrimination against any potential nonpeptidic adducts. Collectively, these measures guarantee the high specificity of our chemoproteomic workflow, thereby providing a reliable substrate atlas for future functional analyses.

Fig. 3. Chemoproteomic identification and characterization of the hepatic acetoacetylome.

(A) Schematic workflow for mapping Kacac sites in mouse liver. (B and C) Representative MS/MS spectra of the unmodified KacacQTALVELLK peptide (B) and Aca-Bio–labeled KAca-BioQTALVELLK peptide (C). (D) Semiquantitative analysis of Kacac peptide abundance in unenriched versus enriched Kacac-BSA samples. (E) Reactome pathway enrichment analysis of identified Kacac-modified proteins demonstrating metabolic pathway clustering. (F) Functional classification of the Kacac-modified proteins. (G) Subcellular distribution analysis of Kacac-modified proteins revealing mitochondrial enrichment.

To validate the sensitivity and specificity of this workflow, we designed a model verification experiment using Kacac-BSA. The Kacac-BSA was mixed with unmodified BSA in a ratio of 10:1, and the resulting sample was divided into two groups, one directly digested with trypsin and analyzed by liquid chromatography (LC)–MS/MS (non-enriched group), and the other processed through the complete workflow (enriched group) (fig. S2). The results revealed that Kacac signature peptides were detectable in both sample groups (Fig. 3, B and C). However, semiquantitative analysis demonstrated greatly enhanced signal intensity in the enriched group (Fig. 3D and data S1), conclusively confirming the enrichment efficiency and analytical reliability of our method.

Functional characterization of Kacac sites in mammalian cells

To elucidate the physiological importance of Kacac, we systematically investigated the Kacac substrate landscape in mammals. Building on previous observations that fasting activates ketone body metabolism (9), we carried out an in-depth analysis of liver tissues from C57BL/6 mice subjected to 48-hour fasting, identifying 260 distinct Kacac sites across 125 proteins (data S2). Notably, five Kacac sites were identified in histones, including our previously reported histone H3 lysine 56 acetoacetylation (H3K56acac) and histone H3 lysine 79 acetoacetylation (H3K79acac) (6), while previously unidentified sites at H2BK5, H2BK108, and H4K31 suggest potential previously unidentified epigenetic regulatory mechanisms involving Kacac. To explore the sequence characteristics of Kacac, we performed sequence logo analysis and compared the resulting motif with the previously reported motif of lysine β-hydroxybutyrylation (Kbhb) (fig. S3A) (9). Both PTMs exhibit similar exclusion of basic residues (Arg and His) at the immediate flanking positions (±1 and ±2). However, they differ in their preferences at these sites: Kacac demonstrates a marked preference for neutral aliphatic amino acids such as alanine and valine, whereas Kbhb displays broader tolerance encompassing both neutral and acidic residues. In addition, in the downstream region (+3 to +6), both PTMs show enrichment for lysine residues. Another distinguishing feature is the pronounced preference for an acidic residue at the −6 position in the Kacac motif, a characteristic that is absent in the Kbhb motif.

Furthermore, three Kacac sites (K310, K427, and K447) were identified on hydroxymethylglutaryl–coenzyme A (CoA) synthase (HMGCS2), a mitochondrial enzyme that serves as a rate-limiting step in ketogenesis. Our examination of previously reported Kbhb sites confirmed that the residues K310 and K447 could also undergo Kbhb, whereas K427 cannot. Sequence alignment of the surrounding amino acid regions (±7 residues) against the canonical Kbhb motif revealed that the sequences flanking K310 and K447 closely conform to the preferred Kbhb recognition motif (fig. S3B), whereas the motif surrounding K427 diverges substantially. This observation suggests that Kacac and Kbhb may be mediated by specialized enzymatic systems with unique substrate recognition specificities.

To further investigate the biological process associated with Kacac, we performed enrichment analysis on the Kacac substrates using the Reactome database. The results revealed significant enrichment of Kacac substrates in metabolic-related pathways, such as tricarboxylic acid cycle and respiratory electron transport [false discovery rate (FDR) = 3.27 × 10⁻²⁰], metabolism of amino acids and derivatives (FDR = 3.99 × 10⁻¹⁰), fatty acid metabolism (FDR = 7.90 × 10⁻⁰⁶), and gluconeogenesis (FDR = 2.6 × 10⁻⁰⁵) (Fig. 3E). Further functional annotation showed that ~50% of the Kacac-modified proteins were metabolic enzymes, with 13% representing metabolism-related transporters (Fig. 3F). Critically, multiple Kacac sites were localized within functional protein domains, such as adenosine triphosphate (ATP)–binding sites in medium-chain acyl-CoA ligase ACSF2 (ACSF2) and acyl-coenzyme A synthetase ACSM1 (ACSM1), and the catalytic domain of argininosuccinate synthase (table S1), suggesting that Kacac may regulate enzymatic activity through steric hindrance or charge effects.

Subcellular localization analysis further revealed that 65 Kacac-modified proteins were localized to mitochondria (Fig. 3G), including respiratory chain complex components and key tricarboxylic acid cycle enzymes. Collectively, these results indicate that Kacac likely plays a substantial role in energy metabolism regulation by targeting mitochondrial metabolic enzymes.

Development of GCE system for Kacac incorporation

While system biology pathway analysis can elucidate the biological processes involving Kacac, deciphering its site-specific functions necessitates the development of precise site-directed modification techniques. GCE technology, which enables precise site-specific incorporation of noncanonical amino acids (ncAAs) into proteins, offers a powerful tool for functional investigation of novel PTMs (Fig. 4A). The core of this technology relies on engineered aminoacyl-tRNA synthetase (aaRS)/tRNA pairs, with systems derived from Methanosarcina barkeri (Mb) and Methanosarcina mazei (Mm) pyrrolysyl-tRNA synthetase (PylRS) being most widely used, as they specifically recognize termination codons to facilitate ncAA insertion (10). However, since our initial report on Kacac, no dedicated aaRS/tRNA systems for Kacac incorporation has been documented.

Fig. 4. Establishment of the GCE system for site-specific Kacac incorporation.

(A) Mechanistic illustration of site-specific incorporation via orthogonal aaRS/tRNA pairs. (B) The mutation sites across engineered PylRS variants screened for Kacac incorporation efficiency. (C) SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analysis of Ni-NTA–purified HA-tagged Ub-K6acac protein expressed in E. coli. (D) Quadrupole orthogonal acceleration–time-of-flight mass spectrometric analysis of Ub-WT and Ub-K6acac proteins demonstrating the characteristic mass shift. (E) MS/MS spectra of Ub-K6acac peptide confirming site-specific incorporation.

On the basis of the structural similarities among different lysine acylations, we hypothesized that existing aaRS system developed for other lysine acylations might accommodate Kacac incorporation. To test this hypothesis, we used ubiquitin (Ub) K6acac as a model system and systematically evaluated the compatibility of eight distinct engineered aaRS (11–17) variants in Escherichia coli expression systems (Fig. 4B). Quantitative Western blotting demonstrated that Mb-mut-2 type aaRS facilitated moderate expression levels of Ub K6acac (Fig. 4C). The precise incorporation of Kacac at the target site was confirmed by quadrupole orthogonal acceleration–time-of-flight high-resolution mass spectrometer, which detected a characteristic mass shift of +84 Da (Fig. 4D), corroborated by fragment ion matching analysis from MS/MS spectra (Fig. 4E).

Functional characterization of HMGCS2 K310acac

In our Kacac substrate identification experiments using the Aca-Bio probe, we found multiple Kacac-modified proteins functioning as key metabolic enzymes in β-oxidation and ketone body metabolism pathways (Fig. 5A). Among these, the mitochondrial rate-limiting enzyme HMGCS2 was particularly noteworthy. This enzyme catalyzes the condensation of acetoacetyl-CoA with acetyl-CoA to form 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA), representing a critical regulatory node in ketone body biosynthesis. Through in-depth mass spectrometric analysis, we identified three Kacac sites on HMGCS2, specifically at residues K310, K427, and K447. However, only the Kacac at the conserved K310 residue was identified across all four mouse sample groups (Fig. 5B and data S2). Analysis of the HMGCS2-HMG-CoA complex crystal structure [Protein Data Bank (PDB): 2WYA] indicated that the K310 residue directly participates in substrate binding through formation of a hydrogen bond interaction with the phosphate group of HMG-CoA (Fig. 5C). On the basis of these structural insights, we hypothesized that K310acac might inhibit enzymatic activity by disrupting this essential interaction, thereby reducing substrate binding efficiency.

Fig. 5. Functional characterization of HMGCS2 K310acac.

(A) Schematic representation of Kacac-modified substrates within the β-oxidation and ketone body metabolism pathways. (B) Multiple sequence alignment demonstrating evolutionary conservation of K310 in HMGCS2 across species. (C) Crystal structure analysis of HMGCS2 in complex with HMG-CoA (PDB: 2WYA) highlighting K310 position at the substrate binding interface. (D) Aca-Bio labeling and HRP-SA detection of immunopurified HMGCS2-WT and HMGCS2-K310acac from HepG2 cells. (E) Comparative enzyme kinetic analysis of immunopurified HMGCS2-WT versus HMGCS2-K310acac from HepG2 cells, with quantitative determination of K_m and V_max values. Error bars are shown as the mean with SD.

To validate this hypothesis, we established a site-specific HMGCS2 K310acac model in HepG2 cells. After screening nine engineered aaRS/tRNA systems, we determined that Mb-mut-2 also demonstrated optimal incorporation efficiency in this system with a yield of ~1.47 μg per 10 cm dish (20.13% relative to the HMGCS2-WT protein) (Fig. 4B and fig. S4). The incorporation of the K310acac was also confirmed by Western blot and MS/MS analysis (Fig. 5D and fig. S5), achieving sufficient modification levels for functional studies.

To investigate whether the K310 represents the predominant acetoacetylation site on HMGCS2, we constructed a K310R mutant Flag-HMGCS2 plasmid. Both WT and K310R Flag-HMGCS2 were overexpressed in HepG2 cells. Following immunopurification with an anti-Flag antibody, the Kacac levels were assessed using the Aca-Bio probe. The results demonstrated a near-complete loss of the Kacac signal in the K310R mutant compared to the WT protein (fig. S6A). Our previous studies revealed that global cellular Kacac levels positively correlate with extracellular acetoacetate (6). We next examined this relationship for HMGCS2. We treated cells overexpressing Flag-HMGCS2 with 1 mM EAA. Subsequent analysis using the Aca-Bio probe revealed an increase in the Kacac level on HMGCS2 following EAA stimulation (fig. S6B). This result confirms that the HMGCS2 K310acac is dynamically regulated by and positively correlates with the concentration of its precursor, acetoacetate. To directly compare the catalytic activities of WT and K310acac-modified HMGCS2, we performed enzyme kinetic assays on purified proteins using a coupled colorimetric method based on CoA by-product generation. In this assay, CoA released during HMG-CoA synthesis reduces 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) to form 5-thio-2-nitrobenzoic acid, which can be detected by ultraviolet spectrophotometry. The results revealed that the K310acac variant exhibited significant reduced activity compared to the WT protein. Kinetic analysis demonstrated that HMGCS2 K310acac had a Michaelis constants (K_m) 2.5-fold higher than that of the WT protein, while its maximal reaction rates (V_max) was only 44% of the value observed for the WT enzyme (Fig. 5E). These data confirm that Kacac modification significantly inhibits enzymatic activity by attenuating substrate binding capacity.

DISCUSSION

The discovery of Kacac, a type of previously unidentified PTM driven by the ketone body metabolite acetoacetic acid, provides a distinct perspective for elucidating the molecular mechanisms by which ketone bodies regulate cellular functions. However, because of limitations in detection technologies, the substrate landscape and functional networks of Kacac remain largely unexplored. This study establishes a comprehensive chemical biology platform by integrating chemical probe technology with GCE, enabling a complete research pipeline from global substrate profiling to site-specific functional characterization of Kacac.

The development of the Aca-Bio probe, leveraging the unique reactivity of ketone carbonyl groups, illustrates the strategic application of dynamic covalent chemistry for Kacac-specific detection. This probe uses the pH-responsive reversibility of oxime bonds (stable under neutral conditions/dissociable under acidic conditions) to achieve specific covalent labeling and controlled release of Kacac-modified substrates. Compared to antibody-based affinity enrichment techniques, which are susceptible to epitope masking effects, this chemistry-based enrichment strategy establishes a methodological foundation for unbiased global modification detection.

Analysis of mouse liver tissue revealed that Kacac is widely distributed across metabolic enzymes involved in fatty acid metabolism and ketone body synthesis/production. Given that Kbhb has been reported to exert feedback regulation on its own biosynthetic pathways, we sought to explore whether a similar regulatory principle applies to Kacac. Our profiling identified 10 key enzymes modified by Kacac, including Acsl1, Acsf2, Acsl5, Acsm1, Hadha, Acaa2, Acat1, Bdh1, Hmgcs2, and Hmgcl (Fig. 5A). This broad targeting of metabolic enzymes suggests a potential feedback role for Kacac. Notably, functional analysis of the modification sites suggests potential functional consequences. For example, in Acsl1, K208 is located within the adenosine monophosphate–binding domain (207 to 494), while K644 resides within the ACS ligase domain. Kacac at these critical sites could directly interfere with the enzyme’s catalytic activity. In Acsm1, K361, a known ATP-binding residue, was found to be modified by Kacac. This modification is likely to perturb ATP coordination and impair the enzyme’s function in fatty acid activation.

On the other hand, through the engineered GCE system, we successfully established site-specific incorporation of Kacac in mammalian cells. Using this technology, we characterized the function of a conserved K310acac site in HMGCS2, a key enzyme in ketone body metabolism. Our findings reveal that K310acac negatively regulates HMGCS2 enzymatic activity by introducing steric hindrance that disrupts the substrate binding interface, as demonstrated by the observed 2.5-fold increase in K_m and a 56% decrease in catalytic efficiency in the K310acac variant. These results suggest that Kacac may acts as a molecular “brake” on HMGCS2 activity, potentially fine-tuning ketone body production during fasting.

In conclusion, our research establishes reliable methodologies for both deep detection of Kacac and precise manipulation of specific Kacac sites. The Aca-Bio probe enabled systematic identification of Kacac substrates, generating comprehensive Kacac landscape and expanding the known Kacac substrate repertoire. In addition, through GCE technology, we achieved site-specific incorporation of Kacac in mammalian systems, revealing the molecular principles by which Kacac regulates metabolic enzyme activity. These methodological advances establish a complete research system from Kacac discovery to functional validation, providing a versatile technical platform for in-depth exploration of the biological significance of Kacac in metabolic regulation, epigenetics, and other fields.

MATERIALS AND METHODS

Animals

Adult male C57BL/6J mice (14 to 16 weeks of age) were obtained from Beijing Vital River Laboratory and housed in a specific pathogen–free–level Laboratory Animal Room at the Shanghai Institute of Materia Medica, Chinese Academy of Sciences. All animal experiments and protocols were approved by the Animal Ethics Committee of the Shanghai Institute of Materia Medica [Institutional Animal Care and Use Committee (IACUC) number: 2024-12-HH-03]. This study was performed under the guidelines and ethics of the Association for Assessment and Accreditation of Laboratory Animal Care International.

Mice were maintained in a temperature-controlled (22°C) vivarium with regulated light cycles and provided ad libitum access to water and food. After an acclimatization period of 7 days, food was withdrawn at Zeitgeber time 8 for fasting experiments. Liver tissues were harvested 48 hours later and stored at −80°C for Kacac substrate profiling experiments.

Cell culture

Human embryonic kidney 293T and HepG2 cells (National Collection of Authenticated Cell Cultures, China) were cultured in Dulbecco’s modified Eagle’s medium (DMEM; catalog no. L110KJ, Basalmedia) supplemented with 10% (v/v) fetal bovine serum (catalog no. WS500T, Ausbian) and 1% penicillin-streptomycin (catalog no. 15140122, Gibco) at 37°C in a humidified incubator with 5% CO₂.

Labeling of HepG2 cell lysate with probe Aca-Bio

HepG2 cells (1 × 10⁵) were seeded in 6-cm dishes and cultured at 37°C. After 12 hours, cells were treated with 1 mM EAA or an equivalent volume of dimethyl sulfoxide (negative control). After a 24-hour incubation, the culture medium was removed, and the cells were washed twice with phosphate-buffered saline (PBS) containing protease inhibitor cocktail [containing phenylmethylsulfonyl fluoride (PMSF), leupeptin, pepstatin, Na₃VO₄, NaF, sodium butyrate, and nicotinamide].

For specific labeling reactions with Aca-Bio probes in whole cell lysates, cells were lysed with buffer A solution [20 mM Hepes-KOH (pH 8.0), 0.02 M KCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 25% glycerinum, and inhibitor cocktail]. After centrifugation at 16,000g for 10 min, the supernatant was collected, and the precipitate was resuspended in buffer B solution [20 mM Hepes-KOH (pH 8.0), 1.2 M KCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 25% glycerinum, and inhibitor cocktail]. The cell suspension was sonicated for 5 to 10 min and then centrifuged at 16,000g for 10 min to remove cellular debris. Protein concentration was determined using a BCA kit (catalog no. P0010, Beyotime Biotechnology).

Subsequently, 200 μg of lysed protein was diluted to 35 μl with double-distilled water (ddH₂O) and reacted with 65 μl of a freshly prepared specific labeling reaction cocktail containing guanidine hydrochloride [50 μl, 6 M in ddH₂O (pH 4.2)], AcONa [10 μl, 1 M in ddH₂O (pH 4.2)], and Aca-Bio (5 μl, 100 mM in ddH₂O). The reaction was performed at 37°C for the specified duration. Labeling efficiency was assessed by Western blot analysis.

Delabeling process of Aca-Bio

HepG2 cell lysate containing 200 μg of protein (100 μl) was labeled with 5 mM Aca-Bio for 4 hours at 37°C. Then, 28 μl of acetic acid and 14 μl of NH₂OMe·HCl solution (0.5 M in ddH₂O) were added. The mixture was incubated at 37°C for the specified duration. Delabeling efficiency was evaluated by Western blot analysis.

Western blot analysis

Approximately 10 μg of protein lysates was separated by SDS–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. The membranes were blocked in TBST solution (20 mM tris-HCl (pH 7.6), 150 mM NaCl, and 0.1% Tween 20] with 3% BSA for 1 hour at room temperature. Membranes were then incubated in diluted antibody solution (HRP-labeled streptavidin antibody:TBST = 1:10,000) for 1 hour at room temperature, followed by three washes with TBST. Bands were visualized using enhanced chemiluminescence kits, and chemiluminescence was recorded on a Tanon 4600 imaging system.

Lysis of mouse liver tissue

Mouse liver tissue (400 mg) were sectioned into pieces in a 2-ml centrifuge tube and washed with 1 ml of precooled PBS containing protease inhibitor cocktail. After initial centrifugation (300g, 4°C, 3 min) and supernatant removal, tissues underwent 4 to 5 washing cycles with fresh buffer until colorless supernatants were achieved. The resulting tissue was then divided into two portions: one for whole liver protein extraction and another for nucleoprotein extraction.

For whole liver protein extraction, grinding beads were added to the centrifuge tube. The tissue was flash-frozen in liquid nitrogen and further pulverized using a grinder. The pulverized tissue was lysed by adding 500 μl of ice-cold radioimmunoprecipitation assay (RIPA) buffer [50 mM tris, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholic acid, and 0.1% SDS (pH = 7.4)] supplemented with protease inhibitor cocktail, followed by vortex mixing for 10 s and incubation on ice for 30 min. The resulting solution was subjected to ultrasonic disruption for 10 min, followed by centrifugation at 16,000g for 10 min. Protein concentration was determined from the supernatant.

Nuclear isolation was performed using a commercially available nuclei extraction kit (SN0020, Beijing Solarbio Science & Technology Co. Ltd.). Briefly, liver tissue was homogenized in 1 ml of reagent A using a glass Dounce homogenizer. Nuclei were collected after centrifuging the suspension at 700g for 5 min. Subsequently, 200 μl of ice-cold RIPA buffer containing protease inhibitor cocktail was added. The mixture was subjected to ultrasonic cracking for 10 min, followed by centrifugation at a speed of 16,000g for 10 min. Protein concentration was determined from the supernatant.

Profiling of Kacac substrates in mouse liver tissue

Whole liver protein (4 mg) and nucleoprotein (400 μg) were separately processed for Kacac substrate detection. Guanidine hydrochloride (final concentration, 3 M), AcONa (final concentration, 100 mM), Aca-Bio (final concentration, 100 mM) were added to the tissue lysate [final protein concentration, 2 μg/μl (pH 4.2)]. The mixture was incubated for 4 hours at 37°C. After the labeling, methanol and chloroform were added to the reaction mixture [methanol:chloroform:reaction mixture = 4:1:2 (v/v/v)], and the resulting solution was incubated at −20°C overnight to precipitate proteins. The precipitate was collected by centrifugation at 3000g for 5 min, followed by successive washing steps with methanol and acetone (one wash cycle per solvent). The precipitate was air-dried on ice for 30 min to evaporate residual organic solvents. The protein pellet was resuspended in 500 μl of 25 mM tris buffer (pH 8.5), followed by the addition of 40 μg (for whole liver protein sample) or 4 μg (for nucleoprotein sample) of trypsin. The solution was incubated at 37°C overnight. After digestion, the sample was heated at 98°C for 30 s to inactivate trypsin, followed by centrifugation at 16,000g for 5 min. The resulting supernatant was incubated with 30 μl of streptavidin agarose (#17511301, Cytiva) at 4°C for 1 hour. The suspension was centrifuged at 500g for 3 min, and the supernatant was removed. The agarose was washed three times by wash buffer [150 mM NaCl, 50 mM tris-HCl, and 0.05% NP-40 (pH 7)] and three times with PBS buffer.

After enrichment, 100 μl of freshly prepared elution buffer, comprising 20% AcOH (20 μl), NH₂OMe·HCl (10 μl, 0.5 M in ddH₂O), guanidine hydrochloride [50 μl, 6 M in ddH₂O (pH 4.2)], and ddH₂O (20 μl), was mixed with the agarose. The mixture was incubated with shaking at 37°C for 6 hours, followed by centrifugation at 300g for 3 min to collect the supernatant. The agarose was subsequently resuspended in 100 μl of fresh elution buffer and subjected to a secondary 6-hour incubation under identical conditions. A third extended incubation (12 hours) was performed with an additional 100 μl of fresh elution buffer. All three supernatants were combined to constitute the final eluate. Last, the sample was desalted using the C18 tip before LC-MS/MS analysis.

Cell transfection

293T cells were seeded 24 hours before transfection to achieve ~80% confluency. Transfection was performed using Hieff Trans Liposomal Transfection Reagent (40802ES03, Yeasen) according to the manufacturer’s instructions. Briefly, plasmid and lipid transfection reagent were individually diluted in high-glucose DMEM medium at a 1:2 mass-to-volume ratio before being combined. The mixture was incubated at room temperature for 20 min and then added dropwise to 293T cells. Approximately 12 hours later, the culture medium was replaced. Cells were cultured for 48 hours at 37°C in a 5% CO₂ incubator for protein expression.

Establishment of the plasmids for GCE

On the basis of previous reports, mutations were introduced into Mb-PylRS and Mm-PylRS contained in the plasmids pNEU-hMbPylRS-4xU6M15 (Addgene #105830, containing a humanized MbPylRS and tRNAM15) and pEVOL-AckRS (Addgene #137976, containing AckRS and pylTcua), respectively. On the basis of the site-directed mutagenesis information in Fig. 4B, we constructed Mb-mut-1, Mb-mut-2, Mb-mut-3, Mb-mut-4, Mb-mut-5, Mb-mut-6, Mm-mut-1. Mm-mut-2, and Mm-mut-3. Subsequently, the mutated pylRS sequences were individually recombined into the pNEU-hMbPylRS-4xU6M15 and pEVOL-AckRS vectors for expression in eukaryotic and prokaryotic systems, respectively.

Expression and purification of HMGCS2 protein in HepG2 and 293T cells

Human HMGCS2 gene was cloned into a pCMV vector with a C-terminal Flag tag, and site-directed mutagenesis was performed to introduce an amber stop codon (TAG) at lysine-310 (K310TAG). Various mutated pNEU-PylRS-4xU6M15 plasmids (Mb-PylRS, Mb-mut-1, Mb-mut-2, Mb-mut-3, Mb-mut-4, Mb-mut-5, Mb-mut-6, and Mm-mut-1) and 3′-FLAG-HMGCS2 K310TAG were cotransfected into 293T cells using Hieff Trans Liposomal Transfection Reagent as described above. Cells were cultured in the presence of 1 mM acetoacetyllysine for 48 hours.

Transfected cells were washed twice with ice-cold PBS containing protease inhibitor cocktail and lysed in lysis buffer [20 mM tris (pH 7.5), 150 mM NaCl, and 1% Triton X-100, supplemented with the protease inhibitor cocktail] for 20 min at 4°C. Cell lysates were collected in 1.5-ml microcentrifuge tube, sonicated, and centrifuged at 16,000g for 10 min to collect the supernatant. The supernatant was incubated with 5 μl of anti-FLAG antibody–conjugated agarose beads (Sigma-Aldrich, A2220-5ML) for 2 hours at 4°C with rotation. The beads were washed three times with cold lysis buffer to eliminate nonspecifically bound proteins. Competitively elution of proteins from Flag beads was achieved using 1× Flag peptide (200 ng/μl dissolved in PBS) over 4 hours at 4°C. Last, an equal volume of 50% glycerol was added to the protein solution, and the mixture was stored at −80°C.

Expression and purification of Ub in E. coli

The various mutated pEVOL-AckRS plasmids (Mb-mut-1, Mb-mut-2, Mb-mut-3, Mb-mut-4, Mb-mut-5, Mm-mut-1, Mm-mut-2, and Mm-mut-3) and 3′-His-UB K6TAG were cotransfected into DH10B and incubated overnight at 37°C with shaking at 220 rpm in LB medium containing ampicillin (50 μg/ml) and chloromycetin (34 μg/ml). Bacterial cultures were transferred to fresh 250-ml medium and incubated at 37°C for 3 to 4 hours until the optical density at 600 nm reached 0.6 to 0.8. l-arabinose (final concentration, 0.2%) and acetoacetyllysine (1 mM) were added to the bacterial culture, which was then incubated at 23°C for an additional 20 hours. The bacteria were harvested by centrifugation and resuspended in lysis buffer [His lysis buffer: 50 mM NaH₂PO_4, 300 mM NaCl, 20 mM imidazole, 10% glycerol, 1% Triton X-100, and 0.2 mM PMSF (pH 8.0)] followed by ultrasonic disruption. After centrifugation at 8000g for 30 min, the supernatant was collected and incubated with 1 ml of nickel-nitrilotriacetic acid (Ni-NTA) beads at 4°C for 2 hours. The beads were washed three times with cold wash buffer [50 mM NaH₂PO₄, 300 mM NaCl, 50 mM imidazole, and 0.2 mM PMSF (pH 8.0)] to eliminate nonspecifically bound proteins. Subsequently, proteins were eluted with elution buffer [50 mM NaH₂PO₄, 300 mM NaCl, 500 mM imidazole, and 0.2 mM PMSF (pH 8.0)]. Last, glycerol was added to the protein solution to achieve a final concentration of 25%, and the resultant mixture was stored at −80°C. The sequence of Ub used in this research is MTSMQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGLEHHHHHHHH.

In vitro HMGCS2 activity assay

Flag-HMGCS2 and Flag-HMGCS2 K310acac proteins were immunoprecipitated from 293T cells. The assay mixture consisted of 100 mM tris-HCl (pH 8.0), 130 mM DTNB (A_{412 nm} = 13.6 mM⁻¹), 62.5 to 2000 μM acetyl-CoA, 10 μM acetoacetyl-CoA, and 0.5 μg of enzyme in a total volume of 200 μl. Kinetic data were fitted to the Michaelis-Menten equation by nonlinear regression analysis to determine the maximal reaction velocity (V_max) and Michaelis constant (K_m). For each concentration, data represent the mean of two to three independent measurements, with error bars showing the SD. A unit of enzyme activity is defined as the amount of enzyme that catalyzes the transformation of 1 μmol of substrate into product per minute.

High-performance LC-MS/MS analysis

The sample analysis was carried out on an EASY-nLC 1200 ultra high performance liquid chromatography (UHPLC) system (Thermo Fisher Scientific) coupled to a Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific). Peptides were dissolved in 2.5 μl of solvent A (0.1% formic acid in water, v/v) and injected to a homemade packed capillary C18 column (20 cm length by 75 μm inside diameter, 1.9 μm particle size, Dr. Maisch GmbH, Germany). The samples were run in 120-min gradient, respectively, from 8 to 90% solvent B (A, 0.1% formic acid; B, 80% acetonitrile in 0.1% formic acid). Full mass scans were acquired with 350 to 1200 mass/charge ratio at a mass resolution of 60,000. Ions with 2+, 3+, and 4+ charges were selected for MS/MS analysis. The 20 most intensive ions were fragmented with 28% normalized collision energy, and tandem mass spectra were acquired with a mass resolution of 15,000.

Protein sequence database searching

After LC-MS/MS acquisition, the raw files were qualitatively analyzed by the Thermo Proteome Discoverer (version 2.4) against the UniProt mouse database (released in January 2020). Trypsin/P was designated as the digestive enzyme, with a maximum of two missed cleavages and a minimum peptide length of seven amino acids. Aminomethylation on cysteine (+57.022 Da) was set as a fixed modification. Protein N-terminal acetylation (+42.0106 Da), methionine oxidation (+15.995 Da), and Kacac-Bio on lysine (+113.048 Da) were designated as variable modifications. The reverse sequences were appended for an FDR evaluation. Peptides with site localization probability less than 0.75 were excluded from further analysis.

Supplementary Materials

The PDF file includes:

Supplementary Text

Figs. S1 to S6

Table S1

NMR and MS spectra of the synthesized compounds

Legends for data S1 and S2

Other Supplementary Material for this manuscript includes the following:

Data S1 and S2