Identification of the Regulatory Elements and Protein Substrates of Lysine Acetoacetylation

Reviewer #2 (Public review):

In the manuscript by Fu et al., the authors developed a chemo-immunological method for the reliable detection of Kacac, a novel post-translational modification, and demonstrated that acetoacetate and AACS serve as key regulators of cellular Kacac levels. Furthermore, the authors identified the enzymatic addition of the Kacac mark by acyltransferases GCN5, p300, and PCAF, as well as its removal by deacetylase HDAC3. These findings indicate that AACS utilizes acetoacetate to generate acetoacetyl-CoA in the cytosol, which is subsequently transferred into the nucleus for histone Kacac modification. A comprehensive proteomic analysis has identified 139 Kacac sites on 85 human proteins. Bioinformatics analysis of Kacac substrates and RNA-seq data reveal the broad impacts of Kacac on diverse cellular processes and various pathophysiological conditions. This study provides valuable additional insights into the investigation of Kacac and would serve as a helpful resource for future physiological or pathological research.

Comments on revised version:

The authors have made efforts to revise this manuscript and address my concerns. The revisions are appropriate and have improved the quality of the manuscript.

Reviewer #3 (Public review):

Summary:

This paper presents a timely and significant contribution to the study of lysine acetoacetylation (Kacac). The authors successfully demonstrate a novel and practical chemo-immunological method using the reducing reagent NaBH4 to transform Kacac into lysine β-hydroxybutyrylation (Kbhb).

Strengths:

This innovative approach enables simultaneous investigation of Kacac and Kbhb, showcasing its potential in advancing our understanding of post-translational modifications and their roles in cellular metabolism and disease.

Weaknesses:

The experimental evidence presented in the article is insufficient to fully support the authors' conclusions. In the in vitro assays, the proteins used appear to be highly inconsistent with their expected molecular weights, as shown by Coomassie Brilliant Blue staining (Figure S3A). For example, p300, which has a theoretical molecular weight of approximately 270 kDa, appeared at around 37 kDa; GCN5/PCAF, expected to be ~70 kDa, appeared below 20 kDa. Other proteins used in the in vitro experiments also exhibited similarly large discrepancies from their predicted sizes. These inconsistencies severely compromise the reliability of the in vitro findings. Furthermore, the study lacks supporting in vivo data, such as gene knockdown experiments, to validate the proposed conclusions at the cellular level.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public review):

Summary

Lysine acetoacetylation (Kacac) is a recently discovered histone post-translational modification (PTM) connected to ketone body metabolism. This research outlines a chemo-immunological method for detecting Kacac, eliminating the requirement for creating new antibodies. The study demonstrates that acetoacetate acts as the precursor for Kacac, which is catalyzed by the acyltransferases GCN5, p300, and PCAF, and removed by the deacetylase HDAC3. AcetoacetylCoA synthetase (AACS) is identified as a central regulator of Kacac levels in cells. A proteomic analysis revealed 139 Kacac sites across 85 human proteins, showing the modification's extensive influence on various cellular functions. Additional bioinformatics and RNA sequencing data suggest a relationship between Kacac and other PTMs, such as lysine βhydroxybutyrylation (Kbhb), in regulating biological pathways. The findings underscore Kacac's role in histone and non-histone protein regulation, providing a foundation for future research into the roles of ketone bodies in metabolic regulation and disease processes.

Strengths

(1) The study developed an innovative method by using a novel chemo-immunological approach to the detection of lysine acetoacetylation. This provides a reliable method for the detection of specific Kacac using commercially available antibodies.

(2) The research has done a comprehensive proteome analysis to identify unique Kacac sites on 85 human proteins by using proteomic profiling. This detailed landscape of lysine acetoacetylation provides a possible role in cellular processes.

(3) The functional characterization of enzymes explores the activity of acetoacetyltransferase of key enzymes like GCN5, p300, and PCAF. This provides a deeper understanding of their function in cellular regulation and histone modifications.

(4) The impact of acetyl-CoA and acetoacetyl-CoA on histone acetylation provides the differential regulation of acylations in mammalian cells, which contributes to the understanding of metabolic-epigenetic crosstalk.

(5) The study examined acetoacetylation levels and patterns, which involve experiments using treatment with acetohydroxamic acid or lovastatin in combination with lithium acetoacetate, providing insights into the regulation of SCOT and HMGCR activities.

We thank all the reviewers for their positive, insightful comments which have helped us improve our manuscript. We have revised the manuscript as suggested by the reviewers.

Weakness

(1) There is a limitation to functional validation, related to the work on the biological relevance of identified acetoacetylation sites. Hence, the study requires certain functional validation experiments to provide robust conclusions regarding the functional implications of these modifications on cellular processes and protein function. For example, functional implications of the identified acetoacetylation sites on histone proteins would aid the interpretation of the results.

We agree with the reviewer that investigating the functional role of individual histone Kacac sites is essential for understanding the epigenetic impact of Kacac marks on gene expression, signaling pathways, and disease mechanisms. This topic is out of the scope of this paper which focuses on biochemical studies and proteomics. Functional elucidation in specific pathways will be a critical direction for future investigation, ideally with the development of site-specific anti-Kacac antibodies.

(2) The authors could have studied acetoacetylation patterns between healthy cells and disease models like cancer cells to investigate potential dysregulation of acetoacetylation in pathological conditions, which could provide insights into their PTM function in disease progression and pathogenesis.

We appreciate the reviewer’s valuable suggestion. In our study, we measured Kacac levels in several types of cancer cell lines, including HCT116 (Fig. 2B), HepG2 (Supplementary Fig. S2), and HeLa cells (data not shown in the manuscript), and found that acetoacetate-mediated Kacac is broadly present in all these cancer cell lines. Our proteomics analysis linked Kacac to critical cellular functions, e.g. DNA repair, RNA metabolism, cell cycle regulation, and apoptosis, and identified promising targets that are actively involved in cancer progression such as p53, HDAC1, HMGA2, MTA2, LDHA. These findings suggest that Kacac has significant, non-negligible effects on cancer pathogenesis. We concur that exploring the acetoacetylation patterns in cancer patient samples with comparison with normal cells represents a promising direction for next-step research. We plan to investigate these big issues in future studies.

(3) The time-course experiments could be performed following acetoacetate treatment to understand temporal dynamics, which can capture the acetoacetylation kinetic change, thereby providing a mechanistic understanding of the PTM changes and their regulatory mechanisms.

As suggested, time-course experiments were performed, and the data have been included in the revised manuscript (Supplementary Fig. S2A).

(4) Though the discussion section indeed provides critical analysis of the results in the context of existing literature, further providing insights into acetoacetylation's broader implications in histone modification. However, the study could provide a discussion on the impact of the overlap of other post-translational modifications with Kacac sites with their implications on protein functions.

We appreciate the reviewer’s helpful suggestion. We have added more discussions on the impact of the Kacac overlap with other post-translational modifications in the discussion section of the revised manuscript.

Impact

The authors successfully identified novel acetoacetylation sites on proteins, expanding the understanding of this post-translational modification. The authors conducted experiments to validate the functional significance of acetoacetylation by studying its impact on histone modifications and cellular functions.

We appreciate the reviewer’s comments.

Reviewer #2 (Public review):

In the manuscript by Fu et al., the authors developed a chemo-immunological method for the reliable detection of Kacac, a novel post-translational modification, and demonstrated that acetoacetate and AACS serve as key regulators of cellular Kacac levels. Furthermore, the authors identified the enzymatic addition of the Kacac mark by acyltransferases GCN5, p300, and PCAF, as well as its removal by deacetylase HDAC3. These findings indicate that AACS utilizes acetoacetate to generate acetoacetyl-CoA in the cytosol, which is subsequently transferred into the nucleus for histone Kacac modification. A comprehensive proteomic analysis has identified 139 Kacac sites on 85 human proteins. Bioinformatics analysis of Kacac substrates and RNA-seq data reveals the broad impacts of Kacac on diverse cellular processes and various pathophysiological conditions. This study provides valuable additional insights into the investigation of Kacac and would serve as a helpful resource for future physiological or pathological research.

The following concerns should be addressed:

(1) A detailed explanation is needed for selecting H2B (1-26) K15 sites over other acetylation sites when evaluating the feasibility of the chemo-immunological method.

The primary reason for selecting the H2B (1–26) K15acac peptide to evaluate the feasibility of our chemo-immunological method is that H2BK15acac was one of the early discovered modification sites in our preliminary proteomic screening data. The panKbhb antibody used herein is independent of peptide sequence so different modification sites on histones can all be recognized. We have added the explanation to the manuscript.

(2) In Figure 2(B), the addition of acetoacetate and NaBH4 resulted in an increase in Kbhb levels. Specifically, please investigate whether acetoacetylation is primarily mediated by acetoacetyl-CoA and whether acetoacetate can be converted into a precursor of β-hydroxybutyryl (bhb-CoA) within cells. Additional experiments should be included to support these conclusions.

We appreciate the reviewer’s valuable comments. In our paper, we had the data showing that acetoacetate treatment had very little effect on histone Kbhb levels in HEK293T cells, as observed in lanes 1–4 of Fig. 2A, demonstrating that acetoacetate minimally contributes to Kbhb generation. We drew the conclusion that histone Kacac is primarily mediated by acetoacetyl-CoA based on multiple pieces of evidence: first, we observed robust Kacac formation from acetoacetyl-CoA upon incubation with HATs and histone proteins or peptides, as confirmed by both western blotting (Figs. 3A, 3B; Supplementary Figs. S3C– S3F) and MALDI-MS analysis (Supplementary Fig. S4A). Second, treatment with hymeglusin—a specific inhibitor of hydroxymethylglutaryl-CoA synthase, which catalyzes the conversion of acetoacetyl-CoA to HMG-CoA—led to increased Kacac levels in HepG2 cells (PMID: 37382194). Third, we demonstrated that AACS whose function is to convert acetoacetate into acetoacetyl-CoA leads to marked histone Kacac upregulation (Fig. 2E). Collectively, these findings strongly support the conclusion that acetoacetate promotes Kacac formation primarily via acetoacetyl-CoA.

(3) In Figure 2(E), the amount of pan-Kbhb decreased upon acetoacetate treatment when SCOT or AACS was added, whereas this decrease was not observed with NaBH4 treatment. What could be the underlying reason for this phenomenon?

In the groups without NaBH₄ treatment (lanes 5–8, Figure 2E), the Kbhb signal decreased upon the transient overexpression of SCOT or AACS, owing to protein loading variation in these two groups (lanes 7 and 8). Both Ponceau staining and anti-H3 results showed a lower amount of histones in the AACS- or SCOT-treated samples. On the other hand, no decrease in the Kbhb signal was observed in the NaBH₄-treated groups (lanes 1–4), because NaBH₄ treatment elevated Kacac levels, thereby compensating for the reduced histone loading. The most important conclusion from this experiment is that AACS overexpression increased Kacac levels, whereas SCOT overexpression had no/little effect on histone Kacac levels in HEK293T cells.

(4) The paper demonstrates that p300, PCAF, and GCN5 exhibit significant acetoacetyltransferase activity and discusses the predicted binding modes of HATs (primarily PCAF and GCN5) with acetoacetyl-CoA. To validate the accuracy of these predicted binding models, it is recommended that the authors design experiments such as constructing and expressing protein mutants, to assess changes in enzymatic activity through western blot analysis.

We appreciate the reviewer’s valuable suggestion. Our computational modeling shows that acetoacetyl-CoA adopts a binding mode similar to that of acetyl-CoA in the tested HATs. This conclusion is supported by experimental results showing that the addition of acetyl-CoA significantly competed for the binding of acetoacetyl-CoA to HATs, leading to reduced enzymatic activity in mediating Kacac (Fig. 3C). Further structural biology studies to investigate the key amino acid residues involved in Kacac binding within the GCN5/PCAF binding pocket, in comparison to Kac binding—will be a key direction of future studies.

(5) HDAC3 shows strong de-acetoacetylation activity compared to its de-acetylation activity. Specific experiments should be added to verify the molecular docking results. The use of HPLC is recommended, in order to demonstrate that HDAC3 acts as an eraser of acetoacetylation and to support the above conclusions. If feasible, mutating critical amino acids on HDAC3 (e.g., His134, Cys145) and subsequently analyzing the HDAC3 mutants via HPLC and western blot can further substantiate the findings.

We appreciate the reviewer’s helpful suggestion. In-depth characterizations of HDAC3 and other HDACs is beyond this manuscript. We plan in the future to investigate the enzymatic activity of recombinant HDAC3, including the roles of key amino acid residues and the catalytic mechanism underlying Kacac removal, and to compare its activity with that involved in Kac removal.

(6) The resolution of the figures needs to be addressed in order to ensure clarity and readability.

Edits have been made to enhance figure resolutions in the revised manuscript.

Reviewer #3 (Public review):

Summary:

This paper presents a timely and significant contribution to the study of lysine acetoacetylation (Kacac). The authors successfully demonstrate a novel and practical chemo-immunological method using the reducing reagent NaBH4 to transform Kacac into lysine β-hydroxybutyrylation (Kbhb).

Strengths:

This innovative approach enables simultaneous investigation of Kacac and Kbhb, showcasing their potential in advancing our understanding of post-translational modifications and their roles in cellular metabolism and disease.

Weaknesses:

The paper's main weaknesses are the lack of SDS-PAGE analysis to confirm HATs purity and loading consistency, and the absence of cellular validation for the in vitro findings through knockdown experiments. These gaps weaken the evidence supporting the conclusions.

We appreciate the reviewer’s positive comments on the quality of this work and the importance to the field. The SDS-PAGE results of HAT proteins (Supplementary Fig. S3A) was added in the revised manuscript. The cellular roles of p300 and GCN5 as acetoacetyltransferases were confirmed in a recent study (PMID: 37382194). Their data are consistent with our studies herein and provide further support for our conclusion. We agree that knockdown experiments are essential to further validate the activities of these enzymes and plan to address this in future studies.

Reviewer #1 (Recommendations for the authors):

This study conducted the first comprehensive analysis of lysine acetoacetylation (Kacac) in human cells, identifying 139 acetoacetylated sites across 85 proteins in HEK293T cells. Kacac was primarily localized to the nucleus and associated with critical processes like chromatin organization, DNA repair, and gene regulation. Several previously unknown Kacac sites on histones were discovered, indicating its widespread regulatory role. Key enzymes responsible for adding and removing Kacac marks were identified: p300, GCN5, and PCAF act as acetoacetyltransferases, while HDAC3 serves as a remover. The modification depends on acetoacetate, with AACS playing a significant role in its regulation. Unlike Kbhb, Kacac showed unique cellular distribution and functional roles, particularly in gene expression pathways and metabolic regulation. Acetoacetate demonstrated distinct biological effects compared to βhydroxybutyrate, influencing lipid synthesis, metabolic pathways, and cancer cell signaling. The findings suggest that Kacac is an important post-translational modification with potential implications for disease, metabolism, and cellular regulation.

Major Concerns

(1) The authors could expand the study by including different cell lines and also provide a comparative study by using cell lines - such as normal vs disease (eg. Cancer cell like) - to compare and to increase the variability of acetoacetylation patterns across cell types. This could broaden the understanding of the regulation of PTMs in pathological conditions.

We sincerely appreciate the reviewer’s valuable suggestions. We concur that a

deeper investigation into Kacac patterns in cancer cell lines would significantly enhance understanding of Kacac in the human proteome. Nevertheless, due to constraints such as limited resource availability, we are currently unable to conduct very extensive explorations as proposed. Nonetheless, as shown in Fig. 2A, Fig. 2B, and Supplementary Fig. S2, our present data provide strong evidence for the widespread occurrence of acetoacetatemediated Kacac in both normal and cancer cell lines. Notably, our proteomic profiling identified several promising targets implicated in cancer progression, including p53, HDAC1, HMGA2, MTA2, and LDHA. We plan to conduct more comprehensive explorations of acetoacetylation patterns in cancer samples in future studies.

(2) The paper lacks inhibition studies silencing the enzyme genes or inhibiting the enzyme using available inhibitors involved in acetoacetylation or using aceto-acetate analogues to selectively modulate acetoacetylation levels. This can validate their impact on downstream cellular pathways in cellular regulation.

We appreciate the reviewer’s valuable suggestions. Our study, along with the previous research, has conducted initial investigations into the inhibition of key enzymes involved in the Kacac pathway. For example, inhibition of HMGCS, which catalyzes the conversion of acetoacetyl-CoA to HMG-CoA, was shown to enhance histone Kacac levels (PMID: 37382194). In our study, we examined the inhibitory effects of SCOT and HMGCR, both of which potentially influence cellular acetoacetyl-CoA levels. However, their respective inhibitors did not significantly affect histone Kacac levels. We also investigated the role of acetyl-CoA, which competes with acetoacetyl-CoA for binding to HAT enzymes and can function as a competitive inhibitor in histone Kacac generation. Furthermore, inhibition of HDAC activity by SAHA led to increased histone Kacac levels in HepG2 cells (PMID: 37382194), supporting our conclusion that HDAC3 functions as the eraser responsible for Kacac removal. These inhibition studies confirmed the functions of these enzymes and provided insights into their regulatory roles in modulating Kacac and its downstream pathways. Further in-depth investigations will explore the specific roles of these enzymes in regulating Kacac within cellular pathways.

(3) The authors could validate the functional impact of pathways using various markers through IHC/IFC or western blot to confirm their RNA-seq analysis, since pathways could be differentially regulated at the RNA vs protein level.

We agree that pathways can be differentially regulated at the RNA and protein levels. It is our future plan to select and fully characterize one or two gene targets to elaborate the presence and impact of Kacac marks on their functional regulation at both the gene expression and protein level.

(4) Utilize in vitro reconstitution assays to confirm the direct effect of acetoacetylation on histone modifications and nucleosome assembly, establishing a causal relationship between acetoacetylation and chromatin regulation.

We appreciate this suggestion, and this will be a very fine biophysics project for us and other researchers for the next step. We plan to do this and related work in a future paper to characterize the impact of lysine acetoacetylation on chromatin structure and gene expression. Technique of site-specific labelling will be required. Also, we hope to obtain monoclonal antibodies that directly recognize Kacac in histones to allow for ChIP-seq assays in cells.

(5) The authors could provide a site-directed mutagenesis experiment by mutating a particular site, which can validate and address concerns regarding the specificity of a particular site involved in the mechanism.

We agree that validating and characterizing the specificity of individual Kacac sites and understanding their functional implications are important for elucidating the mechanisms by which Kacac affects these substrate proteins. Such work will involve extensive biochemical and cellular studies. It is our future goal to select and fully characterize one or two gene targets in detail and in depth to elaborate the presence and impact of Kacac on their function regulation using comprehensive techniques (transfection, mutation, pulldown, and pathway analysis, etc.).

(6) If possible, the authors could use an in vivo model system, such as mice, to validate the physiological relevance of acetoacetylation in a more complex system.

We currently do not have access to resources of relevant animal models. We will conduct in vivo screening and characterization of protein acetoacetylation in animal models and clinical samples in collaboration with prospective collaborators.

Minor Concerns

(1) The authors could discuss the overlap of Kacac sites with other post-translational modifications and their implications on protein functions. They could provide comparative studies with other PTMs, which can improvise a comprehensive understanding of acetoacetylation function in epigenetic regulation.

We have expanded the discussion in the revised manuscript to address the overlap between Kacac and other post-translational modifications, along with their potential functional implications.

(2) The authors could provide detailed information on the implications of their data, which would enhance the impact of the research and its relevance to the scientific community. Specifically, they could clarify the acetoacetylation (Kacac) significance in nucleosome assembly and its correlation with RNA processing.

In the revised manuscript, we have added more elaborations on the implication and significance of Kacac in nucleosome assembly and RNA processing.

Reviewer #3 (Recommendations for the authors):

Major Comments:

(1) Figures 3A, 3B, Supplementary Figures S3A-D

I could not find the SDS-PAGE analysis results for the purified HATs used in the in vitro assay. It is imperative to display these results to confirm consistent loading amounts and sufficient purity of the HATs across experimental groups. Additionally, I did not observe any data on CBP, even though it was mentioned in the results section. If CBP-related experiments were not conducted, please remove the corresponding descriptions.

We appreciate the reviewer’s valuable suggestion. The SDS-PAGE results for the HAT proteins have been included, and the part in the results section discussing CBP has been updated according to the reviewer’s suggestion in the revised manuscript.

(2) Knockdown of Selected HATs and HDAC3 in cells

The authors should perform gene knockdown experiments in cells, targeting the identified HATs and HDAC3, followed by Western blot and mass spectrometry analysis of Kacac expression levels. This would validate whether the findings from the in vitro assays are biologically relevant in cellular contexts.

We appreciate the reviewer’s valuable suggestion. Our identified HATs, including p300 and GCN5, were reported as acetoacetyltransferases in cellular contexts by a recent study (PMID: 37382194). Their findings are precisely consistent with our biochemical results, providing additional evidence that p300 and GCN5 mediate Kacac both in vitro and in vivo. In addition, inhibition of HDAC activity by SAHA greatly increased histone Kacac levels in HepG2 cells (PMID: 37382194), supporting the role of HDAC3 as an eraser responsible for Kacac removal. We plan to further study these enzymes’ contributions to Kacac through gene knockdown experiments and investigate the specific functions of enzyme-mediated Kacac under some pathological contexts.

Minor Comments:

(1) Abstract accuracy

In the Abstract, the authors state, "However, regulatory elements, substrate proteins, and epigenetic functions of Kacac remain unknown." Please revise this statement to align with the findings in Reference 22 and describe these elements more appropriately. If similar issues exist in other parts of the manuscript, please address them as well.

The issues have been addressed in the revised manuscript based on the reviewer's comments.

(2) Terminology issue

GCN5 and PCAF are both members of the GNAT family. It is not accurate to describe "GCN5/PCAF/HAT1" as one family. Please refine the terminology to reflect the classification accurately.

The description has been refined in the revised manuscript to accurately reflect the classification, in accordance with the reviewer's suggestion.

(3) Discussion on HBO1

Reference 22 has already established HBO1 as an acetoacetyltransferase. This paper should include a discussion of HBO1 alongside the screened p300, PCAF, and GCN5 to provide a more comprehensive perspective.

More discussion on HBO1 alongside the other screened HATs has been added in the revised manuscript.