https://www.nature.com/articles/s41392-026-02757-7
I find the issue of PTMs really interesting. Obviously lactylation is more likely when there is more lactate around. I am not sure this paper identifies the electrostatic changes from lactylation. I will perhaps try separately to do that and add it to this.
chatGPT(5.5paid):
Paper overview
Yang et al., “Protein lactylation in health and diseases: molecular mechanisms, biological significance, and clinical implications” is a 2026 narrative review in Signal Transduction and Targeted Therapy. It surveys lysine lactylation from its biochemical origins through its proposed physiological, pathological and therapeutic roles. It is not an experimental paper or systematic review.
Summary
1. Lactylation links metabolism to protein regulation
Lysine lactylation, abbreviated Kla, is the covalent attachment of a lactyl group to a lysine residue. The central proposition is that lactate is not merely a glycolytic product or signalling metabolite: it can also influence cell behaviour by altering histones and numerous non-histone proteins.
The authors distinguish two chemically and mechanistically different processes:
- Enzymatic L-lactylation, generally associated with L-lactate and either lactyl-CoA or a lactate–AMP intermediate.
- Non-enzymatic D-lactylation, arising principally from the reactive glycolytic detoxification intermediate S-D-lactoylglutathione.
This distinction is important because “lactylation” is not necessarily one uniform biological process.
2. The machinery is broader than originally thought
The review catalogues four classes of regulator.
Writers
These include familiar acetyltransferases that can apparently use lactyl-CoA:
- p300/CBP
- KAT2A/GCN5 and KAT2B
- TIP60/KAT5
- HBO1/KAT7
- KAT8
- NAA10
A particularly important recent development is the identification of AARS1 and AARS2 as ATP-dependent lactyltransferases. These enzymes can activate lactate as lactate–AMP and transfer the lactyl group directly, bypassing the need for lactyl-CoA. Examples include lactylation of p53, YAP, TEAD, cGAS, PDHA1 and CPT2.
Readers
Only two relatively convincing readers are listed:
- BRG1, recognising H3K18la
- DPF2, recognising H3K14la
This remains one of the least developed parts of the field.
Erasers
Several deacetylases also act as delactylases:
- HDAC1–3, particularly on histones
- SIRT1–3, especially on non-histone or compartment-specific substrates
Lactyl-CoA-generating enzymes
The paper highlights recent evidence that:
- ACSS2 can generate lactyl-CoA and cooperate with KAT2A.
- Mitochondrial GTPSCS may generate lactyl-CoA and, after nuclear translocation, cooperate with p300.
The review itself acknowledges that intracellular lactyl-CoA concentrations are much lower than acetyl-CoA concentrations and that the conditions determining whether these enzymes catalyse acetylation or lactylation remain poorly established.
3. Lactylation has effects far beyond histone transcription
The paper’s most useful conceptual contribution is to organise the literature by molecular consequence, rather than treating lactylation merely as another histone mark.
Reported effects include:
- altered gene transcription and chromatin accessibility;
- stabilisation or degradation of proteins;
- activation or inhibition of metabolic enzymes;
- altered protein–protein interactions;
- nuclear, mitochondrial, lysosomal or extracellular trafficking;
- competition or cooperation with acetylation, methylation, phosphorylation and ubiquitination;
- regulation of RNA methylation and alternative splicing;
- DNA repair and genome stability;
- modulation of liquid–liquid phase separation.
The diagram on page 7 provides a particularly effective map of these functions, covering DNA, RNA and protein-level consequences.
Examples include:
- p53 K120/K139 lactylation, which reduces DNA binding, transcriptional activity and phase separation.
- cGAS lactylation, which inhibits DNA sensing and cGAS phase separation.
- PDHA1 and CPT2 lactylation, which suppresses pyruvate oxidation and fatty-acid oxidation.
- VPS34 lactylation, which enhances autophagy-related kinase activity.
- MRE11, NBS1 and XRCC1 lactylation, which can facilitate DNA repair.
- TFEB lactylation, which reduces ubiquitination and increases TFEB stability and autophagic activity.
- Tufm lactylation, which can impair mitochondrial localisation and mitophagy.
Thus lactylation is presented not simply as a readout of glycolysis but as a possible feedback mechanism that remodels metabolism, stress responses and cell fate.
4. Interaction with RNA processing is an emerging theme
The review includes a valuable section on RNA regulation.
Proposed mechanisms include:
- H3K18la-induced expression of splicing factor SRSF10;
- lactylation of nucleolin, altering MADD pre-mRNA processing;
- alternative splicing of ACLY generating isoforms with different lactylation patterns;
- lactylation or transcriptional regulation of RNA methylation machinery, including METTL3, METTL16, NSUN2, IGF2BP3, FTO, ALKBH3 and YTHDF2.
However, most examples represent individual disease-specific pathways rather than evidence that lactylation is a general regulator of spliceosome assembly or global splicing fidelity.
5. Physiological roles
The review describes lactylation in:
- somatic-cell reprogramming;
- early embryonic development and implantation;
- neural development;
- cochlear development;
- immune adaptation and tissue repair.
A recurring model is:
increased glycolysis → increased lactate → increased histone lactylation → altered developmental or repair-gene expression.
For example, H3K18la is proposed to promote pluripotency-associated genes during induced pluripotent stem-cell reprogramming. Nevertheless, the authors appropriately concede that the mechanistic evidence in normal physiology is much less complete than the expanding disease literature.
6. Disease associations
The review covers a very large range of conditions, especially:
- solid and haematological cancers;
- neurodegenerative, traumatic and psychiatric disorders;
- cardiovascular and vascular disease;
- retinal and other ophthalmic disease;
- infection, sepsis and inflammatory disease;
- metabolic liver disease, obesity and kidney disease;
- fibrosis, osteoporosis and intervertebral-disc degeneration.
Cancer occupies the largest section. The authors present lactylation as a mechanism through which the Warburg phenotype may promote:
- oncogenic transcription;
- metabolic adaptation;
- stemness;
- DNA repair;
- invasion and metastasis;
- suppression of antitumour immunity;
- treatment resistance.
An important theme is positive feedback. For example:
glycolysis produces lactate → lactylation activates a glycolytic or oncogenic pathway → still more glycolysis and lactate.
However, the direction of effect is highly context-dependent. Lactylation can promote tumour survival in one setting but enhance autophagy, DNA repair, differentiation or tissue recovery in another.
7. Biomarkers and therapies
Large lactyl-proteomic studies have identified thousands of putative sites. The review suggests that site-specific lactylation patterns might eventually serve as diagnostic or prognostic biomarkers. It also discusses machine-learning signatures built from “lactylation-related genes”, although these are usually expression signatures rather than direct measurements of protein lactylation.
Therapeutic approaches are grouped into:
- reducing glycolysis or LDH activity;
- blocking lactate transport through MCTs;
- modulating PDH or PKM2;
- inhibiting p300/CBP or other regulatory enzymes;
- targeting individual lactylated sites or proteins with peptides or antibodies;
- combining lactate/lactylation inhibition with chemotherapy or immunotherapy.
Table 3 is valuable as a catalogue, but almost all specifically lactylation-directed approaches remain preclinical. AZD3965, an MCT1 inhibitor, has entered clinical testing, but it alters lactate transport broadly and is not a selective “lactylation drug”.
What is novel about the paper?
Because this is a review, its novelty is conceptual synthesis rather than new experimental discovery.
1. It incorporates the rapidly expanded 2024–2026 enzymology
Earlier reviews commonly described lactylation mainly as p300-mediated histone modification. This paper integrates newer findings on:
- AARS1/AARS2-mediated lactate–AMP chemistry;
- ACSS2 and GTPSCS as proposed lactyl-CoA synthetases;
- recently identified readers BRG1 and DPF2;
- a much larger range of site-specific non-histone modifications.
That produces a more complete and current model of the lactylation machinery.
2. It treats lactylation as a whole-proteome regulatory system
The paper goes substantially beyond the original histone-centred account. It frames lactylation as affecting:
- transcription;
- RNA processing;
- metabolism;
- DNA repair;
- protein trafficking;
- mitophagy and autophagy;
- protein condensation and phase separation.
This is probably its strongest contribution.
3. It explicitly integrates L-lactylation and D-lactylation
The distinction between enzyme-mediated L-lactylation and reactive, non-enzymatic D-lactylation is chemically important and often blurred in less comprehensive accounts.
4. It highlights bidirectional feedback
Rather than proposing only that metabolism controls epigenetics, the review shows how lactylation may feed back onto glycolysis, mitochondrial oxidation and metabolic enzyme stability. The implied model is:
metabolism → lactylation → transcription/protein function → altered metabolism.
5. It links lactylation to alternative splicing, RNA methylation and phase separation
These are relatively recent branches of the literature and distinguish this review from narrower cancer- or histone-focused articles.
6. It provides unusually broad reference tables
Table 1 catalogues writers, erasers, readers, substrates and sites; Table 2 links individual sites to physiological and pathological settings; and Table 3 connects possible interventions to their proposed lactylation targets. These tables are useful reference resources even though the underlying evidence varies greatly in quality.
Critique
Strengths
Exceptionally comprehensive
The paper brings together a very large and fast-moving literature and is particularly strong as a reference source for individual enzymes, proteins, lysine sites and disease models.
Good molecular organisation
Organising the evidence by molecular effects—stability, activity, interactions, localisation, RNA processing and phase separation—makes the field easier to understand than a simple disease-by-disease catalogue.
Recognises mechanistic heterogeneity
The distinction among lactyl-CoA-dependent transfer, AARS-mediated lactate–AMP transfer and non-enzymatic D-lactylation is scientifically important.
Does not limit itself to histones
The review correctly reflects that much of the most interesting recent work concerns non-histone proteins, including metabolic enzymes and signalling factors.
Main weaknesses
1. It is descriptive rather than evidence-weighted
The review frequently gives similar narrative status to findings derived from:
- purified-enzyme experiments;
- cultured tumour cells;
- genetically modified mice;
- human tumour correlations;
- proteomic site assignments;
- computational gene signatures.
These forms of evidence are not equivalent. A formal hierarchy distinguishing biochemical demonstration, cellular causality, animal validation and human evidence would have made the conclusions much more reliable.
2. Detection of a modification is often treated too readily as functional proof
Mass spectrometry can identify a modified peptide, but this does not establish that the modification:
- occurs at functionally important stoichiometry;
- is dynamically regulated;
- is enzymatically written;
- is required for the phenotype;
- rather than merely accompanying high glycolytic flux.
For many sites, the likely occupancy is unknown. Thousands of detected lactylated residues do not necessarily imply thousands of meaningful regulatory switches.
3. Causality is frequently difficult to separate from lactate metabolism
Interventions such as:
- LDH inhibition;
- glycolysis inhibition;
- MCT blockade;
- lactate addition;
- p300 inhibition;
- HDAC or sirtuin manipulation
alter many processes besides lactylation. They affect ATP production, NADH/NAD⁺ balance, intracellular pH, metabolite transport, acetylation and signalling.
Therefore:
“reducing lactate changed the phenotype” does not by itself prove “reduced lactylation caused the phenotype”.
The strongest studies use site-specific lysine mutants, direct biochemical assays and rescue experiments, but this standard is not consistently applied across the literature reviewed.
4. Competition with acetylation is under-analysed
Many proposed writers and erasers are better established as acetylation enzymes. Lactyl-CoA is reportedly much less abundant than acetyl-CoA, and lactylation kinetics are slower.
This raises several unresolved questions:
- Under physiological conditions, when does lactyl-CoA successfully compete with acetyl-CoA?
- Does apparent lactyltransferase activity require unusually high experimental lactate or lactyl-CoA concentrations?
- Are some phenotypes attributed to lactylation actually caused by reciprocal loss of acetylation at the same lysine?
- Do p300 or HDAC inhibitors act mainly through acetylation rather than lactylation?
The paper acknowledges substrate competition and unknown enzyme specificity but does not critically quantify how serious this problem is.
5. The nomenclature risks conflating distinct chemistries
L-lactyllysine and D-lactyllysine may have different precursors, stereochemistry, kinetics and biological interpretation. Pan-lactyllysine antibodies or standard proteomics workflows may not always distinguish them satisfactorily.
A stronger review would separate:
- enzymatic L-lactylation;
- non-enzymatic D-lactylation;
- direct lactyl-CoA transfer;
- AARS-mediated lactate–AMP transfer
throughout every disease section, rather than sometimes discussing them under the broad label “Kla”.
6. Reader biology is still extremely thin
Only BRG1 and DPF2 are listed as readers, and even these require broader validation. Until specific recognition modules are better established, it remains uncertain whether many histone lactyl marks constitute a true, independently interpreted epigenetic code or whether they mainly alter charge, nucleosome structure or competition with other acylations.
7. Disease coverage is broad but sometimes becomes catalogue-like
The extensive disease sections demonstrate scope, but the resulting narrative can obscure repeated mechanistic patterns. A smaller number of carefully evaluated causal pathways would have been more informative than presenting virtually every published association.
8. The translational discussion is too optimistic
Most proposed therapies act upstream on glycolysis or lactate handling rather than selectively on lactylation. Such interventions may have substantial effects on:
- normal oxidative tissues;
- exercise metabolism;
- immune-cell function;
- tissue repair;
- neuronal lactate shuttling;
- developmental processes.
Because lactylation may be protective in some settings and harmful in others, global inhibition could be counterproductive. There is not yet a validated clinical assay, disease-specific therapeutic window or highly selective writer/reader/eraser drug for most applications.
9. Biomarker claims need stronger validation
Many proposed “lactylation signatures” are derived from gene-expression databases. Expression of lactylation-associated genes is not the same as measurement of lactylated protein sites.
For clinical use, candidate biomarkers would require:
- analytically validated site-specific assays;
- prospective cohorts;
- comparison with standard clinical predictors;
- independent replication;
- evidence of incremental discrimination or clinical utility.
The review does not apply these criteria rigorously.
10. It is not a systematic review
The paper does not provide a reproducible search strategy, inclusion criteria, risk-of-bias assessment or evidence grading. Its breadth is valuable, but publication bias is likely: positive mechanistic findings will be disproportionately represented, while failed replication and null studies may be absent.
Overall assessment
This is a highly useful and unusually current reference review. Its main scientific message—that lactylation may form a reversible metabolic regulatory layer extending from chromatin to RNA processing, metabolic enzymes, DNA repair and protein organisation—is persuasive as a research framework.
However, the evidence does not yet justify treating lactylation as a single, well-defined master regulatory system or a clinically validated therapeutic target. The field remains dominated by cell and animal experiments, low or uncertain modification stoichiometry, pleiotropic metabolic interventions and incomplete chemical discrimination among different forms of lactylation.
The most defensible conclusion is:
Lactylation is a genuine and potentially important family of metabolically sensitive protein modifications, but only a subset of the thousands of reported sites has yet been shown to be causal, physiologically abundant and therapeutically actionable.
The highest-priority next steps are quantitative site occupancy, stereoisomer-specific detection, direct measurement of compartmental lactyl-CoA or lactate–AMP, site-specific genetic validation, comparison with competing acetylation, and prospective human studies.
electrostatic etc
Not explicitly. I found no discussion in the paper using terms such as “electrostatic”, “charge neutralisation”, or “loss of lysine positive charge.”
The paper does say that lactylation can alter:
- protein conformation and stability;
- protein–protein interactions;
- DNA binding;
- RNA binding;
- phase separation;
- chromatin accessibility.
For example, it describes lactylation of p53 as impairing DNA binding and liquid–liquid phase separation, and lactylation of various proteins as either strengthening or weakening molecular interactions. But it generally reports these as observed functional effects rather than explaining them through electrostatics.
Chemically, an electrostatic mechanism is quite plausible. Unmodified lysine normally carries a positively charged ε-amino group at physiological pH. Lactylation converts that amine into a neutral amide:
Lys−NH3+→Lys−NH−CO−CHOH−CH3
This removes a positive charge and adds a larger polar group. It could therefore:
- weaken attraction to negatively charged DNA or RNA;
- alter salt bridges within proteins;
- change protein–protein binding surfaces;
- reduce histone–DNA attraction;
- affect condensate formation and phase separation.
The paper’s examples are compatible with such effects, but the review does not clearly identify charge neutralisation as the underlying mechanism. Indeed, near its conclusion it says that future molecular studies should investigate how lactylation changes protein structure and the precise mechanisms by which those structural effects alter function.
So the electrostatic consequence is an important mechanistic omission from an otherwise broad review.