Lactylation in cancer: Unveiling new layers of complexity

Main text

Since the debut of the Warburg effect, our understanding of lactate in cancer has evolved from a metabolic waste of "low-efficient" glucose metabolism, an acidification factor reshaping tumor microenvironments, to key molecular signals modulating signaling pathways, thereby influencing cell fates. Recently, Zhang’s work introduced a novel post-translational modification (PTM), lactylation, revealing a previously unidentified identity of lactate. Further findings in Plantae and Bacteria have projected lactylation as a common PTM among biological kingdoms.¹

Current lactylation writers can be categorized into lactyl-coenzyme A (CoA)-dependent or -independent manners (Figure 1, left). In the presence of lactyl-CoA, some histone acetyltransferases function as lactyltransferases, such as p300 and HBO1. They added lactyl lysine residues mainly in the promoter regions. Lactylation elevated the expression of METTL3 and "m6A reader" YTHDF2, indicating intricate lactylation regulation. This complexity of lactylation in gene expression regulation can be further highlighted by its impact on translational elongation and protein synthesis through lactylated EEF1A2 and on alternative splicing regulation by lactylated non-histone proteins such as Nucleolin. Furthermore, non-histone proteins lactylated by KAT5/TIP60 and CBP were involved in DNA damage response and chemotherapy resistance. These findings highlight a versatile regulatory mechanism of lactylation. These "non-lactylation-specific" writers and substrates project a potential sophisticated interaction with other types of PTMs, acting as context (e.g., cellular lactate level)-dependent orchestrators of cellular function.

Figure 1.

Mechanisms of lysine lactylation and potential role of lactylation in microtubule-related mitotic abnormalities in cancer cells

Left: enzymatic and non-enzymatic mechanisms of lactylation. Right: investigation of lactylation in microtubule-related mitotic abnormalities in cancer cells. Mass spectrometry is widely used to identify PTMs, including lactylation, with lysine lactylation being the most extensively studied. In addition, previous mass spectrometry analyses have identified lysine lactylation as well as non-lysine lactylation. We have successfully synthesized certain non-lysine lactylated peptides (e.g., lactylation of asparagine [N]), supporting the potential existence of this modification. In the analysis of four lactylation datasets, lactylated proteins have been consistently identified as enriched in mitotic spindle organization, suggesting that the lactylation of these proteins may play a role in cancer cell mitosis. This modification could potentially disrupt microtubule dynamic instability, affect chromosome segregation, and contribute to centrosome amplification.

Lately, two studies have uncovered that specific pairs of acetyltransferases and lactyl-CoA synthetase added lactyl groups to lysine residues, GTPSCS/p300 and ACSS2/KAT2A. This resolves the long-standing question of how lactyl-CoA is synthesized, which also highlights the complexity of regulating lactylation processes. Other than lactyltransferases classified as histone acetyltransferases, a novel type of lactyltransferase has been identified as alanyl-tRNA synthetase 1 (AARS1) in the cytosol and AARS2 in mitochondria. Instead of lactyl-CoA, they lactylated non-histone proteins using lactate and ATPs. Notably, AARS1 lactated p53, impairing its lipid-lipid phase separation and DNA-binding capacity, and lactated YAP to influence its distribution. AARS2 lactylated PDHA1 and CPT2 to modify metabolic pathways. Hence, lactylation further emphasizes the intricate and dynamic nature of protein regulation across biological cellular processes.

Lactylation shows a preference for lactyl substrate, as all the above lactylation types are L-lactyl (L-la) modifications. However, isomers of L-la contribute to this PTM mainly via the glyoxalase pathway. Non-enzymatic covalent lactylation gives rise to D-lactyl (D-la), where lactoylglutathione serves as a donor of D-la to modify lysine residues, primarily on key glycolytic enzymes regulating metabolic outputs. Another isomer is carboxyethyl. The reactive dicarbonyl compound, methylglyoxal, reacts with lysine residues to form carboxyethyl lysine. These modifications also impact protein stability and function, influencing cellular metabolism and signaling pathways.

Despite these advancements, novel lactyltransferases, crucial for fully understanding the biological functions of this modification, are still awaiting discovery. Currently, they are predominantly identified by exploring acetyltransferases, which are hindered by the vast number of proteins with acetyltransferase activity and the lack of clarity regarding the lactylation-specific catalytic domains/motifs. Advances in artificial intelligence, such as AlphaFold, render great potential to identify novel lactyltransferases based on structural similarity, as large-scale AI foundation models exhibit strong generalization capability to be transferrable to different structures. Additionally, the site-specific antibody for lactylation hampers precise analysis of substrate functions, which could be addressed by wet-lab screening methods.

Despite being in its nascent stages, lactylation research provides some insights into a broader role in cellular functions. Recently, two well-known microtubule-associated proteins (MAPs), NUSAP1 and CENPA, were lactylated in pancreatic ductal adenocarcinoma (PDAC) and hepatocellular carcinoma (HCC), respectively. However, how lactylation impacts the cytoskeleton is untouched. Interestingly, lactylated cytoskeletal proteins are shared in colorectal, gastrointestinal, gastric cancers, and HCC,^2,3,4,5 suggesting a conserved role of cytoskeleton lactylation in cancer cells. Intriguingly, mass spectrometry data also revealed lactylation on non-lysine sites of various proteins, including some tubulin isoforms and MAPs across several types of cancer cells (Figure 1, right).² While further investigation of these proteins is necessary, lactylation of cytoskeletal proteins could interpret abnormalities during cancer cell mitosis (e.g., defects in spindle formation and chromosome segregation). These mitotic irregularities might be influenced by the disruption of the dynamic instability of microtubules and the mechanical force generated from the polymerization and depolymerization of microtubules where lactylation modulates the interaction/binding among tubulin heterodimers and other cytoskeletal proteins such as MAPs.

One of the cytoskeletal proteins that was lactylated in the aforementioned four cancer types is TPX2, a microtubule nucleation factor crucial for spindle assembly during mitosis. Additionally, NUMA1, which is closely associated with centrosome amplification when dysregulated, was shown to be lactylated at multiple sites. Coincidentally, centrosome amplification is one of the main characteristics of nearly all solid tumor cells. So, is there a causal relationship between centrosome abnormality in tumor cells and accumulated lactate leading to the lactylation of various MAPs? This lactate-driven modification also extends to other key centrosome-associated proteins, such as CEP135 (lactylated at K983) and CEP170 (lactylated at K412 and K767) in multiple gastrointestinal cancers, highlighting the potential broad impact of lactylation on the integrity of spindle formation and the overall fidelity of cell division in cancer cells.

TPX2 plays a key role in both microtubule nucleation and dynamics, both of which are regulated by RAN. The identification of lactylated RAN further suggests lactylation on cytoskeletal proteins, potentially influencing mitotic spindle assembly. The effect of lactylation on TPX2 may be apparent in the “search-and-capture” model through regulating spindle assembly rather than in spindle pole formation. To effectively locate chromosomes and facilitate their oscillation for proper alignment at the metaphase plate, spindle microtubules must undergo dynamic regulation through polymerization and depolymerization. Therefore, disruption of this instability—via lactylation of MAPs—may impair proper chromosome alignment. A recent study reported an elongated maximum length of microtubules caused by α-tubulin lactylation. Consequently, similar effects on chromosome motions might be observed if lactylation on tubulins enhances microtubule growth. In our study, we also identified lactylation in several tubulin isotypes at both lysine and non-lysine residues. This suggests that tubulin lactylation may regulate the interaction strength at different residues by enhancing or reducing it, thereby enabling a context-dependent mechanism for regulating microtubule dynamics.

For later faithful chromosome segregation in mitosis, kinetochores on two sister chromatids must be securely attached to spindles in a bioriented fashion. Unstable kinetochore-microtubule (K-T) attachment is thought to provide an opportunity for "self-correction." The presence of lactylated WAPL and CHAMP1 suggests that lactylation also plays a role in chromosome segregation. WAPL facilitates the release of cohesin from chromosomes, enabling proper chromosome segregation during mitosis, while CHAMP1 regulates K-T attachments to ensure accurate chromosome alignment and segregation. Hence, “self-correction” mechanisms could be disrupted by the lactylation-induced alterations in K-T attachments, leading to merotelic, syntelic, or monotelic attachments, the phenomenal phenotypes commonly observed in cancer cells. Consequently, aneuploidy and chromosome instability in cancer cells may be partially attributable to lactylation-caused defects through tubulins, MAPs, and other factors participating in K-T attachments.

Due to the pivotal role of microtubules in cancer progression, drugs that disrupt their dynamics have been among the most promising cancer treatments, either by stabilization (taxanes and epothilones) or deconstruction (vinca alkaloids and eribulin). Unfortunately, patients often develop resistance to these drugs, potentially due to the efflux of drugs from cancer cells. However, the impact of lactylation cannot be ruled out, as drug resistance is frequently associated with elevated lactate levels in tumor tissues. Here, we interpret some characteristics of cancer cell mitosis from the perspective of lactylation-induced changes in microtubule dynamic instability and the interactions among cytoskeletal proteins. The influence of lactylated cytoskeletal proteins is undoubtedly complex, considering the variation of lactylated sites, modified levels of proteins, and potential competition with other types of PTMs. Future investigations into lactylation will provide valuable insights into cancer progression and drug resistance, thereby developing more effective therapies or combination treatments.

Funding and acknowledgments

This work was supported by the National Natural Science Foundation of China (NSFC) (T2225006, T2488301, and 82272948 to M.L.) and the Beijing Municipal Natural Science Foundation (Key Program Z220011 to M.L.).

Declaration of interests

The authors declare no competing interests.

Published Online: April 22, 2025