DPF2 reads histone lactylation to drive transcription and tumorigenesis

Abstract

Lysine lactylation (Kla) is a new type of histone mark implicated in the regulation of various functional processes such as transcription. However, how this histone mark acts in cancers remains unexplored due in part to a lack of knowledge about its reader proteins. Here, we observe that cervical cancer (CC) cells undergo metabolic reprogram by which lactate accumulation and thereby boosts histone lactylation, particularly H3K14la. Utilizing a multivalent photoaffinity probe in combination with quantitative proteomics approach, we identify DPF2 as a candidate target of H3K14la. Biochemical studies as well as CUT&Tag analysis reveal that DPF2 is capable of binding to H3K14la and colocalizes with it on promoters of oncogenic genes. Notably, disrupting the DPF2-H3K14la interaction through structure-guided mutation blunts those cancer-related gene expression along with cell survival. Together, our findings reveal DPF2 as a bona fide H3K14la effector that couples histone lactylation to gene transcription and cell survival, offering insight into how histone Kla engages in transcription and tumorigenesis.

Keywords: chemical proteomics; epigenetics; histone modifications; histone reader; lysine lactylation.

Fig. 1.

Histone lactylation correlates with cell proliferation and tumorigenesis. (A) The relative LDH levels of CESC patients and normal controls analyzed by GEPIA database. (B) Kaplan–Meier plots showing the survival of CESC patients with low and high LDH levels. (C) Cell proliferation assay of Hela cells treated with control (vehicle) or LDH inhibitor (LDHi). Cells were counted for 4 d after seeding. (D) Western blotting analysis of Pan-Kla and site-specific histone lactylation (H3K14la) and acetylation (H3K14ac) from cells feeding with or without NaLa. (E) Quantification and statistical analyses of the data in (D).

Fig. 2.

Identification of DPF2 as H3K14la target using multivalent photoaffinity probes. (A) Schematic workflow for capture and identification of H3K14la binders by self-assembly multivalent photoaffinity probes in combination with comparative proteomics. (B) Volcano plot of protein fold enrichment and statistical analysis from quantitative proteomics experiments (n = 3) by leveraging probes H3K14la and H3K14. Red hits are proteins that meet the criteria (ratio ≥ 1.5 and P-value ≤ 0.05). (C) Molecular function analysis of H3K14la candidate binders by gene ontology (GO). Several proteins involved in transcription are highlighted in the box.

Fig. 3.

DPF2 binds H3K14la through its DPF domain in vitro. (A and B) Isothermal titration calorimetry (ITC) fitting curves of the DPF2 titrated with unmodified (H3K14un) or lactylated histone H3 (1-20) peptides (H3K14la) (A), and ITC fitting curves of WT-DPF2 and its mutants with H3K14la peptide (B). (C) Overall structure of DPF2 bound to H3K14la in ribbon view. The Kla and its two key surrounding residues are shown in the close-up window. (D) Electrostatic potential surface view of the DPF2 space-filled by H3K14la peptide. (E) Hydrogen bonding network between H3K14la peptide and DPF2. Hydrogen bonds are illustrated as yellow dashes; Key residues of DPF2 are exhibited as sticks and labeled white.

Fig. 4.

DPF2 is intracellular associated with H3K14la. (A) Schematic of DPF2–histone interaction using immunoprecipitation (IP) assays in combination with western blotting (WB) analysis. (B) In vivo analysis of interaction between DPF2 and histone lactylation. Hela cells are transfected with Flag-WT-DPF2 or its mutants, and the lysates are immunoprecipitated using anti-Flag beads, followed by WB analysis of H3K14la. (C) Representative immunofluorescence (IF) images of cells reacted with antibodies against Flag (red) or H3K14la (green). Cells transfected with Flag-tagged DPF2 or its mutants are used for IF analysis. (D) Quantification of the Flag-DPF2 or its mutants colocalized with H3K14la via the pearson’s correlation analysis.

Fig. 5.

DPF2 colocalizes with H3K14la across genome and mediates gene transcription. (A) Venn diagram showing the overlap of H3K14la-and DPF2-enriched genes in Hela cells. (B) Distribution of overlapped peaks at annotated genomic regions. (C) Heatmap illustrating the genomic occupancy of H3K14la and DPF2 at promoter regions. The genes shown in rows are classified in descending order via signal intensity. (D) GO enrichment analysis of genes occupied by H3K14la and DPF2. (E) Averaged genome-wide occupancies of WT-DPF2 and Mut-DPF2 around gene promoters. (F) The genome browser view of WT-DPF2, Mut-DPF2, and H3K14la at indicated gene locus. The y axis represents normalized read intensity. Gene model is shown below the profiles. (G) qRT-PCR analysis of the expression of indicated genes in cells as in (F). **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 6.

DPF2 links histone lactylation to cell survival. (A) Immunoblotting analysis showing the silencing efficiency of siRNA targeting DPF2. (B) Cell proliferation assay in control (shNC plus vector) and DPF2-KD (sh2) cells expressing WT-DPF2 or its indicated mutants. (C) Clonogenic formation assay of cells as in (B). (D) Quantification analyses of the data in (C). **P < 0.01; ***P < 0.001; ****P < 0.0001. (E) Schematic illustration of findings of the study. The DPF2 is recruited to H3K14la-marked chromatin at gene promoters, thereby driving oncogenes expression. Disrupting the H3K14la–DPF2 recognition or LDH inhibition leads to the suppressed transcription and tumorigenesis.