Metabolic reprogramming regulates histone lactylation during zebrafish caudal fin regeneration

Summary

Tissue regeneration relies on precise molecular mechanisms controlling cell-fate transitions, with metabolism emerging as a key regulator. Lactate-derived histone lactylation has recently been identified as an epigenetic modification regulating gene expression across various biological processes. Here, we report an increase in global histone lactylation in the mesenchyme and osteoblasts of the zebrafish caudal fin during early regeneration. Our findings demonstrate that this epigenetic modification is functionally regulated by increased lactate levels, while the inhibition of glycolysis and lactate production significantly reduces histone lactylation. Transcriptomic profiling under reduced lactylation revealed the downregulation of proliferative and chromatin-remodeling programs. This suggests a model in which injury-induced, lactate-driven histone lactylation sustains chromatin accessibility and promotes proliferative transcription during early regeneration, potentially modulating gene expression essential for cell plasticity and proliferation. This study identifies histone lactylation as a metabolic-epigenetic regulator of regenerative programs, providing mechanistic insights for the development of novel therapeutic strategies to enhance tissue repair.

Graphical abstract

Highlights

• •Histone lactylation increases during early zebrafish caudal fin regeneration
• •Glycolysis-derived lactate is required to maintain histone lactylation
• •Lactylation inhibition downregulates proliferative and chromatin-remodeling genes

Natural sciences; Biological sciences; Physiology; Developmental biology

Introduction

During tissue and organ regeneration, there is a precise control over cell fate changes, like dedifferentiation to generate proliferative progenitors and redifferentiation to restore lost tissues. Recently, cell metabolism has emerged as a key driver of these transitions, with differentiated cells relying on oxidative phosphorylation for an energy source, while stem and proliferative cells favor aerobic glycolysis with lactate production.¹ This change between metabolic states is commonly described as metabolic reprogramming. Several reports, including our previous study, have shown that metabolic reprogramming plays a vital role in endogenous regenerative programmes.^2,3,4,5,6,7 We identified a critical role for metabolism in triggering early events of zebrafish caudal fin regeneration, with bone-producing osteoblasts prioritizing glycolysis and lactate production during dedifferentiation.⁷ Mature osteoblasts dedifferentiation is triggered as early as 6 h post-amputation (hpa), alongside the initial wound healing response.^8,9,10 Osteoblasts re-enter the cell cycle and incorporate the blastema, a cluster of undifferentiated proliferating cells, by 12–24 hpa, and redifferentiate into mature osteoblasts during the outgrowth phase, from 48 hpa to 20 days post-amputation, completing fin regeneration.^11,12,13 While the importance of metabolic reprogramming in regeneration is increasingly recognized, its molecular links to cell fate determination remain poorly understood.

Epigenetic modifications, such as DNA methylation and histone post-translational modifications (PTMs), influence chromatin accessibility, thereby regulating spatiotemporal gene expression during tissue regeneration.¹⁴ Injury-induced gene expression relies on the activation of regeneration-associated enhancers, highlighting the importance of the mechanisms underlying gene activation post-injury in tissue regeneration.¹⁵ In zebrafish, DNA and histone methylation levels decrease during cellular dedifferentiation and proliferation, such as during the blastema formation in caudal fin regeneration, and latter increase to silence proliferative genes as cells redifferentiate during caudal fin^16,17 and heart regeneration.¹⁸ Conversely, histone acetylation activates pluripotency-related genes early in fin regeneration, later decreasing to facilitate cell redifferentiation.¹⁹ DNA and histone modifications often rely on metabolites derived from glucose metabolism, linking metabolic activity to gene regulatory programmes and cell fate decisions.²⁰ Recently, lactate-derived histone lysine lactylation has emerged as a novel PTM, promoting gene transcription in various processes, including inflammation,²¹ embryonic development,^22,23 tumor growth^24,25,26 and somatic cell reprogramming.²⁷ Lactate also acts as signaling molecule in various biological processes, including tissue repair by promoting an anti-inflammatory macrophage phenotype during ischemic muscle regeneration in mice.^28,29 However, the role of lactylation in regulating cell fate transitions during regeneration has not been addressed, making it crucial to understand how histone lactylation influences tissue repair and regeneration through metabolic adaptation and epigenetic modulation.

This study is the first to report histone lactylation in zebrafish regeneration, demonstrating a significant increase in global lactylation levels in mesenchyme and osteoblast cells during the early stages of caudal fin regeneration. Our findings suggest that this increase is functionally regulated by glycolysis and lactate production, placing lactate-derived histone lactylation as a potential orchestrator of the initial stages of regeneration.

Results

Histone lactylation is increased during caudal fin regeneration

We have previously reported that cells responding to fin amputation reprogram their metabolic profile by increasing glycolytic activity and lactate production, which is required for blastema formation during the initial stage of caudal fin regeneration.⁷ Considering that this metabolic shift leads to the accumulation of lactate, we hypothesized that histone lactylation could play a role in the regenerative process. As histone lactylation had not been studied before in zebrafish regeneration, to address its potential function we started by performing a Western Blot (WB) analysis on caudal fin tissue using a pan anti-Kla antibody, which targets lactylated lysines (Kla) across all proteins, including histones.²¹ Full protein extraction showed low intensity in the histone molecular weight range, suggesting lower levels of lactylated lysines in histones relative to other proteins (Figure 1A). However, by enriching the histone fraction from the protein extract, we successfully detected Kla in the molecular weight interval that includes histones, confirming the presence of histone lactylation in zebrafish caudal fin tissue (Figure 1A).

Figure 1.

Histone lactylation increases during caudal fin regeneration

(A) Western blot (WB) for Kla in full protein extract and isolated histone fraction, from uninjured zebrafish caudal fin samples. Kla protein, approximately from 11 to 17 kDa in size, is detected in the corresponding histone-associated region in the enriched histone fraction.

(B) Schematic representation of the caudal fin amputation assay, with the different time points of tissue collection.

(C–N) Representative transverse cryosections of caudal fins immunostained for H3 (magenta) and Kla (green) antibodies and counterstained with DAPI (gray), at different time points: 0 hpa (uninjured condition) (C–E); 6 hpa (F–H); 24 hpa (I–K); and 48 hpa (L–N). White dashed boxes delineate magnified panels in D′, E′, G′, H′, J′, K′, M′, and N′. Arrows indicate H3+Kla+ cells in the epidermis. White arrowheads indicate H3+Kla+ cells in the mesenchyme. Yellow arrowheads indicate the amputation plane. e, epidermis; b, bone; m, mesenchyme. Hpa, hours post-amputation. Scale bar represents 50 μm (C) and 10 μm (D′) in magnified panels.

(O–R) Graph shows the quantification of the average intensity of nuclear Kla, normalized to the average intensity of H3, at 0 hpa, 6 hpa, 24 hpa, and 48 hpa, in the whole caudal fin tissue (O), mesenchyme (P), epidermis (Q), and osteoblasts (R). Statistical analysis corresponds to the Mann-Whitney test with Mean ± SD (for 0, 6, 24, and 48 hpa, n = 11 blastemas from 4 fish). ns, not significative; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001. See also Figure S1.

Considering the direct effect of lactate in lysine lactylation,²¹ and our previous findings that lactate levels increase within the first 24 hpa,⁷ we examined whether histone lactylation is altered during the initial stages of caudal fin regeneration. For this purpose, we analyzed tissue lactylation levels, measuring Kla intensity by immunofluorescence at 6, 24, and 48 hpa, compared to uninjured fins (0 hpa) (Figure 1B). Since Kla detects all lactylated proteins, we used DAPI to restrict Kla signal quantification to the nucleus.³⁰ Additionally, we normalized Kla signal to the levels of histone 3 (H3).¹⁹ We observed a significant increase in nuclear lactylation intensity in the whole fin tissue at 6 and 24 hpa, compared to uninjured control (Figures 1C–1K and 1O), but not at 48 hpa, when lactylation levels returned to those observed in uninjured fins (Figures 1C–1E and 1L–1O). This result suggests that histone lactylation increases during the early stages of regeneration, when cell dedifferentiation and migration are necessary for blastema formation to occur, and then returns to basal levels by the second day of regeneration, when the blastema is fully formed. Interestingly, when analyzing the main fin tissues independently, lactylation levels increase in the mesenchymal compartment at 6 and 24 hpa (Figure 1P), compared to uninjured control, while remaining unchanged in the epidermis across all time points analyzed (Figure 1Q). Given our previous findings that metabolic reprogramming toward glycolysis is critical for the dedifferentiation of mature osteoblasts during regeneration,⁷ we also examined histone lactylation in this cell type at 24 hpa, when lactylation levels in the mesenchyme are more elevated. Using osx:mCherry reporter fish, which labels osteoblasts, we observed that lactylation levels also increase in these cells, compared to uninjured fins (0 hpa) (Figure 1R; Figure S1).

These findings suggest that histone lactylation is important during early regeneration, increasing during the initial 24 h of the caudal fin regeneration process.

Histone lactylation is driven by lactate

To determine whether lactate levels directly contribute to increased histone lactylation during the first hours post-amputation (Figure 1), we administered glucose and lactate to uninjured wild-type fish to assess whether lactylation levels could be induced, mimicking those seen in a regenerative setting. To minimize variability in lactate levels caused by normal feeding, fish were fasted for 24 h prior to injection (Figure 2A). Immunofluorescence analysis of Kla at 24 h post-injection revealed no significant difference in the average nuclear Kla intensity levels between glucose-injected and control PBS-injected fish (Figures 2B–2G′ and 2K). However, fish injected with lactate showed a significant increase in nuclear Kla levels, compared to control fish (Figures 2B–2D′ and 2H–2K), confirming that histone lactylation is responsive to changes in lactate levels.

Figure 2.

Histone lactylation is driven by lactate

(A) Schematic representation of the experimental design used to administer glucose and lactate. Fish were placed in a fasting condition 24 hbi and administered via IP injection with glucose and lactate. Fish are kept in fasting condition until 24 hpi, upon which the caudal fins were collected.

(B–J) Representative transverse cryosections of 24 hpi caudal fins immunostained for H3 (magenta) and Kla (green) antibodies and counterstained with DAPI (gray), in fish treated with: PBS (control condition) (B–D); glucose (E–G); and lactate (H-J). White dashed boxes delineate magnified panels in C′, D′, F′, G′, I′, J′. White arrowheads indicate H3+Kla+ cells. e, epidermis; b, bone; m, mesenchyme. hours before injections (hbi); hours post-injection (hpi). Scale bar represents 100 μm (B) and 20 μm (C′) in magnified panels.

(K) Quantification of the average signal intensity of nuclear Kla, normalized to the H3 signal. Statistical analysis corresponds to the Mann-Whitney test with Mean ± SD (for control PBS and lactate, n = 8 blastemas from 3 fish; for glucose, n = 9 blastemas from 3 fish). n.s., not significative; ∗p < 0.05.

These results provide evidence that exogenous lactate can stimulate lactylation. By administering lactate to uninjured zebrafish, which naturally have lower lactate levels in the fin compared to regenerating conditions,⁷ we induced an increase in Kla levels (Figures 2B–2E), similar to what is observed during regeneration (Figure 1O). This suggests that higher levels of histone lactylation during regeneration likely stem from elevated lactate levels associated with metabolic reprogramming. In contrast, exogenous glucose did not affect Kla levels (Figures 2B–2G′), as in the uninjured condition where no metabolic reprogramming is triggered, suggesting that glucose is mainly diverted toward oxidative phosphorylation rather than glycolysis, making it insufficient to increase lactate production.

Overall, these findings suggest that the early increase in histone lactylation during regeneration is likely driven by elevated lactate levels, consistent with findings reported in other biological systems.^21,31,32

Glycolysis and lactate generation inhibition impair histone lactylation

To further investigate the functional relevance of lactate in histone lactylation, we inhibited glycolysis and assessed its impact on histone lactylation levels following injury. We used a previously established glycolysis inhibition protocol⁷ (Figure 3A), administering the glucose analogue 2-Deoxy-D-glucose (2DG), which competes with glucose for the binding of the hexokinase catalytic domain, the first enzyme in the glycolysis pathway. We treated osx:mCherry reporter fish with 2DG from 0 hpa to 24 hpa, a period of significant increase in histone lactylation levels (Figure 1O). Immunofluorescence analysis for Kla revealed a major decrease in nuclear Kla intensity across the entire fin tissue adjacent to the amputation plane in 2DG-treated fins, compared to control fins (Figures 3B–3J). When analysing each major cell population individually, this reduction was observed in the epidermis, mesenchyme, and osteoblast populations (Figures 3K–3M).

Figure 3.

Glycolysis inhibition impairs histone lactylation

(A) Schematic representation of the experimental design used to inhibit glycolysis. Fish were administered, via IP injection, with vehicle (PBS) or glycolytic inhibitor, 2DG, every 12 h, from fin amputation (0 hpa) until 24 hpa.

(B–I) Representative transverse cryosection of 24 hpa osx:mCherry (yellow) caudal fins immunostained for Kla (green) and H3 (magenta) antibodies and counterstained with DAPI (gray), in fish treated with: PBS (control condition) (B–E) and 2DG (F–I). White dashed boxes delineate magnified panels in C′–E′, G′–I′. Arrows indicate H3+Kla+ cells in the epidermis. White arrowheads indicate H3+Kla+ cells in the mesenchyme. Orange arrowheads indicate H3+Kla+osx+ osteoblasts. Yellow arrowheads indicate the amputation plane. e, epidermis; b, bone; m, mesenchyme. hpa, hours post-amputation. Scale bar represents 50 μm (B) and 10 μm (C′) in magnified panels.

(J–M) Quantification of the average signal intensity of nuclear Kla, normalized to the H3 signal, in 2DG-treated fins compared to control in the whole tissue (J), mesenchyme (K), epidermis (L), and OBs (M). Statistical analysis corresponds to the Mann-Whitney test with mean ± SD (n = 11 blastemas from 4 fish). ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

To further validate these findings, we administered galloflavin, a known inhibitor of LDHA, to block lactate production.³³ Since blastema formation is a prerequisite for caudal fin regeneration and considering that we previously demonstrated that lactate production is essential for blastema growth,⁷ we first assessed the impact of galloflavin on blastema growth. Galloflavin was administered at two time points post-amputation, 0 hpa and 24 hpa, and the total fin regenerated area measured at 48 hpa (Figure 4A; Figures S2A–S2E). We observed a significant reduction in the caudal fin regenerated area in fish injected with 0.12 mg/g/dose (Figures 4B–4D; Figures S2B–S2F), compared to control fins, confirming the requirement of lactate formation for effective tissue regeneration. Subsequently, we treated osx:mCherry reporter fish with galloflavin and analyzed Kla immunofluorescence at 24 hpa, the time point in which we observe a significant increase in histone lactylation (Figure 1O). Consistent with the results of glycolysis inhibition, galloflavin-treated fins showed a decrease in nuclear Kla intensity throughout the entire fin tissue, compared to control fins (Figures 4E–4M). The analysis of each major cell population individually also revealed that nuclear Kla intensity was decreased in the epidermis, mesenchyme, and osteoblast populations (Figures 4N–4P). To verify that these changes specifically reflected loss of lactate-derived histone modification, we performed rescue experiments by co-injecting galloflavin with exogenous lactate (Figure S3A). Co-treatment restored nuclear Kla intensity to control (DMSO/PBS) levels in all tissues, and in epidermal and mesenchymal compartments (Figures S3B–S3P). Within osteoblasts, galloflavin induced a mild but nonsignificant decrease in Kla signal, likely reflecting biological variability (Figure S3Q). These findings demonstrate that exogenous lactate can rescue galloflavin-induced decrease in histone lactylation, confirming that lactate production downstream of glycolysis is both necessary and sufficient to sustain histone lactylation during early fin regeneration.

Figure 4.

Lactate inhibition impairs histone lactylation and redirects gene expression from proliferation to tissue remodeling

(A) Schematic representation of the experimental design used to inhibit lactate. Fish were administered, via IP injection, with vehicle (DMSO) or lactate inhibitor, Galloflavin, at fin amputation (0 hpa) and at 24 hpa.

(B and C) Representative images of 48 hpa whole fins treated with vehicle (DMSO) (B) or Galloflavin (C).

(D) Quantification of the total fin regenerated area at 48 hpa, after vehicle (DMSO) or Galloflavin injection. Statistical analysis corresponds to the Mann-Whitney test with mean ± SD (for control DMSO, n = 6 fish; for Galloflavin, n = 7 fish). ∗p < 0.05.

(E–L) Representative transverse cryosection of 24 hpa osx:mCherry (yellow) caudal fins immunostained for Kla (green) and H3 (magenta) antibodies and counterstained with DAPI (gray), in fish treated with: DMSO (control condition) (E–H) and Galloflavin (I–L). White dashed boxes delineate magnified panels in F′–H′, J′–L′. Arrows indicate H3+Kla+ cells in the epidermis. White arrowheads indicate H3+Kla+ cells in the mesenchyme. Orange arrowheads indicate H3+Kla+osx+ osteoblasts (orange). Yellow arrowheads indicate the amputation plane. e, epidermis; b, bone; m, mesenchyme. hpa, hours post-amputation. Scale bar represents 100 μm (E) and 20 μm (F′) in magnified panels.

(M–P) Quantification of the average signal intensity of nuclear Kla, normalized to the H3 signal, in Galloflavin-treated fins compared to control in the whole tissue (M), mesenchyme (N), epidermis (O), and OBs (P). Statistical analysis corresponds to the Mann-Whitney test with mean ± SD (for control DMSO, n = 9 blastemas from 3 fish; for Galloflavin, n = 10 blastemas from 4 fish). ∗p < 0.05; ∗∗p < 0.01.

(Q) Transcriptomic analysis of regenerating fins at 24 hpa following galloflavin treatment versus DMSO controls. Volcano plot highlighting differentially expressed genes (B-statistic >0), with downregulated genes (blue) and upregulated genes (orange). Vertical dotted lines indicate ±0.5 log₂ fold-change thresholds.

(R) Hallmark Gene Set Enrichment Analysis (GSEA). Bar plots display significantly enriched gene sets ordered by normalized enrichment score (NES, FDR £ 0.05), with negative enrichment in blue and positive enrichment in orange. Only the top 20 gene sets are shown for positive NES values. GSEA was run as described in detail in the STAR Methods section. See also Figures S2–S4, and Tables S1 and S2.

Finally, to exclude any effect of lactate inhibition on cell survival that could interfere with our observations, TUNEL assays at 24 hpa revealed no significant increase in apoptosis in epidermal and mesenchymal compartments following galloflavin treatment (Figures S2G–S2N). This indicates that lactate inhibition specifically impairs histone lactylation and blastema regeneration without substantially affecting cell survival.

Together, these results demonstrate that metabolic reprogramming toward glycolysis is crucial for maintaining histone lactylation during the initial stages of the regeneration process. Furthermore, the observed decrease in lactylation levels within osteoblasts suggests that this modification may regulate transcriptional programs underlying the transition from mature to less differentiated states during blastema formation.

Lactate-derived histone lactylation regulates proliferative and chromatin-remodeling programs during regeneration

To start identifying the cellular processes influenced by histone lactylation during regeneration, we performed RNA sequencing (RNA-seq) of regenerating caudal fins at 24 hpa, under control conditions (DMSO) and following lactate dehydrogenase inhibition with galloflavin. This time point was selected because our data show that lactylation levels peak within the first 24 h and it corresponds to active blastema formation. This differential gene expression analysis identified 69 genes as significantly differentially expressed (B statistic>0) upon lactate inhibition relative to DMSO controls (Figure 4Q; Table S1). Several genes involved in cell-cycle progression and chromatin regulation were downregulated under low lactylation, including cdca7a, a key regulator of the G1-S transition that forms a chromatin remodeling complex with HELLS, required for nucleosome organization and DNA methylation maintenance.^34,35 In cancer contexts, it has been shown to promote proliferation through the upregulation of ccna2.³⁶ Similarly, bcl11bb, a zinc-finger transcription factor involved in chromatin accessibility and lineage commitment, was reduced, further supporting a link between lactylation and proliferative chromatin states.³⁷ Conversely, genes associated with immune signaling (saa, mfap4.12), extracellular matrix remodeling (col9a3), and lipid/membrane dynamics (tgm5l, mboat1, plpp1a, acsl4b) were upregulated.

Gene Set Enrichment Analysis (GSEA) further supported these findings. Biological Process (GO-BP) categories downregulated upon lactate inhibition, and therefore positively associated with lactylation, included DNA replication, chromosome segregation, cell-cycle checkpoint signaling, methylation-related, and chromatin organization pathways (Figure S4A; Table S2). In contrast, upregulated GO-BP involved membrane remodeling, vesicle-mediated trafficking, protein turnover, and metabolic adaptation. Hallmark GSEA, which broadly recapitulate the GO-BP results, showed negative enrichment for E2F targets, MYC targets, G2/M checkpoint, and mitotic spindle gene sets, and a positive enrichment for TNFa/NFkB signaling, IL6-JAK/STAT3 signaling, Complement, and Epithelial-to-Mesenchymal Transition program (Figure 4R; Table S2). At the Molecular Function (GO-MF) level, lactylation inhibition reduced the activity of histone-modifying enzymes, lysine/RNA methyltransferases, DNA polymerases, helicases, and ATP-dependent chromatin remodelers. Conversely, collagen binding, protease-regulator activity, cytokine receptor binding, and lipid transporter activity were upregulated (Figure S4B; Table S2).

Together, these data indicate that histone lactylation sustains chromatin accessibility and proliferative gene expression during early regeneration, whereas its reduction promotes chromatin stabilization, inflammatory activation, and ECM remodeling. These findings support a model in which lactylation acts as a metabolic-epigenetic regulator, coordinating chromatin remodeling and transcriptional reprogramming during tissue regeneration.

Discussion

Zebrafish caudal fin regeneration is driven by an early metabolic shift, with osteoblasts prioritizing lactate-producing glycolysis. This metabolic adaptation contributes to the activation of a regenerative genetic program that governs mature osteoblast dedifferentiation and cell cycle re-entry, supporting blastema formation.⁷ This dynamic process reprograms specialized cells into a less differentiated state, enabling proliferation and differentiation during the different phases of fin regeneration.¹⁹ Understanding these epigenetic mechanisms is crucial for identifying key players and regulatory networks, contributing to a full understanding of the regenerative processes.

Histone lactylation is a recently identified modification implicated in mediating tumorigenesis and inflammation by regulating gene transcription.²⁶ In this study, we provide new evidence supporting the role of glycolysis as an epigenetic regulator of histone lactylation during caudal fin regeneration. Specifically, we observed an elevation of global lactylation levels within the initial 24-h post-amputation period, returning to baseline by 48 hpa. This increase is driven by lactate, with exogenous lactate directly influencing global histone lactylation levels in uninjured fins, confirming that lactate availability can modulate this modification. Furthermore, we demonstrate that glycolysis-derived lactate is crucial for histone lactylation, as inhibiting either glycolysis or lactate production led to an overall decrease in histone lactylation during the early stages of regeneration. These observations are consistent with our previous work showing increased ldha expression and lactate production within the first 24 hpa of caudal fin regeneration, as well as the requirement of lactate production for blastema formation.⁷

In line with these observations, our transcriptomic analysis following lactate dehydrogenase inhibition supports a role for histone lactylation in regulating proliferative and chromatin remodeling programs. Pathway-level analyses revealed that lactylation supports the transcriptional programs and enzymatic activities required for DNA replication and chromatin remodeling, thereby promoting genome accessibility and proliferative potential in the regenerating blastema. Epigenetic regulatory pathways, including those associated with methylation and chromatin organization, were particularly sensitive to lactate inhibition, indicating that lactylation is tightly coupled to the maintenance of a permissive chromatin state. These findings suggest that lactylation cooperates with other histone modifications to establish a transcriptionally permissive chromatin landscape. The upregulation of genes related to inflammatory signaling and ECM remodeling, when lactate dehydrogenase in inhibited, suggests that lactylation is involved in the downregulation of these processes. Both inflammation and ECM remodeling have a dynamic behavior during regeneration; they are activated at the initiation of regeneration to clear debris and prepare the tissue for blastema formation, but as the blastema matures, inflammation starts to be resolved and ECM remodeling is reduced.^38,39 Taken together, these results suggest that histone lactylation may have multiple roles in the epigenetic regulation of fin regeneration, being relevant not only to promote the proliferation of blastemal cells, but also to control the adequate levels of inflammation and ECM remodeling.

By integrating these molecular findings with our cell-type specific analyses, we propose that the chromatin-remodeling function of histone lactylation underlies the reprogramming of mesenchymal and osteoblast cells during blastema formation. Our findings show a significant increase in histone lactylation within both mesenchymal and osteoblast populations during the early stages of regeneration. This increase in lactylation in the mesenchymal compartment suggests a role in the reconfiguration of the epigenetic landscape of various cell types, including mature osteoblasts. This reconfiguration could be essential for promoting cellular plasticity, enabling osteoblasts to re-enter the cell cycle, proliferate, and subsequently differentiate into new osteoblasts, thus contributing to new bone formation.¹² Altogether, our data support the hypothesis that histone lactylation is a key regulator of mature osteoblast plasticity during the initial stages of regeneration, potentially driving essential processes such as cell dedifferentiation, migration, and blastema proliferation.

Overall, in this study, we present new evidence that glycolysis regulates histone lactylation during zebrafish caudal fin regeneration, highlighting the potential role of lactylation in coordinating proliferation, chromatin dynamics, and gene expression programs essential for bone regeneration. Future research should aim to further elucidate the pathways regulated by lactate-derived histone lactylation and assess how these influence regenerative outcomes.

Limitations of the study

This study establishes a link between glycolysis-derived lactate and histone lactylation during zebrafish fin regeneration and provides transcriptomic evidence linking lactylation to proliferative and chromatin-remodeling programs. However, while the RNAseq and GSEA identify pathways associated with changes in lactylation, they do not establish direct genomic targets or causal gene regulation by histone lactylation. Future studies at multiple time points, combining chromatin profiling approaches, such as CUT&Tag using antibodies that label lactylated histones, together with transcriptomic and functional analysis, will be required to determine whether lactylation marks regeneration-specific promoters or enhancers during different stages of regeneration. In addition, the enzymatic mechanism connecting lactate to histone lactylation is not explored in this work. Addressing these questions will be important to fully define the regulatory scope of lactate-derived histone lactylation in tissue regeneration.

Resource availability

Lead contact

Requests for further information and reagents should be directed to and will be fulfilled by the lead contact, António Jacinto (antonio.jacinto@unl.pt).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •Original WB image related to Figure 1 is available at https://doi.org/10.6084/m9.figshare.30920576.
• •RNA-seq data, Fastq files, and counts per million normalized count matrix, have been deposited to the NCBI Gene Expression Omnibus database with an accession number GEO: GSE311919.
• •The code generated for RNA-seq analysis is available at https://anonymous.4open.science/r/borbinha-et-al-332F/.
• •Any additional information required to reanalyze the data reported in this work article is available from the lead contact upon request.

Acknowledgments

We are grateful to Ana Teresa and Lara Carvalho for advice and for reading the article, Telmo Pereira for data analysis and Fior Lab from CF for providing laboratory space and materials. We thank Petra Pintado and Fábio Valério from the NMS Fish Facility, and Catarina Certal, Joana Monteiro, Inês Oliveira, Pedro Seco et al., from the CF Fish Platform for animal care; Ana Farinho from the NMS Histology facility, and Sérgio Casimiro, Inês Romano and Ana Quitéria from the CF Histopathology Platform for assistance in tissue processing and cryosectioning; Telmo Pereira from NMS Microscopy Facility, and Pedro Campinho and Ann Pezzarossa from the CF Advanced BioImaging and BioOptics Experimental Platform for technical assistance.

This work was supported by funding from Fundação para a Ciência e a Tecnologia in the context of a program contract to RL (4, 5, and 6 of article 23. of D.L. no. 57/2016 of August 29, as amended by Law no. 57/2017 of 19 July); PTDC/BIM-MED/0659/2014 in the context of a grant project; SFRH/BD/51990/2012 to A.B.; SFRH/BD/131929/2017 to J.B. This work was also funded by FCT, through complementary funding of EU- Horizon Europe (project 101060346). Zebrafish used as animal model were reproduced and maintained in the NMS Fish Facility and Champalimaud Fish Platform, with the support from Congento LISBOA-01-0145-FEDER-022170, co-financed by FCT (Portugal) and Lisboa2020, under the PORTUGAL2020 agreement (European Regional Development Fund).

Author contributions

Conceptualization, J.B., R.L., and A.S.B.; methodology, J.B., R.L., A.S.C., D.R., and A.S.B.; investigation, J.B. and R.L.; software, D.R.; writing – original draft, R.L.; writing – review and editing, R.L., J.B., A.S.C., R.M., and A.J.; funding acquisition, A.J.; resources, R.M. and A.J.; supervision, A.S.B., R.L., and A.J.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Pan anti-Kla	PTM-BIO	Cat# PTM-1401, RRID:AB_2868521
Anti-Histone H3	Santa Cruz Biotechnology	Cat# sc-517576, RRID:AB_2848194
Anti-mCherry	Thermo Fisher Scientific	Cat# M11217, RRID:AB_2536611
Anti-Histone H3	Milipore	Cat# 06-755, RRID:AB_11211742
Alexa Fluor 488 Goat anti-Rabbit IgG (H + L)	Thermo Fisher Scientific	Cat# A-11070, RRID:AB_2534114
Alexa Fluor 568 Goat anti-Mouse IgG (H + L)	Thermo Fisher Scientific	Cat# A-11031, RRID:AB_144696
Alexa Fluor 568 Goat anti-Rat IgG (H + L)	Thermo Fisher Scientific	Cat# A-11077, RRID:AB_2534121
Alexa Fluor 647 Donkey anti-Mouse IgG (H + L)	Jackson ImmunoResearch	Cat# 715-605-151, RRID:AB_2340863
Horseradish peroxidase-conjugated Goat anti-Rabbit IgG (H + L)	Bio-Rad	Cat# 170-6515, RRID:AB_11125142
Horseradish peroxidase-conjugated Goat anti-Mouse IgG (H + L)	Bio-Rad	Cat# 170-6516, RRID:AB_11125547
Chemicals, peptides, and recombinant proteins
2-Deoxy-D-glucose	Sigma-Aldrich	Cat# D8375
Galloflavin	Cayman	Cat# 14846; CAS: 568-80-9
Glucose	Sigma-Aldrich	Cat#G8270
Sodium L-lactate	Sigma-Aldrich	Cat#L7022
Trizol reagent	Invitrogen	Cat# 15596026
Critical commercial assays
In situ Cell Death Detection Kit, Fluorescein	Roche	Cat# 11684795910
Quick-RNA™ Microprep Kit	Zymo Research	Cat# R1050
KAPA mRNA HyperPrep kit	Roche	Cat# 08098123702
Pierce BCA protein assay kit	Thermo Fisher Scientific	Cat# 23225
Deposited data
RNA-Seq data	Database	GEO: GSE311919
Code	Anonymous GitHub	https://anonymous.4open.science/r/borbinha-et-al-332F/
Experimental models: Organisms/strains
Zebrafish: AB strain	ZIRC	ZDB-GENO-960809-7
Zebrafish: Tg(osterix:mCherry-NTRo)pd46	Singh et al.40	ZFIN: ZDB-ALT-120503-4
Software and algorithms
ZEN Microscopy Software (version 3.12)	ZEISS	RRID:SCR_013672
Fiji	Schindelin et al.50	RRID:SCR_002285
IMARIS (version 10.2.0)	Oxford Instruments	RRID:SCR_007370
GraphPad Prism 8	GraphPad Software	RRID:SCR_002798
FastQC (0.11.9)	Babraham Institute	RRID:SCR_014583
STAR (2.7.10a)	Dobin et al.46	RRID:SCR_004463
conda (25.1.1)	Conda Organization	RRID:SCR_018317
Python (3.11.4)	Python Software Foundation	RRID:SCR_008394
R (4.2.3)	R Foundation	RRID:SCR_001905
edgeR (3.40.2)	Robinson et al.47	RRID:SCR_012802
limma (3.54.2)	Ritchie et al.48	RRID:SCR_010943
clusterProfiler (4.6.2)	R package	RRID:SCR_016884
msigdbr (25.1.1)	R package	RRID:SCR_022870
matplotlib (3.9.1)	Python package	RRID:SCR_008624
NumPy (1.23.5)	Python package	RRID:SCR_008633
pandas (2.3.1)	Python package	RRID:SCR_018214

Experimental model and study participant details

Zebrafish strains and maintenance

All handling and experiments involving animals were approved by the ORBEA-NMS, Animal User and Ethical Committees at Centro de Estudos de Doenças Crónicas (CEDOC), ORBEA-Champalimaud and accredited by the Direcção Geral de Alimentação e Veterinária (DGAV), under protocol 0421/000/000/2022 (2022-03-28, ref. 004980), according to the directives from the EU (Directive 2010/63/UE) and National legislation (Directive 113/2013) for animal experimentation and welfare. Wild-type (WT) AB (obtained from ZIRC in 2019 and refreshed in 2024) and the transgenic zebrafish line Tg(osterix:mCherry-NTRo)pd46 (referred as osx:mCherry), kindly provided by Kenneth Poss,⁴⁰ were maintained in a circulating system with a 14 h/day and 10 h/night cycle at 28°C.⁴¹ Fish housing and husbandry were performed according to standard protocols⁴² with feeding adjustments⁴³: L-type rotifers were provided as live feed until 60 days post-fertilization (dpf), followed by Artemia sp. nauplii thereafter. Dry feed consisted of a 1:1 mixture of GemmaMicro (Skretting) and Zebrafeed (Sparos) up to sexual maturation (60–90 dpf), followed by Zebrafeed only. The AB colony was maintained using a multiple paired-cross scheme to raise at least two stocks per generation.⁴⁴ Equal numbers of embryos from 15 to 30 clutches of distinct pairs were pooled to establish each new stock, and the procedure was repeated over two consecutive weeks. Subsequent generations were produced by crossing females and males from different stocks. All experiments were performed in 3–12 months-old fish, in equal amounts of male and females, and transgenics used as heterozygotes.

Method details

Caudal fin amputation

Caudal fin amputations were performed in fish subjected to analgesia with 4 mg/L Lidocaine 2% Braun (Braun) and anesthetized with 160 mg/mL MS-222 (Sigma, E10521), using a sterile scalpel. Amputations were made 1 or 2 segments below the first bone-segment bifurcation, removing approximately one-half of the fin, as previously described.⁴⁵ Fish were left to regenerate in an incubator at 33°C with water from the circulating system and fins collected at predetermined time-points post-amputation. Regenerated fins were collected from anesthetized fish, and either processed for protein extraction, RNA extraction or cryosectioning.

Protein extraction

For protein extraction, 10 caudal fins from uninjured conditions were collected. Caudal fins were lysed with Tris buffer (100 mM Tris-HCl pH 9.5, 1% SDS), homogenised through sonication using Sonifer SFX 150–10 cycles, 10 s each at 30% power, and centrifuged at 16,000g for 10 min at 4 °C. Total protein concentration from supernatants was determined using Pierce BCA protein assay kit (Thermo Fisher Scientific). Samples were then used for Western blot (WB).

Histone enrichment

For caudal fin samples histone enrichment protocol, to identify the profile of lactylated histones through WB, 4 pools of 60 uninjured caudal fins were collected into PBS at 4 °C. Lysis buffer (10 mM Tris-HCl pH = 6.5; 50 mM Sodium Bisulphate; 0.1% Triton X-100, 10 mM MgCl; 8.6% sucrose – all from Sigma-Aldrich) was added and cells homogenised by applying 20 strokes in a tight fitting Dounce homogeniser. Released nuclei were pelleted by centrifugation at 2500g, for 10 min, at 4 °C. Pellet was then centrifuged in lysis buffer, washed in washing buffer (10 mM Tris-HCl; 13 mM EDTA pH = 7.4 – all from Sigma Aldrich), and centrifuged again at 2500g for 10 min, each. Remaining pellet was resuspended in cold H2SO4 (Sigma-Aldrich, 0.4 M) and incubated for 1 h at 4 °C. For final nuclear precipitation, mixture was centrifuged, and 1 mL of cold acetone added to the supernatant, followed by an overnight (ON) incubation at −20 °C. Samples were then centrifuged, pellet air-dried and resuspended in distilled water. Nuclear protein concentration was determined using Pierce BCA protein assay kit (Thermo Fisher Scientific).

Western blot

For each sample, 20 μg of protein was denatured in Laemmli buffer and heated at 95 °C for 5 min. Total protein was resolved on a 15% SDS gel (for 10 mL: 2.3 mL MiliQ water; 5 mL 30% Acrylamide/Bis Solution (37:5:1); 2.5 mL of separating buffer (90 g Tris, 2 g SDS in 500 mL water, pH = 8.8); 0.1 mL of Ammonia Persulfate 10% and 10 μL of TEMED) and transferred to a nitrocellulose membrane at 100 V during 75 min. Membranes were blocked with Tris-buffered saline (TBS) with 0.1% Tween 20 (TBST) containing 5% non-fat dry milk, followed by incubation with primary antibodies (for antibody details see Table S1), ON at 4 °C. Membranes were washed in TBST and incubated with HRP-conjugated secondary antibodies for 1 h at RT. For imaging, membranes were incubated with ECL TM Prime Western Blotting Detection Reagent (Cytiva) and images acquired using ChemiDoc Touch System (BioRad) within the linear range.

Pharmacological and chemical treatments

For pharmacological treatments fish were subjected to intraperitoneal (IP) injections at designated time-points with either 2DG (0.5 mg/g diluted in 1x Phosphate Buffered Saline (PBS)); Galloflavin (0.12 mg/g diluted in DMSO); Glucose (3 mg/g); Sodium Lactate (3 mg/g) or the corresponding vehicle (Control). For rescue experiments, fish received combined injections of Galloflavin (0.12 mg/g) and Sodium Lactate (3 mg/g), or matched vehicle controls. IP injections were performed with an insulin syringe U-100 G 0.3 mL and a 30 G needle (BD Micro-fine) inserted close to the pelvic girdle. For all experiments water was replaced daily and fish left to regenerate until the desired time-point.

RNA isolation and RNA-Seq

For RNA-Seq analysis, caudal fin composed of the regenerated tissue and one bony-ray segment proximal to the amputation plane were collected. Pools from 10 caudal fins were used per biological replicate, and three biological replicates were used per condition. Briefly, samples were homogenized in Trizol reagent for cell disruption and RNA extraction. Chloroform was added, and the homogenate allowed to separate into a clear upper aqueous layer. RNA was precipitated and purified from the aqueous phase by adding an equal amount of 100% ethanol and loading the mixture into a Zymo-Spin IC Column (ZYMO Research). The remaining procedure was done following the Quick-RNA Microprep Kit manufacturer’s protocol. RNA integrity number (RIN) and concentration used for next-generation sequencing was checked on a TapeStation 4200 and Nanodrop ND-1000, respectfully. The library construction of cDNA molecules was carried out using a Stranded mRNA Library Preparation Kit (Roche/KAPA mRNA HyperPrep kit). The generated DNA fragments (DNA library) were sequenced in the lllumina Novaseq 6000 platform, using 150 bp paired-end sequencing reads and the quality control of raw data generated was performed. All library preparation, quality control and next-generation sequencing was performed by Stab Vida (Caparica, Portugal).

Pre-processing of RNA-seq

Provided FASTQ files with RNA-seq data were checked for the overall quality of sequencing reads using the tool fastqc (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ accessed on 6 November 2025), all samples had good quality. STAR aligner⁴⁶ was used with default settings and ‘--quantMode TranscriptomeSAM GeneCounts’ for alignment and quantification of gene expression using the zebrafish genome assembly GRCz11. Raw gene counts were extracted from ‘∗_ReadsPerGene.out.tab’ output per sample. The raw RNA-seq data were deposited in the Sequence Read Archive. See data and code availability section to retrieve the associated Gene Expression Omnibus (GEO) accession.

Differential gene expression analysis

To prepare count data for differential gene expression (DGE) analysis, the count matrix was converted to a DGEList object and normalization factors to scale raw library sizes were obtained using the function calcNormfactors from package edgeR.⁴⁷ Next genes with less than 0.3 counts per million (CPM) were filtered out of the normalized count matrix. The voom method⁴⁸ was then applied to count data to obtain gene expression estimates in log2-counts per million (logCPM). DGE analysis was performed with linear models using the limma R package.⁴⁸ The expression each gene was fitted with a design matrix parameterized with DMSO and Galloflavin treatments (group) and technical batch correction between the 3 biological replicates (batch), coded as model.matrix(∼0 + group + batch). Then sample comparison was performed using voom transformed values, followed by linear modeling, moderated t-statistic and global empirical Bayes statistics calculated for each gene, allowing for the identification of genes significantly differentially expressed as those with an associated positive log-odds ratio (B-statistic >0) of the gene is differentially expressed and with the magnitude of difference in expression measured in log2 fold-change between the two conditions.

Gene set enrichment analysis

The clusterProfiler R package⁴⁹ was used to perform Gene Set Enrichment Analysis (GSEA) on the t-statistics of DGE genes. Gene symbols were checked for duplicates, and when found, the symbol with the highest absolute t-statistic value was kept. The Gene Ontology (GO) Biological Process and Molecular Function, and Hallmarks pathway genes and terms for Danio rerio were retrieved with the package msigdbr. The GSEA algorithm was run with the default settings, except for the following: maxGSSize = 1500, pvalueCutoff = Inf, eps = 0, seed = TRUE. We considered terms with adjusted p-value <0.05 (Benjamini-Hochberg) to be enriched in our analysis. Normalized enrichment score (NES) bar plots were generated with matplotlib for significant terms. For clarity, only up to 20 highest NES in each direction were plotted together in the same graph, i.e., up to 20 positive/enriched and negative/depleted terms in Galloflavin versus DMSO treatment.

Immunofluorescence and image acquisition

Tissue processing for cryosections was performed as preciously described.⁷ Shortly, fins were collected, fixed overnight (ON) in 4% paraformaldehyde (in 1x PBS) and stored in 100% methanol (MeOH) at −20 °C, until required. They were then gradually rehydrated in a series of MeOH/1x PBS (75%, 50% and 25%) and incubated ON in 30% sucrose (Sigma-Aldrich, in 1x PBS). Subsequently, fins were embedded in 7.5% gelatin (Sigma-Aldrich)/15% sucrose in 1x PBS and subsequently frozen in isopentane at −70 °C and stored at −80 °C. Longitudinal caudal fins sections were obtained at 12 μm using a Microm cryostat (Cryostat Leica CM3050 S) and slides stored at −20 °C. For immunofluorescence on cryosections, slides were thawed for 15 min at room temperature (RT), washed twice in 1x PBS at 37 °C for 10 min and subjected to an antigen retrieval step consisting in a 15 min incubation at 95 °C with heated sodium citrate buffer (10 mM Tri-sodium citrate with 0.05% Tween 20, pH6). Slides were then incubated in 0.1 M glycine (Sigma-Aldrich, in 1x PBS) for 10 min, permeabilized with acetone for 7 min at −20 °C and incubated for 20 min in 0.2% PBST (1x PBS with 0.2% Triton X-100). For TUNEL labeling assay, slides were permeabilized in a sodium citrate solution (0.1% sodium citrate and 0.1% Triton X-100 in 1x PBS) and labeled according to the manufacturer’s protocol. Afterward, they were incubated in a blocking solution (10% non-fat dry milk in PBST) for at least 2 h at RT. Slides were incubated with primary antibodies diluted in blocking solution, ON at 4 °C (for primary antibody details see Table S1). Next day slides were washed with PBST 6 times, 10 min each, and incubated with secondary antibodies diluted in blocking solution, for 2 h at RT, in the dark. Slides were then washed in PBST three times, 10 min each, and counterstained with 4′,6-diamidino-2-phenylindole (DAPI; 0.001 mg/mL in 1x PBS, Sigma-Aldrich) for 5 min in the dark, for nuclei staining. Slides were washed three times with 1x PBS, 10 min each, mounted with fluorescent Mounting Medium (DAKO) and stored at 4°C protected from light until image acquisition.

Image acquisition and processing

For regenerated area measurements, images of live anesthetised WT adult caudal fins were acquired using a Zeiss Discovery.V8 stereoscope equipped with a Zeiss AxioCam ICc1 camera using an objective Achromat S 1.0x FWD 63 mm (at 1x zoom) controlled by Zen 2 PRO blue software. Images were acquired using transmitted light filters and assembled using the Fiji software.⁵⁰

Immunolabeled cryosections were imaged on a Zeiss confocal microscope LSM 980 controlled by ZEN 3.3. Image acquisition was done using an LD LCI Plan-Apochromat 40x/1.2 water objective with 0.6× zoom, a step size of 1 μm, and 488 (2.80%, gain 727 V, offset 1), 561 (1.80%, gain 780 V, offset 0), 639 (1.80%, gain 777 V, offset 1) and 385/DAPI (0.5%, 670 V, offset 0) nm excitation wavelengths. Sequential images were acquired to capture the first segment below the amputation plane and the entire regenerated region. For image analysis and processing, composite maximum intensity z stack projections were made using the Fiji software.⁵⁰ When required, concatenation of several images along the proximal-distal axis of the same longitudinal section was performed using the Fiji plugin 3D Pairwise Stitching.⁵⁰

For all cryosections of manipulated fish and corresponding control, images were acquired employing identical settings (magnification, contrast, gain and exposure time) and in identical/comparable regions. All Images were then processed using the Inkscape (open source vector graphics editor).

Quantification and statistical analysis

Total regenerated fin area

Measurements of total regenerated fin area were obtained by delineating distal end of the regenerated area using the Area tool in Fiji. These were normalized to the total caudal fin width, to avoid discrepancies between different fish sizes, thus resulting in one measurement value per animal. Statistical analysis corresponds to Mann-Whitney test with Mean ± SD. ∗p < 0.05. In Figure 4, for control DMSO n = 6 fish; for Galloflavin n = 7 fish. In Figure S2 for control DMSO n = 14 fish; for 0.04 mg/g Galloflavin n = 8 fish; for 0.08 mg/g Galloflavin n = 10 fish; for 0.12 mg/g Galloflavin n = 10 fish.

TUNEL assay quantification

Total number of TUNEL positive cells was assessed by quantifying the number of labeled cells within each fin compartment (mesenchyme and epidermis), in relation to the corresponding compartment area (per 100 μm²), determined using the Area tool on Fiji. Statistical analysis corresponds to Mann-Whitney test with Mean ± SD. ns, not significant. In Figure S2 for control DMSO, n = 7 blastemas from 2 fish; for Galloflavin, n = 12 blastemas from 4 fish.

Immunostaining average intensity measurement

To determine average intensity of KLac and H3 in complete longitudinal cryosections of individual regenerating bony-rays, we used IMARIS software (version 10.2.0). We used the surface tool for DAPI channel, with a nucleus surface diameter detail of 0.5 μm. This gives us the medium intensity values of all channels colocalizing with the nucleus, including KLac and H3. To independently address nuclear KLac and H3 average intensity values of the epidermis and mesenchyme, we used the surface tool, with manual drawing option to delineate the desired area (either the epidermis or mesenchyme), and then used the mask tool to create a duplicated mask of selected area. We then used the surface tool for DAPI channel on the duplicated mask of selected area, with a nucleus surface diameter detail of 0.5 μm. This gives us the medium intensity values of all channels colocalizing with the nucleus, including KLac and H3, in the duplicated selected area. To address nuclear KLac and H3 average intensity values in osteoblasts, we used surface tool for mCherry channel, with a surface detail of 0.5 μm and then used the mask tool to create a duplicated mask of selected osx:mCherry area. We then used the surface tool for DAPI channel on the duplicated mask of osx:mCherry area, with a nucleus surface detail of 0.5 μm. This gives us the medium intensity values of all channels colocalizing with the nucleus, including KLac and H3, in the duplicated osx:mCherry selected area. For each experiment the average intensity of KLac was normalized for the average intensity of H3. Statistical analysis corresponds to Mann-Whitney test with Mean ± SD. ns, not significative; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. In Figure 1, for 0, 6, 24 and 48 hpa, n = 11 blastemas from 4 fish. In Figure 2, for control PBS and lactate, n = 8 blastemas from 3 fish; for glucose, n = 9 blastemas from 3 fish. In Figure 3, for control PBS, n = 11 blastemas from 4 fish; for 2 DG, n = 11 blastemas from 4 fish. In Figure 4, for control DMSO, n = 9 blastemas from 3 fish; for Galloflavin, n = 10 blastemas from 4 fish. In Figure S1, for uninjured, n = 6 blastemas from 3 fish; for 24 hpa, n = 11 blastemas from 5 fish. In Figure S3, for control DMSO/PBS, n = 10 blastemas from 4 fish; for Galloflavin, n = 12 blastemas from 5 fish; for Galloflavin/Lactate, n = 13 blastemas from 5 fish.

RNA-seq analysis

All statistical methods and tests used for RNA-seq analysis are explained in the relevant sections.

Published: January 24, 2026