Abstract

   Aberrant metabolic enzymes drive glucose metabolism reprogramming,
   which plays a crucial role in tumor malignancy and metastasis. Protein
   acetylation is one of the key regulatory mechanisms of metabolic enzyme
   function, yet its precise role requires further clarification. In the
   present study, we reported that the deacetylation and low expression of
   transaldolase 1 (TALDO1) mediated by HDAC6 weakened the inhibitory
   effect of TALDO1 on tumor proliferation and metastasis in
   nasopharyngeal carcinoma (NPC). Mechanistically, highly expressed HDAC6
   induced lysine 7 (K7) deacetylation of TALDO1, which could inhibit
   SMURF1-mediated K63-linked ubiquitination, thus reducing the protein
   stability of TALDO1. Notably, TALDO1 deacetylation inhibited its
   nuclear translocation and interaction with BRCA1, thereby reducing the
   inhibition of c-Myc transcriptional activation, promoting the
   expression of HK2/LDHA/PDK1, and further enhancing glycolysis
   independent of TALDO1 enzyme activity. This research elucidated the
   regulatory mechanism of TALDO1 from the perspective of acetylation
   modification, clarified the moonlighting functions of TALDO1 in
   metabolic reprogramming, and provided novel biomarkers and intervention
   strategies, such as HDAC inhibitors, for the clinical treatment of NPC.

   Subject terms: Acetylation, Cancer metabolism

Introduction

   Nasopharyngeal carcinoma (NPC) is a common malignant tumor of the head
   and neck in Southeast Asia and South China. With the innovation and
   development of radiation and combined therapy strategies, local NPC
   lesions have been effectively controlled [[38]1, [39]2]. However,
   recurrence and distal metastases remain the main cause of treatment
   failure and death for NPC patients [[40]3]. Therefore, uncovering the
   molecular mechanisms of NPC malignant progression and metastasis is
   necessary to guide to develop more effective treatment strategies.

   Abnormal metabolism is a hallmark of tumors, with various pathways of
   glucose metabolism such as glycolysis, oxidative phosphorylation, and
   the pentose phosphate pathway (PPP) being reprogrammed to provide
   energy and biomolecules, maintain redox stability, and promote tumor
   proliferation and metastasis [[41]4–[42]6]. Studies have shown that
   aberrant expression and activity of metabolic enzymes within tumors are
   key drivers of metabolic reprogramming [[43]5]. For example, the
   elevated HIF-1α upregulated the transcription and expression of
   pyruvate dehydrogenase (PDK1), promoted the conversion of pyruvate to
   lactate, thereby enhancing the glycolytic phenotype of metastatic cells
   and subsequently facilitating liver metastasis of breast cancer cells
   [[44]7]. Glutamine deficiency promoted the palmitoyl transferase
   ZDHHC18 to mediate the palmitoylation of the tricarboxylic acid (TCA)
   cycle enzyme malate dehydrogenase 2 (MDH2), which enhanced the enzyme
   activity and mitochondrial respiration, and then promoted ovarian
   cancer cell proliferation [[45]8]. O-GlcNAcylation of the PPP
   rate-limiting enzyme glucose-6-phosphate dehydrogenase (G6PD) enhanced
   its enzyme activity, further promoted synthesis of nucleotide and lipid
   precursors and reduced antioxidant defense equivalents, thus promoting
   the cell proliferation and survival of lung cancer [[46]9]. However,
   the mechanisms underlying metabolic enzymes regulation during cancer
   malignant progression and metastasis are far from being clarified.

   Protein acetylation is a post-translational modification that is
   reversibly regulated by acetyltransferases and deacetylases [[47]10].
   Acetylation modification is involved in gene transcription regulation
   through modulating histone acetylation status [[48]11]. Notably,
   non-histone protein acetylation affects protein functions through
   diverse mechanisms, including the regulation of protein stability,
   enzymatic activity, subcellular localization, and protein-protein
   interactions [[49]12]. For example, lysine 5 (K5) acetylation
   negatively regulated lactate dehydrogenase A (LDHA) enzyme activity and
   inhibited the proliferation and migration of pancreatic cancer
   [[50]13]. PCAF-mediated acetylation of ATP-citrate lyase (ACLY)
   increased its stability by preventing ubiquitination degradation, and
   then enhanced de novo lipid synthesis, thereby promoting cell
   proliferation and tumor growth [[51]14]. HDAC4-mediated deacetylation
   of glutaminase (GAC) at K311 promoted the interaction between TRIM21
   and GAC, thereby facilitating K63-linked ubiquitination of GAC, which
   suppressed its enzymatic activity, consequently promoting the
   progression of lung cancer [[52]15]. SIRT6-mediated deacetylation of
   pyruvate kinase M2 (PKM2) inhibited its nuclear translocation and made
   it lose the transcriptional coactivator function of oncogenes, thus
   inhibiting the proliferation and metastasis of liver cancer [[53]16].
   These studies indicate that acetylation modification is involved in the
   regulation of metabolic enzymes to impact their roles in tumor
   progression and metastasis [[54]17].

   Previously, we used tandem mass tags (TMT)-based quantitative
   acetylation proteomics to analyze differentially modified proteins in
   NPC cells [[55]18]. In this study, low acetylated and expressed
   transaldolase 1 (TALDO1) was further clarified, and revealed that the
   K7 deacetylation of TALDO1 could suppress its protein stability and
   nuclear localization. The results also showed that the moonlighting
   function of TALDO1 induced by K7 deacetylation promoted glycolysis,
   thereby enhancing the proliferation and metastasis of NPC.

Materials and methods

Clinical samples

   Paraffin-embedded NPC tissue microarrays (containing 126 cases of NPC
   and 14 cases of corresponding adjacent tissues) were purchased from
   Superboitek (NPC 1401, Shanghai, China), and the follow-up data and
   clinical details are shown in Table [56]S1. A total of 35 primary NPC
   formalin-fixed paraffin-embedded tumor samples were retrospectively
   collected from Xiangya Hospital, Central South University (Changsha,
   China) [[57]18]. The 82 serum samples of NPC patients were obtained
   from Xiangya Hospital, Central South University, and the
   clinicopathological information for these patients is summarized in
   Table [58]S2.

Plasmids and transfection

   Blank vectors such as pcDNA3.1 (VT1001) and PGEX-4T-1 (VT1253) were
   purchased from Youbio (Changsha, China). pcDNA3.1-Flag (#20011),
   pCDH-CMV (#167463), pcDNA3.1-His (#52534), and pLKO.1 (#8453) were
   obtained from Addgene. pLVX-shRNA1 (632177) was purchased from Clontech
   (Mountain View, CA). The pCMVβ-p300-Myc (#30489), pcDNA3β-Flag-CBP-HA
   (#32908), pRK5-HA-Ubiquitin-WT (#17608), pRK5-HA-Ubiquitin-K48
   (#17604), pRK5-HA-Ubiquitin-K63 (#17606), and pBABEpuro-HA-BRCA1
   (#14999) plasmids were obtained from Addgene. pcDNA3.1-GCN5-His
   (40507), pcDNA3.1-PCAF-His (40506), and pcDNA3.1-TIP60-His (40505)
   plasmids were purchased from Youbio. The luciferase reporter vector
   pGL3-Basic (E1751) and Renilla luciferase reporter gene control vector
   pRL-TK (E2241) were obtained from Promega (Madison, WI).
   pcDNA3-HDAC6-DN-Flag plasmid was provided by Professor Wei Liu (School
   of Medicine, Zhejiang University) and pCDH-CMV-c-Myc was provided by
   Professor Hongjuan Cui (Medical Research Institute, Southwest
   University). Regarding plasmid construction, a full-length TALDO1
   coding sequence(CDS) was respectively cloned into pcDNA3.1,
   pcDNA3.1-Flag and pCDH-CMV vectors to generate pcDNA3.1-TALDO1,
   pcDNA3.1-TALDO1-Flag and pCDH-CMV-TALDO1, pcDNA3.1-TALDO1 and
   pcDNA3.1-TALDO1-Flag were used for transient transfections, while
   pcDH-CMV-TALDO1 was used for stable cell line construction. The CDS of
   HDAC4, HDAC5, HDAC6, SMURF1, and SYVN1 were cloned into pcDNA3.1-His to
   generate pcDNA3.1-HDAC4-His, pcDNA3.1-HDAC5-His, pcDNA3.1-HDAC6-His,
   pcDNA3.1-SMURF1-His, and pcDNA3.1-SYVN1-His plasmids. pcDNA3.1-His
   vector was digested with ApaI (1604, Takara, Tokyo, Japan), and
   HDAC6-DN coding sequence from HDAC6-DN-Flag plasmid was inserted into
   the digested vector to obtain pcDNA3.1-HDAC6-DN-His plasmid. The
   TALDO1-WT/K7R/K7Q sequences were cloned into the PGEX-4T-1 vector to
   create GST-TALDO1-WT/K7R/K7Q plasmids. The HK2, PDK1, and LDHA promoter
   sequences were gained from PCR amplification and cloned into the
   pGL3-basic vector to generate pGL3-HK2/PDK1/LDHA plasmids. The
   construction of expression plasmids using the Recombination Kit
   (C112-02, Vazyme, Nanjing, China) and the primers are listed in Table
   [59]S3-[60]1. The pcDNA3.1-TALDO1(K7R)-Flag and
   pcDNA3.1-TALDO1(K7Q)-Flag mutants were generated from
   pcDNA3.1-TALDO1-Flag. pCMVβ-p300.DY-myc mutant was generated from
   pCMVβ-p300-Myc. K48R and K63R mutants were generated using
   pRK5-HA-Ubiquitin-WT. Point mutations were introduced using a point
   mutation kit (C215, Vazyme) and primers for the construction of mutated
   plasmids are listed in Table [61]S3-[62]2. The shRNAs targeting TALDO1
   and HDAC6 were cloned into pLVX-shRNA1, while shRNAs targeting BRCA1
   and c-Myc were cloned into pLKO.1. The primers for the construction of
   shRNA plasmids are listed in Table [63]S3-[64]3, and siRNA sequences
   for HDAC4 and HDAC5 are listed in Table [65]S3-[66]4.

   Lipofectamine 2000 (12566014, Invitrogen, Carlsbad, CA) was used for
   all plasmid transfections. Lipofectamine 3000 (L3000075, Thermo Fisher
   Scientific, Waltham, MA) was used for all siRNA transfections.
   Transfection experiments were performed according to the manufacturer’s
   instructions.

Cell culture

   All NPC cell lines (HK1, SUNE1, HONE1, 5-8F, 6-10B, C666-1), normal
   human nasopharyngeal epithelial cell lines (NP69, NP460), and 293T
   cells were sourced from the Cell Center of Central South University.
   NP69 cells were cultured in keratinocyte serum-free medium (17005042,
   Gibco, Grand Island, NY). NP460 cells were cultured in EpiLife Medium
   (MEPI500CA, Gibco). HONE1, 5-8F, 6-10B, and C666-1 cells were
   maintained in RPMI 1640 (C11875500BT, Gibco) with 10% bovine calf serum
   (BCS) (B7446, Sigma-Aldrich, St. Louis, MO). SUNE1 and HK1 cells were
   cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS)
   (F8687, Sigma-Aldrich). 293T cells were cultured in DMEM (C11995500BT,
   Gibco) with 10% BCS. All cells were incubated at 37 °C with 5% CO[2].

   The 293T cells were co-transfected with shRNA or overexpression
   plasmids and lentivirus packaging plasmids (gag/pol, #14887, pRSV-Rev,
   #12253, pCMV-VSV-G, #8454, Addgene) using Lipofectamine 2000 to produce
   recombinant lentivirus. After 48 h, the viral supernatant was collected
   and used to infect NPC cells, which were selected with puromycin
   (ST551, Beyotime, Shanghai, China).

Antibodies and reagents

   Primary antibodies used were TALDO1 (12376-1-AP), Flag-tag
   (80010-1-RR), HDAC4 (17449-1-AP), HDAC5 (16166-1-AP), HDAC6
   (12834-1-AP), His-tag (66005-1-Ig), HA-tag (66006-2-Ig), Myc-tag
   (16286-1-AP), p300 (20695-1-AP), BRCA1 (22362-1-AP), and c-Myc
   (10828-1-AP) from Proteintech (Chicago, IL); SMURF1 (sc-100616) and
   SYVN1 (sc-293484) from Santa Cruz Biotechnology (Santa Cruz, CA); Actin
   (AC026) from Abclonal (Wuhan, China). K63-Ub (#5621), K48-Ub (#8081),
   PDK1 (#3062), LDHA (#3582), PKM2 (#4053), and Histone H3 (#4499) from
   Cell Signaling Technology (Danvers, MA); HK2 (ab209847), and Acetyl
   Lysine (ab21623) from Abcam (Cambridge, MA). LBH589 (S1030), PXD101
   (S1085), 3-MA (S2767), and CAY10603 (S7596) from Selleck Chemicals
   (Houston, TX); LMK-235 (HY-18998), MS-275 (HY-12163), MG132 (HY-13259),
   CHX (HY-12320), CHQ (HY-17589A), and 2-DG (HY-13966) from
   MedChemExpress (Monmouth Junction, NJ).

Western blot analysis and immunoprecipitation (IP)

   Cell samples were lysed in IP buffer (P0013, Beyotime) with protease
   inhibitor cocktail (4693116001, Roche, Basel, Switzerland). Protein
   concentration was measured using BCA assay kit (AR0197, Boster, Wuhan,
   China). 50 μg of cell lysate was separated by SDS-PAGE and transferred
   to a PVDF membrane (IPVH00010, Merck Millipore, Billerica, MA). After
   blocking with 5% milk for 1 h, the membrane was incubated with primary
   antibody at 4 °C overnight. Following TBST washes, the membrane was
   incubated with peroxidase-linked secondary antibody for 1 h and then
   detected with luminol substrate (32209, Thermo Fisher Scientific).
   Proteins were visualized using the Bio-Rad ChemiDoc XRS system
   (Hercules, CA). As for IP, cell lysates were prepared using IP buffer
   with protease inhibitor cocktail. 30 μL of magnetic beads ([67]B23201,
   Selleck Chemicals) were added to 500 μL of lysate and incubated for 2 h
   at 4 °C with rotation. After bead removal, the supernatant was
   incubated overnight at 4 °C with 2 μg of antibody. The mixture was then
   incubated with 30 μL of magnetic beads for 2 h at 4 °C with rotation.
   Bead-antibody-protein complexes were washed three times with cold IP
   buffer and boiled for 10 min with 5× SDS-PAGE loading buffer for
   subsequent western blot.

Quantitative PCR (qPCR)

   Total RNA was extracted with TRIzol reagent (15596026, Thermo Fisher
   Scientific). 2 μg RNA was reverse transcribed to cDNA using a Reverse
   Transcription Kit (K1621, Thermo Fisher Scientific). qPCR was performed
   in triplicate with SYBR Green (4309155, Life Technologies Corporation,
   Gaithersburg, MD) on a CFX Connect Real-Time PCR System (Bio-Rad).
   Relative mRNA expression was calculated using the 2^−ΔΔCT method with
   β-actin as the reference. Primers are listed in Table [68]S3-[69]5.

In vivo ubiquitination assay

   293T cells were co-transfected with the indicated plasmids. After 48 h,
   cells were treated with 20 μM MG132 for 6 h and lysed in IP buffer with
   protease inhibitor cocktail. Lysates were boiled for 10 min and then
   diluted with 900 μL of buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl,
   2 mM EDTA, 1% Triton X-100) and incubated at 4 °C for 30 min with
   rotation. After centrifugation, the supernatant was incubated overnight
   at 4 °C with anti-Flag or anti-TALDO1 antibodies, and then with 30 μL
   of magnetic beads for 2 h at 4 °C with rotation. Bead-antibody-protein
   complexes were washed three times with cold IP buffer, boiled with 5×
   SDS-PAGE loading buffer, and analyzed by immunoblotting using
   anti-K63Ub, anti-K48Ub, or anti-HA antibodies.

Immunofluorescence (IF)

   HK1 cells were transfected with the indicated plasmids for 48 h. Cells
   were fixed with 4% paraformaldehyde for 20 min, washed with 1× PBS, and
   permeabilized with 0.3% Triton X-100 in PBS for 15 min. After blocking
   with 5% goat serum for 60 min at 37 °C, cells were incubated overnight
   at 4 °C with primary antibodies diluted in 5% goat serum: Flag-tag
   (1:200, 80010-1-RR, Proteintech); His-tag (1:200, 66005-1-Ig,
   Proteintech); Flag-tag (1:200, F1804, Sigma-Aldrich); Myc-tag (1:200,
   16286-1-AP, Proteintech). Following washes, cells were incubated with
   secondary antibodies in 5% goat serum for 2 h at 37 °C: Anti-Rabbit-CF
   488 A (1:2000, SAB4600234, Sigma-Aldrich), and Anti-Mouse-Alexa Fluor
   594 (1:2000, A-11005, Thermo Fisher Scientific). DNA was stained with
   DAPI. Cells were washed with 1× PBS and imaged using a Confocal
   Microscope (TCS SP8, Leica, Wetzlar, Germany).

ELISA examination

   The concentrations of TALDO1 in NPC patient clinical serum samples were
   tested using an ELISA kit (HM11727, Bioswamp, Wuhan, China) according
   to the manufacturer’s protocol.

CCK-8 assay

   Cell proliferation activity was measured by the CCK-8 kit (C0005,
   TargetMol, Boston, MA). 1 × 10^3 NPC cells were cultured in 96-well
   plates and allowed to attach overnight at 37 °C. Then, at a certain
   time, 10 μL CCK-8 was added to each well. After being cultured at 37 °C
   for 2 h, the absorbance was measured at 450 nm on a plate reader
   (ELx800, Biotek, Winooski, VT).

Colony formation assay

   1 × 10^3 NPC cells were seeded in each well of 6-well plates. After 10
   to 14 days of culture, cells were washed three times with PBS, then
   stained with 0.05% crystal violet (V5265, Sigma-Aldrich). Colonies in
   each well were scanned, and the data were analyzed.

Apoptosis detection

   Cells of appropriate density were harvested, resuspended in binding
   buffer, and then incubated with Annexin V-FITC and PI (KGA1102, Keygen,
   Jiangsu, China) for 10 min, and the cell apoptosis was detected by flow
   cytometry (MoFlo XDP, Beckman Coulter, Miami, FL) and analyzed with
   FlowJo software.

Transwell assay and wound healing

   Transwell invasion assays were performed using 24-well Cell Culture
   Inserts (353097, Corning, NY) coated with Matrigel (354234, Corning).
   After 24 h of incubation at 37 °C, non-migrated cells in the upper
   chamber were removed. Cells were then fixed with 4% paraformaldehyde
   for 15 min, stained with crystal violet (V5265, Sigma-Aldrich), and
   counted from three random fields. For wound-healing assays, cells were
   cultured in 24-well plates and wounds were created using a 10 µL
   pipette tip. Cell migration was monitored under a microscope (DMI
   3000B, Leica, Wetzlar, Germany), and the wound-healing rate was
   analyzed using ImageJ software.

Recombinant TALDO1 protein and enzyme activity assay

   GST-TALDO1-WT/K7R/K7Q fusion proteins were expressed in E. coli BL21.
   BL21 were cultured at 37 °C to an OD600 of 0.6, then induced overnight
   with 0.1 mM IPTG at 18 °C. Proteins were purified using the GST-tag
   Protein Purification Kit (P2262, Beyotime). For the enzyme activity
   assay, GST tags were cleaved from the fusion proteins.TALDO1 enzyme
   activity was tested in the presence of 3.2 mM D-fructose 6-phosphate,
   0.2 mM erythrose-4-phosphate, 0.1 mM NADH, 5 mM EDTA, and 10 μg of
   α-glycerophosphate dehydrogenase/triosephosphate isomerase at a 1:6
   ratio in 40 mM triethanolamine at room temperature by continuous
   absorbance reading at 340 nm for 20 min [[70]19].

Targeted metabolomics

   1 × 10^6 SUNE1-Vec or SUNE1-TALDO1 stably expressing cells were
   extracted for targeted metabolites as previously described with slight
   modification [[71]20]. Then the cells were sent to Metware (Wuhan,
   China) for targeted metabolomics. Each group included six parallel
   samples. Results were normalized to the BCA protein of the cells in
   each sample. log2FC (fold change) ≥ 1.5 and FDR-adjusted p < 0.05 were
   used to screen for differential metabolic substrates. Orthogonal
   partial least squares discriminant analysis (OPLS-DA) was used to
   assess the overall data structure and group discrimination. Based on
   the OPLS-DA model, the Variable Importance in Projection (VIP) scores
   (VIP ≥ 1) derived from the multivariate model were used to identify
   significant differential metabolites between the WT and KR groups.
   Specifically, the screening criteria were as follows: log2FC (fold
   change) ≥ 1.5, VIP ≥ 1, and FDR-adjusted p < 0.05.

Lactate production assay

   The cell culture medium of NPC cells was collected, and the lactate
   concentration was determined using a lactate assay kit (MAK064,
   Sigma-Aldrich) according to the manufacturer’s protocol.

Glucose uptake assay

   The cell culture medium of NPC cells was collected, and the levels of
   glucose consumption were examined according to the instructions of the
   glucose uptake colorimetric assay kit (MAK083, Sigma-Aldrich).

Seahorse assay

   Extracellular Acidification Rate (ECAR) was measured by using a
   Seahorse XF96 instrument (Seahorse Biosciences, Billerica, MA)
   according to the manufacturer’s protocol. Cells were counted and seeded
   into the Seahorse 96-well plate at a density of 1 × 10^4 cells/well,
   and cultured for 12 h. After that, 10 mM glucose, 1 μM oligomycin and
   50 mM 2-DG were added to different ports of the Seahorse cartridge.
   Results were normalized to the BCA protein of the cells in each well.

Liquid chromatography mass spectrometry (LC-MS)

   HK1 cells were transfected with Flag-TALDO1 plasmid for 48 h, and the
   total protein was lysed with IP buffer containing protease inhibitor
   cocktail. Subsequently, IP experiment was performed with Flag-tag
   antibody, and magnetic beads-antibody-protein complexes were sent to
   Huaying Biotechnology (Shanghai, China) for LC-MS/MS analysis using the
   Orbitrap Fusion Lumos Tribrid mass spectrometer (Thermo Scientific).

Nuclear/cytoplasmic fractionation

   Nuclear and cytoplasmic lysates were prepared using the NE-PER Nuclear
   and Cytoplasmic Extraction Reagent kit (78833, Pierce, Rockford, IL)
   according to the manufacturer’s protocol.

Dual luciferase reporter assay

   pGL3-HK2/PDK1/LDHA promoters, pRL-TK and other indicated plasmids were
   co-transfected into SUNE1 cells. After 48 h of culture, the Dual
   Luciferase Reporter Assay Kit (RG027, Beyotime) was used according to
   the manufacturer’s protocol. Relative luciferase activity was
   calculated by normalizing to the quantified pRL-TK activity.

Animal experiments

   BALB/c female nude mice were randomly grouped. To verify the role of
   TALDO1 in NPC tumorigenesis, 5 × 10^6 SUNE1-Vec or SUNE1-TALDO1 cells
   suspended in 100 μL of PBS were injected subcutaneously into
   five-week-old mice to establish xenografts (n = 5). To investigate the
   role of HDAC6 in NPC progression, 5 × 10^6 SUNE1-shVec or SUNE1-shHDAC6
   cells were similarly injected (n = 5). Tumor size was measured every
   two days and calculated using the formula V = a × b^2/2, where a and b
   were the longest and shortest diameters, respectively. To assess the
   role of TALDO1 in NPC metastasis, 1 × 10^6 SUNE1-Vec or SUNE1-TALDO1
   cells in 50 μL of PBS were injected into the tail vein of five-week-old
   mice (n = 5). Additionally, to explore the effect of CAY10603 on NPC
   metastasis in vivo, 1 × 10^6 SUNE1 cells were injected into the tail
   vein of five-week-old mice (n = 5), followed by treatment with CAY10603
   (20 mg/kg, intraperitoneally, every three days for one month). After
   6–8 weeks, the mice were sacrificed, and lung tissues were collected
   for Immunohistochemistry (IHC) and hematoxylin and eosin (H&E)
   staining.

IHC

   Paraffin-embedded tumor tissues were sectioned into 5–10 μm slices and
   deparaffinized. Antigen retrieval was performed in 10 mM citrate buffer
   (pH 6.0) at 95 °C for 30 min. After quenching endogenous peroxidase
   using a universal two-step kit (PV-9000, ZSGB-BIO, Beijing, China) and
   blocking with 5% normal goat serum, slices were incubated with primary
   antibodies followed by peroxidase-linked secondary antibodies. Sections
   were displayed using a DAB Kit (ZLI-9017, ZSGB-BIO) and re-stained with
   hematoxylin (E607317, Sangon Biotech, Shanghai, China). Protein
   expression was semi-quantitatively evaluated using the method
   previously described [[72]21].

Bioinformatics analysis

   The TCGA-HNSC database and [73]GSE12452 database were used for TALDO1
   mRNA level analysis. The TCGA-HNSC database was used for survival curve
   analysis. The UniProt database ([74]https://www.uniprot.org/) was used
   to analyze the conservation of potential acetylation sites in TALDO1.

Statistical analysis

   According to statistical parameters, each experiment was repeated at
   least three times independently. GraphPad Prism 8 was used for
   statistical analysis. Data were expressed as the mean ± SD and analyzed
   using two-tailed t-tests, whereas for Kaplan–Meier analysis, the
   Log-rank (Mantel-Cox) test was used, and correlations were calculated
   using the parametric two-tailed Pearson correlation test. p < 0.05 (*),
   p < 0.01 (**) and p < 0.001 (***) were considered statistically
   significant.

Results

TALDO1 is deacetylated and downregulated in NPC

   Due to tumor suppressor molecules are usually deacetylated to lose
   their function in most tumors, histone deacetylase inhibitor (HDACi)
   can inhibit the malignant progression of tumors by restoring their
   acetylation [[75]12, [76]22]. In order to investigate tumor suppressor
   proteins directly regulated by deacetylation in NPC, TMT-based
   acetylation quantitative proteomic analysis was performed in NPC HK1
   cells treated with pan-HDACi LBH589 [[77]18]. KEGG analysis of
   upregulated acetylated proteins showed significant enriched in cellular
   energy metabolic pathways, such as carbon metabolism, citrate cycle,
   pyruvate metabolism, and glycolysis/gluconeogenesis (Fig. [78]1A).
   Among them, 15 metabolic enzymes were upregulated acetylated after
   LBH589 treatment (Fold change > 1.5, p < 0.05) (Fig. [79]1B, Table
   [80]S4). TALDO1 was the metabolic enzyme with the most significant
   upregulation of acetylation. IP experiments with a pan-acetylated
   lysine antibody confirmed increased TALDO1 acetylation and expression
   in LBH589-treated cells (Fig. [81]1C, D). Western blot analysis
   demonstrated elevated TALDO1 expression in NPC cells treated with
   LBH589 or another pan-HDACi PXD101 (Fig. [82]1E). Additionally, TALDO1
   was lower expressed in NPC cells compared to normal nasopharyngeal
   cells (Fig. [83]1F). Consistently, analyses of TALDO1 mRNA expression
   using TCGA and GEO datasets showed that TALDO1 was lowly expressed in
   HNSC and NPC, while the mRNA levels were not significantly associated
   with prognosis in HNSC patients (Fig. [84]S1). However, IHC staining
   analysis of NPC tissue microarrays showed that lower TALDO1 expression
   correlated with poorer overall survival (Fig.[85]1G, H). These results
   suggest its protein expression may have more obvious biological
   functions and the importance of post-transcriptional modification of
   TALDO1. Further, recent studies have reported that TALDO1 protein could
   be detected in the serum of patients with colorectal cancer [[86]23].
   Therefore, ELISA analysis of NPC serum samples with and without
   metastasis showed lower TALDO1 in metastatic cases (Fig. [87]1I). In
   summary, these findings suggest that the reduction of TALDO1 protein
   expression in NPC may be due to deacetylation, and the decrease of
   TALDO1 may contribute to tumor progression and metastasis.

Fig. 1. TALDO1 is deacetylated and downregulated in NPC.

   [88]Fig. 1
   [89]Open in a new tab

   A, B LC-MS/MS assays in HK1 cells with HDACi LBH589 treatment (100 nM,
   24 h). A KEGG pathways enrichment analysis of the upregulated
   acetylated non-histone proteins using Ingenuity Pathway Analysis (IPA).
   B The heatmap of the upregulated acetylated metabolism enzymes in the
   LBH589-treated groups compared with DMSO groups (Fold change > 1.5,
   p-value < 0.05). C After treatment with LBH589 (100 nM, 24 h) in HK1
   and SUNE1 cells, TALDO1 acetylation was detected by IP with an
   anti-acetylated lysine antibody and western blot. Ac-K acetylated
   lysine. D After transfected Flag-TALDO1 in 293T cells and treating with
   LBH589, TALDO1 acetylation was examine by IP and western blot. E
   Western blot analysis of TALDO1 in HK1, SUNE1, and HONE1 cells treated
   with LBH589 (100 nM, 24 h) or PXD101 (50 nM, 24 h). F Western blot
   analysis of TALDO1 in NPC cell lines (SUNE1, HONE1, 5-8F, 6-10B and
   HK1) and two normal nasopharyngeal mucosal/epithelial cell lines (NP460
   and NP69). IHC analysis of TALDO1 in NPC tissue microarrays. G The
   TALDO1 expression in NPC tissues by IHC staining (n = 126). H Overall
   survival rates in NPC patients with low or high TALDO1 expression
   assessed by Kaplan–Meier analysis. I TALDO1 protein levels in the serum
   of patients with metastatic or non-metastatic NPC were measured using
   the TALDO1 ELISA kit. Data were shown as the mean ± SD of at least
   three independent experiments.

TALDO1 inhibits proliferation and metastasis of NPC

   To investigate the function of TALDO1 in NPC, we constructed a
   TALDO1-overexpressing plasmid and confirmed its effect by qPCR and
   western blot (Fig. [90]S2). CCK-8 (Fig. [91]2A) and colony formation
   assays (Fig. [92]2B) showed that TALDO1 overexpression significantly
   suppressed NPC cell proliferation. Flow cytometry indicated that
   overexpression of TALDO1 induced cell apoptosis (Fig. [93]2C), and
   wound-healing and Transwell assays demonstrated reduced cell migration
   and invasion after overexpression of TALDO1 (Fig. [94]2D, E). In
   xenograft experiments, tumor growth was significantly slower in mice
   injected with TALDO1-overexpressing SUNE1 cells (Fig. [95]2F–H), and
   IHC staining revealed a decrease in Ki-67 expression (Fig. [96]2I).
   Metastasis experiments showed fewer and smaller metastatic lung foci in
   the TALDO1-overexpressing group (Fig. [97]2J, K). These findings
   collectively suggest that TALDO1 inhibits NPC proliferation and
   metastasis in vitro and in vivo.

Fig. 2. TALDO1 inhibited the proliferation and metastasis of NPC.

   [98]Fig. 2
   [99]Open in a new tab

   A–E After transfections of pcDNA3.1-TALDO1 plasmid into NPC cells for
   48 h. A, B CCK-8 assay and colony forming assay were performed to
   examine the cell proliferation. C The cell apoptosis was analyzed by
   flow cytometry. D Wound-healing assay was performed to detect cell
   migration. E Transwell assays were carried out to detect cell invasion.
   Scale bar, 50 μm. F–I After 5 × 10^6 SUNE1-Vec and SUNE1-TALDO1 stable
   cells were injected subcutaneously into 5-week-old nude mice (n = 5),
   the tumor growth (F), tumor volume (G) and tumor weight (H) were
   monitored. I Representative images of IHC staining of TALDO1 and Ki-67
   in xenograft tumor tissues. J, K After 1 × 10^6 SUNE1-Vec and
   SUNE1-TALDO1 stable cells were injected into the lateral tail vein of
   5-week-old nude mice (n = 5), J Images and quantitative data of lung
   metastatic foci, K Representative images of H&E staining of lung foci
   and IHC staining of TALDO1 in lung foci. Data were shown as the mean ±
   SD of at least three independent experiments. *p < 0.05 and **p < 0.01
   and ***p < 0.001.

HDAC6 mediates deacetylation of TALDO1

   To identify the potential deacetylase for TALDO1, two selective HDACi
   were used. The data showed that TALDO1 acetylation was significantly
   induced by LMK-235 (specific HDAC4/5/6/11/8 inhibitor) but not MS-275
   (specific HDAC1/2/3 inhibitor), and LMK-235-treated NPC cells exhibited
   higher TALDO1 protein expression (Fig. [100]S3A, B). IP assays in 293T
   cells with His-tagged HDAC4, 5, and 6 suggested that TALDO1 interaction
   with these deacetylases (Fig. [101]3A). However, only ectopic HDAC6
   expression significantly induced TALDO1 deacetylation (Fig. [102]3B,
   C). Further, IP experiments in NPC cells transfected with Flag-TALDO1
   confirmed TALDO1 coprecipitation with HDAC6 (Fig. [103]3D), and
   extensive colocalization of HDAC6 with TALDO1 was observed in HK1 cells
   (Fig. [104]3E). Consistent with this, deacetylase-dead HDAC6 (HDAC6-DN)
   [[105]24] did not alter TALDO1 acetylation compared to wild type in
   shHDAC6 293T cells (Fig. [106]3F). Therefore, we established stable NPC
   cell lines with HDAC6 knockdown, and found that HDAC6 knockdown (Fig.
   [107]3G), but not HDAC4 or HDAC5 (Fig. [108]S3C), increased TALDO1
   protein in NPC cells.

Fig. 3. HDAC6 mediates deacetylation of TALDO1.

   [109]Fig. 3
   [110]Open in a new tab

   A–C After indicated plasmids were transfected into 293T cells for 48 h,
   A Interaction between HDAC4/5/6 and TALDO1 was detected by IP and
   western blot. B, C TALDO1 acetylation was detected by IP and western
   blot. D After transfected Flag-TALDO1 in NPC cells, interaction between
   exogenous Flag-TALDO1 and endogenous HDAC6 was analyzed by IP and
   western blot. E Indicated plasmids were transfected into HK1 cells, and
   the colocalization of TALDO1 and HDAC6 was investigated by confocal
   microscopy. Scale bar, 5 μm. F Indicated plasmids were transfected into
   293T cells, TALDO1 acetylation was detected by IP and western blot. G
   After established HK1, SUNE1 and HONE1 knockdown HDAC6 stable cell
   lines through transfected shHDAC recombinant lentivirus, western blot
   analysis of TALDO1 expression in HDAC6 stable knockdown NPC cells. H
   After co-transfected Flag-TALDO1 and Myc-p300 plasmids in 293T cells,
   and interaction between TALDO1 and p300 was detected by IP and western
   blot. I The endogenous interaction between TALDO1 and p300 was detected
   by IP and western blot. J After co-transfected Flag-TALDO1 and Myc-p300
   plasmids in HK1 cells, the colocalization of TALDO1 and p300 was
   investigated by confocal microscopy. Scale bar, 5 μm. K After
   co-transfected Flag-TALDO1 and Myc-p300 plasmids in 293T cells, TALDO1
   acetylation was detected by IP and western blot. L After co-transfected
   Flag-TALDO1 and Myc-p300 WT or DY (a catalytic inactivation mutant of
   p300) plasmids in 293T cells, TALDO1 acetylation was analyzed by IP and
   western blot. M Western blot analysis of TALDO1 in p300 overexpression
   NPC cells. N After co-transfected Flag-TALDO1 and Myc-p300 followed by
   transfection with His-HDAC6 expression plasmid for 48 h, TALDO1
   acetylation was detected by IP and western blot. O The expression and
   correlation of TADLO1 and HDAC6 or p300 were analyzed by IHC staining
   of clinical NPC tissues. Sample size (n = 35), Pearson correlation
   coefficient (r) and p-value (two-tailed) are indicated.

   To identify the acetyltransferase acting on TALDO1, we first
   investigated the interactions between TALDO1 and five common
   acetyltransferases. IP experiments showed that TALDO1 directly
   interacted with p300 in 293T cells (Fig. [111]S4A), and p300
   overexpression could promote TALDO1 protein expression (Fig. [112]S4B).
   Further exogenous and endogenous IP experiments confirmed TALDO1
   coprecipitation with p300 (Fig. [113]3H, I), and IF showed strong
   colocalization of two molecules (Fig. [114]3J). Consistent with this,
   IP experiments in 293T cells demonstrated that overexpression of
   p300-wild type (WT), but not acetyltransferase-deficient p300 (p300-DY)
   [[115]25], induced TALDO1 acetylation (Fig. [116]3K, L). Furthermore,
   p300 overexpression increased TALDO1 protein in NPC cells (Fig.
   [117]3M). Importantly, our results showed that HDAC6 disrupted
   p300-mediated upregulation of TALDO1 acetylation (Fig. [118]3N).
   Additionally, IHC staining of NPC patient tissue samples showed a
   strong negative correlation between HDAC6 and TALDO1 expression, and a
   positive correlation between p300 and TALDO1 expression (Fig. [119]3O).
   These findings suggest that HDAC6-induced deacetylation decreases
   TALDO1 protein expression, whereas p300-mediated acetylation enhances
   TALDO1 expression.

HDAC6 promotes proliferation and metastasis of NPC through TALDO1

   To investigate the function of HDAC6 in NPC, the expression of HDAC6
   was examined, and qPCR and western blot data showed that HDAC6 was
   highly expressed in NPC cells (Fig. [120]S5A, B). CCK-8, colony
   formation and Transwell assays indicated that HDAC6 knockdown
   significantly inhibited NPC cell proliferation and invasion (Fig.
   [121]S5C–F). Further, xenograft experiments demonstrated that the tumor
   growth was significantly slower in the shHDAC6 group (Fig. [122]4A–C),
   and IHC staining showed a decrease in HDAC6 and Ki-67, while an
   increase in TALDO1 expression (Fig. [123]4D). Meanwhile, metastatic
   experiments indicated fewer lung metastatic foci in the shHDAC6 group,
   with reduced HDAC6 and increased TALDO1 expression in metastases (Fig.
   [124]4E, F). Consistently, NPC cells treated with the HDAC6-specific
   inhibitor CAY10603 exhibited a concentration-dependent increase in
   TALDO1 expression (Fig. [125]S6A). Functional assays further showed
   that CAY10603 significantly inhibited NPC proliferation and invasion
   (Fig. [126]S6B–D), and reduced metastasis in vivo (Fig. [127]S6E, F).
   To determine whether HDAC6 affects NPC proliferation and invasion by
   inhibiting TALDO1, HK1 and HONE1 cells with double knockdown of HDAC6
   and TALDO1 were generated (Fig. [128]4G). CCK-8 (Fig. [129]4H), colony
   formation (Fig. [130]4I) and Transwell assays (Fig. [131]4J) indicated
   that TALDO1 knockdown rescued the proliferation and invasion defects
   caused by HDAC6 silencing. Collectively, these results demonstrate that
   HDAC6 promotes NPC proliferation and metastasis by downregulating
   TALDO1.

Fig. 4. HDAC6 promotes the proliferation and metastasis of NPC through
TALDO1.

   [132]Fig. 4
   [133]Open in a new tab

   A–C After SUNE1-shVec and SUNE1-shHDAC6 stable cells were injected
   subcutaneously into 5-week-old nude mice (n = 5), the tumor growth (A),
   tumor volume (B) and tumor weight (C) were monitored. D Representative
   images of IHC staining of HDAC6, TALDO1 and Ki-67 in xenograft tumor
   tissues. E, F After SUNE1-shVec and SUNE1-shHDAC6 stable cells were
   injected into the lateral tail vein of 5-week-old female nude mice
   (n = 5), E images and quantitative data of lung metastatic foci, F
   representative images of H&E staining of lung foci and IHC staining of
   HDAC6 and TALDO1 in lung foci. G Western blot analysis of TALDO1
   expression after TALDO1 knockdown in HK1-shHDAC6 and HONE1-shHDAC6
   stable cells. H, I CCK-8 assays (H) and colony forming assay (I) were
   performed to examine the cell proliferation. J Transwell assays were
   carried out to detect cell invasion. Scale bar, 50 μm. Data were shown
   as the mean ± SD of at least three independent experiments. *p < 0.05
   and **p < 0.01 and ***p < 0.001.

K7 deacetylation attenuates TALDO1 stability

   Previous acetylation quantitative proteomic data showed a 5.7-fold
   increase in K7 acetylation of TALDO1 in LBH589-treated cells compared
   to controls (Fig. [134]5A), and K7 is highly conserved across species,
   indicating its significance as the main acetylation site (Fig.
   [135]5B). To verify this, Flag-TALDO1 K7R (mimicking deacetylation) and
   K7Q (mimicking acetylation) [[136]26] mutants were transfected into
   293T or NPC cells. The K7R mutant exhibited reduced TALDO1 acetylation
   compared to wild-type (WT) TALDO1 (Fig. [137]5C, D). To analyze whether
   acetylation is involved in the regulation of protein expression,
   cycloheximide assays in NPC cells showed that HDAC6 inhibition
   increased TALDO1 protein half-life (Figs. [138]5E, [139]S7A),
   demonstrating that acetylation enhances TALDO1 stability. Further,
   treatment with the proteasome inhibitor MG132 or autophagy inhibitor
   chloroquine did not alter TALDO1 protein levels (Figs. [140]5F,
   [141]S7B), indicating that acetylation did not affect the degradation
   of TALDO1 through ubiquitination or autophagy. Given that K48-linked
   ubiquitination leads to protein degradation, and K63-linked
   ubiquitination is predominantly associated with protein stability and
   signal transduction [[142]27]. Therefore, we analyzed whether
   K63-linked ubiquitination is involved in the stability regulation of
   TALDO1. IP experiments confirmed the existence of ubiquitination
   modification in TALDO1 (Fig. [143]5G). HDAC6 knockdown, LBH589
   treatment and p300 overexpression increased TALDO1 ubiquitination
   (Figs. [144]5H, I, [145]S7C). Consistent with this, TALDO1 K7R showed
   lower ubiquitination than TALDO1 WT (Fig. [146]5J). Furthermore, IP
   experiments with ubiquitin plasmids (K48 and K63) and their mutants
   (K48R and K63R) revealed K63-linked ubiquitination as the major type
   for TALDO1 (Fig. [147]S8A, [148]5K), and endogenous IP experiments also
   confirmed this finding (Fig. [149]S8B). Thus, K7 deacetylation
   attenuates TALDO1 stability through inhibiting K63-linked
   ubiquitination. Moreover, biological function assays showed that the
   TALDO1 K7 mutation inhibited NPC cell proliferation and invasion (Fig.
   [150]S9).

Fig. 5. K7 deacetylation attenuates TALDO1 stability.

   [151]Fig. 5
   [152]Open in a new tab

   A Acetylated TALDO1 K7 was identified by LC-MS/MS. B The sequences
   around TALDO1 K7 from different species were aligned. C Indicated
   plasmids were transfected into 293T cells, the acetylation was analyzed
   by IP and western blot. D Indicated plasmids were transfected into NPC
   cells, western blot analysis of TALDO1. E After NPC shHDAC6 stable
   cells were treated with CHX (0.1 μM), TALDO1 protein was analyzed by
   western blot. The amount of TALDO1 protein was quantitated and
   calculated by ImageJ software. F HK1 and SUNE1 cells were treated with
   MG132 (10 μM) or chloroquine (CHQ) (50 μM) for 12 h, western blot
   analysis was used to detect TALDO1 expression. G 293T cells were
   transfected with HA-Ub together with Vec or Flag-TALDO1. Anti-HA was
   used for IP and western blot was performed with Flag antibody to detect
   TALDO1 ubiquitination. H Indicated plasmids were transfected into 293T
   cells, the ubiquitination was analyzed by IP with a Flag antibody,
   followed by western blot for anti-HA. I Indicated plasmids were
   transfected into 293T cells, the ubiquitination was detected. J
   Flag-TALDO1 WT, K7R, or K7Q were transfected into 293T cells, the
   ubiquitination was examined. K 293T cells were transfected with
   Flag-TALDO1 together with HA-Ub, HA-K48R, or HA-K63R. The
   ubiquitination was detected.

Deacetylation inhibits SMURF1-mediated K63-linked ubiquitination of TALDO1

   To identify the E3 ligase for TALDO1, potential candidates were
   screened and ranked using the UbiBrowser database (Fig. [153]6A). IP
   experiments showed that SMURF1 overexpression decreased TALDO1
   ubiquitination, while SYVN1 had no effect (Fig. [154]6B). Further IF
   and IP assays confirmed an interaction between SMURF1 and TALDO1 (Fig.
   [155]6C, D), and SMURF1 could significantly inhibit K63-linked
   ubiquitination and slightly promote K48-linked ubiquitination of TALDO1
   (Fig. [156]6E). In addition, IP experiments revealed that the K7R
   mutation enhanced the interaction between TALDO1 and SMURF1 (Fig.
   [157]6F) and inhibited SMURF1-mediated K63-linked ubiquitination while
   promoting K48-linked ubiquitination (Fig. [158]6G). These results
   suggest that K7 deacetylation reduces SMURF1-mediated K63-linked
   ubiquitination and increases K48-linked ubiquitination of TALDO1,
   thereby decreasing TALDO1 stability.

Fig. 6. Deacetylation inhibits SMURF1-mediated K63-linked ubiquitination of
TALDO1.

   [159]Fig. 6
   [160]Open in a new tab

   A The E3 ligases associated with TALDO1 were screened using the
   UbiBrowser database ([161]http://ubibrowser.ncpsb.org/), and ranked
   them by binding score and confidence level. B Indicated plasmids were
   transfected into 293T cells, the ubiquitination was examined. C After
   co-transfected Flag-TALDO1 and His-SMURF1 plasmids in HK1 cells, the
   colocalization of TALDO1 and SMURF1 was investigated by confocal
   microscopy. Scale bar, 5 μm. D After transfected Flag-TALDO1 and
   His-SMURF1 in 293T cells, interaction between TALDO1 and SMURF1 was
   detected by IP and western blot. E Indicated plasmids were transfected
   into 293T cells, the ubiquitination was detected. F His-SMURF1 and
   Flag-TALDO1-WT or K7R were transfected into 293T cells, interaction
   between TALDO1 and SMURF1 was examined by IP and western blot. G
   Indicated plasmids were transfected into 293T cells, the ubiquitination
   was detected.

K7 deacetylation of TALDO1 promotes glycolysis in NPC cells independent of
enzyme activity

   TALDO1 is an important metabolic enzyme for the balance of metabolites
   in the PPP and glycolysis pathways [[162]28]. To assess how TALDO1
   affects the energy metabolism of NPC cells, targeted metabolomics
   analysis was performed on TALDO1-overexpressing NPC cells. Principal
   Component Analysis (PCA) showed clear separation between groups (Fig.
   [163]S10A). TALDO1 overexpression led to increased upstream substrates
   such as Glucose-6-phosphate (G6P) and 6-phosphogluconic acid (6PGA) in
   glucose metabolism, but resulted in a decrease in downstream glycolysis
   and TCA cycle substrates, notably 3-phosphoglycerate (3PGA),
   phosphoenolpyruvate (PEP) and isocitric acid. However, the changes in
   its direct substrates such as erythrose-4-pyosphate (E4P) were not
   significant (Figs. [164]7A–C, [165]S10B). These results suggest that
   TALDO1 may reduce glycolysis and TCA cycle in NPC cells. Further,
   glucose uptake and lactate production assays demonstrated that TALDO1
   inhibited both processes (Fig. [166]7D, E). Meanwhile, Seahorse assays
   showed that TALDO1 overexpression decreased glycolysis and glycolytic
   capacity in NPC cells, and this effect was reversed by the K7R mutant
   (Fig. [167]7F). Consistent with this, the glycolysis and glycolytic
   capacity were also reduced in shHDAC6 NPC cells (Fig. [168]S10C).
   However, enzyme activity assays indicated K7 acetylation did not
   significantly affect TALDO1 activity (Fig. [169]7G), suggesting that
   TALDO1 may influence glucose metabolism independent of its enzymatic
   activity but rather through moonlighting functions. Furthermore, we
   assessed the expression of key rate-limiting enzymes in the glycolysis
   and TCA cycle. Western blot analysis showed that the glycolytic
   rate-limiting enzymes HK2 and LDHA, as well as the TCA cycle enzyme
   PDK1, were downregulated in TALDO1-overexpressing cells (Fig. [170]7H).
   Taken together, these results demonstrate that deacetylation of TALDO1
   enhances glycolysis of NPC cells independently of its enzymatic
   activity.

Fig. 7. K7 deacetylation of TALDO1 promotes glycolysis in NPC cells
independent of enzyme activity.

   [171]Fig. 7
   [172]Open in a new tab

   A–C Targeted metabolomics with TALDO1-overexpressing and control SUNE1
   cells was performed. A Differentially enriched pathways of SUNE1 cell
   with overexpressed TALDO1. Differential metabolites were identified
   based on Log2FC (fold change) ≥1.5, VIP ≥ 1, and FDR-adjusted p < 0.05.
   VIP values were extracted from the OPLS-DA results. B Altered
   metabolite levels in the main energy fluxes in TALDO1-overexpressing
   cells compared to the control cells. C Heatmap of most significant
   changed metabolites by the targeted metabolomics analysis. The
   differential metabolites were determined by Log2FC ≥ 1.5 and
   FDR-adjusted p < 0.05. D, E After transfection with pcDNA3.1-TALDO1
   plasmids for 48 h, glucose uptake (D) and lactate production (E) in
   cell culture medium were measured using the glucose or lactate
   colorimetric assay kit. F After transfection with Vec,
   pcDNA3.1-TALDO1(WT), pcDNA3.1-TALDO1(K7R) plasmids into HK1 and SUNE1
   cells, the extracellular acidification rate (ECAR) was measured using
   Seahorse XF assay Kit. The glycolysis and glycolysis capacity were
   analyzed. G Wild type, K7R, K7Q acetylated mutant recombinant proteins
   were used to detect TALDO1 enzyme activity in vitro. H Western blot
   assay was performed to assess the glycolysis rate-limiting enzyme
   levels in TALDO1 overexpression NPC cells. Data were shown as the mean
   ± SD of at least three independent experiments. *p < 0.05 and
   **p < 0.01 and ***p < 0.001.

   Additionally, we aimed to clarify whether TALDO1-mediated glycolysis
   inhibition is involved in the proliferation and invasion of NPC cells.
   CCK-8, colony formation and Transwell assay results showed that stable
   knockdown of TALDO1 significantly promoted the proliferation and
   invasion of NPC cells, and further treatment with 2-DG (competitive
   glycolysis inhibitor) reversed the effect (Fig. [173]S11A–D). Further,
   glycolysis enzyme expression was significantly increased in the TALDO1
   knockdown group, but this increase was mitigated by 2-DG treatment
   (Fig. [174]S11E). Collectively, these results demonstrate that TALDO1
   inhibits the proliferation and invasion of NPC cells by suppressing
   glycolysis.

Deacetylation inhibits nuclear translocation of TALDO1 and promotes
glycolysis of NPC cells

   Traditionally, metabolic reprogramming was linked to enzyme activity
   and substrate concentrations. Recent findings suggest that modulating
   enzyme moonlighting functions also enables tumor cells to adapt during
   tumorigenesis and progression [[175]29, [176]30]. The two TALDO1
   isoforms, generated by alternative translation initiation, show
   distinct nuclear and cytoplasmic localization. The long isoform is
   crucial for nuclear targeting, but the mechanisms regulating its
   nuclear localization and its metabolic role remain unclear [[177]31].
   We analyzed the protein domains of TALDO1 and found that its
   acetylation site, K7, is located within the nuclear localization signal
   (NLS) (Fig. [178]8A). IF and nuclear-cytoplasmic fractionation
   experiments showed that wild-type TALDO1 primarily localizes in the
   nucleus, while the K7R mutant predominantly localizes in the cytoplasm
   (Fig. [179]8B, C), indicating that deacetylation inhibits TALDO1
   nuclear translocation. To explore the mechanism by which nuclear TALDO1
   exerts its tumor-suppressive function, we performed mass spectrometry
   to analyze TALDO1-binding proteins in NPC cells. The results suggested
   that the tumor suppressor protein BRCA1, known for its transcriptional
   regulatory properties, might interact with TALDO1 (Fig. [180]8D). IP
   experiments confirmed this, and K7R significantly weakened the
   TALDO1-BRCA1 interaction (Fig. [181]8E). Previous studies have
   demonstrated that BRCA1 can directly inhibit the transcriptional
   activity of c-Myc [[182]32], thereby reducing the expression of
   glycolysis-related metabolic enzymes such as HK2, LDHA, and PDK1
   [[183]33, [184]34]. We knocked down BRCA1 expression in SUNE1 cells
   overexpressing either wild-type or K7R mutant TALDO1. The results
   showed that wild-type TALDO1 overexpression decreased the mRNA and
   protein expression of HK2, LDHA, and PDK1, while the K7R mutation
   reversed this effect. Furthermore, BRCA1 knockdown alleviated the
   decrease in metabolic enzyme mRNA and protein expression caused by
   TALDO1 overexpression, but had no significant effect in the K7R group
   (Fig. [185]8F, G). We also constructed reporter plasmids for HK2, LDHA,
   and PDK1, and dual-luciferase assays demonstrated that wild-type TALDO1
   overexpression inhibited the transcriptional activity of these genes,
   and this effect was reversed by the K7R mutation. BRCA1 knockdown
   counteracted the changes in transcriptional activity induced by TALDO1
   overexpression, whereas knocking down c-Myc expression negated the
   effect of BRCA1, with no significant changes observed in the K7R group
   (Fig. [186]8H, I). Western blot experiments further confirmed these
   findings (Fig. [187]8J). These results suggest that
   acetylation-mediated nuclear translocation of TALDO1 inhibits
   glycolysis by interacting with BRCA1, which suppresses c-Myc-mediated
   transcriptional activation of glycolytic enzymes. IHC staining of NPC
   tissue microarrays revealed that TALDO1 expression was lower in tumor
   tissues compared to paired adjacent tissues. Moreover, primary tumors
   with metastases exhibited lower TALDO1 expression than those without
   metastases. Notably, as the malignancy of the tissue increased, nuclear
   localization of TALDO1 decreased significantly (Fig. [188]8K, L). Taken
   together, these results indicate that inhibition of glycolysis by
   TALDO1 is dependent on acetylation-mediated nuclear translocation.

Fig. 8. Deacetylation inhibits nuclear translocation of TALDO1 and promotes
glycolysis of NPC cells.

   [189]Fig. 8
   [190]Open in a new tab

   A Schematic diagram of the domain of TALDO1 protein. B After
   transfecting SUNE1 cells with Flag-TALDO1-WT or Flag-TALDO1-K7R
   plasmids for 48 h, cytoplasmic and nuclear fractions were obtained
   using the nuclear cytoplasmic extraction reagent kit, TALDO1
   distribution in the cytoplasm and nucleus was analyzed by western blot,
   with H3 as the nuclear marker and actin as the cytoplasmic marker. C
   The subcellular localization of TALDO1 was analyzed by IF assay 48 h
   after transfection of the indicated plasmid in HK1 cells. D 48 h after
   HK1 cells transfected Flag-TALDO1 plasmid, the proteins interacting
   with TALDO1 were analyzed by protein mass spectrometry. E HA-BRCA1 and
   Flag-TALDO1-WT or K7R were transfected into 293T cells, interaction
   between TALDO1 and BRCA1 was examined by IP and western blot. F The
   mRNA expression of HK2, LDHA and PDK1 was detected by qPCR after
   indicated plasmid was transfected into SUNE1 cells for 48 h. G The
   protein expression of HK2、LDHA and PDK1 was detected by western blot
   after indicated plasmid was transfected into SUNE1 and HONE1 cells for
   48 h. H, I After the indicated plasmids were transfected into SUNE1
   cells for 48 h, the transcriptional activities of HK2, LDHA and PDK1
   were detected by Dual luciferase reporter assay. J After the indicated
   plasmids were transfected into SUNE1 and HONE1 cells for 48 h, the
   protein expression of c-Myc, HK2, LDHA and PDK1 was detected by western
   blot. K, L The expression and subcellular localization of TALDO1 were
   assessed by immunohistochemistry in NPC and adjacent nasopharyngeal
   mucosa/epithelial tissues (K), as well as in metastatic and
   non-metastatic NPC tissues (L). Data were shown as the mean ± SD of at
   least three independent experiments. *p < 0.05 and **p < 0.01 and
   ***p < 0.001.

Deacetylation disrupts nuclear translocation of TALDO1 to promote NPC
proliferation and migration via the BRCA1/c-Myc axis

   To investigate the role of the BRCA1/c-Myc axis in TALDO1-mediated
   glycometabolic inhibition of NPC cells, biological functional assays
   revealed that BRCA1 knockdown in wild-type TALDO1-overexpressing NPC
   cells reversed the inhibition of cell proliferation and migration
   induced by TALDO1 overexpression. However, this effect was absent in
   the K7R mutant group (Fig. [191]9A–C). Consistently, glucose uptake and
   lactate production assays demonstrated that TALDO1 inhibits glycolysis
   in NPC cells via BRCA1, a loss of function in the K7R deacetylation
   mutant (Fig. [192]9D, E). Further biological functional assays showed
   that c-Myc knockdown significantly suppressed the cell proliferation
   and migration promotion effects caused by BRCA1 downregulation in
   wild-type TALDO1-overexpressing NPC cells. This suppression was not
   observed in the K7R mutant group (Fig. [193]9F–H). Additionally, IHC
   staining of NPC tissues revealed a significant negative correlation
   between TALDO1 expression and glycolysis-related enzymes HK2, LDHA, and
   PDK1 (Fig. [194]9I). These results suggest that deacetylation-mediated
   disruption of TALDO1 nuclear translocation promotes glycolysis through
   the BRCA1/c-Myc axis, thereby enhancing NPC cell proliferation and
   migration.

Fig. 9. Deacetylation-mediated nuclear translocation of TALDO1 promotes NPC
proliferation and migration via the BRCA1/c-Myc axis.

   [195]Fig. 9
   [196]Open in a new tab

   A–E After transfections of the indicated plasmids into NPC cells for
   48 h. A, B CCK-8 assay and colony forming assay were performed to
   examine the cell proliferation. C Transwell assays were carried out to
   detect cell invasion. Scale bar, 50 μm. D, E After transfection with
   indicated plasmids for 48 h, glucose uptake (D) and lactate production
   (E) in cell culture medium were measured using the glucose or lactate
   colorimetric assay kit. F–H After transfection with indicated plasmids
   for 48 h, F, G CCK-8 assay and colony forming assay were performed to
   examine the cell proliferation. H Transwell assays were carried out to
   detect cell invasion. Scale bar, 50 μm. I The expression and
   correlation of TADLO1 and HK2, LDHA or PDK1 were analyzed by IHC
   staining of clinical NPC tissues. Sample size (n = 35), Pearson
   correlation coefficient (r) and p-value (two-tailed) are indicated.
   Data were shown as the mean ± SD of at least three independent
   experiments. No significant (ns), *p < 0.05 and **p < 0.01 and
   ***p < 0.001.

Discussion

   Several studies have revealed the essential role of TALDO1 in tumor
   progression. The co-expression of SLC1A5 and TALDO1 in estrogen
   receptor-positive breast cancer was associated with poor clinical
   outcomes in patients treated with endocrine therapy [[197]35]. TALDO1
   has also been associated with drug resistance in HER2-positive breast
   cancer, and deletion of TALDO1 and HER2 inhibition improved
   HER2-positive breast cancer treatment [[198]36]. Higher TALDO1
   expression in upper tract urothelial carcinoma could predict poor
   prognosis [[199]37]. In contrast, TALDO1 knockout mice spontaneously
   develop hepatocellular carcinoma (HCC), suggesting that TALDO1 can
   inhibit HCC carcinogenesis [[200]38]. Consistently, the TALDO1 deletion
   mutant phenotype in humans may lead to early HCC formation [[201]39].
   Increased TALDO1 expression in HNSC predicted better recurrence-free
   survival [[202]40]. These results imply that TALDO1 may have a dual
   role in tumors, and that it acts as an oncogene or as a tumor
   suppressor depending on the tumor type. In this study, we found that
   TALDO1 expression was downregulated in NPC and associated with poor
   prognosis. Ectopic expression of TALDO1 significantly suppressed NPC
   proliferation and metastasis in vitro and in vivo. Our findings
   demonstrate that TALDO1 functions as a tumor suppressor in NPC.

   Post-translational modifications are key regulators of protein
   expression [[203]41]. Here, we demonstrate for the first time that the
   deacetylase HDAC6 interacts with TALDO1, mediating its deacetylation
   and reducing its protein expression, while the acetyltransferase p300
   enhances TALDO1 acetylation, increasing its protein expression.
   Meanwhile, we found that K7 deacetylation inhibits E3 ligase
   SMURF1-mediated K63-linked ubiquitination and promotes K48-linked
   ubiquitin-mediated degradation of TALDO1, thereby reducing its protein
   stability. Our study elucidated the regulatory mechanism of TALDO1
   expression from the perspective of the interaction between protein
   acetylation and ubiquitination.

   TALDO1 is the rate-limiting enzyme of PPP that converts the glycolytic
   intermediate GAP to F6P and plays an important role in linking
   glycolysis and PPP [[204]42]. Our metabolic analysis demonstrated that
   TALDO1 overexpression significantly reduced glycolysis and TCA
   metabolism. Meanwhile, we observed that the TALDO1 K7R deacetylation
   mutant enhanced the glycolytic capacity of NPC cells. However, in vitro
   enzyme activity assays demonstrated that the TALDO1 K7R deacetylation
   mutant did not alter its enzymatic activity. It is suggested that
   deacetylation of TALDO1 may play a role in promoting the glycolysis of
   tumor cells through moonlighting functions.

   In recent years, it has been found that the metabolism reprogramming
   can be realized not only by regulating the classical function of
   metabolic enzyme activity, but also by regulating its moonlighting
   function to give tumor cells adapt to different demands in the
   development of tumor [[205]5, [206]43]. Research has found that the
   main modes of the moonlighting function action of metabolic enzymes in
   tumors includes regulating gene transcription: direct interaction with
   and modulation of transcription factors and transcriptional
   co-regulators, even some enzymes may directly serve as transcription
   factors or regulators after nuclear translocation, whereas subcellular
   localization of some other enzymes may regulate gene transcription
   through histone modification; The intracellular re-distribution of some
   enzymes may establish novel protein interactions, leading to changes in
   the functionality of signaling pathways; And multiple studies revealed
   the direct regulation of protein substrates by metabolic enzymes with
   kinase activity, etc [[207]44, [208]45]. For example, ERK2 mediated
   phosphorylation of PKM2 at Ser37, resulting in PKM2 binding to importin
   α5 and undergoing nuclear translocation. Nuclear PKM2 functions as a
   coactivator of β-catenin, inducing c-Myc expression, which led to the
   upregulation of GLUT1 and LDHA, thereby enhancing glycolysis and
   promoting glioblastoma progression [[209]46]. NAT10 stabilizes ACLY via
   K468 acetylation, thereby increasing nuclear acetyl-CoA levels, which
   upregulate H3K27ac-mediated transcription of CYP2C9 and PIK3R1,
   consequently driving chemoresistance in HCC [[210]47]. Additionally,
   the elevated expression of enolase-1 (ENO1) facilitated its interaction
   with choline kinase α (CHKα), which competitively disrupted the
   CHKα-TRIM25 binding, increased the stability of CHKα, thereby enhancing
   choline metabolism and accelerating brain tumor growth [[211]48].
   Moreover, Creatine kinase B (CKB) phosphorylates GPX4 at S104,
   functioning as a protein kinase, which suppresses HSC70-mediated
   autophagic degradation of GPX4, thereby attenuating ferroptosis and
   advancing HCC progression [[212]49].

   The TALDO1 gene produces two isoforms: a full-length 337-amino acid
   protein and a shorter version missing the first 10 amino acids, which
   lacks nuclear localization. While subcellular localization of TALDO1 is
   not crucial for its activity in the PPP, it influences NIH/3T3 cell
   metabolism, including the TCA cycle, nucleotide metabolism, and
   glycolysis [[213]31]. In this study, unlike nuclear-translocating
   metabolic enzymes such as PKM2 and ACLY that directly act as
   transcriptional coactivators or remodel histone acetylation,
   nuclear-translocated TALDO1 suppresses the c-Myc transcriptional
   activation through binding to the transcriptional corepressor BRCA1.
   Consistently, recent studies demonstrate that nuclear-localized
   glutamate-cysteine ligase modifier subunit (GCLM) competitively
   interacts with the NF-κB repressing factor (NKRF), thereby promoting
   NF-κB activity and augmenting chemoresistance in colorectal cancer
   [[214]50].

   In conclusion, our study highlighted the critical regulatory role of
   deacetylation of TALDO1 in regulating its stability and nuclear
   translocation, as well as its impact on glycolysis. Meaningfully, we
   found that nuclear TALDO1 possessed a moonlighting function, which
   interrupted c-Myc-mediated transcriptional activation of HK2/LDHA/PDK1
   through binding to BRCA1 (Fig. [215]10). This study provides a novel
   biomarker and intervention strategy for NPC from the perspective of
   protein acetylation modification.

Fig. 10. Graphical model.

   Fig. 10
   [216]Open in a new tab

   Schematic diagram illustrates that the potential mechanisms by which
   deacetylation-dependent regulation of TADLO1 stability and nuclear
   translocation promotes glycolysis and malignant progression of NPC.

Supplementary information

   [217]Supplementary Materials^ (3.7MB, docx)
   [218]original western-blots^ (15.6MB, docx)

Acknowledgements