Abstract

   Isoniazid (INH) is a first-line anti-tuberculosis drug used for nearly
   70 years. However, the mechanism underlying the side effects of INH has
   remained elusive. Here, we report that INH and its metabolites induce a
   post-translational modification (PTM) of histones, lysine
   isonicotinylation (K[inic]), also called 4-picolinylation, in cells and
   mice. INH promotes the biosynthesis of isonicotinyl-CoA (Inic-CoA), a
   co-factor of intracellular isonicotinylation. Mass spectrometry reveals
   26 K[inic] sites in histones in HepG2 cells. Acetyltransferases
   CREB-binding protein (CBP) and P300 catalyse histone K[inic], while
   histone deacetylase HDAC3 functions as a deisonicotinylase. Notably,
   MNase sensitivity assay and RNA-seq analysis show that histone K[inic]
   relaxes chromatin structure and promotes gene transcription.
   INH-mediated histone K[inic] upregulates PIK3R1 gene expression and
   activates the PI3K/Akt/mTOR signalling pathway in liver cancer cells,
   linking INH to tumourigenicity in the liver. We demonstrate that
   K[inic]is a histone acylation mark with a pyridine ring, which may have
   broad biological effects. Therefore, INH-induced isonicotinylation
   potentially accounts for the side effects in patients taking INH
   long-term for anti-tuberculosis therapy, and this modification may
   increase the risk of cancer in humans.

   Subject terms: Post-translational modifications, Post-translational
   modifications
     __________________________________________________________________

   Isoniazid (INH) is an anti-tuberculosis drug; however the underlying
   cause of its various side effects are not understood. Here the authors
   identified lysine isonicotinylation as a histone modification on
   chromatin that is metabolically regulated by INH and its metabolites.

Introduction

   Epigenetic mechanisms, including DNA methylation and histone PTMs,
   possess biomarker potential for tumour classification and prognosis of
   targeted therapy. Histone PTMs in eukaryotic cells play a crucial role
   in malignant transformation, tumour development, and progression by
   regulating chromatin structure and transcriptional activity of genes.
   Short-chain fatty acids are the primary metabolites of bacteria and
   cell metabolism. In recent years, a variety of histone acylations have
   been identified, and these acylations appear to use their corresponding
   short-chain CoAs also short-chain fatty acids as co-factors and donors,
   respectively. Cell concentrations of short-chain CoAs and histone
   acylation levels are affected by short-chain fatty acids^[42]1–[43]5.

   The influence of food and drugs on histone PTMs has attracted attention
   in recent years. For example, aspirin acetylates lysine residues in
   vivo and in vitro and preservative sodium benzoate benzoylated
   histones^[44]6–[45]8. The above evidence suggests that compounds that
   we consume orally may also have important effects on epigenetic
   modifications of chromatin.

   Tuberculosis (TB) is a global issue, and INH was first reported to be
   effective for the treatment of TB in 1952^[46]9. INH remains the
   first-line drug approved for TB therapy due to its high efficacy. The
   standard therapies for TB include a combination treatment of INH,
   rifampicin, pyrazinamide, and ethambutol^[47]10, INH can also be used
   alone^[48]11. However, INH is also widely known for its hepatotoxicity
   and even liver failure risk^[49]12,[50]13. Several enzymes are involved
   in INH metabolism and INH metabolites, primarily acetyl-hydrazine
   (AcHz), hydrazine (Hz), and acetyl-isoniazid (AcINH), are thought to be
   responsible for the INH-induced liver injury^[51]14–[52]25. However,
   the molecular mechanisms underlying the side effects of INH remain
   unclear.

   In this report, we describe the identification and certification of
   K[inic] as an INH-induced, evolutionarily conserved PTM in cells. We
   observed that mice and cellular histone K[inic] levels were
   significantly increased after treatment with INH, which promoted the
   generation of Inic-CoA. Furthermore, this INH-induced modification is
   dynamically regulated by CBP and P300, two widely used
   acetyltransferases, and HDAC3, a Zn^2+-dependent deacetylase. Notably,
   our study shows that histone K[inic] relaxes chromatin structures and
   affects expression of hundreds of genes. Thus, our study identifies a
   type of histone acylation, which may lead to other investigations of
   pyridine-containing histones in the epigenetic regulation of gene
   expression.

Results

Identification and characterisation of histone isonicotinylation

   Recently, lysine benzoylation (K[bz]) has been identified as a new type
   of PTM. K[bz] is an abundant, evolutionarily conserved, and dynamic
   histone mark^[53]8. To identify K[bz] substrates, we performed a
   co-immunoprecipitation assay to enrich proteins from HepG2 cells using
   a pan anti-K[bz] antibody. Enriched proteins were treated with trypsin
   and peptides were subjected to HPLC-MS/MS analysis, followed by further
   analysis using PTMap, a software used for analysing the mass shift
   caused by PTMs^[54]26. Surprisingly, we observed a mass shift of
   +104.0268 Da caused by K[bz] at low abundance, but we also identified
   several modified histone peptides, such as KQLATK[+105.0215]AAR, which
   contained a mass shift of +105.0215 Da at the lysine residue position
   H3K23 (Supplementary Fig. [55]1a). The same mass shift of +105.0215 Da
   was also found in other histone lysine residues, including H3K122,
   H4K77, and H4K91, as well as non-histone lysine residues including
   HNRNPC K243 and ACTB K61 (Supplementary Fig. [56]1b–f). These findings
   suggest that this mass shift of +105.0215 is likely caused by a
   modification of histones and non-histones different from K[bz].
   Furthermore, these results indicate that the antibody against K[bz]
   could also be used to enrich peptides with structurally related
   acylations, as previously reported in cases where the acetylation
   antibody recognises propionylation and
   butylation^[57]2,[58]3,[59]27–[60]29 (Supplementary Fig. [61]1g). The
   most probable elemental composition formula for this modification group
   was C[6]H[3]NO, based on the accurately determined mass shift.
   According to the chemical formula, there are three possible lysine
   modifications that produce the isomers matching C[6]H[3]NO: lysine
   picolinylation (K[pic]), lysine nicotinylation (K[nic]), and lysine
   isonicotinylation (K[inic]; Fig. [62]1a). However, which one exists in
   cells remained unknown.

Fig. 1. Identification and characterization of histone isonicotinylation.

   [63]Fig. 1
   [64]Open in a new tab

   a Chemical structures of lysine acetylation, lysine benzoylation and
   three structural isomers modifications (lysine picolinylation, lysine
   nicotinylation and lysine isonicotinylation) that have the elemental
   compositions of C[6]H[3]NO and may induce a mass shift of +105.0215 Da.
   b Dot-blot assay was used to verify the specificity of pan antibody
   against K[inic] modification. Nitrocellulose membrane was spotted by
   different amounts of unmodified peptide (DAVTYTEHAKR), lysine
   nicotinylated peptide (DAVTYTEHAK[nic]R), lysine picolinylated peptide
   (DAVTYTEHAK[pic]R), lysine isonicotinylated peptide (DAVTYTEHAK[inic]R)
   and detected with the pan-K[inic] antibody. c Detection of histone
   K[inic] marks in HepG2 and HeLa cells. Core histones acid-extracted
   from HepG2 and HeLa cells was tested using pan-K[inic], pan-K[ac], and
   H3 antibodies by Western blot. Total histones were visualized with
   Coomassie blue staining. d Subcellular localization of K[inic] marks is
   mainly in nucleus. HepG2 and HeLa cells were stained with pan-K[inic]
   rabbit (green) and pan-K[ac] mouse (red) antibodies. Nuclei were
   stained with DAPI (blue), followed by visualization with confocal
   microscopy. Scale bar, 20 μm. e Detection of histone K[inic] marks in
   Sf-9, HepG2, HeLa, HEK293T, HCT-116, HT-29, NCI-H157, NCI-H1299 cells.
   Core histones acid-extracted from those cells were tested using
   pan-K[inic], pan-K[ac], and H3 antibodies by Western blot. Total
   histones were visualized with Coomassie blue staining. f Detection of
   K[inic] marks in normal tissues. Human normal thyroid, oesophageal
   epithelium, fundic gland, colon, and lung tissues were performed by
   immunohistochemical staining with pan-K[inic] antibody respectively,
   followed by visualization with light microscopy. Scale bar, 50 μm, and
   quantifications of the scores of immunohistochemical staining (upper,
   n = 3 samples, values were expressed as mean ± SEM). Detection of
   K[inic] marks in oesophageal carcinoma tissue. Scale bar, 100 μm
   (bottom).

   To determine which isomer causes the +105.0215 Da mass shift, we first
   determined whether the pan anti- K[bz] antibody recognises the above
   three isomers. We designed and synthesised four peptides based on the
   histone H4 amino acid sequence containing the lysine with
   a + 105.0215 Da mass shift: DAVTYTEHAKR, DAVTYTEHAK[pic]R,
   DAVTYTEHAK[nic]R, and DAVTYTEHAK[inic]R. Dot-blot analysis was used to
   examine the reactivity of the anti-K[bz] antibody with the four
   peptides. The results showed that the anti-K[bz] antibody recognised
   all three modified peptides in an in vitro experimental setup
   (Supplementary Fig. [65]2a), confirming that the anti-K[bz] antibody
   could also recognise lysine picolinylation, lysine nicotinylation, and
   lysine isonicotinylation in vitro. Intriguingly, when we analysed the
   structures of the three isomers, we found that the structure of K[inic]
   is similar to INH. Therefore, we wanted to characterise K[inic] based
   on histone modification. To identify this PTM on histones, a
   customer-designed pan anti-K[inic] antibody was generated, which showed
   a high specificity for K[inic] but not for K[pic] and K[nic], as shown
   by the Dot-blot assay (Fig. [66]1b). This antibody was able to detect
   endogenous K[inic] signals of core histones and non-histones using
   lysates from HepG2 and HeLa cells in a Western blot assay (Fig. [67]1c,
   Supplementary Fig. [68]2b). Additionally, immunofluorescent analysis
   using this antibody demonstrated that K[inic] was primarily distributed
   in the nuclei of living cells (Fig. [69]1d).

   Using this K[inic] antibody, we also detected histones and non-histones
   from insect cells Sf-9 (Spodoptera frugiperda 9), human HepG2, HeLa,
   HEK293T, HCT-116, HT-29, NCI-H157, NCI-H1299, mouse embryonic
   fibroblasts MEF cells, human HK-2, MCF-7, SUM159, SW1116, HEK293A, and
   MCF-10A cells. K[inic] signals were detected at various levels. These
   results indicate that K[inic] is a widely existed modification in
   eukaryotic cells (Fig. [70]1e, Supplementary Fig. [71]2c). Furthermore,
   we performed immunohistochemical staining in a variety of human normal
   tissues and found that in normal thyroid, colon, and lung tissues, the
   K[inic] levels were low; however, in normal oesophageal epithelium and
   fundic gland tissues, the K[inic] levels were high (Fig. [72]1f). We
   also noticed an interesting phenomenon in the normal oesophageal
   epithelium tissue: the isonicotinylation is higher in the
   well-differentiated outer layer than in the inner layer (Fig. [73]1f).

INH stimulates histone K[inic] in cells and mice

   Acylations are usually modulated by fatty acid
   metabolites^[74]1,[75]2,[76]4,[77]30; for example, acetate induces
   acetylation^[78]31. Therefore, we determined whether isonicotinic acid
   (INA) or sodium isonicotinate (SIN) could induce K[inic]. We applied a
   series of concentrations of INA or SIN to induce K[inic] in HepG2,
   HEK293A, and HCT-116 cells for 24 h. Western blot analysis showed that
   histone K[inic] levels were induced in a dose-dependent manner in these
   cells, and the same induction for K[inic] was also seen in non-histone
   proteins. In contrast, no significant change was observed in overall
   histone or non-histone acetylation (Fig. [79]2a–c, Supplementary
   Fig. [80]3a–j). These findings clearly indicate that K[inic] can be
   induced by both intracellular INA and SIN in vivo. Given that INA is a
   downstream metabolite of INH, we evaluated whether INH can also induce
   K[inic] in cells because they share the same structural backbone. To
   test this hypothesis, we detected histone and non-histone K[inic]
   levels in the above three cell lines after increasing the concentration
   of INH, and the results showed that both histone and non-histone
   K[inic]levels were obviously induced by INH (Fig. [81]2d, Supplementary
   Fig. [82]3k–o). These results were also confirmed by immunofluorescent
   staining, in which K[inic] levels in the cell nuclei substantially
   increased after treatment with INH, INA, and SIN, whereas no
   significant change was observed in lysine acetylation levels
   (Fig. [83]2e, Supplementary Fig. [84]3p). To examine whether INH could
   induce K[inic] in animals, we administered gavage feeding mice with
   50 mg/kg/day of INH or normal saline as a control for ten days. The
   regular dose of INH during TB therapy is 5 mg/kg/day in humans and the
   50 mg/kg/day INH dose in mice mimics the pharmacological dose in
   human^[85]13. Interestingly, we found that both histone and non-histone
   K[inic] levels were increased in mouse liver tissues after INH
   treatment (Fig. [86]2f). This result indicated that INH could induce
   K[inic] in mice and maybe also do so in humans who take INH for TB
   therapy.

Fig. 2. INH stimulates histone K[inic] in cells and mice.

   [87]Fig. 2
   [88]Open in a new tab

   a–e INA, SIN, and INH stimulate HepG2 histone K[inic]. a Chemical
   structures of isonicotinic acid (INA), sodium isonicotinate (SIN), and
   isoniazid (INH). b Core histones acid-extracted from HepG2 cells that
   treated with increasing concentration of INA for 24 h and tested using
   pan-K[inic], pan-K[ac], and H3 antibodies by Western blot analysis.
   Total histones were visualized with Coomassie blue staining. “INA”
   indicates isonicotinic acid. c Core histones acid-extracted from HepG2
   cells that treated with increasing concentration of SIN for 24 h and
   tested using pan-K[inic], pan-K[ac], and H3 antibodies by Western blot
   analysis. Total histones were visualized with Coomassie blue staining,
   “SIN” indicates sodium isonicotinate. d Core histones acid-extracted
   from HepG2 cells that treated with increasing concentration of INH for
   24 h and tested using pan-K[inic], pan-K[ac], and H3 antibodies by
   Western blot analysis. Total histones were visualized with Coomassie
   blue staining, “INH” indicates Isoniazid. e HepG2 cells were treated
   with 10 mM INA, SIN, and INH respectively for 24 h and stained with
   pan-K[inic] rabbit (green) and pan-K[ac] mouse (red) antibodies. Nuclei
   were stained with DAPI (blue), followed by visualization with confocal
   microscopy. Scale bar, 20 μm. f INH stimulates mice liver K[inic]
   levels. Liver samples from mice treated with normal saline (NS) or
   50 mg/kg/day of INH for ten days were collected (n = 10 mice each
   group). And core histones and non-histones were extracted and tested
   using pan-K[inic], GAPDH and H3 antibodies by Western blot. Total
   histones were visualized with Coomassie blue staining. “NS” indicates
   normal saline treated mice, “INH” indicates isoniazid treated mice. g
   INH is like to stimulate histone K[inic] directly. Schematic diagram of
   INH metabolism in vivo (left), core histones acid-extracted from normal
   HepG2 cells, 1 mM ABT (1-Aminobenzotriazole), 5 mM APAP (Acetaminophen)
   or 15 mM BNPP (Bis-p-nitrophenyl phosphate) pretreated HepG2 cells
   respectively, and also treated with 10 mM INH for 24 h, then tested
   using pan-K[inic], pan-K[ac] and H3 antibodies by Western blot
   analysis. Total histones were visualized with Coomassie blue staining
   (right). h–i INH stimulates the production of Inic-CoA. h CoA was
   extracted from HepG2 cells treated with or without D4-INH (PH = 7.4)
   for 24 h, then tested the Ac-CoA (left) and D4-Inic-CoA (right) levels
   using HPLC-MS/MS assay, “CON” indicates untreated HepG2 cells; “D4-INH”
   indicates D4-isoniazid treated HepG2 cells, quantifications of the
   chromatographic peak area, data presented are the mean ± SEM from three
   biological replicates (n = 3), as determined by unpaired two-tailed
   Student’s t-test. i CoA was extracted from HepG2 cells treated with
   increasing concentration of INH (PH = 7.4) for 24 h, and tested the
   Inic-CoA levels using HPLC-MS/MS assay, quantifications of the
   chromatographic peak area, data presented are the mean ± SEM from three
   biological replicates (n = 3), as determined by one-way ANOVA followed
   by Bonferrroni’s multiple comparisons test.

   It is known that if two peptides have the same MS/MS fragmentation
   patterns and approximately identical retention times, they are
   considered to have an identical structure. Therefore, to scrutinise
   whether the INH-induced PTM in vivo is the genuine K[inic], we analysed
   in vivo peptide and the in vitro synthetic peptide DAVTYTEHAK[inic]R by
   HPLC-MS/MS. Results showed that they have the same MS/MS fragmentation
   patterns and almost identical retention times (21.50 min vs. 21.49 min;
   Supplementary Fig. [89]3q–r). Therefore, we concluded that there are
   peptides DAVTYTEHAK[inic]R in vivo with K[inic], indicating that the
   INH-induced PTM in cells is K[inic].

   Next, we investigated whether INH-induced K[inic] is caused by INH
   directly or by its metabolite, INA. It was known that INH can be
   converted into INA by two enzymes, N-acetyltransferase 2 (NAT2) and
   isoniazid hydrolase^[90]32. As shown in the left panel of Fig. [91]2g,
   we separately blocked intracellular INH metabolism through NAT2
   inhibitors 1-aminobenzotriazole (ABT) and acetaminophen (APAP), two
   newly identified NAT2 inhibitors^[92]33,[93]34, and isoniazid hydrolase
   inhibitor bis-p-nitrophenyl phosphate (BNPP)^[94]35. Results showed
   that when INH metabolism was separately blocked, the intracellular
   levels of histone K[inic] were not affected (Fig. [95]2g, right panel),
   indicating that INH likely induces histone K[inic] directly, without
   the requirement of INH turnover into INA. Furthermore, we found that
   histone K[inic] levels reached a steady state after INH treatment for
   12 h, while there was no change in histone acetylation levels
   (Supplementary Fig. [96]3s).

INH is a donor for generating intracellular isonicotinyl-CoA

   Accumulating evidence has shown that acylation donors stimulate the
   production of the corresponding cellular acyl-CoAs, resulting in
   elevated levels of histone
   acylations^[97]3,[98]5,[99]8,[100]36,[101]37. Although there is no data
   on Inic-CoA, it is the most likely co-factor for the K[inic] reaction
   in cells, similar to acetyl-CoA for the lysine acetylation reaction. To
   determine whether the Inic-CoA exists intracellularly, we chemically
   synthesised Inic-CoA as a control, based on the chemical formula shown
   in Supplementary Fig. [102]3t. We metabolically labelled HepG2 cells
   with 5 mM isotopic D4-INH for 24 h, followed by mass spectrometry-based
   quantification. The results showed that there was an obviously increase
   in D4-Inic-CoA upon D4-INH treatment (Fig. [103]2h, right panel), while
   no change was observed for Ac-CoA as a control (Fig. [104]2h, left
   panel). Moreover, after treatment with increasing concentrations of
   INH, the HPLC-MS/MS measurements showed an elevated cellular
   concentration of Inic-CoA, which was dose-dependent (Fig. [105]2i).
   Overall, we showed the existence of Inic-CoA in living cells and
   demonstrated the histone mark, K[inic,] with the pyridine ring, which
   is initiated by INH-stimulated production of Inic-CoA.

Identification of K[inic] sites in histones of HepG2 cells

   It is important to identify all K[inic] sites in histones and compare
   them with the other histone marks. To this end, HepG2 cells were
   treated with 10 mM INH for 24 h, and the histones were purified and
   subjected to trypsin digestion. Peptides were enriched by the pan
   anti-K[inic] antibody and analysed by HPLC-MS/MS analysis, followed by
   a protein sequence database search. Results showed that a total of 26
   unique histone K[inic] sites were identified (Fig. [106]3). Notably,
   the histone K[inic] sites in HepG2 cells are widely distributed,
   resembling the distribution of histone K[ac]. In contrast, histone
   K[bz]sites are primarily located at the N-terminal tails of histones,
   according to a previous report^[107]8. These findings indicate that
   K[inic], as a histone mark, may play a significant role in the
   regulation of gene expression alone or in cooperation with other
   histone modifications.

Fig. 3. Identification of K[inic] sites in histones of HepG2 cells.

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

   The detected K[inic] sites in HepG2 cells treated with 10 mM INH for
   24 h using HPLC-MS/MS assay are shown in red. And histone K[ac] and
   K[bz] sites that were reported previously are shown for comparison.
   “K[ac]” indicates lysine acetylation, “K[bz]” indicates lysine
   benzoylation, “K[inic]” indicates lysine isonicotinylation.

CBP/P300 function as isonicotinyltransferases and HDAC3 as deisonicotinylase
for histone K[inic]

   Given that we identified the donor and co-factor for the generation of
   histone isonicotinylation, we attempted to find the transferases that
   transfer the isonicotinyl group to lysine on histones. Based on
   previous studies on histone acylation modifications, the histone
   acylations and deacylations are regulated commonly by histone
   acetyltransferases (HATs) and deacetylases (HDACs) in vivo. We
   transfected HepG2 cells with four HAT plasmids, including CBP, P300,
   PCAF, and hMOF, and found that the histones were obviously
   isonicotinylated by CBP and P300 but weakly isonicotinylated by PCAF
   and hMOF (Fig. [110]4a). When HepG2 and HeLa cells were transfected
   with CBP and P300, histone isonicotinylation and acetylation levels
   increased (Fig. [111]4b, Supplementary Fig. [112]4a). Moreover, when
   endogenous CBP and P300 were knocked down by siRNA or inhibited by
   A-485, a newly reported inhibitor of CBP/P300^[113]38, and SGC-CBP30, a
   widely used inhibitor of CBP/P300^[114]39, the levels of histone
   isonicotinylation and acetylation were decreased, indicating that CBP
   and P300 are responsible for the “writing” during histone
   isonicotinylation (Supplementary Fig. [115]4b–c). In addition, an in
   vitro isonicotinylation reaction experiment was performed to confirm
   the direct role of CBP/P300 in isonicotinylating histones in the
   presence of synthetic Inic-CoA and purified CBP/P300. The results
   showed that CBP/P300 potently isonicotinylated histone K[inic] in vitro
   (Fig. [116]4c).

Fig. 4. CBP/P300 function as isonicotinyltransferases and HADC3 as
deisonicotinylase for histone K[inic].

   [117]Fig. 4
   [118]Open in a new tab

   a–c CBP/P300 catalyze histone K[inic] in vivo and in vitro. a HepG2
   cells were transfected with 3×HA, HA-CBP, HA-P300, HA-hMOF and HA-PCAF
   plasmids respectively, core histones acid-extracted and tested using
   pan-K[inic], pan-K[ac] and H3 antibodies. Total histones were
   visualized with Coomassie blue staining. And total proteins were
   extracted and tested using HA-tag and β-actin antibodies by Western
   blot analysis. b HepG2 cells were transfected with 3×HA, HA-CBP, or
   HA-P300 and stained with pan-K[ac] or pan- K[inic] rabbit (red) and
   HA-tag mouse (green) antibodies. Nuclei were stained with DAPI (blue),
   followed by visualization with confocal microscopy. Scale bar, 20 μm. c
   HA-CBP and HA-P300 plasmids were transfected into HEK293T cells, then
   the protein was purified with HA beads and subjected to in vitro
   isoniconitylation assay with or without Inic-CoA using free core
   histones from HepG2 cells as substrates. Tested using pan-K[inic],
   pan-K[ac], and H3 antibodies. Total histones were visualized with
   Coomassie blue staining. And total proteins were extracted and tested
   using HA-tag and β-actin antibodies by Western blot analysis. d–f HDAC3
   removes histone K[inic.] d Core histones acid-extracted from HepG2
   cells that treated with 3 μM TSA (Trichostatin A) or 5 mM NAM
   (Nicotinamide) for 12 h and tested using pan-K[inic], pan-K[ac], and H3
   antibodies by Western blot. Total histones were visualized with
   Coomassie blue staining. e HepG2 cells were transfected with 3×Flag or
   Flag-HDAC3 plasmid and stained with pan-K[ac] or pan-K[inic] rabbit
   (green) and Flag-tag mouse (red) antibodies. Nuclei were stained with
   DAPI (blue), followed by visualization with confocal microscopy. Scale
   bar, 20 μm. f HepG2 cells were transfected with 3×HA, HA-HDAC3
   plasmids, core histones acid-extracted and tested using pan-K[inic],
   pan-K[ac], and H3 antibodies. Total histones were visualized with
   Coomassie blue staining. Total proteins were extracted and tested using
   HA-tag, HDAC3, and β-actin antibodies by Western blot analysis.

   Next, HepG2 cells were treated with trichostatin A (TSA), a broad
   inhibitor of HDAC family deacetylases, or nicotinamide (NAM), an
   inhibitor of SIRT family deacetylases. We found that histone
   isonicotinylation levels increased after TSA treatment, while no change
   was observed after NAM treatment, indicating that HDAC family members
   influence histone isonicotinylation (Fig. [119]4d). HepG2 cells were
   transfected with a variety of expression vectors encoding HDAC family
   members, including HDAC1-9, and when HepG2 cells were transfected with
   only HDAC3, the isonicotinylation level obviously decreased, as
   measured by immunofluorescent staining, but not with other HDAC family
   members (Fig. [120]4e, Supplementary Fig. [121]4d). Furthermore, the
   Western blot results showed that HDAC3 markedly deisonicotinylates
   histones among the HDAC family members (Fig. [122]4f, Supplementary
   Fig. [123]4e). When endogenous HDAC3 was knocked down by siRNA, the
   levels of histone isonicotinylation and acetylation were increased
   (Supplementary Fig. [124]4f). Based on the above results, we believe
   that HDAC3 is a histone deisonicotinylase. Collectively, these results
   indicate that CBP and P300 act as the histone isonicotinyltransferases,
   and deacetylase HDAC3 acts as histone deisonicotinylase.

Histone K[inic] leads to higher chromatin accessibility and promotes gene
transcription

   It is known that histone acylations can influence nucleosome
   configuration and chromatin accessibility; therefore, we investigated
   the role of isonicotinylation in the regulation of chromatin
   accessibility using an MNase sensitivity assay. The results showed that
   INH-induced histone isonicotinylation leads to higher chromatin
   accessibility by displaying nucleosomal DNA sensitivity to MNase
   treatment in HepG2 cells (Fig. [125]5a). In addition, Sat-2 is a marker
   for alteration of the heterochromatin structure, and its transcription
   is generally repressed. However, when the centromeric region is
   relaxed, Sat-2 gene transcription is increased^[126]40,[127]41. Sat-2
   gene expression during INH induction was determined. The results showed
   that Sat-2 gene expression was upregulated as determined by real-time
   RT-PCR assay in HepG2 cells, indicating that INH-induced histone
   isonicotinylation also leads to heterochromatin relaxation in HepG2
   cells (Fig. [128]5b). Furthermore, to determine which genes are
   potentially regulated by histone K[inic] and which physiological
   functions may be linked to histone K[inic], we performed RNA-seq for
   total RNAs from INH-treated HepG2 cells. In this analysis, 313 genes
   were found to be upregulated, whereas 121 genes were downregulated
   (Fig. [129]5c). The subsequent KEGG pathway enrichment analysis showed
   that the genes upregulated by histone isonicotinylation were primarily
   correlated with the TNF signalling pathway, hematopoietic cell lineage,
   complement and coagulation cascades, GABAergic synapse, and cAMP
   signalling pathway (Fig. [130]5d). Among these pathways, the TNF
   signalling pathway is closely related to inflammation and cell death.
   Next, we chose five genes to validate the correlation between histone
   isonicotinylation and gene expression using real-time RT-PCR assays:
   CR2, CAMK, PLAU, IL1B, and CACNA1D. Real-time RT-PCR results showed
   that the mRNA expression of these selected genes was upregulated by
   1.50- to 8.49-fold (CR2: 8.49-fold, CAMK: 8.08-fold, PLAU: 7.37-fold,
   IL1B: 1.50-fold, CACNA1D: 2.50-fold; Fig. [131]5e). These findings
   indicate that histone isonicotinylation, like other acylations,
   promotes gene transcription by loosening the chromatin structure in the
   genome.

Fig. 5. Histone K[inic] leads to higher chromatin accessibility and promotes
gene transcription.

   [132]Fig. 5
   [133]Open in a new tab

   a, b Histone K[inic] induces chromatin accessibility. a HepG2 cells
   were treated with or without 10 mM INH for 6 h, nucleosomal DNA was
   extracted and analyzed by MNase sensitivity assay (left). “CON”
   indicates untreated HepG2 cells; “INH” indicates isoniazid treated
   HepG2 cells, MW: Molecular Weight. Quantifications of lane signal
   intensity (right, upper bands), data presented are the mean ± SEM from
   three biological replicates (n = 3), as determined by unpaired
   two-tailed Student’s t-test b Relative mRNA expression of Sat-2 in
   HepG2 cells treated with or without 10 mM INH for 12 h and determined
   by real-time RT-PCR assay, “CON” indicates untreated HepG2 cells; “INH”
   indicates isoniazid treated HepG2 cells, data presented are the
   mean ± SEM from three biological replicates (n = 3), as determined by
   unpaired two-tailed Student’s t-test. c Histone K[inic] upregulates 313
   genes and downregulates 121 genes. Volcano plot analysis of pairwise
   comparison of RNA-seq results from HepG2 cells with or without 10 mM
   INH for 12 h (n = 3 samples, values were expressed as mean ± SEM). d
   KEGG pathway analysis of mRNA levels elevated genes. Some genes
   involved in each pathway are labelled. e Real-time RT-PCR analysis of
   the genes labelled in KEGG pathway analysis. Relative expression is
   normalized to GAPDH, “CON” indicates untreated HepG2 cells; “INH”
   indicates isoniazid treated HepG2 cells, data presented are the
   mean ± SEM from three biological replicates (n = 3), as determined by
   unpaired two-tailed Student’s t-test.

INH-induced histone isonicotinylation varies in tumours and upregulated
cancer-related signalling pathway

   INH is a very effective first-line anti-TB drug. However, it is known
   that patients who take INH to cure TB suffered side effects. Thus, many
   investigations were carried out in the 1970s with a focus on the
   toxicity of INH^[134]42. INH is known to induce liver cancer in
   rats^[135]43, but it is unknown whether INH can cause cancer in humans.
   To examine whether INH-induced isonicotinylation is correlated with
   cancer in humans, we collected ten different kinds of tumour tissues
   and normal tissues adjacent to tumours and used immunohistochemical
   staining to determine the levels of isonicotinylation in these tissues.
   Results showed that the total K[inic] level was lower in gastric
   carcinoma tissues than in normal tissues, while it was higher in lung
   adenocarcinoma and thyroid carcinoma tissues than in normal tissues.
   However, the total K[inic] level in renal carcinoma, rectal carcinoma,
   colonic adenocarcinoma, mammary adenocarcinoma, pancreatic carcinoma,
   oesophageal carcinoma, and hepatocellular carcinoma tissues were not
   significantly different from those in normal tissues. (Fig. [136]6a, b,
   Supplementary Fig. [137]5a). These findings suggest that
   isonicotinylation levels may be linked to certain types of cancer.
   Therefore, isonicotinylation of a particular protein, either histone or
   non-histone proteins, might be correlated with cancer development.

Fig. 6. INH-induced histone isonicotinylation varies in tumours and
upregulated cancer-related signalling pathway.

   [138]Fig. 6
   [139]Open in a new tab

   a, b Detection of K[inic] mark in normal adjacent tumour tissues and
   tumour tissues. a Normal adjacent tumour tissue and tumour tissue of
   gastric carcinoma were performed by immunohistochemical staining with
   pan-K[inic] antibody, followed by visualization with light microscopy.
   Scale bar, 50 μm (left). Quantifications of the scores of
   immunohistochemical staining (right) (n = 6 samples, values were
   expressed as mean ± SEM, as determined by paired two-tailed Student’s
   t-test). b Normal adjacent tumour tissues and tumour tissues of lung
   adenocarcinoma and thyroid carcinoma were performed by
   immunohistochemical staining with pan-K[inic] antibody respectively,
   followed by visualization with light microscopy. Scale bar, 50 μm.
   Quantifications of the scores of immunohistochemical staining (right)
   (n = 6 samples, values were expressed as mean ± SEM, as determined by
   paired two-tailed Student’s t-test). c Relative mRNA expression of
   PIK3R1 in HepG2 cells treated with or without 10 mM INH for 12 h and
   determined by real-time RT-PCR assay. Relative expression is normalized
   to GAPDH, “CON” indicates untreated HepG2 cells; “INH” indicates
   isoniazid treated HepG2 cells, data presented are the mean ± SEM from
   three biological replicates (n = 3), as determined by unpaired
   two-tailed Student’s t-test. d HepG2 cells were treated with or without
   10 mM INH for 18 h, core histones acid-extracted and tested using
   pan-K[inic], and H3 antibodies. Total histones were visualized with
   Coomassie blue staining. And total proteins were extracted and tested
   using PIK3R1, Akt, p-Akt, mTOR, p-mTOR, and β-actin antibodies by
   Western blot analysis. e Heatmap displaying differentially expressed
   genes involved in PI3K/Akt/mTOR pathway in response to INH in RNA-seq
   data. “CON” indicates untreated HepG2 cells; “INH” indicates isoniazid
   treated HepG2 cells; p < 0.05, log FC > 1. f A graphical model of
   histone isonicotinylation.

   Intriguingly, in HepG2 cells treated with INH, RNA-seq analysis
   revealed that the PIK3R1 gene was upregulated. This gene encodes the
   protein PIK3R1 (also called P85α), which is the regulatory subunit of
   PI3K and was previously reported to promote HCC progression via the
   PI3K/Akt/mTOR pathway^[140]44–[141]46. We found that the mRNA and
   protein levels of PIK3R1 were greatly increased when histones were
   isonicotinylated by INH stimulation. Notably, the levels of p-Akt and
   p-mTOR also increased, indicating that the mTOR signalling pathway was
   activated by histone K[inic] (Fig. [142]6c, d). In addition, we
   enriched PI3K/Akt/mTOR signalling pathway enrichment based on the
   results of RNA-seq analysis and found that histone isonicotinylation
   promotes the expression of multiple genes of the PI3K/Akt/mTOR pathway,
   including FLT3LG, IL7, PIK3R, and CHRM1 (Fig. [143]6e, Supplementary
   Fig. [144]5b).

   This finding demonstrated that INH-induced histone isonicotinylation
   activates the PI3K/Akt/mTOR signalling pathway, which maybe the
   evidence of histone isonicotinylation linked to cancer development.
   Subsequently, we randomly selected ten paired HCC patient samples and
   detected the profile of total isonicotinylation in tumours and adjacent
   normal tissues. The results showed that two of the ten patients
   (patients #3 and #6) displayed higher levels of isonicotinylation in
   tumour tissues than in the adjacent normal tissues, and the levels of
   PIK3R1 were increased concomitantly in tumour tissues of the same
   patients (patients #3 and #6; Supplementary Fig. [145]5c).

Discussion

   In the present study, we evaluate the histone mark K[inic], a type of
   acylation. This histone modification is stimulated by INH, the
   first-line anti-tuberculosis drug discovered 70 years ago that is still
   effective and widely used clinically. INH and its metabolites, INA and
   SIN, are able to induce K[inic] in cells. However, INH is likely to
   directly induce K[inic] without the need to metabolise to INA or SIN.
   Many studies have shown that BNPP, a reversible carboxyesterase
   inhibitor, can also reduce APAP deacetylation, but the functional
   changes caused are unknown^[146]47,[147]48. As for ABT, as a recently
   reported inhibitor of NAT2, although there is no direct evidence to
   prove the effect of BNPP on ABT, BNPP is an effective carboxyesterase
   inhibitor and deacetylation inhibitor. It is also very likely to affect
   the function of ABT, similar to APAP. Therefore, we did not do the
   experiment of adding APAP/ABT and BNPP together in the cells. So far,
   K[inic] is the only identified histone PTM with a pyridine acyl ring
   structure exposed to lysine. Therefore, we revealed the importance of
   histone K[inic] in the epigenetic regulation of chromatin structure and
   gene expression. Figure [148]6f illustrates a working model for the
   role of K[inic] in histone modification. In this model, INH is like to
   directly induces isonicotinylation of histone and non-histone proteins
   by generating Inic-CoA. CBP and P300 function as “writers” to add
   isonicotinyl group on lysines. In contrast, HDAC3 removes the
   isonicotinyl group as an “eraser.” Once lysines on histones are
   isonicotinylated, the binding ability between histones and genomic DNA
   is weakened, thereby loosening the chromatin structure and promoting
   the transcription of hundreds of genes. Among them, PIK3R1 gene
   expression is upregulated by K[inic], which leads to increased PIK3R1
   protein levels and activates the PI3K/Akt/mTOR pathway related to HCC.

   Histone acetylation with a short chain linked to lysine exhibits
   definite biological functions. The histone K[inic] contains a more
   complicated side chain linked to a lysine, and K[inic] is structurally
   similar to K[bz], the latter with a -CH in the benzene ring being
   replaced by an N atom in the pyridine ring. Despite their structural
   similarity, K[inic] and K[bz] as histone marks are distinct in many
   aspects. K[inic] and K[bz] are generated via different metabolic
   pathways. First, K[inic] has a pyridine acyl group, and K[bz] has an
   aromatic acyl group, on the lysine sidechain, and the pyridine ring and
   benzene ring are unique in chemical reactivities and accessibility. For
   example, the hydrogenation of the pyridine ring is easier than that of
   the benzene ring. Second, K[bz] is stimulated by sodium benzoate, the
   primary component of preservatives, by increasing the concentration of
   benzoyl-CoA; however, K[inic] is induced by INH, an anti-tuberculosis
   drug, which promotes the generation of Inic-CoA. Third, lysine residues
   in histones modified by K[bz] primarily exist at the N-terminal tails,
   whereas the lysine residues in histones modified by K[inic] are widely
   distributed on histones from the N- to C-terminals. The discrepancy in
   the distribution of lysine modification sites may reflect their
   functional distinction in the epigenetic regulation of chromatin and
   gene expression. Fourth, histone K[inic] is dynamically regulated by
   the acetyltransferases CBP and P300 and the deacetylase HDAC3, whereas
   histone K[bz] is removed by deacetylase SIRT2. So far, the
   acyltransferase for generating K[bz] is unknown, or perhaps the
   acyltransferase is not required for the generation of K[bz]
   modification on histones. Finally, histone K[inic] is found to regulate
   the expression of specific genes associated with TNF, GABAergic
   synapse, and cAMP signalling pathways, which differ significantly from
   K[bz]-regulated signalling pathways, such as the phospholipase D
   signalling pathway^[149]8. Hence, we anticipate that histone K[inic]
   may regulate more extensive biological functions than K[bz].

   In this report, we determined that K[inic] exists endogenously in
   living cells, in both histone and non-histone proteins. Notably, we
   also proved the existence of Inic-CoA in cells, and it serves as a
   co-factor of CBP or P300 to generate K[inic] intracellularly. Exogenous
   INH promotes endogenous K[inic] by promoting the generation of
   Inic-CoA; thus, a clear metabolic map linking K[inic] with INH has been
   established. INH is widely used for the treatment of TB. The metabolism
   of INH and its role in INH-induced liver injury have been extensively
   studied, but the underlying mechanism remains controversial. INH itself
   is thought to be directly bioactivated to a reactive metabolite, which,
   in some patients, leads to an immune response and liver injury^[150]49.
   NAT2 is now known to be the primary enzyme that metabolises INH, and it
   has been shown that NAT2 deficiency increases the risk of liver injury
   induced by INH^[151]50. However, the detailed mechanisms remain
   unknown. In our study, when the enzymatic activity of NAT2 was
   inhibited, the levels of histone K[inic] increased (Fig. [152]2g).
   Intriguingly, histone K[inic] promotes the activation of the
   PI3K/Akt/mTOR signalling pathway by upregulating the PI3K p85 subunit.
   Oral feeding of INH stimulates elevated levels of isonicotinylation in
   the liver tissue of mice (Fig. [153]2f), and INH-stimulated histone
   K[inic] activates the PI3K/Akt/mTOR signalling pathway, regulating a
   variety of biological functions that contribute to the side effects of
   INH. Therefore, this notion is a reasonable explanation for the
   induction of hepatotoxicity and tumourigenicity in mice. The mechanism
   is also reminiscent of INH enhancing cancer risk in humans, which needs
   to be examined in the future for TB patients under INH treatment.

   INH contains a pyridine ring structure, and K[inic] is a PTM that could
   bring in a pyridine structure in histone and non-histone proteins.
   Pyridine is widely used in the chemical and pharmaceutical industries
   and is also a component of some vitamins, enzymes, and various drugs,
   including INH. The pyridine ring-containing K[inic] histones may endow
   unique biological functions and play important epigenetic regulatory
   roles. The readers of K[inic] may be quite different from those of
   K[ac]and K[bz]. Therefore, the biological functions regulated by
   K[inic]for histones and non-histone proteins should be investigated
   further.

   K[inic] identification and characterisation represent a type of PTM on
   histone and non-histone proteins. However, this also raises many
   interesting questions. For example, where does the endogenous K[inic]
   come from? How is intracellular Inic-CoA synthesised? Are
   isonicotinylated non-histone proteins widely present in cells? Who are
   readers of K[inic]? What can K[inic] do for cells? Does high K[inic]
   cause cancer? Does K[inic] influence other PTMs on histones, for
   example, acetylation and methylation? These important questions warrant
   further investigation. In addition, the role of site-specific K[inic]
   on histones in the epigenetic regulation of gene transcription should
   be investigated thoroughly, and site-specific K[inic] antibodies need
   to be produced in mechanistic studies.

   In summary, we demonstrated that K[inic] is a histone mark on chromatin
   that is metabolically regulated by INH and its metabolites. Due to the
   wide use of pyridine-containing chemicals in the modern world, we
   anticipate that histone isonicotinylation may be involved in various
   biological effects and diseases related to public health. This
   investigation may also shed light on the complicated histone code and
   epigenetic control of gene expression in combination with the
   metabolism of therapeutic drugs.

Methods

Chemicals and antibodies

   Unless otherwise noted, all chemical reagents were purchased from
   Sigma-Aldrich. Cocktail (78443) and phosphatase inhibitor cocktail
   (78428) were purchased from Thermo Fisher Scientific. A485 (HY-107455),
   SGC-CBP30 (HY-15826), Acetaminophen (APAP, HY-66005),
   1-Aminobenzotriazole (ABT, HY-103389), Trichostatin A (TSA, HY-15144)
   and Nicotinamide (NAM, HY-B0150) were purchased from MedChemExpress.
   Bis-p-nitrophenyl phosphate (BNPP, N3002) was purchased from
   Sigma-Aldrich. D4-INH was purchased from Shanghai ZZBIO Co., Ltd.,
   isonicotinyl-CoA was made by Shanghai Nafu Biotechnology Co., Ltd.,
   Rabbit-pan-K[bz] (PTM-762) was purchased from PTM Biolabs: WB,1:1000.
   Mouse-pan-K[ac] (sc-32268) were purchased from Santa Cruz, pan-K[ac]
   (9441 S): IF,1:200. Rabbit-total histone H3 (9715 S), Rabbit-CBP
   (7389), Rabbit-P300 (86377), Rabbit-mTOR (2972 S), Rabbit-p-mTOR
   (Ser2448) (2971 S), Rabbit-Akt (9272 S), Rabbit-p-Akt (Ser473) (9271 T)
   antibodies were purchased from Cell Signaling Technology: WB,1:1000.
   Rabbit-anti-pan-K[inic] antibody was customer designed and prepared by
   PTM Biolabs Inc. (China): WB,1:1000; IF,1:200. Mouse-anti-Flag (F1804)
   and Rabbit-anti-HA (H3663) antibodies were purchased from
   Sigma-Aldrich: WB,1:2000. IF,1:400. Mouse-β-actin (66009-1-lg) and
   Mouse-GAPDH antibodies (60004-1-lg) were purchased from Proteintech:
   WB,1:3000. Mouse-anti-PIK3R1 antibody (EM1701-62) was purchased from
   Huabio (China): WB,1:1000. Rabbit-HDAC3 antibody ([154]T55595) was
   purchased from Abmart (China): WB,1:1000. 2-step plus Poly-HRP
   anti-rabbit IgG detection system (PV6001) was purchased from Zhong Shan
   Jin Qiao, China. HiScript II Q Select RT SuperMix Kit (R223) (+gDNA
   wiper) and ChamQ SYBR qPCR Master Mix (Q311) were purchased from
   Vazyme, China.

Dot blot analysis

   Synthetic peptides of different qualities were spotted on
   nitrocellulose membranes. After incubation with 5% milk for 1 h, the
   membranes were incubated with the first antibodies at 4 °C for 4 h,
   then washed three times with TBST (10 mM Tris-HCl, 150 mM NaCl, 0.1%
   Tween, PH = 7.5), and incubated with the secondary antibodies
   (anti-mouse or anti-rabbit) at room temperature for 1 h and washed
   three times with TBST. Then membranes were developed using enhanced
   chemiluminescence detection system according to protocol from the
   manufacturer. SageCapture Imaging Software (v2.19.12.20190311) was used
   for scanning.

Cell culture

   Sf-9 cells were cultured in SIM-SF without FBS. HepG2, HeLa, HEK293T,
   HCT-116, HT-29, NCI-H157, NCI-H1299, MEF, HK-2, MCF-7, SUM159, SW1116,
   HEK293A, MCF-10A cells were cultured in DMEM with 10% FBS, 100 units/ml
   penicillin and 100 mg/ml streptomycin were added, and incubated at
   37 °C with 5% CO[2].

HPLC-MS/MS analysis and database search

   The tryptic peptides were dissolved in 0.1% formic acid (solvent A),
   directly loaded onto a home-made reversed-phase analytical column
   (15-cm length, 75 μm i.d.). The gradient was comprised of an increase
   from 6% to 23% solvent B (0.1% formic acid in 98% acetonitrile) over
   26 min, 23% to 35% in 8 min and climbing to 80% in 3 min then holding
   at 80% for the last 3 min, all at a constant flow rate of 400 nL/min on
   an EASY-nLC 1000 UPLC system. The peptides were subjected to NSI source
   followed by tandem mass spectrometry (MS/MS) in Q ExactiveTM Plus
   (Thermo) coupled online to the UPLC. The electrospray voltage applied
   was 2.0 kV. The m/z scan range was 350 to 1800 for full scan, and
   intact peptides were detected in the Orbitrap at a resolution of
   70,000. Peptides were then selected for MS/MS using NCE setting as 28
   and the fragments were detected in the Orbitrap at a resolution of
   17,500. A data-dependent procedure that alternated between one MS scan
   followed by 20 MS/MS scans with 15.0 s dynamic exclusion. Automatic
   gain control (AGC) was set at 5E4. Fixed first mass was set as 100 m/z.
   The resulting MS/MS data were processed using Maxquant search engine
   (v.1.5.2.8). Tandem mass spectra were searched against SwissProt Human
   database concatenated with reverse decoy database. Trypsin/P was
   specified as cleavage enzyme allowing up to 4 missing cleavages. The
   mass tolerance for precursor ions was set as 20 ppm in First search and
   5 ppm in Main search, and the mass tolerance for fragment ions was set
   as 0.02 Da.

Acid extraction of histones

   Cell pellets were lysed in hypotonic buffer (10 mM Tris–Cl pH 8.0, 1 mM
   KCl, 1.5 mM MgCl[2], and 1 mM DTT, add protease and phosphatase
   inhibitors before use), then resuspended in 0.4 N sulfuric acid, and
   the supernatants were precipitated by trichloroacetic acid. After
   centrifuging the histone pellets were washed with cold acetone twice
   and dissolved in ddH[2]O.

Extraction assay of total cell proteins

   Cells were collected and lysed in NP-40 buffer (50 mM Tris-HCl, 150 mM
   NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA)
   supplemented with protease inhibitor cocktail and phosphatase inhibitor
   cocktail. A Model 680 microplate reader was used for absorbance
   detected in a 96 well plates for BCA protein quantification.

Western blot analysis

   Protein extracts were separated by SDS-PAGE gel electrophoresis and
   transferred to PVDF membranes. After incubation with 5% milk, the
   membranes were incubated with the first antibodies at 4 °C overnight.
   The next day membranes were washed three times with TBST, then
   incubated with the secondary antibodies (anti-mouse or anti-rabbit) at
   room temperature for 1 h and washed three times with TBST. Then
   membranes were developed using enhanced chemiluminescence detection
   system according to protocol from the manufacturer. SageCapture Imaging
   Software (v2.19.12.20190311) was used for SDS-PAGE scanning.

Immunofluorescent staining

   HepG2 and HeLa cells were seeded on round coverslip before experiments.
   The cells were washed three times with phosphate buffered saline (PBS,
   137 mM NaCl, 2.7 mM KCl, 10 mM Na[2]HPO[4], and 2 mM KH[2]PO[4]) and
   fixed in 4% paraformaldehyde at room temperature for 15 min. After
   washing with PBS three times (10 min for each time), the coverslips
   were incubated with 0.1% Triton X-100 at room temperature for 15 min,
   then blocked with 5% BSA at room temperature for 2 h, after washing
   with PBS three times the cells were incubated with first antibodies
   mixture at 4 °C overnight in wet box. The next day coverslips were
   washed three times with PBS, followed by incubation with secondary
   antibodies at room temperature for 2 h. Then, the coverslips were
   counterstained with 4,6-diamidino-2-phenylindole and mounted onto glass
   slides with nail polish. A LSM-780 confocal laser-scanning microscope
   (with Zeiss confocal Zen Software (v2011) using a100´/1.4 NA objective
   lens) was used to capture images from immuno-stained cells.

Immunohistochemical staining

   The paraffin-embedded tissue sections were dewaxed in xylene and
   gradually rehydrated. Then the sections were incubated in 3% H[2]O[2]
   for 30 min to quench endogenous peroxidase. To restore the antigen,
   slice in 10 mM sodium citrate buffer (pH = 6.0) for 20 min in 95 °C
   water. When incubated with pan-K[inic] antibody overnight at 4 °C, wash
   with PBS three times (10 min for each time), then the 2-step plus
   Poly-HRP anti-rabbit IgG detection system was performed. Finally,
   streptavidin-biotin-peroxidase method was used. All immunohistochemical
   staining were investigated independently by two pathologists. The
   assessment was classified into four grades: low reactivity marked as
   1+, faint reactivity as 2+, moderate reactivity as 3+, and strong
   reactivity as 4+.

Animal experiment

   Mice are on the C57BL/6 background and were maintained under a standard
   12 h dark/light cycle with water and chow diet ad libitum. These mice
   (7–8 weeks old, male, n = 10 for each group) were oral gavage with
   normal saline or 50 mg/kg/day concentration of INH for ten days, and
   the mice were sacrificed and liver tissues were harvested for further
   analysis. All above studies related to animals were approved by the
   Peking University Health Science Center Institutional Animal Care and
   Use Committee.

Patient tumour samples

   A total of ten liver cancer patients were recruited among the
   hospitalized patients. Patients were of both sexes at the age of 25 to
   70 years. Ten paired liver cancer samples were obtained from Peking
   University People’s Hospital from September 2018 to June 2020. All
   samples were obtained with informed consent at the Peking University
   People’s Hospital. All samples were obtained with informed consent at
   the Peking University People’s Hospital. The studies were approved by
   the Peking University Health Science Center Ethics Committee under
   IRB00001052-12088, in accordance with the declaration of Helsinki.

In vitro isonicotinylase reaction experiment

   Transfected the HA-CBP and HA-P300 plasmids into HEK293A cells and
   enriched them with HA beads, then put the beads and histones extracted
   from HepG2 cells into HAT buffer (50 mM Tris-HCl with PH = 8.0, 50 mM
   NaCl, 4 mM MgCl[2], 0.1 mM EDTA, 1 mM DTT, 10% glycerine) at 30 °C
   water bath for 1 h, with the presence of 5 mM isonicotyl-CoA. Then
   transferred the mixture on ice for 10 min to stop the reaction. Western
   blot analysis was applied to detect the isonicotinylation.

CoA extraction and HPLC-MS/MS analysis

   HepG2 cells were cultured in full media with various concentrations of
   INH or D4-INH treatment for 24 h, and removed the media quickly, placed
   the dishes on the top of dry ice. Then 2 ml 80% ethanol (v/v) was
   immediately added to the dishes and transferred the dishes to -80 °C
   freezer overnight. The next day collected the cell lysate/methanol
   mixture on dry ice and centrifuged at 4 °C, the supernatants were dried
   for HPLC-MS/MS analysis. The HPLC-MS/MS system was a 6500plus QTrap
   mass spectrometer coupled with ACQUITY UPLC H-Class system. An ACQUITY
   UPLC HSS T3 column (2.1 × 100 mm, 1.8 μm) was employed with mobile
   phase A: water with 5 mM ammonium acetate, and mobile phase B:
   methanol. Linear gradient is: 0 min, 0% B; 1.5 min, 0% B; 6 min, 95% B;
   7.4 min, 95% B; 7.5 min, 0% B; and 10 min, 0% B. Flow rate was
   0.3 mL/min. Column chamber and sample tray were held at 40 °C and
   10 °C, respectively. Data were acquired in multiple reaction monitor
   (MRM) mode for Ac-CoA and INH-CoA with transitions of 873.1/428.1 in
   positive mode. The ion transitions were optimized using chemical
   standards. The nebulizer gas (Gas1), heater gas (Gas2), and curtain gas
   were set at 55, 55, and 30 psi, respectively. The ion spray voltage was
   5500 v. The optimal probe temperature was determined to be 550 °C, and
   the column oven temperature was set to 40 °C. The SCIEX OS 1.6 software
   was applied for metabolite identification and peak integration.

MNase sensitivity assay

   HepG2 cells were resuspended in permeabilization solution-I (150 mM
   sucrose, 80 mM KCl, 35 mM HEPES with pH = 7.4, 5 mM K[2]HPO[4], 5 mM
   MgCl[2], 0.5 mM CaCl[2]) with 0.05% Triton X-100, and incubated at room
   temperature for 5 min, then washed with permeabilization solution-I
   without Triton X-100 and resuspended in permeabilization solution-II
   (150 mM sucrose, 50 mM Tris-HCl with pH 7.5, 50 mM NaCl, 2 mM CaCl[2])
   containing micrococcal nuclease and incubated for 5 min at room
   temperature. Equal volume of 2×TNESK solution (20 mM Tris with pH 7.4,
   0.2 M NaCl, 2 mM EDTA, 2% SDS, 0.2 mg/ml proteinase K) and lysis
   dilution buffer (150 mM NaCl, 5 mM EDTA) were added and incubated at
   37 °C to stop the enzymatic reaction. Isolation of MNase digested
   chromatin from highly permeabilized parasite was performed using
   ethanol precipitation. Cold ethanol and 3 M sodium acetate (pH = 5.2)
   were added and incubated at -20 °C overnight. The next day after
   centrifuging, the pellet was washed with 80% ethanol and air dried. The
   chromatin was dissolved in TE buffer (pH = 8.0). Tanon GIS Software
   (v1.100) was used for gel scanning, and the intensity of the MNase
   sensitivity results was analyzed by densitometry using ImageJ software
   (v1.8.0).

Real-time RT-PCR

   Total RNAs were extracted by TRIZOL reagent according to the
   manufacturer’s protocol. Complementary DNA was prepared using HiScript
   II Q Select RT SuperMix Kit (+gDNA wiper) and real-time RT-PCR was
   performed using ChamQ SYBR qPCR Master Mix according to the
   manufacturer’s protocol. The sequences for real-time RT-PCR primer
   pairs are supplied in Supplementary Table [155]1.

RNA-seq analysis

   The sequencing libraries were prepared using the Next ^® mRNA Sample
   Prep Reagent Set for Illumina (NEB) as per the manufacturer’s
   instruction. The library preparations were sequenced on an Illumina
   NovaSeq 6000 platform and 150 bp paired-end reads were generated. Align
   Reads to the Reference Genome, we took the GRch38 genome as the
   reference genome for this project. Sequencing reads were aligned to the
   reference genome sequence using Tophat2 (v2.1.1) and bowtie2 (v2.2.2)
   in default parameter. Genes and isoforms expression level are
   quantified by a software package: RSEM (RNA-Seq by Expectation
   Maximization v1.3.0). RSEM computes maximum likelihood abundance
   estimates using the Expectation-Maximization (EM) algorithm as its
   statistical model. EdgeR (v3.6.8) package method was used for screening
   differentially expressed genes. GSEA analysis of KEGG pathway database
   was implemented using the Diamond (v0.8.33) package.

Statistics and reproducibility

   All data presented were repeated at least three times independently
   with similar results. All statistical data are presented as mean ± SEM.
   GraphPad Prism 8.0 (GraphPad Software Inc., La Jolla, CA) was used to
   plot and analyze the data, p- values were calculated by two-tailed
   Student’s t-test and one-way ANOVA followed by Bonferrroni’s multiple
   comparisons test. p < 0.05 was considered statistically significant,
   and marked with N.S. for no significance, * for p < 0.5, ** for
   p < 0.01, *** for p < 0.001.

Reporting summary

   Further information on research design is available in the [156]Nature
   Research Reporting Summary linked to this article.

Supplementary information

   [157]Supplementary Information^ (2.3MB, pdf)
   [158]Peer Review File^ (1.4MB, pdf)
   [159]Reporting Summary^ (283.9KB, pdf)

Acknowledgements