SIRT5‐mediated deacetylation of LDHB promotes autophagy and tumorigenesis in colorectal cancer

Lactate dehydrogenase B (LDHB) is a glycolytic enzyme that catalyses the conversion of lactate and NAD + to pyruvate, NADH and H+. Protons (H+) generated by LDHB promote lysosomal acidification and autophagy in cancer, but how this role is regulated has not been defined. In this study, we identified an important post‐translational mechanism by which LDHB is regulated during autophagy in cancer cells. Mass spectrometry revealed that protein sirtuin 5 (SIRT5) is a binding partner of LDHB that deacetylated LDHB at lysine‐329, thereby promoting its enzymatic activity. Deacetylated LDHB increased autophagy and accelerated the growth of colorectal cancer (CRC) cells. Notably, SIRT5 knockout or inhibition by GW5074 increased LDHB acetylation at K329 and inhibited LDHB activity, which downregulated autophagy and CRC cell growth in vitro and in vivo. Clinically, the LDHB‐Ac‐K329 staining score in CRC tissues was lower than that in corresponding peritumour tissues. Low LDHB‐Ac‐K329 status was associated with malignant progression of human CRC and served as a potential prognostic indicator for patients with CRC. Altogether, we conclude that SIRT5‐induced deacetylation of LDHB triggers hyperactivation of autophagy, a key event in tumorigenesis. Thus, the SIRT5/LDHB pathway may represent a novel target for treating CRC.

Lactate dehydrogenase B (LDHB) is a glycolytic enzyme that catalyses the conversion of lactate and NAD + to pyruvate, NADH and H + . Protons (H + ) generated by LDHB promote lysosomal acidification and autophagy in cancer, but how this role is regulated has not been defined. In this study, we identified an important post-translational mechanism by which LDHB is regulated during autophagy in cancer cells. Mass spectrometry revealed that protein sirtuin 5 (SIRT5) is a binding partner of LDHB that deacetylated LDHB at lysine-329, thereby promoting its enzymatic activity. Deacetylated LDHB increased autophagy and accelerated the growth of colorectal cancer (CRC) cells. Notably, SIRT5 knockout or inhibition by GW5074 increased LDHB acetylation at K329 and inhibited LDHB activity, which downregulated autophagy and CRC cell growth in vitro and in vivo. Clinically, the LDHB-Ac-K329 staining score in CRC tissues was lower than that in corresponding peritumour tissues. Low LDHB-Ac-K329 status was associated with malignant progression of human CRC and served as a potential prognostic indicator for patients with CRC. Altogether, we conclude that SIRT5-induced deacetylation of LDHB triggers hyperactivation of autophagy, a key event in tumorigenesis. Thus, the SIRT5/LDHB pathway may represent a novel target for treating CRC.

Introduction
Colorectal cancer (CRC) is the third most common cancer and the second leading cause of cancer-related death worldwide (Bray et al., 2018). Recent studies have shown that autophagy is robustly activated in many types of malignant tumours and contributes to the development and progression of cancers, including CRC (White, 2012). During the autophagic process, intracellular proteins, lipids and organelles are isolated to form an autophagosome (White, 2012). Then, the lysosome fuses with the autophagosome, and this process provides protons and acid-activated proteases and allows cellular organelles to be broken down and reused for biosynthesis and energy metabolism (Maclean et al., 2008). Autophagy is activated in the intestinal epithelium of patients and in a mouse model of CRC (Levy et al., 2015). Autophagosomes are actively produced and promptly consumed in human CRC cell lines (Sato et al., 2007). Autophagy is essential for maintaining cellular homoeostasis and supporting cell survival and growth (Ryter et al., 2013). However, the underlying molecular mechanisms of autophagy in CRC remain elusive.
Lactate dehydrogenase B (LDHB) is a crucial glycolytic enzyme that catalyses conversion of lactate and NAD + to pyruvate, NADH and H + (Quistorff and Grunnet, 2011). A recent study has shown that LDHB plays an important role in autophagy in cancer cells (Brisson et al., 2016). LDHB interacts with V-ATPase at the lysosomal surface; thus, H + protons generated by LDHB directly promote lysosomal acidification and autophagy (Brisson et al., 2016). LDHB-mediated autophagy benefits both oxidative and glycolytic cancer cells. In oxidative cancer cells, monocarboxylate transporter 1 (MCT1), which is highly expressed on the surface of oxidative cancer cells, facilitates the entry of lactate into tumour (Hui et al., 2017;Sonveaux et al., 2008). LDHB catalyses the oxidation of lactate to pyruvate, with the transformation of NAD + to NADH, and H + . Pyruvate then fuels the tricarboxylic acid (TCA) cycle. H + generated by LDHB promotes lysosomal acidification and autophagy which offers cytoprotection by recycling oxidized proteins and organelles when oxidative cancer cells face redox stress (Navarro-Yepes et al., 2014;White, 2012). In glycolytic cancer cells, a large amount of intracellular lactate was catalysed from pyruvate by lactate dehydrogenase A (LDHA) (Shi et al., 2014). This provides sufficient substrate for LDHB. Autophagy sustains glycolytic cancer cell survival in metabolically restricted microenvironments by providing an additional source of energetic and biosynthetic precursors (Rabinowitz and White, 2010). LDHB-mediated autophagy transfers protons to lysosomes and thus contributes to pH homoeostasis in glycolytic cancer cell (Spugnini et al., 2015). Thus, understanding the mechanisms that control LDHB-mediated autophagy may help to identify new targets for treating CRC.
Herein, we report that SIRT5 is an important regulator of autophagy due to its interaction with LDHB. The underlying mechanism involves deacetylation modification and subsequent promotion of LDHB enzymatic activity. Our in vivo and in vitro data indicate that the specific deacetylation of LDHB by SIRT5 is essential for cancer development. Moreover, the lysine-329 acetylation status of LDHB negatively correlates with the malignant progression of human CRC. These data provide a rationale for the use of a SIRT5/ LDHB pathway as the potential target for therapeutic intervention in CRC with an abnormally activated autophagy status.

RNA extraction and quantitative real-time polymerase chain reaction
Total RNA was isolated using TRIzol reagent (Omega, Norcross, GA, USA) and transcribed to cDNA using a cDNA synthesis kit (Takara, Dalian, China) according to the manufacturer's protocol. Quantitative real-time PCR was carried out by using SYBR Green PCR Master Mix (Takara, Shiga, Japan). The StepOnePlus Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) was used to detect the transcript levels of the genes. The gene-specific primers were as follows: LDHB (sense 5ʹ-ATGGC AACTCTAAAGGATCAGC-3ʹ, antisense 5ʹ-CCAACC CCAACAACTGAATCT-3ʹ) and CB (sense 5ʹ-AGATG TAGGCCGGGTGATCT-3ʹ, antisense 5ʹ-CCGCCCTG GATCATGAAGTC-3ʹ).

Confocal immunofluorescence microscopy
HEK293T and HCT116 cells were grown in 24-well plates on glass coverslips. Cells were harvested, washed with PBS and fixed using 4% paraformaldehyde for 10 min. After permeabilization with 0.25% Triton X-100 for 10 min, the cells were washed with PBS and blocked with 1% bovine serum albumin (BSA)/PBS for 30 min, followed by incubation with primary antibodies overnight at 4°C and subsequently with a secondary antibody at room temperature for 1 h. Images were acquired on a confocal laser-scanning microscope (Zeiss, Oberkochen, Germany). The results were from two independent experiments, performed in triplicate.

Mass spectrometry
Cell lysate including Flag-tagged vector/SIRT5 protein was incubated with 15 lL of anti-Flag agarose (Sigma-Aldrich) overnight at 4°C. Then, the samples were subjected to electrophoresis followed by Coomassie Blue staining. Then, the gel was destained with 50% acetonitrile containing 50 mM NH 4 HCO 3 and reduced with 100 mM NH 4 HCO 3 . Then, gel pieces were dehydrated with 100% acetonitrile for 5 min. After the liquid was removed, the gel pieces were rehydrated in dithiothreitol (10 mM) at 37°C for 60 min. Gel pieces were again dehydrated in 100% acetonitrile. After the liquid was removed, gel pieces were rehydrated with iodoacetamide (55 mM). Samples were incubated for 45 min at room temperature (RT) in dark place. Gel pieces were washed with NH 4 HCO 3 (50 mM) and dehydrated again with 100% acetonitrile. Finally, the pellet was resuspended using NH 4 HCO 3 (50 mM) containing 10 ngÁlL À1 trypsin and incubated on ice for 1 h. Excess liquid was removed and gel pieces were digested with trypsin at 37°C overnight. Peptides were sequentially extracted with 50% acetonitrile/5% formic acid and 100% acetonitrile. The sample was then dried and further analysed by LC-MS/MS. In LC-MS/MS analysis, peptides were dissolved in 0.1% formic acid (solvent A). Solvent B was 100% acetonitrile containing 0.1% formic acid. Gradient separation setting: 0-1 min 3-7% solvent B; 1-46 min 7-22% solvent B; 46-55 min 22-35% solvent B; and 55-56 min 35-80% solvent B, then holding at 80% for 4 min. Flow rate is maintained at 300 nLÁmin À1 . The peptides were injected into an NSI source for ionization and then analysed by Thermo Scientific VELOS PRO mass spectrometry (Waltham, MA, USA). Voltage: 2.0 kV. The peptide precursor and its secondary fragments were detected and analysed using high-resolution Orbitrap. The m/z scan range was 350-1800 for full scan, and intact peptides were detected at a resolution of 60 000. The fragments were detected in the Orbitrap at a resolution of 17 500. The dynamic exclusion time of the tandem mass spectrometry scan was set to 15.0 s. Automatic gain control (AGC) was set at 5E4. The resulting MS/MS data were processed using Proteome Discoverer 2.0. Database: human identification (Thermo Scientific).

GST pull-down
GST-tagged SIRT5 and His-tagged LDHB were expressed in BL21(DE3) cells (Sangon Biotech, Shanghai, China). GST-tagged proteins were purified with Glutathione Sepharose 4B beads (GE Healthcare, Chicago, IL, USA) according to the manufacturer's instructions. His-tagged proteins were prepared and purified using Ni-affinity resins (GE Healthcare). Purified GSTtagged SIRT5 protein was incubated with His-tagged LDHB protein at 4°C for 1 h. The beads were washed 5-10 times and boiled in SDS loading buffer. Then, samples were analysed by western blotting.

Immunoprecipitation
To analyse endogenous protein-protein interaction, whole lysates were incubated with antibody against LDHB or SIRT5 and 20 lL protein A/G agarose (Pierce, Waltham, MA, USA) overnight at 4°C. For exogenous co-IP assay, cell lysate containing Flagtagged SIRT5 or HA-tagged LDHB was incubated with anti-Flag (Sigma-Aldrich) or anti-HA agarose (Sigma-Aldrich) overnight at 4°C. Then, 59 SDS/PAGE sample loading buffer was added to the agarose and boiled for 10 min. The resulting samples were analysed by western blotting.

Western blotting assay
Cells were washed with cold PBS and lysed in the RIPA buffer containing protease inhibitors by incubating for 30 min on ice, followed by centrifugation at 15 000 g for 30 min. Samples were boiled and then loaded on 10% or 15% SDS/PAGE, separated by electrophoresis and transferred to PVDF membranes, which were blocked and then incubated with the secondary antibodies for 1 h at room temperature. The immunoreactive bands were visualized by an ECL Plus system (Tanon, Shanghai, China).

Autophagic flux assay
An autophagic flux assay was performed using an mRFP-GFP-LC3 adenoviral vector-encoding construct (HanBio Technology, Shanghai, China) to monitor autophagosome maturation, which was used according to the manufacturer's instructions (Zhou et al., 2012). The images were visualized using a confocal laser-scanning microscope (Zeiss).

In vitro deacetylation
Acetylated HA-LDHB protein was immunopurified using an anti-HA antibody from TSA-treated shSIRT5 HEK293T cells and then incubated with purified SIRT5 protein with or without 100 lM NAD + or 5 mM nicotinamide, as indicated in the deacetylation reaction buffer (50 mM Tris/HCl, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl 2 , 1 mgÁmL À1 BSA and 200 nM TSA, pH 8.0) for 1 h at 37°C. After that, the samples were analysed by western blotting.

Immunohistochemistry
CRC samples were obtained from surgical patients who provided signed informed consent at Renji Hospital, Shanghai, China. The experiment was approved by the Ethics Committee of Renji Hospital. Subcutaneous tumour tissues of mice fixed in 4% paraformaldehyde were dehydrated, embedded in paraffin and cut into 4-lm sections. Human colorectal tumour tissue samples or subcutaneous mouse tumour tissues were dewaxed, hydrated and washed. Antigens were retrieved with 10 mM sodium citrate buffer, and then, the slides were treated with 2% H 2 O 2 in methanol to block endogenous peroxide, after which primary antibody was added and incubated at RT for 2 h. HRP-conjugated goat antirabbit IgG (Cell Signaling Technology) and DAB [3,30diaminobenzidine solution (DAKO, Copenhagen, Denmark)] were used, and counterstaining was performed with haematoxylin. The signal intensity of IHC was independently evaluated by two researchers without prior knowledge about the patients and samples. The signal intensity was divided into 0 = negative, 1 = weak, 2 = moderate and 3 = strong. The staining frequency was categorized as follows: 0 = no staining, 1 < 25%, 2 = 25-50% and 3 > 50%. The final scores for LDHB-Ac-K329 in those colorectal tissues were on a scale of 0-9, in which a score ≤ 3 was defined as representing low expression and a score of > 3 as representing high expression.

Lysosomal pH measurement
Lysosomes were loaded overnight with 0.5 mgÁmL À1 of pH-sensitive FITC/dextran (Sigma) prepared in DMEM. FITC/dextran was removed and washed in HBSS, followed by 2 h of incubation in HBSS with the indicated treatments. The emitted fluorescence was recorded using Axiovision software on a microscope equipped with an ApoTome module for structured illumination (Zeiss). The pH values were derived from the linear standard curve, which was obtained with the addition of 20 lM monensin (Sigma) and 10 lM nigericin (Invitrogen) in HEPES-buffered HBSS adjusted to different pH values between 3.5 and 7.5. Fluorescence intensities of six images per condition were transformed into pH values using the standard curve.

XF24 extracellular flux analysis
We used an XF24 extracellular flux analyser to evaluate the oxygen consumption rate (OCR). In total, 1 9 10 4 cells were plated in XF24 cell culture plates (Seahorse Bioscience, Billerica, MA, USA) and incubated at 37°C overnight. Then, the cells were incubated in bicarbonate-free DMEM medium in a 37°C non-CO 2 incubator for 1 h, and the XF assay was performed at the designated time points. The values are presented as the means AE SE of the mean.

ATP and lactate assay
ATP Colorimetric Assay Kit (Biovision, Milpitas, CA, USA) was used to determine ATP production according to the manufacturer's instructions. Lactate levels in culture medium and in cells were measured according to the instructions of the Lactate Assay Kit (Njjcbio). These readouts were normalized by cell numbers.

Cell proliferation and colony-formation assay
For all cell proliferation experiments, 5.0 9 10 4 cells per well were seeded onto a 6-well plate. Cells were collected every 24 h, and live cells were counted using a haemocytometer after trypan blue exclusion. For the plate colony-formation assay, 200 cells per well were seeded onto a 6-well plate. For the soft-agar colony-formation assay, 100 cells were plated onto a 24-well plate. Cells were suspended in culture medium containing 0.4% agarose (Sigma) and placed on top of solidified 0.6% agarose in DMEM. The medium was changed regularly. After culture for 14 days, colonies were stained with crystal violet and counted.

Xenograft tumour studies
All in vivo experiments were performed in accordance with the guidelines provided by the Animal Ethics Committee of Renji Hospital. Five-week-old male BALB/c nude mice (Shanghai Laboratory Animal Center, Shanghai, China) were injected subcutaneously with LDHB-KO HCT116/DLD1 cells (5 9 10 6 cells) stably expressing EGFP-3Flag-LDHB WT/K329R/K329Q . All mice were killed at 21 days after injection, and tumour tissues were collected and weighed. Five-weekold male BALB/c nude mice (Shanghai Laboratory Animal Center) received subcutaneous injections of 5 9 10 6 LDHB-KO HCT116 cells that were stably transfected with EGFP-3Flag-LDHB WT or EGFP-3Flag-LDHB K329Q in 100 lL of saline at the right dorsal flank. Fourteen days after tumour implantation, the mice were randomly assigned to a treatment group receiving GW5074 (5 mgÁkg À1 ) or a control group receiving saline three times per week intraperitoneally for 21 days. At the end of the experiment, animals were sacrificed, and tumour tissues were collected for weighing and immunohistochemical analysis.

Statistics
All data are presented as the means AE SEM and were statistically analysed using GraphPad Prism 6 (Graph-Pad Software, La Jolla, CA, USA) or SPSS 17 software (SPSS, Dallas, TX, USA). A two-tailed unpaired t-test or ANOVA was used to analyse the difference between two or more groups. The chi-square test was applied for rate comparisons. The overall survival curves were calculated using the Kaplan-Meier method and compared by log-rank test. P-values < 0.05 were considered as statistically significant.

SIRT5 interacts with LDHB
Based on the importance of SIRT5 in tumour progression, we asked which regulatory signalling factors were directly regulated by SIRT5. Most previous studies have mainly focused on the PTM functions of SIRT5 in glycolysis, ammonia metabolism, FAO and oxidative stress (Colak et al., 2015;Lin et al., 2013;Nakagawa et al., 2009;Nishida et al., 2015;Wang et al., 2018;Zhou et al., 2016). To search for novel interacting partners of SIRT5, we overexpressed C-terminal Flag-tagged SIRT5 protein in HEK293T cells (Fig. 1A) and then performed mass spectrometry to identify the associated proteins. We identified LDHB as a novel interacting partner Immunoprecipitates and whole-cell extracts were subsequently separated by SDS/PAGE, followed by western blotting with an anti-LDHB or SIRT5 antibody. (C) HEK293T cells were transfected with HA-LDHB for 48 h. Total protein extracts were immunoprecipitated with anti-HA antibody (IgG as a control) and separated by SDS/PAGE, followed by western blotting with an anti-SIRT5 or HA antibody. (D) HEK293T cells were transfected with Flag-SIRT5 for 48 h. Total protein extracts were immunoprecipitated with anti-Flag antibody (IgG as a control) and separated by SDS/PAGE, followed by western blotting with an anti-LDHB or anti-Flag antibody. (E, F) HEK293T cells were cotransfected with HA-LDHB and Flag-SIRT5 for 48 h. Total protein extracts were immunoprecipitated with anti-HA or anti-Flag antibody (IgG as a control) and separated by SDS/PAGE, followed by western blotting with an anti-HA or anti-Flag antibody. (G) Recombinant GST-SIRT5 (or GST-vector as a control) was incubated with purified His-LDHB protein for 3 h. Bound proteins were eluted and analysed with SDS/PAGE and Coomassie Blue staining. Anti-His antibody was used to detect LDHB in eluates. (H) The colocalization (yellow) of endogenous SIRT5 (red) and LDHB (green) in HEK293T and HCT116 cells was analysed using an immunofluorescence and confocal laser-scanning microscopy. Scale bar, 10 lm.
protein of SIRT5. The SIRT5-LDHB interaction was then confirmed by direct endogenous and exogenous endogenous co-immunoprecipitation (co-IP) in HEK293T cells (Fig. 1B, endogenous interaction; Fig. 1C-F, exogenous interaction). In the GST pulldown assay, we further found that GST-tagged SIRT5 interacted with recombinant LDHB protein (Fig. 1G). Immunofluorescence staining showed that SIRT5 and LDHB colocalized in the cytoplasm in HEK293T and HCT116 cells (Fig. 1H). Therefore, LDHB functions as a physiological interacting partner of SIRT5.

SIRT5 deacetylates LDHB at lysine-329
Given that SIRT5 has been demonstrated to regulate metabolic enzymes by direct PTM (Yang et al., 2015), we asked whether this interaction could affect the lysine acetylation of LDHB. We found that SIRT5 overexpression decreased LDHB acetylation in HCT116 cells, while the succinylation, malonylation or glutarylation of LDHB remained largely unaffected ( Fig. 2A and Fig. S1). Supporting this observation, SIRT5 knockdown using specific siRNA increased LDHB acetylation (Fig. 2B). Furthermore, we stably knocked down SIRT5 in HEK293T cells (Fig. 2C, left) and found that only SIRT5-WT, but not enzymatically inactive SIRT5-H158Y, deacetylated LDHB (Fig. 2C, right). Finally, in vitro deacetylation assays confirmed that affinity-purified acetyl-HA-LDHB protein can be directly deacetylated by SIRT5 protein in a NAD + -dependent manner (Fig. 2D). Collectively, these data indicated that LDHB is a deacetylation substrate of SIRT5 in cells and in vitro.
Four putative acetylation sites were identified in LDHB by mass spectrometry (Fig. 2E) (Choudhary et al., 2009). We then mutated each of the four putative acetylation sites individually to arginine (R). The mutation of K329, but not other lysine residues, to arginine resulted in a significant reduction in LDHB acetylation (Fig. 2F). This result indicates that K329, which is evolutionarily conserved from Danio rerio to mammals (Fig. 2G), is a major acetylation site in LDHB. Afterwards, we generated an antibody specifically recognizing the K329-acetylated LDHB. We found that SIRT5 had no significant effect on the acetylation of LDHB K329R and LDHB K329Q [a mimic LDHB acetylation mutant created by substituting K329 with glutamine (Q)] mutants (Fig. 2H,I). Thus, we demonstrated that LDHB was deacetylated by SIRT5 at K329.

SIRT5-mediated deacetylation increases LDHB activity
To determine whether the SIRT5-mediated deacetylation modification regulated LDHB expression or function, we initially knocked down SIRT5 in CRC cells using specific siRNA. SIRT5 knockdown had no effects on the mRNA or protein levels of LDHB ( Fig. 3A) but resulted in a decrease in LDHB activity in HCT116 and DLD1 cells ( Fig. 3B and Fig. S2A). SIRT5 KO had similar effects on LDHB activity in HCT116 cells (Fig. 3C,D). In contrast, we verified that siSIRT5 or SIRT5 KO did not alter LDHA activity in CRC cells (Fig. S2B,C). Furthermore, re-expression of wild-type SIRT5, but not the inactive H158Y mutant, increased LDHB enzyme activity in HEK293T shSIRT5 cells (Fig. 3E).
To test the effect of K329 acetylation, the activities of LDHB K329R and LDHB K329Q mutants were compared with that of wild-type LDHB. We found that LDHB K329Q displayed only 50% of the wild-type activity, whereas the LDHB K329R mutation produced an opposite effect on LDHB activity in HCT116 and DLD1 cells (Fig. 3F and Fig. S2D). SIRT5 regulated the activity of the LDHB WT , but not LDHB K329R and LDHB K329Q mutants (Fig. 3G,H), indicating that SIRT5 stimulated LDHB activity mostly via deacetylation of K329. These data demonstrated that LDHB Fig. 3. SIRT5-mediated deacetylation increases LDHB activity. (A) HCT116 cells were transfected with siRNA-SIRT5 (or siRNA-NC as a control) for 48 h. The level of LDHB mRNA was analysed with qPCR. LDHB protein was analysed with western blotting. b-Tubulin was used as a loading control for western blotting. Error bars represent means AE SEM (n = 3). (B) HCT116 cells were transfected with HA-LDHB and siRNA against SIRT5 or negative control siRNA for 48 h. In addition, control cells were treated with the LDHB inhibitor oxamate (10 mM) for the last 24 h. HA-LDHB was immunoprecipitated, and LDHB activity was assayed. LDHB activity was normalized against protein levels. Error bars represent means AE SEM (n = 3). (C) LDHB mRNA and protein levels were analysed in SIRT5 WT HCT116 cells or SIRT5 KO HCT116 cells. Error bars represent means AE SEM (n = 3). (D) SIRT5 WT/KO HCT116 cells were transfected with HA-LDHB for 48 h. In addition, WT cells were treated with the LDHB inhibitor oxamate (10 mM) for the last 24 h. LDHB was immunoprecipitated, and LDHB activity was assayed. LDHB activity was normalized against protein levels. Error bars represent means AE SEM (n = 3). (E) HEK293T shSIRT5 cells were cotransfected with HA-LDHB and vector/SIRT5 wild-type (WT)/SIRT5 inactive mutant (H158Y). LDHB was immunoprecipitated, and LDHB activity was assayed. LDHB activity was normalized against the protein levels. Error bars represent means AE SEM (n = 3). (F) HA-tagged wild-type and mutant LDHB proteins were expressed in HCT116 cells and purified by immunoprecipitation. The enzyme activity was measured and normalized against the protein level. Error bars represent means AE SEM (n = 3). (G) HCT116 cells were cotransfected with Flag-SIRT5 (or empty vector as control) and HA-LDHB WT , HA-LDHB K329R or HA-LDHB K329Q for 48 h. LDHB was immunoprecipitated, and LDHB activity was assayed. LDHB activity was normalized against the protein levels. Error bars represent means AE SEM (n = 3). (H) HCT116 cells were cotransfected with siRNA-SIRT5 (or siRNA-NC as a control) and HA-LDHB WT , HA-LDHB K329R or HA-LDHB K329Q for 48 h. LDHB was immunoprecipitated, and LDHB activity was assayed. LDHB activity was normalized against the protein levels. Error bars represent means AE SEM (n = 3). *P < 0.05, **P < 0.01. P-values were based on Student's t-test. activity was regulated by SIRT5 via deacetylation modification at K329.

SIRT5 promotes autophagy through LDHB deacetylation
A previous study showed LDHB activity to be necessary for basal autophagy and cancer cell proliferation since it directly controls lysosomal acidification ( Fig. 4A; Brisson et al., 2016). SIRT5-mediated deacetylation of LDHB increased LDHB activity in our study. Thus, we examined possible autophagy regulation by SIRT5. Autophagic impairment often results in accumulation of the autophagic substrate optineurin (Korac et al., 2013). Silencing SIRT5 increased optineurin accumulation and apoptosis (cleaved caspase-3) levels ( Fig. 4B) and decreased lysosome acidification in HCT116 cells (Fig. 4C). To demonstrate that SIRT5 controls autophagic flux in cancer cells, we determined the LC3-II protein level, a marker of the autophagic flux (Rubinsztein et al., 2009). SIRT5 knockdown induced leupeptin-sensitive LC3-II protein accumulation in HCT116 cells, which represents a potent inhibition of the autophagic flux (6.17-2.87) (Fig. 4D). Furthermore, we transiently transfected a tandem mRFP-EGFP-LC3 reporter to estimate the autophagic flux by measuring the abundance of mature autolysosomes (Kimura et al., 2007) and found that SIRT5 KO decreased the number of mature autolysosomes in HCT116 cells (Fig. 4E).
To examine whether the SIRT5-mediated regulation of autophagy was LDHB-dependent, LDHB was stably knocked down in SIRT5 WT/KO HCT116 cells by shLDHB. Knocking out SIRT5 caused LC3-II protein accumulation and autophagic flux inhibition with no additive effect from shLDHB (Fig. 4F). LDHB overexpression abolished the siSIRT5-mediated autophagy inhibition and pro-apoptotic effect (Fig. 4G). Moreover, SIRT5 overexpression significantly promoted lysosomal acidification (Fig. 4H). However, LDHB knockdown by specific siRNA almost completely compromised SIRT5-induced lysosomal acidification in the HCT116 cells (Fig. 4H). Thus, our results show that the effect of SIRT5 on autophagy is largely LDHBdependent.
Furthermore, SIRT5 KO and shLDHB HCT116 cells were subsequently transfected with HA-LDHB WT , HA-LDHB K329R , HA-LDHB WT +Flag-SIRT5 or HA-LDHB K329Q +Flag-SIRT5. Cells expressing the deacetylated LDHB protein (HA-LDHB K329R or HA-LDHB WT +Flag-SIRT5 group) exhibited a decrease in apoptosis and an accumulation of autophagic substrate (Fig. 4I), and promotion in lysosome acidification (Fig. 4J), and autolysosome formation (Fig. 4K). Similar effects were observed in DLD1 cells ( Fig. S3A-C). GW5074 is an inhibitor of SIRT5 with pharmacologically suitable properties (Suenkel et al., 2012). We observed that GW5074 treatment reduced SIRT5 protein levels and increased LDHB-K329 acetylation (Fig. 4L), followed by an inhibition of (L) HCT116 cells were transiently transfected with HA-LDHB WT for 48 h and treated or not treated with 100 lM GW5074 for the last 24 h, followed by immunoprecipitation, LDHB enzyme activity assay and LDHB acetylation assay by western blotting. Error bars represent means AE SEM (n = 3). (M) Autophagic and apoptotic protein abundance in HCT116 cells treated or not treated with 100 lM GW5074 for 24 h. (N) LDHB was knocked out in HCT116 cells, and LDHB and b-tubulin levels were determined by western analysis using anti-LDHB and b-tubulin antibodies. LDHB-KO HCT116 cells were transfected with HA-LDHB WT or HA-LDHB K329Q for 48 h and treated or not treated with 100 lM GW5074 for the last 24 h. Autophagic and apoptotic protein abundance is shown. *P < 0.05, **P < 0.01, ***P < 0.001. P-values were based on Student's t-test.
autophagy and an induction of apoptosis (Fig. 4M). When SIRT5 was deregulated by GW5074, ectopic expression of LDHB WT and LDHB K329Q in LDHB-KO HCT116 cells resulted in a comparable level of autophagy and apoptosis (Fig. 4N). These results suggest, for the first time, that the acetylation status of LDHB, as directed by SIRT5, determines its enzymatic activity and regulates autophagy as well as apoptosis in CRC cells.

LDHB deacetylation by SIRT5 promotes cell respiration, proliferation and tumour growth
As LDHB is a glucose metabolic enzyme oxidating lactate to pyruvate, we tested the effect of LDHB-K329 acetylation on metabolism. HCT116 LDHB-KO cells were infected with EGFP-3Flag-LDHB WT/K329R/K329Q to stably overexpress these genes (Fig. S4A). We found that LDHB K329 deacetylation significantly decreased intracellular and extracellular lactate (LDHB reaction substrate) levels in HCT116 cells (Fig. S4B). We also observed that oxygen consumption rate (OCR) reflecting cell respiration and ATP production were increased in cells transfected with the LDHB K329R deacetylated mutant and decreased with the LDHB K329Q-acetylated mutant (Fig. S4C,D). Together, these data show that LDHB K329 deacetylation promotes lactate-fuelled cell respiration.
It is well known that increased autophagic flux and lactate-fuelled cell respiration are related to oxidative cancer cell proliferation (Brisson et al., 2016;Sonveaux et al., 2008). We further assessed the contribution of the SIRT5/LDHB axis to these events. As shown in Fig. 5A, SIRT5 knockdown by specific siRNA significantly decreased cell number in HCT116 cells. Conversely, overexpression of SIRT5 in HCT116 cells increased cell number (Fig. 5B). Furthermore, SIRT5induced cell proliferation effect was completely compromised by LDHB knockdown (Fig. 5B), suggesting that the SIRT5-mediated cell proliferation advantage in HCT116 cells was LDHB-dependent. To further examine the role of LDHB K329 deacetylation by SIRT5 in cell proliferation, we performed plate and soft-agar colony-formation assay. Compared with the LDHB WT -transfected cells, we observed significantly more colonies in the nonacetylatable LDHB K329Rtransfected cells while less colonies in the LDHB K329Qtransfected cells (Fig. 5C). SIRT5 KO caused a significant decrease of cell number in HCT116 cells transfected with LDHB WT , whereas the number of cells expressing LDHB K329R or LDHB K329Q was comparable to that in the control group, regardless of the expression of SIRT5 (Fig. 5D). After the cell proliferation advantage afforded by SIRT5-mediated LDHB deacetylation to cancer cells in vitro was verified, we performed xenograft studies in vivo. In xenograft mouse models, LDHB-KO HCT116 or DLD1 cells with stable expression of EGFP-3Flag-LDHB WT/K329R/ K329Q were inoculated into nude mice. As shown in Fig. 5E,F and Fig. S5A-C, replacement of endogenous LDHBWT by LDHBK329Q reduced tumour growth and replacement of endogenous LDHBWT by LDHBK329R promotes tumour growth. In case of GW5074 treatment, ectopic expression of either LDHB WT or K329Q resulted in comparable tumour volumes and weights (Fig. 5G,H). Consistent with our previous in vitro results, IHC analysis of tumours also revealed that GW5074 reduced SIRT5 protein level and resulted in a significant downregulation of autophagic flux and cell proliferation (Fig. 5I). Collectively, these in vivo and in vitro results indicated the K329 deacetylation of LDHB by SIRT5 promoted cancer cell proliferation and tumour growth.  3). (E) A total of 5 9 10 6 LDHB-KO HCT116 cells stably expressing EGFP-3Flag-LDHB WT/K329R/K329Q were injected subcutaneously into the right dorsal flank of nude mice. Image of tumours isolated from nude mice (n = 5 for each group). (F) The weight of tumours when mice were sacrificed. (G-I) A total of 5 9 10 6 LDHB-KO HCT116 cells stably expressing EGFP-3Flag-LDHB WT/K329Q were injected subcutaneously into the right dorsal flank of nude mice. Fourteen days after tumour implantation, mice were randomly assigned to a treatment group receiving GW5074 (5 mgÁkg À1 ) or saline as a control group three times per week intraperitoneally for 35 days (n = 5 for each group). (G) Image of tumours isolated from nude mice. (H) The weight of tumours when mice were sacrificed. (I) Haematoxylin and eosin (H&E) staining and immunohistochemical detection of Ki67, LC3 and SIRT5. Scale bar, 50 lm. *P < 0.05, **P < 0.01. P-values were based on Student's t-test.

Decreased LDHB-K329 acetylation in human CRC samples and correlation with poor prognosis
The finding that the deacetylation of LDHB by SIRT5 promotes autophagy and cancer cell proliferation led us to examine K329 acetylation status in human CRC tissues. A total of 54 pairs of CRC samples with adjacent normal colorectal tissues were examined by immunostaining with antibody against LDHB acetylated at K329 (LDHB-Ac-K329; Fig. 6A). High K329 acetylation was detected in 69% (37 of 54) of normal tissues, and only 33% (18 of 54) of tumour tissues exhibited high K329 acetylation status (P < 0.001; Fig. 6B). The mean LDHB-Ac-K329 staining score in CRC tissues was lower than that in corresponding normal tissues (P < 0.001; Fig. 6C). Univariate statistical analysis found that LDHB-Ac-K329 levels were negatively associated with tumour size (P = 0.004) and histological grade (P = 0.031; Table 1). Finally, we found that low LDHB-Ac-K329 staining predicted poor overall patient survival (Fig. 6D). These results suggest that LDHB K329 acetylation status might be used to predict tumour progression in human CRC.

Discussion
Lactate dehydrogenase B, catalysing the conversion of lactate and NAD + to pyruvate, NADH and H + , is a key contributor to lysosomal acidification and autophagy in cancer (Brisson et al., 2016). LDHB is highly expressed in multiple types of cancer, including medulloblastoma, cholangiocarcinoma, oesophageal squamous cell carcinoma and breast cancer (de Haas et al., 2008;Isozaki et al., 2012;McCleland et al., 2012;Yoo et al., 2009). Furthermore, elevated LDHB expression had a significant impact on cancer progression and metastasis (Koshiyama et al., 2013;Li et al., 2016;McCleland et al., 2012). Therefore, LDHB has an important role in cancer progression. PTM regulates the function of multiple metabolic enzymes and the consequent cellular biochemical metabolism and biological status. Interestingly, our data proved that LDHB is an acetylated protein and verified that K329 of LDHB is the major deacetylation residue modified by SIRT5. Although functional LDH is a homo-or heterotetramer composed of LDHA and LDHB subunits, LDHA does not harbour a motif comparable to K329 in LDHB and our research proved that SIRT5 only regulated LDHB activity but not LDHA. Thus, in this study, we uncovered a mechanism of LDHB regulation that contributes to its increased activity by SIRT5 on R329 in CRC cells.
SIRT5 is a member of the sirtuin family, which can sense nutrient requirements and optimize cellular pathways to maintain metabolic homoeostasis and cellular survival under limited nutritional conditions (Guedouari et al., 2017). Recent studies have revealed that SIRT5 is involved in tumorigenesis via its PTM activities (Wang et al., 2018;Xiangyun et al., 2017). In our study, when we used Flag-SIRT5 as the bait to find its possible partners, LDHB was identified from the SIRT5 interaction network. We further found that SIRT5 regulates autophagy, apoptosis and proliferation of CRC cells via its association with LDHB. The deacetylation of LDHB by SIRT5 powerfully enhanced LDHB enzymatic activity. However, SIRT5 has also been reported as a suppressor of autophagy in the human breast cancer cell line MDA-MB-231 (Polletta et al., 2015), indicating that SIRT5 might act as both autophagy promoter and suppressor in a context-dependent manner. Notably, MCT1, which facilitates lactate uptake, is not expressed in MDA-MB-231 (Hussien and Brooks, 2011). Lactate is the metabolic substrate of LDHB. In addition, MDA-MB-231 cells expressed mainly LDHA, not LDHB (Hussien and Brooks, 2011). Therefore, the function of SIRT5 in autophagy is context-dependent.
Lactate dehydrogenase B interacts with V-ATPase at the lysosomal surface; as a result, the protons (H + ) generated by LDHB promote lysosomal acidification. Lysosomal acidification is important for the following vesicle maturation and protease activation during autophagy. In our study, LDHB acetylation or SIRT5 KO prevented the lysosomal acidification and decreased the abundance of mature autolysosomes. LDHB acetylation or SIRT5 KO also induced leupeptin-sensitive LC3-II protein accumulation, which represented a potent inhibition of the autophagic flux (Rubinsztein et al., 2009) and prevented degradation of autophagic substrate optineurin. Thus, it was reasonable to believe that SIRT5deacetylated LDHB could promote the lysosomal acidification and, subsequently, autophagy, which are essential for cancer development (Fig. 6E).
It has been shown that inhibiting autophagy with agents, such as chloroquine, may lead to cell death, which implies a promising therapeutic approach against cancer (Maclean et al., 2008;Morgan et al., 2014). The modulation of the key elements of the autophagy pathway is also being exploited as a novel therapeutic option for CRC treatment (Mokarram et al., 2017). The deletion of the autophagy-related gene may prevent the development and progression of CRC in genetically predisposed patients (Levy et al., 2015). In another study on human CRC cells, inhibiting autophagy increases anticancer drug-induced apoptotic cell death (Gurkan et al., 2018). Thus, the promotion of autophagy by SIRT5 via LDHB suggested that they might affect cancer growth. Besides being an important regulator in autophagy, LDHB oxidates lactate to pyruvate, which fuels the TCA cycle and benefits for oxidative cancer cells (Sonveaux et al., 2008). Our results showed that LDHB K329 deacetylation significantly increased cell respiration and ATP generation in oxidative HCT116 cells. These findings imply that cell respiration mediated by LDHB deacetylation might also contribute to cancer cell growth. In this study, our data show that SIRT5 KO or LDHB knockdown inhibited proliferation of CRC cells. Moreover, SIRT5-mediated regulation of cell proliferation was dependent on the deacetylation of LDHB at the K329. Furthermore, our in vivo growth assays showed that cancer cells overexpressing LDHB WT grew quicker than cells overexpressing LDHB K329Q , whereas the growth advantage induced by LDHB WT was blocked by the presence of SIRT5 inhibitor GW5074. Although GW5074 could impair clonogenic and tumorigenic capacity of CRC cells by its inhibitory effect on c-Raf kinase activity (Borovski et al., 2017), our studies demonstrated it could significantly reduce the expression level of endogenous SIRT5 and blocked autophagic flux and tumour cell proliferation via SIRT5/LDHB axis. Considering the side effects of lysosomotropic agent chloroquine (Maycotte et al., 2012) and a decrease in chloroquine activity by tumour acidity (Pellegrini et al., 2014), SIRT5-mediated LDHB deacetylation pathway might be a new therapeutic target for treating CRC. In addition, we found that LDHB K329 acetylation was lower in CRC tissues than in normal colorectal tissues. Significantly, low levels of LDHB-Ac-K329 in tumours may predict short survival time in patients with CRC. Hence, the LDHB-Ac-K329 status is a potential prognostic indicator for patients with CRC and might be useful for identifying the patients with CRC who are suitable for anti-autophagy treatment.

Conclusions
Overall, these results suggested a model where SIRT5 directly interacts with LDHB and yields strong deacetylation of LDHB at the K329 site. This effect increases LDHB activity and favours lysosomal acidification and autolysosomal maturation. Enhanced autophagy along with reduced apoptotic cell death contributes to colorectal tumour progression. This confirms that targeting SIRT5/LDHB pathway might be a feasible and effective strategy in the treatment of CRC.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. SIRT5 does not affect the succinylation, malonylation or glutarylation of LDHB. Fig. S2. SIRT5 increases LDHB activity. Fig. S3. SIRT5 promotes autophagy via LDHB deacetylation. Fig. S4. LDHB deacetylation promotes cell respiration. Fig. S5. LDHB deacetylation promotes tumour growth in vivo.