Fig 1.
Salmonella infection changes host protein acetylation.
(A) Lysine acetylomics workflow is shown. Cell pellets were lysed, combined, reduced, alkylated, and digested with trypsin to generate peptides with 5 mg of proteins input per state. Then, digested peptides were subjected to anti-KAc antibody enrichment and LC/MS analysis. (B) Annotation enrichment analysis of proteins with regulated KAc sites identified in U937 and TPH-1 cells infected with S. Typhimurium for 1, 3, and 24 h. The bar graphs show significantly gene ontology biological processes (GOBPs), gene ontology cellular compartments (GOCCs) and gene ontology molecular functions (GOMFs). (C) Associations among proteins with regulated Kac sites based on STRING. (D) Representative mass spectrometry analysis of the peptide containing CDC42 K153Ac. Shown is the spectrum covering the region from 100 to 1200 m/z which includes the peptide containing the acetylated lysine 153. Acetylation of CDC42 K153 was downregulated by ~2-fold at 1 h. (E) Verification of K153 acetylation level after Salmonella infection. Flag-CDC42 was transfected into HEK293T cells for 48 h, followed by S. Typhimurium infection for 1 h. Cell lysates were used for IP with anti-Flag antibody, and K153 acetylation was determined by western blot (WB) assay with CDC42-specific acetyl-K153 antibody (anti-K153Ac). (F) Alignment of CDC42 amino acid sequences from various species, red box indicates the conserved K153.
Fig 2.
Acetylation of K153 is essential for CDC42 interaction with PAK4.
(A) Ribbon diagram showing the structure of CDC42-PAK4 (PDB-ID: 5UPK). (B) S. Typhimurium infection drastically decreased the binding between CDC42 and PAK4. Flag-CDC42 was co-transfected with HA-PAK4 or HA-PAK1 into HEK293T cells for 48 h, followed by S. Typhimurium infection for 1 h. Protein interactions was analyzed by immunoprecipitation (IP) with anti-Flag antibody and followed by WB. K153R mutant attenuated the interaction between CDC42 and PAK4. Flag-tagged CDC42-WT, K153Q or K153R was co-transfected with HA-PAK4 (C) or HA-PAK1 (D) into HEK293T cells with CDC42 knockdown. The association between CDC42 and PAK4 or PAK1 was determined by IP/WB with indicated antibodies. (E) K153 acetylation did not affect the activation of CDC42. Expression level of the active GTP-bound form of CDC42 WT or K153QR mutants in HEK293T cells with endogenous CDC42 knockdown was detected by CDC42 Activation Assay Kit (CST).
Fig 3.
CDC42 K153 is deacetylated by SIRT2.
(A) CDC42 K153 acetylation is regulated by SIRT family deacetylases. HEK293T cells transfected with Flag-CDC42 were treated with the deacetylase inhibitors TSA (2 μM) and NAM (10 mM) for 16 h before harvesting. K153 acetylation was determined by IP/WB. (B) SIRT2 decreases K153 acetylation in vivo. Overexpression of SIRT2 significantly decreases K153 acetylation. Flag-CDC42 was co-transfected with or without HA-SIRT2 into HEK293T cells, and K153 acetylation was analyzed by IP with anti-Flag antibody and WB with anti-K153Ac antibody. (C) SIRT2 decreases K153 acetylation in vitro. Purified GST-CDC42 protein was incubated with the cell lysate of HEK293T cells expressing HA-SIRT2 or the control vector. K153 acetylation was determined by the GST pull-down assay, followed by WB with anti-K153Ac antibody. (D) A reduced acetylation of K153 was shown in AK7-treated cells. Flag-CDC42 was co-transfected with or without HA-SIRT2, followed by treatment with the SIRT2 specific inhibitor AK7 (10 μM) or dimethyl sulfoxide (DMSO) (as a negative control) for 16 h. K153 acetylation was then determined by IP/WB. SIRT2 knockdown increases K153 acetylation. Flag-CDC42 was transfected with or without HA-PAK4 into HEK293T cells with SIRT2 knocked down, and the cells were infected with S. Typhimurium. Salmonella infection failed to reduce K153 acetylation and binding between CDC42 with PAK4 in the SIRT2 knockdown cells. Cell lysates were used for IP with anti-Flag antibody. K153 acetylation was then analyzed by WB with anti-K153Ac antibody (E), and the association between CDC42 and PAK4 was determined by WB (F).
Fig 4.
CDC42 K153 is acetylated by p300/CBP.
CBP/p300 increases the acetylation level of K153. HA-CBP (A) or HA-p300 (B) was co-transfected with Flag-CDC42 into HEK293T cells, followed by IP with anti-Flag antibody and K153 acetylation level was examined by WB with anti-K153Ac antibody. (C) The p300/CBP-specific inhibitor A-485 attenuated K153 acetylation. Flag-CDC42 was co-transfected with HA-PAK4 into HEK293T cells, followed by treatment with A-485 (3 μM) or DMSO for 24 h before harvesting. The acetylation levels of K153 and the association between CDC42 and PAK4 were analyzed by IP/WB. (D) p300/CBP increases K153 acetylation in vitro. Purified GST-CDC42 protein was incubated with the cell lysate of HEK293T cells expressing HA-p300, HA-CBP, or the control vector. K153 acetylation was determined by GST pull-down and followed by WB.
Fig 5.
Acetylation of CDC42 K153 affects JNK and p38 phosphorylation.
(A) HCT116 cells stably expressing Flag-CDC42-WT, Flag-CDC42 K153Q, or Flag-CDC42 were established. pCDH-flag vector (EV) was used as a control, and the cells were assessed by WB with anti-Flag antibody and anti-CDC42 antibodies. (B) Phosphorylation of JNK and p38 in K153R mutant. Phosphorylation of JNK and p38 were analyzed by WB in HCT116 cells stably expressing Flag-CDC42-WT, Flag-CDC42 K153Q, or Flag-CDC42 K153R, with pCDH-Flag vector (EV)-expressing cells as a control. (C) Determination of p44/42 (ERK) and AKT phosphorylation by WB in HCT116 cells stably expressing Flag-CDC42-WT, Flag-CDC42 K153Q, or Flag-CDC42 K153R, with pCDH-Flag vector (EV)-expressing cells as a control. (D) The phosphorylation of JNK and p38 were analyzed by WB in HCT116 cells stably expressing Flag-CDC42-WT, Flag-CDC42 K153Q, or Flag-CDC42 K153R treated by S. Typhimurium for 1 h. (E) Phosphorylation of PAK1 and PAK4 were determined in HCT116 cells stably expressing CDC42-WT, CDC42 K153Q or CDC42 K153R by WB (Left), and densitometry analysis was shown (Right). Phosphorylation levels of JNK and p38 were decreased in PAK4 knockdown cells, and a lower phosphorylation level of p38 in PAK4 knockdown cells after Salmonella infection. HCT116 cells with PAK4 knocked down were infected without (F) or with (G) S. Typhimurium for 1 h, and wild-type HCT116 cells was used as a control. JNK and p38 phosphorylation were analyzed by WB (Left), and densitometry analysis was shown (Right). *p<0.05, **p<0.01.
Fig 6.
Unacetylated CDC42 K153 is correlated with enhanced tumorigenesis phenotype of HCT116 cells.
HCT116 cells stably expressing CDC42-WT, CDC42 K153Q or CDC42 K153R were analyzed for apoptosis as followed: (A) The level of full-length or cleaved caspase-3 or PARP in cells treated with 4 μM staurosporine (STS) for 4 h was determined by WB. (B) Cells were treated with 0.4 μM STS for 24 h, and the level of surface exposure of the apoptosis signature phospholipid phosphatidyserine (PS) was determined by Annexin V (AV) staining followed by flow cytometry analysis. Propidium iodide (PI) was used to exclude necrotic cells. (C) Percentage of AV+PI- (gray bars) and AV+PI+ (white bars) cells. Asterisks represent p values for all AV+ cells, and octothorpes represent p values for AV+ and PI- cells (**p<0.01, ***p<0.001, ##p<0.01). (D) Migration and invasion assays of HCT116 cells infected with S. Typhimurium. (E) Migration and invasion assays of HCT116 cells stably expressed CDC42-WT, CDC42 K153Q, or CDC42 K153R. (F) Determination of the mRNA expression level of mmp-2, mmp-9, and e-cadherin by quantitative real-time PCR in HCT116 cells infected with S. Typhimurium. (G) Determination of MMP-2, MMP-9, and E-cadherin by WB in HCT116 cells during S. Typhimurium infection, and densitometry analysis was shown (Right). **p<0.01. (H) Determination of the mRNA expression level of mmp-2, mmp-9, and e-cadherin by quantitative real-time PCR in HCT116 cells stably expressing WT or QR mutants of CDC42 K153. (I) Determination of MMP-2, MMP-9, and E-cadherin by WB in HCT116 cells stably expressing CDC42-WT, CDC42 K153Q or CDC42 K153R, and densitometry analysis was shown (Right). *p<0.05, **p<0.01.
Fig 7.
K153 acetylation is down-regulated in CRCs.
(A) Scheme of the experimental design. Arrow symbols indicate azoxymethane (AOM), 10mg/kg body weight, administered through intraperitoneal injection. The lightning symbol indicates that mice were infected with Salmonella through a blunt gavage needle. Rhombus symbols indicate the administration of 1.5% dextran sodium sulfate (DSS) in drinking water. Triangle symbols indicate that mice were sacrificed, and their colorectal tissues were collected. (B) Fecal Salmonella was detected by PCR with Salmonella-specific primers after 14 weeks post-infection. (C) Colonic tumors in situ. Representative colons from the indicated groups 14 weeks after Salmonella infection. Tumors are indicated by red arrows. (D) Localization of Salmonella in the tissues of the indicated groups was assessed by immunofluorescence staining with anti-Salmonella-specific LPS antibodies. (E) Control mucosa and colonic tumors in each group were stained by hematoxylin-eosin (HE) staining, and the expression levels of CDC42, CDC42 K153 acetylated and Ki67 in the indicated groups were determined by immunohistochemistry (IHC) and quantitated by modified H-score (left panel). Quantitative imaging of CDC42 K153 acetylation (upper right panel) and CDC42 (lower right panel) in untreated (control) mucosa and colonic tumors. Data are expressed as mean ± SD of control mice (n = 3) or AOM+DSS mice (n = 3) or AOM+DSS+STM mice (n = 4) in each group. Mouse tissues were harvested after 14 weeks post-infection, and levels of CDC42 K153 acetylation (F), p-PAK4, PAK4, p-PAK1 and PAK1 (G), MMP-2, MMP-9 and E-cadherin (H) were analyzed by WB (Left), and densitometry analysis was shown (Right). (I) p-ERK, ERK, p-AKT, AKT, p-JNK, and JNK in control and AOM/DSS induced tumors were analyzed by WB (Left), and densitometry analysis was shown (Right). *p<0.05, **p<0.01.
Fig 8.
Low level of CDC42 K153 acetylation predicts poor overall survival of patients with CRC.
(A) Human CRC specimens were confirmed by HE staining, and the expression levels of CDC42 and CDC42 K153 acetylation and the localization of bacteria in tissues of 17 human CRC specimens were detected by IHC. CDC42 K153 acetylation level was significantly lower in the colorectal adenocarcinoma tissues than in the adjacent normal colorectal tissues as determined by IHC. The average of IHC intensity ± SD were quantitated by modified H-score from two groups of 17 patient samples (B) and 69 patients with different CRC stages (C) is shown. (D) Kaplan–Meier analysis of overall survival of 69 patients with CRC according to CDC42 K153 acetylation level. (E) Model of CDC42 K153 acetylation effect on colorectal tumorigenesis. Under physiological conditions, acetylated CDC42 K153 can maintain the normal physiological function of cells through MAPK phosphorylation (including ERK, JNK and p38) mediated by the CDC42-PAK4 signaling pathway. When Salmonella infects a host cell, SIRT2 is activated to deacetylate CDC42 K153, which causes an impaired binding of its downstream effector PAK4 and an attenuated phosphorylation of p38 and JNK, consequently reduces cell apoptosis. Moreover, low acetylation level of CDC42 K153 may contribute to the migration and invasion abilities of CRC cells, and promoting tumorigenesis, which may activate tumors mainly through the CDC42-PAK signaling axis.