The prognostic role of ACSL4 in postoperative adjuvant TACE-treated HCC: implications for therapeutic response and mechanistic insights

Abstract

Background: The response of hepatocellular carcinoma (HCC) to transarterial chemoembolization (TACE) treatment and its underlying mechanisms remain elusive. This study investigates the role of enzymes involved in fatty acid activation, specifically Acyl-CoA synthetase long chain 4 (ACSL4), in HCC patients treated with postoperative adjuvant TACE (PA-TACE) and in nutrient-deprived HCC cells.

Methods: We examined the expression of ACSL4 and its family members in HCC clinical samples and cell lines. The clinical significance of ACSL4, particularly regarding the prognosis of patients treated with PA-TACE, was assessed using two independent HCC cohorts. We further explored the role of ACSL4 in glucose starvation-induced cell death in HCC cells and xenograft mouse models.

Results: Among the family members, ACSL4 is the most up-regulated enzyme, associated with poor survival in HCC patients, particularly in post-recurrent TACE-treated patients in a Singapore cohort. ACSL4 is essential for HCC cell survival in response to glucose starvation, rather than to hypoxia or to the combination of hypoxia with doxorubicin or cisplatin. ACSL4-mediated arachidonic acid (AA) metabolism supports mitochondrial β-oxidation and energy production. CCAAT/enhancer binding protein α (CEBPA) transcriptionally regulates ACSL4 by binding 3 motifs (-623 to -613, -1197 to -1187 and -1745 to -1735) of ACSL4 upstream promoter region, enhancing its pro-survival effects. Furthermore, canagliflozin (Cana), a clinical-approved drug for type 2 diabetes, mimics glucose starvation and inhibits the growth of ACSL4-low xenograft tumors. Moreover, high ACSL4 or CEBPA expressions correlate with increased recurrence susceptibility after PA-TACE in the China-Guangxi HCC cohort.

Conclusions: The CEBPA-ACSL4 pathway is critical in protecting HCC cells from glucose starvation-induced cell death, suggesting that ACSL4 and CEBPA could serve as valuable prognostic indicators and potential therapeutic targets in the context of PA-TACE treatment for HCC.

Keywords: ACSL4; Canagliflozin; Glucose starvation; HCC; TACE.

Fig. 1

ACSL4 up-regulation in HCC was associated with poorer survival particularly among those recurrent patients who had TACE as post-recurrent treatment. a Expression of ACSL and ACSS family members were determined by qRT-PCR in HCC and adjacent normal liver tissues (N = 20). b ACSL4 was determined by IHC in tissue microarrays collected from the Singapore HCC cohort (N = 170). The expression levels were divided according to a four-tiered grading system. c A pie chart of the difference of ACSL4 protein expression between non-tumor (NT) and HCC tumor tissue (Tu). d A Kaplan–Meier survival comparison between ACSL4-present and ACS4-abssent HCC patients. e–f Kaplan–Meier survival comparisons between ACSL4-present and ACSL4-absent recurrent HCC patients who had TACE/TAE as post-recurrent treatment (e) or who had other treatments except TACE/TAE (f). ***, P < 0.001

Fig. 2

ACSL4 protected HCC cells against glucose starvation induced cell death. a ACSL1, ACSL3 and ACSL4 expression in 7 HCC cell lines. b The indicated HCC cells were treated with full medium (Full), glucose & glutamine double deprivation (DN), DN with 4 mM glutamine (Gln) or DN with 4.5 g/L glucose (Glu) for 24 h. The proportion of survival cells were determined by PI assay. c Huh7 cells were treated with ACSL4 inhibitor triacsin C (2 μM) or rosiglitazone (20 μM) in full medium or glucose starvation for 24 h. The proportion of survival cells is shown in the right panel. d ACSL4-knockout Huh7 cells (ACSL4-sg1, -sg2, upper panel) were treated with full medium or glucose starvation for 60 h. The proportion of survival cells is shown in the lower panel. e ACSL4-overexpressing N1S1 cells were treated with full medium or glucose starvation for 36 h. f-g ACSL4-knockout huh7 cells were treated with cisplatin (f) or doxorubicin (g) with or without hypoxia (Hpx) for 36 h or 24 h respectively. ###, compared with negative control (NC), P < 0.001; **, P < 0.01; ***, P < 0.001; NS, not significant

Fig. 3

ACSL4-KO obstructed maintenance of cellular ATP levels by inhibiting β-oxidation under glucose starvation. a ACSL4-KO Huh7 cells were treated with glucose starvation for 48 h. Then cellular ATP were detected. b ACSL4-KO Huh7 cells were treated with glucose starvation for 12 h, mitochondrial membrane potential was detected using JC-1 dye by flow cytometry. The ratios of red/green florescence were shown. c-e ACSL4-KO Huh7 cells were treated with glucose starvation for 24 h. The cellular lipid droplets were stained with oil red O dye (c) or Nile red dye (d), Nile red fluorescence was assessed by flow cytometry. The cellular triglycerides were detected (e). f ACSL4-KO Huh7 cells were treated with glucose starvation for 48 h. Then cellular β-hydroxybutyrate (β-HB) were detected. ###, compared with NC, P < 0.001; ***, P < 0.001

Fig. 4

ACSL4 switched on the utilization of AA in β-oxidation to supply energy. a-c ACSL4-KO Huh7 cells and negative control cells were treated with glucose starvation medium with/without BSA or different BSA-fatty acid conjugates (AA-CoA, AA, LA, PA, OA, SA; 25 μM for each) for 60 h. Then the proportion of survival cells (a), cellular ATP (b) and β-HB (c) were detected. d ACSL4-KO Huh7 cells were treated with glucose starvation medium supplied with AA-CoA (25 μM) and/or etomoxir (ETO, 40 μM). The proportion of survival cells were detected. e–h ACSL4-KO Huh7 cells were treated with 25 μM AA or BSA after incubated in glucose-limited medium for 12 h. Then mitochondrial OCRs were determined by a Seahorse XF analyzer (e). The basal OCR (f), ATP-linked OCR (g) and maximal OCR (h) were summarized, respectively. Abbreviations: BSA, bovine serum albumin; AA-CoA, arachidonoyl coenzyme A; AA, arachidonic acid; LA, linoleic acid; OA, oleic acid; PA, palmitic acid; SA, stearic acid. ETO, etomoxir. # compared with NC, P < 0.05; ##, compared with NC, P < 0.01; ###, compared with NC, P < 0.001; **, P < 0.01; ***, P < 0.001; NS, not significant

Fig. 5

Canagliflozin inhibited ACSL4-KO HCC cells by mimicking glucose restriction in vitro and in vivo. a-b The 3D Spheroid tumor developed by indicated cells were treated with glucose starvation (a) or canagliflozin (Cana, 20 μM) in glucose-present (0.9 g/L) medium (b) for 60 h. The dead cells were stained with PI (red fluorescence) while live cells were stained with calcein (green fluorescence) before observation under a fluorescent microscopy (left panel). The volume of spheroid (middle panel) and cell death (the ratio of red to green, right panel) were measured. c The mice bearing xenograft developed from ACSL4-KO or control Huh7 cells were treated with control PBS or Cana (30 mg/kg) for 33 days once every other day. The resected tumors were photographed (left panel). The tumor volumes (middle panel) and tumor weights (right panel) were measured. d-f Representative images of hematoxylin & eosin staining (HE), IHC for ACSL4 and Ki67 are shown (d). The proportion of area of necrotic tissue (e) and the Ki-67 positive score (f) were summarized. ##, compared with NC, P < 0.01; ###, compared with NC, P < 0.001; ***, P < 0.001. NS, not significant

Fig. 6

ACSL4 is positively regulated by transcriptional factor CEBPA. a Correlation of mRNA expression between CEBPA and ACSL4 in HCC (Timer 2.0 online database, http://timer.cistrome.org/). b Correlation of mRNA expression between CEBPA and ACSL4 in 6 HCC cells. c CEBPA-knockout Huh7 cells (CEBPA-sg1, CEBPA-sg2) and its negative control (NC) were treated with glucose starvation for 24 h. The proportion of survival cells is shown in the right panel. d Indicated cells were treated with glucose starvation for 12 h, the cellular lipid droplets were stained with Nile red dye. e CEBPA-sg1 cells (CEBPA-KO) were transfected with ACSL4-overexpression (ACSL4-OE) lentivirus, then were treated with glucose starvation for 24 h. The proportion of survival cells is shown in the right panel. f Putative motifs of ACSL4 promoter binding with CEBPA predicted by JASPAR database (https://jaspar.elixir.no/). Indicated cells were transfected with a variety of truncated ACSL4 promoter constructs, and the luciferase activity normalized by Renilla luciferase activity were determined. g Indicated cells were transfected with single or simultaneous deletion mutated ACSL4 promoter plasmids at motif 1, 3, 7, the relative luciferase activity was determined. h-i ChIP-qPCR assays to measure the enrichment of ACSL4 promoter on CEBPA or IgG in Huh7 cells (h). PCR products were detected by agarose gel electrophoresis (i)

Fig. 7

Higher ACSL4 and CEBPA expression in HCC was associated with higher recurrence of HCC patients with TACE treatment. a ACSL4 protein expression (measured as the average optical density (AOD) using ImageJ software) and CEBPA positivity (the ratio of CEBPA-positive cells to all cells) in HCC (TU) and adjacent nontumor (NT) tissue samples of Guangxi cohort (n = 80) were summarized on the middle and right panel respectively. The representative pictures were shown in the left panel, CEBPA-positive cells were marked using red arrow. b The correlation between ACSL4 and CEBPA in expression of HCC tissue samples in Guangxi cohort. c-d Comparison of recurrence between ACSL4-High and ACSL4-Low HCC patients (c) or CEBPA-High and CEBPA -Low HCC patients (d) who received PA-TACE treatment