LPCAT1 overexpression promotes the progression of hepatocellular carcinoma

Abstract

Background

Hepatocellular carcinoma (HCC) remains one of the most common malignant neoplasms. Lysophosphatidylcholine acyltransferase 1 (LPCAT1) plays a key role in the lipid remodelling and is correlated with various neoplasms. Nonetheless, the biological functions and molecular mechanisms of LPCAT1 underlying HCC remain obscure.

Methods

In the present study, we investigated the role of LPCAT1 in the progression of HCC. In-house RT-qPCR, tissue microarrays, and immunohistochemistry were performed to detect the expression levels and the clinical value of LPCAT1 in HCC. External datasets were downloaded to confirm the results. Proliferation, migration, invasiveness, cell cycle, and apoptosis assays were conducted to reveal the biological effects LPCAT1 has on SMMC-7721 and Huh7 cells. HCC differentially expressed genes and LPCAT1 co-expressed genes were identified to explore the molecular mechanisms underlying HCC progression.

Results

LPCAT1 showed upregulated expression in 3715 HCC specimens as opposed to 3105 non-tumour specimens. Additionally, LPCAT1 might be an independent prognostic factor for HCC. LPCAT1-knockout hampered cellular proliferation, migration, and metastasis in SMMC-7721 and Huh7 cells. More importantly, the cell cycle and chemical carcinogenesis were the two most enriched signalling pathways.

Conclusions

The present study demonstrated that increased LPCAT1 correlated with poor prognosis in HCC patients and fuelled HCC progression by promoting cellular growth, migration, and metastasis.

Background

Hepatocellular carcinoma (HCC), characterized by high molecular heterogeneity, serves as the predominant subtype of liver cancer [1, 2]. HCC is the third most frequent cause of cancer deaths worldwide, and ranks fifth and seventh in male and female malignancy [3], respectively. According to American Cancer Statistics 2021, there will be approximately 40,000 new cancer cases and 30,000 deaths per year, with a 5-year relative survival rate of only 20% [3]. HCC is a complex and multistage disease that may be explained by genetic and epigenetic changes [4,5,6,7,8]. Despite the development of surgery and combined treatment, as well as progress in early diagnosis and interventional therapy, most patients are still diagnosed in the advanced stage of HCC with a low 5-year survival rate and poor prognosis [9,10,11]. Therefore, clarifying the pathogenic mechanisms of HCC is essential for establishing novel potential therapeutic strategies and improving the prognosis of advanced stage HCC patients.

As a basic material of life, phospholipids play an important role in maintaining normal cellular structures, homeostasis, and fundamental cellular biological functions [12]. However, abnormal lipid metabolism disturbs cellular activities and drives cancer formation by affecting cell differentiation, proliferation, and apoptosis [13]. Lysophosphatidylcholine acyltransferase 1 (LPCAT1), encoded by LPCAT1, is a phospholipid biosynthesis/remodelling enzyme that plays a key role in the lipid remodelling pathway referred to as Lands’ cycle [14, 15]. LPCAT1 participates in phospholipid metabolism by catalysing the conversion from lysophosphatidylcholine to phosphatidylcholine and has been demonstrated to be correlated with various neoplasms, including hepatoma, acute myeloid leukaemia [13], breast cancer [16], clear cell renal cell carcinoma [17], colorectal cancer [18], esophageal cancer, lung adenocarcinoma [19], oral squamous cell carcinoma [20], and prostate cancer [21]. Thus far, only one study, limited in sample size, demonstrates that LPCAT1 is upregulated in 37 HCC tissue samples [22]. During the process of tumour initiation and progression, increased LPCAT1 promotes the synthesis and remodelling of lipids in the lipid-dependent membrane structures of rapidly proliferating cancer cells [16]. Additionally, LPCAT1 induces metastasis of cancer through complicated mechanisms, such as the lipid metabolism pathway and PI3K/ AKT/ MYC pathway [23]. Moreover, the alteration of the phospholipid component regulated by LPCAT1 promotes cell proliferation and membrane fluidity, thus causing the occurrence and development of clear cell renal cell carcinoma [17]. Collectively, such findings indicate the vital roles LPCAT1 play in tumour onset and progression. Nonetheless, the exact biological functions and precise molecular mechanisms of LPCAT1 underlying HCC remain obscure. More experiments must be conducted to determine the key roles of LPCAT1 in HCC and to design novel therapeutic targets.

In this study, the authors aimed to verify the expression of LPCAT1 in large numbers of clinical HCC tissue samples and to explore the biological functions of LPCAT1 in HCC cells. Furthermore, the potential molecular mechanisms and therapeutics of LPCAT1 underlying HCC were probed.

Methods

HCC-derived cell lines and tissue samples

HCC-derived SMMC-7721 and Huh7 cell lines were obtained from the Cell Bank of the Chinese Academy of Sciences, Shanghai, People’s Republic of China [24]. The cellular identity was confirmed by short tandem repeat profiles. All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS) (Sigma), as formerly reported [25].

To detect the mRNA expression level of LPCAT1, 204 HCC tissue specimens and 204 non-tumour hepatic tissue specimens were obtained from the Affiliated Cancer Hospital of Guangxi Medical University, Nanning, People’s Republic of China. The clinicopathological parameters and follow-up information of each patient were recorded. Informed consent was provided. To determine the protein expression level of LPCAT1 in HCC, a total of four tissue microarrays, containing LVC1505, LVC1531, LVC1601, and LVC1602, were purchased from Fanpu Biotech, Inc. (http://www.fanpu.com), Guilin, People’s Republic of China. A total of 315 HCC tissue specimens and 195 non-HCC tissue specimens were prepared to perform immunohistochemistry (IHC) staining. The inclusion criteria for patients in this study are as follows, (I) the pathological diagnosis should be primary HCC; (II) there should be sufficient tissue specimens for RNA extraction or IHC. To avoid confounding effect, HCC patient with intrahepatic cholangiocarcinoma was excluded in this study. The present research was authorised by the ethics committee of the First Affiliated Hospital of Guangxi Medical University.

In-house real-time reverse transcription-polymerase chain reaction, tissue microarrays, and immunohistochemistry

Fresh HCC tissue specimens and adjacent normal hepatic tissue specimens were prepared to extract total RNA by employing the AxyPrep Multisource Total RNA Miniprep Kit (AP-MN-MS-RNA-250G, AXYGEN, China). RNA was reversely transcribed into cDNA using the ‘Takara PrimeScript RT reagent’ kit (Takara, Nanning, People’s Republic of China, as we previously reported [26]. Beta-actin (ACTB) was used as the internal reference gene (forward sequence: “5′-CAGGCACCAGGGCGTGAT-3′”; reverse sequence: “5′-TAGCAACGTACATGGCTGGG-3′”) [27]. The primer of the target gene was designed on the mRNA sequence between the two gRNA binding sites (forward sequence: 5′-GTGACCATGACGATGTCCTCC-3′; reverse sequence: 5′-GTTCCCCAGATCGGGATGTC-3′). An algorithm of 2^−Δcq was used to determine the relative mRNA expression level of LPCAT1 [28,29,30].

The protein expression status of LPCAT1 in HCC was detected by conducting IHC, which was based on tissue microarrays. All the experimental operations concerning tissue microarrays and IHC staining follow the instructions of the manufacturer. The interpretation of staining intensity was performed by three expertized pathologists, according to the previously reported criteria [31].

In-silico data mining of HCC RNA sequencing and gene chip datasets

The mRNA expression profiles of HCC submitted before October 2020 were queried and filtered from public biomedical databases, including Gene Expression Omnibus, ArrayExpress, The Cancer Genome Atlas, Sequence Read Archive, The Genotype-Tissue Expression, and scientific literature. The inclusion criteria were as follows: (I) the subjects included in the studies should be Homo sapiens; (II) the patients should be diagnosed with HCC; (III) the platform information should be provided to annotate each probe. The exclusion criteria were as follows: (I) the subjects were animals; (II) valid data related to LPCAT1 expression was not presented; (III) duplicated data was reported; (IV) publications were in the form of case reports, reviews, and editorials. Three authors performed datasets screening independently, and an agreement was reached. A log₂(x + 1) transformation was considered if the expression matrices had not been normalized. The included datasets were classified according to their affiliated sequencing agencies. Those datasets that belonged to the same sequencing agency were integrated into a larger expression matrix, and the inter-batch differences were removed by using Limma-Voom and sva packages in R v.3.6.1 software. The batch-removed expression matrices were prepared for gene analysis.

In vitro cellular biological experiments

Construction of recombinant plasmids with the clustered regularly interspaced short palindromic repeats (CRISPR)/associated protein 9 (Cas9) system

We designed two gRNA sequences gRNA1: 5′-GGCGTCGAAGTAGGACGAGTG-3′ and gRNA2: 5′-GTCTGACCGGGACACGAACAC-3′. By synthesizing the oligo-DNAs of these gRNAs and annealing them, then cloning them into backbone vectors ‘pSpCas9(BB)-2A-Puro (PX459) V2.0’ (Plasmid #62988, addgene), two gRNA-expressing plasmids were formed [32].

Transfection with plasmids

HCC-derived cells (SMMC-7721 and Huh7) were stably transfected through lipo3000. After transfection for 24 h, the complete medium containing 2 μg/ml puromycin was screened for 48 h, and then the medium was transferred to the complete medium without puromycin to continue the culture. Transfected cells were used for further experiments.

Knockout effect verification

RT-qPCR was used to check the gene knockout efficiency. ACTB was used in combination as the internal reference gene. The primers of ACTB and the target gene were aforementioned in the “In-house real-time reverse transcription-polymerase chain reaction, tissue microarrays, and immunohistochemistry” section.

Cell viability

Proliferation assay was used to detect the effect of LPCAT1 knockout on cell viability. Total 2 × 10³ transfected cells were seeded in each well of a 96-well plate. The cells were counted every individual day and lasted for five days. The numbers of cells between the experimental group and the blank control group were compared, and the cell growth curve was drawn.

Migration assay

Cell migration ability was measured using a transwell cell in a 24-well plate (corning, Costar 3422). A total of 100 μl medium (0.5 × 10⁵ cells) was added to the upper chamber, and 500 μl of medium was added to the lower chamber. The mediums in the upper and lower chambers were DMEM containing 5% FBS. After incubation at 37 °C for 24 h, the cells on the upper surface of the chamber were gently wiped with cotton swabs [20], washed twice with phosphate buffered solution (PBS), fixed in methanol for 30 min, dried and dyed in 0.1% crystal violet for 30 min, and washed with PBS to remove the excess dye solution. The dyeing solution was observed under a microscope after drying. A total of five fields per chamber were subjected for the migrated cell counting.

Invasiveness assay

Cell invasiveness was measured using a transwell cell in a 24-well plate (corning, Costar 3422). Matrigel (BD, 356234) was diluted with serum-free DMEM. The ratio of Matrigel to DMEM was 1:8. 60 μl of the diluted Matrigel was added to the chamber. After incubation at 37 °C for 1 h, the chamber was taken out and the liquid was aspirated before DMEM was used to wash it twice. A total of 100 μl of serum-free DMEM (0.5 × 10⁵ cells) was added into the chamber, and 500 μl of DMEM containing 5% FBS was added into the lower chamber and cultured at 37 °C for 24 h. A total of five fields per chamber were subjected for the invaded cell counting.

Cell cycle and apoptosis

The cell cycle distribution and apoptotic rate of the LPCAT1-knockout group and control blank group were detected using a flow cytometer in SMMC-7721 and Huh7 cells.

Prospective molecular mechanisms of LPCAT1 in HCC

HCC differential expression genes (DEGs, |log₂FoldChange|> 1 and P < 0.05) were initially filtered from the batch-removed expression matrices by using the Limma package in R software v3.6.1. Standardized mean differences (SMD, |SMD|>0 and P < 0.05) were then quantitively calculated to identify DEGs, which were more reliable [33]. LPCAT1 co-expressed genes (CEGs, Pearson Correlation Coefficient, |PCC|≥ 0.3 and P < 0.05) were filtered from the batch-removed expression matrices. Specifically, one gene with a PCC value ≥ 0.3 were defined as positive CEGs, while genes with a PCC value ≤ − 0.3 were defined as negative CEGs. DEGs and CEGs were intersected to analyse the clustered Gene Ontology (GO) terms, Disease Ontology (DO), reactome, and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways by using the clusterprofiler package. The Search Tool for the Retrieval of Interacting Genes (STRING) database was used to analyse protein–protein interaction (PPI) [34]. Functional module analysis was conducted by using an algorithm of vertex-weighted MCODE in Cytoscape v3.6.1 software [35]. The hub gene in the PPI network was identified using Cytoscape. The genetic alterations and mutations of specific genes enriched in the first ranked KEGG pathways were analysed using the maftools package and cBioportal [36].

Potential small molecule agents for HCC treatment

Although increasing novel cancer targets are reported, their practical implications are abrogated by poor druggability of such targets [37]. However, a connectivity map (Cmap) provides an opportunity to repurpose Food and Drug Administration-approved drugs according to specific gene signatures, which is a potential strategy for precision treatment of cancer [38,39,40]. By targeting at the gene regulatory network of the DEGs that were co-expressed with LPCAT1 in HCC, several small molecule drugs for HCC treatment were forecasted.

Data analysis

All the data were analysed under IBM Statistical Product and Service Solutions Statistics v19.0, STATA v12.0, and R v3.6.1 software. An independent samples t-test was used to compare the LPCAT1 expression data in HCC tissues and non-tumour hepatic tissues. To combine SMD and hazard ratio (HR), a fixed-effect model was first considered when the I² < 50%. A randomized-effect model should be used to handle the significant heterogeneity (I² ≥ 50%). Sensitive analysis was adopted to detect the source of heterogeneity. Begg’s funnel plot was used to explore publication bias. The receiver operating characteristic curve (ROC) was plotted to determine the sensitivity and specificity. Summary receiver operating characteristic (SROC) curves were utilized to describe the area under the curve (AUC), pooled sensitivity, and pooled specificity. The likelihood ratio was used to determine the discriminatory ability of LPCAT1 between HCC and non-HCC tissues. Kaplan–Meier analysis was performed to assess the prognostic value of LPCAT1. P < 0.05 signified significance.

Results

Increased LPCAT1 mRNA expression in HCC tissues

The result of in-house RT-qPCR indicated that LPCTA1 mRNA was dominantly increased in 204 HCC tissue specimens as opposed to 204 non-tumour liver tissue specimens (Fig. 1a; Additional file 1: Table S1). Increased LPCAT1, however, possessed a weak ability to discriminate between HCC and normal liver tissues (Fig. 1b), which was due to a relatively small sample size. Eight out of ten external datasets confirmed the upregulated LPCAT1 expression level (Additional file 2: Figure S1), where LPCAT1 showed a moderate or strong ability to discriminate between HCC and normal liver tissues in four subsets. The basic information of such datasets was presented in Additional file 1: Table S1 and Additional file 3: Table S2. Collectively, a combined SMD value of 0.63 [0.35, 0.90] integrated in-house RT-qPCR results and external datasets and verified the increased mRNA expression levels of LPCAT1 in 3715 HCC tissue samples in comparison with 3105 non-HCC tissue samples (Additional file 4: Figure S2a). A randomized-effects model was utilized to handle the significant heterogeneity (P < 0.01). The result of the sensitive analysis failed to identify a heterogeneity source (Additional file 4: Figure S2b). No publication bias was detected (Additional file 4: Figure S2c). The result of SROC indicated a moderate ability of LPCAT1 in differentiating between HCC and non-HCC (AUC = 0.80 [0.77, 0.84]; sensitivity = 0.59 [0.42, 0.74]; specificity = 0.85 [0.74, 0.92]; combined diagnostic likelihood ratio positive = 3.98 [2.32, 6.82]; and combined diagnostic likelihood ratio negative = 0.48 [0.34, 0.69]) (Additional file 5: Figure S3a–c).

Fig. 1

Upregulation of LPCAT1 mRNA in hepatocellular carcinoma based on in-house real-time reverse transcription-polymerase chain reaction. a LPCAT1 mRNA expression levels are markedly increased in hepatocellular carcinoma tissues. b Increased LPCAT1 possessed a weak ability to discriminate between HCC and normal liver tissues. ****P < 0.0001

Increased LPCAT1 protein expression level in HCC tissues

As opposed to normal liver tissue specimens (Fig. 2a–b), LPCAT1 protein expression levels were markedly increased in the HCC tissue specimens (Fig. 2c–d), which was consistent with LPCAT1 mRNA expression. Increased LPCAT1 protein expression possessed a strong ability to discriminate between HCC and normal liver tissues (Fig. 2e–f).

Fig. 2

Upregulation of LPCAT1 protein in HCC based on immunohistochemistry. Compared with normal liver tissues. a, b LPCAT1 protein expression levels are markedly increased in HCC tissues (c, d) (magnification × 40 and × 400). e, f Increased LPCAT1 protein expression possessed a strong ability to discriminate between HCC and normal liver tissues (P < 0.0001). HCC, hepatocellular carcinoma; ****P < 0.0001

Poor prognosis of HCC patients with LPCAT1 upregulation

The result of Kaplan–Meier survival analysis indicated that higher LPCAT1 expression forecasted worse overall survival outcomes in HCC patients (Fig. 3a). Increased LPCAT1 expression patterns varied from gender (male: 3.28 ± 3.10 vs. female: 1.73 ± 1.66; P = 0.001), age (< 60: 3.33 ± 3.22 vs. ≥ 60: 2.17 ± 1.65; P = 0.002) and microvascular tumour thrombus (with thrombus: 3.50 ± 3.20 vs. without thrombus: 2.64 ± 2.71; P = 0.040) (Fig. 3b–d). External HCC cohorts were used to confirm the prognostic value of LPCAT1 in HCC (Additional file 6: Figure S4a). The GSE10143 cohort was probed to be the primary source of heterogeneity and had been removed (Additional file 6: Figure S4b). Collectively, a combined HR value of 2.21 [1.64, 2.99] indicated that LPCAT1 might serve as an independent risk factor for HCC (Additional file 6: Figure S4c).

Fig. 3

Increased expression of LPCAT1 is correlated with poor prognosis of HCC patients. a Higher LPCAT1 expression predicts worse overall survival outcomes. b–d Among HCC patients, elevated expression levels of LPCAT1 were significantly different in terms of gender, age, and microvascular tumor thrombus. ***P < 0.001; **P < 0.01; *P < 0.05. HCC, hepatocellular carcinoma

Establishment of LPCAT1 knockout HCC cells

To reveal the cellular biological functions of LPCAT1, LPCAT1 knockout cells were successfully established and were transfected into two HCC-derived cell lines (i.e., SMMC-7721 and Huh7) [20]. The result of the RT-qPCR verification experiment indicated lower LPCAT1 mRNA expression levels in LPCAT1 knockout SMMC-7721 cells (P = 4.0E–05) and LPCAT1 knockout Huh7 cells (P = 3.3E–06) than those in the control group (Fig. 4a–b).

Fig. 4

LPCAT1 induced the proliferation of Huh7 and SMMC-7721 cells. The LPCAT1-knockout Huh7 and SMMC-7721 cells were constructed by using the clustered regularly interspaced short palindromic repeats (CRISPR)/associated protein 9 (Cas9) system. a, b Validated the transfection efficiency of LPCAT1-knockout group and control group in Huh7 and SMMC-7721 cell lines, respectively. c, d Proliferation assay was conducted to evaluate the effect of LPCAT1-knockout on cellular proliferation. e–j Flow cytometry was used to detect the cycle distribution and apoptosis of LPCAT1-knockout SMMC-7721 cells. ****P < 0.0001; ***P < 0.001; **P < 0.01

Inhibited proliferation, migration, and invasion abilities in LPCAT1 knockout HCC cells

It was found that compared with the blank control group, LPCAT1 knockout significantly hampered cellular growth in SMMC-7721 and Huh7 cells (Fig. 4c–d). Therefore, LPCAT1 might promote the proliferation of HCC cells. Compared with the LPCAT1 non-knockout transfected cells in the two cell lines, there was a significant decrease in the number of penetrating LPCAT1 knockout SMMC-7721 and Huh7 cells (Figs. 5, 6). Therefore, LPCAT1 might enhance the migration and invasion abilities of SMMC-7721 and Huh7 cells.

Fig. 5

The ability of cell migration to LPCAT1-knockout group and control group in SMCC-7721 and Huh7 cell lines. LPCAT1 might promote the migration of SMCC-7721 and Huh7 cells. ****P < 0.0001; ***P < 0.001

Fig. 6

The ability of cell invasion to LPCAT1-knockout group and control group in SMCC-7721 and Huh7 cell lines. LPCAT1 might promote the invasion of SMCC-7721 and Huh7 cells. ****P < 0.0001

Cell cycle and apoptosis of LPCAT1 knockout cells

It was observed that compared with the distribution of 56.45% cells in the control group, about 53.32% of LPCAT1 knockout cells were distributed in the G0/ G1 period. Additionally, about 31.16% of LPCAT1 knockout cells were distributed in the S period, higher than that of 28.45% cells in the control group (Fig. 4e–g). Therefore, LPCAT1 knockout might arrest the cell cycle at the S phase. Additionally, the apoptotic rates of LPCAT1 knockout SMMC-7721 cells slightly increased when compared with the blank control group (Fig. 4h–j). However, the statistical significance cannot be determined because of a lack of duplicate samples.

Prospective signal transduction pathways of LPCAT1 in HCC

A total of 5399 HCC DEGs were identified, containing 3736 overexpressed DEGs and 1663 under expressed DEGs, all with significant SMD. Additionally, a total of 1223 LPCAT1 CEGs were determined, including 564 positive CEGs and 659 negative CEGs. Upregulated DEGs were intersected by positive CEGs, and 473 genes were obtained whose expression level was increased in HCC and were positively correlated with LPCAT1. Downregulated DEGs were intersected by negative CEGs, and 367 genes were obtained. Enrichment analysis indicated that the upregulated LPCAT1 CEGs were dominantly clustered in the cell cycle pathway according to the KEGG (Additional file 7: Figure S5a) and reactome (Additional file 7: Figure S5b). Additionally, embryonal cancer was determined to be the most enriched disease (Additional file 7: Figure S5c). Moreover, mitotic nuclear division, cell-substrate adherens junction, and cadherin binding were found to be the most prominent “Biological Process,” “Cellular Component,” and “Molecular Function,” respectively (Additional file 8: Figure S6). On the other hand, the downregulated LPCAT1 CEGs were primarily clustered in chemical carcinogenesis (Additional file 9: Figure S7a) and biological oxidations (Additional file 9: Figure S7b). Accordingly, inherited metabolic disorder was found to be the most associated disease (Additional file 9: Figure S7c). Furthermore, small molecule catabolic process, plasma lipoprotein particle, and monooxygenase activity were identified as the most aggregated GO terms according to the downregulated genes negatively related to LPCAT1 in HCC (Additional file 10: Figure S8). The mutation oncoplot of the cell cycle and chemical carcinogenesis is presented in Additional file 11: Figure S9a–b.

Construction of the genetic regulatory network of LPCAT1 in HCC

LPCAT1’s genetic regulatory network in HCC was constructed based on the intersected genes that had been described in the “Prospective signal transduction pathways of LPCAT1 in HCC” section (Fig. 7). Two functional modules (i.e., Cluster 1 and Cluster 2) were identified for the upregulated CEGs and downregulated CEGs, respectively. PPI of the cell cycle and chemical carcinogenesis pathways were explored further. CCNB1, which was significantly correlated with LPCAT1 (Spearman R = 0.58, P = 2E–48), was established as the hub gene in the cell cycle pathway. Moreover, a total of three core genes were established in the chemical carcinogenesis pathway, containing UGT2B10 (Spearman R = –0.50, P = 3.2E–34), GSTA2 (Spearman R = –0.30, P = 2.4E–12), and CYP1A1 (Spearman R = – 0.37, P = 3.1E–18).

Fig. 7

The genetic regulatory network of LPCAT1’s differentially co-expressed genes in hepatocellular carcinoma. Gene in blue and red represent downregulation and upregulation, respectively

Potential therapeutic agents for HCC by targeting at LPCAT1’s regulatory network

By constructing the gene regulatory network based on the DEGs related to LPCAT1 in HCC, several potential therapeutic agents for HCC were identified by targeting at LPCAT1’s regulatory network (Table 1). The molecular formulas of such drugs are presented in Additional file 12: Figure S10. Surprisingly, the anti-HCC effects of LY-294002 [41], trichostatin A [42], sirolimus [43], vorinostat [44], and harmine [45] have been demonstrated by in vitro or in vivo experiments.

Table 1 Potential therapeutic agents for hepatocellular carcinoma by targeting LPCAT1’s regulatory network

Discussion

Lipid metabolisms have been verified to play profound roles in the onset and deterioration of HCC through lipid metabolic reprogramming and metabolism rearrangements [46,47,48]. Mounting evidence demonstrated that LPCAT1 participates in the pathophysiological process of cancer by dysregulating lipid metabolisms [16, 48]. In this study, the authors found that increased LPCAT1 induced the proliferation, migration, and metastasis of SMMC-7721 and Huh7 cells. A regulatory network of LPCAT1 was constructed to analyse the molecular mechanisms behind the initiation and progression of HCC. Several small molecule drugs were forecasted by targeting at the network.

There was only one small study that elucidated the upregulation of LPCAT1 in HCC [22]. By contrast, the present study highlights the biological functions of LPCAT1 and verifies its upregulated expression patterns in 3715 HCC specimens as opposed to 3105 non-HCC specimens. Furthermore, the prognostic value of LPCAT1 was comprehensively assessed by in-house and external HCC cohorts. More importantly, the cell cycle and chemical carcinogenesis were identified as the two most enriched signalling pathways. Potential drugs targeting at LPCAT1 in HCC were predicted.

LPCAT1 mRNA and protein were abundant in HCC tissues. In-house RT-qPCR, immunohistochemistry, external RNA sequencing, and gene chip were used to confirm the increased expression levels of LPCAT1 in HCC tissues. A SMD index was used to combine the effect size, which further supported the upregulation of LPCAT1 (SMD = 0.63 [0.35, 0.90]). Therefore, the results of the present study are more convincing. Additionally, increased LPCAT1 presages poor prognosis of HCC patients. Higher LPCAT1 was correlated with worse overall survival conditions in 125 HCC patients, suggesting LPCAT1 might be a potent risk factor in HCC (combined HR = 2.21 [1.64, 2.99]). Also, LPCAT1 expression shows an ascending pattern in male, age < 60, and microvascular tumour thrombus HCC patients. Collectively, it is conceivable that upregulated LPCAT1 might play crucial roles in the unfavourable prognosis of HCC.

LPCAT1 might promote HCC progression by accelerating cell growth, migration, and metastasis. LPCAT1 knockout SMMC-7721 and Huh7 cells were established using the CRISPR/Cas9 system [49]. The proliferation, migration, and metastasis abilities of LPCAT1 knockout transfected SMMC-7721 and Huh7 cells were significantly hampered, thus indicating that LPCAT1 might participate in the deterioration of HCC cells by promoting the mechanisms of cell growth, migration, and metastasis. This hypothesis can be supported by a previous study, where only one HCC-derived cell line was used [22]. Interestingly, another study reported that LPCAT1 is required for the miRNA-205-induced proliferation of SMMC-7721 cells [50]. Furthermore, LPCAT1 knockout might arrest the cell cycle at the S phase and increase the cell apoptotic rates in SMMC-7721 cells, which was reported for the first time. Therefore, the authors speculate that LPCAT1 might induce the DNA synthesis process and abrogate cell apoptosis to survive HCC cells. However, the influence LPCAT1 has on the cell cycle distribution and the apoptotic rate of HCC cells has not been verified by duplicate samples. More experimental evidence must be supplemented in the future to support the findings in the present study.

Since LPCAT1 was important in HCC initiation and progression, the authors sought to shed light on its potential molecular mechanisms underlying HCC. A genetic regulatory network of LPCAT1 was constructed, where upregulated LPCAT1 CEGs were prominently clustered in the cell cycle pathway. As is known, aberrant regulation of the cell cycle induces uncontrollable cell proliferation and handles various cancers, including HCC [51,52,53]. The result of flow cytometry unveils an intimate association between LPCAT1 and cell cycle regulation; however, the precise regulatory mechanisms had never been studied before. In this study, upregulated LPCAT1 CEGs were found to participate in mitotic nuclear division, regulation of mitotic cell cycle phase transition, cell cycle checkpoints, and mitotic G1-G1/S phases in terms of GO and reactome. During mitotic nuclear division, the primary event of the M phase, mitotic genes are prone to be activated when the chromatin structures change, which creates favourable conditions for the proliferation of HCC cells [54]. Interestingly, CCNB1, the hub gene in the cell cycle pathway, encodes an important regulatory protein in mitosis (cyclin B1). Increased cyclin B1, prominently working as a “switch” in the process of cell cycle, results in malignant proliferation of HCC cells by binding to cyclin-dependent kinases [55]. Furthermore, there was a positive association between the mRNA expression levels of LPCAT1 and CCNB1 (Spearman R = 0.58, P = 2E–48), indicating that there might be a synergetic effect between LPCAT1 and CCNB1 on the pathogenesis of HCC. Additionally, the cell cycle checkpoint is verified to be one of the most important cell cycle regulatory mechanisms [56]. Nonetheless, dysregulation of the cell cycle checkpoint might lead to radio resistance of HCC cells, thus causing unfavourable prognosis in HCC patients [57]. Therefore, the upregulated LPCAT1 CEGs might induce abnormal regulation of the cell cycle and promote the development and deterioration of HCC. On the other hand, downregulated LPCAT1 CEGs provide important clues on metabolism mechanisms, which might propel chemical carcinogenesis. Downregulated LPCAT1 CEGs are involved in small molecule catabolic processes, plasma lipoprotein particles, and biological oxidations. Surprisingly, increased LPCAT1 is likely to induce metabolic disturbance and fuel the process of liver damage and carcinogenesis. In liver injury mouse models, LPCAT1 was upregulated in the carbon tetrachloride-treated mouse, along with increased saturated fatty acylphospholipid in liver tissues and peripheral plasma [58]. Furthermore, another study verified the lipid composition in Huh7 cells by using mass spectrometry imaging [22]. Collectively, it was conceivable that upregulated LPCAT1 might disturb the metabolism of lipids and promote liver carcinogenesis by co-expressing with downregulated CEGs. More studies are required to further confirm our hypothesis.

Moreover, several therapeutic drugs were successfully predicted by targeting at the genetic regulatory network of LPCAT1. As is known, the clinical efficiency of HCC patients was largely limited by chemotherapy resistance and liver dysfunction. It is urgently needed to develop more potent agents with hepatoprotective and anti-HCC properties [59]. In the present study, LY-294002, trichostatin A, sirolimus, vorinostat, and harmine were forecasted to be potential small molecule drugs for treating HCC. LY-294002, an inhibitor of phosphatidylinositol 3-kinase, has proved useful in hampering HCC cell growth, migration, and invasion [60]. Additionally, histone deacetylase inhibitor trichostatin A has seen anti-HCC efficacy by activating cell death signalling cascades [61]. More importantly, a single-centre study demonstrated the favourable effect sirolimus has on HCC patients, whose survival time were longer than in the control group [62]. Therefore, such evidence indicates that repurposing FDA-approved small molecule drugs is a promising direction for HCC treatment. In vivo experiments and clinical trials will promote the applications of these small molecule drugs.

There are several limitations. First, a high level of heterogeneity was detected in the present study. The author used a randomized-effect model to minimize the influence of heterogeneity. In the future, more single cell RNA sequencing experiments will help to reduce intercellular heterogeneity. Second, the influence LPCAT1 has on the cell cycle distribution and the apoptotic rate of HCC cells has not been verified by different cell lines. Third, the complicated molecular mechanisms and practical effects of the predicted small molecule drugs are unclear. However, the present study sheds light on several novel signalling pathways and reveals potential therapeutic directions for HCC. Further studies are needed to validate the conclusions.

Conclusions

LPCAT1 was upregulated and correlated with poor prognosis in HCC patients. Increased LPCAT1 fuelled HCC progression by promoting cell proliferation, migration, and metastasis. LPCAT1 co-expressed genes and HCC differentially expressed genes prominently participated in the cell cycle and chemical carcinogenesis pathways.