Dysregulation of genes involved in the long-chain fatty acid transport in pancreatic ductal adenocarcinoma

Abstract

BACKGROUND

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive lethal malignancy with limited options for treatment and a 5-year survival rate of 11% in the United States. As for other types of tumors, such as colorectal cancer, aberrant de novo lipid synthesis and reprogrammed lipid metabolism have been suggested to be associated with PDAC development and progression.

AIM

To identify the possible involvement of lipid metabolism in PDAC by analyzing in tumoral and non-tumoral tissues the expression level of the most relevant genes involved in the long-chain fatty acid (FA) import into cell.

METHODS

A gene expression analysis of FASN, CD36, SLC27A1, SLC27A2, SLC27A3, SLC27A4, SLC27A5, ACSL1, and ACSL3 was performed by qRT-PCR in 24 tumoral PDAC tissues and 11 samples from non-tumoral pancreatic tissues obtained via fine needle aspiration or via surgical resection. The genes were considered significantly dysregulated between the groups when the p value was < 0.05 and the fold change (FC) was ≤ 0.5 and ≥ 2.

RESULTS

We found that three FA transporters and two long-chain acyl-CoA synthetases genes were significantly upregulated in the PDAC tissue compared to the non-tumoral tissue: SLC27A2 (FC = 5.66; P = 0.033), SLC27A3 (FC = 2.68; P = 0.040), SLC27A4 (FC = 3.13; P = 0.033), ACSL1 (FC = 4.10; P < 0.001), and ACSL3 (FC = 2.67; P = 0.012). We further investigated any possible association between the levels of the analyzed mRNAs and the specific characteristics of the tumors, including the anatomic location, the lymph node involvement, and the presence of metastasis. A significant difference in the expression of SLC27A3 (FC = 3.28; P = 0.040) was found comparing patients with and without lymph nodes involvement with an overexpression of this transcript in 17 patients presenting tumoral cells in the lymph nodes.

CONCLUSION

Despite the low number of patients analyzed, these preliminary results seem to be promising. Addressing lipid metabolism through a broad strategy could be a beneficial way to treat this malignancy. Future in vitro and in vivo studies on these genes may offer important insights into the mechanisms linking PDAC with the long-chain FA import pathway.

Key Words: Carcinoma; Pancreatic ductal; Fatty acid transport; Gene expression; Biomarkers

Core Tip: In this original article, we show preliminary results of a case-control study in which we analyzed the expression level of nine relevant genes involved in the long-chain fatty acid (FA) import in pancreatic ductal adenocarcinoma (PDAC). We found that three FA transporters (SLC27A2, SLC27A3, and SLC27A4) and two long-chain acyl-CoA synthetases genes (ACSL1 and ACSL3) were significantly upregulated in the PDAC tissue compared to the non-tumoral tissue. These data suggest that addressing lipid metabolism through a broad strategy that may impact both tumor cells and the tumor microenvironment could be a beneficial way to treat this malignancy.

INTRODUCTION

Lipids are essential biological components that function as building blocks for membranes and are involved in energy metabolism, signaling cascades, and cell transport. The preservation of lipid homeostasis is essential for life, and prolonged exposure to excess lipids may have harmful effects and lead to serious lipid-related illnesses[1]. In particular, reprogramming of lipid metabolism is a characteristic of different types of cancer[2], as tumor cells can increase de novo lipogenesis, fatty acid (FA) absorption, and FA oxidation represent mechanisms to produce energy and accumulate lipids for the synthesis of plasma membranes[3].

Dysregulated FA metabolism is driven by specific carcinogenic signaling pathways, including the B-Raf kinase and the epidermal growth factor receptor, and affects the composition and saturation of lipid membranes[4]. This, in turn, modulates cancer cell tolerance to reactive oxygen species, enabling their survival, invasion, and metastasis[5]. Consequently, targeting lipid metabolism using a multifaceted strategy that can affect tumor cells as well as the tumor microenvironment would appear to represent a valuable approach to treat cancer[6]. Notably, alterations in cellular lipid metabolism also play a significant role in tumor drug resistance[7].

One of the most lethal cancers is pancreatic ductal adenocarcinoma (PDAC), an aggressive solid tumor with a 5-year survival rate of 11% in the United States. The poor survival rate is attributed to several characteristics, including aggressive tumor biology, late symptom presentation, tumor localization, difficult surgical care, and an absence of effective systemic drugs[8]. When adjusting for age, countries with the highest projected growth in the number of individuals above 65-years-old are the same as those with the highest rates of pancreatic cancer. Unfortunately, only approximately 15%-20% of patients are diagnosed with surgically resectable PDAC[9]. Thus, as the lifespan increases worldwide, the overall burden of pancreatic cancer will also increase. Current estimates indicate that pancreatic cancer will become the second leading cause of cancer-related death after lung cancer.

A growing number of studies now indicate a relationship between PDAC and dysregulated lipid metabolism[10-13]. Specialized transporters, such as CD36 or the FA transport protein family [FATPs; often referred to as solute carrier protein family 27 (SLC27)], are necessary for the effective passage of exogenous FAs across the plasma membrane[14].

CD36 (also known as FA translocase) is a cell surface scavenger receptor involved in metastasis, immune evasion, and drug resistance[15]. Kubo et al[16] identified elevated CD36 expression as a major adverse prognostic factor in PDAC that affects gemcitabine resistance by influencing anti-apoptosis protein activity. Conversely, a low expression of CD36 in PDAC has been found to be associated with low Tumor, Node, Metastasis (TNM) staging and CA19-9 levels, even in the presence of a large tumor size and a poor prognosis for survival[17].

Another factor that affects several aspects of PDAC is SLC27. It consists of a family of six members, from SLC27A1 through SLC27A6, that are involved in the uptake of long-chain FAs. These transporters are also implicated in the emergence of several malignancies, as they can modify FA metabolism, cell growth, and cell proliferation[18]. For example, while normal cells obtain FAs exogenously, tumor cells obtain FAs both exogenously and by de novo synthesis via FASN[19]. FASN functions as a major regulator of lipid metabolism, is essential for the development and survival of tumors exhibiting lipogenic characteristics, and is also involved in the metabolism of amino acids and glycolysis in cells[19].

The bioconversion of exogenous or newly generated FAs to fatty acyl-CoA is catalyzed by five distinct long-chain acyl-CoA synthetase (ACSL) isoforms that make up the ACSL family and are implicated in the growth of malignant tumor cells[20,21]. For example, Cai and Ma demonstrated that ACSL1 expression levels were elevated in human prostate cancers, and that ACSL1 enhanced triglyceride production, lipid accumulation in cancer cells, and the manufacture of fatty acyl-CoAs, such as C16: 0-, C18: 0-, C18: 1-, and C18: 2-CoA[22].

In the present study, our aim was to investigate the role of genes involved in long-chain FA transport in PDAC by comparing the gene expression levels of CD36, SLC27A1, FASN, SLC27A2, SLC27A3, SLC27A4, SLC27A5, ACSL1, and ACSL3 in 24 PDAC tumoral tissues to the expression levels in 11 pancreatic non-tumoral tissues.

MATERIALS AND METHODS

Sample collection and gene expression analysis

Tumoral pancreatic tissue from 24 patients was obtained via fine needle aspiration (n = 19) or surgical resection (n = 5) and were diagnosed with PDAC at the Clinical Emergency Hospital of Bucharest. During surgical resection, 11 samples of non-tumoral pancreatic tissue were also obtained from the peritumoral area of patients with PDAC (n = 6), neuroendocrine tumor (n = 3), and metastasis of colorectal carcinoma (n = 2) as a control group (CTRL). The CTRL tissues were verified by a pathologist and only tissues without tumoral cells were included in the study. The samples were collected between January 2020 and March 2023. Only adults aged 18 years or older with a diagnosis of PDAC were included in the study; patients who underwent chemotherapy before sample collection were excluded. When possible, information on the lifestyle habits, presence of diabetes, and months of survival after PDAC diagnosis, were collected (Table 1). All patients provided written informed consent to participate in the study.

Table 1 Sociodemographic data, lifestyle habits, and presence of diabetes in pancreatic ductal adenocarcinoma patients and controls.

Parameter	PDAC, n = 24	CTRL, n = 11	P value
Age	64.75 ± 10.17	58.00 ± 13.94	0.115
Sex as %	54.2 (F), 45.3 (M)	54.5 (F), 45.5 (M)	0.983
Smokers as %	25	33.31	0.632
Coffee consumers as %	63.22	55.51	0.700
Alcohol consumers as %	4.1	01	0.557
Diabetes as %	20.8	01	0.160
Survival in months	4.81 ± 3.253

¹Available for 9 control group patients.

²Available for 19 pancreatic ductal adenocarcinoma patients.

³Available for 16 pancreatic ductal adenocarcinoma patients.

CTRL: Control group; PDAC: Pancreatic ductal adenocarcinoma.

For the gene expression analysis, we selected the most relevant genes involved in long-chain FA import into cells (GO: 0044539), which included CD36, SLC27A1, FASN, SLC27A2, SLC27A3, SLC27A4, SLC27A5, ACSL1, and ACSL3. IPO8 was used as the reference gene for qRT-PCR, as a previous study demonstrated its stability in human pancreatic cancer[23].

The miRNeasy Mini Kit (Qiagen, Hilden, Germany) was used to isolate the total RNA and the RT2 First Strand Kit (Qiagen) to reverse transcribe 300 ng of total RNA. PCR was performed in a volume of 25 μL on an ABI-7500 fast instrument (Thermo Scientific, Waltham, MA, United States) using 12.5 μL of RT2 SYBR® Green qPCR Mastermix (Qiagen), 2 μL of cDNA, 9.5 μL of RNase free water, and 1 μL of primer (RT2 qPCR Primer Assay 200 Cat. No./ID: 330001). The primers were purchased at Qiagen and the Gene Globe IDs were PPH01356A, (CD36, NM_000072) PPH17902A (SLC27A1, NM_198580), PPH07803A (SLC27A2, NM_003645), PPH20806A (SLC27A3, NM_024330), PPH00471A (SLC27A4, NM_005094), PPH16732B (SLC27A5, NM_012254), PPH19272A (ACSL1, NM_001995), PPH15368F (ACSL3, NM_004457), and PPH16835A (IPO8, NM_006390). The mRNA expression data are reported both as 2^−ΔCT values for each group and as fold change (FC = 2^−ΔΔCT; Table 2, Figure 1 and Figure 2), and were calculated as previously described by Livak and Schmittgen[24]. A gene was considered differentially expressed between the groups when FC > 2 or FC < 0.5 and P < 0.05.

Figure 1 Gene expression results. A and B: Genes found differentially expressed between non-tumoral pancreatic tissue and pancreatic ductal adenocarcinoma tissue, comprising fatty acid transporters (A) and long-chain acyl-CoA synthetases (B). Bars indicate the mean ± standard error. NT: Non-tumoral; PDAC: Pancreatic ductal adenocarcinoma.

Figure 2 Expression difference of SLC27A3 between pancreatic ductal adenocarcinoma tissue with lymph nodes invasion (Lym+) and pancreatic ductal adenocarcinoma tissue without (Lym-). Bars indicate the mean ± standard error.

Table 2 Gene expression results.

Gene	Our study (24 PDAC vs 11 NT)				OncoDB (200 PDAC vs 178 normal)
	PDAC (2−ΔCT ± SEM)	NT (2−ΔCT ± SEM)	FC	P value1	FC	P value2
CD36	0.527 ± 0.004	0.499 ± 0.009	1.06	0.001	2.893	< 0.001
SLC27A1	0.462 ± 0.003	0.462 ± 0.002	1.00	ns	2.836	< 0.001
FASN	1.325 ± 0.233	0.986 ± 0.222	1.34	ns	1.86	< 0.001
SLC27A2	0.763 ± 0.252	0.135 ± 0.036	5.66	0.033	3.689	< 0.001
SLC27A3	2.384 ± 0.548	0.888 ± 0.168	2.68	0.04	2.948	< 0.001
SLC27A4	0.996 ± 0.367	0.318 ± 0.059	3.13	0.033	2.366	< 0.001
SLC27A5	0.087 ± 0.026	0.045 ± 0.010	1.94	ns	2.385	< 0.001
ACSL1	14.063 ± 2.103	3.424 ± 0.725	4.10	< 0.001	19.043	ns
ACSL3	6.925 ± 1.397	2.593 ± 0.651	2.67	0.012	7.242	< 0.001

¹Mann–Whitney test.

²t-test Student mean.

FC: Fold change; NT: Non-tumoral; PDAC: Pancreatic ductal adenocarcinoma; SEM: Standard error of the mean.

In silico validation of gene expression results

Genes that showed significant dysregulation in our comparison of the 24 PDAC and 11 CTRL tissues were searched in public databases that report RNA-seq data from large case-control studies. Data from the TCGA database, including 178 patients with pancreatic adenocarcinoma and 200 controls (obtained from GTEx study), were downloaded from OncoDB (http://oncodb.org/index.html) and analyzed with Student’s t-test.

Statistical analysis

Differences in categorical variables (sex, lifestyle habits, and presence of diabetes) and continuous variables between the groups were assessed using the χ² test and Student’s t-test, respectively. The normality of the gene expression data (in terms of 2^−ΔCT) was evaluated using the Shapiro–Wilk test. Because the gene expression data were not normally distributed (Shapiro–Wilk test, P < 0.05), the nonparametric Mann–Whitney test was used to assess the statistical differences between the two groups. The statistical analyses were performed using SPSS version 20.0, and the graphs were generated using GraphPad Prism 8.4.3.

RESULTS

The groups of patients and controls were age-matched (P = 0.115) and sex-matched (P = 0.983). No difference between patients and controls was found in terms of number of smokers (P = 0.632), coffee consumers (P = 0.700), alcohol consumers (P = 0.557), and presence of diabetes (P = 0.160; Table 1).

Most of the patients presented PDAC localized in the head of the pancreas (n = 18, 75%), followed by 2 in the body (8.3%), 2 in the body-tail (8.3%), 1 patient in the neck (4.2%), and 1 patient in the tail of the pancreas (4.2%). At the time of sample retrieval, 17 of 24 patients presented with lymph node involvement, 9 patients had hepatic metastasis, and 3 patients had lung metastasis.

The gene expression analysis showed that five of the nine investigated genes were significantly dysregulated within the PDAC tissue compared to the CTRL. The dysregulated genes belonged to two categories: FA transporters and Long-Chain ACSL. Our gene expression results obtained by qRT-PCR and those obtained by RNA-seq available in OncoDB are reported in Table 2.

The five genes that showed significant differences in expression between the groups were SLC27A2, SLC27A3, SLC27A4, ACSL1, and ACSL3. All five genes were upregulated (FC > 2 and P < 0.05) in the pathological condition: SLC27A2 (FC = 5.66; P = 0.033), SLC27A3 (FC = 2.68; P = 0.040), SLC27A4 (FC = 3.13; P = 0.033), ACSL1 (FC = 4.10; P < 0.001), and ACSL3 (FC = 2.67; P = 0.012); (Figure 1).

We further investigated any possible association between the levels of the analyzed mRNAs and the specific characteristics of the tumors, including the anatomic location, the lymph node involvement, and the presence of metastasis. A significant difference in the expression of SLC27A3 was found comparing patients with and without lymph nodes involvement with an overexpression of this transcript in 17 patients presenting tumoral cells in the lymph nodes (FC = 3.27; P = 0.040; Figure 2). When correlating the expression levels of the analyzed genes with the patient’s survival, we observed a significant negative correlation between the months of survival and the FASN expression (P = 0.035, r = -0.528).

DISCUSSION

In this study, we report a general upregulation of key genes involved in long-chain FA import in PDAC. We found that three FA transporters and two long-chain ACSLs genes were significantly upregulated in the tumoral tissue compared to the non-tumoral pancreatic tissue. Specifically, we found upregulation of SLC27A2, SLC27A3, and SLC27A4, as well as ACSL1 and ACSL3.

The SLC27A2, SLC27A3, and SLC27A4 genes encode FATP2, FATP3, and FATP4, respectively. FATP2 and FATP4 transport exogenous FAs into the cell[25,26], whereas FATP3 does not have a transporter function, but instead acts as an acyl-CoA ligase, facilitating the ATP-dependent formation of fatty acyl-CoA[27]. The existing literature indicates that FATPs contribute to the development of various cancers by altering FA metabolism.

FATP2 plays a role in reprogramming neutrophils in cancer to mediate the acquisition of immunosuppressive activity[28]. In the kidneys, FATP2 regulates proximal tubule lipoapoptosis[29], and its upregulation suppresses the proliferation and invasion of renal cancer[30]. In lung cancer, reduced levels of FATP2 induce cisplatin resistance and are correlated with poor patient survival[31]. In thyroid cancer, high levels of FATP2 promote tumor proliferation and migration[32].

FATP3 expression is involved in glioblastoma, where it supports glioblastoma stem cell maintenance and tumorigenicity[33]. FATP3 expression is increased in lung cancer[27], and our research group has observed similar increases in FATP3 expression in colorectal cancer[34], as well as upregulation of FATP4.

Elevated levels of FATP4 have also been associated with tumorigenesis and tumor progression in clear cell renal cell carcinoma[35], and with tumor progression in non-muscle-invasive bladder cancer[36].

To our knowledge, our study is the first to report upregulation of expression of genes encoding these three transporters in PDAC tissue. Moreover, we also found higher levels of SLC27A3 (FATP3) in the tumors from patients with lymph node involvement than in those without. This upregulation is not surprising since FATP3 can mediate the levels of long-chain FAs that have been found to accumulate in metastatic sites. Targeting this mechanism to reduce this accumulation has been suggested to improve anti-tumor immune responses[37].

We also found that ACSL1 and ACSL3 were upregulated in PDAC tissue compared to non-tumoral pancreatic tissue. The ACSL family enzymes, which comprise ACSL1 and ACSL3-6, play key roles in activating free FAs to form fatty acyl-CoAs, which are needed for FA incorporation into phospholipids.

ACSLs are involved in endoplasmic reticulum stress, ferroptosis, drug resistance, and in perpetuating the tumor inflammatory microenvironment[38]. In liver cancer, ACSL1 promotes tumor growth[39], and its upregulation in hepatocellular carcinoma induces intracellular lipid accumulation[40]. In colorectal cancer, ACSL1 is considered a prognostic biomarker because its high expression is associated with poor clinical outcomes in stage-II cancer[41] and with promotion of tumor invasion in ovarian cancer[42].

ACSL3 expression is also associated with different cancers and shows an involvement in the metastasis of melanoma[43] and prostate cancer, where it also promotes tumor growth and proliferation[44]. In lung cancer, ACSL3 seems to promote tumor survival and chemosensitivity, and its high expression correlates with worse outcome in patients[45].

In pancreatic cancer, evidence linking ACSL family members has come from studies that investigated ACSL3, ACSL4, and ACLS5. In line with our findings, Sebastiano et al[46] found that ACSL3 mRNA levels were higher in primary ductal adenocarcinomas and metastasis than in healthy epithelium, and their results were confirmed by interrogating public databases analyzing PDAC human samples. The authors, conducting a study in mice, also demonstrated that ACSL3 deletion delayed PDAC progression and reduced fibrosis[46].

Another study showed that depletion of extracellularly derived lipids by restriction of ACSL3 could trigger autophagy and reduce PDAC cell proliferation[47]. For ACSL4, most studies indicate that it has an oncogenic function in most cancers; however, in pancreatic cancer ACSL4 expression appears to facilitate cell sensitivity to chemotherapy[48]. For ACSL5, progression-free survival is significantly shorter in patients with low expression than with high expression[49].

In the present study, although we identified an association between PDAC and the expression of three FA transporters and two long-chain ACSLs, our results cannot explain the molecular mechanism underpinning this relation. We can hypothesize that, as for other cancers, the overexpression of these transporters induces the elevated uptake of exogenous FAs that are subsequently stored as lipid droplets and ultimately undergo β-oxidation to generate ATP. This mechanism may produce enough ATP to support the enhanced growth and progression of tumor cells. The limitations of our study included its relatively small sample size and our investigation of our panel of genes only at the mRNA level, without confirming expression at the protein level.

CONCLUSION

The results reported in this article highlight the involvement of genes implicated in long-chain FA import in PDAC. We identified five transcripts (SLC27A2, SLC27A3, SLC27A4, ACSL1, and ACSL3) that were differentially expressed between PDAC and non-tumoral pancreatic tissue, with significant upregulation observed in the pathological condition. Despite the small number of patients analyzed, the preliminary results are promising. Future in vitro and in vivo studies on these genes may offer important insights into the mechanisms linking PDAC with the long-chain FA import pathway.