Background
Pancreatic cancer is one of the most lethal digestive tract malignancies, is difficult to detect early and has a poor prognosis. The 5-year overall survival rate has remained relatively stable, at close to 8%, over the past decade [
1,
2]. Chemotherapy has become one of the primary and indispensable therapeutic methods. However, chemoresistance to gemcitabine (GEM), which is the first-line adjuvant therapy for pancreatic cancer, is prevalent and has become one of the key reasons that patients with pancreatic cancer show a poor prognosis. Therefore, elucidating the mechanism of chemoresistance to GEM may contribute to enhancing the effects of therapy and improving the prognosis of pancreatic cancer.
Reprogrammed energy metabolism has become an emerging hallmark of cancer in recent years [
3]. Previous studies indicated that most metabolites did not show significant changes in pancreatic cancer cells exposed to GEM; however, chemoresistance to GEM induces metabolic reprogramming [
4,
5]. Therefore, metabolic reprogramming must be involved in the regulation of chemoresistance. Glutamine and glucose are the primary nutrient sources for cancer cells and are crucial to the biosynthesis of products such as nucleic acids and nonessential amino acids (NEAAs) [
6]. Otherwise, glutamine metabolism is involved in both glycolysis and the tricarboxylic acid (TCA) cycle. In addition, other studies show that glutamine can activate mechanistic target of rapamycin (mTOR) [
7] and inhibit apoptosis to induce chemoresistance [
8]. Thus, glutamine transport and metabolism might play an important role in chemoresistance in pancreatic cancer.
L-type amino acid transporter 2 (LAT2), encoded by the SLC7A8 gene, belongs to the SLC7 subfamily of the LAT family [
9]. LAT2 is a Na
+-independent neutral amino acid transporter and is responsible for transporting neutral amino acids, including Gln, Gly, Ser, Ala, Thr, Asn, Met, Val, Phe, Tyr, Leu, Ile, Trp and His [
9,
10]. LAT2 is upregulated in 9 different cancer types according to Oncomine data [
9]; however, few studies have proved its specific role in cancer, and the expression level and role of LAT2 in pancreatic cancer remain uncertain and elusive. In our previous study, we identified a cluster of dysregulated mRNAs associated with GEM chemoresistance via mRNA microarray analysis in the established gemcitabine-resistant pancreatic cancer cell line AsPC-1-GEM(data unpublished). Among these dysregulated mRNAs, LAT2 showed a higher expression level in AsPC-1-GEM cells (≥10-fold change) than in the parental cells. Thus, we presumed that LAT2 might be involved in the development and regulation of chemoresistance in pancreatic cancer. Given the previously described role of LAT2 in glutamine transport and metabolism, we hypothesized that LAT2 could promote GEM chemoresistance in pancreatic cancer cells through glutamine-dependent mTOR activation.
Methods
Patients, sample collection and immunohistochemistry (IHC)
Formalin-fixed, paraffin-embedded pancreatic cancer (n = 87) and matched paracancerous (n = 71) tissue microarrays were obtained from Peking Union Medical College Hospital. None of the patients received neoadjuvant therapy. Tumor staging relies on the 8th edition of TNM system designed by the American Joint Committee on Cancer (AJCC). Follow-up depended on medical records and telephone. The primary end point was overall survival. Mouse anti-human LAT2 monoclonal antibodies (GTX83618, GeneTex, 1:100) and rabbit anti-human LDHB polyclonal antibodies (14824–1-AP, Proteintech, 1:100) were used for the standard immunohistochemical staining procedures. The results of immunohistochemistry were scored by adding the percentage score and the intensity score, which were assessed according to the percentage of positive cells and intensity of staining respectively. The percentage score was classified into 4 grades using the percentage of positive cells (1, < 25%; 2, 25–50%; 3, 50–75%; 4, > 75%). The intensity score was classified into 4 grades using the intensity of stained cells (0, none; 1, weak; 2, moderate; 3, strong). LAT2 and LDHB expression was considered to be low if the total score was equal to or less than the median, and considered high if the score was greater than the median.
Cell culture
The human pancreatic cancer cell lines MIA PaCa-2 and PANC-1 were donated by Dr. Freiss H (University of Heidelberg, Heidelberg, Germany). MIA PaCa-2 and PANC-1 cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM, Logan, UT, USA). All Media were supplemented with 10% fetal bovine serum (FBS, Hyclone, Logan, UT, USA) at 37 °C and 5% CO2.
Plasmids and antibodies
Complementary DNA encoding LAT2 was synthesized and subcloned into the pENTER-FLAG vector (Vigene Bioscience, MD, USA) according to the manufacturer’s instructions. For PANC-1 cell lines that stably overexpressed LAT2, the lentivirus plasmid pLVX-LAT2 was constructed by subcloning the LAT2 coding sequence (Gene ID: 23428) into the modified pLVX-puro backbone (Clontech, CA, USA). The forward primer was 5'-GCGGCTAGCATGGAAGAAGGAGCCA-3', and the reverse primer was 5'-CGCGGATCCGGGCTGGGGCTGCC-3'. The PCR products of LAT2 coding sequence were inserted into the pLVX-puro plasmids with Nhe I and BamH I. The pLVX-LAT2 was packaged into lentiviruses. PANC-1 cells were infected with the pLVX-LAT2 lentivirus, and the cell line was got after 2μg/ml puromycin screening.
Antibodies against LAT2 (GTX83618, GeneTex), LDHB (14824–1-AP, Proteintech), p-mTOR (Ser2448) (5536, Cell Signaling Technology), mTOR (2983, Cell Signaling Technology), PKM2 (4053, Cell Signaling Technology), GAPDH (10494–1-AP, Proteintech), Glutaminase (ab156876, Abcam) and Glutamine synthetase (sc-74430, Santa Cruz Biotechnology) were purchased commercially.
RNA interference
SiLAT2 and siNC were purchased from RiboBio (Guangzhou, China). The sequences for LAT2 siRNA are listed as below: the forward primer was 5'-GAAGAUGAUGAUGCCAAUUTT-3', and the reverse primer was 5'-AAUUGGCAUCAUCAUCUUCTT-3'. MIA PaCa-2 and PANC-1 cells were transfected with siRNA at 50-100 nM packaged by Lipofectamine 3000(Invitrogen, Carlsbad, CA, USA) in 6-well plates (5 × 105cells/well).
Lentivirus production
HEK293T cells were cultured in individual 15 cm dishes for each viral construction. When the cells reached approximately 70–80% confluency (~ 1.5 × 107 cells), they were cotransfected with 15 μg of the pLVX-LAT2 plasmid, 15 μg of the psPAX2 plasmid (Addgene #12260, Cambridge, MA, USA) and 10 μg of the pMD2.G plasmid (Addgene #12259, Cambridge, MA, USA) plasmid with Lipofectamine 2000 transfection reagent. At 18 h post-transfection, the medium was replaced with 25 mL of fresh complete medium. The supernatants were harvested at 48 h and 72 h, centrifuged at 1000 rpm for 10 min at 4 °C to remove the cells, and filtered through a 0.45 μm filter to remove the debris. Finally, the supernatant was ultracentrifuged at 120,000×g for 2 h at 4 °C, dissolved in PBS after the removal of the supernatant, and stored at − 80 °C.
CRISPR KO cell line construct
The target sequence was cloned into the lentiCRISPR v2 backbone (Addgene #52961, Cambridge, MA, USA), and two oligonucleotides were generated after digestion with BsmBI. The design of the target sequences was performed using
http://crispr.mit.edu to obtain the gRNA sequence. The LAT2 gRNA forward primer sequence was 5’-CACCGTGACATCGGCCTCGTCGCAC-3′, and the reverse primer sequence was 5’-AAACGTGCGACGAGGCCGATGTCAC-3′. LentiCRISPR-LAT2sgRNA was packaged into the lentiviruses. MIA PaCa-2 and PANC-1 cells were infected with the LentiCRISPR-LAT2sgRNA lentivirus for more than 48 h, and the cell lines were obtained 48 h after screening with 2 μg/ml puromycin. Then, the monoclonal cell lines were screened by identifying those with a knockout of LAT2 protein expression and a disturbance of the target DNA sequence.
Cell proliferation and growth inhibition assay
At 24 h after transfection, MIA PaCa-2 and PANC-1 cells were plated into 96-well culture plates (3000 cells/well) for cell proliferation assays. All the plates were cultured at 37 °C with 5% CO2. For proliferation assays, 10 μL/well cell count kit (CCK-8) reagent was added at 0, 24, 48 and 72 h after plating. After an additional 2.5-h incubation with CCK-8 reagent at 37 °C, the optical density (OD) at the 450-nm wavelength (OD450) was measured using a microplate reader (Wellscan MK3, Thermo/Labsystems, Finland). OD630 served as a reference, and the OD in the blank well was used as the base level. For growth inhibition assays, 4000 cells/well were plated into 96-well culture plates at 24 h after transfection. After incubation for 4–6 h for cell adherence, a gemcitabine (Eli Lilly and Company) concentration gradient from 100 nM to 1 mM or control PBS buffer was added into each well. Cell count kit (CCK-8) reagent (10 μL/well) was added after an additional 48-h incubation at 37 °C. Then, the inhibition rate was calculated as follows: OD = OD450-OD630, inhibition rate = 1-(ODGEM-ODblank) / (ODPBS-ODblank).
Apoptosis assay
MIA PaCa-2 and PANC-1 cells were transfected into 6-well plates and treated with 10 μM gemcitabine twenty-four hours later. After treating for 48 h, the cells were collected and resuspended in binding buffer. Next, the cells were stained with Annexin V-FITC and propidium iodide (PI) (Beyotime, China) according to the manufacturer’s instructions. Analyze was carried out using flow cytometry (FACScan, BD Biosciences, USA).
The OCR and ECAR of cells were measured with an XF96 Extracellular Flux Analyzer (Seahorse Bioscience, North Billerica, MA, USA). Cells were plated (10,000 cells per well for MIA PaCa-2; 5000 cells per well for PANC-1) in at least triplicate for each condition the day before the experiment. The energy phenotype test and glycolytic rate assay were performed as described in the user guides for the Agilent Seahorse XF Cell Energy Phenotype Test Kit (103325–100, Agilent Technologies) and the Agilent Seahorse XF Glycolytic Rate Assay Kit (103344–100, Agilent Technologies). OCR and ECAR were normalized to the cell number as determined by CellTiter-Glo analysis at the end of the experiments.
Measurement of intracellular metabolites
The intracellular level of glutamine was determined by ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) with glutamine as the target. A standard solution of L-glutamine was prepared, and calibration curves were drawn. The UHPLC separation was performed using an Agilent 1290 Infinity II series UHPLC system (Agilent Technologies) equipped with a Waters ACQUITY UPLC BEH amide column (100 × 2.1 mm, 1.7 μm). An Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies) equipped with an AJS electrospray ionization (AJS-ESI) interface was used for assay development. If the concentrations of the metabolites were beyond the linear range, the samples were diluted accordingly to put their concentrations within the range. The final concentration in nmol/L equals the calculated concentration multiplied by the dilution factor.
Immunofluorescence assay
PANC-1 cells were transfected in slide chambers (NUNC, Denmark) with Flag-LAT2 vector for 24 h. The cells were fixed in methanol, blocked with 10% FBS and then incubated with mouse anti-LAT2 or p-mTOR(Ser2448) antibody and rabbit anti-LDHB antibody. The LAT2 and p-mTOR(Ser2448) staining were detected with an Alexa Fluor 488-labeled goat anti-mouse antibody and the LDHB staining was detected with an Alexa Fluor 647-labeled goat anti-rabbit antibody. DAPI was used for nuclear staining. Images were made using a Leica TCS SP2 microscope.
Immunoblotting and immunoprecipitation
After transfection for 48 h, total cellular protein was extracted with RIPA lysis buffer (Applygen, Beijing). Total protein (40 μg) was separated on an SDS-PAGE gel and then transferred to a PVDF membrane (Millipore, Billerica, MA). After blocking with 5% nonfat milk at room temperature for 1 h, the membrane was probed with primary antibodies (1:1000, Danvers, MA) overnight at 4 °C. The next day, the membrane was incubated with a horseradish peroxidase-conjugated secondary antibody (1:3000, Applygen, Beijing) at room temperature for 1 h. Proteins were detected by ECL reagents (Millipore, Billerica, MA). For immunoprecipitation, stable PANC-1 transfectants of the pENTER and pENTER LAT2 vectors were seeded in 6-well plates. The cells were then lysed in modified RIPA buffer. The cell lysates were incubated with antibody for 12 h at 4 °C on a rotating plate. The proteins were immunoprecipitated by protein A/G agarose beads (Santa Cruz, USA). The samples were resolved by SDS-PAGE and were subjected to immunoblot analysis.
Animal experiments
PANC-1 cells stably transfected with LAT2-lentiviral vectors or control lentiviral vectors were injected subcutaneously into the right flank of 6-week-old female BALB/c mice (Shanghai, Chinese Academy of Sciences, China) (5 × 106 cells in 250 μl of PBS per mouse). Each experimental group included five mice. Gemcitabine of 50 mg/kg or equal PBS were administered by intraperitoneal injection two weeks after tumor formation (tumor size between 100 and 200mm3), followed by periodic booster shots every three days for two weeks. Tumor size was measured twice a week using a caliper to measure two perpendicular tumor diameters. Tumor volume (mm3) was calculated based on the formula: volume (mm3) = 1/2 × length × width2. Tumor growth inhibition rate was calculated based on the formula: TGI % = (1-(Ti-T0) / (Ci-C0)) × 100. Ti = tumor volume after treatment with GEM, T0 = tumor volume prior to treatment with GEM, Ci = tumor volume after treatment with PBS, C0 = tumor volume prior to treatment with PBS. The tumor-bearing mice were euthanized on the 27th day.
Statistical analysis
Statistical analysis was performed using SPSS (version 19) and GraphPad PRISM 6 software. Statistical methods included the Wilcoxon signed rank test, Pearson chi-square test, two-tailed Student’s t-test, Kaplan-Meier survival analysis, and Cox proportional hazards regression analysis. All statistical tests included a two-way analysis of variance. Statistical significance was assumed when P < 0.05.
Discussion
In humans, the specific role of LAT2 in pancreatic cancer has not been reported and is still elusive. Our previous study has demonstrated that the expression level of LAT2 in AsPC-1-GEM cells is higher than that in the parental cells, which might indicate that LAT2 plays an important role in chemoresistance in pancreatic cancer. Chemoresistance is regarded as one of the primary causes of poor prognosis and high mortality in pancreatic cancer. In this study, we identified LAT2 as a novel oncogenic protein in pancreatic cancer. We found that the expression levels of both LAT2 and LDHB in pancreatic cancer tissues were higher than those in the paracancerous tissues and that high levels of both LAT2 and LDHB were associated with poor prognosis. In addition, LAT2 could decrease GEM sensitivity, promote proliferation, inhibit apoptosis and activate glycolysis in pancreatic cancer cells, and these effects could be reversed by RAD001 treatment. Mechanistically, LAT2 could regulate the glutamine/p-mTORSer2448/glutamine synthetase feedback loop to keep mTOR activated and to upregulate LDHB in order to activate glycolysis, thereby promoting chemoresistance in pancreatic cancer cells. We also demonstrated that the decreased apoptosis and increased ECAR and glycoPER induced by the LAT2-mTOR-LDHB pathway in pancreatic cancer cells were dependent on glutamine. Consequently, our data indicate that the LAT2-mTOR-LDHB pathway might be a valuable prognostic predictor and promising therapeutic target in pancreatic cancer.
LAT2 has been reported to play different roles in multiple tumor types. Barollo et al. confirmed that LAT2 was overexpressed in neuroendocrine tumors, including pheochromocytoma and medullary thyroid carcinoma, compared with normal tissues and that it was responsible for dihydroxyphenylalanine uptake [
11]. Luo et al. found that LAT2 might play important roles in the proliferation of uterine leiomyoma cells [
12]. In breast cancer, SLC7A8 mRNA expression was elevated in samples from estrogen receptor alpha-positive breast cancer patients. However, a high level of SLC7A8 mRNA was significantly associated with longer relapse-free survival in estrogen receptor alpha-positive and lymph node-positive breast cancer [
13]. Our study demonstrated that a high LAT2 level was associated with poor overall survival in pancreatic cancer. In terms of chemosensitivity, Rumiato et al. found that single nucleotide polymorphisms mapping to the SLC7A8 gene, in combination with clinical variables, could contribute to the accuracy of the predicted response to platinum chemotherapy in esophageal cancer patients [
14]. Meanwhile, another study showed that the expression of the SLC7A8 gene was significantly decreased in the paclitaxel-resistant W1 human ovarian cancer cell line compared with W1 parental cells; thus, SLC7A8 might be involved in reversing drug resistance in ovarian cancer [
15]. However, in our previous study, LAT2 showed a higher expression level in AsPC-1-GEM cells than in the parental cells. In this study, we also demonstrated that LAT2 played an oncogenic role in pancreatic cancer and could decrease GEM sensitivity by regulating the glutamine-dependent LAT2-mTOR-LDHB pathway.
Glutamine addiction is a feature of many cancer types. Glutamine, a major source of carbon and nitrogen for sustaining the proliferation of cancer cells, has been reported to be critical for proliferation in pancreatic cancer [
16‐
18]. Otherwise, oncogenic alterations in genes such as MYC and KRAS could reprogram glutamine metabolism in cancer cells [
8]. For example, c-MYC has been demonstrated to bind to the promoter regions of high-affinity glutamine importers, thereby causing increased glutamine uptake, and oncogenic KRAS could upregulate enzymes involved in glutamine metabolism [
19,
20]. Jewell et al. have demonstrated that glutamine could activate mTORC1, which is important for the growth of both normal and tumor cells [
7]. Recent studies have indicated that transporters act as amino acid sensors involved in controlling mTOR recruitment and activation at the surface of multiple intracellular compartments. We have demonstrated that LAT2 could bind to p-mTOR to regulate the glutamine/p-mTOR
ser2448/glutamine synthetase feedback loop, resulting in an increased intracellular level of glutamine and the sustained activation of mTOR. In contrast, LAT2 could downregulate glutaminase to further increase the intracellular level of glutamine. Activated mTOR could bind to and upregulate LDHB to increase glycolysis and decrease chemosensitivity. Glutamine deprivation could then cause mTOR activation failure and decreased glycolysis. We also demonstrated that RAD001 treatment could downregulate LAT2, which indicated that mTOR activation might contribute to maintaining LAT2 stability. Therefore, LAT2 could bind to p-mTOR
Ser2448, and they could interact with each other to form a positive feedback loop. Next, p-mTOR
Ser2448 could bind to and upregulate glutamine synthetase to increase endogenous glutamine levels, thereby forming another positive feedback loop (glutamine/p-mTOR
ser2448/glutamine synthetase). Meanwhile, mTOR activation depends on exogenous glutamine. Therefore, both positive feedback loops are dependent on exogenous glutamine.
RAD001, a mTOR inhibitor, has been approved for the treatment of pancreatic cancer; however, treatment with single mTOR inhibitors can lead only to disease stabilization rather than regression because mTOR contributes to cellular proliferation rather than cell survival [
21]. In this study, we have demonstrated that RAD001 treatment could reverse the decrease in GEM sensitivity caused by LAT2 overexpression in pancreatic cancer. These results indicate that GEM combined with RAD001 might improve chemosensitivity in pancreatic cancer patients with LAT2 overexpression.
In terms of the relationship between mTOR and LDHB, Zha et al. demonstrated that LDHB, which is critical for oncogenic mTOR-mediated tumorigenesis, is a downstream target of mTOR [
22]. Our study also confirmed that LAT2 could target mTOR activation to regulate LDHB, which is critical for the conversion of pyruvate to lactate. Then, we assessed the role of the LAT2-mTOR-LDHB pathway in glycolysis, and the results indicated that both LAT2 KD and mTOR inhibition could downregulate LDHB, decrease glycoPER and inhibit glycolysis. In contrast, increased glycoPER has been reported to be associated with decreased chemosensitivity. In addition, we have demonstrated that RAD001 treatment could reverse the decrease in GEM sensitivity induced by LAT2 OE. Therefore, we speculated that LAT2 could regulate the above two positive feedback loops to upregulate LDHB and activate glycolysis, which would promote chemoresistance in pancreatic cancer. However, Cui et al. found that LDHB might act as a suppressor of glycolysis and as a tumor suppressor gene in pancreatic cancer due to promoter hypermethylation [
23]. Conversely, a recent study indicated that positive LDHB protein expression was associated with progression and poor prognosis in patients with pancreatic cancer [
24]. Otherwise, LDHB has been reported to be an oncogenic protein in multiple tumor types, including uterine cancer, ovarian cancer, colon cancer and breast cancer [
25‐
27]. In this study, we demonstrated that a high LDHB level is associated with poor prognosis in pancreatic cancer.