Metabolomic Investigations into Hypoxia-mediated Metabolic Reprogramming of Pancreatic Cancer Cells

Abstract

Hypoxia is a crucial microenvironmental factor that defines tumor cell growth and aggressiveness. Cancer cells adapt to hypoxia by altering their metabolism. These alterations impact various cellular and physiological functions, including energy metabolism, vascularization, invasion and metastasis, genetic instability, cell immortalization, stem cell maintenance, and resistance to chemotherapy (1). Hypoxia-inducible factor-1α (HIF-1α) is known to be a critical regulator of glycolysis that directly regulates the transcription of multiple key enzymes of the glycolysis pathway. Moreover, HIF-1α stabilization can be directly modulated by TCA-derived metabolites, including 2-ketoglutarate, and succinate (2). Overall, the molecular mechanisms underlying the adaptation of cellular metabolism to hypoxia impact the metabolic phenotype of cancer cells. Such adaptations include increased glucose uptake, increased lactate production, and increased levels of other metabolites that stabilize HIF-1α, leading to a vicious circle of hypoxia-induced tumor growth.

1. Introduction

Hypoxic microenvironments, which are defined as regions with reduced oxygen levels, are highly prevalent in pancreatic ductal adenocarcinoma (PDAC) and other solid tumors due to desmoplasia-mediated poor blood flow (1, 3). Under hypoxia, cancer cells reprogram their metabolism to glycolysis as an ATP source. Though less energy-efficient than oxidative phosphorylation, glycolysis generates metabolic intermediates for cell growth, proliferation, and adhesion molecule expression (4-6). The key regulator orchestrating the adaptation to low oxygen concentration is hypoxia-inducible factor-1α (HIF-1α), which acts as a transcription factor (7). The stability and activity of HIF-1α are regulated by various post-translational modifications, including hydroxylation, acetylation, and phosphorylation. Under normoxia, the HIF-1α subunit is rapidly degraded via the von Hippel-Lindau tumor suppressor gene product (pVHL)-mediated ubiquitin-proteasomal pathway (8). Hydroxylation of two critical prolines by specific hydroxylases (PHDs) targets the HIF-1α subunit for proteasomal degradation (9).

PHDs use molecular oxygen as a substrate for catalysis. In addition to oxygen, PHDs require α-ketoglutarate (α-KG) as well as the cofactors Fe2+ and ascorbate (8, 10). As α-KG is an intermediate metabolite of the mitochondrial TCA cycle, a decrease in the concentration of α-KG or accumulation of subsequent metabolites following α-KG, such as succinate, fumarate, and malate, could affect the activity of α-KG-dependent dioxygenases (7, 11). The accumulation of succinate and fumarate by loss-of-function mutations in genes encoding for succinate dehydrogenase (SDH) subunits and fumarate hydratase (FH) leads to competitive inhibition of the α-KG-dependent PHD catalytic reaction (12).

The switch to a HIF-1-regulated phenotype promotes selection for hundreds of genes, many of which are associated with a more malignant phenotype [12]. Additionally, several oncogenes, such as MUC1, regulate stabilization and transcriptional activity of HIF-1α (3-5, 13). In most cell types, metabolic genes have been identified that are transactivated by HIF-1α, such as lactate dehydrogenase A (LDHA) and other glycolytic enzyme genes, including hexokinase 1 (HK1), hexokinase 2 (HK2), phosphofructokinase (PFK), and phosphoglycerate kinase 1 (PGK1) (14).

This chapter describes a detailed methodology of metabolomic analysis using pancreatic adenocarcinoma cells, Capan-2 exposed to normoxic and hypoxic conditions for 24 hours. We describe this method in a series of steps starting from cell culture in hypoxic conditions, polar metabolite extraction, setting up the LC-MS/MS, and analysis of the generated data along with the inferences derived from our hypoxia-metabolomic experiment.

2. Materials

2.1. Cell Culture Reagents

1. Dulbecco’s Modified Eagle’s Medium (DMEM) (13.4 g/L) (with phenol red).

2. Fetal bovine serum (FBS).

3. 1× phosphate buffered saline (PBS) w/o calcium and magnesium.

4. 0.9% Sodium chloride solution.

5. 0.25% Trypsin solution.

6. Penicillin-streptomycin solution (100×) (penicillin 10,000 units/mL; streptomycin 10,000 μg/mL).

2.2. LC-MS/MS reagent

1. 100% Methanol (LC-MS grade).

2. Water (LC-MS grade).

3. Choline-d9 chloride.

4. 10 ng/ml of Choline Chloride-d9 in methanol for metabolite extraction: (Stock solution: mix 400 mL of LC-MS Grade 100% methanol with 100 mL of LC-MS grade water and 5 mg choline chloride-d9 in a glass bottle. Store in a −80 °C freezer for further use)

5. Acetonitrile (ACN) (LC-MS grade).

6. 50% (w/v) Ammonium hydroxide.

7. Ammonium acetate.

2.3. Western blot analysis

1. Protein Assay Dye Reagent Concentrate.

2. NP-40 lysis buffer: 50mM Tris pH 7.4, 1% NP-40, 150 mM NaCl, 5 mM EDTA supplemented with protease inhibitor cocktail, 5mM NaF and 1 mM Na₂VO₄.

3. Milli-Q water.

4. 30% acrylamide and bis-acrylamide solution, 19:1.

5. 1 M Tris pH 8.8.

6. 0.5 M Tris pH 6.8.

7. 10% SDS solution.

8. 10% APS solution.

9. TEMED (N,N,N',N'-Tetramethylethylenediamine).

10. Tris-glycine-SDS running buffer.

11. Tris-glycine-SDS transfer buffer.

12. 1× Tris-buffered saline (TBS) with 0.05% Tween.

13. Not-Fat Skim milk.

14. anti-HIF-1α antibody BD610959 (BD Bioscience).

15. anti-mouse HRP conjugated antibodies.

16. ECL Western blotting substrate.

2.4. Equipment

1. 96-well plate.

2. 6 cm dish.

3. 10 cm dish.

4. O₂/CO₂ cell culture incubator.

5. Conical tubes.

6. Serological pipets: 5ml, 10ml, 25ml.

7. Gilson pipets.

8. Pipette tips: 0.1–20 μL, 200 μL, 1000 μL.

9. Automated cell counter or hemocytometer.

10. Cell scrapers.

11. Microfuge tubes, 1.5 ml.

12. Mini vortex mixer.

13. Refrigerated microfuge.

14. SpeedVac.

15. UPLC: ACQUITY UPLC H-Class system (Waters).

16. Guard column: Waters Acquity UPLC BEH Amide 1.7 um, vanguard pre-column (130Å, 2.1mmx5mm), Part no: 186004799.

17. Column: Waters Acquity UPLC BEH Amide Column, 130Å, 1.7 μm, 2.1 mm X 100 mm, Part number; 186004801.

18. Waters Xevo TQ Absolute triple quadrupole mass spectrometer system.

19. HPLC glass vials with 100 μL insert and silicone Screw Cap 8mm Slit PTFE.

20. Microplate reader to measure 595 nm absorbance.

21. Orbital-shaker.

22. Imaging system or X-ray films.

2.5. Software

1. MassLynx (V4.1) used for configuration and operation of Waters UPLC coupled with triple quadrupole mass spectrometer and raw data acquisition.

2. Skyline version 22.2 software used for data processing and reporting.

3. GraphPad Prizm.

4. MetaboAnalyst 5.0 (https://www.metaboanalyst.ca/) online statistical tool.

3. Methods

3.1. Mobile phase preparation

1. Mobile phase A: 100% ACN.

2. Mobile phase B: 95:5 H₂0:ACN with 10mM Ammonium acetate, pH 9.0 (see Note 1 and Note 2).

3.2. Cell culture and hypoxia

1. Seed cells at approximately 2 × 10⁶ cells in T75 cell culture flasks and culture in DMEM high glucose with 10% heat inactivated FBS at 37°C and 5% CO₂ with saturating humidity. Passage cells when they reach 70-80% confluency with 0.25% Trypsin(see Note 3) .

2. For the experiment, seed 2.33×10⁴ Capan-2 cells per cm² in 10 cm dishes overnight.

3. The following morning, replace the culture medium with 10 ml fresh medium, and divide the cells into two experimental groups: normoxic and hypoxic.

4. To study the hypoxia condition, place cells in a tri-gas incubator with 1% O₂ atmosphere, while keeping control normoxic cells in the conventional cell incubator (21% O₂) for 24 h (see Note 4).

3.3. Metabolites extraction

1. After the 24 hour incubation, aspirate culture media and wash cells quickly with cold 0.9% NaCl solution (see Note 5 and Note 6).

2. Place the culture dishes on dry ice.

3. Add 2 ml of ice-cold 80% methanol with choline chloride-d9 (Internal standard) to the plate on dry ice and transfer it to −80°C freezers for at least two hours.

4. Scrape the cells and transfer the methanol extracts to 1.5 ml tubes to lyse overnight at −80°C (see Note 7).

5. Next morning, centrifuge the samples at 15000 x g, 4°C for 10 min.

6. Collect the supernatants in new tubes and evaporate liquids using Eppendorf Concentrator Plus (Eppendorf), temperature must be maintained and not exceed evaporation time, usually 3-5 hours.

7. Store dried samples at −80°C until the LC-MS/MS analysis (see Note 8).

8. Reconstitute samples for analysis by adding 75 μl of ACN:H2O solution in 50:50 ratio, centrifuge at 15000g, 4°C for 10 min.

9. Transfer 30 μl of the sample to a HPLC vial, and inject 8 μl into LC-MS/MS.

10. Pipette 3 μl of each sample and mix in one tube for the QC analysis and inject before every batch as a readout of instrument consistency and reproducibility.

3.4. Liquid Chromatography- Coupled Tandem Mass Spectrometry

1. Perform liquid chromatography using the following UPLC parameter gradient method at a constant flow rate of 0.3 ml/min. The mobile phase gradient parameters are presented in Table 1.

2. MS parameters: Source temperature should be 150 °C, desolvation temperature 500 °C, cone gas flow 150 L/h, desolvation gas flow 1000 L/h, and collision gas flow 0.24 mL/min.

Table 1.

Mobile phase gradient parameters.

Time (min)	0	3	7	12	15	17	19	22	26	30
%A	85	84	65	60	55	50	50	70	85	85
%B	15	16	35	40	45	50	50	30	15	15

3.5. Data processing

1. Open the raw file in Skyline software and follow the steps for integrating individual peaks as per the instructions (see Note 9).

2. Export the final data containing peak areas, retention times, sample, and individual metabolite labels, and organize the data by grouping the samples by replicates of cells treated in normoxic and hypoxic conditions.

3.6. Selecting the metabolite peaks for comparative analysis

1. Check the variance among the individual metabolite peak areas within the biological replicates of a group by using statistical variance analysis.

2. Metabolites with a fold variance value >2 between the replicates should be excluded from the analysis.

3. Metabolites with low abundance (peaks with low signal/noise, S/N) are identified with their very low peak area values and excluded from analysis (see Note 10).

3.7. Data formatting for fold change analysis

1. Normalize the data by using the factor obtained from the ratios of the choline chloride-d9 peak area in this sample and peak areas obtained from normoxic and hypoxic group samples. Then choose one random choline chloride-d9 peak area value and normalize all peak areas to this value.

2. Then normalize data by the number of cells seeded. Estimate the factor obtained from the cells number ratio in control and experimental conditions inside the group, e.g., control cells cultured in normoxia and hypoxia (see Note 11).

3. Obtain the statistical mean for each individual metabolites in one group (e.g., normoxia or hypoxia).

4. Divide each normalized metabolite peak area value in experimental replicate samples from hypoxia group and normoxia group with the mean area value obtained for the same metabolite and corresponding sample in normoxia.

5. Use the fold change values obtained in steps 3 and 4 for the graphical data analysis (using GraphPad Prism or MetaboAnalyst 5.0), and represent the data as relative values.

3.8. Data analysis using MetaboAnalyst and data interpretation

1. Organize the normalized peak area data in the rows for the samples and in the columns for the analytes. MetaboAnalyst 5.0 website (http://www.metaboanalyst.ca/MetaboAnalyst).

2. Apply Principal Component Analysis (PCA) from MetaboAnalyst tools to obtain the clustering patterns among the replicates and the groups under study as shown in Fig. 1.

3. The close clustering of the biological replicates in the heatmaps and the PCA analysis (Fig.1) is an indication of low variance among the biological replicates of the samples under each condition.

Figure 1:

Principal component analysis (PCA) of pancreatic cancer cell metabolomes under normoxia and hypoxia. PCA plot showing the clustering for Capan-2-normoxia (green) and Capan-2-hypoxia (red) groups based on their polar metabolite contents. Each colored circle indicates one replicate of the group.

3.9. Western blotting

1. After the 24 hour incubation, aspirate cell medium and wash cells quickly with cold 0.9% NaCl solution (see Note 12).

2. Place the 6 cm culture dishes on ice.

3. Add 250 μl of NP-40 lysis buffer, vortex and allow cells to lyse for 20 min on ice.

4. Centrifuge cell lysates at 12000 × g, 4°C for 20 min and transfer supernatants to new tubes.

5. Estimate the protein concentration by reference to a standard BSA curve using Bradford assay according to the manufacturer protocol.

6. Load 35 μg of total protein for each lane. Resolve samples on 10% polyacrylamide gel in 1× SDS-PAGE running buffer and transfer proteins to the PVDF membrane.

7. Block membranes with 5% skim milk in TBST for 1h at RT.

8. Incubate membranes with anti-HIF1α (1:2000) overnight at 4°C, and with anti-β-actin antibody 1h at RT.

9. Wash membranes 3 times with TBST and incubate with anti-mouse HRP-conjugated antibodies (1:10000) 1h at RT.

10. Wash membranes 3 times with TBST and then incubate for 30 seconds with ECL reagent.

11. Visualize and analyze the results (Fig. 2).

Figure 2:

Western blot analysis of HIF-1α expression in Capan-2 cells subjected to 24 h normoxia and hypoxia treatments. Luminescent signal from HRP-conjugated antibodies was detected in cell lysates after 24 h hypoxia (HX) at a molecular weight of approximately 120 kDa and not in cells exposed to normoxia (NX). α-β-actin is shown as a loading control.

4. Notes

1) Measure the pH of the buffers before the acetonitrile is added.

2) Measure the pH of the buffers in aliquots in a different tube from the mobile phase bottle. Do not dip electrodes into the mobile phase bottle.

3) Cells must be in the exponential (log) growth phase to keep the uniform metabolic activity.

4) Maintain the hypoxia incubator at 1% O2 using nitrogen gas.

5) Wash cells with cold (+1°C to +4°C) 0.9% NaCl solution to immediately slow down all metabolic processes in the studied cells.

6) To minimize the sample-derived contamination and improve the MS data quality, replace the 0.9% NaCl solution to milli Q water. Aspirate water within 5 s to avoid osmotic lysis of cells.

7) Adding the methanol for each plate after transferring to dry ice makes it easier to scrape the cells rather than adding the methanol after aspirating 0.9% NaCl solution from the plates.

8) Dried metabolite samples can be stored at −80 °C for several weeks.

9) MRM (multiple reaction monitoring) transitions for individual metabolites can be obtained either from published reports or mass bank (QuantPedia) databases.

10) Exclude peaks with areas <10,000 from analysis. However, if the metabolite peak area is high (>10,000) among all the replicates in one condition (e.g., in normoxia) but low in other condition, such data can be retained for further comparative analyses.

11) Normalization can also be performed with the protein or DNA content from the cells cultured under normoxia and hypoxia.

12) For western blotting, wash the cells as quickly as possible, add the lysis buffer, and scrape the cells immediately. HIF-1 alpha is rapidly degraded under normoxic conditions.