1. Introduction
Brain and other central nervous system (CNS) tumours are the second largest cause of cancer-related deaths in children, after leukaemia. Medulloblastomas (MBs) account for 20% of paediatric brain tumours, making them the most common type of solid, malignant childhood brain tumour [
1]. MB manifests in the cerebellum and has a tendency to spread to the spinal cord. Four main subgroups of MB have been identified and are now recognised by the World Health Organisation (WHO): as Wnt-activated, Shh-activated TP53 wild-type, Shh-activated TP53 mutant and non-Wnt/non-Shh, the latter of which has been subdivided into “Group 3” and “Group 4” [
2,
3]. Recent studies reveal even further divisions [
4]. Each group has disparate features including demographics, clinical outcomes, characteristic genetic abnormalities and rates of metastasis. Although the current five-year event-free survival is >80% for low-risk children, and 60–70% for high-risk children [
5], the quality of life of individuals after treatment is often significantly reduced.
Regardless of oxygen availability, cancer cells commonly use aerobic glycolysis for ATP production; this is known as the Warburg effect [
6]. Pyruvate is diverted away from the mitochondria, where it normally undergoes oxidative phosphorylation (OXPHOS) to generate ATP. Instead, it is converted to lactate by LDHA and nicotinamide adenine dinucleotide (NAD+) is regenerated from (NAD)H in the process [
7]. NAD+ is required by glyceraldehyde 3-phosphate dehydrogenase to maintain glycolysis and ATP production. This reaction also occurs occasionally in some specialised non-neoplastic cells during rapid proliferation and other non-neoplastic cells when oxygen is unavailable. Nevertheless, as this is the preferential method of ATP production used by cancer cells, it is an attractive target for cancer therapies. We previously reviewed the regulation and function of LDHA and its therapeutic potential in brain tumours [
8].
Mouse models have shown that during normal development of the cerebellum, cerebellar granule neural precursors (CGNPs) are stimulated by Shh to undergo aerobic glycolysis. Shh and phosphoinositide 3-kinase signalling has been shown to stimulate aerobic glycolysis in CGNPs in a hexokinase-2 (HK2)-dependent manner [
9]. Furthermore, inhibition of aerobic glycolysis by deleting HK2 in Shh MB mouse models reduced MB malignancy by promoting differentiation and reducing proliferation [
9]. Shh MBs also exhibit elevated lipid synthesis implicated in aerobic glycolysis [
10].
Using magnetic resonance spectroscopy and 18fluorodeoxyglucose positron emission tomography, MBs in patients have been shown to have a glycolytic metabolic phenotype and a recent study has shown elevated lactate levels were associated with a Group 3/4 subgroup [
11,
12]. MBs also have elevated expression of monocarboxylate transporter 1 (MCT1), which transports lactate and pyruvate across the plasma membrane, also indicating a glycolytic phenotype [
13]. Moreover MCT1 is inhibited by miR-124, which is commonly downregulated in MB, and ectopic expression of miR-124 has been shown to inhibit MB proliferation by blocking cell cycle progression at G1.
Cellular metabolism can be influenced by many pathways, including several implicated in MB. The Wnt signalling network has been shown to increase glycolysis and decrease oxidative phosphorylation [
14,
15]. C-Myc, which is overexpressed in Group 3 MB and activated downstream of Wnt signalling pathways, is also known to promote glycolysis in many cancers [
2,
16] and upregulates LDHA expression [
17]. Although MB is made up of four distinct subgroups, together these studies indicate that aerobic glycolysis is a common feature of most if not all MBs.
LDHA has not yet been explored as a therapeutic target for MB. We used oxamate, a structural analogue of pyruvate, which competes with pyruvate to inhibit LDHA activity (not expression) [
18] and LDHA siRNA to investigate the therapeutic potential of targeting LDHA in MB. We hypothesised that LDHA inhibition would result in a decrease in lactate concentrations and a change from a glycolytic to an oxidative phosphorylation metabolic phenotype, leading to decreased medulloblastoma, proliferation and motility.
2. Materials and Methods
Cell culture: Human MB cell lines UW402, Res256 (Dr. John Silber, University of Washington, Seattle, WA, USA) were cultured in Dulbecco’s Modified Eagle’s medium (Gibco, Paisley, UK), 10% FBS. Human MB cell line DAOY (ATCC, London, UK) was cultured in Eagle’s minimum essential medium (ATCC), 10% FBS. Cells were maintained in a 37 °C, 95% air, 5% CO2 in a humidified incubator (NUAIRE, Caerphilly, UK). Cells were routinely screened for mycoplasma using the MycoAlert™ mycoplasma detection kit (Lonza, London, UK).
Analysis of patient data: Gene expression datasets were obtained and analysed using the R2 genomics and visualisation platform (
http://r2.amc.nl) [
19] LDHA expression was analysed from a selection of CNS tissues available from the Normal Various-Roth-504–MAS5.0–u133p2 dataset and in MBs using the Tumor Medulloblastoma-Gilbertson-76–MAS5.0–u133p2 dataset.
Western blot: Cells were lysed with M-PER® (Thermo Scientific, Basingstoke, UK) and Halt™ protease and phosphatase inhibitor cocktail (Thermo Scientific, Basingstoke, UK, 1:100). 50 µg of protein was separated by SDS-PAGE in Tris-Glycine 4–20% gels (Thermo Scientific, Basingstoke UK) and transferred onto Immun-Blot® PVDF membrane (Bio-Rad, Oxford, UK). Blots were incubated overnight at 4 °C in 5% milk blocking buffer containing LDHA (Novus Biologicals, Oxford, UK) or Cyclophilin A antibody (Abcam, Cambridge, UK). The blots were washed in TBST and incubated with Goat anti-rabbit IgG horseradish peroxidase antibody (Promega, Southampton, UK) for 1 h at room temp. The blots were incubated with Luminata™ Forte Western HRP Substrate (Merck Millipore, Watford, UK) and imaged using the G-Box chemiluminescent imaging system (Syngene, Cambridge, UK). Semi-quantitative downstream analysis of protein expression was performed using Image J software.
Flow Cytometry: Cells were harvested and centrifuged at 300 RCF at 4 °C, the supernatant removed and pellet washed in PBS (Gibco, Poole, UK). The pellet was re-suspended in of BD cytofix/cytoperm™ fixation and permeabilisation solution (BD Biosciences, Oxford, UK), and incubated for 20 min at 4 °C in the dark. The cells were washed with wash buffer (PBS with 0.2% saponin, and 1% goat serum) and re-suspended with LDHA antibody (Novus Biologicals) and incubated for 30 min at 4 °C. Cells were washed with wash buffer and re-suspended with goat anti-rabbit IgG Alexa fluor 488 (Life Technologies, Warrington, UK), and incubated for 15 min at 4 °C. Cells were washed with wash buffer and re-suspended in staining buffer (PBS with 1% goat serum and 0.09% w/v sodium azide pH to 7.4–7.6) and 20,000 events from each sample analysed using a BD FACSCalibur™. Acquisition and analysis were performed using CellQuestPro software to determine cell line expression of antigen.
Oxamate: Cells were seeded and left to adhere overnight before treatment with oxamate (0 h). Oxamate (Sigma Aldrich, Poole, UK) stock was prepared at 0.5 M concentrations in dH2O and filtered, using a Millex®-GP 22µm filter unit (Merck Millipore, Watford, UK).
Small interfering RNA (siRNA) knockdown: Stock LDHA Silencer® select pre-designed siRNA (Ambion, Basingstoke, UK, 4392420) and Silencer® select Negative control #1 siRNA (Ambion, UK) in nuclease-free water (Ambion, UK) were diluted in Opti-MEM (Gibco, UK). The siRNA stock was mixed with jetPRIME® buffer and jetPRIME® reagent (Polyplus transfection®, Source Bioscience Life Sciences, Nottingham, UK) according to the manufacturer’s instructions. 15 nmol siRNA was added to 70,000 cells in 12-well plates and the media changed after 24 h. LDHA knockdown was verified by Western blots parallel to every functional siRNA experiment.
Lactate assay: Lactate concentrations were measured using a lactate assay kit (MAK064, Sigma Aldrich, UK) according to the manufacturer’s instructions. Cells were lysed and protein concentration determined by BCA. 20 µL of lysates were analysed at 570 nm using POLARstar OPTIMA plate reader (BMG Labtech, Aylesbury, UK).
Glycolysis stress test and Mito stress test: Glycolysis and mitochondrial activity was measured using the Seahorse Bioanalyser 96XF with Seahorse Bioanalyser Glycolysis stress test kit and Mito stress test kit according to the manufacturer’s instructions respectively. Cells were treated with oxamate 24 h prior to analysis and data analysis was conducted using Wave software.
Trypan blue exclusion assay: Cells were harvested from six-well plates, transferred into Vi-Cell cups and counted using the Vi-Cell® (BeckMan Coulter, High Wycombe, UK) automated trypan blue dye exclusion assay. Seeding density was optimised for each cell line to ensure that no more than 85% confluence was achieved by the 72 h time point in the control wells.
Cell cycle analysis: Cells were harvested from six-well plates and centrifuged at 500 RCF at room temperature, the supernatant discarded then washed and re-suspended in PBS (Gibco, UK). The cells were fixed with cold 70% ethanol and stored at 4 °C for seven days. The ethanol-suspended cells were centrifuged at 800 RCF at 4 °C and the supernatant removed. The cells were washed with PBS and re-suspended in Chemometec Solution 3 (1 µg/mL DAPI, 0.1% triton x-100 in PBS). 10 µL of each sample was loaded into a chamber of a NC-Slide A8™ and >10,000 cells analysed using the NucleoCounter® NC-3000.
Live cell imaging: A scratch was made through a monolayer of cells in a 24-well plate using a pipette tip. The media was removed and replaced with 500 µL of appropriate media (with/without oxamate) and secured into a Ziess Axiovert 200 M microscope stage enclosed in a temperature (tempcontrol 37-2 digital, Meyer instruments, Houston, TX, USA), CO
2 and oxygen (Bold line, Okolab, Indigo Scientific, Baldock, UK) controlled incubator set to 37 °C, 95% air, 5% CO
2. Images were taken every 30 minutes using Volocity
® acquisition software (PerkinElmer, Chiltern, UK). The area of the gap was measured using T Scratch software developed by the Koumoutsakos group (CSE Lab, Zürich, Switzerland), at ETH Zürich [
20]. 10 cells from the edge of each scratch were tracked manually for the first 24 h using Volocity
® analysis software (PerkinElmer, Chiltern, UK).
Statistics: All experiments were completed together and independently in triplicate. All statistical analysis was carried out using GraphPad Prism 6 software (6.0, GraphPad Software, San Diego, CA, USA). p-values less than 0.05 were considered significant. t tests were used to analyse the WB and FC data. Ordinary one-way ANOVA was used to analyse the R2 database genomic data and to analyse 24 h of cell tracking data. Two-way ANOVA was used for all other experiments.