Lactate Metabolism Regulates Tumour Growth and Progression in Glioblastoma

Background. Tumor microenvironment (TME) plays a pivotal role in establishing malignancy and it is associated with high glycolytic metabolism and increased lactate production accumulating in TME through monocarboxylate transporters (MCTs). Several lines of evidence suggest that lactate also serves as a signalling molecule through its receptor HCAR1thus functioning as a paracrine and autocrine signalling molecule in TME. The aim of the present study was to investigate the role of lactate in glioblastoma (GBM) progression and metabolic reprogramming in an in vitro and in vivo model. Methods. Cell proliferation, migration and clonogenicity assay were performed in vitro on three different human GBM cell lines. Protein expression of MCT1, MCT4 and pharmacological lactate receptor (GPR81) were evaluated both in vitro and in a zebrash GBM in vivo model. These results were further validated in patient-derived GBM biopsies. Results. Our results showed that lactate signicantly increased cell proliferation, migration and colony formation capacity of GBM cells, both in vitro and in vivo. We also showed that lactate increased MCT1 and HCAR1 expression. Moreover, lactate modulated epithelial-mesenchymal transition protein markers E-Cadherin and β-Catenin. Interestingly, lactate induced mitochondrial mass and OXPHOS gene suggesting an improved mitochondrial tness. Similar effects were observed after treatment with 3,5-Dihydroxybenzoic acid, a known agonist of GPR81. Consistently, GBM zebrash model exhibited an altered metabolism and increased expression of MCT1 and HCAR1 leading to high levels of extracellular lactate and thus supporting tumor cell proliferation. Our data from human GBM biopsies also showed that in high proliferative GBM biopsies, Ki67 positive cells expressed signicantly higher levels of MCT1 compared to low proliferative GBM cells. Conclusions. Our data suggest favours neighbourhood by cooperating with their glycolytic metabolism, sensing and removing extracellular lactate. particular, lactate its transporter GBM following 72 h of lactate treatment. Figures presented are the representative of at least four independent experiments and values represent the means ± SEM of experiments performed in quadrupled. HCAR1 gene expression (d) in U-87 MG, A-172 and U-251 MG cells following 24 h of lactate treatment. HCAR1 (e), MCT1 (f) and MCT4 (g) gene expression in U-87 MG, A172 and U-251 MG cells following 24 h of 3,5-DHBA treatment. Values means experiments performed in quadrupled. p values < 0.05 were considered to be statistically signicant (*p < < vs gene expression in U-87 MG cells (a-b), A-172 cells (c-d) and U-251 MG cells (e-f) following 24 and 48 h of treatment. Computerized analysis of mitotracker uorescence intensity on the control versus lactate 18 hours after treatment (g-h). Figures presented are the representative of at least three independent experiments. Values represent the means ± SEM of experiments performed in quadrupled. p values < 0.05 were considered to be statistically signicant (*p < 0.05; **p < 0.01; ***p < 0.001 vs untreated).


Background
Glioblastoma (GBM) represents the most common primary brain tumor in the adult population and is classi ed by WHO as a grade IV glioma. Current therapeutic approach for newly diagnosed GBM relies on surgical resection, radiotherapy and chemotherapy (i.e. temozolomide) [1]. However, despite aggressive therapeutic regimens, these tumors still have a dismal prognosis with medial overall survival of 12 to 15 months. Histologically, GBM is a highly cellular glioma composed by glial cells with signi cant pleomorphism and nuclear atypia [2]. Such cellular features are coupled with microvascular proliferation and palisading necrosis characterized by regular areas of necrosis and dense accumulation of GBM cells [2]. GBM characteristics are related to cell proliferation, usually assessed by evaluating KI-67 expressing cells classifying high proliferative index (HPI, KI-67 positive cells > 30%) and low proliferative index (LPI, KI-67 positive cells < 30%). Furthermore, GBM cell proliferation, migration and invasiveness are closely related to availability of blood-derived nutrients and oxygen. Indeed, two niches have been described in GBM in relation to availability of oxygen, the so-called perivascular niches, in which GBM cell receive glucose and oxygen from blood stream and oxidative phosphorylation in these cells determines highly e cient metabolism, and the GBM hypoxic niches, for example tumor core, in which low oxygen levels shapes metabolisms towards a glycolytic state inducing lactate accumulation [3]. Indeed, such tumors have a rapid rate of glucose consumption and convert large amounts of glucose into lactic acid, even in the presence of oxygen [4]. This metabolic phenotype, known as Warburg effect contrasts sharply with that observed in normal tissues in which glycolysis occurs mainly in hypoxic conditions [5].
To maintain enhanced glycolytic ow, glioblastomas require rapid out ow of lactic acid into the tumor microenvironment (TME), facilitated by a series of plasma membrane transporters called monocarboxylate transporters (MCTs) [6]; among these only four (MCT 1-4) are known to play a role in lactic acid transport in mammalian tissues, including cancers [7] and MCT1 and MCT4 have been implicated in multiple aspects of GBM progression including angiogenesis, cell proliferation and immunity modulation [8]. Glycolytic cancer cells are known to upregulate lactate export by increasing MCT4 expression to better adapt to lactate accumulation. In contrast, tumor cells of oxidative tumors have been reported to upregulate MCT1 expression to mediate lactate uptake from the extracellular environment to fuel metabolism [9]. A recent report suggests that this dynamic may create a metabolic symbiosis between the two GBM subpopulations maintaining a favorable environment for both subtypes [8,10].
Besides having a role as end-product metabolite of glycolysis and being utilized by cellular metabolic programs to produce energy, lactate also acts as signalling molecule through its receptor HCAR1 (GPR81) [11]. Therefore, extracellular lactate is not a simple bystander causing milieu acidi cation but it also serves as a paracrine and autocrine signalling molecule in TME [12]. Elevated expression of HCAR1/GPR81 was found in carcinomas of the breast, pancreas and cervix, despite negligible expression in the corresponding benign epithelium [12,13]. Several groups have identi ed autocrine roles for HCAR1/GPR81 in TME, where lactate produced by tumor cells activates HCAR1/GPR81 and confers cancer-promoting phenotypes [14], including upregulation of transporter MCT1 and MCT4 and the secretion of factors that promote angiogenesis and tumor progression [15].
The aim of the present study was to assess the role of lactate metabolism in cancer growth and progression in several glioblastoma cell lines, in pathological specimens and in an in vivo model.

Clonogenic assay
Colony assays performed by seeding cells in 6-well plates at low density (2000 cells/well) and allowing growth for 10 days. Colonies were xed, stained with crystal violet and colonies were quanti ed with Operetta high content screening (HCS) System (Perkin Elmer). The experiments were done in quadruplicates.

Real-Time Monitoring of Cell Proliferation
xCELLigence experiments were performed using the RTCA (Real-Time Cell Analyser) DP (Dual Plate) instrument according to manufacturers' instructions (Roche Applied Science, Mannheim, Germany and ACEA Biosciences, San Diego, CA). The RTCA DP Instrument includes three main components: (i) RTCA DP Analyser, which is placed inside a humidi ed incubator maintained at 37°C and 5% CO2, (ii) RTCA Control Unit with RTCA Software preinstalled and (iii) E-Plate 16 for proliferation assay. First, the optimal seeding number was determined by cell titration and growth experiments. After seeding the optimal cell number (3000 cells/well), cells were treated and automatically monitored every 15 min for 24h. Optimal cell number was determined in a preliminary set of experiments (data not shown) to obtain a signi cant cell index value and a constant cell growth during the entire duration of the experiment.

Cell Migration
Cell migration was studied by employing the "wound healing" assay. Brie y, cells were seeded in 24 wells dishes and cultured until con uence. Cells were treated with vehicle, lactate or 3,5 -DHBA and were then scraped with a 200 µl micropipette tip and monitored at 0, 24, and 48 h. The uncovered wound area was measured and quanti ed at different intervals with ImageJ 1.37v (NIH).
Real-time RT-PCR for gene expression analysis RNA was extracted by Trizol® reagent (Invitrogen, Carlsbad, CA, USA). First-strand cDNA was then synthesized with Applied Biosystem (Foster City, CA, USA) reverse transcription reagent. Quantitative realtime PCR was performed in Step One Fast Real-Time PCR System Applied Biosystems, using the SYBR Green PCR MasterMix (Life Technologies, Monza, Italy). The speci c PCR products were detected by SYBR Green uorescence. The relative mRNA expression level was calculated by the threshold cycle (Ct) value of each PCR product and normalized with that of actin by using a comparative 2 − ΔΔCt method. The sequence of primers used are presented in Table 1. Mione. Fishes with somatic and germline expression of oncogenic HRAS were generated as described [17,18].

Gene expression analysis
The analysis of expression of genes involved in glycolysis in zebra sh brain tumors was performed on previously generated data (GSE74754, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE74754 ). The heatmap was generated using the web application heatmapper (http://www.heatmapper.ca/) .
For gene expression analysis of further samples, total RNA was extracted from larval heads and brains/tumors with TRIzol reagent (Invitrogen). Total RNA was cleaned up using RNeasy Mini Kit (Qiagen) following the manufacturer's instructions and treated twice with DNase I (1 unit/µg RNA, Qiagen). The RNA concentration was quanti ed using nanodrop2000 (Thermo Fisher) and VILO superscript KIT (Thermo Fisher) was used for First-strand cDNA synthesis according to the manufacturer's protocol. qRT-PCR analysis was performed using qPCR BIO Sygreen Mix (Resnova -PCR Biosystem) using a standard ampli cation protocol. The primers used for zebra sh mct1 were: forward 5'-GTCACCATTGTGGAATGTGC-3' and reverse 5'-TCATCATAGATATCGTTGAGTCGTC-3'; for hcar1 zebra sh were: forward 5'-CATCGTCATCTACTGCTCCAC-3' and reverse 5'-GCTAACACAAACCGCACA-3'; for zebra sh rps11 (housekeeping): forward: 5'-ACAGAAATGCCCCTTCACTG-3' and reverse: 5'-GCCTCTTCTCAAAACGGTTG-3'. Real-time PCR was performed with a CFX96 Real-Time PCR Detection System (Bio-Rad) machine. Q-PCR analysis was performed with Microsoft Excel and Graphpad Prism. In all cases, each PCR was performed with triplicate samples and repeated with at least two independent samples.

Immuno uorescence in zebra sh
Adult zebra sh resulting from crosses between zic:Gal4 and UAS:RAS, or from somatic expression of UAS:RAS [19], were screened under a uorescent stereomicroscope for the presence of GFP-HRAS G12V brain masses. Positive sh (over 90% of screened sh) were sacri ced by MS222 overdose, their brains removed, xed and sectioned as previously described (Mayrhofer et al., 2020).
Sections were then washed in PBS (pH 7.4) and incubated primary antibodies diluted in PBS containing 5% normal goat serum and 0.1% triton x-100 at 4°C overnight. The antibody used and their dilutions were as follows: MCT1(abcam, 1:100) and HCAR1 (abcam, 1:100), Phospho Histone 3 (Abcam, 1:1000). A secondary antibody conjugated with Alexa 546 (Abcam, 1:250) was used for 2 hour at room temperature, and nuclei were counterstained with DAPI. Images were acquired using an inverted Leica TSP8 confocal microscope. For whole-mount immuno uorescence of 5 day postfertilization (dpf) zebra sh, larvae of the zic: Gal4 line (controls) or zic:Gal4 x UAS:RAS line (tumor) were treated with 20 mM lattate or 10 mM AZ3965 in 1% DMSO in E3, or with 1% DMSO alone. Solutions with the drugs were changed every day starting at 1dpf till 5dpf, when the larvae were culled by anaestetic overdose, xed in 4% PFA for 2 to 12 hrs at 4 C, their brains carefully removed under a stereomicoscope and processed with Ph3 antibody, diluted 1:1000 in 5% NGS, 0.5% Triton X100 in PBS overnight. A secondary antibody conjugated with Alexa 546 was used for 6 hour at room temperature. Images were acquired using an inverted Leica TSP8 confocal microscope, after equilibrating the brains in 100% glycerol.

Seahorse on zebra sh
For Seahorse analysis, tumors from adult sh or control brains were dissociated with a pipette tip in the assay medium provided by the manufactor, passed through a 40 mM sieve and counted. 50K cells were seeded on poly-L-Lysin coated Seahorse XFP plates and incubated for 20 min in the absence of CO2 before adding medium up to a nal volume of 180 ul.
XF mitostress test kit including oligomycin, carbonyl cyanide p-tri uoromethoxy-phenylhydrazone (FCCP), and Rotenone A were obtained from Seahorse Bioscience Inc. (Billerica, MA, USA). XFp cell culture plates, sensor cartridges and XF base medium were also purchased from Seahorse Bioscience Inc. The Agilent Seahorse XFp Sensor Cartridge is hydrated in Agilent Seahorse XF Calibrant at 28°C in a non-CO2 incubator overnight. Control and tumor zebra sh brain cells are plated in the Agilent Seahorse XFp Cell Culture Miniplate at the desired density (50 K per well) using the appropriate cell culture growth medium. PBS 1X is added to the chambers to prevent evaporation of the culture medium. Within 1 hour from plating the Agilent Seahorse XFp Cell Culture Miniplate is put into a 28°C non-CO2 incubator for 1 hour prior to the assay.

Mito stress test assay
Assay medium is prepared by supplementing Agilent Seahorse XF Base Medium with 1 mM pyruvate, 2 mM glutamine, and 10 mM glucose bringing the pH to 7.4 with 0.1 N NaOH. Cells are placed in 28°C incubator with 5% CO2.
Injections of oligomycin, FCCP and Rotenone A were diluted in the assay medium following Agilent Seahorse XFp Mito Stress Test User Guide and loaded into ports A, B and C, respectively. The machine was calibrated, and the assay was performed using mito stress test assay protocol as suggested by the manufacturer (Seahorse Bioscience, Billerica, MA, USA). ECAR was measured under basal conditions followed by the sequential addition of oligomycin, FCCP and Rotenone A.

Data analysis
The XF reports of mito stress data were analysed with the freeware Wave and exported to Excel and Prism for further analysis and visualization. Human gene expression

Dataset selection
The NCBI Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/) [20] was used to select transcriptomes datasets of interest. Mesh terms "human", "glioblastoma", and "tumor grade", were used to identify the datasets. We sorted the datasets by the number of samples (High to Low), age and sex of the participants and by the clinical data made available by the authors. We selected the GSE108474 dataset [21] over the others available for the number of subjects recruited (541), for the availability of clinical data (tumor staging) and for the variety of tumors analyzed (glioblastoma, oligodendrocytoma, astrocytoma and normal subjects). Data processing, experimental design and statistics To process and identify Signi cantly Different Expressed Genes (SDEG) within the datasets, we used the MultiExperiment Viewer (MeV) software (The Institute for Genomic Research (TIGR), J. Craig Venter Institute, La Jolla, USA). In cases where multiple genes probes have insisted on the same GeneID NCBI, we used those with the highest variance.
For GSE108474 (Table 2) we performed a statistical analysis with GEO2R, applying a Benjamini & Hochberg (False discovery rate) [22][23][24]. In Table 2 the sample detection from GSE dataset. Correlations were determined using Pearson correlation. All tests were two-sided and signi cance was determined at adjusted p value 0.05. The dataset selected was transformed for the analysis in Z-score intensity signal. Z score is constructed by taking the ratio of weighted mean difference and combined standard deviation according to Box and Tiao (1992) [25]. The application of a classical method of data normalization, z-score transformation, provides a way of standardizing data across a wide range of experiments and allows the comparison of microarray data independent of the original hybridization intensities. The z-score it is considered a reliable procedure for this type of analysis and can be considered a state-of-the-art method, as demonstrated by the numerous bibliography [26][27][28][29][30][31][32][33][34][35][36][37].
The e ciency of each biomarker across the different tumor grade was assessed by the receiver operating characteristic (ROC) curve analyses [38][39][40]. The ROC curves analyzed brain biopsies of healthy subjects (NT) vs glioblastoma patients, astrocytoma vs glioblastoma, and oligodendroglioma vs glioblastoma.
The area under the ROC curve (AUC) and its 95% con dence interval (95% CI) indicates diagnostic e ciency. The accuracy of the test with the percent error is reported [41].

Results
We rst analysed the effects of lactate on 3 human GBM cell lines (i.e. U-87 MG, A-172 and U-251 MG) by performing a clonogenic assay on lactate exposed cells ( gure S1  24 and 48 hours in all tested cells lines ( gure 1d-g). We also con rmed the effects of HCAR stimulation through 3,5-DHBA on cell migration, nding a signi cantly reduced % of wideness of scratch assay test at 48 hours in all tested cells lines ( gure 1d-g).
In an effort to link lactate, as a positive modulator of cell proliferation and migration, to the underlying molecular mechanisms activated in GBM cell lines, we performed western blot analysis for lactate transporters MCT1 and MCT4, and for β-Catenin and E-Cadherin on control and lactate treated U-87 MG, A-172 and U-251 MG cells.
We found that U-87 MG cells responds to increased levels of extracellular lactate by increasing the levels of MCT1 transporter of about 2.5-fold as compared to control cultures and slightly, but signi cantly, reducing MCT4 expression levels ( gure 2a).
Importantly, the β-Catenin protein levels were found to be signi cantly increase of about 6-fold in lactate exposed U-87 MG cells and such a modulation was coupled with reduced expression levels of E-Cadherin ( gure 2a).
Notably, analysis of A-172 and U-251 MG exposed to increased extracellular lactate levels, revealed some differences in cellular responses as compared to U-87 MG cells. Indeed, we con rmed that exposure to lactate increased MCT1 and β-Catenin expression levels in both A-172 ( gure 2b) and U-251 MG ( gure 2c) but showed that both cell lines respond to lactate also inducing signi cantly higher MCT4, increased of about 1.2-fold in both cell lines, and E-Cadherin expression levels ( gure 2b-c).
Given the evidence on cellular modulation exerted by increased extracellular levels of lactate, we sought to link molecular mechanisms underlying these phenomena with the activation of lactate receptor HCAR1. We rst investigated HCAR1 mRNA expression levels on U-87 MG, A-172 and U-251 MG cell lines after exposure to lactate, nding a signi cant increase of HCAR1 mRNA levels in all tested cells at 24 hours ( gure 2d). We then moved to evaluate the effects of 3,5-DHBA, con rming a signi cant increase of HCAR1 mRNA levels at 24 hrs post 3,5-DHBA incubation in all tested cell lines ( gure 2e).
To nd whether HCAR1 selective stimulation was able to increase lactate transporters MCT1 and MCT4 we also checked the mRNA expression levels of these transporters, nding that 3,5 To further expand our evidences on molecular mechanisms induced by the increase of extracellular lactate, we analyzed a panel of mRNAs of genes involved in mitochondrial activity and energy metabolism. Our data show that U-87 MG signi cantly increase of about 4-fold the relative mRNA levels of transcription factor A mitochondrial (TFAM), PPARG coactivator 1 alpha (PGC1a) and sirtuin 1 (SIRT1) ( gure 3a-b), coupled with an overall increase of ATP synthase (ATP syn), cytochrome c oxidase subunit 4 (COX IV) and COX II, mitochondrial Cytochrome b (CYTB) and mitochondrial NADH-ubiquinone oxidoreductase chain 4 (ND4, gure 3a), when exposed to lactate for 24 or 48 hours as compared with untreated cells ( gure 3a-b). These observations were con rmed in A-172 ( gure 3c-d) and U-251 MG cell lines ( gure 3e-f). Speci cally, we observed superimposable effects on A-172 as compared to U-87 MG cells, where U-251 MG showed an increase of about 15-fold of TFAM, PGC1a and SIRT1 at 48 hours as compared to untreated cells ( gure 3f), coupled with a slight reduction of COX IV mRNA at the same timepoint (-1.87 ± 0.1 log 2 fold change over control, gure 3e). We also performed a computer-assisted analysis of mitotracker uorescence intensity on control versus lactate treated cells, nding that lactate was able to signi cantly increase cytoplasmic mitotracker intensity 18 hrs post treatment ( gure 3g-h).
To link intracellular mediators of mitochondrial tness with HCAR stimulation, we performed an mRNA expression level analysis of PGC1a, TFAM, SIRT1, ATP syn, COX II and COX IV on 3,5-DHBA stimulated cells. Our analysis revealed that U-87 MG cells exposed to 3,5-DHBA recapitulate the molecular mRNA activation observed with lactate ( gure 4a-b). Indeed, all tested genes, except for TFAM, were signi cantly increased in cultures exposed to HCAR stimulation ( gure 4b). These data were con rmed in A-172 cells that showed increased levels of all tested genes upon 3,5-DHBA stimulation ( gure 4c-d). Finally, U-251 MG showed a very similar mRNA expression pro les, but we observed that HCAR stimulation through 3,5-DHBA did not modulate PGC1a expression at tested timepoint on this cell line ( gure 4e-f).
To nally link HCAR stimulation with the effects on mitochondria observed on GBM cell lines exposed to increased extracellular lactate levels, we performed a mitotracker analysis, nding a signi cant increase of normalized intensity in 3,5-DHBA stimulated cells as compared to control cultures ( gure 4g).

Given the capability of extracellular lactate to modulate β-Catenin and E-Cadherin expression levels, we performed a western blot analysis on 3,5-DHBA stimulated A-172 cells. Our analysis revealed that HCAR1
activation induces a signi cant increase of β-Catenin protein expression levels as compared to control cultures and this phenomenon was coupled with a signi cant reduction of E-Cadherin ( gure 4h), revealing that lactate may also act via additional mechanisms that induce E-Cadherin not related to HCAR1 activation.
To investigate whether lactate accumulation, resulting from increased glycolysis, may have similar effects in vivo, we used a zebra sh model of glioblastoma [17] ( gure 5a), and analysed the metabolic phenotype of these tumors. Comparison of the expression levels of 29 genes encoding for enzymes and transporters involved in the glycolytic pathway acquired through RNA-Seq (GSE74754, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE74754), revealed increased expression (log2 FC >1.2, P-value <0.001 or adjusted P-value <0.05) of 26 out of 29 genes, with aldh1a3, hk2 and hcar1-3 being the most upregulated in tumors ( gure 5b). We then performed a mitostress test on freshly dissociated zebra sh control and tumor brains using the Seahorse XFp apparatus. This test con rmed that upon blockage of energy production through mitochondrial respiration, zebra sh GBM cells experience a huge increase of the extracellular acidi cation rate (ECAR), indicating a prominent role of anaerobic glycolysis in energy production, accompanied by increased proton leak (leading to high ROS production) and lower ATP yield ( gure 5c). Staining for MCT1 and HCAR1 in sections of zebra sh brain tumors revealed an increase in the number of both MCT1+ and HCAR1+ cells ( gure 5d), whereas q-PCR analysis of mRNA expression for mct1 and hcar1 revealed a signi cant increase in expression for mct1 in adult tumors compared to control brain, and a signi cant increase in expression of hcar1 in both adult brain tumors and in 5days post-fertilization (dpf) larvae expressing oncogenic RAS ( gure 5e).
Then, we evaluated the effects of exposing to lactate or to the MCT1 inhibitor, AZ3965 (AZD), on the proliferation rate of control brains and brains expressing oncogenic RAS, using immunostaining for a mitotic marker (phospho-serine 10 on histone 3, PH3). Incubation of developing larvae from 1 to 5 dpf with 20 uM lactate induced a signi cant increase in proliferation in brains expressing oncogenic RAS, but not in control brains, while treatment with 10 uM AZD did not affect the proliferation rate in either control or RAS expressing brains ( gure 5f).

MCT1 gene expression analysis as a diagnostic and prognostic marker of glioma
The MCT1 gene expression analysis obtained from the GSE108474 dataset showed that there were signi cant differences when the expression levels obtained from brain biopsies of glioblastoma patients were compared to the other brain tumors stages. ( gure 6). Speci cally, patients with glioblastoma expressed signi cantly higher levels of the MCT1 messenger in the brain than patients with oligodendrocytoma (p<0.0001), astrocytoma (p<0.0001), or healthy subjects (p<0.0001) ( gure 6a). This nding was con rmed by the signi cantly positive correlation between MCT1 expression levels and tumor grade (r = 0.4026; p = 0.0223) ( gure 6b). According to these results, we investigated the prognostic potential of MCT1 expression in the progression of main brain tumors. Currently the expression analysis of Isocitrate Dehydrogenase (NADP (+)) 1 (IDH1) and the identi cation of its main mutations (e.g. R132H) are used for glioma diagnosis and prognosis [42]. By carrying out a Pearson correlation analysis between MCT1 and IDH1 brain tumor expression levels, we highlighted that in glioblastoma patients the expression levels of the two genes were signi cantly closely inversely correlated (r = -0.4163, p <0.0001) ( gure 6c). Furthermore, in order to evaluate the potential diagnostic ability of MCT1 gene expression to discriminate against the brain tumors stages, we performed a Receiver operating characteristic (ROC) analysis. We con rmed the diagnostic ability of MCT1 to discriminate the glioblastoma patients from healthy subjects (AUC=0.7558, p<0.0001) ( gure 6d) or from the patients affected to astrocytoma (AUC=0.7775, p<0.0001) ( gure 6e) or oligodendrocitoma (AUC=0.8104, p<0.0001) ( gure 6f).

Discussion
Cell metabolism and its related intercellular signalling has been shown to be of great importance in a number of physiological and pathological processes [43]. In the present study, we rst evaluated the effects of lactate on three human GBM cell lines, nding that it increases both migration and cell proliferation. Such a phenomenon was linked to a potential lactate dependent HCAR1 activation, as observed using 3,5-DHBA, a selective HCAR1 agonist.
Several authors showed that stimulation of HCAR1 leads to the activation of cell survival signalling promoting cell proliferation via the inhibition of apoptosis and stimulates the secretion of several angiogenic factors in a PI3K/Akt-CREB signalling pathway-dependent manner [44]. Interestingly, an essential part of the repair process after a neonatal brain injury is the generation of new cells by increase of proliferation and differentiation of stem cells, Lauritz H. K. et al., by neurosphere assays, demonstrated that the cells lacking HCAR1 had reduced proliferation ability [45].
Moreover, MCT1 is mainly used by oxidative cells to intake extracellular lactate and MCT4 is mainly used to release accumulated lactate into the extracellular milieu, in many cases by hypoxic and/or highly glycolytic cells [46][47][48]. Our data support the hypothesis that lactate leads GBM cells to increase HCAR1, acting as a sensor, levels and MCT1, mediating lactate intake from the extracellular milieu. This phenomenon is coupled with increased mitochondrial content and tness, thus prompting GBM cells towards oxidative metabolism. It is worth noticing that this mechanism is not related to the increased lactate level itself but is dependent on the agonism on HCAR1 receptor. Indeed, we were able to reproduce these metabolic reshaping using the selective HCAR1 agonist 3,5-dhba. Consistently, a study performed in GPR81-silenced pancreatic cancer cells led to reduced mitochondrial activity and survival in several cancer cell [49]. In particular, several cancer cell types, including colon, breast, lung, cervical, and pancreatic showed an increase of HCAR1 expression and functional studies indicated that it is important for lactate regulation of genes involved in lactate uptake and metabolism. Moreover, HCAR1 is critical for cancer cell survival only when glucose was absent and in the presence of lactate [50].
Interestingly, we observed critical differences in cell response to HCAR1 activation analyzing MCT4 levels. Indeed, we observed that 24 hrs of exposure to lactate mediated a reduction of MCT4 protein levels in U-87 MG, whether we found a signi cant MCT4 increase in both U-251 MG and A-172. Such differential response to lactate among tested cell lines, may be linked to the metabolic reshaping of these cells. Given the insights coming from in vitro experiment on relevant human GBM cell lines, we enrolled a HRAS overexpressing zebra sh model of GBM to test whether similar metabolic changes are taking place in this model. Our data con rmed a widespread upregulation of glycolytic enzymes, with upregulation of HCAR1, thus indicating a prominent role for lactate signalling. In the tumor microenvironment HCAR1 upregulation was coupled with a signi cantly increased proton leak and less e cient ATP production.
The increased expression of lactate transporters (mct1) and sensor (hcar1) was already present at 5 dpf, when tumors start to grow.
Lactate exposure determined a signi cant increase in proliferating PH3 positive cells in RASoverexpressing zebra sh brain, but not in control brains, and this was reverted by selective inhibition of MCT1. This evidence suggests that lactate intake support cell proliferation in cancer and that metabolic reshaping is a critical stimulus in GBM microenvironment. Thus, both cell culture and in vivo studies, using different approaches and different genes, converge toward the same conclusion, i.e. that glycolysis is prominent in GBM and leads to massive lactate production which shapes the microenvironment towards an aggressive phenotype and represent a valid therapeutic target.
Our data from human GBM biopsies were also consistent with preclinical evidence we are providing herein. We observed that in high proliferative GBM biopsies, KI67 negative cells were expressing Further con rmation of our study was obtained by analyzing the human GSE108474 dataset. The analysis allowed us to highlight that MCT1 is signi cantly modulated during the progression of the disease. In particular, signi cant expression changes were highlighted with the increase in the degree of malignancy. Furthermore, our results showed that MCT1 can potentially be used in order to discriminate patients with glioblastoma versus those with astrocytoma and oligodendrocytoma. These data are in agreement with the current bibliography which considers MCT1 a new prognostic biomarker and potentially target in human glioblastoma [60]. Interestingly, the correlation analysis between MCT1 and IDH1 brain expression levels in glioblastoma patients was inversely proportional, further con rming recently obtained data in which mutant IDH1 expression is associated with down-regulation of monocarboxylate transporters [61].

Conclusion
In conclusion, we showed that lactate is involved in various mechanisms favoring tumor development and progression. In particular, lactate possesses a dual role being involved in the metabolic changes of tumor cells and acting as a molecule promoting cellular signaling through its membrane receptors. The ability to metabolically shift from glycolytic to oxidative metabolism and vice versa, is likely to confer an advantage in survival, progression and drug resistance. A glycolytic metabolism (Warburg effect) certainly in the rst phase of disease expansion, determines an advantage in tumor proliferation. The lactate thus produced in the tumor microenvironment favors, on the one hand, the immune escape mechanisms, on the other, it may modify the metabolism of the adjacent tumor cells, becoming more oxidative and therefore more resistant also to antiblastic therapies. Therefore, lactate metabolism may be considered as a therapeutic target to develop novel pharmacological strategies to GBM therapy and improve the outcome and quality of life of such patients. Availability of data and materials

List Of Abbreviations
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests      MCT1 expression analysis from the human brain tumor GSE108474 dataset Analysis of MCT1 gene expression in brain biopsies of patients with astrocytoma, oligodendrocytoma, glioblastoma, and healthy subjects. b) Pearson correlation analysis between MCT1 expression levels and tumor grade of brain biopsies obtained from patients affected by main brain tumors. c) Pearson's correlation between MCT1 and IDH1 expression levels in brain biopsies of patients with glioblastoma. d) Receiver operating characteristic (ROC) analysis between MCT1 brain expression levels in healthy subjects vs glioblastoma patients, between glioblastoma patients vs astrocytoma patients (e), and vs oligodendrocytoma (f). Data are expressed as mean ± SD of at least four independent experiments. (*p<0.05; **p<0.005; ***p<0.001; ****p<0.0001).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.