Pharmacological inhibition of the amino-acid transporter LAT1 demonstrates in vitro anti-neoplastic activity towards medulloblastoma

Most cases of medulloblastoma (MB) occur in young children. While the overall survival rate can be relatively high, current treatments combining surgery, chemo- and radiotherapy are very destructive for patient development and quality of life. Moreover, aggressive forms and recurrences of MB cannot be controlled by classical therapies. Therefore, new therapeutic approaches yielding good efficacy and low toxicity for healthy tissues are required to improve patient outcome. Cancer cells sustain their proliferation by optimizing their nutrient uptake capacities. The L-type amino acid transporter 1 (LAT1) is an essential amino acid carrier overexpressed in aggressive human cancers that was described as a potential therapeutic target. In this study, we investigated the therapeutic potential of JPH203, a LAT1-specific pharmacological inhibitor, on two independent MB cell lines belonging to subgroups 3 (HD-MB03) and Shh (DAOY). We show that while displaying low toxicity towards normal cerebral cells, JPH203 disrupts AA homeostasis, mTORC1 activity, proliferation and survival in MB cells. Moreover, we demonstrate that a long-term treatment with JPH203 does not lead to resistance in MB cells. Therefore, the present study suggests that targeting LAT1 with JPH203 is a promising therapeutic approach for MB treatment.


INTRODUCTION
Medulloblastoma (MB) is the most prevalent pediatric brain tumor [1]. Recent advances in genetic characterization of the disease have led to an international consensus classification of MB into four biologically and clinically relevant subtypes: the wingless (Wnt), the sonic hedgehog (Shh), and the more similar though molecularly distinguishable groups 3 and 4 [2,3]. The most recent studies combining genetics, epigenetics and RNA expression data have refined and further delineated several sub-groups within these groups [3][4][5][6]. MB subgrouping is used to orient the therapeutic approach that comprises surgical removal of the tumor, cranio-spinal radiation therapy (RT) and chemotherapy. Shh and Wnt subgroups show the highest overall survival while group 4 and, to a higher extent, group 3 MBs are of poor prognosis with a strong tendency to form metastasis [2,3]. Although the overall survival at 5 years reaches 70% in all MB types taken together, even peaking over 90% in the Wnt group, toxic effects due to the treatments often lead to irreversible damages that severely hamper the children's cognitive development and general quality of life. In particular, RT dose and age of exposure to RT are well-documented risk factors for long-term cognitive impairments in MB survivors [7][8][9]. Reducing RT doses has been proposed however it is unsuitable for high risk MBs [8,9]. Moreover, recurrences are fatal in a large majority of cases and are characterized by strong divergence of the relapsed tumors as compared to the initial tumors [10]. A challenge therefore lies in finding new treatments that combine high efficacy and little toxicity. Personalized targeted therapy appears to be a promising approach in this context [11,12]. Clinical trials involving Shh pathway targeted inhibitors have enrolled Shh group MB patients based on this assertion. To date, these trials have not led to the expected results as they have induced unbearable toxicity and tumor resistance to the treatment [13][14][15]. All these features support the need to develop and design novel therapies against alternative targets.
Rapidly growing tumors experience an increased demand for nutrients in order to sustain their proliferative metabolism. Up-regulation of processes related to the supply of glucose, amino acids and lipids is thus a hallmark of cancer metabolism [16,17]. Essential amino acids (EAAs), which cannot be synthetized de novo by human cells, are absolutely required for cancer cell proliferation. Indeed, some of these nutrients can also be converted to essential metabolite intermediates of the TCA cycle, further participating in tumor energy metabolism and macromolecule synthesis. Notably, catabolism of the branched chain EAA subclass (leucine, isoleucine, and valine) is required for growth of some brain tumors [18]. In addition, leucine is an essential signaling molecule required to sustain the activation of mTORC1, the master kinase of protein, lipid, nucleotide syntheses and cell proliferation [19,20]. Therefore, considering the high expression of LAT1 in MB cells [21], we hypothesized that targeting EAA uptake might be an innovative strategy for MB treatment. The L-type amino acid transporter 1, LAT1 (SLC7A5), is a 12-transmembrane protein responsible for Na +independent transport of large neutral EAA (Leu, Val, Ile, Phe, Trp, His, Met, Tyr) in tight association with chaperone CD98 (SLC3A2). LAT1 is an obligatory exchanger with the uptake of one amino acid (AA) being coupled to the efflux of another. LAT1 is overexpressed in aggressive human cancers including MB and has been described as a potential therapeutic target [21,22]. In a previous study, we demonstrated that LAT1 activity is essential for tumor growth [23]. Indeed, genetic disruption of LAT1 in colorectal and lung adenocarcinoma cell lines leads to EAA starvation, mTORC1 inactivation and growth arrest [23,24]. In addition, our lab with others demonstrated that treatment with JPH203, a specific inhibitor that targets LAT1 and no other LAT, recapitulates these effects in several cancer cell lines, including glioma cell lines [21,23,25,26]. However, the efficiency of this compound had never been addressed in MB. In the present study, we demonstrate that LAT1 is the main leucine transporter in two independent cell lines isolated from Shh (DAOY) and Group 3 (HD-MB03) MBs [27,28]. We subsequently show that JPH203 treatment disrupts AA homeostasis, mTORC1 activity as well as proliferation and survival of both MB cell lines. Further, in contrast to its effect on MB cell lines, JPH203 displays low toxicity on two normal cerebral cell types. Finally, we show that although MB cells try to adapt to a JPH203 chronic treatment by upregulating the expression of AA transporters, no resistant clones could be isolated from the cell populations. Altogether our results strongly suggest that JPH203 is a promising therapeutic candidate for treating MB.
Immunoreactive bands were detected with horseradish peroxidase-coupled anti-mouse or antirabbit antibodies (Promega) using the ECL system (Merck Millipore WBKLS0500).
Acquisition of the immunoblot images were performed using LI-COR Odyssey Imaging System.

Proliferation and viability assays
The different cell lines (2.5 ×10 4 cells for 7 days, 5 × 10 4 cells for 3 days) were seeded onto 6well plates in triplicates. We measured proliferation by trypsinizing the cells and counting them daily with a Coulter Z1 (Beckman) after 48 hours. The cell proliferation index was calculated as "fold increase" by standardizing each measurement to the cell number obtained 24 hours after seeding (day 0). Viability assays were performed using an ADAM-MC automatic cell counter (Nanoentek) according to the manufacturer's protocol. Briefly, cells were seeded in 12-well plates, treated for 48 hours with the indicated concentrations of JPH203 or the corresponding amounts of DMSO, trypsinized and the samples were analyzed according to the manufacturer's protocol.

Three-dimensional growth assay
DAOY and HD-MB03 3D cultures were prepared using ultra-low attachment 96-well plates (Corning). 5000 cells were seeded in 200µL medium per well. The spheroids were cultured for 8 days and pictures were taken with an AMG Evos microscope 40x objective (Thermo Fisher Scientific Inc). Spheroid areas were measured using ImageJ 1.51j8 software (National Institute of Health) [30].

Scratch assay
800000 DAOY cells were seeded in 60mm diameters and grew until they reached confluency (2 days). A wound was then done ine the monolayer by scratching it with the tip of a pipette.
The culture medium was then replaced by fresh medium containing DMSO or the indicated concentrations of JPH203.Paired measurements of the wound width were done at 0 and 8 hours at 6 different places for each condition.

Statistical analysis
Data are expressed as mean ± SD. Each experiment was performed at least three times.
Statistical analysis was done with the unpaired Student t test. Differences between groups were considered statistically significant when P < 0.05 in a Student's t-test.

LAT1 is the main leucine transporter in HD-MB03 and DAOY MB cell lines and is essential for AA homeostasis and mTORC1 activity
We first demonstrated that LAT1 and its chaperone CD98 are expressed in HD-MB03 and DAOY cell lines (Fig. 1A). Functional activity of LAT1 was quantified by measuring the Na + -independent rate of leucine transport in the presence or absence of JPH203, a specific LAT1 inhibitor (Fig. 1B). JPH203 completely abolished leucine uptake (Fig. 1B), suggesting that LAT1 is the main functional leucine transporter in these two MB cell lines. Next, we investigated the effects of LAT1 inhibition on the two AA-sensing pathways: GCN2 and mTORC1 (Fig.1C) [31]. In both cell lines, LAT1 inhibition resulted in the activation of the AA stress response pathway GCN2, observed through increased phosphorylation of GCN2 and EIF2α and upregulation of ATF4 expression (Fig.1C). Moreover, JPH203 treatment resulted in a strong decrease in mTORC1 activity, scored by the phosphorylation of its two effectors: p70-S6K1 and the ribosomal protein S6 (Fig.1C). Altogether these results demonstrate that JPH203 treatment leads to AA starvation and suggest that LAT1 activity is required for AA homeostasis in cells belonging to different MB subgroups.

Pharmacological inhibition of LAT1 impairs MB cell proliferation, survival and migration abilities
We next assessed the effect of JPH203-induced LAT1 pharmacological inhibition on cell proliferation and cell viability. Two concentrations of the inhibitor (20 and 30µM) strongly decreased the proliferation of HD-MB03 and DAOY cell lines ( Fig.2A). Moreover, while having a cytostatic effect at 20⎧M, JPH203 was cytotoxic at 30µM in both cell lines (Fig.   2B). This effect was stronger in HD-MB03 (30%) than in DAOY cells (7%) suggesting that the HD-MB03 cell line, belonging to the most aggressive subgroup of MBs and expressing the highest level of LAT1/CD98 complex (Fig. 1A), is also the most sensitive to LAT1 inhibition. The effect of 30µM of JPH203 was tested on murine primary cortical neurons (PCN) and non-tumoral cerebellar astrocytes (C8-D1A). The treatment had no significant effect on PCN viability and only slightly impaired astrocyte viability (Fig. 2C). The effect of JPH203 was then tested on spheroids generated with HD-MB03 and DAOY cells to assess the effect of LAT1 inhibition on the 3-dimensional (3D) growth of tumor cells. As found in 2D, JPH203 completely abolished HD-MB03 and DAOY spheroid growth at two different concentrations ( Fig. 2D and E). These results demonstrate that LAT1 activity is crucial for

MB cell proliferation and survival.
Finally, we tested the effect of LAT1 on the migration capacities of MB cells using a scratch assay on a confluent layer of DAOY cells. JPH203 treatment resulted in a 25% reduction of scratch closure after 8 hours compared to cells treated with DMSO ( Fig. 3A and B). This suggests that LAT1 activity is required to sustain MB cell motility. Altogether, these findings demonstrate that LAT1 through its EAA transport activity promotes some of the key malignant features of MB cells while bearing seemingly no or low toxicity towards healthy cerebral cells.

Chronic treatment of MB cells with JPH203 induces cellular adaptation but no resistance
In order to test the development of resistance mechanisms, HD-MB03 and DAOY cell lines were incubated for more than 120 days in JPH203-supplemented medium (20 and 30 ⎧ M).
The levels of LAT1 and its chaperone CD98 were dramatically increased in response to chronic exposure to JPH203 (Fig. 4A). Consistently, the mRNA levels of LAT1 and LAT3 were also increased (Fig.S1). Nonetheless, JPH203 still displayed strong anti-proliferative and cytotoxic effects on both cell lines ( Fig.4B and C). These results strongly suggest that this 4-month adaptation was insufficient to induce resistance to LAT1 inhibition although the chronic treatment with JPH203 led to upregulation of some components of the amino acid transport machinery.

DISCUSSION
The most aggressive forms of MBs as well as local or metastatic relapses generally show a high degree of resistance to classical treatments. It is therefore crucial to discover new treatments presenting important properties such as selective toxicity towards cancer cells and low toxicity towards normal tissues. In the present study, we show that JPH203-induced blocking of LAT1-dependent EAA transport efficiently disrupts AA homeostasis, mTORC1 activity, proliferation and survival of two independent MB cell lines isolated from MBs of different subgroups (high risk Shh group and group 3) [32]. This suggests that JPH203 may represent an efficient treatment for MBs of different genetic backgrounds. The low toxicity of JPH203 towards primary murine astrocytes and cortical neurons suggests reduced toxic effects in a clinical setup. However, impairment of the transport of branched-chain amino acids at the blood-brain barrier caused by a LAT1 deficiency has also been described to cause autism spectrum disorders in mice [33]. Hence, JPH203 may lead to psycho-cognitive disorders in patients. Yet, according to the results from the first phase 1 clinical trial of JPH203 reported at the American Society of Clinical Oncology Gastrointestinal Cancers Symposium 2018 ("ASCO GI"), JPH203 is in fact well tolerated in adult patients [34].
Development of resistance by cancer cells has proven to be one of the most frequent causes of targeted therapy failure and may account for the low efficacy of therapies targeting the Shh pathway in MB [35,36]. mTORC1 pharmacological inhibition can overcome the acquired resistance to Shh targeted therapy in Shh group MBs [37]. Our study shows that inhibition of LAT1 leads to a strong decrease in mTORC1 activity, suggesting that the use of JPH203 may be relevant to bypass resistance to Shh targeted therapies. We also show that despite an adaptive response via an upregulation of EAA transporter expression, MB cells treated for several months with JPH203 never acquired the capacity to overcome the cytotoxic and cytostatic effects of this compound. This result suggests that MB cells may not have the ability to adapt to EAA import inhibition and subsequent mTORC1 inactivation. Even if encouraging, our results still need in vivo validation. In particular, the ability of JPH203 to pass the blood-brain barrier and its toxicity towards developing organisms still require thorough investigation.     The expression levels of CD98 and LAT1 in HD-MB03 and DAOY cells chronically exposed for 120 days to the indicated concentrations of JPH203 were analyzed by immunoblot.

ACKNOWLEDGMENTS
Tubulin was used as a loading control.