SLC6A14 mediates Na+/Cl−-coupled concentrative uptake of a broad-spectrum of amino acids. It is expressed at low levels in many tissues but up-regulated in certain cancers. Pharmacological blockade of SLC6A14 causes amino acid starvation in estrogen receptor positive (ER+) breast cancer cells and suppresses their proliferation in vitro and in vivo. In the present study, we interrogated the role of this transporter in breast cancer by deleting Slc6a14 in mice and monitoring the consequences of this deletion in models of spontaneous breast cancer (Polyoma middle T oncogene-transgenic mouse and mouse mammary tumour virus promoter-Neu-transgenic mouse). Slc6a14-knockout mice are viable, fertile and phenotypically normal. The plasma amino acids were similar in wild-type and knockout mice and there were no major compensatory changes in the expression of other amino acid transporter mRNAs. There was also no change in mammary gland development in the knockout mouse. However, when crossed with PyMT-Tg mice or MMTV/Neu (mouse mammary tumour virus promoter-Neu)-Tg mice, the development and progression of breast cancer were markedly decreased on Slc6a14−/− background. Analysis of transcriptomes in tumour tissues from wild-type mice and Slc6a14-null mice indicated no compensatory changes in the expression of any other amino acid transporter mRNA. However, the tumours from the null mice showed evidence of amino acid starvation, decreased mTOR signalling and decreased cell proliferation. These studies demonstrate that SLC6A14 is critical for the maintenance of amino acid nutrition and optimal mammalian target of rapamycin (mTOR) signalling in ER+ breast cancer and that the transporter is a potential target for development of a novel class of anti-cancer drugs targeting amino acid nutrition in tumour cells.
Cancer cells grow rapidly, a feature that places increased demand for nutrients. Uptake of nutrients into tumour cells is increased by up-regulation of specific transporters; these transporters have potential as drug targets for cancer therapy. Interference with the entry of nutrients into tumour cells would starve these cells to death. If the transporters that are specifically induced in tumour cells compared with normal cells are identified, blocking the function of the induced transporters would offer a logical and effective strategy for cancer therapy with little effect on normal cells.
Tumour cells need an increased supply of amino acids to support protein and nucleotide synthesis. Mammalian cells cannot synthesize essential amino acids and must obtain them via specific transporters. Glutamine, though a non-essential amino acid, is critical for tumour cell proliferation. Tumour cells are ‘glutamine-addicted’ [1,2] because glutamine is coupled to mechanistic target of rapamycin (mTOR) signalling, which integrates signals from growth factors, energy status and amino acid nutrition and co-ordinates these signals with cell growth, cell cycle progression and antioxidant machinery . There is a growing interest in amino acid transporters in cancer cells; recent studies have highlighted the importance of three amino acid transporters, namely SLC1A5, SLC7A5 and SLC7A11, as potential drug targets for cancer therapy [4–8].
We have identified a novel amino acid transporter, SLC6A14, as a drug target for cancer treatment; this transporter has functional features superior to those of SLC1A5, SLC7A5 and SLC7A11 to support tumour growth. It transports all essential amino acids and glutamine [9,10]. The transport via SLC6A14 is almost unidirectional based on its energetics mediating the concentrative uptake of its amino acid substrates into cells; in contrast, the other three transporters are obligatory exchangers, meaning that influx of an amino acid into cells via these three transporters is always coupled to efflux of some other amino acid out of the cells. Further, the uphill uptake of glutamine via SLC6A14 is coupled to SLC7A5 and SLC7A11 by concentrating glutamine inside the cells to serve as an exchangeable amino acid for SLC7A5 (glutamine) and SLC7A11 (glutamate) . SLC6A14 is induced in colon cancer , cervical cancer , pancreatic cancer  and estrogen receptor positive (ER+) breast cancer . SLC6A14 drives ‘glutamine addiction’ in these cancers and blockade of SLC6A14 function starves these tumour cells to death in vitro and in vivo .
To establish unequivocally the tumour-promoting role of SLC6A14, we generated a mouse line with the deletion of the gene coding for the transporter, characterized the phenotype of Slc6a14-null mice and examined the impact of the transporter deletion on breast tumour in two different mouse models of spontaneous breast cancer.
MATERIALS AND METHODS
The Slc6a14 gene is located on the X chromosome. We constructed a targeting vector by deleting a ~2.6 kb region containing exons 1 and 2 and ~1 kb of the sequence upstream of exon1 and used the vector to generate the knockout mouse. Deletion of the gene was not lethal and there was no overt phenotype either in hemizygous males (-/y) or in homozygous females (-/-).
PyMT-Tg (Polyoma middle T oncogene-transgenic) mice and MMTV/Neu-Tg (mouse mammary tumour virus promoter-Neu-transgenic) mice were from Jackson Laboratories. The mice were crossed with Slc6a14−/− mice to generate these two transgenic mouse lines on Slc6a14+/+and Slc6a14−/− backgrounds.
The Georgia Regents University Institutional Animal Care and Use Committee and Biosafety Committee approved the animal experiments reported in the present study.
Analysis of plasma amino acids and cytokines
Plasma amino acids were measured in three wild-type mice and three Slc6a14−/− mice using the Metabolic Lab Core Facility at Vanderbilt University. Plasma levels of various cytokines were measured in three wild-type mice and three Slc6a14−/− mice by ELISA.
Analysis of mammary tumours
PyMT-Tg female mice on Slc6a14+/+ and Slc6a14−/− backgrounds were monitored for breast tumour development. PyMT-Tg female mice typically develop tumours at the age of ~10 weeks. The size of the palpable tumours was measured once in 6 days and the volume was calculated using the formula (W2 × L)/2. At the age of 16 weeks, mice were killed and the tumours excised and weighed. The lungs were checked for metastasis using India ink . A similar approach was used for the analysis of mammary tumours in MMTV/Neu-Tg female mice on Slc6a14+/+ and Slc6a14−/− backgrounds.
RT-PCR and qPCR
The expression profiles of 13 different amino acid transporters in lung and colon tissues from wild-type and Slc6a14−/− mice and also in Slc6a14+/+ tumours and Slc6a14−/− tumours from age-matched PyMT-Tg mice were monitored by reverse transcriptase (RT)-PCR. For the analysis of mRNA levels for asparagine synthetase (ASNS), CHOP (CCAAT/enhancer-binding protein homologous protein), vascular endothelial growth factor (VEGF), glucose transporter 1 (Slc2a1) and monocarboxylate transporter 4 (Slc16a3), RT-PCR as well as qPCR were used.
RNA was extracted from Slc6a14+/+ and Slc6a14−/− tumours from PyMT-Tg mice. The labelled cDNA was prepared using this RNA according to the Ambion WT Expression Kit (Life Technologies) and GeneChip WT Terminal Labeling kit (Affymetrix) and then hybridized to an AffymetrixGeneGhip Mouse Gene 1.0 ST Array. The analysis was done with biological triplicates.
Mammary tumours from PyMT/Slc6a14+/+ and PyMT/Slc6a14−/− mice were preserved in 10% buffered formalin. Paraffin-embedded tissue sections were cut into 5-μm thickness, fixed and stained with haematoxylin and eosin (H&E). Deparaffinized tissue sections were incubated with primary antibodies against Ki67 (Abcam), phospho-mTOR (Cell Signaling) or HIF1α (hypoxia inducible factor 1α) (Novus Biologicals). Following biotinylated secondary antibody incubation and avidin–biotin complex treatment, immunopositive signals were visualized by diaminobenzidine staining. Sections were then counterstained in Mayer's haematoxylin and mounted under coverslips. Tissue sections were examined using an Olympus bright-field microscope. Frozen sections of lung and colon tissues from wild-type mice and Slc6a14−/− mice were also used for immunofluorescence analysis of Slc6a14 protein expression with an antibody raised against a mouse peptide sequence .
Whole mount analysis of mammary glands
Inguinal mammary glands from wild-type and Slc6a14−/− female mice (virgin) were spread on microscopic slides, fixed in Carnoy's fixative for 48 h, hydrated and stained with Carmine Alum stain overnight. The tissue was dehydrated, incubated in xylene for 48 h to remove fat and then mounted with permount for microscopic analysis.
Statistical analyses were conducted with the software SAS 9.3 (SAS Institute Inc.). A P<0.05 was considered statistically significant.
Phenotype of the knockout mice
Deletion of Slc6a14 in male (-/y) or female (-/-) mice, as documented by genotype analysis of the pups (Figure 1A), did not affect viability. Knockout males and females were fertile and there was no overt phenotype. Slc6a14 mRNA was not detected in colon and lung from both knockout males and knockout females whereas the expression was evident in these tissues in wild-type mice (Figure 1A). This corroborated with the absence of Slc6a14 protein in colon and lung from knockout mice as evident from immunofluorescence (Figure 1B). The mRNA was not detectable in normal wild-type mouse mammary gland (Figure 1A). In colon and lung of the knockout mice, there were no major compensatory changes in the expression of any of the 12 amino acid transporter mRNAs examined (Figure 1C). There was also no detectable change in the morphology and development of mammary gland in knockout female mice (Figure 1D).
Generation of Slc6a14-null mice
We monitored the plasma levels of amino acids in wild-type and knockout mice and found no significant differences between the two genotypes (Table 1). Since SLC6A14 is up-regulated in patients with ulcerative colitis indicating potential association between inflammation and the transporter , we measured the plasma levels of various cytokines, but again failed to detect major changes in most of the cytokines analysed (result not shown). Only four cytokines [interleukin (IL)-1α, IL-4, RANTES (regulated on activation, normal T cell expressed and secreted) and GMCSF (granulocyte macrophage colony-stimulating factor)] were affected to a significant extent; the levels of the first three cytokines decreased and those of the third increased in knockout mice.
Influence of Slc6a14 deletion on mammary tumour development in PyMT-Tg mice
PyMT-Tg mice develop breast cancer spontaneously at ~10 weeks of age . We examined the expression of Slc6a14 in breast tumours obtained from 4-month-old PyMT-Tg female mice and in mammary glands obtained from age-matched wild-type female mice. We found the transporter expression to be undetectable in wild-type mammary gland but up-regulated in PyMT-induced tumours (Figure 2A). We then assessed the consequences of Slc6a14 deletion on PyMT-driven breast cancer in these mice. At 3 months of age, all Slc6a14+/+ females with PyMT transgene developed tumours in multiple mammary glands which grew to sizes of 2000–2500 mm3 at 4 months of age (Figure 2B). In contrast, no Slc6a14−/− female with PyMT transgene developed tumours at 3 months of age; however, at 4 months of age, small tumour nodules became detectable (Figure 2B), indicating a marked delay in mammary tumour development in PyMT-Tg mice as a consequence of Slc6a14 deletion. In 4-month-old PyMT-Slc6a14+/+ mice, metastatic nodules were detectable in the lung but there was no metastasis in age-matched PyMT-Slc6a14−/− mice (Figure 2B). Furthermore, the growth of PyMT-driven breast tumours was markedly reduced on Slc6a14−/−background compared with Slc6a14+/+ background (Figure 2C).
Deletion of Slc6a14 markedly delays the development of mammary tumours in PyMT-Tg mice
Molecular differences in tumours between PyMT-Slc6a14+/+ and PyMT-Slc6a14−/− mice
Deficiency of amino acids, especially glutamine and leucine, interferes with cellular signalling via mTOR . mTOR lies upstream of HIF1α. Therefore, we expected that PyMT-driven mammary tumours in Slc6a14−/− mice would exhibit a decrease in cell proliferation, mTOR signalling and HIF1α. This was indeed the case. Tumours from PyMT-Slc6a14−/− mice showed markedly reduced Ki67, mTOR phosphorylation and HIF1α (Figure 3A). Amino acid deficiency in Slc6a14−/− tumours was evident from the increased expression of ASNS and CHOP (Figure 3B) and the deficiency of HIF1α was evident from the decreased expression of its target genes coding for the glucose transporter [GLUT1 (Slc2a1)], lactate transporter MCT4 (monocarboxylate transpoter 4) (Slc16a3) and VEGF (Figures 3C and 3D). These data show that amino acid deficiency and consequent suppression of mTOR and HIF1α signalling underlie the decreased tumour growth caused by Slc6a14 deletion. Phosphorylation of mTOR is increased in response to leucine , an excellent substrate for SLC6A14; as such, our findings that mTOR phosphorylation is markedly decreased in Slc6a14−/− tumours indicate that the tumours suffer from amino acid deprivation.
Molecular phenotype of Slc6a14−/− and Slc6a14+/+ mammary tumours from PyMT-Tg mice
Lack of compensatory changes in the expression of other amino acid transporters in PyMT-driven mammary tumours in Slc6a14−/− mice
We then examined whether the Slc6a14−/− tumours up-regulated the expression of other amino acid transporters to compensate for the lack of Slc6a14. With 12 different amino acid transporters examined by RT-PCR, none of the transporters differed in expression between PyMT-driven tumours with Slc6a14+/+ and Slc6a14−/− backgrounds (result not shown). We also performed microarray analysis of PyMT-Slc6a14+/+ and PyMT-Slc6a14−/− mammary tumours to evaluate the expression of amino acid transporters without any biased preconception of the identity of the transporters that might be subject to changes in expression in association with Slc6a14 deletion. The microarray data (Gene Expression Omnibus accession No. GSE56612) contained information on 391 solute transporters, a majority of them showing no significant change in expression between the two groups. There were only four transporters with 2-fold or higher up-regulation in Slc6a14−/− tumours; among these, only two were amino acid transporters: Slc1a3 and Slc7a2. Neither of the transporters is likely to activate mTOR because they do not transport leucine or glutamine.
We analysed the microarray data for other genes that are differentially regulated in Slc6a14−/− tumours compared with Slc6a14+/+ tumours. A surprising finding was a marked up-regulation of immunoglobulin genes in the knockout tumours. The up-regulation was noticed for various immunoglobulin subtypes including γ and μ heavy chains and κ light chain. It seems likely that deletion of Slc6a14 boosts tumour immunity, thus potentially implicating immune cell function in the reduced growth of tumours in the absence of the transporter.
We also evaluated the consequences of Slc6a14 deletion in breast cancer development and growth in another model of spontaneous mammary tumour (MMTV-Neu-Tg mouse). Tumours that arise in these mice were also associated with up-regulation of Slc6a14. We then monitored mammary tumour incidence and progression in this mouse line on Slc6a14+/+ and Slc6a14−/− genetic backgrounds. Deletion of the transporter resulted in a significant decrease in tumour incidence, tumour volume and the time needed for tumour appearance (Figure 2D).
The studies reported in the present study demonstrate that Slc6a14 plays an important role in promoting the development and growth of PyMT-driven and Neu-driven mammary tumours in mice. Deletion of the transporter in mice markedly suppresses the incidence and growth of these tumours. This effect is not associated with compensatory changes in the expression of any other amino acid transporter. Under normal conditions, deletion of the transporter does not lead to any noticeable phenotype, suggesting that pharmacological or biological blockade of the transporter for therapeutic purposes is not likely to elicit significant off-target effects in normal tissues. These studies highlight the potential of SLC6A14 as a target for development of a novel class of anti-cancer drugs.
Nutrient deprivation represents a logical strategy to kill cancer cells; yet none of the currently available anti-cancer drug development programmes has seized the opportunity to utilize this strategy. The present studies with Slc6a14 knockout mice firmly establish the therapeutic potential of this transporter in treatment of certain selective types of cancer. The therapeutic approach involving the blockade of SLC6A14 with either small molecules or biologicals would induce cell death in SLC6A14-positive tumour cells by at least three different mechanisms: (a) it would prevent the entry of essential amino acids; (b) it would target the ‘glutamine addiction’ behaviour; and (c) it would interfere with mTOR signalling. These studies support the idea that nutrient transporters can be exploited as potential drug targets for cancer therapy and usher us into a new era of novel cancer therapeutics beyond kinase inhibitors and growth factor receptor blockers.
mTOR complex 1 (mTORC1) is a known target for activation by amino acids. Three amino acids, leucine, glutamine and arginine, are potent activators of mTORC1. SLC6A14 transports all these three amino acids in a concentrative manner. Therefore, it makes sense that deletion of this transporter leads to decreased mTORC1 signalling in breast cancer. The contribution of decreased mTORC1 signalling to the blockade of tumour growth is in addition to the logically expected contribution of amino acid deprivation and consequent decreased protein and nucleotide synthesis to the phenomenon. Recent studies have shown that the pathways for mTORC1 activation by leucine and glutamine might be distinct [21,22]. There is also another mechanism by which cellular amino acid nutrition impacts on mTOR signalling; this involves the amino acids generated within the lysosomes by intracellular protein degradation and autophagosomal digestion [23,24]. This pathway implicates SLC38A9, an amino acid transporter expressed on the lysosomal membrane, in mTOR activation. However, no connection has yet been made between the lysosomal amino acid transporters and increased mTOR activity in tumour cells.
SLC6A14 is not a universal drug target for all types of cancers as is the case with all of currently available anti-cancer drugs. Some cancers do not up-regulate SLC6A14 and therefore would not respond to SLC6A14 blockers. But successful demonstration of SLC6A14 as an effective drug target for treatment of selective subtypes of cancer would provide new hope even for patients with other types of cancers. The SLC6A14-negative cancer subtypes must up-regulate some other glutamine transporters to support their growth. If these transporters are identified, a similar approach can be used to establish the therapeutic potential of such transporters in other cancer subtypes as well.
Ellappan Babu, Yangzom Bhutia and Sabarish Ramachandran performed most of the experiments; Jaya Gnanaprakasam did the ELISA for cytokines; Puttur Prasad and Muthusamy Thangaraju were responsible for the design of the targeting vector used in the generation of the knockout mouse and also for the analysis of plasma amino acids; Vadivel Ganapathy was responsible for the design of the study, interpretation of the data, and writing of the manuscript.
This work was supported in part by the Welch Endowed Chair in Biochemistry [grant number BI-0028] at Texas Tech University Health Sciences Center.