Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 May 2;114(18):E3602-E3611.
doi: 10.1073/pnas.1617066114. Epub 2017 Apr 17.

SNAT7 Is the Primary Lysosomal Glutamine Exporter Required for Extracellular Protein-Dependent Growth of Cancer Cells

Affiliations
Free PMC article

SNAT7 Is the Primary Lysosomal Glutamine Exporter Required for Extracellular Protein-Dependent Growth of Cancer Cells

Quentin Verdon et al. Proc Natl Acad Sci U S A. .
Free PMC article

Abstract

Lysosomes degrade cellular components sequestered by autophagy or extracellular material internalized by endocytosis and phagocytosis. The macromolecule building blocks released by lysosomal hydrolysis are then exported to the cytosol by lysosomal transporters, which remain undercharacterized. In this study, we designed an in situ assay of lysosomal amino acid export based on the transcription factor EB (TFEB), a master regulator of lysosomal biogenesis that detects lysosomal storage. This assay was used to screen candidate lysosomal transporters, leading to the identification of sodium-coupled neutral amino acid transporter 7 (SNAT7), encoded by the SLC38A7 gene, as a lysosomal transporter highly selective for glutamine and asparagine. Cell fractionation confirmed the lysosomal localization of SNAT7, and flux measurements confirmed its substrate selectivity and showed a strong activation by the lysosomal pH gradient. Interestingly, gene silencing or editing experiments revealed that SNAT7 is the primary permeation pathway for glutamine across the lysosomal membrane and it is required for growth of cancer cells in a low free-glutamine environment, when macropinocytosis and lysosomal degradation of extracellular proteins are used as an alternative source of amino acids. SNAT7 may, thus, represent a novel target for glutamine-related anticancer therapies.

Keywords: cancer; cell nutrition; glutamine; lysosome; transporter.

Conflict of interest statement

Conflict of interest statement: The authors have filed a European patent application on potential therapeutic uses of SNAT7-directed tools.

Figures

Fig. 1.
Fig. 1.
Live-cell assay reveals that SNAT7 selectively exports asparagine and glutamine from lysosomes. (A) Principle of the TFEB-based assay. (Left) In untreated HeLa cells, the transcription factor TFEB mainly localizes to the cytosol. (Center) Treating cells with an amino acid (AA) ester results in intralysosomal overload of the corresponding amino acid, which in turn induces TFEB translocation to the nucleus. (Right) Overexpression of a lysosomal transporter that exports the amino acid out of lysosomes reduces lysosomal stress and, consequently, TFEB nuclear translocation. (B) Proof-of-concept experiment. HeLa cells were transiently cotransfected with plasmids encoding mRFP-TFEB and one of the three well-characterized lysosomal transporters fused to EGFP, sialin (Si), LYAAT1 (LY), and PQLC2 (PQ), which export sialic acid, alanine, and lysine, respectively, from lysosomes. In the case of sialin, we used a nontransporting mutant (P334R) to serve as a negative control. Two days after transfection, cells were treated with 100 mM sucrose or with the indicated amino acid esters at an optimal concentration (Results and Materials and Methods) for 2 h before fixation. The mRFP-TFEB intracellular localization (cytosolic, nuclear, or mixed) was determined under a fluorescence microscope, as shown in the photograph, for at least 50 EGFP-positive cells per condition. (Scale bar: 10 μm.) The graph shows the mean outcome of the number of experiments indicated on the bars. Treating cells with sucrose or amino acid esters induces nuclear translocation of TFEB in ∼50% of cells. However, TFEB translocation was reduced to 25% and 10%, respectively, in LYAAT1- and PQLC2-expressing cells treated with an ester of a translocated amino acid (alanine and lysine, respectively). (C) HeLa cells transiently expressing mRFP-TFEB and either rPQLC2-EGFP (PQ) or EGFP-mSNAT7 (S7) were treated with sucrose or one of the indicated amino acid esters for 2 h before fixation, and the TFEB localization measurement was performed as in B. The graph displays the mean outcome of three independent experiments. SNAT7 overexpression selectively decreases the lysosomal stress induced by asparagine and glutamine overload. Error bars represent SD. P values are shown above the bars.
Fig. S1.
Fig. S1.
Dose dependence of the TFEB nuclear translocation induced by amino acid esters. HeLa cells transiently expressing mRFP-TFEB were treated during 2 h with 100 mM sucrose or different amino acid ester concentrations before fixation. TFEB localization was observed and quantitated as in Fig. 1.
Fig. 2.
Fig. 2.
SNAT7 is the major lysosomal transporter for glutamine and asparagine. (A) Genomic sequences of SNAT7 KO HeLa cell clones generated by the CRISPR/Cas9 method. The WT protein sequence of the targeted region is shown at the top. Protospacer adjacent motif (PAM) sequences are indicated in red, and target specific guide sequences are indicated in blue. The two KO clones display homozygous disruption of the SLC38A7 gene: a 26-bp deletion and a 19-bp deletion + 205-bp insertion (ins) for clones 2.22 and 1.24, respectively. The control clone (CT2) shows no modification relative to the HeLa cell line (CL). (B) Characterization of the anti-hSNAT7 antibody by Western blot. SNAT7 appears as a strong 40-kDa band (red arrow) detected in lysates from WT, but not KO, HeLa cells. (C) Scheme for subcellular fractionation of HeLa cells. Centrifugation is expressed as 1,000 × g × min. L, light mitochondrial fraction; M, heavy mitochondrial fraction; N, nuclear fraction; P, peroxisomal and microsomal fraction; S, soluble fraction; MLPS, LPS and PS, supernatants containing the corresponding fractions. (D) Fractions from differential and isopycnic centrifugations were analyzed for SNAT7 immunoreactivity and for diverse organellar markers by Western blot and enzymatic measurements, respectively. Equal protein amounts were loaded on each lane of the blots shown. β-Galactosidase was used as a lysosomal marker (other cell compartments are shown in Fig. S3). For the differential centrifugation, graphs show the enrichment factor (Relative Specific Signal) of the protein of interest against the percentage of total proteins recovered in each fraction. For isopycnic centrifugation, the frequency (Materials and Methods) of the protein of interest is plotted against sucrose density. (E) Principle of the countertransport assay of SNAT7. Lysosomes partially purified from HeLa cells are selectively loaded with glutamine using glutamine-tert-butyl (Gln-Bu) ester. After washing, fractions are incubated with [3H]Gln, which accumulates into lysosomes in exchange for intralysosomal glutamine. (F) SNAT7 mediates lysosomal countertransport of glutamine. Lysosome-enriched fractions, prepared from WT (CT2 clone, black squares) or SNAT7 KO (clone 1.24, red squares) HeLa cells, were preloaded with Gln, washed, and incubated with [3H]Gln for diverse durations. [3H]Gln uptake, determined by filtration and scintillation counting, increased linearly over time in a SNAT7-dependent manner. This uptake occurred in lysosomes as lysosomal lysis induced by 10 mM GME (arrow) released the radioactivity. A representative experiment (exp.; Left) and average normalized kinetics from four lysosome preparations (prep.; Right) are shown. P value is shown only for the earliest time point. (G) [3H]Gln countertransport was assayed in fractions from WT HeLa cells (black) or SNAT7 KO clones transiently expressing (green), or not (red), mEGFP-SNAT7. Uptake was measured in triplicates at 0 min (open bars) or 30 min before (solid bars) or after (hatched bars) lysosomal lysis. Overexpression of mSNAT7 partially rescued glutamine uptake in SNAT7 KO lysosomes. A representative experiment (Left, mean ± SD of triplicate measurements) and average normalized values from six independent experiments performed on distinct lysosome preparations (Right) are shown.
Fig. S2.
Fig. S2.
Lack of immunofluorescence detection of native SNAT7 with the HPA041777 antibody. WT or SNAT7 KO (CRISPR/Cas9, clone 2.22) HeLa cells were fixed with either paraformaldehyde or methanol and immunolabeled with the SNAT7 antibody diluted 1:100. No specific signal could be detected.
Fig. S3.
Fig. S3.
Supplementary analyses of the subcellular fractionation of HeLa cells. Enzyme markers were quantitated in the differential centrifugation and isopycnic centrifugation fractions and displayed for the indicated organelles as in Fig. 2D.
Fig. S4.
Fig. S4.
SNAT7 interacts with glutamine and asparagine in a similar manner. (A) [3H]Asn (Left) or [3H]Gln (Right) countertransport was measured over time in lysosome-enriched HeLa cell fractions preloaded, or not (red curve), with (w) Asn (green curve) or Gln (black curve) using ester precursors. Arrows indicate selective lysosomal lysis by 10 mM GME. The graphs show a representative experiment of three. (B) Dependence of Gln countertransport on SNAT7 does not reflect a higher vulnerability of SNAT7-defective lysosomes to osmotic lysis. Lysosome-enriched fractions from WT (black curve) or KO (red curve) HeLa cells were treated for 15 min with increasing GME concentrations. Lysosomal integrity was then measured using a β-galactosidase latency assay. The graph shows mean ± SD values from three independent experiments.
Fig. 3.
Fig. 3.
Functional properties of SNAT7. (A) Substrate selectivity on the cytosolic side. Lysosome-enriched fractions from WT and SNAT7 KO HeLa cells were preloaded with glutamine and assayed for countertransport of various [3H]AA for 30 min. Specific lysosomal uptake, determined based on GME-induced lysis and SNAT7 dependence (Materials and Methods), was calculated for each substrate and normalized to the [3H]Gln signal. (B) Substrate selectivity on the luminal side. [3H]Gln countertransport was assayed in lysosome-enriched fractions preloaded with various amino acids using ester precursors. (C) [3H]Gln countertransport was assayed as in A (ctrl) and either in the absence of ATP (Left) or in the presence of 500 nM bafilomycin A1 (baf; Right). (D) Lysosomal v-ATPase creates an electrochemical H+ gradient (acidic and positive inside) across the lysosomal membrane. Ionophores can selectively disrupt the chemical [pH gradient (ΔpH)] or electrical component [voltage gradient (ΔΨ)], or both, as depicted in the scheme. The [3H]Gln countertransport was assayed as in A in the presence of ATP and 3 μM nigericin (nig), valinomycin (val), or FCCP. Disruption of the pH gradient, but not the voltage gradient, abolished [3H]Gln uptake. The graph in A shows mean values from two independent experiments. Other graphs are representative experiments of two (B and D) or three (C) experiments performed on distinct lysosome preparations. Error bars represent SD from triplicate measurements.
Fig. S5.
Fig. S5.
SNAT7 dependence of lysosomal countertransport of asparagine and glutamine. Lysosome-enriched fractions from WT (black bars) and SNAT7 KO (red bars) HeLa cells were preloaded with Gln and assayed for countertransport with the indicated [3H] amino acid (AA) as in Fig. 2G. Mean ± SD uptake values are shown at 0 min (open bars) or at 30 min before (solid bars) or after (hatched bars) GME-induced lysosomal lysis for four representative experiments. Lysosome-specific (GME-dependent) uptake of [3H]Asn and [3H]Gln strictly depends on SNAT7. Tritiated Arg, Asp, Glu, His, Phe, and Pro accumulated at a low level in an SNAT7-independent manner, in contrast to the transport activity reported in a previous study (30). Note that [3H]Ala and [3H]Val accumulated to high levels, yet in a GME-insensitive, SNAT7-independent manner, suggesting that this uptake occurs into a nonlysosomal compartment.
Fig. 4.
Fig. 4.
SNAT7 transport activity is required for extracellular protein-dependent growth of cancer cells in a low free-glutamine environment. (A) BSA supplementation (2%) rescues MIA PaCa-2 cell growth in low free-glutamine medium. Growth was measured in triplicate at diverse Gln concentrations by cell counting and expressed as the mean ratio ± SD between final and initial cell numbers. One representative experiment of three is shown. (B) SNAT7 expression was silenced in MIA PaCa-2 cells using three distinct SLC38A7-targeted siRNAs (S7-1, S7-2, and S7-3) or two control siRNAs (CT-1 and CT-2) as negative controls. (Inset) SLC38A7 silencing strongly reduced SNAT7 expression (Western blot). Cells were then transiently transfected with mSNAT7 (S7) or, as a negative control, a nontransporting sialin construct (Si) and assayed for 2% BSA-dependent growth at 0.2 mM Gln (double arrow in A). SNAT7 silencing decreased, and mSNAT7 expression rescued, BSA-dependent growth. The graph shows mean value ± SD from three independent experiments. P values for the difference between control (white bars) and siRNA-treated (gray bar) samples and between siRNA-treated (gray bar) and mSNAT7-reexpressing (gray hatched bar) samples are displayed in red and black, respectively. (C) SNAT7 expression was abolished in MIA PaCa-2 cells using CRISPR/Cas9 gene editing. The resulting clones were assayed for BSA-dependent growth as in A. (Left and Center) Graphs show a representative experiment for a WT clone (CTB) and a KO clone (34I) with (red curve) or without (black) BSA at diverse Gln concentrations. (Right) Graph shows mean BSA-dependent growth value ± SD for three independent experiments at 0.2 mM Gln for the cell line, one control clone (CTB), and two independent KO clones (34B and 34I). (Inset) Loss of SNAT7 expression was verified by Western blot. (D, Left) Fluorescence micrographs of HeLa cells expressing N62H and N62Q mEGFP-SNAT7 (green) show a good overlap with the lysosomal marker LAMP1 (red). (D, Right) Magnifications (2.5×) of the selected areas are shown for each channel. (Scale bar: 20 μm.) (E) Rescue experiments of [3H]Gln countertransport in HeLa cells. Uptake was assayed in fractions from WT cells (black); SNAT7 KO cells (red, 1.24 clone); and KO cells transiently expressing WT, N62Q, or N62H mEGFP-SNAT7 (green) as in Fig. 2F. The graph shows mean ± SD values normalized to WT cells from three independent experiments. P values relative to WT-expressing KO cells are indicated above the bars. The expression of EGFP-mSNAT7 was checked by Western blot. (F) Rescue experiments of BSA-dependent growth in MIA PaCa-2 cells. WT (white bars) and KO (black) clones transiently expressing sialin (Si) or the indicated mEGFP-SNAT7 constructs were assayed for 2% BSA-dependent growth in 0.2 mM Gln over 3 d. WT and N62Q mSNAT7 rescue BSA-dependent growth, in contrast to the transport-defective N62H mutant. The graph shows mean ± SD values from three independent experiments. P values relative to WT-expressing KO cells are indicated above the bars. The expression of EGFP-mSNAT7 was checked by Western blot.
Fig. S6.
Fig. S6.
Additional characterization of SNAT7-silenced cancer cells. (A) Time schedule of SNAT7 silencing experiments in MIA PaCa-2 and A2780 cells. (B) Human ovarian carcinoma cell line A2780 also uses extracellular BSA as a glutamine source in a SNAT7-dependent manner. SNAT7 was silenced (gray bars), or not (white bars), in A2780 cells using distinct siRNAs. Cells were then transiently transfected with mSNAT7 (S7) or, as a negative control, a nontransporting sialin construct (Si) and assayed for 2% BSA-dependent growth at 0.2 mM free glutamine. The graph shows mean ± SD values from three independent experiments. BSA supplementation increased growth of nonsilenced cells by 63% (white bars). This effect decreased to 25–40% when the SLC38A7 gene was silenced (uniform gray bars; red P values relative to white bars). Overexpression of mSNAT7 in SNAT7-silenced cells apparently increased BSA-dependent growth (hatched gray bars), although in a nonsignificant manner (black P values relative to uniform gray bars). This marginal result may be due to a limited expression of mSNAT7 in this cell line. (C) SNAT7 silencing does not impair degradation of internalized BSA in MIA PaCa-2 cells. After silencing with SLC38A7-specific siRNAs (S7-1, S7-2, and S7-3) or with negative controls (CT-1 and CT-2), cells were incubated with 50 μg/mL DQ-red-BSA, a self-quenched fluorescent BSA conjugate that undergoes dequenching upon internalization and degradation in lysosomes. DQ-red-BSA fluorescence dequenching (mean ± SD, n = 10,000) was measured in cells fixed with 4% paraformaldehyde by flow cytometry [Canto II cytometer, Becton Dickinson; excitation at 633 nm and detection at 660 ± 20 nm on 10,000 cells selected in the R1 gate drawn on the forward scatter/side scatter biparametric histogram to consider only the population of cells devoid of aggregates and debris]. SNAT7 silencing did not alter DQ-red-BSA fluorescence. The graph shows a representative experiment of three.
Fig. S7.
Fig. S7.
Genomic sequences of the gene-edited MIA PaCa-2 cell clones. SLC38A7 was edited in MIA PaCa-2 cells with the nickase version of the CRISPR/Cas9 method using the specific guide sequences indicated in blue on the WT sequence. PAM sequences are shown in red. The amino acid sequence of the target region (first coding exon) is shown at the top. The sequence of the negative control clone (CTB) was identical to the WT cell line, whereas the 34B and 34I KO clones showed deleterious compound heterozygous mutations. The 34B clone presents several mutations (indel + missense) on both alleles inducing truncation and modification of the SNAT7 protein after residue Gly-44. The 34I clone has a frameshift mutation on one allele and a 9-aa deletion (Ile-59 to Gly-68) in the middle of first transmembrane domain.
Fig. S8.
Fig. S8.
Alignment of the SNAT7 and SNAT9 amino acid sequences. The hSNAT7 (accession no. Q9NVC3; UniProt) and SNAT9 (accession no. Q8NBW4; UniProt) sequences were aligned using Clustal Omega (www.ebi.ac.uk/Tools/msa/clustalo/). (A) Sequence alignment of the N-terminal region and the first two transmembrane domains (TMs). Identities and similarities are highlighted in black and gray, respectively. Red arrowheads label SNAT9 residues critical for its interaction with the Ragulator/RAG GTPase complex in the lysosomal membrane (16). The green asterisk labels a SNAT7 residue critical for the Gln transport function of SNAT7 (Fig. 4E; N62H mutant). (B) Amino acid identity and similarity are displayed for the full sequence; the N-terminal tail; and the remaining region, including 10 or 11 predicted TMs. The lower similarity of the N-terminal tail reflects the absence of a 60-aa extension in SNAT7.

Similar articles

See all similar articles

Cited by 5 articles

Publication types

LinkOut - more resources

Feedback