Lysosomal metabolomics reveals V-ATPase- and mTOR-dependent regulation of amino acid efflux from lysosomes

Abstract

The lysosome degrades and recycles macromolecules, signals to the cytosol and nucleus, and is implicated in many diseases. Here, we describe a method for the rapid isolation of mammalian lysosomes and use it to quantitatively profile lysosomal metabolites under various cell states. Under nutrient-replete conditions, many lysosomal amino acids are in rapid exchange with those in the cytosol. Loss of lysosomal acidification through inhibition of the vacuolar H⁺-adenosine triphosphatase (V-ATPase) increased the luminal concentrations of most metabolites but had no effect on those of the majority of essential amino acids. Instead, nutrient starvation regulates the lysosomal concentrations of these amino acids, an effect we traced to regulation of the mechanistic target of rapamycin (mTOR) pathway. Inhibition of mTOR strongly reduced the lysosomal efflux of most essential amino acids, converting the lysosome into a cellular depot for them. These results reveal the dynamic nature of lysosomal metabolites and that V-ATPase- and mTOR-dependent mechanisms exist for controlling lysosomal amino acid efflux.

Fig. 1. LysoIP method for rapid immunoisolation of intact lysosomes for absolute quantification of their metabolite content

A) Localization of Tmem192-3xHA fusion protein to lysosomes. Tmem192-3xHA and lysosomes were detected by immunofluorescence with antibodies to the HA epitope tag and the lysosomal marker LAMP2, respectively. Scale bars, 10 μm. Insets represent selected fields that were magnified 3.24X. B) Schematic of the workflow for the LysoIP method. Control-Lyso and HA-Lyso cells refer to cells stably expressing 2xFlag-tagged TMEM192 or 3xHA-tagged Tmem192, respectively. C) The LysoIP method isolates pure lysosomes. Immunoblotting for protein markers of various subcellular compartments in whole cell (whole-cell) lysates, purified lysosomes, or control immunoprecipitates. Lysates were prepared from cells expressing the 2xFlag-tagged TMEM192 (Control-Lyso cells) or 3xHA-tagged Tmem192 (HA-Lyso cells). ER, endoplasmic reticulum. D, E and F) Purified lysosomes are intact and retain their contents. (D) Cathepsin D activity was measured in whole-cell lysates and lysosomes, and immunoprecipitates from Control-Lyso cells served as a negative control (Control IP) (mean ± SEM, n=3). (E) Purified lysosomes take up radiolabeled arginine (Arginine [³H]). Lysosomes treated with a detergent were used as a control (mean ± SEM, n=3). (F) Calculations of the amounts of captured lysosomes (mean ± SEM, n=6, p > 0.05, N.S., not significant, ANOVA) were similar whether determined by tracking a membrane protein (LAMP2), the activity of the lysosomal protease cathepsin D (CatD), or a lysosome-specific small molecule (LysoTracker). Data are presented as the fraction of the material in the initial cell lysate. G) Absolute quantification of lysosomal metabolites. Comparison of concentrations of lysosomal metabolites across two biological replicates, with R-squared value shown. H) Metabolite concentrations in lysosomes and whole cells. Metabolites above the dotted blue line are enriched in lysosomes. Cys, cystine; Uri, uridine; Gua, guanosine; Ade, adenosine; Cyt, cytidine; Ino, inosine; GA, glucuronic acid. I) Whole-cell and lysosomal concentrations of 57 metabolites in HEK-293T cells (mean ± SEM, n=5). n indicates the number of independent biological replicates.

Fig. 2. The efflux from lysosomes of most non-essential, but not essential, amino acids requires the proton gradient

A) Changes in metabolite concentrations in whole-cells and lysosomes upon V-ATPase inhibition. Principal component analyses of changes in metabolite concentrations in whole-cells (circle) or lysosomes (square) after treatment for 1 hour with 200 nM Bafilomycin A1 (BafA1, blue) or Concanamycin A (ConA, purple). DMSO vehicle-treated cells were used as control (vehicle, green). B) Most metabolites accumulate in lysosomes upon V-ATPase inhibition while their levels in whole-cells are not affected. P-values are for comparisons between metabolite concentrations in whole-cell (triangle) or lysosome (circle) samples shown in (A) (n=3 for each treatment; dotted line represents p-value = 0.05). Lower panel, heat map of fold changes (log₂) in metabolite concentrations after V-ATPase inhibition relative to vehicle-treatment. Gray boxes indicate undetected metabolites. C and D) Accumulation of most non-essential, but not essential, amino acids in lysosomes upon V-ATPase inhibition. Fold changes in whole-cell and lysosomal concentrations of amino acids in BafA1- or ConA-treated cells relative to vehicle-treated cells (mean ± SEM, n=3, *p<0.05). E) Tracing of exogenously added alanine and isoleucine in live cells. Cells were incubated in medium containing ¹⁵N-labeled alanine and isoleucine for the indicated time points and then subjected to LysoIP. Data are presented as the fraction of the total pool of the amino acid that is ¹⁵N-labeled in whole-cells (black) or lysosomes (red) (mean ± SEM, n=3 in each time point). F) Dependence of lysosomal efflux of alanine but not that of isoleucine on the proton-gradient. Cells treated with or without ConA were incubated in medium containing ¹⁵N-labeled alanine and isoleucine for 1 hour (pulse period), which was then replaced with medium containing the natural ¹⁴N-containing isotope for the indicated time points (chase period). The fold change in the fraction of ¹⁵N-labeled amino acid remaining in the whole cells (circle) or lysosomes (square) was measured (mean ± SEM, n=3; k (in min⁻¹) is the rate constant for the decay of the ¹⁵N-labeled amino acid from the lysosome and * indicates non-overlapping 95% confidence intervals of the calculated k values between the treatments). Two-tailed t tests were used for comparisons between groups.

Fig. 3. mTOR regulates the lysosomal levels of essential non-polar amino acids in an autophagy-independent manner

A) mTORC1 regulates the abundance of amino acids in lysosomes upon amino acid starvation. A heat map shows fold changes (log₂) in amino acid (AA) concentrations in whole cell samples or lysosomes of wild-type and DEPDC KO cells after amino acid starvation for 60 minutes relative to cells cultured in medium with all amino acids (n=2 for each time point). B) Amino acid starvation inhibits mTORC1 signaling. Immunoblotting was used to monitor the levels and phosphorylation state of S6 kinase (S6K1) in the same samples as in (A). Raptor served as a loading control. C) Pharmacological inhibition of mTOR leads to the accumulation of many metabolites in lysosomes. Cells were treated with 250 nM Torin1 or DMSO (vehicle) for 1 hour and the lysosomal metabolite concentrations were determined and compared (n=3). Red lines indicate three-fold change in lysosomal concentration in Torin1- relative to vehicle-treated cells. D) Lysosomal accumulation of non-polar essential amino acids and tyrosine in an autophagy-independent manner after mTOR inhibition. Fold changes in the whole-cell and lysosomal concentrations of amino acids in wild-type and Atg7-null cells treated with Torin1 relative to vehicle-treated cells (mean ± SEM, n=3, *p<0.05; N.S, non significant). Histidine and serine served as examples of autophagy-dependent amino acids. E) Upon mTOR inhibition, nucleosides accumulate in lysosomes in a mostly autophagy-dependent manner. Fold changes in whole-cell and lysosomal concentrations of nucleosides in the same cells as in (D) (mean ± SEM, n=3, *p<0.05). F) Proteasome activity is dispensable for the lysosomal accumulation of amino acids upon mTOR inhibition. Fold changes in the whole-cell and lysosomal concentrations of amino acids in cells treated with 250 nM Torin1 or Torin1 together with 5 uM Bortezomib relative to DMSO- or Bortezomib-treated cells, respectively (mean ± SEM, n=3, *p<0.05; N.S, non significant). G) Proteogenic amino acids can be divided into those whose lysosomal levels are regulated by V-ATPase- or mTOR-dependent mechanisms. Two-tailed t tests were used for comparisons between groups.

Fig. 4. mTOR controls the efflux of non-polar essential amino acids and tyrosine from lysosomes

A) Tracing of exogenously added leucine, tyrosine, phenylalanine, and serine in live cells. Cells were incubated in medium containing the indicated ¹⁵N-labeled amino acids for various times and subjected to the LysoIP method. Data are presented as the fraction of the total pool of the amino acid in whole cells (black) or lysosomes (red) that is ¹⁵N-labeled (mean ± SEM, n=3 in each time point). B) Independently of protein synthesis, mTOR inhibition leads to the accumulation in lysosomes of leucine, tyrosine, phenylalanine, and isoleucine, but not serine. Data are presented as the fold change in the whole-cell and lysosomal abundance of the indicated ¹⁵N-labeled amino acid after Torin1-treatment relative to the DMSO vehicle treatment. The experiment was performed as indicated in the figure in the presence of 50 ug/mL cycloheximide (CHX) (mean ± SEM, n=3, p<0.005). This experiment was performed using Atg7-null cells. C) mTOR controls the lysosomal efflux of non-polar essential amino acids as well as tyrosine. Cells treated with 250 nM Torin1 or DMSO (vehicle) were incubated in medium containing the indicated ¹⁵N-labeled amino acids for 1 hour (pulse period). This medium was then replaced with media containing the natural ¹⁴N-isotope of the amino acid for the indicated time points (chase period). Data are presented as the fold change in the fraction of ¹⁵N-labeled amino acid remaining in whole cells (triangle) or lysosomes (circle) at each time point (mean ± SEM, n≥3; k (in min⁻¹) is the rate constant for the decay of the ¹⁵N-labeled amino acid from lysosomes and * indicates non-overlapping 95% confidence intervals of the calculated k between the treatments). This experiment was performed using Atg7-null cells. D) mTORC1 regulates the lysosomal abundance of essential non-polar amino acids and tyrosine in an SLC38A9-dependent manner. Fold changes in the lysosomal concentrations of indicated amino acids in wild-type and SLC38A9-null cells treated with Torin1 relative to vehicle-treated wild-type cells (mean ± SEM, n=3). E) Loss of SLC38A9 impairs the fitness of cells under starvation conditions. Wild-type, SLC38A9-null, and SLC38A9-null addback (SLC38A9-null+SLC38A9) cells were seeded in medium containing all amino acids (full media) or lacking leucine, isoleucine, tyrosine, and phenylalanine (-LIYF) for 3 days at which point cell numbers were measured (mean ± SD, n=3, *p<0.001). F) SLC38A9 is required for inhibiting the GCN2 pathway after a prolonged amino acid starvation period. Immunoblotting was used to monitor the levels and phosphorylation state of eIF2α. Raptor served as a loading control. LIYF indicates leucine, isoleucine, tyrosine, and phenylalanine. G) A model proposing a role for mTORC1 in regulating the efflux of amino acids from lysosomes. Two-tailed t tests were used for comparisons between two groups.