Loss of Sirtuin 1 Alters the Secretome of Breast Cancer Cells by Impairing Lysosomal Integrity

SUMMARY

The NAD⁺-dependent deacetylase Sirtuin 1 (SIRT1) is down-regulated in triplenegative breast cancer. To determine the mechanistic basis by which reduced SIRT1 expression influences processes related to certain aggressive cancers, we examined the consequences of depleting breast cancer cells of SIRT1. We discovered that reducing SIRT1 levels decreased the expression of one particular subunit of the vacuolar-type H+ ATPase (V-ATPase), which is responsible for proper lysosomal acidification and protein degradation. This impairment in lysosomal function caused a reduction in the number of multi-vesicular bodies (MVBs) targeted for lysosomal degradation and resulted in larger MVBs prior to their fusing with the plasma membrane to release their contents. Collectively, these findings help explain how reduced SIRT1 expression, by disrupting lysosomal function, and generating a secretome comprised of exosomes with unique cargo and soluble hydrolases that degrade the extracellular matrix, can promote processes that increase breast cancer cell survival and invasion.

Graphical Abstract

eTOC Blurb

Sirtuin 1 (SIRT1) expression is down-regulated in triple-negative breast cancer. Latifkar et al. show how reducing SIRT1 levels inhibits proper lysosomal function, and in doing so, results in the generation of a secretome with unique components, i.e. exosomes and resident lysosomal hydrolases, that promotes the aggressiveness of breast cancer cells.

INTRODUCTION

Sirtuins are NAD⁺-dependent deacylases that play important roles in a number of physiological processes and diseases (Chalkiadaki and Guarente, 2015). This family of enzymes consists of 7 members, many of which differ in their location and function (Jing and Lin, 2015). One of the most extensively studied members of the family is SIRT1, largely because its ectopic expression in yeast and mammals results in lifespan extension (Cohen et al., 2004; Lin et al., 2000). However, SIRT1 has been suggested to play multiple, and in some cases, contradictory roles in cancer (Chalkiadaki and Guarente, 2015). Some studies (Chung et al., 2015; Wu et al., 2012) suggest SIRT1 potentiates cancer phenotypes, while others indicate SIRT1 functions as a tumor suppressor, such as in highly aggressive breast cancers, where decreased SIRT1 expression is correlated with tumor expansion and metastatic spread (Simic et al., 2013; Wang et al., 2008a; Wang et al., 2008b). Given these findings, we were interested in probing how reduced SIRT1 expression enhances cellular phenotypes that underlie breast cancer progression. As described below, this led us to uncover a connection between SIRT1 and lysosomal function. Deregulation of this process results in the generation of a secretome with unique components, including exosomes and resident lysosomal hydrolases, that promote cell survival and invasive activity.

Exosomes are a type of non-classical secretory vesicle referred to as extracellular vesicles (EVs) (Desrochers et al., 2016a). They are attracting a good deal of attention because they contain various proteins, RNA transcripts, and microRNAs, and impact a wide range of diseases, including cancer. Exosomes can be distinguished from the other major type of EV, microvesicles (MVs), based on their size and biogenesis. MVs range from 0.2–2.0 μm in diameter and directly bud off from the plasma membrane, whereas, exosomes are ~30–150 nm in diameter and are contained within multi-vesicular bodies (MVBs). The fusion of MVBs with the plasma membrane, results in the release of their exosome content into the extracellular space.

Both types of EVs generated by cancer cells can engage and transfer cargo to neighboring cancer cells, stimulating their growth and survival. However, EVs from cancer cells can also affect normal cells, conferring upon them several characteristics of cancer cells, including the ability to exhibit anchorage-independent growth (Antonyak et al., 2011; Li et al., 2012a). EVs derived from highly aggressive cancer cells also promote chemotherapy resistance (Kreger et al., 2016; Qu et al., 2016), tumor angiogenesis (Feng et al., 2017), and metabolic reprogramming (Zhao et al., 2016). Exosomes, in particular, have been implicated in the formation of the pre-metastatic niche and enhancing organ-specific metastasis (Costa-Silva et al., 2015; Hoshino et al., 2015).

It has been suggested that lysosomal function can impact exosome biogenesis by altering the fate of MVBs (Miao et al., 2015; Alvarez-Erviti et al., 2011). However, how this happens is unclear. Here, we describe a mechanism by which reductions in SIRT1 expression in breast cancer cells alter lysosomal activity, resulting in increased numbers of exosomes shed from the cells and significant changes in the composition of their cargo. Specifically, we show that SIRT1 knock down, or pharmacological inhibition of this enzyme, destabilizes the mRNA encoding the A subunit of the lysosomal V-ATPase proton pump (ATP6V1A), causing a reduction in its expression. This decrease in ATP6V1A levels impairs lysosomal degradative activity and causes the enlargement of MVBs, which then fuse with the plasma membrane and release exosomes that contain distinct cargo and strongly promote cell survival and migration. We further demonstrate that, upon reduction of SIRT1 expression, there is a marked increase in the secretion of soluble lysosomal luminal proteins, i.e. Cathepsins, which degrade the extracellular matrix, allowing tumor cells to invade surrounding tissues (Gocheva and Joyce, 2007; Mitrović et al., 2017). Taken together, these findings show how SIRT1 plays an important role in a fundamental aspect of cell biology by ensuring proper lysosomal function, and in doing so, influences the secretome of cells. Moreover, they provide an explanation for how reducing SIRT1 expression contributes to the aggressiveness of breast cancer cells.

RESULTS

Decreasing SIRT1 expression levels promotes exosome release

SIRT1 has been suggested to be a tumor suppressor in breast cancer, as its expression is down-regulated in aggressive forms of the disease (Shi et al., 2018; Simic et al., 2013; Wang et al., 2008a). When we used Cancer RNA-Nexus (GSE58135, Varley et al., 2014) to examine triple-negative breast cancers (TNBCs), compared to normal tissues found adjacent to the TNBC tumors (NTNBCs), more than 80% (34/42) of the tumor samples showed a marked reduction in SIRT1 transcript levels (Figure 1A). The protein expression levels of SIRT1 also tended to be lower in TNBC cell lines, compared to non-TNBC cell types (Figure S1A). Similar reductions in the transcript levels of the highly related SIRT6 and SIRT7 proteins were not observed (Figures S1B and S1C).

Figure 1. Decreasing SIRT1 expression levels promotes exosome release.

(A) SIRT1 transcript levels were determined in Triple-Negative Breast Cancer (TNBC) tumors, and normal tissues adjacent to the TNBC tumors (NTNBC) using RNA-Seq Nexus (GEO accession: GSE58135). (B) Western blot analysis of SIRT1 and β-Actin levels in whole cell lysates (WCL) of sham shRNA expressing control (CTRL), and SIRT1 knock down (KD), MDA-MB-231 cells. (C) Fluorescence microscopy images of the cells in (B) immunostained for CD63 (green), and stained with DAPI (blue). The periphery of each cell is outlined (dashed lines), and insets are higher magnifications of boxed areas. Scale bar, 4 μm. (D) Quantification of MVB diameter for each condition in (C). (E) Western blot analysis of nuclear sirtuins, and heat shock protein 90 (HSP90), levels in WCL of sham shRNA expressing control (CTRL), SIRT1 KD, SIRT6 KD, and SIRT7 KD MDA-MB-231 cells. The expression level of each sirtuin was quantified relative to HSP90 and included in the blot. (F) Nano-particle tracking analysis (NTA) was performed on the conditioned media collected from an equivalent number of serum starved sham shRNA expressing control (CTRL), SIRT1 KD, SIRT6 KD, and SIRT7 KD cells. (G) Quantification of exosomes generated for each of the conditions in (F). (H) Western blot analysis of IκBα, CD63, and HSP90 levels in WCL, microvesicles (MV), exosomes (EXO), and vesicle free medium (VFM) prepared from sham shRNA expressing control (CTRL), and SIRT1 KD, MDA-MB-231 cells. The data shown in (A), (D) and (G) represent means ± SD; ****p<0.0001, **p<0.01, and not-significant (ns). See also Figure S1.

We obtained a clue regarding how reduced SIRT1 expression affects the behavior of breast cancer cells, when examining the intra-cellular features of SIRT1 knock down MDA-MB-231 cells, versus control MDA-MB-231 cells (Figure 1B). Specifically, we found differences in the MVBs between the two cell types, when immunofluorescence microscopy was performed using an antibody against the MVB marker protein CD63. While a comparable number of MVBs were detected in each sample, many MVBs in cells with reduced SIRT1 expression were noticeably larger (Figures 1C and 1D).

We further showed that shRNA-mediated knock down of SIRT1 in normal human astrocytes, and primary human dermal fibroblasts (Figure S1D), caused similar increases in MVB size (Figures S1E and S1F). We also deleted SIRT1 from MDA-MB-231 cells using CRISPR/Cas9. Clones that lost a single copy (SIRT1 +/−), or both copies (SIRT1 −/−), of the SIRT1 gene were shown to have enlarged MVBs, compared to wildtype (SIRT1 +/+) cells (Figures S1G-S1I).

These observations raised the question of whether the changes in MVB size observed in SIRT1 knock down cells reflected differences in their ability to generate exosomes. Therefore, we examined whether SIRT1, as well as the related SIRT6 or SIRT7 proteins, could impact exosome formation in MDA-MB-231 cells. shRNAs that specifically target each of these enzymes, or a control sham shRNA, were introduced into MDA-MB-231 cells, which are capable of generating EVs (Antonyak et al., 2011; Kreger et al., 2016; Li et al., 2012a). The expression levels of SIRT1, SIRT6, and SIRT7 were each knocked down by at least 85%, compared to control cells (Figure 1E). The amount of EVs in the conditioned media from an equivalent number of control cells, or cells depleted of SIRT1, SIRT6, or SIRT7, was determined using nano-particle tracking analysis (NTA). Knocking down SIRT6 and SIRT7 caused only modest changes in the amount of EVs produced by the cells, whereas, knock downs of SIRT1 resulted in a significant increase in EVs, with a greater than 3-fold increase in the number of exosome-sized vesicles (i.e. 40-150 nm vesicles) (Figures 1F and 1G). Depleting SIRT1 in U87 glioma cells and human kidney (HK)-2 cells led to similar increases in exosome production (Figures S1J and S1K).

The exosomes and MVs released into the media by control, or SIRT1 knock down, MDA-MB-231 cells were then isolated using a combination of filtration and ultracentrifugation steps (Figure S1L). The isolation of each type of EV, as well as the media depleted of EVs (i.e. the vesicle free medium; VFM), was verified by immunoblotting for specific EV cargo and selected soluble proteins (Figure 1H). HSP90, which is expressed in both classes of EVs, was detected in each fraction (bottom panel) (Li et al., 2012), while the exosome marker CD63, was only found in the exosome fractions (middle panel). IKBα was only detected in the cell lysates (top panel), demonstrating that the EV and media fractions do not contain cytosolic contaminants. Electron Microscopy carried out on the exosomes isolated from control and SIRT1 depleted MDA-MB-231 cells (Figure S1M) showed that these vesicles were similar in size and morphology.

Exosomes from SIRT1 knock down cells contain distinct protein cargo

Exosomes generated by cells depleted of SIRT1 contain a protein composition distinct from exosomes produced by control cells, as evident when performing SDS-PAGE and Coomassie-Blue staining of their protein cargo. Whereas the whole cell lysates (WCL) showed no obvious differences in protein expression between control MDA-MB-231 cells and cells lacking SIRT1 (Figure S2A, lanes labeled WCL), the amounts of some proteins were noticeably reduced (blue arrows), while others were significantly increased (red arrow), in exosomes derived from SIRT1 knock down cells. This was further indicated when stable isotope labelling with amino acids in cell culture (SILAC) was performed on exosomes from control and SIRT1 knock down MDA-MB-231 cells (Figure S2B). Figure S2C lists proteins whose levels were either the most enriched (left table) or reduced (right table) in exosomes derived from SIRT1 knock down cells, compared to exosomes from control cells. Immunoblotting experiments confirmed the SILAC results; two examples are presented in Figure S2D, where the increased levels of 14-3-3 zeta/delta, and the decreased levels of CD81, as indicated by SILAC (Figure S2C proteins in red), were also observed in immunoblots. DAVID GO analysis performed on the proteins enriched in exosomes from SIRT1 depleted cells showed they are involved in diverse cellular processes and come from different cellular localizations (Figure S2E).

MVBs and their contents are either directed to the lysosome and degraded or reach the cell surface where they fuse with the plasma membrane. A critical step in MVB formation, where protein sorting occurs, involves the maturation of intraluminal vesicles (Villarroya-Beltri et al., 2014). Endosomal sorting complexes required for transport (ESCRTs) are responsible for recognizing and importing ubiquitinated proteins into a subset of intraluminal vesicles that are eventually degraded when MVBs fuse with lysosomes. Thus, we examined whether knocking down SIRT1 in MDA-MB-231 cells altered the amount of ubiquitinated cargo in their exosomes. While no significant differences in the levels of ubiquitinated proteins were detected in whole cell lysates upon knock down of SIRT1 (Figure 2A), there was a clear increase in the amounts of ubiquitinated cargo in exosomes isolated from cells depleted of SIRT1 (Figures 2B and 2C). Treatment of MDA-MB-231 cells with the SIRT1 inhibitor, EX-527, or deleting SIRT1 from cells by CRISR/Cas9, also resulted in the generation of exosomes enriched with ubiquitinated proteins (Figures 2D and 2E), while knocking down SIRT6 or SIRT7 expression did not have the same effect (Figures 2B and 2C). Histone 2A (H2A) and 2B (H2B), which are known to be ubiquitinated, were identified by SILAC to be enriched in exosomes derived from MDA-MB-231 cells depleted of SIRT1 (Figure S2C) and shown to exit the cells as ubiquitinated species in these vesicles (Figure 2F, top and middle panels). Exosomes isolated from SIRT1 knock down cells were also highly enriched in Survivin (Figure 2F, bottom panel), which is specifically expressed in aggressive cancer cells and degraded in a ubiquitin-dependent manner (Li et al., 2014).

Figure 2. Exosomes from SIRT1 knockdown cells contain distinct protein cargo.

(A) Western blot analysis of ubiquitinated protein and β-Actin levels in WCL of sham shRNA expressing control (CTRL), and SIRT1 KD, cells. The amount of ubiquitinated proteins detected in each lysate was quantified relative to β-Actin and included in the blot. (B) Western blot analysis of ubiquitinated protein and HSP90 levels in exosomes (EXO) from sham shRNA expressing control (CTRL), SIRT1 KD, SIRT6 KD, and SIRT7 KD cells. The amount of ubiquitinated proteins detected in each lysate was quantified relative to HSP90 and included in the blot. (C) Quantification of ubiquitinated protein levels for each condition in (B). (D) Western blot analysis of ubiquitinated protein levels in exosomes (EXO) from cells treated with DMSO, or EX-527 (50 μM), for 16 hours. (E) Western blot analysis of ubiquitinated protein and Flotillin-2 levels in exosomes (EXO) from wildtype MDA-MB-231 cells (SIRT1 +/+), or cells in which both copies of the SIRT1 gene were genetically deleted (SIRT1 −/−). The amount of ubiquitinated protein detected in each lysate was quantified relative to Flotiliin-2 and included in the blot. (F) Western blot analysis of ubiquitinated Histone 2A (UB-H2A), ubiquitinated Histone H2B (UB-H2B), and Survivin levels in exosomes (EXO) from sham shRNA expressing control (CTRL), and SIRT1 KD, cells. (G) Western blot analysis of RAB27A levels in endolysosomal fractions immunoprecipitated from sham shRNA expressing control (CTRL), and SIRT1 KD, cells ectopically expressing Flag-tagged TMEM192. (H) Western blot analysis of RAB7, SIRT1, and β-Actin levels in the WCL of sham shRNA expressing control (CTRL), and RAB7 KD MDA-MB-231, cells. (I) Fluorescence microscopy images of the cells in (H), immunostained for CD63 (green), and stained with DAPI (blue). Insets are higher magnifications of boxed areas. Scale bar, 4 μm. (J) Quantification of MVB diameter for each condition in (I). (K) Western blot analysis of ubiquitinated protein (left) and Survivin (right) levels in exosomes (EXO) from control (CTRL), and RAB7 KD, cells. (L) DAVID GO-Cellular Component analysis of proteins enriched in vesicle free medium (VFM) collected from SIRT1 KD cells, compared to control cells. (M) Western blot analysis of Cathepsin B and HSP90 levels in microvesicles (MV), exosomes (EXO), and vesicle free media (VFM) fractions from sham shRNA expressing control (CTRL), and SIRT1 KD, cells. The unprocessed and processed forms of Cathepsin B are indicated. (N) Levels of Cathepsin B activity in the conditioned medium (CM) from control (CTRL), and SIRT1 KD, MDA-MB-231 cells treated without or with CA-074 (10 μM) for 30 minutes. The data shown in (C) and (J) represent means ± SD; ****p<0.0001, **p<0.01, and ns. See also Figure S2.

Since reducing SIRT1 levels in cells increases the amount of ubiquitinated cargo present in their exosomes, we next examined whether the ectopic expression of SIRT1 in cells could reverse this effect. The triple-negative MDA-MB-453 breast cancer cell line was used for this experiment, because it expresses low levels of SIRT1 and generates exosomes containing considerable amounts of ubiquitinated cargo (Figure S2F). When SIRT1 was ectopically expressed in MDA-MB-453 cells, the levels of ubiquitinated proteins present in exosomes isolated from these cells was reduced (Figure S2F).

To further examine how decreased SIRT1 levels result in more MVB content (i.e. exosomes) being released into the extracellular environment, we used an approach that isolates intact lysosomes and MVBs from cells (Abu-Remaileh et al., 2017). The endolysosomal fractions immunoprecipitated using a FLAG antibody from control, and SIRT1 knock down, MDA-MB-231 cells expressing a FLAG-tagged form of the MVB/lysosomal resident protein TMEM192 (FLAG-TMEM192), were subjected to Western blot analysis using antibodies against RAB27A and the MVB marker Mannose-6-phosphate Receptor (M6PR). Under conditions where equivalent amounts of M6PR-positive vesicles were immunoprecipitated (Figure 2G, bottom panel), a greater amount of RAB27A associated with MVBs immunoprecipitated from SIRT1 depleted cells (Figure 2G, top panel). Consistent with the role of RAB27A in promoting late endosomal trafficking to the plasma membrane (Ostrowski et al., 2010), knocking down RAB27A in cells lacking SIRT1 led to the accumulation of large MVBs (Figures S2G-S2I). These findings support the idea that RAB27A is recruited to MVBs in cells depleted of SIRT1 to help mediate their transport and fusion with the plasma membrane.

Next, we considered the possibility that reduced SIRT1 levels could interfere with the ability of MVBs to be degraded in lysosomes. If so, knocking down Rab7 (Figure 2H), which mediates MVB-lysosome fusion (Vanlandingham and Ceresa, 2009) and causes a significant enlargement in the size of MVBs (Figures 2I and 2J), should have a similar effect on exosome production. Immunoblotting the exosomes generated by cells depleted of Rab7 showed that they were highly enriched in ubiquitinated cargo and Survivin (Figure 2K, left and right panels), similar to when SIRT1 expression was knocked down. These findings suggest that reduced SIRT1 levels cause MVBs that normally would be degraded in lysosomes to instead become enlarged and fuse with the cell surface to release their contents.

We then examined whether SIRT1 knock down cells also secrete soluble proteins that reside in MVBs. The conditioned media depleted of exosomes and MVs (i.e. the vesicle-free-media; VFM) from control and SIRT1 knock down MDA-MB-231 cells were resolved by SDS-PAGE and stained with Coomassie-Blue. The resulting gel (Figure S2J) showed that the VFM collected from SIRT1 depleted cells contained both increased (red arrows) and decreased (blue arrow) amounts of various proteins. SILAC analysis performed on these same preparations identified several lysosomal luminal proteins, namely members of the cathepsin hydrolase family, that were enriched in the VFM collected from cells depleted of SIRT1 (Figure 2L), with Cathepsin B showing the greatest increase (H/L~30). This result was confirmed by immunoblotting, which showed that smaller processed (i.e. active), as well as larger unprocessed (i.e. less active), forms of this enzyme were detectable in the VFM isolated from SIRT1 knock down cells (Figure 2M, lanes labeled VFM). A smaller amount of unprocessed Cathepsin B was found in the VFM from control cells, while it was absent in MVs or exosomes isolated from either control or SIRT1 knock down cells (Figure 2M). Increases in Cathepsin B levels were also detected in the VFM collected from cells treated with the SIRT1 inhibitor EX-527 (Figure S2K), and conditioned media from SIRT1 depleted cells exhibited significantly higher Cathepsin B activity, compared to conditioned media from control cells (Figure 2N). This increase in activity was completely blocked upon treatment of the medium derived from SIRT1 knock down cells with CA-074, a Cathepsin B-specific inhibitor.

Experiments examining whether Cathepsin B secretion was increased to the same extent in cells depleted of SIRT6 and SIRT7 showed that this was not the case (Figure S2L). Similarly, non-lysosomal metalloproteases MMP7 and MMP9 were not enriched in the VFM collected from SIRT1 depleted cells (Figure S2M).

SIRT1 depletion causes lysosomal impairment

Given the effects of reduced SIRT1 expression levels on MVB maturation and the contents of the secretome, we then examined whether SIRT1 influences lysosomal function. Immunofluorescent microscopy performed on control MDA-MB-231 cells showed LAMP1-positive lysosomes located near the nucleus of each cell (Figure 3A, top panel). However, lysosomes in SIRT1 knock down cells contained significantly larger LAMP1-positive structures (Figure 3A, bottom panel, and Figure 3B), suggesting lysosomal impairment (Bret Evers et al., 2017; Kilpatrick et al., 2015). Treatment of cells with either Chloroquine or Bafilomycin-A to inhibit lysosomal activity yielded similar effects (Figures S3A and 3C). Treating MDA-MB-231 cells with either of these inhibitors also resulted in cells with enlarged CD63-positive MVBs (Figures S3B and 3D), and gave rise to exosomes with increased amounts of ubiquitinated proteins (Figure 3E) and Survivin (Figure S4A). Moreover, treatment of MDA-MB-231 cells with Bafilomycin-A not only led to increased exosome release (Figure 3F), but also to an increase in the levels of Cathepsin B in the medium (i.e. the VFM) (Figure S4B).

Figure 3. SIRT1 depletion resembles lysosomal impairment.

(A) Fluorescence microscopy images of sham shRNA expressing control (CTRL), and SIRT1 KD, cells immunostained for LAMP1 (red), and stained with DAPI (blue). Insets are higher magnifications of boxed areas. Scale bar, 4 μm. (B) Quantification of lysosome diameter for each condition in (A). (C) Quantification of lysosome diameter, and (D) Quantification of MVB diameter, for cells treated with DMSO, Chloroquine (CQ, 50 μM), or Bafilomycin-A (Baf-A, 200 nM) as described in Figures S3A and S3B. (E) Western blot analysis of ubiquitinated protein and Flotillin-2 levels in exosomes (EXO) from cells treated with DMSO, Bafilomycin-A (Baf-A, 200 nM), or Chloroquine (CQ, 50 μM), for 16 hours. The amount of ubiquitinated protein detected in each lysate was quantified relative to Flotillin-2 and included in the blot. (F) Quantification of exosomes generated by cells treated with DMSO, or Bafilomycin-A (Baf-A, 200 nM), for 16 hours. (G) Fluorescence microscopy images of sham shRNA expressing control (CTRL), and SIRT1 KD, cells immunostained for Cathepsin B (red) and LAMP1 (green). The cells were also stained with DAPI (blue). Scale bar, 8 μm. The images of the SIRT1 KD cells are a composite of two separate pictures. Insets are higher magnifications of boxed areas. The data shown in (B), (C), (D) and (F) represent means ± SD; ****p<0.0001. See also Figures S3 and S4.

Additional evidence that lysosomes in cells with limiting amounts of SIRT1 are not functioning properly came from an experiment where Cathepsin B maturation was examined. Cathepsins undergo a maturation process, along the endocytic pathway, that involves multiple glycosylation and cleavage events (Olson and Joyce, 2015). Thus, we immunoblotted whole cell lysates (WCL) from control MDA-MB-231 cells, and MDA-MB-231 cells depleted of SIRT1, for Cathepsin B. While Cathepsin B was detected primarily as a mature, fully processed enzyme, in control cells, their SIRT1 knock down counterparts had little detectable fully processed enzyme (Figure S4C). Instead they predominantly expressed a partially processed form of Cathepsin B, while parental MDA-MB-231 cells treated with increasing amounts of Bafilomycin-A showed a complete loss of the processed forms of Cathepsin B (Figure S4D). Immunofluorescence microscopy carried out on SIRT1 knock down cells using a Cathepsin B antibody showed a marked difference in the localization of this hydrolase. Specifically, Cathepsin B predominantly localized with LAMP1-positive lysosomes in control cells (Figure 3G, top panel), as expected. However, in cells depleted of SIRT1, there was less localization of Cathepsin B with LAMP1-positive lysosomes (Figure 3G, bottom panel); instead, it predominantly localized with CD63-postitive MVBs (Figure S4E). Such changes were previously shown to be associated with its secretion from cancer cells, where its hydrolase activity played an important role in degrading the extracellular matrix and promoting tumor invasiveness (Gocheva and Joyce, 2007; Mitrović et al., 2017; Olson and Joyce, 2015).

SIRT1 loss inhibits lysosomal acidification

We next determined whether SIRT1 directly impacted lysosomal function by assaying the pH of lysosomes in control, and SIRT1 knock down, MDA-MB-231 cells using LysoSensor Yellow-Blue Dextran® ratiometric dye. Based on the calibration curve shown in Figure S5A, we found that the pH of lysosomes in control cells was 4.7 (Figure 4A), matching the reported pH of properly functioning lysosomes (Diwu et al., 1999), while the pH of lysosomes in cells depleted of SIRT1 was slightly above 5.6 (Figure 4A).

Figure 4. SIRT1 loss disrupts lysosomal acidification.

(A) Lysosomal pH measurements were determined for sham shRNA expressing control (CTRL), and SIRT1 KD, MDA-MB-231 cells. (B) Scheme of the lysosome re-acidification assay (top), and the percent of lysosomal re-acidification determined using LysoTracker Green DND-26® in sham shRNA expressing control (CTRL), and SIRT1 KD, cells. (C) SILAC results showing the ratio of ATP6V1A protein levels in SIRT1 KD cells, compared to control cells (top). Western blot analysis of ATP6V1A, ATP6V0D1, SIRT1, and HDAC6 protein levels in control (CTRL) and SIRT1 KD cells (bottom). The expression levels of ATP6V1A in each lysate was quantified relative to HDAC6 and included in the blot. (D) Western blot analysis of ATP6V1A, SIRT1, and HSP90 levels in control (CTRL), and ATP6V1A KD, cells. The expression levels of ATP6V1A was quantified relative to HSP90 and included in the blot. (E) NTA was performed on the conditioned media from an equal number of serum starved sham shRNA expressing control (CTRL), and ATP6V1A KD, cells. (F) Quantification of exosomes generated for each condition in (E). (G) Western blot analysis of ubiquitinated protein, Survivin, and Flotillin-2 levels in exosomes (EXO) generated by the cells in (D). The amount of ubiquitinated proteins detected in each lysate was quantified relative to Flotillin-2 and included in the blot. (H) Western blot analysis of SIRT1, ATP6V1A, and β-Actin levels in sham shRNA expressing control (CTRL), and SIRT1 KD, cells ectopically expressing either the vector alone (Vector) or ATP6V1A (ATP6V1A OE). (I) Quantification of exosomes generated for each condition in (H). (J) Western blot analysis of ubiquitinated protein, Cathepsin B, and HSP90 levels in exosomes (EXO; left panel) and vesicle free media (VFM; right panel) collected from these cells. The amounts of ubiquitinated protein and Cathepsin B detected in each lysate were quantified relative to HSP90 and included in the blots. The unprocessed and processed forms of Cathepsin B are also indicated. The data shown in (A), (B), (F) and (I) represent means ± SD; ***p<0.001, **p<0.01, and ns. See also Figure S5.

We then examined whether the increase in lysosomal pH associated with SIRT1 knock down cells is due to a defective proton pump. Lysosomal re-acidification assays were performed (Figure 4B, diagram), where the pH of lysosomes in control and SIRT1 knock down cells was first increased by Bafilomycin-A treatment, followed by the addition of LysoTracker to determine how quickly the pH of the lysosomes recovered. The pH of lysosomes in control cells fully recovered within 60 minutes (Figure 4B, graph). In contrast, the pH recovery of lysosomes in cells lacking SIRT1 was significantly slower.

Based on these findings, we turned our attention to the proton pump. V-ATPases are multi-subunit enzymes that are responsible for the acidification of late endosomes and lysosomes (Settembre et al., 2013), and alterations in the expression of any component of this pump can disrupt the ability of lysosomes to maintain proper pH (Lee et al., 2010; Meo-Evoli et al., 2015). Analysis of our SILAC results performed on SIRT1 knock down cells revealed that the expression of subunit A of V1, (ATP6V1A), is down-regulated by 65%, compared to control cells. This was confirmed by Western blot analysis (Figure 4C), whereas the expression of the ATP6V0D1 subunit, was unchanged by SIRT1 knock down (Figure 4C). Knocking down SIRT1 in two other breast cancer cell lines, i.e. Hs-578T and MDA-MB-453 cells, similarly affected ATP6V1A protein levels (Figure S5B), while knocking down the expression of ATP6V1A in MDA-MB-231 cells (Figure 4D), mirrored the effects of knocking down SIRT1 and increased the release of exosomes, with these vesicles being enriched in Survivin and ubiquitinated proteins (Figures 4E-4G). However, the ectopic expression of ATP6V1A in cells depleted of SIRT1 (Figure 4H) rescued the effects of knocking down SIRT1, thereby reducing exosome numbers (Figure 4I), and the amount of ubiquitinated proteins detected in their exosomes, as well as the level of Cathepsin B in the VFM (Figure 4J).

SIRT1 regulates ATP6V1A mRNA stability

The reduction in ATP6V1A protein levels observed in cells lacking SIRT1 was accompanied by a corresponding decrease in its mRNA levels, as determined by performing RT-qPCR using two independent primer sets that target the ATP6V1A mRNA (Figure 5A). The expression levels of a number of lysosomal genes are known to be regulated by the transcription factor EB (TFEB), which binds to CLEAR elements in the promoter regions of these genes and promotes their transcription (Sardiello et al., 2009). Because there is a putative CLEAR motif upstream of the ATP6V1A transcriptional start site (Figure 5B, diagram), and given a recent study identifying SIRT1 as a positive regulator of the transcriptional activity of TFEB (Bao et al., 2016), we examined whether knocking down SIRT1 inhibited TFEB transcriptional activity and reduced ATP6V1A mRNA levels. However, using a Dual-Luciferase reporter whose expression was under the control of the ATP6V1A promoter (Figure 5B, diagram), no differences in luciferase luminescence between control and SIRT1 knock down cells were detected (Figure 5B, graph). Moreover, the mRNA levels of several known TFEB targets were not affected to the same extent as ATP6V1A levels under conditions where SIRT1 expression was inhibited using shRNAs (Figure S6A).

Figure 5. SIRT1 Regulates ATP6V1A mRNA Stability.

(A) RT-qPCR was performed using two primer sets to determine ATP6V1A transcript levels (relative to actin) in sham shRNA expressing control (CTRL), and SIRT1 KD, cells. The transcript levels of SIRT1 were also determined as a control. (B) Dual-Reporter Luciferase assays were performed on the promoter region of the ATP6V1A gene. Top panel: a schematic of the luciferase reporter used. Bottom panel: the ratio of luciferase luminescence to renilla luminescence in sham shRNA expressing control (CTRL), and SIRT1 KD, MDA-MB-231 cells expressing the reporter construct. (C) ATP6V1A mRNA stability assays were performed on Actinomycin-A treated MDA-MB-231 cells treated with DMSO, or EX-527 (20 μM), or ectopically expressing SIRT1 (SIRT1 OE), for the indicated times. (D) ATP6V1A mRNA stability assays were performed on cells ectopically expressing ATP6V1A mRNA containing (UTR), or lacking (CDS), its 3′UTR sequence and treated with either DMSO, or EX-527 (20 μM), for the indicated times. (E) ATP6V1A transcript levels were determined in TNBC tumors, and normal tissues adjacent to the TNBC tumors (NTNBC) using RNA-Seq Nexus (GEO accession: GSE58135). (F) Correlation of ATP6V1A and SIRT1 mRNA levels in the tumor samples in (E), as well as in (G) the GDC TCGA Breast Cancer dataset. The data shown in (A), (B), (C), (D) and (E) represent means ± SD; ****p<0.0001, ***p<0.001, **p<0.01, and ns. See also Figure S6.

Mechanistic Target of Rapamycin Complex 1 (mTORC1) is another known regulator of lysosomal function, based on its ability to inhibit TFEB activity and reduce the levels of several components of the proton pump (Peña-Llopis et al., 2011). However, treatment of SIRT1 knock down cells with rapamycin or Torin 1, two inhibitors of mTORC1, had no effect on the enrichment of ubiquitinated cargo in exosomes or ATP6V1A mRNA levels (Figures S6B and S6C).

While this manuscript was in preparation, another study was published showing that SIRT1 was capable of promoting the transcription of genes involved in autophagy and lysosomal biogenesis by inhibiting a negative regulator of transcription, bromodomain-containing protein 4 (BRD4). To investigate whether BRD4 binding might be responsible for the reduced ATP6V1A mRNA levels in SIRT1 knock down cells, we used a specific inhibitor of BRD4, (i.e. JQ1). While JQ1 treatment of SIRT1 knock down cells increased the mRNA levels of SQSTM1 mRNA, as previously reported (Sakamaki et al., 2017), the inhibitor had no effect on ATP6V1A or CLCN7 mRNA levels in control, or SIRT1 depleted, cells (Figure S6D).

We then examined whether SIRT1 influences the turnover of the ATP6V1A RNA transcript. A key factor that affects the stability of RNA transcripts is the presence of A-U rich elements (AREs) and U-stretches in their 3′-untranslated-regions (3′UTRs). Certain RNA-binding proteins bind AREs and U-stretches in mRNA transcripts and alter their half-lives (Schoenberg and Maquat, 2012). Analysis of the 3′UTR of the ATP6V1A mRNA revealed an unusually large number of AREs and U-stretches, compared to other lysosomal genes (Figure S6E). This led us to determine whether ATP6V1A mRNA stability was affected by SIRT1. MDA-MB-231 cells transcriptionally inhibited by actinomycin-D treatment, were further treated with either DMSO, or the SIRT1 inhibitor EX-527, for increasing lengths of time. The RNA was then isolated from these cells and analyzed for ATP6V1A transcript levels. The results showed that the ATP6V1A mRNA in cells treated with EX-527, had a much shorter half-life, compared to cells treated with DMSO (Figure 5C). Ectopic expression of SIRT1 in the same cells resulted in the further stabilization of the ATP6V1A transcript. Moreover, EX-527 treatment did not have detrimental effects on the stability of an ectopically expressed ATP6V1A construct that lacked its 3’UTR. In contrast, the inclusion of the 3’UTR in this construct resulted in a significant reduction in the stability of the ectopically expressed ATP6V1A transcript, upon treatment of the cells with EX-527 (Figure 5D).

Because SIRT1 is frequently down-regulated in TNBC (Figure 1A), we examined whether there was a corresponding decrease in ATP6V1A mRNA levels in these tumor samples. Indeed, ATP6V1A transcript levels were significantly lower in TNBC tissues, compared to the normal adjacent tissues (Figure 5E). Moreover, there was a positive correlation between SIRT1 and ATP6V1A mRNA levels within the same tumors in this dataset (Figure 5F), as well as when analyzing a larger dataset of breast tumors using the TCGA database (GDC TCGA Breast Cancer, Figure 5G).

The secretome of SIRT1 depleted breast cancer cells promotes cell survival and invasive activity

To determine the biological effects of exosomes derived from SIRT1 knockdown cells, we first wanted to demonstrate the transfer of cargo in exosomes to recipient cells. A yellow fluorescent protein (YFP)-tagged form of Survivin (YFP-SURV), when ectopically expressed in MDA-MB-231 cells, could be detected in their exosomes (Figure S7A). When treating the non-invasive MCF10AT1 breast cancer cell line with these exosomes, YFP-tagged Survivin was detected within the cells, as indicated by Western blot analysis (Figure S7B) and immunofluorescent microscopy (Figure S7C). We then investigated whether the exosomes and Cathepsins present in the secretome of cancer cells with reduced expression of SIRT1, work together to promote a cancer cell phenotype. To examine this possibility, spheres of non-invasive MCF10AT1 cells were generated and embedded in a collagen matrix (Figure 6A). The cells were cultured in the absence (untreated control), or presence, of equivalent amounts of exosomes and/or VFM isolated from control, and SIRT1 depleted, MDA-MB-231 cells. The addition of exosomes or VFM from either of these two cell types had only minimal effects on MCF10AT1 sprouting (invasion) (Figures 6B and 6C). However, the combination of exosomes and VFM derived from the SIRT1 knock down MDA-MB-231 cells strongly promoted this invasive phenotype (i.e. Day 4 in Figure 6D).

Figure 6. The secretome of SIRT1 depleted breast cancer cells promote invasion.

(A) Diagram of the invasion assay. Spheroids of MCF10AT1cells were prepared, embedded in collagen matrix, cultured under different conditions, and the extent to which they migrated was determined. (B-D) Images of invasion assays performed on MCF10AT1 cells treated with equivalent amounts of (B) exosomes (EXO), (C) vesicle-free media (VFM), or (D) EXO and VFM from either sham shRNA expressing control (CTRL), or SIRT1 KD, MDA-MB-231 cells (SIRT1 KD). Images show the extent of cell outgrowths on days 0,2, and 4 of the experiment. Arrows highlight areas of invasion and the insets are higher magnifications of boxed areas. Scale bar, 0.3 mm. (E) Fluorescent microscopy and second-harmonic generation (SHG) images of invasion assays performed on GFP-expressing MCF10AT1 cells treated with an equivalent amount of exosomes (EXO) and vesicle free media (VFM) collected from sham shRNA expressing control (CTRL), and SIRT1 KD, MDA-MB-231 cells, or from SIRT1 KD MDA-MB-231 cells ectopically expressing ATP6V1A (SIRT1 KD/V1A OE). Some of the cells were also treated with CA-074 (10 μM). Arrow heads indicate areas where cells have invaded into the collagen. Scale bar, 100 μm. (F) Quantification of invasion area for each condition in (E). (G) Cell death assays were performed on MCF10AT1 cells that were left untreated (Serum Starved), or were treated with exosomes isolated from either sham shRNA expressing control (EXO-CTRL), or SIRT1 KD, cells (EXO-SIRT1 KD). As a control, cells were cultured with media containing 2% fetal bovine serum (FBS). (H) Quantification of wound healing (migration) assays performed in Figure S7F, where MCF10AT1 cells were left untreated, or were treated with exosomes from either control (CTRL), or SIRT1 KD, MDA-MB-231 cells. The data shown in (F), (G), and (H) represent means ± SD; ****p<0.0001, ***p<0.001, **p<0.01, and ns. See also Figure S7.

We also ectopically expressed a green fluorescent protein (GFP) construct in the cells to visualize them embedded in the collagen matrix by fluorescent microscopy. Again the addition of exosomes and VFM from SIRT1 depleted MDA-MB-231 cells increased the invasion of GFP-expressing MCF10AT1 cells, compared to cells treated with exosomes and VFM from control MDA-MB-231 cells (Figures 6E, top panels and 6F). This approach also allowed for the visualization of areas where collagen was invaded by the cells (Figure 6E, see arrows in the middle and bottom panels). The increase in cell invasion caused by exosomes and VFM from SIRT1 knock down cells was significantly reduced when MCF10AT1 cells were treated with the Cathepsin B inhibitor CA-074, or when ATP6V1A was ectopically expressed in the SIRT1 knock down MDA-MB-231 cells prior to collecting their exosomes (Figures 6E and 6F). Consistent with our findings that SIRT1 inhibits lysosomal function to generate a unique secretome, an increase in the invasiveness of GFP-expressing MCF10AT1 cells was promoted by exosomes and VFM derived from MDA-MB-231 cells treated with Bafilomycin-A (Figures S7D and S7E).

Survival and wound-healing (scratch) assays were also performed. Exosomes isolated from MDA-MB-231 cells depleted of SIRT1 were more effective at promoting the survival of serum-deprived MCF10AT1 cells, compared to an equivalent amount of exosomes from control cells (Figure 6G), and they strongly stimulated MCF10AT1 cell migration, as determined in wound closure assays (Figures 6H and S7F). Because exosomes derived from SIRT1 knock down cells are enriched with Survivin, a protein known to promote cell survival and migration (Li et al., 2015; Mckenzie et al., 2010), we determined whether it was important for mediating these effects. Control and SIRT1 knock down MDA-MB-231 cells were treated with phosphate buffer saline (PBS) as a control, or with YM155, a small molecule that inhibits Survivin expression (Figure S7G), such that it is absent from exosomes generated by SIRT1 knock down cells (Figure S7H). Exosomes from control cells treated without (PBS treated) or with YM155 caused only a modest enhancement in the survival and migration of MCF10AT1 cells (Figures S7I-S7K). However, exosomes from SIRT1 knock down cells treated with YM155 no longer promoted cell survival and migration as effectively as exosomes from SIRT1 knock down cells treated with only PBS (Figures S7I- S7K).

DISCUSSION

Lysosomes help maintain cellular homeostasis by degrading unwanted proteins, RNA, and DNA. This occurs as MVBs containing intraluminal vesicles and soluble proteins are trafficked to, and fuse with, lysosomes, exposing their contents to the acidic and hydrolase-rich environment of the lysosomal lumen (Davidson and Vander Heiden, 2017). However, some MVBs are directed to the cell surface, where they give rise to a class of EVs referred to as exosomes. Research involving EVs has been attracting increasing attention from diverse fields of biology and the pharmaceutical industry, primarily because it is now recognized that virtually all cells form and shed distinct classes of EVs that contain cargo reflecting their cellular origin. EVs can be transferred to other cells, resulting in phenotypic changes that have an impact on several biological processes and diseases and have been extensively studied in the context of cancer (Desrochers et al., 2016a). Exosomes derived from highly aggressive cancer cells have been shown to promote cell growth and survival, as well as invasive and metastatic activities (Costa-Silva et al., 2015; Hoshino et al., 2015; Kreger et al., 2016). Still, there remain large gaps in our understanding of EVs, especially regarding the mechanisms that regulate exosome biogenesis.

Our findings, summarized in Figure 7, identify a connection between SIRT1, lysosomal activity, and the formation of a secretome with unique characteristics. Specifically, we discovered that knocking down SIRT1 in breast cancer cells decreases the protein levels of the ATP6V1A subunit of the V-ATPase proton pump located on lysosomal membranes. This results in a poorly functioning pump and the inability of lysosomes to maintain the low pH needed for degradative activity. Thus, in cells lacking sufficient amounts of SIRT1 expression or activity to sustain proper lysosomal function, MVBs that would typically be degraded in lysosomes, instead fuse with the plasma membrane and release their contents (i.e. exosomes with unique cargo and hydrolases that normally reside in the lysosomes to degrade proteins) into the extracellular environment.

Figure 7. SIRT1 downregulation alters the secretome of breast cancer cells by impairing lysosomal function.

Model describing the role of SIRT1 in regulating exosome biogenesis and hydrolase secretion. Decreasing SIRT1 levels in breast cancer cells reduces the stability of the ATP6V1A transcript and causes a corresponding loss in the expression of the ATP6V1A protein. This impairs lysosomal function and results in MVBs that would normally be degraded in the lysosomes to instead fuse with plasma membrane and release their content, i.e. exosomes and hydrolases.

A possible explanation for the increased release of exosomes in SIRT1 depleted cells involves the V-ATPase machinery directly impacting exocytosis. The V0 and V1 domains of the V-ATPase assemble together on late MVBs and lysosomes to acidify the lumens of these structures. However, it has been recently shown that the V0 domain can dissociate from the V1 domain to promote exocytosis (Liégeois et al., 2006; Poëa-Guyon et al., 2013). Since knocking down SIRT1 decreases the expression of ATP6V1A, a subunit of the V1 domain, it was plausible that this could be sufficient to cause the dissociation of the V0 domain and increase exosome release. Although we cannot completely rule out this possibility, inhibiting lysosomal function using Chloroquine or Bafilomycin-A, or interfering with the ability of MVBs to fuse with lysosomes (by knocking down RAB7), showed the same effects as reducing SIRT1 levels. It is also worth noting that treating cells with Chloroquine has been shown to enhance the assembly of the V1-V0 subunits (Stransky and Forgac, 2015), further suggesting that inhibiting the acidification of lysosomes is sufficient to produce a unique secretome.

Using 3D cultures of non-invasive MCF10AT1 breast cancer cells, we showed that distinct components within the secretome of SIRT1-deficient breast cancer cells act synergistically to promote cell survival and invasive activity. Specifically, the increased amounts of lysosomal hydrolases secreted by SIRT1 KD breast cancer cells, particularly Cathepsin B, degrade extracellular matrix components, while exosomes enriched in Survivin derived from these cells strongly promote cell survival and migration. The release of exosomes and soluble factors by highly aggressive cancer cells may be important for cancer progression, as tumors are composed of a heterogenous collection of cells. Thus, more aggressive cells may produce a secretome that alters the behavior of less aggressive cells within the tumor, making them more invasive and drug-resistant.

While it is well-accepted that the multi-subunit V-ATPase proton pump is essential for maintaining the acidification and function of lysosomes (Cotter et al., 2015), how the components of this pump are regulated, as well as the consequences of deregulating lysosomal activity in cancer (Kallunki et al., 2013), is poorly understood. Thus, the regulation of lysosomal function by SIRT1 offers new insights into these questions. The expression of the ATP6V1A subunit of the V-ATPase proton pump was reduced by ~65% in SIRT1 KD MDA-MB-231 cells, compared to control cells, due to decreased ATP6V1A transcript levels. Based on previous findings (Jing and Lin, 2015; Shan et al., 2017), we initially suspected that knocking down SIRT1 would inhibit the transcription of the ATP6V1A gene. However, the transcription of this gene was similar in both control and SIRT1 KD cells. Instead, we found that depleting cells of SIRT1 decreased the half-life of the ATP6V1A transcript. One likely possibility for this effect is that depleting cells of SIRT1 leads to the acetylation and inactivation of a protein that binds to and stabilizes the ATP6V1A transcript and we are currently attempting to identify such a SIRT1 substrate.

SIRT1 is best known for its role in extending lifespan (Cohen et al., 2004; Lin et al., 2000). More recently, there have been suggestions that it also can function as a tumor suppressor (Chalkiadaki and Guarente, 2015), as reducing SIRT1 expression levels frequently occurs in TNBCs. However, how SIRT1 might act to inhibit cancer progression has been an open question. Our findings showing that knocking down SIRT1 expression in cancer cells impairs lysosomal activity, and results in the generation of a secretome capable of strongly promoting cell survival and invasive activity, now offer a plausible explanation. This previously unappreciated connection between SIRT1 and lysosomal function may also help to shed light on how reducing SIRT1 levels can negatively impact aging and certain neurodegenerative disorders that are characterized by the loss of SIRT1 expression (Kim et al., 2007; Min et al., 2010).

Contact for Reagent and Resource Sharing

Further information and requests for reagents may be directed to and will be fulfilled by the Lead Contact, Richard Cerione (rac1@cornell.edu).

Experimental Model and Subject Details

Cell Lines

Human embryonic kidney (HEK)-293T, MDA-MB-231, MDA-MB-453, and Hs-578T breast cancer cells, U87 glial cells, and Human Kidney-2 (HK-2) cells were obtained from the ATCC (https://www.atcc.org/), while the primary Human Dermal Fibroblast were purchased from Zenbio. The MCF10AT1 cells were provided to us by Claudia Fischbach-Teschl, Cornell University, and the Normal Human Astrocytes were from Ichiro Nakano, University of Alabama. HEK-293T cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% calf serum (CS). MDA-MB- 231, MDA-MB-453, Hs-578T, ZR-75-1, BT-474, CAMA-1, SK-BR-3, T-47D, MCF7, MDA-MB-468, TSE, and U87 cells were cultured in Roswell Park Memorial Institute 1640 (RPMI-1640) medium supplemented with 10% fetal bovine serum (FBS). MCF10AT1 were cultured in DMEM/F12 medium supplemented with 5% horse serum, 10 μg/mL insulin, 0.5 μg/mL hydrocortisone, 100 ng/mL cholera toxin, 20 ng/mL EGF, 100 I.U./mL penicillin, and 100 μg/mL streptomycin. Primary Human Dermal Fibroblast were cultured in DMEM supplemented with 10% FBS, 100 I.U./mL penicillin, 100 μg/mL streptomycin, and 0.25 μg/mL Amphotericin B. Normal Human Astrocytes were cultured in DMEM/F-12 supplemented with 10% FBS. HK-2 cells were cultured in Keratinocyte Serum Free Medium (K-SFM) supplemented with 0.05 mg/ml bovine pituitary extract (BPE), and 5 ng/mL EGF. All cell lines were maintained at 37°C with 5% CO₂.

Cells stably expressing constructs of interest were selected for, and maintained by, supplementing the growth medium with 2 μg/mL puromycin.

Method Details

Plasmid Generation, Virus Production, and Cell Infection

The pLJM1-LAMP1-mRFP-FLAG (#34611) construct was purchased from Addgene, and the ATP6V1A transcript, lacking or containing its 3′UTR, were cloned into the pLJM1 plasmid using the following primers: ATP6V1A-CDS-Forward (CGTCAGATCCGCTAGCATGGATTTTTCCAAGCTACCC), ATP6V1A-CDS-Reverse (TCGAGGTCGAGAATTCCTAATCTTCAAGGCTACGGAATGC), ATP6V1A-UTR-Forward (CGTCAGATCCGCTAGCATGGATTTTTCCAAGCTACC), ATP6V1A-UTR-Reverse (TCGAGGTCGAGAATTCTGTTAATTTAAATCCACTTTTTATT).

For the luciferase reporter assay, the 500 bp region immediately upstream of the ATP6V1A transcription start site was cloned into the pGL3-Luciferase reporter vector (E1751), and for transfection efficiency normalization the pRL-Renilla reporter (E2231) was purchased from Promega. All shRNA constructs were from Sigma.

Lentiviruses were generated by transfecting HEK-293T cells with the shRNA plasmids (Sigma) and the packaging plasmids (#12259 and #12263, Addgene) using Fugene 6 (Promega). The viruses shed into the medium by the cells were harvested 24 and 48 h after transfection. The viruses were then used to infect the target cells using Polybrene (8 μg/mL).

To generate SIRT1 knockout MDA-MB-231 cells, CRISPR/Cas9 was used. Two sgRNAs were used to induce double-strand DNA breaks and clones that lost one copy of SIRT1 gene, or both copies of this gene, were derived by selection with 5 μg/mL blasticidin. The results were confirmed with PCR.

RNA Isolation and Quantitative (q) RT-PCR Analysis

Total RNA was isolated from cells using the PureLink RNA Mini Kit (Invitrogen), and the mRNA transcripts were converted to cDNA using Superscript III Reverse Transcriptase (Invitrogen) and oligo dT₂₀. The cDNA was then used to determine the expression levels of the indicated transcripts using SYBR Green Supermix (Bio-Rad) and the Applied Biosystems® 7500 Real-Time PCR System with the T method (ABI). The following primers were used for the RT-qPCR analyses:

SIRT1-FW CGTCAGATCCGCTAGCATGGATTTTTCCAAGCTACCC

SIRT1-RW TCGAGGTCGAGAATTCCTAATCTTCAAGGCTACGGAATGC

ATP6V1A-S1-FW CGTCAGATCCGCTAGCATGGATTTTTCCAAGCTACC

ATP6V1A-S1-RW TCGAGGTCGAGAATTCTGTTAATTTAAATCCACTTTTTATT

ATP6V1A-S2-FW ATGTCTAATGGTTATGAAGACCACATGGCC 312032410c2

ATP6V1A-S2-RW TTATTTGTGGGACTTGTTGGTTTTGAAGGAAAC

ATP6V1A-CDS and UTR-FW CAAAGACGATGACGACAAGa

ATP6V1A-CDS and UTR-RW CCCACTCTCACCAGCTCATA

GBA-FW ATGGAGCGGTGAATGGGAAG

GBA-RW GTGCTCAGCATAGGCATCCAG

CTSB-FW ACAACGTGGACATGAGCTACT

CTSB-RW TCGGTAAACATAACTCTCTGGGG

CLCN7-FW CCCACACAACGAGAAGCTCC

CLCN7-RW ACTTGTCGATATTGCCCTTGATG

ATP6V1H-FW CAGAAGTTCGTGCAAACAAAGTC

ATP6V1H-RW TCAGGGCTTCGTTTCATTTCAA

ATP6V0D1-FW GCATCACCTCTGACGGTGTC

ATP6V0D1-RW CTCCTTAATGTCACGCACGAT

ACTB-FW CATGTACGTTGCTATCCAGGC

ACTB-RW CTCCTTAATGTCACGCACGAT

SQSTM1-FW GACTACGACTTGTGTAGCGTC

SQSTM1-RW AGTGTCCGTGTTTCACCTTCC

RNA Stability Assay

To determine the stability of the endogenously or exogenously expressed ATP6V1A transcripts in control or SIRT1 KD MDA-MB-231 cells, the RNA from these cells was collected at different times (up to 3 h) following their treatment with Actinomycin D (Cayman Chemicals, 4 μg/mL). The levels of each transcript were then determined by RT-qPCR using the primer sets listed in the RNA Isolation and Quantitative (q) RT-PCR Analysis section.

Dual Reporter Luciferase Assay

The pGL3-Luciferase reporter construct containing the ATP6V1A promoter (5 μg) and the pRL-Renilla reporter construct (0.5 μg) were transfected into control and SIRT1 KD MDA-MB-231 cells using Fugene 6. After 16 h, the luciferase and the renilla bioluminescence was measured using a BioTek Synergy 2 plate reader according to the manufacturer’s instructions.

EV and Vesicle Free Medium Preparation

The conditioned medium collected from 2.0 × 10⁶ serum starved cells was subjected to two consecutive centrifugations at 700 × g to clarify the medium of cells and cell debris. The partially clarified medium was filtered using a 0.22 μm pore size Steriflip PVDF filter (Millipore). The filter was rinsed two times with 5 mL of Phosphate buffered saline (PBS) to remove any remaining exosome sized EVs (less than 0.22 μm) from the filter. The EVs larger than 0.22 μm retained by the filter were lysed using 250 μL of lysis buffer (25 mM Tris, 100 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM DTT, 1 mM NaVO₄, 1 mM β-glycerol phosphate, and 1 μg/ml each aprotinin and leupeptin). This was considered the microvesicle (MV) lysate. The filtrate was then subjected to ultracentrifugation at 100,000 × g for 8 h. The pelleted exosomes were either lysed using 250 μL of lysis buffer, or resuspended in 500 μL of serum free medium for cell-based assays. The supernatant depleted of MVs and exosomes was concentrated using 10 KDa centricons (Amicon). This was considered the vesicle free medium (VFM). Whole cell lysates (WCL) were prepared by rinsing dishes of cells with PBS, adding 800 μL of lysis buffer, and scraping the cells off the plate. The resulting MV, exosome, and cell lysates were centrifuged at 16,000 × g for 10 min, and then the supernatants were subjected to Western blot analysis.

SILAC and Mass Spectrometry

Quantitative proteomics using SILAC was performed on samples as described in Zhang et al., 2016. Briefly, MDA-MB-231 cells were cultured in SILAC RPMI-1640 media supplemented with Dialyzed FBS (HyClone) and either [¹³C₆,¹⁵N₂]-L-lysine (Sigma) and [¹³C₆,¹⁵N₄]-L-arginine (Sigma) or L-lysine (Sigma) and L-arginine (Sigma) for five generations. The MDA-MB-231 cell line that was grown with heavy L-lysine and L-arginine was then treated with lentivirus containing shRNA targeting SIRT1, whereas the MDA-MB-231 cell line grown in normal L-lysine and L-arginine was treated with control lentivirus shRNA. After 48 h, the cells, and the exosomes released by these cells, were lysed in lysis buffer, while the vesicle free media (VFM) samples were prepared as described above. The protein concentrations of the samples were determined by Bradford assay, and an equivalent amount of each sample (30 μg) was processed as outlined in Zhang et al, 2017. The resulting lyophilized peptides were then analyzed using nano LC-MS/MS (Cornell University, Proteomics Facility). All data was acquired using Xcalibur 2.2 operation software.

Endolysosomal Immunoprecipitation

Endolysosomal immunoprecipitations were performed as described in Abu-Remaileh et al., 2017. Briefly, 30 million cells stably expressing FLAG-tagged TMEM192 were infected with control shRNA, or SIRT1 targeting shRNA. After 48 hours, the cells were washed twice with PBS before being removed from the dish using a cell scraper and resuspended in 1.0 mL KPBS (136 mM KCl, 10 mM KH₂PO₄, pH 7.25). The cells were centrifuged at 1000 × g for 2 min at 4°C, and the cell pellets were resuspended in 950 μL of KPBS, 25 μL of which was used as a loading control. After homogenization of the remaining cells using a dounce homogenizer, the samples were centrifuged at 1000 × g for 2 min at 4°C. The resulting supernatant was the n incubated with 100 μL of magnetic anti-FLAG beads for 15 minutes. The beads were captured using a magnet and washed three times with KPBS before being lysed with 100 uL of lysis buffer. The loading controls, as well as the various immunoprecipitates, were analyzed using Western blot analysis.

Lysosomal pH measurement

Lysosomal pH was determined as described previously (Zoncu et al., 2011) with slight modifications. 1.0 × 10⁶ sham shRNA expressing control, or SIRT1 knock down, MDA-MB-231 cells were treated with 50 μg/mL of Lysosensor yellow-blue Dextran® for 12 hours, before being amino acid starved for an additional 2 hours. The cells were then rinsed twice with PBS, and resuspended in physiological buffer (136mM NaCl, 2.5mM KCl, 2mM CaCl2, 1.3mM MgCl₂, 5mM Glucose, 10mM HEPES pH 7.4), and transferred to individual wells of a black 96-well plate. Lysosensor fluorescence emission was recorded at 460 nm and 540 nm upon excitation at 360 nm using a BioTek Synergy 2 Plate Reader. To measure the lysosomal pH, the 460/540 fluorescence emission ratios were interpolated to a calibration curve that was established by resuspending the cells containing lysosensor in 200 μL aliquots of pH calibration buffers (145 mM KCl, 10 mM glucose, 1 mM MgCl2, and 20 mM of either HEPES, MES, or acetate supplemented with 10 μg/ml nigericin), buffered to pH ranging from 3.5 to 8.0.

Lysosomal pH Recovery Assay

Lysosomal pH was determined as described previously (Chen et al., 2017) with slight modifications. Sham shRNA expressing control, and SIRT1 KD, MDA-MB-231 cells (5 × 10³) plated in 24 well plates were treated with either DMSO, or 100 nM Bafilomycin-A, for 1 h, at which point the cells were rinsed extensively with media and allowed to recover for 1 h. The cells were then treated with LysoTracker® Green DND 26 (Thermo Fisher) for 1 h, washed twice with PBS, and were lysed in 200 μL of RIPA buffer (10 mM Tris, 140 mM NaCl, 1% Triton X-100, 1 mM EDTA, 0.1% SDS, 0.1% sodium deoxycholate and 1 mg/ml each aprotinin and leupeptin). The fluorescence of each sample was measured in a 96-well plate using a BioTek Synergy 2 Plate Reader; EX: 485 nm and EM: 520 nm. The percent recovery was calculated as follows:

$$\begin{gathered} 100\%:\text{DMSO for the whole duration of the assay} \\ 0\%:\text{Baf-A for the whole duration of the assay} \\ \%\text{recovery}=(A_R-A_0)∕(A_{100}-A_0) \end{gathered}$$

Western Blot Analysis

The protein concentrations of cell, microvesicle, and exosome lysates, as well as of vesicle free media samples, were determined using the Bradford protein assay. The lysates were normalized by protein concentration, resolved by SDS-PAGE, and the proteins were transferred to PVDF membranes. Membranes were blocked with 5% bovine serum albumin (Sigma) in TBST (19 mM Tris Base, 2.7 mM KCl, 137 mM NaCl, and 0.5 % Tween-20) for 1 h and the membranes were incubated with the indicated primary antibodies overnight, followed by detection with HRP-conjugated secondary antibodies (Cell Signaling Technology) and exposure to ECL reagent (Pierce).

Immunofluorescence Microscopy

Cells grown on glass coverslips were treated as indicated, fixed and permeabilized with methanol. The slides were then blocked with 10% bovine serum albumin diluted in PBS. For each antibody used, the cells were incubated for 90 min using the following dilutions: CD63 (1:200, Abcam), LAMP1 (1:100, Cell Signaling Technologies), Cathepsin B (1:800, Cell Signaling Technologies), GFP (1:100, Cell Signaling Technologies). Anti-Mouse IgG-Alexa 488 Conjugate antibody (1:400, Thermo Fisher) and anti-Rabbit IgG-Alexa 568 Conjugate antibody (1:400, Thermo Fisher). DAPI (Sigma) was used to label nuclei, and conjugated Phalloidin (1:2000) was used to label actin. The cells were visualized with Super Resolution Structured Illumination Microscopy using a Zeiss Elyra Super Resolution Microscope with a 63x oil objective lens (Cornell University, Biotechnology Resource Center). Image processing and quantification was performed with ImageJ software.

Nanoparticle Tracking Analysis

The sizes and concentrations of EVs in a given sample were determined using a NanoSight NS300 (Malvern, Cornell NanoScale Science and Technology Facility) as described in Kreger et al., 2016. Briefly, the samples were diluted in serum free RPMI-1640 and injected into the beam path to capture movies of EVs as points of diffracted light moving rapidly under Brownian motion. Five 45-s digital videos of each sample were taken and analyzed to determine the concentration and size of the individual EVs based on their movement, and then results were averaged together.

Electron Microscopy:

Transmission Electron Microscopy (TEM) on exosomes was performed as described in Desrochers et al., 2016b. Briefly, 5 μL of an exosome preparation derived from either control or SIRT1 KD MDA-MB-231 cells were diluted in PBS, added to a carbon-coated 300-mesh copper grid, and then stained with 1.75% uranyl acetate. Once dry, the samples were imaged using the FEI T12 Spirit 120 kV Field Emission Transmission Electron Microscope at Cornell’s Center for Materials Research (CCMR), supported by NSF MRSEC award number: NSF DMR-1120296.

Cathepsin B Activity Assay

The Magic Red substrate for the Cathepsin B Activity Assay Kit (Immunochem Technologies) was diluted in PBS to a 20X concentration. Then, 95 μL of concentrated conditioned media (prepared using 10 KDa centricons (Amicon)) collected from an equivalent number of either sham shRNA expressing control, or SIRT1 KD, MDA-MB 231 cells were added to each well of a 96-well plate containing 5 μL of 20X Magic Red. In some cases, 10 μM of CA-074 was added to the samples for 30 minutes. The resulting fluorescence that occurred over time was readout using a Cary Eclipse Fluorescence Spectrophotometer; EM: 530 nm and EX: 595 nm.

Cell Death Assays

MCF10AT1 cells grown in 6-well dishes were placed in serum-free medium supplemented with nothing (serum starved) or an equivalent amount (5.0 × 10⁷ exosomes/mL) exosomes from sham shRNA expressing control, or SIRT1 KD, MDA-MB-231 cells. Approximately 40 h later, the cells were stained with DAPI to label nuclei. The cells were visualized by fluorescent microscopy and nuclear condensation and/or blebbing was used to identify dead or dying cells.

Wound Healing Assay

Confluent cultures of MCF10AT1 cells were placed in serum free medium supplemented with nothing, or equivalent amounts of exosomes (5.0 × 10⁷ exosomes/mL) from control or SIRT1 KD MDA-MB-231 cells that had been left untreated, or were treated with either PBS or YM155, for 12 h. Wounds were then struck through the cells using a pipette tip and the medium on the cells was replaced. Approximately 8 h later, the cells were fixed with 4% paraformaldehyde and then imaged using phase contrast microscopy. The extend of wound closure for each condition was determined using ImageJ software.

Invasion Assay

Poly(dimethylsiloxane) (PDMS, Dow Corning) was cast onto a petri dish (150 mm) to form a 1.0 mm thick layer. Biopsy punches of 6 and 8 mm were generated in the PDMS to create a ring pattern. The PDMS rings were treated with plasma cleaner and covalently bond to a glass coverslip followed by treatment with 1% [v/v] polyethyleneimine (Sigma) and 0.1% [v/v] glutaraldehyde (Fisher). To generate spheroids of MCF10AT1 cells, the wells of a 96-well plate were coated with 50 μL of 1.5% agarose diluted in PBS to form a non-adhesive layer. 5 × 10³ MCF10AT1, or GFP expressing MCF10AT1, cells were added to each well, and the plates were placed in a shaking incubator at 37°C overnight. The resulting spheroids that formed were individually selected using a glass pasteur pipette, mixed with collagen (Corning), and cast into the center of each PDMS ring. The samples were then subjected to consecutive temperature changes from ice cold, to room temperature, to 37°C at 15 min intervals. After the collagen solidified, each well was treated with various combinations of exosomes (5.0 × 10⁷ exosomes/mL), VFM (400 μg/mL), and inhibitors, as indicated for 4 days and fixed with paraformaldehyde. Bright field images of the cells were taken every day and their media was changed every other day. The GFP expressing MCF10AT1 cells, and the collagen matrix adjacent to the cells, were subjected to fluorescent and second-harmonic generation (SHG) imaging microscopy, respectively. The extent of invasion was calculated as the area of sprouting and outgrowth for each spheroid.

Quantification and Statistical Analysis

Quantitative data are presented as means ± SD. All experiments were independently performed at least three times. Statistical significance was calculated by ANOVA (Tukey’s test) for experiments involving comparing more than two conditions, and student’s t test for experiments involving comparing two conditions. Error bars represents the mean ± SD. *p ≤ 0.05, **p ≤ 0.01, ***p<0.001, ****p<0.0001, ns = non-significant.

KEY RESOURCE TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-SIRT1 antibody	Cell Signaling Technology	2493S; RRID: AB_2188359
Anti-SIRT6 antibody	Cell Signaling Technology	12486S; RRID: AB_2636969
Anti-SIRT7 antibody	Cell Signaling Technology	5360S; RRID: AB_2716764
Anti-RAB27A antibody	Cell Signaling Technology	95394S
Anti-RAB7A antibody	Cell Signaling Technology	9367S; RRID: AB_1904103
Anti-IkBα antibody	Cell Signaling Technology	4812S; RRID: AB_10694416
Anti-β-Actin antibody	Cell Signaling Technology	3700S; RRID: AB_2242334
Anti-Survivin antibody	Novus Biologicals	NB500-201; RRID: AB_10001517
Anti-GFP antibody	Cell Signaling Technology	2956S; RRID: AB_1196615
Anti-HDAC6 antibody	Cell Signaling Technology	7558S; RRID: AB_10891804
Anti-Flotillin-2 antibody	Cell Signaling Technology	3436S; RRID: AB_2106572
Anti-Flag antibody	Cell Signaling Technology	8146S; RRID: AB_10950495
Anti-HSP90 antibody	Cell Signaling Technology	4877S; RRID: AB_2233307
Anti-LAMP1 antibody	Cell Signaling Technology	9091S; RRID: AB_2687579
Anti-LDHA	Cell Signaling Technology	3582S; RRID: AB_2066887
Anti-CD63 antibody	Abcam	ab59479; RRID: AB_940915
Anti-M6PR Antibody	Cell Signaling Technology	14364S
Anti-CD81 antibody	Millipore	MABF2061
Anti-ATP6V1A antibody	Abcam	ab137574; RRID: AB_2722516
Anti-ATP6V0D1 antibody	Abcam	ab56441; RRID: AB_940402
Anti-MMP7 antibody	Cell Signaling Technology	71031S
Anti-MMP9 antibody	Cell Signaling Technology	13667S
Anti-Cathepsin B antibody	Cell Signaling Technology	31718S; RRID: AB_2687580
Anti-Ubiquitin antibody	Santa Cruz Biotechnology	sc-8017; RRID: AB_628423
Anti-UB H2A antibody	Cell Signaling Technology	8240S; RRID: AB_10891618
Anti-UB H2B antibody	Cell Signaling Technology	5546S; RRID: AB_10693452
Anti-14-3-3 Zeta/Delta antibody	Cell Signaling Technology	7413S; RRID: AB_10950820
Anti-Rabbit IgG-HRP Conjugate antibody	Cell Signaling Technology	7074S; RRID: AB_2099233
Anti-Mouse IgG-HRP Conjugate antibody	Cell Signaling Technology	7076S; RRID: AB_330924
Anti-Mouse IgG-Alexa 488 Conjugate antibody	Thermo Fisher	A-11029; RRID: AB_2534088
Anti-Rabbit IgG-Alexa 568 Conjugate antibody	Thermo Fisher	A-11036; RRID: AB_10563566
Bacterial and Virus Strains
E.coli: Stellar Competent Cells	Clonetech	636763
E.coli: One Shot Stbl3 Chemically competent cells	Thermo Fisher	C737303
Chemicals, Peptides, and Recombinant Proteins
Leupeptin	Sigma	L9783
Aprotinin	Sigma	10236624001
Dithiothreitol (DTT)	Sigma	10197777001
Dimethylsulfoxide (DMSO)	Sigma	D8418
Chloroquine	Cayman Chemicals	14194
Bafilomycin-A	Cayman Chemicals	11038
LysoTracker™ Green DND-26	Thermo Fisher	L7526
LysoSensor™ Yellow/Blue dextran, 10,000 MW	Thermo Fisher	L22460
EX-527	Cayman Chemicals	10009798
Rapamycin	Cayman Chemicals	13346
Torin 1	Cayman Chemicals	10997
YM155	Tocris	6491
Nigercin	Sigma	N7143
Monesin	Sigma	M5273
Actinomycin D	Cayman Chemicals	11421
Puromycin	Sigma	P9620
Alexa Fluor™ 594 Phalloidin	Thermo Fisher	A12381
Keratinocyte Serum Free Medium (K-SFM)	Thermo Fisher	17005042
Bovine Pituitary Extract (BPE)	Thermo Fisher	13028014
Gibco™ Amphotericin B	Thermo Fisher	15290026
DAPI	Sigma	D9542
CA-074	Tocris	4863
EGF	Millipore	01-107
DMEM	Gibco	11965-092
RPMI	Gibco	11875-093
DMEM/F12	Gibco	12634-010
SILAC RPMI	Thermo Fisher	89984
Fetal Bovine Serum	Gibco	10437028
Calf Serum	Gibco	16010159
Horse Serum	Gibco	16050-122
Pen-Strep	Gibco	15140122
Dialyzed FBS	HyClone	SH30079.02HI
[13C6,15N2]-L-lysine	Sigma	608041
[13C6,15N4]-L-arginine	Sigma	608033
Insulin	Sigma	I2643-50MG
Hydrocortisone	Sigma	H4001-1G
Choleratoxin	Sigma	C8052-1MG
Critical Commercial Assays
RNA Isolation Kit	Invitrogen	12183018A
Dual-Reporter Luciferase Assay	Promega	E1910
InFusion Cloning Kit	Clonetech	638909
Superscript III Reverse Transcriptase	Invitrogen	18080044
Cathepsin B activity assay kit	Immunochem Technologies	937
Experimental Models: Cell Lines
Human: HEK-293T	ATCC	N/A
Human: MDA-MB-231	ATCC	N/A
Human: MDA-MB-453	ATCC	N/A
Human: Hs-578T	ATCC	N/A
Human: ZR-75-1	ATCC	N/A
Human: BT-474	ATCC	N/A
Human: CAMA-1	ATCC	N/A
Human: SK-BR-3	ATCC	N/A
Human: T-47D	ATCC	N/A
Human: MCF7	ATCC	N/A
Human: MDA-MB-468	ATCC	N/A
Human: TSE	ATCC	N/A
Human: Human Dermal Fibroblasts	Zenbio	DF-F
Human: Normal Human Astrocytes	Provided by Ichiro Nakano, University of Alabama	N/A
Human: U87	ATCC	HTB-14
Human: HK-2	ATCC	CRL2190
Human: MCF10AT1	Provided by Claudia Fischbach, Cornell Univeristy	N/A
Human: GFP-MCF10AT1	This Paper	N/A
Human: ATP6V1A-CDS MDA-MB-231	This Paper	N/A
Human: ATP6V1A-UTR MDA-MB-231	This Paper	N/A
Oligonucleotides
Control shRNA Sequence: CCGGCAACAAGATGAAGAGCACCAACTCGAGTTGGTGCTCTTCATCTTGTTGTTTTT	Sigma Mission ShRNA	SHC002
shRNA Targeting Sequence: SIRT1 : GTACCGGCATGAAGTGCCTCAGATATTACTCG AGTAATATCTGAGGCACTTCATGTTTTTTG	Sigma Mission ShRNA	TRCN0000218734
shRNA Targeting Sequence: SIRT6 : CCGGGAAGAATGTGCCAAGTGTAAGCTCGAG CTTACACTTGGCACATTCTTCTTTTTG	Sigma Mission ShRNA	TRCN0000232528
shRNA Targeting Sequence: SIRT7 : CCGGGTCCAGCCTGAAGGTTCTAAACTCGAG TTTAGAACCTTCAGGCTGGACTTTTTG	Sigma Mission ShRNA	TRCN0000359663
shRNA Targeting Sequence: RAB27A : GTACCGGGATCTTCTCTATGATTGATACCTCG AGGTATCAATCATAGAGAAGATCTTTTTTG	Sigma Mission ShRNA	TRCN0000380306
shRNA Targeting Sequence: RAB7A : GTACCGGGGTTATCATCCTGGGAGATTCCTCG AGGAATCTCCCAGGATGATAACCTTTTTTG	Sigma Mission ShRNA	TRCN0000380577
shRNA Targeting Sequence: ATP6V1A : CCGGGCTGTCCAACATGATTGCATTCTCGAGA ATGCAATCATGTTGGACAGCTTTTT	Sigma Mission ShRNA	TRCN0000029539
sgRNA-1 targeting SIRT1: CACCGGCTCCCCGGCGGGGGACGACG	This Paper	N/A
sgRNA-2 targeting SIRT1: CACCGTCGTACAAGTTGTCGGCCAG	This Paper	N/A
Recombinant DNA
pLJM1-LAMP1-mRFP-FLAG	Zoncu et al, 2011	Addgene Plasmid #34611
pGL3 Luciferase Reporter Vector - Basic	Promega	E1751
pRL Renilla Luciferase Control Reporter Vector	Promega	E2231
pMD2.G-VSV-G-expressing envelope plasmid	From Didier Trono	Addgene Plasmid #12259
pCMV delta R8.2-Lentiviral Packaging plasmid	From Didier Trono	Addgene Plasmid #12263
pUMVC-Retroviral Packaging plasmid	Stewart et al, 2003	Addgene Plasmid #8449
pYESir2-puro plasmid	Vaziri et al, 2001	Addgene Plasmid # 1769
LentiCRISPRv2-Blast	From Mohan Babu	Addgene Plasmid # 83480
LentiCRISPRv2GFP	Walter et al, 2017	Addgene Plasmid # 82416
pLJC5-Tmem192-2xFlag	Abu-Remaileh et al, 2017	Addgene Plasmid # 102929
pLJM1-ATP6V1A-CDS plasmid		This Paper
pLJM1-ATP6V1A-CDS+UTR plasmid		This Paper
Software and Algorithms
Snapgene Viewer	GSL Biotech	snapgene.com
ImageJ	NIH	https://imagej.nih.gov/ij/
Fiji open source image analysis software	Fiji	https://fiji.sc/
DAVID Bioinformatics Resource	Huang et al., 2008, 2009	https://david.ncifcrf.gov/
Prism	Graphpad	https://www.graphpad.com
Xcalibur 2.2	Thermo Fisher
Other
SYBR Green Supermix	Bio-Rad	1725121
Fugene 6 Transfection Reagent	Promega	E2692
Trypsin – Lys-C Mix	Promega	V5073
Sep-Pak C18 Columns	Waters	186000308
0.22 μm Steriflip Filter	Millipore	SEM1M179M6
Amicon Ultra-15 Centrifugal Filter Units – 10 KDa	Millipore	UFC901024
NheI	New England Biolab	R3131S
EcoRI	New England Biolab	R3101S

Highlights.

• SIRT1 expression is frequently decreased in aggressive breast cancers.

• Reducing SIRT1 levels in breast cancer cells impairs lysosomal acidification.

• Loss of SIRT1 gives rise to a secretome that promotes cell invasion and survival.

• SIRT1 mediates these effects by regulating V-ATPase expression.