Chaperone-mediated autophagy is required for tumor growth

Abstract

The cellular process of autophagy (literally "self-eating") is important for maintaining the homeostasis and bioenergetics of mammalian cells. Two of the best-studied mechanisms of autophagy are macroautophagy and chaperone-mediated autophagy (CMA). Changes in macroautophagy activity have been described in cancer cells and in solid tumors, and inhibition of macroautophagy promotes tumorigenesis. Because normal cells respond to inhibition of macroautophagy by up-regulation of the CMA pathway, we aimed to characterize the CMA status in different cancer cells and to determine the contribution of changes in CMA to tumorigenesis. Here, we show consistent up-regulation of CMA in different types of cancer cells regardless of the status of macroautophagy. We also demonstrate an increase in CMA components in human cancers of different types and origins. CMA is required for cancer cell proliferation in vitro because it contributes to the maintenance of the metabolic alterations characteristic of malignant cells. Using human lung cancer xenografts in mice, we confirmed the CMA dependence of cancer cells in vivo. Inhibition of CMA delays xenograft tumor growth, reduces the number of cancer metastases, and induces regression of existing human lung cancer xenografts in mice. The fact that similar manipulations of CMA also reduce tumor growth of two different melanoma cell lines suggests that targeting this autophagic pathway may have broad antitumorigenic potential.

Fig. 1

CMA is up-regulated in cancer cell lines and in human tumors. (A) CMA activity in nontumor and tumor cell lines transfected with a photoactivatable KFERQ-PA-mCherry1 fluorescent reporter. Left: standard pattern in resting or CMA up-regulated (serum−) NIH 3T3 cells and in cancer cells. Right: average number of fluorescent puncta per cell (n = 4 to 6; *P = 0.012, 0.007, 0.004, 0.0036, and 0.002, t test). Serum removal in cultured NIH 3T3 cells (gray bar) is used as a positive control of CMA induction in nontumor cells. (B) Uptake and degradation of radiolabeled proteins by lysosomes in two human lung cancer cell lines and in mouse fibroblasts (NIH 3T3) maintained in the presence or absence of serum and in liver cells from fed mice or mice starved for 48 hours. Values are percentage of proteolysis and are means ± SE (n = 3 to 6; *P = 0.005 and 0.0047 compared to serum+ and 0.002 compared to NIH 3T3 cells, t test). (C) CMA activity in MEFs untransfected (primary) or after transfection with SV40 T antigen determined using the photoswitchable KFERQ-PS-CFP2 reporter. Left: quantification of the percentage of lysosomes positive for the reporter (mean ± SE of >50 cells; *P = 0.006, t test). Right: representative fields. Individual channels for the reporter, LAMP-1, and the merge images are shown. (D) Immunostaining for LAMP-2A of human tumors from the indicated organs. Top: example of LAMP-2A staining in normal tissue (top) and tumor (bottom). Liver samples were observed by immunofluorescence and the rest of the samples by immunohistochemistry. Bottom: quantification of the total LAMP-2A–positive area in normal and tumor samples expressed as fold staining in normal tissue after correction by tissue area. Values are means ± SE of 3 to 15 different tumors (*P = 0.036, 0.0001, and 0.002, t test). (E) Immunostaining for LAMP-2A and LAMP-2B of skin and lung tumors and normal tissue. Left: examples of staining (inset shows higher magnification). Right: quantification of the staining for each protein. Values are means ± SE for 10 and 11 different skin and lung tumors, respectively (*P = 0.002 and 0.016, t test).

Fig. 2

Blockade of CMA in lung cancer cells reduces cell proliferation and increases cell death. (A) Immunoblot for LAMP-2A (L2A) and actin of two human lung cancer cell lines either untreated [control (ctr)] or treated with RNAi to knock down LAMP-2A [L2A(−)]. (B) Fluorescent staining of the CMA reporter in the same cells. (C and D) Incorporation of BrdU (C) and clonogenicity (D) of human lung cancer cells with LAMP-2A knocked down or untreated [n = 6; *P = 0.038 and 0.048 in (C) and 0.043 and 0.036 in (D), t test]. (E) Staining with ethidium bromide/acridine orange of the same cells. Left: representative image. Right: quantification of the number of cells with orange nuclei (late apoptotic cells) per field. Values are means ± SE of four fields for three independent experiments (*P = 0.083, t test). (F and G) Staining for annexin V–PE or annexin V–7ADD in these cells quantified by FACS. Values in (F) indicate the percentage of cells positive for annexin V in each group, and in (G) the percentage of single- and double-stained cells (n = 3). (H) Incorporation of BrdU in the indicated cells treated with the pan-caspase inhibitor ZVAD or untreated (n = 4). (I) Immunoblots for the indicated stress-related proteins in human cancer cell lines: control (ctr) or after LAMP-2A knockdown [L2A(−)]. (J) β-Galactosidase staining of control and L2A(−) A549 and H460 cells. Staining in IMR-90 primary human lung fibroblasts after 45 population doublings (PDL45) is shown as a positive control. (K) Cell cycle distribution of cells labeled with PI and analyzed by FACS. Percentage of cells in each phase (n = 3). (L) Doubling time of control and L2A(−) A549 and H460 cells (n = 4; *P = 0.0453 and 0.0461, t test).

Fig. 3

Blockade of CMA in human cancer cells induces changes in cellular metabolism that limit cell proliferation and induce cell death. (A) Concentrations of oxidized proteins in human lung cancer cell lines {control (ctr) or after LAMP-2A knockdown by RNAi [L2A(−)]} detected by the presence of carbonyl groups. (B) Pattern of polyubiquitinated proteins in the same cells. Stacking of the gel is shown to visualize possible protein aggregates that do not enter the gel. (C) Immunoblot for the indicated effectors and regulators of macroautophagy in the same cells. (D) Autophagic flux in the same cells was determined as the increase in LC3-II levels upon blockade of lysosomal hydrolysis with protease inhibitors (P.I.). Top: representative immunoblots. Increased levels of p53 in the knockdown cells are shown as a reference. Bottom: quantification of steady-state levels of LC3-II (left) and autophagic flux (right) defined as the ratio of LC3-II in cells treated with PI or untreated (n = 4). (E) Concentrations of intracellular ATP (left) and of lactate (right) released in the culture media of H460 human cancer cell lines, either control or cells in which LAMP-2A was knocked down by RNAi [L2A(−)] (n = 4; *P = 0.047 and 0.001, t test). (F) ECAR in A549 and H460 cells, control (ctr) or with LAMP-2A knocked down [L2A(−)]. Left: representative ECAR plots. Where indicated, rotenone was added to the culture media to block mitochondrial oxidative phosphorylation. Right: mean values of ECAR before rotenone addition (basal acidification, left) and average increase of ECAR values after rotenone addition (inducible acidification, right) (n = 6; *P = 0.0002, 0.0001, 0.00008, and 0.0001, t test). (G) OCR in the same cells. Left: representative OCR plots. Right: mean values of OCR before addition of rotenone, and percentage of OCR that is sensitive to rotenone (mitochondrial oxidative phosphorylation) (n = 6; *P = 0.0020 and 0.0066, t test). (H) Rates of β-oxidation of radiolabeled fatty acids in the same cells (n = 3; *P = 0.0001, t test).

Fig. 4

Abnormally increased p53 is responsible for reduced glycolytic metabolism and compromised proliferative capability after CMA blockade in cancer cells. (A) ECAR in A549 and H460 cells, control (ctr) or with LAMP-2A knocked down [L2A(−)], when maintained in complete media (None), media low in glucose (Low G), or depleted of glutamine [Glut(−)] (n = 4; *P = 0.0001, 0.0001, 0.0074, and 0.0007, t test). (B and C) BrdU incorporation in control and L2A(−) cells maintained in complete media alone (none) or supplemented with MPV (B) or the indicated concentrations of adenosine (C) (n = 3 to 5; *P = 0.003 and 0.002, t test). (D) Immunoblots for the indicated glycolytic enzymes in control and L2A(−) human lung cancer cell lines. Left: representative immunoblots. Right: protein concentrations expressed as fold increase over control cells (n = 3 to 5). (E) mRNA levels of the indicated glycolytic enzymes in the same cells. Values are expressed as fold increase over values in control cells after normalization to β-actin. LAMP-2A mRNA levels are shown as reference. (F) Immunoblot for p53 in control and L2A(−) H460 after RNAi knockdown of p53. (G) Mean basal ECAR (top) and OCR (bottom) in the same cells (n = 8; *P = 0.0001 and 0.0001, t test). (H) Changes in mRNA levels for the indicated proteins in L2A(−) cells after p53 knockdown (n = 3). (I) Immunoblot for the indicated proteins in control and L2A(−) cells, either control or with p53 knocked down. Left: representative immunoblot. Right: quantification of the changes in protein levels relative to those in control cells (n = 3). (J) BrdU incorporation in the same cells (n = 3; *P = 0.015, t test). (K) Mean ECAR values in control and L2A(−) A549 cells untreated (none) or treated with pifithrin α (PFTa) (n = 4; *P = 0.0001, t test). (L) BrdU incorporation in the same cells (n = 3; *P = 0.002, t test).

Fig. 5

Blockade of CMA reduces tumor formation in xenograft mouse models and induces shrinkage of preexisting tumors. (A to D) Two human lung cancer cell lines, control (ctr) or with LAMP-2A knocked down [L2A(−)], were subcutaneously injected into nude mice, and their ability to form tumors was monitored. Rates of tumor growth of explants from A549 (A) and H460 (B) cell lines. Insets show immunoblots for LAMP-2A (L2A) in the tumors at the time of resection, and micrographs show examples of immunostaining for L2A in tumors from control cells and cells with LAMP-2A knocked down (n = 6 to 9 mice; *P = 0.0001 to 0.0059 as labeled, t test). Sections of individual tumors of each cell type were stained for H&E (C), TUNEL, Ki-67, CD34, and AIF-1 (D). Double-capped lines mark the area of debris from the border of the tumor. (E to H) Tumors grown from xenografts of human lung cancer cells of sizes up to 100 to 200 mm³ were directly injected with lentivirus carrying shRNA against LAMP-2A on 2 consecutive days. (E) Mean values of rates of tumor growth relative to the size of the tumor at first injection (n = 3 to 7 mice; *P = 0.0001 to 0.032 as labeled, t test). Arrows indicate time of injection. (F) H&E staining of sections of tumors from H460 human lung cancer cell xenografts unmodified (ctr) or after injection with the shRNA against LAMP-2A. Double-capped lines mark the area of debris from the border of the tumor. Staining for TUNEL (G) and immunostaining for Ki-67 (H) in the same tumors are shown. Arrows indicate positive cells.

Fig. 6

Inhibition of CMA reduces formation of human lung tumor metastases. (A) Nude mice were injected in the foot fat pad with A549 and H460 cells, either control or after LAMP-2A knockdown [L2A(−)]. The number of human cancer cells (detected by FACS analysis as GFP fluorescent cells) (top left) and the average number of metastases per lung section (bottom left) was quantified (n = 5 to 6 mice; *P = 0.0001, 0.014, and 0.004, t test). Right: representative H&E-stained sections of lungs. Arrows indicate cancer cell foci. (B to D) Nude mice were injected via the tail vein with H460 human lung cancer cells, either control or with LAMP-2A knocked down, and the lungs were subjected to treatment with collagenase to promote cell dissociation. The number of human cancer cells (detected as GFP fluorescent cells) was determined by FACS. Representative sorting plots of lungs from three to four different mice injected or uninjected with the different cell types (B) and mean values of the percentage of cancer cells in the total amount of cells sorted (C) (n = 3 to 5 animals). (D) Quantification of the number of lung metastases per mouse in sections (n = 3 to 5 mice; *P = 0.010 and 0.009, t test). (E) H&E-stained sections of the lungs from the same animals. Arrows mark areas of metastatic lesions. (F) Time course of wound closing in human lung cancer cells, either control or with LAMP-2A knocked down. Top: representative images of the size of the wound at different times. Bottom: quantification of the size of the wound at 24 hours. Values are expressed as percentage of the initial wound remaining (n = 3 to 4; *P = 0.017 to 0.005, t test). (G) Migration of human lung cancer cells, either control or with LAMP-2A and Atg7 knocked down, in the Transwell migration assay. Left: representative images of cells detected on the bottom side of the filter. Right: quantification of the number of cells detected per field (n = 3; *P = 0.006, 0.0008, and 0.027, t test). (H) Percentage of dead cells detected after preventing attachment of control and L2A(−) human lung cancer cells. Values are expressed as percentage of cells seeded (n = 3; *P = 0.008 to 0.005, t test).