Doxorubicin (DOX) is an effective anti-cancer agent. However, DOX treatment increases patient susceptibility to dilated cardiomyopathy. DOX predisposes cardiomyocytes to insult by suppressing mitochondrial energy metabolism, altering calcium flux, and disrupting proteolysis and proteostasis. Prior studies have assessed the role of macroautophagy in DOX cardiotoxicity; however, limited studies have examined whether DOX mediates cardiac injury through dysfunctions in inter- and/or intra-lysosomal signaling events. Lysosomal signaling and function is governed by transcription factor EB (TFEB). In the present study, we hypothesized that DOX caused myocyte injury by impairing lysosomal function and signaling through negative regulation of TFEB. Indeed, we found that DOX repressed cellular TFEB expression, which was associated with impaired cathepsin proteolytic activity across in vivo, ex vivo, and in vitro models of DOX cardiotoxicity. Furthermore, we observed that loss of TFEB was associated with reduction in macroautophagy protein expression, inhibition of autophagic flux, impairments in lysosomal cathepsin B activity, and activation of cell death. Restoration and/or activation of TFEB in DOX-treated cardiomyocytes prevented DOX-induced suppression of cathepsin B activity, reduced DOX-mediated reactive oxygen species (ROS) overproduction, attenuated activation of caspase-3, and improved cellular viability. Collectively, loss of TFEB inhibits lysosomal autophagy, rendering cardiomyocytes susceptible to DOX-induced proteotoxicity and injury. Our data reveal a novel mechanism wherein DOX primes cardiomyocytes for cell death by depleting cellular TFEB.
Doxorubicin (DOX) is an anthracycline used clinically as an effective cancer chemotherapeutic [1,2]. A significant number of patients receiving anthracycline treatment experience cardiotoxicity in the form of dilated cardiomyopathy [1,2]. DOX-induced cardiotoxicity is often associated with endoplasmic reticulum (ER) stress, mitochondrial dysfunction, calcium signaling abnormalities, and ensuing up-regulation of apoptotic signaling [3–5]. Recent studies have demonstrated that impairment in cellular protein degradation pathways at the lysosomal level can directly contribute to perturbations in ER and mitochondrial function [6,7]. However, studies on the role of proteolytic disruptions in DOX-mediated cardiotoxicity have presented inconsistent findings [8–10]. Nevertheless, a recent study elegantly demonstrated that the adverse effects of DOX in cardiomyocytes involved disruption of lysosomal function . This study suggested for the first time that DOX mediates its detrimental effects on cardiomyocytes by disrupting lysosomal physiology .
Autophagy pathways culminate within the lysosome and are responsible for the degradation of misfolded/damaged proteins [11–13]. Lysosomal autophagy occurs by the process of macroautophagy or chaperone-mediated autophagy (CMA) [14–16]. Macroautophagy involves the de novo synthesis of a double-membrane-bound autophagosome, which is initiated and propagated by unc-51-like kinase 1 (ULK1) and Beclin1 [17,18]. The developing autophagosome utilizes polyubiquitin cargo receptor p62/sequestosome 1 (SQSTM1) to engage in partial selection of bulk protein load and organelle content [17,19,20]. Proper autophagosomal function and lysosomal interaction requires lipidation of microtubule-associated protein light chain 3B (LC3B)-I to LC3B-II, which aids in autophagosomal maturation and lysosomal fusion [18,21,22]. On the other hand, CMA involves the specific degradation of intracellular protein cargo containing a KFERQ consensus sequence via utilization of a chaperone complex containing heat-shock cognate 70 (Hsc70) and heat-shock protein 90 (Hsp90). Transported chaperone cargo is then translocated to the lysosomal lumen via the rate-limiting CMA receptor, lysosome-associated membrane protein 2A (LAMP-2A) [11,20,23–25].
Autophagosomal processing and lysosomal clearance is mediated by transcriptional control. Prior studies have identified transcription factor EB (TFEB) as the master regulator of lysosomal biogenesis and function [23,26]. Indeed, TFEB-mediated action accelerates and enhances lysosomal clearance of protein cargo by co-ordinating encoding of genes critically involved in macroautophagy and lysosomal function [24,26–28]. Therefore, TFEB-mediated action represents a critical signaling hub for proper autophagosomal processing, autophagosome–lysosome fusion, and lysosomal integrity and function [27,29].
Prior studies examining the role of macroautophagy in DOX cardiotoxicity have reported conflicting data [3,10,30,31]. It has been reported that DOX attenuates macroautophagy by increasing the LC3B-II/I ratio and increasing p62 (also known as SQSTM 1) levels in H9C2 cardiomyoblasts . Conversely, others have reported that DOX augments autophagy [30,32,33]. Notably, a recent groundbreaking work has reported that DOX blocks autophagic flux by inhibiting lysosomal acidification . However, to date, limited data are available on DOX's effect on CMA and specifically on the master regulator of lysosomal biogenesis and function, TFEB. Since TFEB action augments lysosomal content and function [26,27,29], it is plausible that DOX-induced impairments in transcriptional regulation of autophagic processes leads to proteotoxic stress and cell death. Indeed, during ischemia/reperfusion injury and during liver aging, deficits in TFEB function are associated with exacerbation of pathology [34–36]. Furthermore, it has been elegantly demonstrated that amyloidic hearts exhibit deficits in TFEB content and that rescuing TFEB and/or restoring TFEB action attenuates amyloidogenic heart failure and ensuing contractile dysfunction . However, it still remains to be investigated whether DOX-induced aberrations in lysosomal autophagy are (1) an outcome of defective TFEB action, and (2) whether these aberrations precede or follow cell death.
In the present study, we examined whether DOX targets TFEB in the cardiomyocyte, leading to disruption of lysosomal autophagy and function. By utilizing in vitro, ex vivo, and in vivo models of DOX cardiotoxicity, we demonstrated that (1) DOX rapidly depletes myocyte TFEB to suppress lysosomal proteolysis, (2) DOX inhibits macroautophagy and autophagic flux with a concomitant decline in cellular viability, and (3) DOX-induced cleaved caspase-3 expression is attenuated by TFEB overexpression in the cardiomyocyte, leading to improved cellular viability. We report a novel mechanism by which DOX negatively targets the transcriptional regulator of lysosomal function, TFEB, to render cardiomyocytes susceptible to cellular injury and death.
Materials and methods
DOX (D00158336), phosphatase inhibitor (524628), and Na4P2O7 (567540) were purchased from Calbiochem. Chloroquine (C6628), Dulbecco's modified Eagle's medium (DMEM)/nutrient mixture Ham's F-12 medium (DMEM/Ham's F12; D6421), protease inhibitor (P8340), cytosine-β-d-arabinofuranoside (ARAC; C1768), N-Suc-Leu-Leu-Val-Tyr-7-AMC (S6510), laminin (11243217001; Roche), bovine serum albumin (BSA fraction V, A7030), 2,3-butanedione monoxime (BDM; B0753), adenosine 5′-triphosphate disodium salt hydrate (ATP; A2383), Media 199 (M5017), streptomycin (S9137), penicillin (P3032), taurine (T0625), and minimum essential medium Eagle with Hanks' salts (MEM/HBS; M1018) were purchased from Sigma (MA, USA). Insulin/transferrrin/selenium (ITS; 25800-CR), penicillin/streptomycin (30-002-CI), gentamycin (30-005-CR), Dulbecco's phosphate-buffered saline (20-031-CV) and Primaria cell culture plates (353801) were purchased from Corning (NY, USA). Torin-1 (4247) was purchased from Tocris. Sprague–Dawley rats were acquired from Charles River Laboratories (Quebec, Canada). C57BL6J mice were acquired from Jackson Laboratories (ME, USA). DMEM-HG (SH30243.01) was purchased from Hyclone Laboratories (UT, USA). DMEM 1× (11966025), Earle's balanced saline solution (EBSS; 14155-063), HEPES (15630-080), Pierce BCA protein assay kit (23255), Pierce MemCode reversible protein stain (24580), glutamine (Glutamax; 35050-061), lipofectamine and P3000 from Lipofectamine 3000 Transfection Reagent kit (L3000008), Hoescht 33342 (62249) and Opti-MEM media (22600) were purchased from Thermo Fisher Scientific (MA, USA). dithiothreitol (DTT; 0281), and Brij 35 solution (0281) were purchased from Amresco (OH, USA). 7-Amino-4-methylcoumarin (26093-31-2) and human cathepsin B (BML-SE198-0025) were purchased from Enzo Life Sciences (NY, USA). Fetal bovine serum (FBS; 1400-500) was purchased from Seradigm. Collagenase-type 2, trypsin and deoxyribonuclease (DNase) were purchased from Worthington Biochemical Corporation (NJ, USA). Quick Start Bradford Assay kit (500-0202) was purchased from Bio-Rad Laboratories (CA, USA). Western Lightning Plus-ECL enhanced chemiluminescence substrate (NELI05001EA) was purchased from Perkin Elmer (MA, USA).
Antibodies used for immunoblotting include anti-TFEB (A303-673a-T) purchased from Bethyl Laboratories; anti-cleaved caspase-3 (9664), anti-phospho-JNK Thr183/Tyr185 (4688), anti-ULK1 (8054), anti-Beclin (3495), anti-LC3B (2775), anti-phospho-mTOR Ser2448 (2971), anti-mTOR (2972), anti-phospho-p70S6K Thr389 (9234), anti-p70S6K (2708), anti-pS6 Ser240/244 (2215), anti-total S6 (2217), and anti-Lamin (2032) were purchased from Cell Signaling Technology (MA, USA); anti-p62 (03-GP62-C) was purchased from American Research Products (MA, USA); anti-pan actin (sc-1616), anti-total JNK (sc-571), anti-v-ATPase (sc-166218), and anti-hemagglutinin (HA) tag (sc-805-G) were purchased from Santa Cruz Biotechnology; anti-α-actin (ab-28052) and anti-LAMP-2A (ab-18528) were purchased from Abcam; anti-Hsp90 (ADI-SPA-830-D) was purchased from Enzo; anti-Ran GTPase (610341) was purchased from BD Biosciences; anti-Myc tag (13-2500) was purchased from Invitrogen; and anti-phospho-TFEB Ser142 was provided as a gift from the laboratory of Dr Gerard Karsenty (Columbia University). Secondary antibodies conjugated to horseradish peroxidase (HRP) were purchased from Santa Cruz Biotechnologies. All primary and secondary antibodies were used at concentrations of 1:1000–2000.
Buffers and media
The following buffers were used: lysis buffer [20 mM Tris/HCl, pH 7.4, 5 mM EDTA, 10 mM Na4P2O7, 100 mM NaF, 1% Nonidet P-40, 2 mM Na3VO4, protease inhibitor (10 µl/ml), and phosphatase inhibitor (10 µl/ml)]; nuclear extraction buffer [1 mM EGTA, 1 mM EDTA, 10 mM HEPES, glycerol 1.25 g/l, 412 mM NaCl, 1.5 mM MgCl2, 1 mM DTT, protease inhibitor and phosphatase inhibitor at pH 7.5]; l-cysteine activation buffer (concentration 8.0 mM; containing 352 mM potassium phosphate monobasic, 48 mM sodium phosphate dibasic, and 4.0 mM EDTA constituted at 40°C); Tyrode buffer for adult rat cardiomyocytes (ARCM) [11.69 mM NaCl, 1.49 mM KCl, 0.33 mM KH2PO4, 1.206 mM MgSO4, 12.51 mM taurine, 3.604 mM dextrose, 4.766 mM HEPES, and 0.396 mM l-carnitine]; mouse perfusion buffer [113 mM NaCl, 4.7 mM KCl, 0.6 mM KH2PO4, 0.6 mM Na2HPO4, 1.2 mM MgSO4·7H2O, 12 mM NaHCO3, 10 mM KHCO3, 10 mM HEPES, 30 mM taurine, 10 mM BDM, and 5.5 mM glucose]; plating medium for adult rat cardiomyocytes (0.987 g/100 ml Media 199, 26.2 mM NaHCO3, 25 mM HEPES, 1.24 mM l-carnitine, 137 µM streptomycin, 280.6 μM penicillin, 10 mM taurine, and 1% fraction V at pH 7.4); DMEM/Ham's F-12 growth media for neonatal rat cardiomyocytes (NRCM) [10% FBS, 10 µM ARAC, ITS and 1% penicillin-streptomycin and 0.05 mg−1 gentamycin]; serum free DMEM 1× media for NRCM [10% FBS, 10 µM ARAC, ITS and 1% penicillin-streptomycin and 0.05 mg−1 gentamycin]; adult mouse cardiomyocyte (AMCM) plating media [535 mg/50 ml, 10% FBS, 10 mM BDM, 100 U/ml penicillin-streptomycin, 2 mM glutamine and 2 mM ATP]; and AMCM culture media [535 mg/50 ml, 0.1% BSA, 10 mM BDM, 100 U/ml penicillin-streptomycin, 2 mM glutamine and 2 mM ATP].
In vivo DOX treatment
Sprague–Dawley rats were injected subcutaneously with either saline or DOX (10 mg/kg) suspended in saline twice daily (total of five doses). Control groups (n = 3) were given the same doses with saline alone (total of five doses). After the treatment regimen, animals were euthanized and hearts were excised and snap-frozen in liquid nitrogen and stored at −80°C for future analysis. Experiments with Sprague–Dawley rats were performed in accordance with the Dalhousie University Institutional Animal Care and Use Committee guidelines.
Hearts isolated from adult rats were ground and homogenized in ice-cold lysis buffer. Tissue homogenate was then centrifuged at 1200 g for 20 min at 4°C to pellet tissue debris and supernatant (total lysate) was stored at −80°C for subsequent biochemical analyses.
Cardiomyocyte isolation and culturing
Neonatal rat ventricular cardiomyocytes
Neonatal rat ventricular cardiomyocytes (NRCMs) were isolated from 1- to 2-day-old Sprague–Dawley rat pups. Briefly, hearts were excised, and ventricles were isolated and subjected to proteolytic digestion with collagenase type 2 (2%, w/v), DNase (0.5%, w/v), and trypsin (2%, w/v) with gentle agitation at 37°C to dissociate heart pieces into single cells. Following digestion, fibroblasts were excluded with a differential plating method by plating the cell suspension in a T75 flask for 2 h. Non-adherent cells were collected and suspended in DMEM/Ham's F-12 growth medium at a density of 1.1 × 106 cells/plate in Primaria cell culture plates. Next day, media was changed to fresh DMEM/Ham's F-12 growth medium and cells were adenovirally transduced as described in the figure legend. After 24 h of viral transduction, cells were washed with DMEM 1× medium, treated with either DOX or vehicle (VEH) in serum free DMEM 1× medium at indicated times as mentioned in the figure legend and cells were harvested for biochemical analysis. For non-viral studies, after 24 h of cell plating, DMEM/Ham's F-12 growth medium was changed to serum free DMEM 1× medium for 24 h, followed by treatment with either DOX or VEH at indicated times as mentioned in the figure legend and cells were harvested for biochemical analysis.
Adult rat cardiomyocytes
Isolated hearts from female adult Sprague–Dawley rats were immersed in cold Tyrode buffer (pH 7.4). Hearts were perfused retrogradely by the Langendorff technique. Ventricular calcium-tolerant myocytes were prepared by a previously described procedure . Cardiomyocytes were plated on laminin-coated plates in ARCM plating medium at 50–75 × 103 cells/plate and incubated at 37°C. After 4 h of cell plating, the medium was replaced with fresh ARCM plating medium, followed by treatment with either DOX or VEH at indicated times as mentioned in the figure legend and cells were harvested for biochemical analysis.
Adult mouse cardiomyocytes
Isolated hearts from male adult C57BL6 mice were immersed in mouse heart perfusion buffer (pH 7.4). Hearts were perfused retrogradely by the Langendorff technique. Ventricular calcium-tolerant myocytes were prepared by a previously described procedure . Cardiomyocytes were plated on laminin-coated plates using mouse cardiomyocyte plating media at a final cell density of 40–70 × 103 cells/plate, and incubated at 37°C. After 4 h, the plating medium was changed to fresh mouse cardiomyocyte culture medium, followed by treatment with either DOX or VEH at indicated times as mentioned in the figure legend and cells were harvested for biochemical analysis.
H9C2 rat embryonic cardiomyoblasts
H9C2 rat embryonic cardiomyoblast (CRL-1446; ATCC) cells were cultured at density of 5–6 × 105 in DMEM-HG supplemented with 10% FBS for 48 h. H9C2 cardiomyoblasts were differentiated for 48 h in DMEM 1× supplemented with 0.5% FBS and 5 mM glucose. After 48 h of differentiation, cells were treated as described in the figure legends. Cells were treated with either DOX or VEH at indicated times as mentioned in the figure legend and cells were harvested for biochemical analysis. For viral studies, after 48 h of cell differentiation, cells were virally transduced for the next 48 h as described in figure legend, followed by treatment with either DOX or VEH at indicated times as mentioned in the figure legend and cells were harvested for biochemical analysis.
NRCMs, ARCMs, AMCMs (adult mouse cardiomyocytes), and H9C2 cells were cultured as described above and treated as described in the figure legends. After treatment, cells were rinsed with ice-cold 1× PBS and spun at 10 000 g for 10 min at 4°C to pellet cells, followed by sonication in lysis buffer (pH 7.4). Lysates were centrifuged at 1500 g at 4°C for 15 min to get the supernatant and pellet, referred to as cytosolic fraction and nuclear pellet, respectively. Nuclear pellets were lysed in nuclear extraction buffer followed by centrifugation at 10 000 g for 20 min at 4°C and supernatant (nuclear fraction) was collected. Both cytosolic and nuclear protein concentrations were determined using the BCA protein assay and Bradford method, respectively, with serum albumin as standard. Lysates were stored at −80°C for subsequent biochemical analyses.
For viral overexpression studies, NRCMs and H9C2 cells were adenovirally infected with HA-tagged human TFEB (SKU # ADV-225358, Lot # 20140903T#9; Vector Biolabs), Myc-tagged human phosphorylation-resistant TFEB (S142A; courtesy of plasmid donated from laboratory of Dr. Shawn M. Ferguson, Yale University School of Medicine, CT, USA) and cytomegalovirus (CMV)-GFP (1060, LOT # 20160126T#6; Vector Biolabs) control virus at the multiplicities of infection (MOIs) and incubation times are described in the figure legends.
For plasmid transfection studies, NRCMs and H9C2 cells were transfected with fluorescent monomeric red fluorescent protein (mRFP)-enhanced green fluorescent protein (EGFP) tandem probe ptfLC3 [ptfLC3 was a gift from Tamotsu Yoshimori (21074; Addgene plasmid) ]. Briefly, an opti-MEM media consisting of Lipofectamine and P3000 from plasmid transfection kit were added to each plate. DNA plasmid concentrations and incubation times are described in the figure legends. Nuclei were counterstained with 1 µl of Hoechst 33342 for 7 min.
H9C2 rat cardiomyoblasts were seeded in a 96-well plate at a density of 10 × 103 per well in 200 µl of DMEM-HG + 10% FBS. Cells were differentiated as described above and were treated as described in the figure legends. After treatment, cells were washed with PBS and terminal deoxynucleotidyltransferase dUTP nick-end labelling (TUNEL) was performed in H9C2 cells according to the manufacturer's specifications (G3250; Promega). Cells were treated in triplicate, and five or six cell images were taken per replicate. Nuclei were counterstained with 1 µl of Hoechst 33342 for 7 min. Green fluorescence of apoptotic cells (fluorescein-12-dUTP) was measured at 520 ± 20 nm using ZEISS microscope. Fluorescence intensity was measured using ZEN Pro software.
For LysoTracker experiments, NRCMs and H9C2 cells were incubated with 75 nM LysoTracker Green DND-26 and LysoTracker Red DND-99, respectively, for 1 h after treatments described in the figure legends. Cells were treated in triplicate, and five or six cell images were taken per replicate. Nuclei were counterstained with 1 µl of Hoechst 33342 for 7 min. Cells were imaged at excitation/emission maxima of approximately 504/511 nm for EGFP and 577/590 nm for RFP using ZEISS microscope. Fluorescence intensity was measured using ZEN Pro software.
Reactive oxygen species (ROS) measurements were performed in H9C2 cells utilizing 5 µM MitoSOX Red, Superoxide Indicator for live cell imaging (ThermoFisher Scientific, M36008) for 15 min after viral infection and DOX treatments as described in the figure legends. Cells were treated in triplicate, and five or six cell images were taken per replicate. Nuclei were counterstained with 1 µl of Hoechst 33342 for 7 min. Cells were imaged at excitation/emission maxima of approximately 510/580 nm using ZEISS microscope. Fluorescence intensity was measured using ZEN Pro software.
Quantitative polymerase chain reaction
Briefly, RNA was extracted from H9C2 cells and adult Sprague–Dawley rat hearts (20–25 mg) using RIBOZOL (N580-CA, Amresco) and chloroform (C2432; Sigma). After Qiaxel quantification and integrity measurement, cDNA was produced using qScript cDNA supermix (CA101414-104; Quanta Biosciences) from 1 µg of RNA. 2 µl of cDNA was mixed with SYBR Green Low ROX PCR supermix (AB1323A; Thermo Fisher Scientific), rat TFEB forward primer 5′–3′ (GTCCAGCAGCCACCTGAACGT), and rat TFEB reverse primer 3′–5′ (ACCAGGGAAGCCGTGACCTG), and quantitative PCR was performed in duplicate in a ViAA7 PCR system under conditions recently described . The gene HPRT (encoding hypoxanthine-guanine phosphoribosyltransferase) was amplified collaterally as a reference gene for transcript measurement. qPCR results were determined using Biogazelle qbase+ software and presented as relative quantities.
Cathepsin B activity assay
Cathepsin B activity was measured in cytosolic fraction using fluorogenic substrate N-Suc-Leu-Leu-Val-Tyr-7-AMC. Standard curves were prepared in the range 300–12.5 nM using 7-amino-4-methylcoumarin. Briefly, standards, samples, and cathepsin B control were added into a 96-well plate. Following the addition of standards, samples and cathepsin B, l-cysteine activation buffer was added. The volume of each well was made up to 150 µl by adding 0.1% of Brij 35 solution [0.1% (v/v) of 30% (w/v) Brij solution in water]. Finally, the fluorogenic substrate (final concentration 0.006 mM) was added to samples and control wells, and the plate was mixed using a thermomixer. The intensity of fluorescence was measured in triplicate at an excitation wavelength of 360 nm and emission wavelength of 440 nm at the 0 min time point. The plate was then incubated at 37–40°C and fluorescence was measured again at the 60 min time point. Relative fluorescence units and nanomolar concentrations of liberated 7-AMC were obtained using linear regression of standards. Data are presented as relative fluorescent units (RFU)/min/µg of proteins.
Presto Blue cell viability assay
NRCMs and H9C2 rat cardiomyoblasts were seeded in a 96-well plate at densities of 30 × 103 and 8–10 × 103 cells per well in 200 µl of NRCM growth medium and 200 µl of DMEM-HG + 10% FBS, respectively. H9C2 cells were differentiated as previously described. Following DOX treatment, Presto Blue reagent (A13261; Invitrogen) was added according to the manufacturer's specifications. Plates were read at time points ranging from 1.5 to 16 h in a Synergy H4 plate reader at a wavelength of 570 nm.
Cell and tissue lysates (10–35 µg) were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS/PAGE) and proteins were transferred onto a nitrocellulose membrane to measure protein phosphorylation and protein expression by immunoblotting. Immunoblots were visualized using a Western Lightning Plus-ECL enhanced chemiluminescence substrate. Densitometric analysis was performed using the Image lab v5.0 software (Bio-Rad), and protein expression data were corrected to reversible Coomassie Blue total protein stain.
Results are presented as means ± SEM. Differences between groups were examined for statistical significance by Student's t-test (designated with +) or one-way ANOVA (designated with *) followed by Tukey's post hoc test for multiple comparisons.
DOX suppresses lysosomal proteolysis in cardiomyocytes
Maintenance of cardiomyocyte health depends on proper protein homeostatic mechanisms, of which lysosomal proteolysis is essential and relies on adequate functioning of lysosomal proteases. In ex vivo and in vivo models of DOX cardiotoxicity, we first assessed lysosomal proteolysis by examining cathepsin B activity. We found that cathepsin B proteolytic activity was suppressed by a 16 h 2 µM DOX treatment in NRCMs, AMCMs, and ARCMs (Figure 1A–C). Despite not being statistically significant, a trend of decreased cathepsin B proteolytic activity was observed in the whole heart of Sprague–Dawley rats administered with 10 mg/kg DOX for two consecutive days (Figure 1D). Thus, utilizing ex vivo and in vivo models of DOX cardiotoxicity, we demonstrate that DOX markedly inhibits lysosomal proteolytic processes, finding that they are in agreement with a recently published report .
Cathepsin B proteolytic activity is suppressed by doxorubicin in ex vivo and in vivo models of DOX cardiotoxicity.
DOX rapidly depletes TFEB in cardiomyocytes
Optimal lysosomal proteolytic processing is dependent on adequate lysosomal content and functioning, both of which are transcriptionally governed. TFEB, a member of the microphthalmia-associated transcription factor family of transcription factors, is an indispensable transcriptional regulator of proteins involved in macroautophagy and is critical for lysosomal biogenesis and function [28,29,41]. DOX suppressed TFEB expression in neonatal and primary cardiomyocytes and in the whole heart. NRCMs and AMCMs treated for 16 h with 2 µM DOX exhibited 65% and 50% reductions in TFEB expression, respectively, with concomitant increases in cleaved caspase-3 content (Figure 2A–D). Furthermore, ARCMs treated with 2 µM DOX for 16 h also demonstrated increases in cleaved caspase-3 and a 23% reduction in TFEB content (Figure 2E,F). The DOX-induced TFEB decline observed in cardiomyocytes was also recapitulated in vivo in Sprague–Dawley rats treated with 10 mg/kg DOX twice daily for 2 days (Figure 2G,H). Therefore, using multiple models, our data consistently and convincingly demonstrate that DOX suppresses cardiomyocyte TFEB content, which could precipitate lysosomal dysfunction in DOX-treated myocytes. Next, we verified the temporal pattern of TFEB decline following acute DOX exposure. In H9C2 rat cardiomyofibroblasts, DOX rapidly suppressed cytosolic TFEB content as early as 6 h and continued to deplete TFEB levels over 24 h treatment (Figure 3A,B). TFEB-mediated action is indispensable for regulation of lysosomal proteolysis, which is primarily mediated by cathepsin B in cardiomyocytes . Indeed, TFEB suppression following DOX occurred concomitantly with decreased cathepsin B activity, signifying inhibition of lysosomal proteolysis (Figure 3C). During nutrient sufficiency, TFEB is localized on the lysosome and is restricted in the cytosol following inhibitory phosphorylation by upstream kinases, extracellular-signal-regulated kinase (ERK) and mTOR (mammalian/mechanistic target of rapamycin) [42,43]. However, following starvation, inactivation of mTOR relieves inhibitory phosphorylation of TFEB, allowing nuclear translocation of TFEB to activate the co-ordinated lysosomal expression and regulation (CLEAR) network that encodes genes critical for cargo recognition for proteolysis, autophagosome formation, and fusion with lysosomes, and lysosomal biogenesis and function . Notably, DOX treatment similarly suppressed nuclear TFEB content at 6–24 h of DOX treatment (Figure 3D,E). DOX is widely known to cause nucleotoxic effects through DNA intercalation, resulting in DNA damage . Therefore, we next assessed whether DOX-induced depletion of TFEB protein was an outcome of decreases in TFEB mRNA expression, plausibly indicating transcriptional suppression following DOX treatment. DOX suppressed TFEB mRNA in the DOX-treated rat heart (2-fold decrease) and markedly suppressed TFEB mRNA (14-fold decrease) in H9C2 cardiomyoblasts (Figure 3F). Overall, these findings demonstrate for the first time that DOX rapidly depletes total cellular TFEB content, which is associated with reduced lysosomal proteolysis, plausibly an outcome of reduced lysosomal content and function.
DOX suppresses TFEB expression across ex vivo and in vivo models of DOX cardiotoxicity.
DOX rapidly depletes cytosolic and nuclear TFEB content with concomitant suppression of cathepsin B activity in H9C2 cardiomyofibroblasts.
DOX-induced TFEB suppression is associated with rapid suppression of macroautophagy and autophagic flux
Next, we investigated whether DOX-induced depletion of TFEB was associated with aberrations in content of proteins indispensable for macroautophagy. Formation and maturation of a preautophagosome is executed by the indispensable proteins ULK1 and Beclin1 [17,20]. DOX treatment significantly suppressed ULK1 and Beclin1 protein expression after 6 h by 48 and 15%, respectively, with gradual reductions in ULK1 and Beclin1 protein expressions at 12 and 24 h, respectively (Figure 4A,B). Early suppression of ULK1 and Beclin1 content suggests that DOX rapidly compromises preautophagosomal assembly, resulting in macroautophagy dysfunction. ULK1 and macroautophagy are negatively regulated by mTOR signaling during states of nutritional surplus. Discordantly, mTOR signaling markers were not augmented until chronic time points, suggesting that DOX-induced defects in autophagosome–lysosome interactions occur independently of mTOR (24 h; Supplementary Figure S1). Since suppression of macroautophagy following DOX occurs independently of upstream regulators of macroautophagy, we further assessed downstream modulators of macroautophagy. We specifically examined the expression of cargo receptor p62/SQSTM1, a bona fide autophagosome-binding partner, and non-lipidated (LC3B-I) and the autophagosome-associated lipidated (LC3B-II) forms of LC3B, alterations in which informs the efficiency of the macroautophagy process. LC3B-II resides on the autophagosomal membrane and is a surrogate marker for autophagosome content [45,46]. We observed a significant down-regulation of LC3B-II after 6 h of DOX treatment that persisted through 24 h of treatment (Figure 4A,B). Interestingly, we also observed a notable increase in LC3B-I at earlier time points (2–6 h) followed by a robust decrease in LC3B-I after 24 h of DOX treatment (Figure 4A,B). Furthermore, DOX-induced a significant suppression of cargo receptor P62/SQSTM1 expression as early as 2 h, an effect that persisted throughout the treatment (24 h; Figure 4A,B), suggesting that changes in p62/SQSTM1 preceded changes in LC3B content. Interestingly, unlike prior studies which have reported an inverse relation of p62/SQSTM1 to LC3B-II protein expression , our data highlight a dissociation between changes in p62/SQSTM1 and LC3B-II, since decreases in p62/SQSTM1 were not associated with increases in the LC3B-II/I ratio following DOX treatment. Prior studies suggest that autophagosome accumulation is attributed to the inability of the lysosome to clear autophagosome load . Therefore, we hypothesized that decreases in LC3B-II could be an outcome of impaired autophagosome maturation and/or fusion with the lysosome (Figure 4A,B). To test this hypothesis, cells were incubated with chloroquine (CQ), a lysosomal deacidifier that is known to block autophagosome turnover and cause autophagosome accumulation, in the presence and absence of DOX. CQ (200 µM) treatment elicited a significant accumulation of LC3B-II in the VEH treatment group (fold change: 4.28), an effect that was diminished in the presence of DOX (fold change: 2.99; Figure 4C,D). Diminished accumulation of LC3B-II (Figure 4C,D) signifies a DOX-induced impairment in preautophagosomal assembly, and/or suppression of autophagosome processing, in lieu of disruptions in autophagosomal clearance. In addition to assessing autophagic flux via immunoblot analysis, we also visualized LC3B fluorescence by transfecting H9C2 cells with a tandem reporter plasmid expressing mRFP–EGFP LC3 followed by incubation with DOX for 16 h. Increases in autophagic flux are indicated by an increase in yellow (autophagosomes) and red (autolysosomes) fluorescence. An absence of yellow puncta was observed in DOX-treated cells, signifying a strong decline in autophagosomal content (Figure 5A,B). Concomitantly, DOX-treated cells displayed robust increases in red puncta, indicating augmented autolysosomal content (Figure 5A,B). This finding highlights an impairment not only in early autophagosomal processing but also in autolysosomal clearance, probably an outcome of compromised lysosomal functioning (Figure 5A,B). To verify DOX-induced dysregulation of lysosomal function and content, we performed lysosome imaging in live H9C2 cardiomyoblasts using LysoTracker Red DND-99 dye. We observed a drastic decline in lysosomal content as indicated by a reduction in lysosomal staining in DOX-treated myocytes, suggesting that cardiac lysosomal content and/or function is negatively targeted by DOX (Figure 5C,D), which is in agreement with a previous study . Furthermore, signaling markers of lysosomal CMA such as Hsp90 and LAMP-2A, an index of lysosomal integrity, were also decreased after 24 h DOX treatment, highlighting the extent of lysosomal damage caused by DOX (Supplementary Figure S2). Recent studies have suggested that changes in lysosomal physiology and function can be an outcome of excess ROS signaling . Indeed, using MitoSOX Red superoxide indicator dye in H9C2 cells treated with DOX, we observed potent up-regulation of ROS production, as reported previously [49,50]. Taken together, these data suggest that DOX rapidly impairs TFEB and macroautophagy signaling to disrupt lysosomal acidification, and that prolonged DOX exposure compromises lysosomal integrity to destabilize lysosomal CMA proteins, a mechanism probably involving ROS overproduction.
DOX-induced depletion of TFEB occurs concomitantly with rapid suppression of autophagosomal markers and autophagic flux in H9C2 cardiomyofibroblasts.
DOX-induced suppression of TFEB and macroautophagy machinery are associated with suppression of lysosome staining, autolysosomal accumulation, and augmented ROS production in H9C2 cardiomyofibroblasts.
Time- and concentration-dependent TFEB repression following DOX parallels augmentation of caspase-3 cleavage, activation of stress signaling and suppression of cell viability
We further examined whether DOX's suppression of TFEB and macroautophagy preceded, followed or occurred concomitantly with DOX-induced caspase-3 cleavage and suppression of cellular viability. We observed significant reductions in cell viability commencing at 4–6 h with progressive reductions in cell viability over 24 h of DOX treatment (Figure 6A). Furthermore, increased caspase-3 cleavage commenced after 6 h of DOX treatment (9.9-fold increase) and this effect persisted for 24 h of DOX treatment (Figure 6B,C). Importantly, we also observed an early (2 h) and sustained activation of stress-activated protein kinase c-Jun N-terminal kinase (JNK) 1 and 2, as observed by increased JNK phosphorylation at Thr183 and Tyr185 residues (Figure 6D,E). Overall, our data are in agreement with prior studies [4,5,51], which indicate that DOX acutely suppresses cell viability that coincides with increased activation of caspase-3 and cell stress pathways. Furthermore, our data suggest a strong association between DOX-induced TFEB suppression, cleaved caspase-3 induction, and reduction in cardiomyocyte viability. Indeed, DOX significantly suppressed TFEB protein expression by 25, 36, 73, and 87% at DOX concentrations of 1, 2, 5, and 10 µM, respectively (Figure 7A,B), which coincided with 0.75, 6.77, 13.8, and 13.8-fold increases in cleaved caspase-3 at concentrations of 1, 2, 5, and 10 µM, respectively (Figure 7A,B). Furthermore, cell viability indices showed a 35% reduction in cell viability at 0.5 µM DOX concentration with further decreases in viability with increasing concentration (1–10 µM) and depletion of TFEB content (Figure 7C). Additionally, DOX-induced cell death in H9C2 cells were confirmed using TUNEL assay, which further recapitulates a DOX concentration-dependent apoptotic activation (Figure 7D,E). These data highlight a consistent association of DOX-induced suppression of TFEB and cardiomyocyte viability coinciding with augmentation of cleaved caspase-3 and apoptosis that occurs in a time- and concentration-dependent manner.
DOX time-dependent suppression of TFEB and autophagy signaling markers is associated with rapid suppression of cardiomyoblast viability and augmentation of caspase-3 cleavage and stress signaling.
DOX-induced concentration-dependent repression of TFEB is associated with increases in cleaved caspase-3 expression and suppression of cardiomyoblast viability.
The metabolic phenotype and cell type also dictate the magnitude of DOX toxicity . H9C2 cells are cardiomyofibroblasts, which have a glycolytic phenotype and do not phenocopy the metabolic profile of a cardiomyocyte, which has a balanced usage of glucose and fatty acids . Therefore, we next assessed DOX's temporal affect on TFEB and caspase-3 activation in NRCMs, which exhibit a more cardiac-like phenotype. DOX was found to suppress TFEB content and concomitantly up-regulate cleaved caspase-3 in a time-dependent manner, recapitulating our findings in H9C2 cells (Figure 8A,B). Furthermore, we next examined whether DOX-induced suppression of TFEB was associated with compromised autophagosomal production and clearance by employing mRFP–EGFP LC3 reporter plasmid in NRCMs. Indeed, DOX was found to repress autophagosomal content by the absence of yellow puncta, suggesting compromised autophagosomal assembly, and DOX was also found to potently up-regulate red autolysosomal puncta content, suggesting compromised autolysosomal clearance (Figure 8C,D). Additionally, LysoTracker staining was also found to be suppressed by DOX in NRCMs, recapitulating H9C2 findings of DOX-induced suppression of lysosome content and/or function (Figure 8E).
DOX suppresses TFEB and induces cleavage of caspase-3 in a time-dependent manner in NRCMs with concomitant reduction in lysosomal staining and increases in autolysosomal content.
Genetic restoration and pharmacological activation of TFEB attenuates cleaved caspase-3 expression, improves cardiomyocyte viability, and reduces ROS overload following DOX treatment
To assess whether DOX mediates its toxic effects on cardiomyocytes through negative targeting of TFEB, human TFEB was overexpressed in NRCMs that were subsequently treated with 2 µM DOX for 16 h. Alternatively, to augment nuclear TFEB levels during DOX exposure, we constructed an adenovirus that expresses a phosphorylation-resistant TFEB mutant in which serine residue 142 is mutated to alanine (S142A; courtesy of the laboratory of Dr Shawn Ferguson for providing the plasmid for the viral construct), rendering TFEB constitutively active. Cleaved caspase-3 expression was increased by 85% upon DOX treatment in GFP-infected NRCMs (Figure 9A,B). However, overexpression with wild-type (WT) TFEB and mutant TFEB (S142A) reduced DOX-induced caspase-3 cleavage by 60 and 33%, respectively (Figure 9A,B), when compared with DOX-treated GFP-expressing cells. Notably, DOX's ability to induce cleaved caspase-3 expression was completely ablated in NRCMs overexpressing WT TFEB and phosphorylation-resistant mutant TFEB (Figure 9A,B). Furthermore, overexpression of WT TFEB and S142A TFEB in DOX-treated NRCMs improved cellular viability by 20 and 21%, respectively (Figure 9C). These data further suggest that DOX predisposes cardiomyocytes to insult by negatively targeting TFEB content, rendering the myocyte susceptible to lysosomal impairments and proteolytic injury. It has been recently reported  that DOX inhibits lysosomal acidification to disrupt autophagic flux and cathepsin protease function. TFEB orchestrates and co-ordinates the expression of proteins essential for proper autophagosomal and lysosomal functioning, including the machinery involved in lysosomal acidification and proteolysis [28,54]. Indeed, overexpression of human TFEB and phosphorylation mutant TFEB abolished DOX's ability to suppress cathepsin B activity (Figure 9D). Therefore, these data indicate the critical involvement of TFEB in DOX-induced lysosomal dysfunction, probably contributing to proteolytic insult and caspase cell death. In an alternative model of DOX cardiotoxicity, using H9C2 rat cardiomyoblasts, we genetically restored TFEB content by adenoviral overexpression of WT TFEB and observed a 15% reduction in the cleavage of caspase-3 and a 16% increase in cell viability upon DOX (2 µM) treatment (Figure 10A,B,E). Furthermore, overexpression of constitutively active phosphorylation-resistant TFEB (S142A) likewise attenuated DOX-induced cleaved caspase-3 expression by 19% and improved cardiomyocyte viability by 27% (Figure 10C–E). Recent reports have shown that mitochondrial ROS regulates TFEB through calcineurin-mediated activation . We next assessed whether protection rendered by TFEB in DOX-treated cardiomyocytes is an outcome of restoring ROS homeostasis, since DOX is known to induce ROS overload and deplete antioxidant capabilities . We performed ROS imaging in live H9C2 cells using the mitochondria-targeting fluorogenic dye MitoSOX. DOX-treated cells overexpressing WT TFEB and phospho-mutant TFEB (S142A) demonstrated a 27 and 36% reduction in ROS accumulation, respectively (Figure 10F,G). Since TFEB is a potent transcriptional regulator of lysosomal content and function, we examined whether TFEB increased lysosomal content, indicated by LysoTracker staining. Forced overexpression of WT TFEB and phospho-mutant TFEB (S142A) rescued DOX suppression of lysosomal staining in H9C2 cells, indicating that DOX negatively targets lysosomal content and/or function by depleting myocyte TFEB (Figure 11A–D). Taken together, these data demonstrate that TFEB-mediated improvement of lysosomal function in a setting of DOX is associated with reducing DOX's ability to overproduce ROS, which probably improves mitochondrial function and cellular viability.
Rescue of TFEB prevents DOX-induced suppression of cathepsin activity and caspase-3 cleavage in NRCMs.
Viral augmentation of TFEB attenuates DOX-induced cleaved caspase-3 expression and improves cell viability in DOX-treated rat cardiomyoblasts.
Forced overexpression of WT TFEB and S142A TFEB ameliorates DOX-induced suppression of lysosomal content in H9C2 cells.
Since TFEB mediates autophagic function through transcriptional up-regulation of CLEAR genes, we pharmacologically activated TFEB. Torin-1 inactivates mTORC1 and relieves its inhibition on TFEB, thereby activating TFEB to increase macroautophagy function [42,43]. Torin-1 significantly increased DOX-treated myoblast viability (Figure 12A) and significantly reduced cleaved caspase-3 expression upon DOX treatment (Figure 12B,C). These data recapitulate the potent involvement of TFEB and macroautophagy in DOX-induced cardiac injury. Our data suggest for the first time that impairments in lysosomal functioning following DOX are due to DOX-induced depletion of the cardiomyocyte reservoir of TFEB, promoting lysosomal dysfunction and cardiomyocyte insult through suppression of autophagy (Figure 12D).
Chemical activation of TFEB ameliorates DOX-induced suppression of autophagy, myocyte viability, and caspase-3 cleavage.
DOX is an effective anti-cancer agent; however, prolonged use of DOX results in cardiac failure, limiting its utility. DOX induces cardiac stress by a plethora of mechanisms, including aberrations in calcium handling, ROS-mediated mitochondrial dysfunction, ER stress, cytoskeletal disorganization, loss of nuclear integrity, and, more recently studied, maladaptations in protein degradation processes [3–5,10]. Proteolytic processing maintains cellular health by clearing proteotoxic load, an operation orchestrated largely by autophagy . Previous studies using in vivo and in vitro models have reported that DOX increases and/or decreases autophagy through peri-autophagosomal and lysosomal events [8,10]. These disparate findings from prior studies may be due to a lack of data examining DOX's direct effect on transcriptional regulation of inter- and intra-lysosomal signaling events within the cardiomyocyte. While this manuscript was in review, a recent study for the first time demonstrated that DOX induces cellular stress by profoundly affecting lysosomal acidification . Importantly, the ability of lysosomes to execute proteolysis is governed by TFEB, a transcriptional regulator of lysosome function and content . Therefore, we hypothesized that the underlying mechanism by which DOX causes lysosomal dysfunction involves negative targeting of TFEB, rendering the myocyte susceptible to proteotoxicity and cell death. Using in vivo and in vitro models, our data demonstrate that (1) DOX potently suppresses lysosomal proteolysis and autophagic flux with concomitant up-regulation of cell death pathways, (2) DOX rapidly depletes total cellular TFEB, which is strongly associated with reduced cardiomyocyte viability, apoptotic activation, augmented ROS, dysregulation of autophagy signaling, and malfunctions in lysosomal proteolysis, and (3) rescue of TFEB in DOX-treated cells augments lysosomal proteolysis, decreases ROS overload, attenuates cleaved caspase-3 expression, and improves cardiomyocyte viability. Collectively, our data demonstrate that depleting TFEB is a novel mechanism by which DOX induces lysosomal dysfunction, compromising autophagy, and rendering myocytes susceptible to injury and failure. Furthermore, replenishing cellular TFEB offers resistance to DOX-mediated lysosomal and mitochondrial toxicity.
It was previously demonstrated that autophagosome number and clearance is controlled by basic helix–loop–helix (bHLH)-leucine zipper TFEB, the master regulator of lysosomal biogenesis and function . Across in vitro, ex vivo, and in vivo models of DOX cardiotoxicity, we consistently demonstrate that DOX rapidly suppresses cellular TFEB and lysosomal proteolytic activity with a concomitant decline in cellular viability. The decrease in lysosomal proteolysis coincided with decreased expression of autophagy signaling proteins, including p62/SQSTM1 and LC3B, which are bona fide transcriptional targets of TFEB . Interestingly, we also observed a robust decline in TFEB mRNA levels in DOX-treated H9C2 cardiomyofibroblasts, but less so in hearts from Sprague–Dawley rats treated with DOX, suggesting that DOX's effect on transcriptional suppression of TFEB is cell type- and/or tissue-specific. Furthermore, DOX's ability to repress transcription of TFEB may be an outcome of dysregulation of the recently described FXR (farnesoid X receptor)/CREB (cyclic AMP-response element-binding protein) transcription axis  and/or up-regulation of micro-RNA 128, a potent suppressor of TFEB mRNA .
Lysosomal TFEB translocates to the nucleus and activates the CLEAR network of genes to regulate autophagosome synthesis, clearance, and lysosomal biogenesis. Importantly, we observed not only reductions in cytosolic TFEB expression but also decreased nuclear TFEB content following DOX treatment. Strategies of increasing nuclear TFEB, such as by overexpressing either WT TFEB or constitutively active mutant TFEB (S142A), attenuated DOX-induced declines in lysosomal cathepsin B activity and cell viability. Restoration of nuclear TFEB was sufficient to offer protection against DOX toxicity independent of decreases in nuclear protein lamin, suggesting that, despite loss of nuclear membrane integrity, transcriptional effects of TFEB remain functional in a setting of DOX. Therefore, our data imply that loss of TFEB mediates toxic effects of DOX on lysosomal integrity and cardiomyocyte survival. Interestingly, a recent study demonstrated that DOX blocks cardiomyocyte autophagy by inhibiting acidification of the lysosomal lumen, resulting in impaired proteolysis . However, our study has now demonstrated that loss of TFEB following DOX is sufficient to impair lysosomal proteolysis and macroautophagy.
Since TFEB signaling generates more autophagosomes and accelerates their delivery to lysosomes by increasing lysosomal number, DOX's effect on TFEB is expected to have profound effects on the process of macroautophagy and CMA. DOX is reported to either increase  or decrease  autophagy processes in cardiomyocytes depending on the type of in vivo and/or in vitro models employed. In spite of the disparity in prior findings, autophagy is physiologically important for mediating DOX's cardiotoxic effects since reversing DOX regulation of autophagy attenuates cardiotoxic outcomes [8,57,58]. In our in vitro model of DOX cardiotoxicity, we observed rapid suppression of p62/SQSTM1, a cargo receptor whose levels are a surrogate for predicting the extent of macroautophagy. Interestingly, DOX induced a reduction in p62/SQSTM1 content, which is not consistent with prior reports [8,31]. We speculate that these differences are attributed to changes in DOX concentration, duration, and the type of cell in which these studies were performed. DOX-induced decline in p62/SQSTM1 content also paralleled repression of preautophagosomal assembly markers, ULK1 and Beclin1, suggesting that initial steps of macroautophagy are impaired in a DOX-treated cardiomyocyte milieu. Indeed, DOX's effect on inhibiting initial steps of autophagy also reflected in a significant decline in LC3B lipidation as observed by diminished LC3B-II content, data consistent with prior findings [10,30].
Since a decline in LC3B-II may be an outcome of inhibited autophagosome formation or an acceleration of autophagosomal clearance , autophagic flux was examined by blocking autophagosomal turnover with CQ and LC3B-II accumulation was analyzed . Consistent with prior studies [10,60], we found that DOX suppressed autophagic flux by diminishing accumulation of LC3B-II in response to CQ treatment, indicating a reduction in mature autophagosome content that was also observed microscopically in cells transfected with tandem fluorescent mRFP–EGFP LC3 reporter plasmid. However, we observed an up-regulation of autolysosomes in DOX-treated NRCMs and H9C2 cells, suggesting that DOX impairs clearance of autolysosomes. Our data are in agreement with a prior study which demonstrated that TFEB facilitates autolysosomal recycling via lysosomal exocytosis at the plasma membrane during the process of lysosomal autophagy ; therefore, loss of TFEB following DOX is the underlying mechanism for autolysosomal accumulation in the myocyte. Additionally, impaired autolysosome clearance could be secondary to lysosomal dysfunction . Indeed, our data from NRCMs and H9C2 cells also demonstrate that reduced LysoTracker staining following DOX is probably due to impaired lysosomal acidification and proteolytic activity. Therefore, following DOX exposure, depletion of TFEB probably results in compromised autophagosomal processing, lysosomal dysfunction, and accumulation of autolysosomes. These effects correspond with augmentation of caspase-3 cleavage and stress signaling, suggesting that impairment in transcriptional regulation of lysosomal functioning predisposes cardiomyocytes to cellular stress and demise. Additionally, impaired lysosomal activity could also be an outcome of mitochondrial dysfunction induced by excess ROS production , suggesting that mitochondria are required to maintain lysosomal function. Indeed, in our model of DOX toxicity, lysosomal dysfunction and TFEB depletion were associated with augmented ROS levels. Interestingly, not only can intact mitochondria restore lysosomal function  but also the reverse is possible since restoring TFEB and ensuing lysosomal function attenuated DOX-induced ROS production, suggesting that restoring lysosomal function can attenuate mitochondrial damage.
Transcriptional effects of TFEB are inhibited by phosphorylation via mTOR at Ser211 or phosphorylation via ERK2 at Ser142, leading to nuclear extrusion of TFEB . However, mTOR activation did not occur until later time points (24 h; Supplementary Figure S1) of DOX treatment, implying that mTOR does not play a role in early regulation of autophagy in DOX-treated myocytes. Proteolytic signaling cascades at the lysosomal level are also known to mediate cardiomyocytes' ability to protect against cellular stressors [64,65]. The recently described process of CMA, involving cytosolic protein turnover by concerted actions of chaperone Hsp90 and LAMP-2A receptor, has been demonstrated to be associated with detrimental cardiac health outcomes [15,66]. In our model of DOX cardiotoxicity, CMA signaling markers were suppressed albeit at a delayed time point (Supplementary Figure S2), suggesting that early changes in macroautophagy and TFEB sensitizes the myocyte for a delayed dysregulation of CMA, all of which could exacerbate cardiac functional outcomes during later stages of DOX cardiotoxicity.
In summary, our data demonstrate for the first time that DOX primes cardiomyocytes for lysosomal dysfunction and cell death by depleting TFEB content in vivo, ex vivo, and in vitro, leading to proteotoxic stress and cell death (Figure 12D). Notably, in vivo exposure to DOX triggered an early decrease (48 h) in cardiac TFEB content, suggesting that this mechanism precedes left ventricular cardiomyopathy, which is usually observed following longer duration of DOX treatment . The present study has identified a novel mechanism by which DOX suppresses macroautophagy and lysosomal proteolysis through its effects on TFEB. Indeed, deficits in TFEB action lead to exacerbation of ischemia/reperfusion injury and lipotoxic injury , which was attenuated by replenishing TFEB, leading to lysosomal homeostasis. Notably, our data recapitulate a pathological mechanism similar to that of amyloidogenic cardiomyopathy involving deficits in TFEB-mediated autophagy, highlighting a plausible critical involvement of TFEB in many types of heart disease . Our studies show that TFEB critically regulates intra-lysosomal functioning and this process is targeted by DOX to cause cardiomyocyte injury and death. Since TFEB maintains lysosomal homeostasis and cellular function , inhibiting TFEB loss or activating TFEB in DOX-treated hearts could prevent or mitigate DOX cardiomyopathy.
adult mouse cardiomyocyte
adult rat cardiomyocyte
co-ordinated lysosomal enhancement and regulation
Dulbecco's modified Eagle's medium
Dulbecco's modified Eagle's high-glucose medium
Earle's balanced saline solution
enhanced green fluorescent protein
fetal bovine serum
green fluorescent protein
heat-shock cognate protein 70
heat-shock protein 90
lysosome-associated membrane protein 2A
microtubule-associated light chain 3B
monomeric red fluorescent protein
mammalian/mechanistic target of rapamycin
multiplicity of infection
neonatal rat cardiomyocyte
relative fluorescence units
reactive oxygen species
transcription factor EB
terminal deoxynucleotidyltransferase dUTP nick-end labeling
unc-51-like kinase 1
T.P., J.J.B., and P.C.T. designed the research. J.J.B. and P.C.T. equally contributed to this work. P.Y. provided rodent heart tissues. T.P. and P.C.K. contributed reagents and analytical tools. J.J.B., T.P., and P.C.T. analyzed the data. P.C.K. provided intellectual inputs, technical assistance, and reviewed and proofread the manuscript before submission. J.J.B., T.P., and P.C.T. wrote the paper.
This work was funded by grants from the Natural Sciences and Engineering Research Council of Canada [RGPIN-2014-03687] and the New Brunswick Health Research Foundation to T.P. and P.C.K. J.J.B. was supported by the government of New Brunswick through a New Brunswick Health Research Foundation summer studentship.
The Authors declare that there are no competing interests associated with the manuscript.
T.P. is the guarantor of this work, had full access to all the data, and takes full responsibility for the integrity of data and the accuracy of data analysis.
We thank the New Brunswick Health Research Foundation for granting a summer studentship for J.J.B. and Dalhousie Medicine New Brunswick for granting a graduate fellowship to P.C.T. We kindly thank the laboratories of Dr Gerard Karsenty for providing us with anti-phospho-TFEB (Ser142) antibody. We thank Dr Shawn M. Fergusson for providing us the phosphorylation-resistant mutant TFEB (S142A) plasmid to construct phosphorylation-resistant TFEB (S142A) adenovirus.
These authors contributed equally to this work.