Nanoparticle‐mediated lysosomal reacidification restores mitochondrial turnover and function in β cells under lipotoxicity

ABSTRACT

Chronic exposure of pancreatic β cells to high concentrations of free fatty acids leads to lipotoxicity (LT)‐mediated suppression of glucose‐stimulated insulin secretion. This effect is in part caused by a decline in mitochondrial function as well as by a reduction in lysosomal acidification. Because both mitochondria and lysosomes can alter one another's function, it remains unclear which initiating dysfunction sets off the detrimental cascade of LT, ultimately leading to β‐cell failure. Here, we investigated the effects of restoring lysosomal acidity on mitochondrial function under LT. Our results show that LT induces a dose‐dependent lysosomal alkalization accompanied by an increase in mitochondrial mass. This increase is due to a reduction in mitochondrial turnover as analyzed by MitoTimer, a fluorescent protein for which the emission is regulated by mitochondrial clearance rate. Mitochondrial oxygen consumption rate, citrate synthase activity, and ATP content are all reduced by LT. Restoration of lysosomal acidity using lysosome‐targeted nanoparticles is accompanied by stimulation of mitochondrial turnover as revealed by mitophagy measurements and the recovery of mitochondrial mass. Remarkably, re‐acidification restores citrate synthase activity and ATP content in an insulin secreting β‐cell line (INS‐1). Furthermore, nanoparticle‐mediated lysosomal reacidification rescues mitochondrial maximal respiratory capacity in both INS‐1 cells and primary mouse islets. Therefore, our results indicate that mitochondrial dysfunction is downstream of lysosomal alkalization under lipotoxic conditions and that recovery of lysosomal acidity is sufficient to restore the bioenergetic defects.—Assali, E. A., Shlomo, D., Zeng, J., Taddeo, E. P., Trudeau, K. M., Erion, K. A., Colby, A. H., Grinstaff, M. W., Liesa, M., Las, G., Shirihai, O. S. Nanoparticle‐mediated lysosomal reacidification restores mitochondrial turnover and function in β cells under lipotoxicity. FASEB J. 33, 4154–4165 (2019). www.fasebj.org

ABBREVIATIONS

AVO

acidic vesicular organelle

BSA

bovine serum albumin

CS

citrate synthase

EF1α

elongation factor‐1α

FACS

fluorescence‐activated cell sorting

FBS

fetal bovine serum

FFA

free fatty acid

F.I.

fluorescence intensity

GSIS

glucose‐stimulated insulin secretion

INS‐1

insulin secreting β‐cell line

LAMP‐1

lysosomal‐associated membrane protein 1

LT

lipotoxicity

MTG

MitoTracker green

NP

nanoparticle

paNP

photoactivable acidic nanoparticle

ROS

reactive oxygen species

RPMI

Roswell Park Memorial Institute

vATPase

vacular ATPase

Neurodegenerative and metabolic diseases are associated with decreased mitochondrial respiratory function (1) as well as by a decrease in lysosomal catabolic activity (autophagy) (3–6). Although both dysfunctions are thought to be causally linked, the respective place of mitochondria and lysosomes in mediating this mechanism is challenging to elucidate in view of the interdependence of these organelles. On the one hand, mitochondria can act upstream of lysosomal dysfunction by generating reactive oxygen species (ROS) that can impair autophagy (7) or by reducing ATP production, which is essential for lysosomal vacuolar ATPase (vATPase)‐mediated acidification. On the other hand, accumulation of damaged mitochondria can occur downstream to lysosomal dysfunction within the mitochondrial quality control mechanism, which, as we have previously reported, consists of targeting fragmented depolarized mitochondria to autophagy for degradation (8). Under lysosomal inhibition, the disruption of mitochondrial quality control leads to accumulation of depolarized mitochondria and to an increase in oxidative damage (8–10). In addition to their effect on mitochondrial turnover, lysosomes can affect mitochondria indirectly. For instance, lysosomal alkalization is reported to alter amino acid storage and induce mitochondrial depolarization, thereby impairing mitochondrial function (3).

Mitochondrial respiratory function and lysosomal acidification are reported to be impaired during lipotoxicity (LT) (5). Although the main assumption is that free fatty acids (FFAs) directly suppress cellular bioenergetics and mitochondrial function (11), here we investigate whether impaired autophagic turnover due to lysosomal alkalization plays a key role in the accumulation of dysfunctional mitochondria, thus promoting a decline in ATP production and ultimately promoting β‐cell dysfunction during LT. In order to accomplish this study, we will use photoactivable acidic nanoparticles (paNPs) that enter in the lysosomes via endocytosis and release their acidic content upon exposure to light. Previously, we showed that treatment of primary β cells and an insulin secreting βcell line (INS‐1) with paNPs followed by their photoactivation reversed the LT‐mediated lysosomal alkalization as well as the autophagic flux and rescued the cells from suppression of glucose‐stimulated insulin secretion (GSIS) (12). Here we hypothesize that this recovery is mediated by an improvement in mitochondrial respiratory function and thus bioenergetics. We test this hypothesis by treating INS‐1 cells and mouse islets exposed to LT with paNPs and examining mitochondrial mass and function.

MATERIALS AND METHODS

Experimental animals

Islet isolation was performed on 12–15 wk‐old C57BL/6J male mice. All animal work was conducted under an approved Institutional Animal Care and Use Committee protocol (16‐018) at University of California, Los Angeles. Animals were fed standard chow and kept in normal housing conditions (19–22°C and a 14:10‐h light/dark cycle).

Isolation and culture of mouse islets

Islets were isolated from mice via collagenase P injection in the bile duct, as previously described (13). The islets were cultured overnight at 37°C and 5% CO₂ in islet medium [11 mM glucose Ros well Park Memorial Institute (RPMI) 1640 medium, 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 µg/ ml streptomycin] prior to respirometry experiments.

Cell culture

INS‐1 (832/13) cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 10 mM HEPES buffer, 1 mM pyruvate, 50 µM β‐mercaptoethanol, 50 U/ml penicillin, and 50 µg/ml streptomycin. The cells were used between passages 70 and 90.

Preparation of palmitate complexed to bovine serum albumin

Palmitate was dissolved in DMSO and then gradually added to warm (45°C) RPMI 1640 medium containing 6.7% fatty acid‐free bovine serum albumin (BSA) (MilliporeSigma, Burlington, MA, USA) to make a 4 mM (10 times) stock. For control BSA conditions, a 10× stock consisting of RPMI 1640 medium with 6.7% BSA and 1% DMSO was used. On the day of the experiment, the stocks were diluted with RPMI 1640 medium containing 1% FBS, 50 U/ml penicillin, and 50 µg/ml streptomycin and 10 mM glucose. The pH was then adjusted to 7.4 followed by sterile filtration before the treatment.

Preparation of oleate‐palmitate complexed to BSA

Mouse islets were exposed to a mixture of 200 µM oleate and palmitate for 48 h. Sodium oleate (MilliporeSigma) and sodium palmitate (MilliporeSigma) were complexed to 16.75% fatty‐acid‐ free BSA in a 150 mM NaCl solution at 37°C at a FA:BSA ratio of 2:1 (1:1 ratio of oleate:palmitate) to produce a 10 mM stock solution (50×). The pH was then adjusted to 7.4 using NaOH, then filtered and stored at −80°C. For the treatment conditions, the 10 × stocks were added to RPMI 1640 medium containing 1% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin and glucose at 10 mM.

Nanoparticle treatment

Synthesis and characterization of acidic nanoparticles (NPs) was done as previously described (12). INS‐1 cells were loaded with NPs at a concentration of 25 µg/ml for 18–24 h together with BSA or palmitate. Photoactivation was performed by exposing the cells for 5 min to a compact UV lamp (UVL‐21 Blak Ray Lamp; Analytic Jena, Upland, CA, USA) at 365 ran, 115 V. Cells and NP stocks were kept in the dark as much as possible to prevent ambient light exposure.

To determine if restoration of lysosomal acidity using the NPs would affect mitochondrial bioenergetics of mouse islets under lipotoxic conditions, the NP dose for mouse islets was calibrated in INS‐1 cell experiments by calculating NP dose as micrograms per cell (instead of micrograms per milliliter), because this is a more accurate indicator of NP exposure per cell, similarly to that reported before (12). We determined that the lowest effective nontoxic dose for 30 islets is 2 µg/ml. This concentration was applied thereafter in the islet assays.

Determination of cell viability

Cells at a density of 2 × 10⁵ were seeded in 12 well plates and then exposed to palmitate or BSA for 24 h. The cells were trypsinized and stained with 0.4% trypan blue, and cellular viability was determined by Countess II Automated Cell Counter (Thermo Fisher Scientific, Waltham, MA, USA).

High‐throughput imaging

Cells expressing MitoTimer were imaged on PerkinElmer Operetta high‐content wide‐field fluorescence imaging system, coupled to Columbus software (Operetta; PerkinElmer, Waltham, MA, USA). We used a ×40 numerical aperture objective lens in a single focal plane across each plate. For the long‐term imaging experiments, the cells were kept under a live‐cell chamber controlling a steady condition of 37°C and 5% CO₂ levels. The bottom of each well was detected automatically by the Operetta focusing laser, and the focal plane was calculated relative to this value. Lysosomal‐associated membrane protein 1 (LAMP‐1) and green MitoTimer were imaged under 460–490 nm excitation and 520–550 nm emission; red MitoTimer was imaged under 500–550 nm excitation and 560–630 nm emission. A total of 15 fields of view were taken per well, with an identical pattern of fields used for every well.

Modified Columbus (PerkinElmer) image analysis software was used to calculate intensity of each probe.

Flow cytometry

Fifty thousand INS‐1 cells per condition were stained for 40 min with 200 nM MitoTracker green (MTG). The cells were washed 3 times and analyzed using Fluorescence‐Activated Cell Sorting (FACS) Canto II (BD Biosciences, San Jose, CA, USA). The data from the FACS analysis were exported as FCS 3.0 files and further analyzed using the De Novo software (De Novo Software, Glendale, CA, USA). For MTG measurement, the 488 ran channel was used, and mean fluorescence intensity (F.I.) was calculated for each condition.

FACS was also used for sorting INS‐1 cell lines expressing MitoTimer. INS‐1 cells were infected with lentivirus encoding MitoTimer. The cells expressing MitoTimer were then sorted using the 488 nm channel according to their level of expression.

Creation of stable cell line expressing constitutive MitoTimer

Lentivirus containing the MitoTimer reporter (15) was generated by cloning the reporter cassette into a pHAGE2 lentivector, with a constitutive elongation factor‐1α (EF1α) promoter and followed by an IRES‐Puro cassette enabling positive selection. Lentiviral particles were generated in human embryonic kidney 293T cells by transfection of the pHAGE2 plasmid together with 4 packaging plasmids (Rev1, Tat1, Hgpm2, and VSV‐G). The virions were collected, and INS‐1 cells were infected for 48 h in the presence of 3 µg/ml polybrene (MilliporeSigma). The cells were sorted according to F.I. into medium‐ and high‐expressing populations with a BD FACSAria III cell sorter (BD Biosciences).

mCherry‐GFP‐Fis1 probe

The mCherry‐GFP‐Fis1 (101–152) construct was a generous gift from Ian Ganley (16). The plasmid was packaged into adenoviral particles under cytomegalovirus promoter (Welgen, Worcester, MA, USA) at a concentration of 1 × 10¹² vp/ml. INS‐1 cells were transduced with the probe at a multiplicity of infection of 300 and imaged after 72 h.

ATP detection

Three hundred thousand cells per condition were plated in 6 well plates. At the end of the treatment the cells were washed twice with cold PBS and resuspended in ATP Assay Buffer according to the manufacturer's instructions (Abcam, Cambridge, United Kingdom). ATP levels were determined with an ATP kit (Thermo Fisher Scientific) according to the manufacturer's instructions. The results were normalized to protein concentration after the Pierce BCA Protein Assay (Thermo Fisher Scientific).

Fluorescent dyes

Cells were plated on glass‐bottomed plates 72 h prior to imaging. Before imaging, the cells were loaded with 200 nM MTG (Thermo Fisher Scientific) for 30 min followed by washout.

For lysosomal and nuclear staining, the cells were stained for 30 min before imaging with 50 nM LysoTracker deep red (Thermo Fisher Scientific) and with 2 µg/ml Hoechst (MilliporeSigma). The imaging was performed with an Operetta high‐throughput imaging system.

Western blotting

Cells were lysed in RIPA buffer [50 mM Tris‐HCl, pH 7.5, 0.1% (w/v) Triton X‐100, 1 mM EDTA, 1 mM EGTA, 50 mM NaF, 10 mM sodium β‐glycerophosphate, 5 mM sodium pyrophosphate, 1 mM sodium vanadate, and 0.1% (v/v) 2‐ME] with protease inhibitors (a 1:1000 dilution of protease inhibitor mixture; MilliporeSigma). The lysates were shaken for 20 min at 4°C, centrifuged (13,500 g, 15 min at 4°C), and the supernatant was collected. Protein concentration was determined using the Pierce BCA Protein Assay. Equal amounts of protein (12 µg) were mixed with 4 times lithium dodecyl sulfate sample buffer before running on 15% SDS‐PAGE and transferred onto a PVDF membrane (GVS Group, Bologna, Italy) using a wet transfer system (Bio‐Rad, Hercules, CA, USA). The membranes were blocked with 5% nonfat dry milk for 1 h and then incubated with Total Oxphos Rodent WB Antibody Cocktail (Abcam) or citrate synthase (CS; Abeam). Antibodies were used according to the manufacturer's instructions. After overnight incubation, membranes were washed with PBS containing 0.1% Tween‐20 and then incubated with anti‐rabbit IgG secondary antibody (Cell Signaling Technology, Danvers, MA, USA) solution (1:5000) for 1 h or with anti‐mouse antibody (Cell Signaling Technology) solution (1:2500) for 1 h. The membrane was again washed as above and then exposed to a chemiluminescent protein detection system (ChemiDoc, MP Imaging System; Bio‐Rad).

As an alternative for β‐actin, stain‐free blotting was used for normalization according to the manufacturer's instructions. Briefly, 0.6% trichloroethanol (MilliporeSigma) was added to the separating gel. At the end of the electrophoresis of the gel, the gel was exposed to UV for 1 min in order to conjugate tryptophans of the samples with the trichloroethanol. After the transfer, the membrane was taken to the ChemiDoc MP system, and an image of the marked proteins was taken. The normalization factor was calculated using the Imagelab software, v.5.2.1 (Bio‐Rad).

Measurement of oxygen consumption rate in INS‐1 cells

Oxygen consumption rate was measured using Seahorse respirometry (XFe24; Agilent, Santa Clara, CA, USA), as previously described (17). INS‐1 cells were plated (120 K. cells/well) and grown on 24 well microplates (Agilent) for 24 h. The cells were then subjected to the different treatments for 20 h. The cell number was determined using the Operetta fluorescence microscope by counting the nuclei using Hoechst staining (2 µg/ml). This measurement was later used for normalization.

The INS‐1 cell medium was replaced with an assay medium [modified DMEM without sodium bicarbonate (MilliporeSigma) with an addition of 2 mM glucose], followed by 60 min incubation at 37°C (no CO₂) before loading the plate into the XFe24 extracellular analyzer. During these 60 min, the ports of the cartridge containing the oxygen probes were loaded with the compounds to be injected during the assay (50 µl/port), and the cartridge was calibrated.

After steady basal consumption rates were obtained, glucose was injected to a final concentration of 10 mM. This was followed by an injection of oligomycin A (MilliporeSigma) at a final concentration of 2 µM, followed by injection of carbonyl cyanide 4‐(trifluoromethoxy) phenylhydrazone (MilliporeSigma) at a final concentration of 3 µM and then 3 µM Rotenone (MilliporeSigma).

Measurement of oxygen consumption rate in islets

A Seahorse extracellular flux analyzer XF96 (Agilent) was used to assess oxygen consumption rate in intact mouse islets as previously described in (13) with minor modifications. Islet respirometry was performed in poly‐d‐lysine‐coated XF96 cell culture plates. Matrigel GFR (1.5 µl/well; Corning, Corning, NY, USA) was first pipetted in the center measurement zone of each well. Three islets per well were deposited in the Matrigel‐coated measurement zone using a pipette. Islets were size‐matched across wells between treatment conditions to minimize variability based on islet size. Matrigel was allowed to solidify for 3.5 min in a non‐CO₂ incubator, before addition of 100 µl of prewarmed basal Seahorse medium, containing 2.8 mM glucose and 0.1% FBS in Seahorse XF Base Medium (Agilent). Islets were incubated in a non‐CO₂ incubator for ∼1 h prior to measurement of oxygen consumption. After measurement of basal respiration, islets were acutely exposed to 20 mM glucose (final concentration) in port A, 4.5 µM oligomycin A (final concentration) in port B, 1 µM carbonyl cyanide 4‐(trifluoromethoxy) phenylhydrazone (final concentration) in port C., and finally 2.5 µM antimycin A (final concentration) in port D. The starting assay volume was 100 µl, and all injections were 25 µl for a final volume of 200 µl upon completion of the assay.

Immunocytofluorescence

INS‐1 cells were cultured on quadrant dishes and fixed with 4% v/v paraformaldehyde for 15 min at room temperature. After washing them 3 times with PBS, the cells were incubated in permeabilization buffer (2 µl/ml Triton X‐100 and 0.5 mg/ml sodium deoxycholate in PBS, pH 7.4) for 15 min at room temperature. Subsequently, the cells were blocked with 3% BSA for 1 h at room temperature and then incubated with 1:200 primary antibody of LAMP‐1 (BioLegend, San Diego, CA, USA) at 4°C overnight. The next day, the cells were washed in PBS and incubated with Anti‐Mouse Alexa Fluor 488 (Thermo Fisher Scientific) for 1 h at room temperature. Samples were kept in PBS and then imaged by Operetta and superresolution microscopy.

Confocal and super resolution microscopy

Confocal imaging was performed on Zeiss LSM710 and LSM880 (Carl Zeiss GmbH, Jena, Germany). Superresolution imaging was performed on a LSM880 microscope with Airyscan mode for LAMP‐1 immunostaining. A 488‐nm laser was used to excite Alexa Fluor 488 for INS‐1 cells stained with LAMP‐1 antibody. Live cell imaging for mitophagy assessment using the mCherry‐GFP‐Fis1 probe was performed with a humidified 5% CO₂ chamber on a temperature‐controlled stage.

Statistics

Statistical significance for all experiments was determined using either an unpaired 2‐tailed Student's t test, or a 1‐way ANOVA test followed by Tukey post hoc analysis. Values of P < 0.05 were considered significant.

RESULTS

Palmitate decreases lysosomal acidification and cell number in a dose‐dependent manner

The lipotoxic effect of FFAs depends on the concentration of its unbound fraction, which is determined by the ratio of BSA:FFA. We have previously shown that palmitate at a concentration of 0.4 mM and at a ratio of 1:5 palmitate:BSA reduces lysosomal acidity and thus suppresses autophagy (5). To further strengthen this association, we performed dose‐response experiments with different ratios.

INS‐1 cells were treated with BSA complexed to palmitate within a range of ratios (1:1, 1:2, 1:3,1:4,1:5) for 24 h and were then subjected to assays quantifying cell survival, acidic vesicular organelle (AVO) mass, and mitochondrial mass. The toxic effect of palmitate was apparent only at 1:4 and 1:5 ratios (Fig. 1A and Supplemental Fig. S1A). Concomitantly, palmitate significantly reduced the number of AVOs at 1:4 and 1:5, with a respective decrease of 23 ± 8 and 27 ± 9% compared with BSA (Fig. 1B, C ). The reduction in the number of AVOs was not accompanied by a change in the number of lysosomes as determined by LAMP‐1 staining (Supplemental Fig. S2A, B), indicating that the decrease in staining is not due to a decrease in lysosomal mass but is rather the result of a decrease in lysosomal acidification.

Figure 1.

LT increases mitochondrial mass. INS‐1 cells were treated for 24 h with either BSA alone or with different BSA complexed to palmitate (Palm) ratios (1:1, 1:2, 1:3, 1:4, 1:5). A) Cell survival as determined by counting the nuclei stained positively with trypan blue. B, C) The number of AVOs was determined after staining the cells with LysoTracker deep red (red) and Hoechst (blue). D–G) Mitochondrial mass was assessed by measuring the total F.I. of the cells stained with MTG and treated with palmitate (D; n = 3–7) and with bafilomycin A (E; n = 3) or by quantifying the expression of mitochondrial protein CS and various subunits of the 5 respiratory complexes (F, G). The latter approach was employed on cells exposed to 1:4 BSA: palmitate ratio only (n = 4–12 per protein). Note that palmitate significantly increased mitochondrial mass, up to the level observed with bafilomycin A (E). Data are means ± sem. *P < 0.05, **P < 0.01, ***P < 0.001.

Palmitate increases mitochondrial mass by reducing mitochondrial clearance rate

A decrease in autophagic flux is expected to reduce mitochondrial clearance, which, in the absence of decreased biogenesis, would result in accumulated mitochondrial mass. To test if this is the case, we examined mitochondrial mass resulting from the same dose‐response using MTG staining. As shown in Fig. 1D , increasing the unbound FFA fraction was associated with an increase in mitochondrial mass, the apex being reached at a 1:5 ratio with a 44 ± 3% increase above control. This increase was close to that observed after blocking lysosomal acidification with bafilomycin A treatment (52.83 ± 7.76% increase above control; Fig. 1E ), pointing to a potent suppression of lysosomal function and mitochondrial clearance by palmitate.

Because 1:4 was the lowest ratio to give a significant effect on all 3 parameters (viability, AVO number, and mitochondrial mass), we employed this ratio in our next experiments.

To corroborate the increase in mitochondrial mass, we measured the content of different subunits of the respiratory complexes (bound to the inner membrane) and that of CS, a matrix protein. Consistent with the results obtained with MTG staining, palmitate significantly increased the content of those mitochondrial proteins compared with the BSA control (Fig. 1F, G ).

To better elucidate how mitochondrial clearance rate is affected by palmitate, we determined changes in mitochondrial turnover by employing 2 different approaches: pulse‐chase analysis with MTG and measurement of the shift in emission of cells stably expressing the fluorescent MitoTimer reporter protein.

MTG binds covalently to mitochondrial proteins (18) and thus serves as a marker for mitochondrial turnover. INS‐1 cells were treated for 30 min with MTG, washed, and then treated with either 0.4 mM palmitate (at 4:1 with BSA) or BSA (controls). After 24 h, the F.I. was measured by flow cytometry (Fig. 2A ). As shown in Fig. 2B , the MTG F.I. at that stage was significantly higher under palmitate complexed to BSA than under BSA alone, indicating de creased clearance of mitochondria under palmitate. The relative F.I. with both palmitate and bafilomycin A (Fig. 2C ) was higher by 47.14 ± 10.52 and 74.31 ± 20.27%, respectively, than their controls.

Figure 2.

LT reduces mitochondrial turnover. A) Scheme depicting the pulse‐chase approach used to measure mitochondrial clearance. B, C) Twenty‐four hours after plating, the INS‐1 cells were stained for 30 min with MTG. The cells were then washed and treated for 24 h with either BSA, palmitate (Palm; at a 4:1 ratio with BSA), or bafilomycin A (200 nM). Mitochondrial clearance rate was subsequently assessed by measuring green F.I. with flow cytometry. Note that mitochondrial clearance rate was significantly decreased in cells exposed to bafilomycin A as well as in cells exposed to palmitate (n = 3 experiments for each condition). D–G) INS‐1 cells stably expressing MitoTimer were treated with either BSA, palmitate (at a 4:1 ratio with BSA), or bafilomycin A (200 nM) for 22 h, and mitochondrial turnover was assessed by measuring the red:green ratio at different time points. D) Representative images depicting the red and green fluorescence following 22 h treatment. Note the increase in red F.L under palmitate compared with BSA. E) Red:green F.L with time after beginning the palmitate treatment. Note that the ratio increases relative to time 0 only after 10 h treatment. F, G) Separate red and green F.L after 22 h treatment with palmitate or with bafilomycin A. The results are expressed as percentage of BSA and as percentage of RPMI, respectively (n = 3 experiments for each condition). FACS, fluorescence‐activated cell sorting. Data are means ± sem [P > 0.05 not significant (n.s.)]. *P < 0.05, **P< 0.01.

MitoTimer is a fluorescent protein targeted to the mitochondrial matrix. Upon translation, it emits green fluorescence, but after ∼16 h its emission shifts to red. Thus, the red: green ratio of cells expressing MitoTimer serves as an indicator for mitochondrial turnover, with an increase in the red: green ratio demonstrating a reduction in turnover (15, 19, 20). INS‐1 cells stably expressing MitoTimer were treated with either BSA alone or with palmitate complexed to BSA followed by measurement of the relative change in red:green F.I. across time. As shown in Fig. 2D, E , during the first 6 h of incubation, palmitate initially reduced the red: green F.I., indicating an induction in mitochondrial turnover. However, after 10 h, the trend changed, and an increase in red:green F.I. was observed. This trend continued until the termination of the experiment at 22 h (Supplemental Movies S1 and S2), reaching a significant increase of 18.4 ± 3.7% above BSA. The increase in red: green ratio was mainly the result of the accumulation of old (red) MitoTimer, indicating impaired clearance (Fig. 2F ). Long‐term treatment with bafilomycin A generated a similar, though more robust pattern (Fig. 2G ). Taken together, these results show that mitochondrial clearance is impaired under LT.

Acidic NPs restore mitochondrial mass in INS‐1 cells exposed to palmitate

To determine the role of lysosomal alkalization in the observed increase of mitochondrial mass and the dysregulation of mitochondrial turnover, we employed paNPs that we previously reported to restore lysosomal pH and autophagy in INS‐1 cells under LT (12).

The induction of lysosomal acidification with paNP proceeds in 2 steps: 1) a long‐term treatment (∼20 h) of the cells with paNPs that is aimed at targeting them to the lysosomes via the endocytotic pathway, and 2) acute (∼5 min) photoexposure of the cells to induce the release of the NPs' acidic content into the lysosomes, thus reducing the pH. Therefore, the use of paNPs is organelle specific and pharmacologically behaves in a time‐dependent and dose‐dependent manner (12). In addition, this method is now enabling us to address the outstanding question of whether the lack of acidification is upstream or down‐stream of bioenergetics dysfunction.

INS‐1 cells were treated with palmitate or BSA in the presence or absence of 25 µg/ml NPs for 20 h to allow endocytosis of inactive NPs. The cells were then exposed to UV light for 5 min to trigger the release of the acid from the NPs into the lysosomes. MTG staining was performed 2 h later to determine the effect on mitochondrial mass (Fig. 3A ).

Figure 3.

Lysosomal acidification with paNPs increases mitochondrial turnover and decreases mitochondrial mass. A) Scheme of experimental protocol for NP experiments. After plating INS‐1 cells, BSA or palmitate complexed to BSA (Palm) were administered with or without NPs. Twenty‐four h later the cells were exposed to UV light to induce photoactivation and after 2 h were subjected to assay for mitochondrial mass, mitophagy, or bioenergetic assays. B) Mitochondrial content measured by flow cytometry using MTG after 20 h exposure to BSA, palmitate (Palm), and palmitate with paNPs at a concentration of 25 µg/ml (n = 9–11). C) Representative Western blot for mitochondrial protein content (subunits of complex I, II, III, IV, V and for CS) under palmitate with and without paNPs. D) Quantification of Western blots (n = 3–5/protein); note that paNPs decrease the content of all proteins assayed. E) Confocal images of INS‐1 cells expressing mCherry‐GFP‐Fis1 targeted to the outer mitochondrial membrane after exposure to BSA, palmitate, or palmitate with paNPs. All cells were exposed to UV light 2 h before imaging. Yellow dots represent mitochondria in a nonacidic environment. Red mitochondria indicate mitochondria inside AVOs. F) Quantification of mitochondria inside AVOs. Note that palmitate decreased the number of mitochondria inside AVOs, whereas treatment with paNPs reversed this effect (n = 3 experiments with 50–80 cells analyzed). Data are means ± sem [P> 0.05 not significant (n.s.)]. *P < 0.05, **P < 0.01.

As shown in Fig. 3B , paNPs reversed mitochondrial mass accumulation in cells exposed to palmitate, reaching a level similar to that of BSA condition.

To confirm the prevention of mitochondrial mass accumulation by paNPs in cells exposed to palmitate, we examined the level of respiratory subunits and of CS following treatment with palmitate in presence and absence of paNPs (Fig. 3C, D ). In accordance with the MTG F.I. results, paNPs significantly reduced the total content of the mitochondrial proteins.

The effect of lysosomal reacidification on mitophagy was addressed using a mCherry‐GFP‐Fis1 probe, a mitochondrial chimeric protein that allows the identification of mitochondria inside autolysosomes (16). Under neutral pH conditions, mitochondria emit red and green fluorescence. When mitochondria are recruited into autolysosomes, mitochondria are exposed to low pH that quenches the green fluorescent protein signal without affecting mCherry fluorescence. As shown in Fig. 3E, F , palmitate reduced the number of red mitochondria indicating a reduction of mitophagy. However, this effect was rescued by treatment with paNPs. Taken together, the results demonstrate the capacity of NPs to recover mitophagy and mitochondrial turnover.

paNPs increase ATP content and improve mitochondrial respiratory function under LT

To determine whether lysosomal alkalization and decreased mitochondrial turnover mediate mitochondrial dysfunction, we established the capacity of paNPs to recover the energetic status of INS‐1 cells after 24 h exposure to palmitate.

As shown in Fig. 4A , palmitate significantly decreased ATP content compared with BSA. Strikingly, lysosomal re‐acidification after 24 h exposure to palmitate and 2 h only prior to the assay not only recovered ATP levels but increased them above the control (BSA) condition.

Figure 4.

paNPs restore mitochondrial function under LT. INS‐1 cells were treated with either BSA alone, palmitate complexed to BSA (Palm) in absence of paNPs, or palmitate complexed to BSA in presence of paNPs. After 20 h, the cells were photoactivated, and 2 h later they were subjected to ATP content assay (A), CS activity (B, C), or respirometry (D–H). A) Cellular ATP content as percentage of BSA. Note that paNPs recover ATP levels after exposure to palmitate compared with non‐UV–exposed palmitate (n = 3–4 experiments). B) Representative traces of CS 5, 5′‐dithio‐bis(2‐nitrobenzoic acid) absorbance assay. C) Quantification of CS‐specific activity. Note that palmitate impairs the enzyme activity compared with BSA but that it was restored under paNP treatment (n = 5–8 experiments). D–H) Oxygen consumption rates (OCRs) of INS‐1 cells treated with BSA, palmitate (Palm) and palmitate with paNPs (n = 4–7 independent experiments). D) Representative trace of OCR over time. E) Proton leak percentages. F) Coupling efficiency (ATP production rate over basal rate). G) ATP‐linked OCR (basal OCR minus oligomycin‐insensitive OCR). H) Mitochondrial spare capacity calculated as the percent increase in OCR over basal upon treatment with carbonyl cyanide 4‐(trifluoromethoxy) phenylhydrazone (FCCP). Note that reduced spare capacity under palmitate is reversed by paNP treatment. Data are means ± sem. *P < 0.05.

To assess if this increase in ATP is associated with an increase in citric acid cycle activity, we measured the activity of CS by 5, 5′‐dithio‐bis(2‐nitrobenzoic acid) absorbance. As shown in Fig. 4B, C , CS activity under palmitate was reduced to 43.3 ± 12.9% compared with BSA. Remarkably, paNPs completely restored this activity.

To further examine the bioenergetics state of the cells, we performed a respirometry experiment of INS‐1 cells treated with BSA and palmitate in the presence or in absence of NPs.

Palmitate significantly increased the proton leak and decreased coupling efficiency, ATP‐linked oxygen consumption rate, and spare respiratory capacity (Fig. 4D–H ).

In the presence of paNPs, we did not detect a significant decrease in proton leak. Strikingly, however, in spite of the observed reduction in mitochondrial mass, paNPs recovered the maximal respiratory capacity under palmitate treatment (Fig. 4D–H ). Taken together, paNPs improved respiration and enhanced mitochondrial function, suggesting the removal of defective mitochondria to maintain the health of the mitochondrial population.

paNPs restore mitochondrial respiratory function in islets exposed to LT

To verify the beneficial effects of lysosomal reacidification observed in INS‐1 cells, we next applied paNPs in a more physiologic model of isolated mouse islets. Islets were isolated from C57BL/6 mice and were exposed to a combination of 200 µM oleate and palmitate (monounsaturated and saturated FFA) conjugated with BSA at a ratio of 2:2:1 for 48 h, in the absence or presence of paNPs (Fig. 5A ). This lipid combination is more physiologically relevant than palmitate alone, because chronic high‐dose palmitate can be toxic (21).

Figure 5.

paNPs restore mitochondrial respiratory capacity in mouse islets exposed to LT. A) Combined respirometry traces of oxygen consumption rate (OCR) for mouse islets treated with BSA, oleate‐palmitate (OP), or OP with paNPs (n = 3 independent experiments, with total n = 9–12/condition). B) Basal and glucose‐stimulated OCR. C) Delta of glucose‐stimulated respiration. D) Proton leak percentages. E) ATP‐linked OCR. F) Mitochondrial spare capacity upon treatment with carbonyl cyanide 4‐(trifluoromethoxy) phenylhydrazone (FCCP). Note that reduced spare capacity under OP is reversed by paNP treatment. Data are means ± sem [P > 0.05 not significant (n.s.)]. *P < 0.05.

As shown in Fig. 5, chronic fatty acid exposure significantly decreased islet glucose‐stimulated respiration (Fig. 5B, C ) and spare respiratory capacity (Fig. 5F ) compared with the BSA control, but it did not significantly alter proton leak (Fig. 5D ) or ATP‐linked respiration (Fig. 5E ). Remarkably, treatment with paNPs at a concentration of 2 µg/ml reversed the lipid‐mediated reductions in glucose‐stimulated mitochondrial respiration (Fig. 5C ) and spare respiratory capacity (Fig. 5F ). These results confirm the beneficial effects of paNPs in primary β cells exposed to lipotoxic conditions and suggest that lysosomal pH is a critical factor for maintaining mitochondrial function.

DISCUSSION

Mitochondria are subject to a tight quality control mechanism that involves mitochondrial fission, fusion, and clearance of damaged mitochondria via autophagy. Several studies document that LT is associated with both mitochondrial and lysosomal dysfunction (5, 23, 24). Although mitochondria may provide ATP essential for lysosomal acidification, lysosomes allow for mitophagy to proceed (3, 5, 8). To address a cause and effect relationship, mitochondrial function must be studied under conditions in which lysosomes are reacidified. Here, we report that restoration of lysosomal acidity recovers mitochondrial function under nutrient excess in β cells, demonstrating that bioenergetic defects are downstream of lysosomal alkalization under LT.

Mitochondrial mass accumulation under LT

LT decreases both mitochondrial fusion and autophagic flux in primary β cells and in INS‐1 cells (5). In this study, we show that under LT mitochondrial clearance is decreased because of suppression of autophagic flux that emanates from lysosomal alkalization. We further demonstrate that this effect is associated with an accumulation in dysfunctional mitochondria, as reflected by the reduction in respiratory capacity.

To measure the effect of LT on mitochondrial mass, two different approaches are used: staining mitochondria with MTG and immunoblot analysis for key mitochondrial respiratory chain proteins. Results from both approaches point to an increase in mitochondrial mass under LT. This observation is in line with a previous report by Anello et al. (25) showing an increase in the levels of several mitochondrial proteins in islets from patients with type 2 diabetes.

Mitochondrial clearance is impaired under LT

In this study we used two complementary approaches to assess the effect of LT on mitochondrial clearance: pulse‐chase with MTG and analysis of MitoTimer. These facile methods allow high‐throughput analysis, but each has its weaknesses. MTG reacts with the free thiol groups of cysteine residues in inner membrane proteins and thus could be affected by the redox state (26). Moreover, changes in the morphology of the inner membrane under different metabolic states may affect the interpolation of the F.I. reading by flow cytometry. Finally, an inherent confounding factor of pulse‐chase with MTG is the probe's recycling and the potential activity of the molecule used for the pulse. However, the complementation of our results with the MitoTimer approach corroborates impairment in mitochondrial turnover by palmitate. MitoTimer's increase in red:green fluorescence upon palmitate exposure supports the hypothesis that mitochondrial clearance is suppressed during LT. The absence of an effect of palmitate on MitoTimer's green fluorescence, which is associated with the generation of new mitochondria, suggests that the mitochondrial biogenesis rate is unaffected. Yet the interpretation of this observation must be carefully considered. Indeed, if mitochondrial biogenesis reflects the level of mitochondrial gene expression, then MitoTimer's green fluorescence, being regulated by a general promoter (EF1α promoter), correlates to general gene expression and not to mitochondrial gene expression or biogenesis. Thus, although the impairment in mitochondrial clearance by FFAs is well supported, further examination should be undertaken to rule out the involvement of mitochondrial biogenesis in the increase in mitochondrial mass.

In line with our previous report (5) showing a dual effect of FFAs on autophagy, with an increase in autophagic flux in the short term and a decrease in the long term, we observed a dual effect on mitochondrial turnover with MitoTimer. This association supports the role of impaired autophagic flux in mediating the effect of chronic FFA exposure on mitochondrial clearance.

Restoration of lysosomal acidification by NPs restores mitochondrial mass and mitochondrial function

If mitochondrial function were not impaired, one would expect that an increase in mitochondrial mass would be reflected by an increase in total function. However, this is not the case during LT. Under LT, mitochondrial respiration is suppressed, and the mitochondrial membrane potential is reduced (5), thus accounting for the reduced ATP content. Our study demonstrates that NP‐mediated lysosomal reacidification prevents mitochondrial mass accumulation and restores mitochondrial function. Although it is possible that the recovery of mitochondrial function is unrelated to the recovery of mitochondrial mass, it is likely that clearance of damaged mitochondria that reside in the autophagosomes affects β‐cell bioenergetics. One possibility is that the engulfed mitochondria affect the mitochondria that are outside the autophagosomes by producing deleterious diffusible molecules such as ROS. Supporting this possibility are studies showing that ROS scavengers rescue β cells from LT (27).

Another interesting possibility is that mitochondria that accumulate inside the autophagosomes under LT are ATP‐hydrolyzing mitochondria, namely, mitochondria whose ATP synthase activity has shifted to an ATPase mode because of the depolarization of the mitochondria. We have previously shown that mitochondria targeted to the autophagosomes are depolarized (8), making this scenario a possibility worth investigating. Following re‐acidification of the lysosomes, those mitochondria are expected to be degraded, leading to recovery of ATP levels and thus to recovery of GSIS, as observed with treatment of the cells with the paNPs. The activation of the acidic NPs was performed only 2 h prior to our endpoint assays, indicating that this time span is sufficient to clear dysfunctional mitochondrial and reduce overall mass to the control level, which in turn significantly increases maximal respiration and ATP levels in INS‐1 cells and glucose‐stimulated respiration and maximal respiration in mouse islets. Although there was a trend of paNPs to restore oxidative phosphorylation capacity (but not efficiency, meaning that more nutrients are needed to produce the same ATP), it is likely that restoring lysosomal acidification independently of vATPase activity might contribute to the recovery of ATP levels. In addition, it is possible that the glycolytic flux generating ATP is improved as well, given the interplay between autophagy and glyceraldehyde 3‐phosphate dehydrogenase activity.

In view of the central role of mitochondria and ATP in mediating GSIS (28), it is likely that the restoration of maximal cellular respiration and ATP content observed with the activated NPs is responsible for the recovery of GSIS previously reported by us (12). Under stimulatory glucose, mitochondria drive the increased ATP: ADP ratio in β cells. Thus, our results suggest that paNP treatment improved mitochondrial ATP synthesis in β cells exposed to excess fatty acids.

CONCLUSIONS

Disruption of autophagy and impaired lysosomal acidification are associated with genetic disorders, neurodegenerative diseases, obesity, and type 2 diabetes. Autophagy is a cellular housekeeping process that is particularly critical in nonproliferating cells and tissues that rely on it to remove damaged material, including mitochondria that accumulate with aging. Our results demonstrate that LT‐induced mitochondrial dysfunction observed in β cells is the result of reduced autophagy‐mediated mitochondrial turnover; this is the first time this sequence of events has been confirmed in this model of diabetes‐related β‐cell dysfunction. Remarkably, restoration of lysosomal acidity by novel paNPs recovers cellular bioenergetics and function. Accordingly, paNP application shows that lysosomal acidification is a potential therapeutic target and represents a novel strategy to enhance mitophagy in diabetic pancreatic β cells and to rescue their function.

AUTHOR CONTRIBUTIONS

E. A. Assali, D. Shlomo, J. Zeng, E. P. Taddeo, K. M. Trudeau, K. A. Erion, and A. H. Colby performed research experiments; E. A. Assali, D. Shlomo, J. Zeng, K. M. Trudeau, K. A. Erion, A. H. Colby, M. W. Grinstaff, M. Liesa, G. Las, and O. S. Shirihai designed the study; E. A. Assali, D. Shlomo, K. M. Trudeau, K. A. Erion, G. Las, and O. S. Shirihai analyzed data and interpreted results; E. A. Assali, D. Shlomo, M. W. Grinstaff, G. Las, and O. S. Shirihai wrote the manuscript; and E. P. Taadeo, K. M. Trudeau, K. A. Erion, M. W. Grinstaff, and M. Liesa edited the manuscript.

Supporting information

Supplementary Material

Supplementary Material

Supplementary Material

Supplementary Material

Assali, E. A. , Shlomo, D. , Zeng, J. , Taddeo, E. P. , Trudeau, K. M. , Erion, K. A. , Colby, A. H. , Grinstaff, M. W. , Liesa, M. , Las, G. , Shirihai, O. S. Nanoparticle‐mediated lysosomal reacidification restores mitochondrial turnover and function in β cells under lipotoxicity. FASEB J. 33, 4154–4165 (2019). www.fasebj.org