Mitochondrial Autophagy Promotes Cellular Injury in Nephropathic Cystinosis

Abstract

The molecular and cellular mechanisms underlying nephropathic cystinosis, which exhibits generalized proximal tubular dysfunction and progressive renal failure, remain largely unknown. Renal biopsies from patients with this disorder can reveal abnormally large mitochondria, but the relevance of this and other ultrastructural abnormalities is unclear. We studied the ultrastructure of fibroblasts and renal proximal tubular epithelial cells from patients with three clinical variants of cystinosis: Nephropathic, intermediate, and ocular. Electron microscopy revealed the presence of morphologically abnormal mitochondria and abnormal patterns of mitochondrial autophagy (mitophagy) with a high number of autophagic vacuoles and fewer mitochondria (P < 0.02) in nephropathic cystinosis. In addition, we observed increased apoptosis in renal proximal tubular epithelial cells, greater expression of LC3-II/LC3-I (microtubule-associated protein 1 light chain 3), and significantly more autophagosomes in the nephropathic variant. The autophagy inhibitor 3-methyl adenine rescued cell death in cystinotic cells. Cystinotic cells had increased levels of beclin-1 and aberrant mitochondrial function with a significant decrease in ATP generation and an increase in reactive oxygen species. This study provides ultrastructural and functional evidence of abnormal mitophagy in nephropathic cystinosis, which may contribute to the renal Fanconi syndrome and progressive renal injury.

Cystinosis is an inherited disorder caused by mutations in the CTNS gene, which encodes cystinosin, a lysosomal transmembrane protein involved in cystine export to the cytosol.¹ A large number of genetic variants have been characterized in CTNS; nevertheless, poor clinical correlation exists between genotyped mutations and the different clinical phenotype of infantile nephropathic cystinosis with proximal Fanconi tubulopathy² and renal failure in the first decade of life³ or adult and ocular cystinosis, which have mild or no renal involvement.⁴ The molecular and cellular basis of disease heterogeneity and mechanisms underlying renal Fanconi syndrome and renal tubulopathy are not well understood. Although cystine accumulation is toxic to the milieu of the cells, renal injury in nephropathic cystinosis may not simply be caused just by cystine accumulation, because renal injury is not seen either in other human forms of cystinosis or in the murine cystinosis knockout model despite high cystine load in the Ctns^−/− mouse kidney⁴ and progressive renal injury occurs despite cystine depletion therapy. In untreated patients, with the generalized accumulation of cystine in various organs, patients develop hypothyroidism, photophobia myopathy and retinal blindness, chronic renal failure, pulmonary dysfunction, and central nervous system calcifications and symptomatic deterioration.⁵ To improve on current treatment options for cystinosis, it is critical to understand the basis of generalized and multisystem tissue injury from lysosomal cystine accumulation, as well as the specific cellular and molecular injury mechanisms in the kidney that result in early Fanconi syndrome and subsequent renal failure, despite the use of aggressive cystine depletion therapy.⁶

Autophagy is the process by which organelles and bits of cytoplasm are sequestered and subsequently delivered to lysosomes for hydrolytic digestion.⁷ Increasing evidence indicates that autophagy of mitochondria occurs selectively, and the term “mitophagy” has been suggested for this process.⁸ As a major source of reactive oxygen species (ROS), mitochondria are especially prone to ROS damage. Oxidative stress and various disease processes cause mitochondrial damage and dysfunction. Timely elimination of aged and dysfunctional mitochondria is essential to protect cells from the harm of disordered mitochondrial metabolism and release of proapoptotic proteins. The mechanism of mitochondrial turnover is predominantly autophagic sequestration and delivery to lysosomes for hydrolytic degradation.

We used individual samples of nephropathic cystinotic renal proximal tubular epithelial (RPTE) cells and skin fibroblasts extracted from patients with three clinical phenotypes of cystinosis—nephropathic (in which the Fanconi syndrome and renal failure is mandatory), intermediate (in which Fanconi syndrome and renal insufficiency are mild and of later onset), and ocular (in which there is no renal involvement and corneal crystals are the main complaint)⁹—to explore the specific injury mechanism in nephropathic cystinosis. Our data demonstrate enhanced apoptosis, abnormal mitochondria, mitophagy, reduced mitochondrial ATP generation, and increased ROS generation in nephropathic cystinosis. We further demonstrate that specific inhibition of autophagy results in significant attenuation of cell death in nephropathic cystinosis. Our findings suggest that mitochondrial autophagy may be a critical mechanism of renal injury in nephropathic cystinosis.

Results

Morphologic Analysis of Cystinotic Fibroblasts by Fluorescence Imaging

With the aim to study the morphology of mitochondria in the nephropathic cystinotic cells, we stained four nephropathic cystinosis fibroblast samples with MitoTracker Red and analyzed them by confocal microscopy. Normal fibroblasts showed intense and definite staining, whereas a diffused staining was observed in the cystinotic cells (Figure 1, A and B). Moreover, in the normal fibroblasts, mitochondria were organized as extended tubular structures, whereas, in the cystinotic fibroblasts, the mitochondrial network was altered to rounded, nonfilamentous, fragmented structure. We further verified the mitochondrial morphologic aberrations in nephropathic cystinotic cells by evaluating the patterns of immunostaining for ATP5H, a mitochondrial membrane protein. The immunofluorescence (IF) images for ATP5H were in agreement with the MitoTracker images and clearly revealed a deteriorated fragmented mitochondrial morphology in cystinosis as opposed to an intact filamentous tubular network in normal fibroblasts (Figure 1, C and D).

Figure 1.

Morphologic anomalies in nephropathic cystinotic fibroblasts are shown. (A and B) MitoTracker Red CMXROS staining reveals abnormal mitochondrial staining and distribution in cystinotic fibroblasts. Live cells (1 × 10⁶ cells/ml) were incubated with MitoTracker Red (50 nM) for 45 min under growth conditions, followed by fixation with 3.7% (vol/vol) formaldehyde. Mitochondria were visualized by confocal microscopy, and the representative images are shown. (C and D) Immunostaining specific for a mitochondrial protein, ATP5H, revealed deteriorated mitochondrial morphology in cystinotic fibroblasts. Cells (1 × 10⁶ cells/ml) were fixed in 4% formalin, permeabilized, and immunostained with antibodies for ATP5H. Slides were viewed by confocal microscopy, and the representative images are shown. As seen, MitoTracker Red is already diffuse in the cytoplasm, suggesting loss of mitochondrial membrane potential in cystinosis. Magnifications: ×63 in A and B; ×100 in C and D.

Ultrastructural Morphologic Alterations in Cystinosis

Electron Microscopy of Cystinotic RPTE Cells.

To characterize better the morphologic aberrations of mitochondria in cystinosis, we used electron microscopy (EM) to image the ultrastructural alterations in cystinotic cells. The renal proximal tubules are the major site of injury in nephropathic cystinosis; therefore, we performed EM on RPTE cells obtained from patients with nephropathic cystinosis and compared them with the normal RPTE cells. We analyzed eight nephropathic cystinotic RPTE cell cultures by EM. Figure 2A represents low-power EM images showing abundant autophagic vesicles (AV) in cystinotic cells. In normal cells, the mitochondria were present as well preserved structures with normal morphology (Figure 2B), whereas the cystinotic cells presented mitochondria with condensed matrix (Figure 2C) and abnormally shaped dark cristae (Figure 2D) in addition to the abundant AVs and autolysosomes (ALs; Figure 2, E and F).

Figure 2.

Electron microscopic evaluation of cystinotic cells is shown. (A) EM of human normal and nephropathic cystinotic RPTE cells at low magnification. Arrows indicate AVs. (B through F) EM of normal RPTE (B) and nephropathic cystinotic RPTE cells (C through F). Abnormal mitochondria in cystinosis are shown by arrows in C through F and an inset in C. AVs and ALs are shown in E and F. Bars = 1 μm. (G through N) Progressive mitochondrial autophagy in nephropathic cystinotic RPTE cells. Structurally abnormal mitochondria (G) are surrounded by typical double membrane of autophagosomes (H and I) or fuse with vacuoles (J and K), followed by fusion with lysosomes (L) to generate ALs (M and N). Bars = 1 μm in G through J and L through N and 0.5 μm in K. (O) Quantification of mitochondria in cystinotic RPTE cells. Number of mitochondria per cell in cystinosis was significantly low as compared with normal cells, whereas no significant decrease was observed in CDME-treated normal cells. (P through S) Electron microscopy on normal human fibroblasts (P) and cystinotic fibroblasts (Q through S). Q and R insets show magnified version of abnormal mitochondria indicated by arrows. Inset in S shows magnified autophagosomes in cystinosis. Bars = 2 μm. (T) Quantification of mitochondria in cystinotic fibroblasts. Number of mitochondria per cell in nephropathic cystinosis was significantly low as compared with normal cells, whereas no significant decrease was observed in CDME-treated normal cells. No significant change was observed in the number of mitochondria in intermediate and ocular phenotype compared with the normal controls. CDME-treated normal fibroblasts and RPTE cells did not exhibit the presence of AV, but morphologically abnormal mitochondria were observed. Only the representative EM images are shown. Magnifications: ×20,500 in B through J and L through N; ×44,000 in K; ×5900 in A and P through S.

Careful EM examination of cystinotic cells at a higher magnification revealed various stages of degradation of mitochondria by autophagy (Figure 2, G through N). Figure 2G demonstrates morphologically abnormal mitochondria with dark cristae and condensed matrix. The abnormal mitochondria were often seen in close association with sequestering membranes during the initial stages of mitophagy (Figure 2H). The sequestering membranes were observed to form a cup-shaped membranous structure that eventually enveloped the aberrant mitochondria (Figure 2I). In addition, mitochondria seemed to be closely associated with the vacuoles (Figure 2J). These vacuolar membrane formations have previously been described to occur in mitophagy.^10,11 Also, the close contacts that looked like fusion between the vacuoles and mitochondria were observed (Figure 2K). The double-membrane vesicles sequestering mitochondria, called autophagosomes, were seen fusing with lysosomes to form ALs (Figure 2, L through N). Selective degradation of organelles has been described to occur via microautophagy or macroautophagy.^12,13 Evidence of the vacuolar membranes fusing with mitochondria and a complete sequence of autophagosome formation engulfing mitochondria strongly suggest that both microautophagy and macroautophagy of mitochondria are taking place in cystinotic RPTE cells. Quantification of mitochondria per cell (Figure 2O) demonstrated a significant decrease in number of mitochondria in nephropathic cystinotic RPTE cells as compared with the normal cells (P = 0.016).

EM of Cystinotic Fibroblasts.

We also performed EM on eight fibroblast cell cultures obtained from patients with cystinosis (all three phenotypes) and compared them with the normal fibroblasts. We observed well-preserved mitochondrial structures in normal fibroblasts (Figure 2P), whereas giant (Figure 2Q) and spherical mitochondria with abnormal cristae (Figure 2R) were present in nephropathic cystinotic fibroblasts. EM also revealed the presence of numerous autophagosomes in nephropathic cystinotic fibroblasts (Figure 2S). Interestingly, we observed very few AVs and mostly morphologically normal mitochondria in the fibroblasts from ocular and intermediate phenotypes (data not shown). Quantification of mitochondria (Figure 2T) demonstrated that there was a significant decrease in the number of mitochondria per cell in the nephropathic cystinotic fibroblasts (P = 0.0005) when compared with the normal fibroblasts.

Specific Mechanism of Cell Injury in Cystinosis

Demonstration of Increased Autophagy in Cystinosis

Expression analyses of autophagy related proteins, LC3 and beclin-1, were conducted to investigate the enhanced autophagy in cystinosis.

Evaluation of Autophagy Markers LC3 and Beclin-1 in Cystinotic Fibroblasts and RPTE Cells by Western Blot.

To characterize further the enhanced autophagy in cystinosis, we analyzed the expression status of LC3 and beclin-1 in the cystinotic cells. LC3 is a mammalian homologue of yeast Apg8p, and during the formation of autophagosomes, the LC3-I isoform is converted into LC3-II, the amount of which correlates with the number of autophagosomes.¹⁴ LC3-II is the only known protein that specifically associates with autophagosomes and not with any other vesicular structures. We used this property of LC3 to monitor the dynamics of autophagy in cystinosis. This lipid-conjugated, final form of LC3, designated LC3-II, migrates faster than LC3-I in SDS-PAGE. Consequently, Western blot analysis with anti-LC3 antibody gives two bands with different relative molecular weights (18 kD for LC3-I and 16 kD for LC3-II). The intensity of the LC3-II bands, when compared with the control cells, was higher in all of the nephropathic cystinotic RPTE and fibroblasts cell lysates that were analyzed by Western blot (Figure 3A, top). Densitometry of LC3 bands, normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH), in both fibroblasts and RPTE cells demonstrated that there was a significant increase in the LC3-II/LC3-I ratio in the nephropathic cystinotic cells compared with the controls (P = 0.039 and 0.021, respectively; Figure 3A, bottom). Cystine dimethyl ester (CDME)-treated normal fibroblasts and RPTE cells did not exhibit any significant difference in the levels of LC3-II (data not shown). This is in agreement with the electron micrographs of CDME-treated cells, which demonstrate the absence of AVs.

Figure 3.

Modification of endogenous LC3 and beclin-1 in nephropathic cystinosis indicating activated autophagy. (A) Fibroblasts and RPTE cell extracts were subjected to PAGE through a 4 to 20% gel, and separated proteins were immunoblotted using a rabbit polyclonal anti-LC3 antibody and subsequently detected with goat anti-rabbit peroxidase and ECL reagent (Amersham Biosciences). The same immunoblots were stripped off and reprobed with a mouse monoclonal anti-GAPDH antibody. LC3 bands were normalized and quantified by Image J, and the LC3-II/LC3-I ratio was plotted. In fibroblasts, the increase in LC3-II signal was detected only in the nephropathic cystinotic fibroblasts and not in intermediate and ocular cystinotic fibroblasts. (B) Nephropathic cystinotic RPTE cell extracts were examined by Western blot specific for beclin-1 and compared with normal RPTE and HK-2 controls. The same immunoblot was stripped off and reprobed with a mouse monoclonal anti-GAPDH antibody. (C) Nephropathic cystinotic RPTE cells immunostained with anti-LC3 antibody and visualized under a confocal microscope revealed a vesicular punctate pattern in contrast to normal RPTE cells, which exhibit more diffuse cytosolic expression of LC3. Bar = 47.62 μm. (D) Quantification of LC3 immunostaining was done by counting and averaging the LC3-positive dots per cell in at least five view fields per specimen under confocal microscope, suggesting enhanced autophagy in cystinosis. The figure is representative of multiple experiments. Magnification, ×63 in C.

Increased autophagy in cystinotic cells was further confirmed by Western blot for another autophagy marker, beclin-1. This mammalian orthologue of the yeast Apg6/Vps30 gene has a key role in autophagy.^15–18 It regulates the autophagy-promoting activity of Vps34¹⁹ and is involved in the recruitment of membranes to form autophagosomes. Beclin-1 can also bind to Bcl-2, an important regulator of apoptosis.²⁰ As shown in Figure 3B, increased beclin-1 expression was observed in nephropathic cystinotic RPTE cells compared with normal RPTE and HK-2 controls. These results demonstrate that beclin-1 is upregulated in cystinotic kidney and may be a critical regulator of autophagy in cystinosis.

Evaluation of LC3 in Nephropathic Cystinotic RPTE Cells by IF.

To assess the enhanced autophagy in cystinosis, we performed IF analysis to compare the distribution of endogenous LC3 in four nephropathic cystinotic RPTE cell cultures. As expected, LC3 was mainly cytosolic with diffused staining in control cells, although occasional LC3-labeled autophagosomes could also be detected (Figure 3C). In contrast, numerous LC3-positive structures with enhanced fluorescence and punctate staining pattern were observed in cystinotic cells, suggesting the translocation of LC3 to autophagosomes and thus indicating constitutive activation of basal autophagy in these cells (Figure 3C). Quantitative analysis revealed that the LC3-positive dots per cell were significantly greater in cystinotic cells than in control cells (P = 2.50E-06; Figure 3D). Altogether, our results indicate an enhancement of autophagy in cystinosis.

Demonstration of Increased Mitophagy in Nephropathic Cystinosis by IF Imaging.

We further investigated mitophagy in nephropathic cystinotic RPTE cells by immunostaining for LC3 (autophagy marker), LAMP2 (lysosomal marker), and ATP5H (mitochondrial marker), as shown in Figure 4, A through F. Merged images of individual immunostaining are shown in Figure 4, G through N. We attempted to ascertain autophagy of mitochondria, as observed in electron micrographs, by evaluating co-localization of ATP5H and LC3. Our data indicated co-localization of LC3 and ATP5H in cystinotic cells, thereby confirming the sequestration of mitochondria inside autophagosomes (Figure 4K). Because autophagy is a constitutive cellular event, our data indicated a basal co-localization of LC3 and ATP5H in control cells (Figure 4G). When autophagy is induced, autophagosomes fuse with lysosomes to form ALs, the site of ultimate degradation of their cargo; therefore, we examined the co-localization of LAMP2 and LC3 to gain insight into this final step of autophagy. The ubiquitously expressed LAMP2 is localized primarily in the late endosomes and lysosomes.²¹ This enrichment of LAMP2 in late AVs is similar to that observed for LAMP1.²² The distribution of LC3 and LAMP2 showed a largely nonoverlapping pattern in normal control cells (Figure 4I). Cystinotic cells showed overlapping structures, indicative of the formation of ALs (Figure 4M). Also, distribution of LAMP2 and ATP5H exhibited overlapping pattern in cystinosis (Figure 4L), whereas no such co-localization was observed in the normal control cells (Figure 4H). Next, as evident in Figure 4, J and N, when the images for all three markers (LC3, LAMP2, and ATP5H) were merged, the cystinotic cells exhibited much more enhanced co-localization as compared with the control cells, indicating autophagy of mitochondria in cystinosis.

Figure 4.

Fluorescence immunostaining for autophagic marker (LC3), lysosomal marker (LAMP2), and mitochondrial marker (ATP5H) in nephropathic cystinotic RPTE cells is shown. The intracellular distributions of LC3, LAMP2, and ATP5H were studied by indirect IF. (A through F) Cells (1 × 10⁶ cells/ml) were fixed in 4% formalin, permeabilized, and co-immunostained with antibodies for LC3, LAMP2, and ATP5H. (G through N) Localization patterns of LC3:LAMP2, LC3:ATP5H, LAMP2:ATP5H, and LC3:LAMP2:ATP5H are shown in nephropathic cystinotic and normal RPTE cells. A significant extent of co-localization is observed in cystinotic kidney cells. Bar = 47.62 μm. (O through Q) High-power images for co-localization of mitochondrial marker (ATP5H, blue) with LC3 (red) in autophagosomes in cystinotic RPTE cells. (Q) White arrowhead in the zoomed image indicates a ring-like shape and a co-localized ATP5H-LC3 signal in cystinosis. Bar = 30 μm. Magnification, ×63 in A through N; ×100 in O through Q.

To validate further mitophagy in nephropathic cystinotic cells, we investigated the co-localization of ATP5H and LC3 at a higher magnification. We immunostained the normal and nephropathic cystinotic RPTE cells for LC3 and ATP5H and captured the fluorescence images at a higher magnification (×100). We observed a weak and diffused cytoplasmic signal of LC3 in normal control cells with a minimal co-localization with ATP5H (Figure 4O). In contrast, LC3 signals in cystinosis were often found to be circular, consistent with their localization on autophagosomal membrane (Figure 4P). The merged images for LC3 and ATP5H in cystinosis (Figure 4Q) clearly indicated circular LC3 signals sequestering ATP5H, suggesting autophagy of mitochondrial structures in these cells. Thus, our fluorescence microscopy data confirmed enhanced autophagy of mitochondria in nephropathic cystinotic RPTE cells.

Functional Significance of Mitophagy in Nephropathic Cystinosis

Mitochondrial Dysfunction in Cystinosis.

We further analyzed mitochondrial function by assaying the rate of ATP synthesis and ROS production in nephropathic cystinotic RPTE cell cultures. Accordingly, we assessed electron transfer capacities of nephropathic cystinotic RPTE cells and measured ATP synthesis in the presence and absence of oligomycin, an inhibitor of mitochondrial ATP synthase, to evaluate mitochondrial ATP production only. The ATP synthesis rate was measured in the presence of either complex I substrates (glutamate and malate) or complex II substrate (succinate) with complex I inhibitor, rotenone. We detected a substantial reduction of ATP synthesis in cystinotic cells with complex I substrate (P = 0.01), whereas ATP level measured in the presence of complex II substrate was not significantly different in cystinotic and control cells (Figure 5A). These results indicate a correlation of the observed reduction of ATP synthesis with an impairment of complex I activity in cystinotic cells.

Figure 5.

Determination of mitochondrial ATP synthesis rate and ROS generation is shown. (A) Mitochondrial ATP synthesis rate was measured in the presence and absence of oligomycin. By providing glutamate and malate, ATP synthesis dependent on complexes I, II, III, IV, and V was stimulated. Using the complex I inhibitor rotenone, succinate ATP production dependent on complexes II, III, IV, and V was stimulated. Significantly decreased mitochondrial ATP generation was observed in nephropathic cystinotic RPTE cells with complex I substrate. (B) ROS production was determined by dichlorodihydrofluorescein diacetate fluorescence, and the data are expressed as fluorescence (arbitrary units) per milligram of protein. ROS levels were significantly higher in the cystinotic cells when compared with the normal control cells. CDME-treated control cells did not reveal any significant change in ATP synthesis and ROS generation. Experiments were conducted on four nephropathic cystinotic cells cultures, and the averages are plotted.

In mitochondria and submitochondrial particles as well as in intact cells, respiration produces ROS such as H₂O₂ and superoxide anion, especially when respiration is inhibited or otherwise disordered.^23–26 Accordingly, We measured ROS generation in the cystinotic cells. DCFH-DA was used for ROS detection. DCFH-DA is cleaved intracellularly by nonspecific esterases to form DCFH, which is further oxidized by ROS to form the fluorescence compound DCF.²⁷ As shown in Figure 5B, ROS levels were significantly higher in the cystinotic cells (P = 0.003).

Autophagy is Pro-death in Nephropathic Cystinosis.

Autophagy has also been more recently linked to the death process itself. Furthermore, apoptosis and autophagy are not mutually exclusive pathways; they may act in synergy or counter to each other.²⁸ To examine whether abnormal induction of autophagy correlated with decrease in cell survival in cystinotic cells, we measured the viability of four nephropathic cystinotic RPTE cell cultures in the presence and absence of 3-MA, an autophagy inhibitor. The viability of cystinotic cells increased significantly when treated with 3-MA (P = 1.35E-07; Figure 6A). We further stained the 3-MA–treated and untreated cystinotic RPTE cells with Annexin V and analyzed them using flow cytometry. Cystinotic cells had a higher percentage of Annexin V–positive cells (Figure 6B, top), with a significant reduction (approximately 14%) in apoptosis when rescued with 3-MA (Figure 6B, bottom).

Figure 6.

Effect of autophagy on cell survival and apoptosis. (A) Nephropathic cystinotic RPTE cells and normal control RPTE cells were seeded in 12-well plates overnight and then treated with 3-MA. After 8 h, viable cells were determined by Trypan Blue Exclusion assay. Percentage of viable cells increased significantly in 3-MA–treated nephropathic cystinotic RPTE cells. The plot is representative of three independent experiments performed in triplicate of each sample. (B) Increase apoptosis in nephropathic cystinosis, with a reduction in apoptosis when treated with 3-MA, as seen by a significant decrease in the percentage of Annexin V-FITC–stained cystinotic RPTE cells with 3-MA treatment. Shown is one representative experiment of four cystinotic RPTE cell cultures, each performed in triplicate.

Discussion

The pathogenesis of proximal renal tubular dysfunction and Fanconi syndrome in nephropathic cystinosis is poorly understood. The most frequent renal symptom in mitochondrial cytopathies is Fanconi syndrome, and why the proximal tubules seem particularly sensitive to mitochondrial injury and its consequences is not clear.²⁹ An obvious answer is that it requires high levels of energy in the form of ATP. In cystinosis, data from fibroblasts and animal models suggested decreased ATP levels but intact mitochondrial electron transport chain activity^30–33; however, the involvement and role of mitochondria in the pathogenesis of nephropathic cystinosis remains unclear. Thus, we sought possible factors causing ATP depletion and perturbation of mitochondrial function, a potential cause for renal Fanconi in cystinosis. Here, for the first time, we describe a striking degree of mitochondrial autophagy specific to the nephropathic phenotype that could possibly be an important event leading to ATP depletion and renal Fanconi with subsequent renal injury and cell death in nephropathic cystinosis.

Abnormal autophagy and mitochondrial aberrations were also recently observed in other types of lysosomal storage diseases.³⁴ This is the first study in which involvement of autophagy in nephropathic cystinosis has been investigated. To date, EM is the gold standard to monitor the formation of autophagosomes, the morphologic hallmark of autophagy. Here, EM showed an increased autophagic activity in nephropathic cystinotic cells that displayed abundant vacuolization, a widely known morphologic indicator for autophagic cell death. Indeed, Western blot and IF evaluation of LC3-II, a widely used marker for autophagy, also revealed an increased expression only in the most severe³⁵ nephropathic phenotype of cystinotic fibroblasts and RPTE cells. Vesicles of autophagic pathway appear as electron-dense dark bodies in electron micrographs.³⁶ Previously reported “mysterious” dark cells of unexplained origin that are unique to cystinotic kidneys may be due to accumulation of vesicular compartments of autophagic origin.³⁷

The next question we addressed was whether this abnormal induction of autophagy observed in the nephropathic cystinotic cells was specific to an organelle in the cells. Our EM and IF co-localization data established that the mitochondria in cystinotic cells were specifically targeted to autophagy. Interestingly, IF results exhibited higher levels of perinuclear staining for LAMP2 in cystinosis. It was previously shown that LAMP2 may be involved in lysosomal biogenesis and/or the fusion between autophagosomes and lysosomes required for the final catabolism of autophagic material.^15,38,39 Thus, increased LAMP2 expression in cystinosis also suggests enhanced active autophagy in these cells. Studies in mouse models of two LSDs showed that the autophagic delivery of bulk cytosolic contents to lysosomes is impaired as a result of decreased ability of lysosomes to fuse with autophagosomes³⁴; however, in the cystinotic cells, we observed that there is a significant degree of co-localization between lysosomes and autophagosomes as evident by LC3 and LAMP2 immunostaining of these cells. This suggests that fusion of AV and lysosomes may be efficient in cystinosis.

Interestingly, EM and IF analysis of CDME-treated normal fibroblasts and RPTE cells exhibited the presence of morphologically abnormal mitochondria, but no AVs were observed, which was also confirmed by LC3-II Western blot of CDME-treated cells. In fact, many dead cells were observed, indicating that CDME treatment was leading to cell death, as analyzed by EM and IF. Many studies have used a CDME in vitro model to study cystinosis in the past, for its ability artificially to load lysosomes with cystine^30,40–43; however, the use of a CDME-loading model in studying pathogenesis of cystinosis can be questioned, because loading with CDME has a direct and acute impact on the viability of the cells as a result of direct toxicity from CDME, independent of cystine accumulation.^44,45 CDME is rapidly (within minutes) converted to cystine in lysosomes and leaves fairly rapidly from normal cells; therefore, drawing conclusions on cystinosis cellular injury on the basis of CDME-loading models may not be accurate. Our results suggest that the mitophagy observed in cystinosis may be independent of the effect of simply cystine accumulation in cystinotic cells.

Increased apoptosis has been reported in cultured cystinotic fibroblasts and RPTE cells.^40,42,46 Apoptosis is thought to involve the activation of caspases as well as a stereotyped pattern of mitochondrial alterations, namely release of cytochrome c and mitochondrial membrane potential dissipation that contribute to the acquisition of the apoptotic morphology. Apoptosis rate determined by caspases-3 activity in cystinotic RPTE cells has been reported to be significantly increased both with and without an apoptosis trigger.⁴² Basal level of apoptosis in cystinotic RPTE cells, as measured by Annexin V staining, was observed to be increased in this study as well. It has been shown that protein kinase Cδ forms disulfide bonds specifically with cystine that is released from lysosomes in cultured cystinotic fibroblasts and RPTE cells during apoptosis.⁴⁶ Also, loss of mitochondria integrity, as measured by cytochrome c release, has been shown to occur downstream of lysosomal permeability in cystinotic cells.⁴⁶ Accumulating evidence suggests that different intracellular organelles contribute synergistically to the initiation of apoptosis by specific stress inducers. It is evident from the electron micrographs that the mitochondria in cystinotic cells are abnormal in structure and vary in size. The abnormal giant mitochondria have previously been reported in cystinotic kidney.⁴⁷ It has been shown that there is a disorder in ATP generation and ROS generation in cystinotic cells.^30,31 Our data show that the mitochondria in cystinotic kidney cells are dysfunctional and independent of the loading effect of cystine, with increased ROS and deficient mitochondrial ATP generation, which may result from an impairment of complex I activity in nephropathic cystinosis.⁴⁸ Mitochondrial dysfunction is instrumental in triggering apoptosis/autophagy.⁴⁹ These alterations in mitochondrial function likely also result in endoplasmic reticulum (ER) stress or vice versa; ER stress was previously demonstrated in various lysosomal storage diseases, including nephropathic cystinosis.⁵⁰ ER stress can further induce an autophagic response by stimulating the assembly of the preautophagosomal structures.⁵¹ Interestingly, it was recently shown that caspase-4 expression, which is known to be activated by ER stress, increases specifically in proximal tubules in cystinotic kidneys⁵² and that ER stress induces autophagy in renal proximal tubular cells.⁵³ Both lysosomal and mitochondrial destabilization may contribute to the initiation stage of apoptosis under ER stress. A number of studies revealed interesting cross-talk between the apoptotic and autophagic pathways, with the identification of several proteins that can play a role in both responses.^54,55 Moreover, these pathways are not mutually exclusive; they may be both synergistic and antagonistic under different conditions. On the basis of our data, autophagy and apoptosis seem to be synergistic in nephropathic cystinosis. Autophagy may act as an enabler of apoptosis, participating in certain morphologic and cellular events that occur during the injury associated with nephropathic cystinosis. Specific and selective inhibition of autophagy in vitro results in a significant attenuation of apoptosis in nephropathic cystinosis. Inhibition of autophagy may thus be an attractive target for the attenuation of tissue injury and, specifically, renal injury in nephropathic cystinosis.

Concise Methods

Cells

Ten primary fibroblast cultures were used for this study; two were form normal fibroblasts (GM02651 and GM00316; Coriell Cell Repositories, Camden, NJ) and eight were from cystinosis patients attending the National Human Genome Research Institute at the National Institutes of Health (Bethesda, MD). Eighteen RPTE cultures were used; one lot of normal primary RPTE cells (Cambrex Biosciences, East Rutherford, NJ), the HK-2 cell line (American Type Culture Collection, Manassas, VA), and 16 primary cultures from cystinotic RPTE cells isolated from the urine of patients with nephropathic cystinosis (a gift from Dr. Lorraine Racusen⁵⁶). The study was controlled by institutional review board approvals from the National Institutes of Health and Stanford University.

The fibroblast cells were cultured in MEM with Earle's salts, supplemented with 15% FBS, 2 mM lglutamine, 2× concentration of nonessential AA, 100 μg/ml penicillin, 100 U/ml streptomycin, and 0.5 μg/ml Fungizone (all from Invitrogen Corp., Carlsbad, CA) at 37°C in a 5% CO₂ atmosphere. The medium was changed every 3 d, and cultured cells were released by 0.05% Trypsin-EDTA (Invitrogen) and passaged. All RPTE cells were cultured in renal epithelial growth medium, made according to the manufacturer's instructions (Cambrex). All cells were passaged with trypsin (0.05%) and were cultured in a 95% air/5% CO₂ Thermo Forma incubator (Waltham, MA) at 37°C. All of the experiments were performed between passages 2 and 5. For in vitro mimicking of cystinosis, the cells were treated with 1 mM CDME (Sigma) for 30 min.⁴⁴

Antibodies

The primary antibodies used were goat polyclonal anti-human LAMP2 purchased from Santa Cruz Biotechnology (Santa Cruz, CA), rabbit polyclonal anti-human LC3 antibody purchased from MBL (Woburn, MA), rabbit polyclonal anti-human GAPDH antibody purchased from Abcam (Cambridge, MA), rabbit polyclonal anti-human beclin-1 antibody purchased from Cell Signaling Technology (Beverly, MA) and mouse monoclonal anti-human ATP synthase subunit d antibody purchased from Mitosciences (Eugene, OR). Secondary antibodies used were peroxidase-conjugated goat ant-rabbit IgG and FITC-conjugated rabbit anti-mouse IgG purchased from Jackson ImmunoResearch Laboratories (West Grove, PA). For co-immunostaining, the secondary antibodies used were Alexa Fluor 555 donkey anti-rabbit IgG, Alexa Fluor 647 donkey anti-mouse IgG, and Alexa Fluor 488 donkey anti-goat IgG purchased from Invitrogen.

Electron Microscopy

Cells were harvested gently using Trypsin EDTA, washed with PBS, fixed at 4°C for 2 h with 2% glutaraldehyde in neutral phosphate buffer, postfixed in osmium tetroxide, and embedded in Epon. Sections were cut at 80 nm, stained with lead citrate and uranyl acetate, and examined under an FEI (Hillsboro, OR) Tecnai10 electron microscope. For quantification of mitochondria, the number of mitochondria per cell was scored. The electron micrographs obtained from multiple distinct low-powered fields (×4200 to ×5900) were used to count the number of mitochondria per cell in at least five different view fields for each cell culture sample, and the average number of mitochondria per cell culture was calculated. Data are presented as means ± SD.

IF Imaging and MitoTracker Staining

MitoTracker Red CMXROS kit (Molecular Probes) was used for MitoTracker Red staining of the cells. MitoTracker Red is a marker that accumulates in the mitochondria regardless of the mitochondrial membrane potential. Briefly, 1 × 10⁶ cells/ml cells were incubated in the prewarmed growth medium containing the MitoTracker Red probe (40 nM) for 45 min. After staining, cells were washed with prewarmed growth medium without MitoTracker Red. The growth medium was then replaced with freshly prepared 4% paraformaldehyde to fix the cells and incubated at 37°C for 20 min, after which the cells were rinsed several times with PBS and mitochondria were visualized at ×63 (oil) magnification using a Leica SP2 AOBS Confocal Laser Scanning Microscope equipped with Leica software.

For immunostaining of mitochondrial ATP5H protein, cells were plated on chamber slides, washed twice in PBS, fixed in 4% formalin (30 min), and permeabilized with 0.5% Triton X-100 (30 min). Cells were incubated in blocking buffer (PBS [pH 7.2] and 3% BSA) for 1 h, then washed twice in PBS followed by incubation with the primary antibody for 2 h at room temperature. FITC-conjugated secondary antibody was used to detect bound primary antibody for 1 h at room temperature. Similarly, for LC3 immunostaining and co-immunostaining of LC3, LAMP2, and ATP5H, cells were processed and co-immunostaining was performed by incubation with the primary antibodies (LC3, LAMP2, and ATP5H) for 2 h at room temperature followed by washing and incubation with secondary antibodies (Alexa Fluor 555, Alexa Fluor 488, and Alexa Fluor 647, respectively) for 2 h at room temperature. Slides were viewed using a Leica SP2 AOBS Confocal Laser Scanning Microscope, and the images were analyzed by Leica Confocal software (version 2.5). For quantification of autophagy in LC3-immunostained RPTE cells, the number of LC3-positive dots per cell was scored. The fluorescence images were used to count the LC3-positive dots per cell in at least five different view fields for each cell strain, and the average number of dots per cell strain was calculated. Data are presented as means ± SD.

Protein Extraction and Immunoblotting

Cells were harvested and lysed in CellLytic-M reagent (Sigma) supplemented with protease inhibitor cocktail (Pierce) according to the manufacturer's instructions. Cell extracts were quantified using QuantiPro BCA Assay Kit (Sigma), and an equal amount of total protein was subjected to SDS-PAGE. Cell extracts were mixed with 6× Laemmli loading buffer and heated to 95°C for 5 min before electrophoresis at 120 V for 1 to 2 h on 4 to 20% SDS-polyacrylamide gels. For immunoblotting, proteins were transferred to Immobilon polyvinylidene difluoride membrane (Millipore, Billerica, MA) for 75 min at 20 V in a MiniProtean III transfer tank (BioRad, Hercules, CA). Membranes were then air-dried before rewetting with 70% methanol followed by blocking with PBS supplemented with 0.1% (vol/vol) Tween-20 and 5% (wt/vol) milk. All primary antibody incubations were done in PBS supplemented with 0.1% (vol/vol) Tween-20 and 1% (wt/vol) milk for a minimum of 1 h at room temperature followed by washing with PBS-Tween (PBS supplemented with 0.1% Tween). Peroxidase-conjugated secondary antibodies were diluted 1:10,000 in PBS-Tween, incubated with the blot for a minimum of 1 h at room temperature, then washed with PBS-Tween and developed using ECL Plus detection reagent (Amersham). Band quantification was performed using the ImageJ software (National Institutes of Health).

Determination of ROS

The cells were seeded and grown in medium for 24 h in 24-well plates (2 × 10⁴/well). ROS production was measured after loading with 5 μmol/L DCFH-DA for 30 min.⁵⁷ After washing, cells were incubated in medium without phenol red. Fluorescence measurements (excitation and emission wave lengths of 485 and 535 nm, respectively) were carried out with a Tecan Infinite M200 (Mannedorf, Switzerland). Fluorescence values were normalized for protein concentration.

ATP Synthesis Rate

The cellular ATP synthesis was determined by using the luciferin/luciferase assay.⁵⁸ Briefly, after trypsinization, cells were resuspended (7 × 10⁶/ml) in buffer A (10 mmol/L KCl, 25 mmol/L Tris-HCl, 2 mmol/L EDTA, 0.1% BSA, 10 mmol/L potassium phosphate, and 0.1 mmol/L MgCl₂ [pH 7.4]), kept for 15 min at room temperature, and then incubated with 50 μg/ml digitonin for 1 min. After centrifugation, the cell pellet was resuspended in buffer A, and aliquots were taken to measure ATP synthesis and protein content. Aliquots of cells were incubated with 5 mmol/L malate plus 5 mmol/L glutamate (complex I–driven substrates) in the presence or absence of 10 μg/ml oligomycin or with 10 mmol/L succinate plus 2 μg/ml rotenone (complex II–driven substrate) and 0.2 mmol/L ADP for 3 min. The values obtained for ATP in the presence and absence of oligomycin were used to estimate the ATP synthesis rate by mitochondrial oxidative phosphorylation in the cystinotic cells. The rate of ATP synthesis was expressed as nanomoles of ATP produced per minute per milligram of protein.

Cell Viability Measurement

Cells were seeded in 12-well plates and then incubated in the presence or absence of 10 mM 3-MA (Sigma) for 8 h. Cell death was determined by the Trypan blue exclusion assay. The cells were scraped and resuspended in the Trypan blue solution (0.4%) and counted in a hemacytometer under a light microscope, and percentage of viable cells was calculated. At least three independent experiments were conducted, and each sample was tested in triplicate.

Flow Cytometry

Apoptosis of nephropathic cystinotic RPTE cells treated with or without autophagy inhibitor (3-MA) was assessed by flow cytometry, analyzing the Annexin V–stained cells (Annexin Apoptosis Detection Kit; Santa Cruz Biotechnology) with FACSCalibur (Becton Dickinson,) cytometry. Cells were seeded in six-well culture plates. After 15 h, the cells were treated with 10 mM 3-MA for 8 h. After harvesting, the cells were washed, incubated for 15 min with Annexin V–FITC, and analyzed by flow cytometry. The percentage of Annexin V–stained cells was assessed using the CellQuest Pro software (Becton Dickinson). At least three independent experiments were performed, and each sample was tested in triplicate.

Statistical Analysis

Unless otherwise indicated, data are presented as means ± SD of three or more experiments. The t test was used to compare two means. Statistical significance was defined at the level of P < 0.05.

Disclosures

None.