Restoration of CTSD (cathepsin D) and lysosomal function in stroke is neuroprotective

ABSTRACT

Stroke is a leading cause of death and disability. The pathophysiological mechanisms associated with stroke are very complex and not fully understood. Lysosomal function has a vital physiological function in the maintenance of cellular homeostasis. In neurons, CTSD (cathepsin D) is an essential protease involved in the regulation of proteolytic activity of the lysosomes. Loss of CTSD leads to lysosomal dysfunction and accumulation of different cellular proteins implicated in neurodegenerative diseases. In cerebral ischemia, the role of CTSD and lysosomal function is not clearly defined. We used oxygen-glucose deprivation (OGD) in mouse cortical neurons and the middle cerebral artery occlusion (MCAO) model of stroke to assess the role of CTSD in stroke pathophysiology. Our results show a time-dependent decrease in CTSD protein levels and activity in the mouse brain after stroke and neurons following OGD, with concurrent defects in lysosomal function. We found that shRNA-mediated knockdown of CTSD in neurons is sufficient to cause lysosomal dysfunction. CTSD knockdown further aggravates lysosomal dysfunction and cell death in OGD-exposed neurons. Restoration of CTSD protein levels via lentiviral transduction increases CTSD activity in neurons and, thus, renders resistance to OGD-mediated defects in lysosomal function and cell death. This study indicates that CTSD-dependent lysosomal function is critical for maintaining neuronal survival in cerebral ischemia; thus, strategies focused on maintaining CTSD function in neurons are potentially novel therapeutic approaches to prevent neuronal death in stroke.

Abbreviations: 3-MA: 3-methyladenine; ACTB: actin beta; AD: Alzheimer disease; ALS: amyotrophic lateral sclerosis; CQ: chloroquine; CTSB: cathepsin B; CTSD: cathepsin D; CTSL: cathepsin L; FTD: frontotemporal dementia, HD: Huntington disease; LAMP1: lysosomal associated membrane protein 1; LSD: lysosomal storage disease; MCAO: middle cerebral artery occlusion; OGD: oxygen glucose deprivation; OGR: oxygen glucose resupply; PD: Parkinson disease; SQSMT1: sequestosome 1; TCA: trichloroacetic acid; TTC: triphenyl tetrazolium chloride.

INTRODUCTION

Lysosomes are essential subcellular organelles critical for maintaining protein and cellular homeostasis via the degradation of unwanted cellular components, including damaged or misfolded proteins [1,2]. Lysosomal function is a vital physiological process in virtually all cell types, including neurons, and any alteration in lysosomal function can lead to harmful effects in a cell. However, the role of lysosomal dysfunction in different diseases was unrecognized for many years, as lysosomes were merely regarded as terminal degradative compartments participating in eliminating cellular waste. Recent studies have demonstrated that lysosomal dysfunction can lead to cell death via the aggregation of proteins and cellular components in many diseases [3–5]. Lysosomal dysfunction is now recognized as one of the major pathological processes in neurodegenerative diseases, including Alzheimer disease (AD), Parkinson disease (PD), Huntington disease (HD), amyotrophic lateral sclerosis (ALS), and frontotemporal dementia (FTD) [5,6]. Defects in lysosomal function lead to neurodegeneration and cognitive decline due to the deposition of pathogenic Abeta and MAPT/Tau in AD [7,8]. Likewise, the inability of lysosomes to clear a toxic form of polyglutamine HTT (huntingtin; polyQ-HTT) is an essential pathological process in HD [9]. Similarly, the accumulation and aggregation of SNCA (synuclein, alpha) lead to the loss of dopaminergic neurons in PD [10,11]. Studies involving ALS, FDT, and GRN/progranulin-mediated neurodegeneration further describe defective lysosomes as a mediator of neurodegeneration [12–14].

Lysosomes require a full range of proteases to carry out their function. CTSD (cathepsin D), an aspartic protease, is found in the lysosomes and is ubiquitously present in almost all tissues with higher expression levels in the brain. CTSD exists as a 53 kDa preform in the Golgi complex, which is processed into an enzymatically active 48 kDa intermediate form in the endosomes [15,16]. The fully mature form of CTSD consists of a 14 kDa light chain and a 28 to 30 kDa heavy chain in the lysosomes [15]. Mature CTSD is present in a majority of cells, including cancer cells [17]. However, in neurons, an intermediate 48 kDa CTSD is the predominant form [18]. CTSD is critical for maintaining lysosomal-dependent protein homeostasis in the brain [19–21]. In models of PD, lysosomal dysfunction due to CTSD haploinsufficiency promotes the cell-to-cell transmission of SNCA aggregates [22]. Defects in the activity or decreases in CTSD expression can lead to pathological protein aggregation in congenital neuronal ceroid lipofuscinosis, ALS, FTD, PD, and AD [23,24]. Interestingly, not all lysosomes are CTSD-positive in neurons [18], but CTSD is considered as an essential factor for lysosomal proteolytic activity. Thus, in neurons, only a small pool of lysosomes is proteolytically active, and their function is important in the brain [18]. The severe phenotype, such as seizures, neurodegeneration, and early post-natal death of ctsd knockout mice further supports the critical role of CTSD in the brain [19,21,25]. Therefore, the dysfunction or loss of CTSD is almost certain to aggravate or induce neuronal impairment or injury in the brain.

Stroke is a leading cause of death and disability in both the United States and the world. The pathology of stroke is very complex and not fully understood. A majority of proteins involved in neurodegenerative diseases aggregate in the models of stroke [26], suggesting that defects in protein clearance pathways may contribute to stroke pathology [26,27]. A study using a global proteomics approach revealed that the expression of CTSD decreases following stroke in rats [28], indicating that the loss of CTSD might play a critical role in protein aggregation in stroke. A recent study in the permanent occlusion model of stroke revealed that loss of the lysosomal protein degradation pathway contributes to stroke pathology [29]. Presumably, a decrease in CTSD following stroke is contributing to lysosomal dysfunction in stroke. However, the role of CTSD in stroke pathogenesis or neuroprotection is not known. In this study, we used oxygen-glucose deprivation (OGD) in mouse cortical neurons and a mouse middle cerebral artery occlusion (MCAO) model to study the role of CTSD in stroke. Our results demonstrate that alteration in CTSD is an early event in the pathophysiology of stroke and that loss of CTSD leads to lysosomal dysfunction, protein aggregation, and cell death. This study suggests that therapeutics targeted to increase CTSD and lysosomal function may represent a robust strategy to limit brain damage following stroke and related neural diseases.

Results

CTSD decreases in neurons following OGD

CTSD decreases in the infarct brain following a stroke in rats [28]. To investigate the role of CTSD in stroke, we used in vitro OGD for 90 min followed by oxygen-glucose resupply (OGR) in mouse cortical neurons, as depicted in Figure 1A. At 90 min, OGD induced approximately 65% cell death in mouse cortical neurons assessed by Alamar Blue 24 h after OGR (Fig. S1A). Hoechst and PI staining also showed a significant increase in cell death (PI-positive cells) similar to the cell death observed in Alamar Blue assay (Fig. S1B). We collected neuronal lysates at indicated time points after OGR or from control cultures to assess the levels of CTSD using western blotting and immunofluorescence. At 0 h following OGD, we did not observe any change in CTSD levels (Fig. S2A-C). However, our data revealed that the levels of both proCTSD (48 kDa) and mature CTSD (34 kDa) were decreased following OGD in neurons at 1 h, 3 h, and 6 h after OGR (Figure 1A-D) as compared to control neurons. At 6 h of OGR, the level of proCTSD was ~0.5-fold to that of the control neurons (Figure 1B). Levels of both pro- and mature-CTSD were decreased significantly in OGD neurons, indicating either increased degradation of both forms of CTSD or less transport of proCTSD to the lysosome resulting in reduced formation of the mature form. However, the quantification of mature or proCTSD suggests that the degradation of both forms of CTSD in OGD, as the ratio of both forms, were decreased in neurons exposed to OGD.

Figure 1.

CTSD levels and activity are reduced in neurons after OGD

(A) Schematic diagram of the experimental design (Top). Representative western blot showing levels of CTSD in neurons after OGD. Neurons were exposed to OGD for 90 min, and cell lysates were collected at indicated time points after OGR to assess CTSD expression. Quantitative analysis of (B) proCTSD, (C) mCTSD and (D) mCTSD/proCTSD levels normalized to the ACTB level (n = 4). (E) Representative immunostaining of CTSD (green) and LAMP1 (red) in control and OGD neurons after 3 h of OGR. Nucleus is stained with DAPI (blue). Scale bar: 20 μm. Quantification of fluorescence intensity of (F) LAMP1-positive CTSD, (G) CTSD-positive LAMP1, (H) CTSD, and (I) LAMP1 puncta in control and OGD neurons (n = 6). (J) Ctsd mRNA levels in control and OGD neurons (n = 3). (K) CTSD activity in control and OGD neurons (n = 4). Data represent mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001 vs control, calculated with Student’s t-test for (F) – (I), and one-way ANOVA followed by Tukey’s post hoc test for (B)-(D) and (J, K). ns, not significant.

Immunostaining data also showed a loss of CTSD in neurons exposed to OGD (Figure 1E-I). The intensity of CTSD in the LAMP1-positive puncta and total CTSD was decreased (Figure 1F-I). However, the intensity of CTSD in LAMP1-positive puncta was lower than the intensity of total CTSD (~47% and ~57% of control, respectively) (Figures 1F–1H). Similarly, the intensity of LAMP1 puncta in CTSD-positive puncta was higher than that of the total LAMP1 intensity (~174% and ~138% of control, respectively) (Figures 1G–1H), indicating that CTSD level decreases as the intensity of LAMP1-positive puncta increases, which is an indication of lysosomal dysfunction [30]. qRT-PCR on RNA collected from control neurons and OGD-treated neurons showed that Ctsd mRNA did not alter in OGD/OGR (Figure 1J), suggesting that the changes in CTSD expression are mainly due to the alteration in protein levels. Since the ubiquitin-proteasomal pathway is a major route of protein degradation, we investigated CTSD protein levels in OGD neurons in the presence of the ubiquitin proteasomal pathway inhibitors (MG132 and lactacystin). We found that treatment with MG132 or lactacystin recovered CTSD levels in neurons exposed to OGD (Fig. S1D and S1E). These results indicate that the ubiquitin proteasomal pathway degrades CTSD in stroke. CTSD is an aspartate protease, and its activity is crucial for lysosomal function. We also measured CTSD activity and observed that the catalytic activity of the enzyme gradually decreased in OGD (Figure 1K). We further observed that the CTSD level and activity were also down in lysosomal fractions (Fig. S3A and S3B) of neurons exposed to OGD treatment. One possibility for the decrease in CTSD levels following stroke is that it may be accumulated in the Triton X-100-insoluble fractions. We monitored CTSD levels in the Triton X-100-insoluble fractions following OGD using western blotting and observed no detectable levels of CTSD in the insoluble fractions (Fig. S1 F). Together, these data suggest that decreases in protein levels and catalytic activity of CTSD are early pathological hallmarks in cortical neurons exposed to OGD.

CTSD decrease in the brain of MCAO mice

To assess whether CTSD also follows a similar trend in vivo, we used the MCAO model of stroke in mice for 90 min and collected samples at different time points, as depicted in Figure 2A. MCAO for 90 min led to consistent infarcts in the brain, as assessed by triphenyl tetrazolium chloride (TTC) staining 24 h after reperfusion (Figures 2B–2C). The evaluation of the MCAO brains revealed that CTSD expression was significantly decreased in the ischemic stroke brain at 1 h, 3 h, and 6 h after reperfusion as compared to the sham brain (Figure 2D-F). As observed in OGD, the decrease in CTSD was not due to its accumulation in the Triton X-100-insoluble fraction (Fig. S1 G). Immunostaining analysis, using RBFOX3/NeuN and GFAP as markers for neurons and astrocytes, respectively, revealed that CTSD was specifically decreased in neurons following MCAO (Figure 2G-I). Analysis of its catalytic activity also revealed that CTSD activity was decreased in stroke brain at the indicated time points after reperfusion compared to the sham brain (Figure 2J).

Figure 2.

Cerebral ischemia reduces CTSD levels and activity

(A) Schematic diagram showing the induction of stroke and reperfusion time. Mice were subjected to MCAO for 90 min followed by reperfusion at indicated time points. (B) Quantification and (C) TTC staining showing infarct volume of MCAO brain after 24 h reperfusion (n = 3). (D) Western blot showing CTSD levels in sham and stroke brains. Brain samples were prepared from the ipsilateral cortex area of sham-operated mice, and mice subjected to MCAO followed by 1, 3, and 6 h of reperfusion. Quantification of (E) proCTSD and (F) mCTSD, normalized to ACTB levels (n = 3). (G) Immunofluorescence images of the ipsilateral cortex of sham- and MCAO-operated brain showing the expression of CTSD (green) in RBFOX3 (red) or GFAP (red) positive cells. Scale bar: 20 μm. Quantification of fluorescence intensity of (H) RBFOX3-positive CTSD, (I) GFAP-positive CTSD puncta in sham and MCAO brain (n = 3). (J) CTSD activity in sham and stroke brains (n = 3). Data represent mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001 vs control, calculated using one-way ANOVA followed Tukey’s post hoc test.

OGD induces lysosomal dysfunction in neurons

Studies have shown that CTSD is required for the proteolytic activity of the lysosomes and any alteration in the CTSD levels or activity can cause lysosomal dysfunction [20,31]. Therefore, we assessed the lysosomal function in neurons after OGD. We used different but complementary methods to evaluate lysosomal function in the control and OGD-treated neurons. Live-cell imaging using LysoTracker Red DND-99 and LysoSensor Blue DND-167 was performed to assess lysosomal function. Confocal image analysis of lysosomal staining with these vital dyes demonstrated that LysoSensor staining was decreased in LysoTracker-positive puncta in neurons exposed to OGD (Figures 3A–3B), suggesting lysosomal dysfunction. Next, we assessed the lysosomal proteolytic capacity of long-lived proteins using a previously described radioactive pulse-chase assay, which is a widely accepted method for evaluating the lysosomal function. This assay uses [³ H]˗leucine to label long-lived proteins to assess the rate of proteolysis over time [32]. Our results showed a significant reduction in the rate of proteolysis in OGD neurons as compared to control neurons (Figure 3C), suggesting an accumulation of long-lived proteins in neurons following OGD. Long-lived protein aggregates contain ubiquitin conjugates [33–35] and are targeted to the lysosomal pathway for degradation [33,36,37]. Therefore, an increase in ubiquitination in the insoluble fractions is an alternative indicator of lysosomal dysfunction [35,38]. We assessed protein ubiquitination in the Triton X-100-insoluble fractions from control and OGD-exposed neuronal cell lysates by western blotting. The results indicated a significant accumulation of ubiquitinated proteins in the Triton X-100-insoluble fractions of neurons exposed to OGD (Figures 3D–3E). Ubiquitin-positive aggregates accumulated as early as 1 h following OGD (Figures 3D–3E), suggesting that lysosomal dysfunction is an early pathological process in OGD. Moreover, LAMP1 (lysosomal-associated membrane protein 1) and LAMP2 accumulate in lysosomal dysfunction [30]. Therefore, we analyzed the levels of LAMP1 and LAMP2 in OGD neurons using western blotting. Our data showed that both LAMP1 and LAMP2 were accumulated in neurons exposed to OGD in a time-dependent manner (Figure 2F-I). Together, these results indicate that lysosomes become dysfunctional in neurons following OGD.

Figure 3.

Lysosomal function is compromised in neurons exposed to OGD

(A) Top: Schematic diagram showing the timeline of live-cell imaging using LysoTracker and LysoSensor. Neurons were exposed to OGD for 90 min. Lysosomal dyes were added to neurons at 2 h of OGR, and images were captured after 1 h of incubation. Below: phase-contrast and confocal images showing staining of lysosomes with LysoTracker (red) and LysoSensor (blue). Scale bar: 20 μm. (B) Quantification of LysoSensor fluorescence intensity in control and OGD neurons (n = 6), Student’s t-test. (C) Measurement of the rate of proteolysis of long-lived proteins in control and OGD neurons through radioactive pulse-chase assay (n = 3). (D) Western blot showing ubiquitinated protein levels in insoluble (Triton X-100-insoluble) fractions of control and OGD neurons. (E) Quantification of ubiquitinated protein levels (n = 3). (F) Western blot and (G) Quantification of LAMP1 levels, and (H) Western blot and (I) Quantification of LAMP2 levels in control and OGD neurons. LAMP1 and LAMP2 levels were normalized to corresponding ACTB (n = 3) for each group. Data represent mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001 vs control, determined with one-way ANOVA followed by Tukey’s post hoc test for (C), (E), (G) and (I).

Autophagy initiation remains intact, but lysosomal clearance is hampered in neurons subjected to OGD

Damaged proteins and cellular components are delivered via autophagosomes to the lysosomes for degradation. Autophagy is a multistep process comprising the formation, maturation, and fusion of the autophagosomes with lysosomes. This dynamic process leads to the formation of LC3-II from LC3-I. A change in LC3-II levels is a hallmark of altered autophagy in the cells. We monitored LC3-II formation in mouse neurons, following OGD at the indicated time points, using western blotting. Our data showed that the levels of LC3-II significantly increased in neurons following OGD (Figures 4A–B), suggesting that OGD induced changes in autophagy. To assess whether it is an increase in autophagy induction (autophagosome formation) or a decrease in lysosomal clearance, we evaluated LC3-II levels in the presence of lysosome inhibitors, ammonium chloride (NH₄ Cl) or chloroquine (CQ) and an inhibitor of autophagy induction, 3-methyladenine (3-MA). Our data showed that with NH₄ Cl and CQ treatment, the levels of LC3-II were accumulated in control neurons, indicating lysosomal inhibition with these inhibitors (Figures 4C-F). Western blot data, at 0 h of OGD, showed that lysosome inhibition led to a significant upregulation of LC3-II level compared to the vehicle group, suggesting an intact autophagic flux at this time point (Fig. S2D-G). However, the levels of LC3-II in the presence of NH₄ Cl or CQ at 6 h of OGR was not further elevated (Figures 4C-F), suggesting inhibition of autophagic flux at this time point. Conversely, 3-MA effectively decreased LC3-II levels in neurons after OGD compared to untreated neurons (Figures 4G–H), indicating that the initiation of autophagy was not affected by OGD.

Figure 4.

OGD induces the accumulation of autophagy-related proteins in neurons

(A) Expression and (B) quantification of LC3-II/I in control and OGD neurons (n = 4). (C) – (H) Expression and quantification of LC3-II/I in control and OGD neurons in the presence or absence of lysosome inhibitors (NH₄Cl or CQ) and autophagosome biogenesis inhibitor, 3-MA. n = 4, for each group. Neurons were processed for WB 6 h after OGR. (I) Western blot and (J) quantification of SQSTM1 in Triton X-100-soluble, and (K) western blot and (L) quantification of SQSTM1 in Triton X-100-insoluble fractions of control and OGD neurons at different time intervals after OGR. SQSTM1 levels were normalized to corresponding levels of ACTB. n = 3 for each group. Immunofluorescence images of (M) CTSD (green) and LC3B (red) puncta and (N) CTSD (green) and SQSTM1 (red) in control and OGD neurons. Images were captured after 6 h of OGR (n = 3). Regression analysis showing the correlation between puncta intensity between CTSD and LC3B in (O) control and (P) OGD neurons. Regression analysis showing the correlation between puncta intensity between CTSD and SQSTM1 in (Q) control and (R) OGD neurons. Data represent mean ± SEM. **P < 0.01 and ***P < 0.001 compared to control using one-way ANOVA followed by Tukey’s post hoc test for (B), (D), (F), (H), (J) and (L), ns, not significant.

Additionally, we observed that the levels of SQSTM1 were decreased progressively in the soluble fractions and increased in the insoluble fractions in a time-dependent manner in OGD neurons as compared to control neurons (Figures 4I-L), further indicating that OGD induced lysosomal impairment in neurons. SQSTM1 is degraded by the autophagy-lysosomal pathway and accumulated as aggregates in cells upon inhibition of autophagic flux [39]. Our data showing the accumulation of SQSTM1 in insoluble fractions of OGD neurons suggest inhibition of autophagic flux in these neurons. Under a basal condition, LC3B and SQSTM1 exhibit diffuse staining pattern but generate punctate structures with increased intensity when autophagy is impaired. Therefore, we used the immunostaining of CTSD, LC3B, and SQSTM1 in neurons to assess whether the accumulation of LC3B or SQSTM1 puncta accumulates in the same cells that show a decrease in CTSD. Our data indicated that either LC3B or SQSTM1 puncta were accumulated in neurons that displayed a loss of CTSD staining (Figures 4M–N). Linear regression analysis further demonstrated that significant correlations exist between the puncta intensity of CTSD and LC3B or SQSTM1 (Figures 4O-R). Together, these results show that OGD does not affect the induction of autophagy but induces lysosomal dysfunction and decreases autophagic flux in neurons.

Autophagy function is impaired in the mouse brain following MCAO stroke

To assess whether the autophagic function is perturbed in cerebral ischemia in vivo, we determined the levels of autophagy-related proteins in the mouse brain following MCAO. Western blots of brain samples collected at different time points after 90 min MCAO demonstrated a significant increase of the autophagic marker LC3-I and LC3-II (Figures 5A–B), as compared to the sham brain, suggesting an alteration in autophagy. We also observed a simultaneous decrease of SQSTM1 in the Triton X-100-soluble fractions and a concurrent increase in the insoluble fractions of the stroke brain (Figures 5C-F). Ubiquitinated proteins in Triton X-100-insoluble fractions also accumulated at 1 h, 3 h, and 6 h after reperfusion in the stroke brains (Figures 5G–H), while LAMP1 and LAMP2 levels in stroke brain, determined by western blotting, also increased (Figures 5I-L). Consistent with the in vitro results, these data suggest that autophagic and lysosomal dysfunction is an early pathological event in stroke.

Figure 5.

Autophagic function is perturbed in the stroke brain

(A) Western blot showing levels of LC3-I and LC3-II in sham and stroke brain after reperfusion at indicated time points. (B) Quantitative analysis of LC3-II/I levels (n = 3). (C) Western blot and (D) quantification of SQSTM1 in Triton X-100-soluble, and (E) western blot and (F) quantification of SQSTM1 in Triton X-100-insoluble fractions of sham and stroke brains at different time intervals after reperfusion. SQSTM1 levels were normalized to corresponding levels of ACTB (n = 3). (G) Western blot showing the ubiquitination of aggregated proteins in sham and stroke brains. Triton X-100-insoluble fractions were analyzed to assess ubiquitination. (H) Quantification of ubiquitination normalized to ACTB (n = 3). ND, not detected. (I) Western blot and (J) quantification of LAMP1, and (K) western blot and (L) quantification of LAMP2 levels in sham and stroke brains. LAMP1 and LAMP2 levels were normalized to corresponding levels of ACTB. n = 3 for each group. Data represent mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001 vs indicated group, calculated with one-way ANOVA followed by Tukey’s post hoc test.

CTSD is indispensable for lysosomal function in neurons

We observed a decrease in CTSD with simultaneous lysosomal dysfunction in neurons following OGD in the stroke brains. To assess whether a reduction in CTSD is sufficient to cause lysosomal dysfunction, we employed a CTSD knockdown in primary neurons using shRNA lentiviral transduction (Figure 6A). Our data showed an efficient knockdown of both the 48 and 34 kDa forms of CTSD (Figure 6A). LysoTracker Red DND-99 and LysoSensor Blue DND-167 staining showed a drastic decrease in LysoSensor Blue DND-167 uptake in LysoTracker Red DND-99-positive puncta in neurons with CTSD knockdown compared to neurons transduced with the control lentiviral shRNA (Figures 6B–C). Proteolysis rates of the long-lived proteins assessed with the radioactive pulse-chase assay also showed a significant decrease in neurons with CTSD knockdown compared to the control neurons (Figure 6D). Moreover, ubiquitinated protein aggregates and LAMP1 and LAMP2 levels increased in CTSD-depleted neurons (Figures 6E-J). Studies have shown that CTSD is the major protease required for lysosomal function. Lysosomes lacking CTSD have lesser degradative capacity [18], while deficiency of other proteases, including CTSB (cathepsin B) or CTSL (cathepsin L), does not affect autophagic flux [40,41]. Previous studies reported alterations in the levels of CTSB and CTSL during stroke [42–44]; therefore, we assessed the levels of CTSB using immunohistochemistry. Our data showed that CTSB was also decreased in neurons following OGD (Fig. S4A and S4B). We also assessed the activity of CTSB, which was also decreased in OGD-exposed neurons (Fig. S4C). To assess whether the loss of CTSB induces lysosomal dysfunction, we used lentiviral-mediated Ctsb shRNA transduction in the mouse cortical neurons to knockdown CTSB (Fig. S4D) and assessed the LAMP1, LC3-I and -II levels, and the accumulation of ubiquitinated proteins in these neurons. Unlike CTSD knockdown, CTSB knockdown did not affect the LAMP1, LC3-II/I, or ubiquitinated protein levels (Fig. S4E-J). Interestingly, the overexpression of CTSD in OGD-exposed neurons restored the activity of CTSB (Fig. S4 K), which is likely due to the restoration of lysosomal function following CTSD recovery. We also assessed the CTSL levels using western blotting. Our data indicated that CTSL did not change after OGD in the mouse cortical neurons (Fig. S4 L and S4 M). These data support the notion that CTSD is indispensable for lysosomal function in neurons, and suggest that lysosomal dysfunction in OGD may be a result of the decrease in CTSD.

Figure 6.

CTSD is essential for lysosomal function

(A) Western blot showing the knockdown of CTSD in neurons using a lentivirus transduction system. Neurons were transduced with control or Ctsd shRNA virus at DIV 6, and proteins were analyzed at DIV 11. ACTB was used as a loading control. (B) Representative images of control and CTSD knockdown neurons stained with LysoTracker (red) and LysoSensor (blue). Neurons were transduced with control or Ctsd shRNA virus at DIV 6 and were incubated with LysoTracker and LysoSensor for 1 h at DIV 11. Scale bar: 20 μm. (C) Quantification of LysoSensor fluorescence intensity in control and CTSD knockdown neurons (n = 6). (D) Rate of proteolysis of long-lived proteins in control and CTSD knockdown neurons (n = 3). (E) Representative western blot of ubiquitinated proteins obtained from Triton X-100-insoluble fractions from control and CTSD knockdown neurons. (F) Quantification of ubiquitinated protein levels, normalized to ACTB (n = 3). (G) Western blot and (H) quantification of LAMP1 levels in control and CTSD knockdown neurons. (I) Western blot and (J) quantification of LAMP2 levels in control and CTSD knockdown neurons. LAMP1 and LAMP2 levels were normalized to corresponding ACTB (n = 3 for each group). Data represent mean ± SEM. One-way ANOVA, followed by Tukey’s post hoc test was used for statistical analyses. *P < 0.05, **P < 0.01 and ***P < 0.001 vs indicated group. ns, not significant.

CTSD protects against the OGD-induced lysosomal dysfunction and cell death

We used the lentiviral-mediated CTSD overexpression in the mouse neurons to assess whether replenishing CTSD in neurons exposed to OGD can restore lysosomal function. Lentiviral transduction in the mouse cortical neurons resulted in approximately 2.2- and 7.8-fold increase in the expression of the 48 and 34 kDa forms of CTSD, respectively (Figure 7A). LysoTracker Red DND-99 and LysoSensor Blue DND-167 staining showed that the uptake of LysoSensor Blue DND-167 in the LysoTracker-positive puncta was significantly higher in neurons with CTSD overexpression after OGD (Figures 7B–C). Proteolysis rates of long-lived proteins in OGD neurons with CTSD overexpression were substantially increased compared to the OGD neurons transduced with the control virus (Figure 7D). Furthermore, CTSD overexpression significantly increased autophagic flux as assessed by LC3-II clearance (Fig. S5A and S5B). Likewise, the accumulation of ubiquitinated proteins (Figures 7E–F) and SQSTM1 (Fig. S5 C and S5D) was significantly decreased in Triton X-100-insoluble fractions of OGD-exposed neurons with CTSD overexpression. In addition, CTSD overexpression decreased the OGD-mediated accumulation of LAMP1 and LAMP2 levels (Figures 7G-J). We also assessed whether CTSD overexpression reduces the LC3B and SQSTM1 puncta in neurons following OGD using immunostaining. Our results demonstrated that the punctate staining of LC3B and SQSTM1 was drastically decreased in CTSD-overexpressing OGD neurons (Fig. S5E-H).

Figure 7.

CTSD conserves lysosomal function in OGD neurons

(A) Representative western blot of overexpression of CTSD in neurons using lentivirus. ACTB was used as a loading control. Neurons were transduced with control (Con Lenti) or CTSD overexpression (Ctsd Lenti) viruses at DIV 6 and neurons were lysed at DIV 11 to analyze CTSD levels. (B) Confocal images of lysosome staining. Neurons were transduced with Con Lenti or Ctsd lentiviruses at DIV 6 and were stained with LysoTracker (red) and LysoSensor (blue) at DIV 11 after 2 h post OGD. Scale bar: 20 μm. (C) Quantification of LysoSensor fluorescence intensity (n = 6). (D) Assessment of the rate of proteolysis of long-lived proteins in control and OGD neurons transduced with control or Ctsd lentiviruses. Neurons were transduced with the virus at DIV 6 and exposed to OGD at DIV 11. Rate of proteolysis was measured at 6 h of OGR using pulse-chase assay (n = 3). (E) Western blot showing expression of ubiquitinated proteins in Triton X-100-insoluble fractions of control and OGD neurons transduced with control or Ctsd lentiviruses. Cell lysates were collected at 6 h of OGR after induction of OGD at DIV 11. (F) Quantification of ubiquitinated protein levels, normalized to ACTB (n = 3). (G) Western blot and (H) quantification of LAMP1, and (I) western blot and (J) quantification of LAMP2 in control and OGD neurons transduced with control or Ctsd lentiviruses. Cell lysates were collected at 6 h of OGR after induction of OGD at DIV 11 to analyze LAMP1 and LAMP2 levels (n = 3). Data represent mean ± SEM. Statistical comparisons were performed using two-way ANOVA followed by Tukey’s post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001 vs indicated group.

We assessed whether CTSD could protect neurons against OGD-mediated cell death. Mouse cortical neurons were transduced with lentiviruses expressing (Ctsd Lenti), knockdown (Ctsd shRNA) CTSD, or the corresponding controls at DIV 6 followed by OGD treatment at DIV 11. To assess that overexpressed CTSD localization is similar to that of the endogenous CTSD, we used the immunostaining of flag-tagged CTSD that we delivered in the mouse cortical neurons via lentiviral transduction. We observed that exogenous CTSD localized to LAMP1, suggesting correct subcellular distribution (Fig. S6). The overexpression of CTSD did not alter the cell viability in control neurons, while knockdown of CTSD induced significant cell death in control neurons (Figure 8A). OGD for 90 min resulted in a significant loss of cell viability, assessed 24 h after OGD by Alamar Blue assay or Hoechst and propidium iodide (PI) double staining (Figures 8B-D and S1A) as previously shown [45]. Together, CTSD overexpression resulted in significant protection against OGD-mediated cell death in the mouse cortical neurons (Figures 8B-D).

Figure 8.

CTSD protects neurons against OGD treatment

(A) Cell viability in neurons transduced with control, Ctsd shRNA, or Ctsd lentiviruses. Neurons were transduced with viruses at DIV 6, and the assay was performed on DIV 11 using Alamar Blue reagent. Results are expressed as a percent of control (n = 5). (B) Schematic diagram of the experimental design (Top). Bellow: Cell viability in control and OGD neurons. Neurons were transduced with viruses at DIV 6 and exposed to OGD at DIV 11. Cell death assay was performed 24 h after OGR. Results are expressed as a percent of control (n = 5). (C) Schematic diagram of the experimental design showing the time point of Hoechst 33,258 (blue) and propidium iodide (PI) (red) staining (Top). Bellow: Representative images showing Hoechst 33,258- and PI-labeled cells in control and OGD neurons. Scale bar: 20 μm. After transduction with viruses at DIV 6 neurons were exposed to OGD on DIV 11 and stained with Hoechst 33,258 and PI 24 h after OGR. (D) PI-positive cells were quantified from 5 independent experiments from each group and calculated as a percent of total cells. Data are mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001 vs indicated groups, calculated with two-way ANOVA followed by Tukey’s post hoc test. ns, not significant.

We also assessed the CTSD neuroprotection in vivo following stereotaxic lentiviral injections of CTSD in the mouse cortex, followed by 90 min MCAO. We collected tissue from the injected site 3 d after injection to assess CTSD overexpression. Western blots and immunofluorescence data showed that CTSD was overexpressed in the brain following stereotaxic lentiviral injections (Fig. S7A and S7B). We subjected these mice to MCAO 4 d after stereotaxic injections and collected mouse brains for Nissl staining. Due to the limited spread of lentivirus in the brain that expressed well along and around the needle track, the evaluation of infarct volume in the mice would be difficult. We used Nissl staining to assess whether CTSD overexpression preserves more neurons along the needle track as compared to control viral injection. Our data showed more Nissl-positive neurons in the brain following MCAO in CTSD-injected mice as compared to the control virus-injected brain (Fig. S7 C). Together, these data indicate that the loss of CTSD leads to lysosomal dysfunction in neurons following OGD and stroke and that the restoration of CTSD in neurons is neuroprotective. These results suggest that strategies or drugs that can increase CTSD expression in neurons can protect against stroke-mediated brain injury.

Discussion

The main findings of this study are that CTSD is required for lysosomal function and survival of neurons and that defects in CTSD following stroke led to lysosomal dysfunction and cell death. The lysosomal function is a critical physiological process in cells, and any impairment in this vital cellular process leads to the accumulation of protein aggregates and toxic waste in cells. Such protein aggregates are the hallmarks of many neurodegenerative diseases, including AD, PD, HD, ALS, and FTD [46–48], and lysosomal dysfunction or the loss of CTSD contributes to protein aggregation in these neurodegenerative diseases. Emerging evidence suggests that protein aggregations also occur in ischemic stroke brain and play a significant role in neuronal death [26,49–51]. Our data show that lysosomal dysfunction via defects in CTSD is the cause of protein aggregation in neurons after stroke. Lysosomal degradation and proteasomal degradation together constitute the major pathways to clear damaged and misfolded proteins [52,53]. Lysosomal degradation supplements the ubiquitin-proteasome pathway to remove damaged proteins in a stroke brain [27,54,55]. Previously, the role of CTSD in regulating lysosomal function in stroke was not known. This study is the first to show that lysosomal function is compromised in neurons after stroke with a concurrent decrease of CTSD activity and protein levels. While other lysosomal cathepsins, such as CTSB or CTSL, may play a role in lysosomal function, our data demonstrate that CTSD is a major lysosomal protease playing a significant role in regulating lysosomal function in neurons after OGD or stroke. This pathological event leads to an increase in the markers of lysosomal dysfunction with subsequent defective proteolysis of long-lived proteins and accumulation of ubiquitinated proteins.

CTSD is a vital protease in neuronal lysosomes. Neuronal lysosomes lacking CTSD are proteolytically inactive. Supporting this notion, our results demonstrate that CTSD is essential for maintaining lysosomal function, as depletion of CTSD alone decreases the proteolytic activity of lysosome, resulting in the accumulation of ubiquitinated proteins in neurons. We further demonstrate that CTSD improves lysosomal function, clears protein aggregates, and provides neuroprotection. Lysosomes contain other cathepsins, but they do not complement each other’s roles due to substrate specificities [56]. CTSD is the only protease that is ubiquitously present in the lysosomes of all cells in the body [56] and plays an essential role in the degradation of long-lived proteins and other contents of the autophagosomes [22]. Consistent with these observations, our data indicate that the loss of CTSB is not sufficient to cause lysosomal dysfunction in neurons; therefore, CTSD is the most credible target to restore lysosomal function in neurons following stroke. A recent study showed that not all lysosomes contain CTSD in neurons. CTSD-positive lysosomes are the only lysosomes capable of performing a proteolytic activity in neurons [18]. This finding further signifies the importance of CTSD in maintaining lysosomal activity in neurons. Our data obtained in OGD-treated neurons and mouse stroke brains suggest that lysosomal dysfunction in neurons is likely due to the depletion of CTSD. Studies show that the loss of CTSD is involved in lysosomal dysfunction and proteinopathies in neurodegenerative disorders. Our results demonstrating a decrease in CTSD, a defect in lysosomal proteolysis, and an accumulation of ubiquitinated proteins, seen as early as 1 h after stroke, confirm the prominence of this pathway in mediating cell death in stroke. Therefore, restoring autophagy in stroke may be therapeutically essential to salvage brain tissue after stroke.

Autophagosomes are an integral part of the lysosomal protein degradation pathway. Autophagosomes engulf cytosolic proteins and organelles during autophagy. In the process of autophagosome formation, LC3-I is conjugated to phosphatidylethanolamine to form LC3-II [57,58]. LC3-II is then recruited to autophagosomes membranes. Intra-autophagosomal content and LC3-II in the autophagosomal lumen are degraded upon the fusion of the autophagosomes to lysosomes [59]. Therefore, the turnover of LC3-II is a reliable marker for autophagy. Alteration in LC3-II is observed in various diseases, including AD, PD, and LSD [60–62]. Our results demonstrate that in neurons exposed to OGD or in the stroke brain, LC3-II accumulates over time. However, an increase in LC3-II could be either due to excessive initiation of autophagy or poor lysosomal clearance. We see defects in lysosomal function after stroke or OGD; thus, it is likely that the accumulation of LC3-II in our study is due to lysosomal dysfunction. Our data that 3-MA in neurons undergoing OGD decreases LC3-II suggest that the initiation of autophagy still occurs. However, the observation that OGD further decreases LC3-II levels in combination with the inhibition of lysosomes with CQ or NH₄Cl suggests that autophagic flux is attenuated after stroke. The accumulation of insoluble SQSTM1 and ubiquitinated proteins further indicates defects in the autophagy-lysosome clearance pathway in stroke or OGD.

The initiation of autophagosome formation in cells with defective lysosomal clearance is detrimental [63]. A recent report that the autophagy-lysosomal pathway is defective in a permanent occlusion stroke model [29] supports our data. Inadequate lysosomal activity and the accumulation of autophagosomal contents can aggravate cell death by depleting cellular energy and nutrients [64–66]. Therefore, the defects in lysosomal clearance via CTSD alteration in stroke should be considered a critical pathological process since the bioenergetic status in the ischemic brain is already in crisis [67]. Supporting the role of CTSD, our data show that replenishing CTSD via overexpression in neurons can restore lysosomal function and clearance of long-lived proteins and protects neurons against OGD. Whereas, the loss of CTSD via shRNA-mediated knockdown aggravates the lysosomal dysfunction and cell death in OGD, suggesting that an intervention targeting to increase CTSD function in stroke may be protective. Supporting this notion, a recent study demonstrates that a single mutation or partial loss of CTSD is sufficient to induce lysosomal dysfunction and cell death in lysosomal storage disease and alpha-synucleinopathies [22,68].

In summary, this study demonstrates that the loss of CTSD in stroke leads to lysosomal dysfunction and subsequent increase in protein aggregates, leading to cell death in stroke. Maintaining CTSD activity by pharmacological intervention can improve the lysosomal function and protect the brain against stroke-induced cell death. Further studies are warranted to know the molecular mechanisms that lead to the loss of CTSD in stroke so that pharmacological molecules can be developed to restore this vital process and protect the brain from dying after a stroke.

Materials and methods

Primary cortical neuronal culture

The Institutional Animal Use and Care Committee at the University of Alabama at Birmingham approved all the experiments in this study. Mouse primary cortical neurons were cultured from embryonic day 15 mouse embryos, as previously described [69]. Briefly, the cortices of the embryonic brains were aseptically dissected, freed of meninges and dissociated in dissecting medium (DMEM + 20% horse serum [Thermo Fisher Scientific, 26050088]). After washing with serum-free DMEM medium (Thermo Fisher Scientific, 11885084) and trypsin digestion at 37°C for 5 min, the cortices were in trituration using a glass pipette in custom Neurobasal medium (Thermo Fisher Scientific, A2477501) supplemented with 10 mM glucose (Sigma-Aldrich, G7528), 1 mM GlutaMax (Thermo Fisher Scientific, 35050061), 1 mM sodium pyruvate (Thermo Fisher Scientific, 11360070), and B-27 (Thermo Fisher Scientific, 17504044). Following straining through a 40-µm filter, the homogeneous cell suspension was plated at a density of 5 × 10⁵ cells/ml on cell culture plates coated with poly-L-ornithine. (Sigma Aldrich, P3655). On the day in vitro (DIV) 2, the cultures were treated with 5-fluoro-2-deoxyuridine (40 μM; Sigma Aldrich, F0503) to inhibit glial growth and proliferation. Mature neurons representing more than 95% of the neurons in the culture were used at DIV 11–12 for OGD studies.

Oxygen-glucose deprivation (OGD) in primary cortical neurons

For OGD, the culture medium from neurons was removed and saved. After that, the neurons were washed two times with phosphate-buffered saline (PBS; Thermo Fisher Scientific, 70011–044) pre-warmed at 37°C. OGD treatment was initiated by incubating the neuronal cultures for 90 min with an OGD buffer (NaCl 116 mM, KCl 5.4 mM, MgSO₄ 0.8 mM, NaHCO₃ 26.2 mM, NaH₂PO₄ 1 mM, CaCl₂ 1.8 mM, glycine 0.01 mM, pH 7.4) pre-bubbled with OGD gas (5% CO₂, 10% H₂ and 85% N₂) in a hypoxia chamber at temperature 37°C, connected to an O₂ sensor/monitor (Biospherix Ltd. NY, USA). The O₂ in the chamber was kept at less than 1% by the continuous flow of the OGD gas through the chamber. OGD was terminated by resupplying the neuronal media that was saved in the beginning before washing the cultures for OGD and transferring the culture back to the regular incubator. Thus, oxygen-glucose resupply (OGR) was achieved by returning neurons to normoxic conditions (5% CO₂ and 95% air) and supplying glucose-containing media for indicated time points. Neurons were treated with lysosome inhibitors, ammonium chloride (NH₄Cl, 10 mM; Sigma Aldrich, 254134), or chloroquine (CQ, 10 µM; Sigma Aldrich, C6628) at the onset of OGR or autophagosome biogenesis inhibitor, 3-methyladenine (3-MA, 10 mM; Sigma Aldrich, 189490) at the onset of OGD.

Middle Cerebral Artery Occlusion (MCAO)

Transient cerebral ischemia was induced by MCAO, as described previously with minor modifications [38]. Briefly, mice were anesthetized with 1–3% isoflurane and maintained at the body temperature was maintained at 37°C. A 7–0 silicone-coated nylon monofilament (Doccol Corporation, MA, USA) was introduced into the right internal carotid artery (ICA) through a nick in the right external carotid artery. The filament was slowly pushed into the anterior carotid artery to block the blood flow to the middle cerebral artery. Laser-Doppler flowmetry probe, placed on the thinned skull over the lateral parietal cortex, confirmed the blockage of blood flow. After 90 min of occlusion, the filament was slowly removed to allow reperfusion. The brains of MCAO mice were collected at different time points for biochemical analysis or stained with the vital dye, TTC (Sigma Aldrich, T8877), to analyze infarct volume 24 h after reperfusion.

Cloning and plasmids

We used mouse cDNA as a template to synthesize Ctsd by PCR. The PCR product was cloned in lentiviral vector pLVX-Puro (Clonetech, 632164) [70] at XhoI and BstBI restriction sites. The sequence of primers with restriction sites are: Forward: 5ʹ- TAATACTCGAGATGAAGACTCCCGGCGTCTTGC-3ʹ Reverse: 5ʹ-

GCGTATTCGAATTAGAGTACGACAGCATTGGCAAAG-3ʹ. Primers were purchased from Thermo Fisher Scientific. To knockdown Ctsd or Ctsb in primary mouse neurons, shRNA against CDS of Ctsd or Ctsb were purchased from Sigma Aldrich. The sequence of shRNA for Ctsd used is 5ʹ-CCGGCCTCTTATCCAGGGTGAGTATCTCGAGATACTCACCCTGGATAAGAGGTTTTT-3ʹ (Mission shRNA, TRCN0000030490) and for Ctsb is 5ʹ-CCGGCCTTTGATGCACGGGAACAATCTCGAGATTGTTCCCGTGCATCAAAGGTTTTTG-3ʹ (Mission shRNA, TRCN0000030635). The shRNAs were supplied in lentiviral vector pLKO.1.

Production and transduction of lentivirus

High titer lentivirus was generated by transfecting HEK293 FT (Thermo Fisher Scientific, R70007) cells with a mixture containing three plasmids and 25 kDa linear polyethyleneimine (PEI; Polysciences, Inc., 239662) [63]. In brief, 2.5 µg pMD2.G (Addgene, 12259: Didier Trono) and 6.5 µg psPAX2 (Addgene, 12260: Didier Trono) vectors were mixed with 3 µg pLVX-Puro plasmid containing the target genes. The mixture was diluted in 1.5 ml Opti-MEM (reduced serum) (Thermo Fisher Scientific, 11058021) medium and incubated for 5 min. In a 15-ml falcon tube, PEI was mixed gently with 1.5 ml of Opti-MEM. The plasmid mixture and diluted PEI were mixed and incubated for 20 min at room temperature. The plasmid-PEI mixture was added to the HEK293 FT cells (70–80% confluence) grown in a T-175 flask and incubated in 5% CO₂ at 37°C overnight. The original medium was replaced with fresh medium (DMEM, 1x MEM Non-essential amino acid [Thermo Fisher Scientific, 11140050] and fetal bovine serum [Thermo Fisher Scientific, 10082147]) 18–20 h after transfection. The supernatant containing the first batch of the lentivirus was collected 24 h after the replacement of the medium. This step was repeated and the second batch of lentivirus was collected after 48 h. The two batches of lentivirus were combined and filtered through a 0.45-μm filter. To concentrate the lentivirus, the filtrate was placed in a centrifuge tube containing Opti-prep (~4 ml; Sigma Aldrich, D1556) at the bottom as the cushion and centrifuged at 50,000 × g for 2 h using an SW32Ti rotor (Beckman Coulter, CA, USA). After centrifugation, a layer containing the lentiviral particles located between the medium and Opti-prep was collected and placed in a 50-ml falcon tube. Culture medium was added to the tube to increase the volume to 50 ml. A second centrifugation was done at 5000 × g overnight at 4°C. The pellet containing the lentiviral particles was resuspended in ice-cold PBS (Thermo Fisher Scientific, 10010023; pH 7.4) and stored as 10-μl aliquots at −80°C. Primary cortical neurons were transduced with the lentivirus at DIV 6. The titers for control virus (empty pLVX vector), Ctsd virus (vector carrying Ctsd CDS), control shRNA virus, Ctsd shRNA virus and Ctsb shRNA virus were 1.4x10⁹, 8x10⁸, 1.6x10⁹, 1.4 × 10⁹ and 2.3 × 10⁹ TU/ml, respectively. For the overexpression of CTSD, neurons at DIV 6 were transduced with control or CTSD virus at a multiplicity of infection (MOI) of 10 and for the knockdown, they were transduced with control, or Ctsd or Ctsb shRNA virus at an MOI of 15.

Subcellular fractionation

For subcellular fractionation, mouse cortical neurons at DIV 11 were suspended in a subcellular fractionation buffer (250 mM sucrose, 20 mM Tris-HCl, pH 7.4, 1 mM EGTA, 1 mM EDTA with protease and phosphatase inhibitor mixture) and passed through a 30 G needle (10x) to break the cells. For subcellular fractionation (cytoplasm and lysosome) of brain ipsilateral cortex from sham and MCAO mouse, the tissue was first homogenized in the subcellular fractionation buffer using a pre-chilled dounce homogenizer. The lysates were centrifuged at 750 × g for 10 min at 4°C to clear cell debris and nuclei. The supernatant was further centrifuged at 20,000 x g for 5 min at 4°C to collect the cytosolic fraction. The pellet was then resuspended in subcellular fractionation buffer and laid on top of discontinuous sucrose (Fisher Scientific, BP220-212) density gradients (from top to bottom, 17, 20, 23, 27 and 30%) followed by centrifugation at 145,000 x g for 2 h at 4°C. The lysosome fractions were collected from 1/10^th of the gradient volume and further mixed with 2 volumes of PBS and centrifuged for 30 min at 18,000 x g at 4°C. The pellet collected contains pure lysosomes. Equal amounts of proteins from each fraction were analyzed by western blotting.

Stereotaxic injection of lentivirus

The stereotactic delivery of the lentivirus into the cortex was performed in anesthetized mice using ketamine-xylazine (75 mg/kg ketamine, 10 mg/kg xylazine i.p.). Lentiviral particles (3 μl of ~10⁹ titer viral particles) were injected using stereotaxic coordinates, 0.5 mm rostral, 1 mm lateral and 2.0 mm ventral from bregma, using a 26 G Hamilton syringe mounted on a stereotaxic stand and coupled with an automatic injector (Stoelting Co., IL, USA). The virus was injected at the rate of 0.5 μl/min followed by 10 min needle wait time before completing the injection. The MCAO surgery on the mice was performed 4 d after the viral injections.

Immunoblotting

The neuronal cells or ipsilateral cortex from MCAO or sham mouse brains were lysed using lysis buffer (1% Triton X-100 [Fisher Scientific, BP151-500], 25 mM Tris HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, supplemented with Halt protease and phosphatase inhibitor cocktail [Thermo Fisher Scientific, 78440]), and the cell lysates were centrifuged at 12,000 × g for 15 min at 4°C. The pellets (Triton X-100-insoluble fractions) were washed three times with lysis buffer and resuspended in 1X Laemmli buffer and centrifuged at 12,000 × g for 15 min. BCA protein assay kit (Thermo Fisher Scientific, 23225) was used to measure total protein concentration. Aliquots of proteins (15 µg) from each fraction were loaded separately in each well of SDS-PAGE gel (4–15%) and resolved using a running buffer (Bio-Rad Laboratories, 1610772) for approximately 90 min at 100 V, and then transferred onto a 0.22 μm nitrocellulose membrane (Bio-Rad Laboratories, 1620112) at a constant voltage (100 V) for 90 min using western transfer buffer (Bio-Rad Laboratories, 1610771). The membranes were then blocked with 5% (w:v) skimmed milk in Tris-buffered saline (Fisher Scientific, L15846) with 0.1% Tween 20 (Fisher Scientific, 12247) (TBST). After blocking, the membranes were washed three times (10 min each) with TBST and then probed with primary antibodies against CTSD (Genetex, GTX62063 and Proteintech, 231327-1-AP), LAMP1 (DSHB, 1D4B), LAMP2 (DSHB, ABL-93), CTSB (Cell Signaling Technology, 31718), ubiquitin (Cell Signaling Technology, 3933), SQSTM1 (Cell Signaling Technology, 23214), LC3-I/II (Cell Signaling Technology, 4108), CTSL, (Novus Biologicals, NB100-1775). Following overnight incubation at 4°C with the diluted antibodies (1:1000 dilution) in TBST containing 5% milk, the membranes were probed with horseradish peroxidase-conjugated secondary antibodies (Cell Signaling Technology, 74074 S and 7077) for 1 h at room temperature. Images were obtained using SuperSignal West Pico Plus Chemiluminescent Substrate (Thermo Fisher Scientific, 34578) via a ChemiDoc MP imaging system (Bio-Rad Laboratories, CA, USA)

Immunocytochemistry

Primary cortical neurons were grown on poly-L-ornithine (Sigma Aldrich, P3655)-coated glass coverslips for immunofluorescence studies. Cells were washed twice with ice-cold PBS then fixed in 4% (w:v) paraformaldehyde for 15 min at room temperature. After washing three times with PBS, cells were permeabilized with 0.2% (v:v) Triton X-100 in PBS with 10% donkey serum (Sigma Aldrich, D9663) for 30 min. Cells were then blocked in PBS with 10% donkey serum for 1 h at room temperature in the dark. Afterward, the cells were incubated in primary antibodies against CTSD (Genetex GTX62063), CTSB (Cell Signaling Technology, 31718), Flag Tag (Proteintech, 66008-2-Ig), LAMP1 (DSHB, 1D4B), LC3B (Santa Cruz Biotechnology, sc-376404), SQSTM1 (Novus Biologicals, H00008878-M01) overnight at 4⁰C. After incubation with primary antibodies, cells were washed three time with PBS and incubated for 1 h in the dark at room temperature with the Alexa Fluor 488- and Alexa Fluor 555-conjugated secondary antibody (Thermo Fisher Scientific, A21206 and A31570; in PBS containing 1% donkey serum. Afterward, cells were washed and counterstained with DAPI (300 nM) for nuclear staining. After final washes with PBS, the cells on glass coverslips were mounted onto glass slides using Immuno-Mount (Thermo Fisher Scientific, 9990402). Images were acquired using Zeiss 710 confocal microscope with a 63x oil immersion 1.35 NA objective (Wetzler, Hassen, Germany). Images were captured by randomly selecting 4 visual fields pooled from 6 biological replicates using the same setting parameters. Intensity measurements were performed using Zen software (Zeiss).

Immunohistochemistry

For immunohistochemistry, the mouse brains were perfusion-fixed with 4% paraformaldehyde (PFA). Following perfusion fixation, the brains were cryoprotected and free-floating sections (40 µm) were collected. The sections were washed in PBS (×3), permeabilized for 1 h with 0.3% Triton X-100 in PBS containing 10% donkey serum followed by overnight incubation at 4°C with primary antibodies against RBFOX3/NeuN (Millipore, MAB377), GFAP (Proteintech, 60190-1-Ig), CTSD (Proteintech, 231327-1-AP), in a mixture of PBS containing 0.3% Triton-X and 10% donkey serum. The sections were then washed in PBS (3 × 10 min) followed by incubation with Alexa Fluor 488- and Alexa Fluor 555-conjugated secondary antibody (Thermo Fisher Scientific, A21206 and A31570; 1:1000 dilution) in 1% donkey serum in PBS for 1 h at room temperature in the dark. Thereafter, the sections were washed and counterstained with DAPI (300 nM) for nuclear staining. After final wash with PBS, the sections were mounted onto glass slides using Immuno-Mount (Thermo Fisher Scientific, 9990402). Images were captured using a Zeiss 800 confocal microscope (Wetzler, Hassen, Germany).

CTSD and CTSB activity assay

CTSD activity was measured using a CTSD activity assay kit (Biovision, Inc., K143-100) following the manufacturer’s instructions. Briefly, lysates from control and OGD/OGR neurons, and sham and stroke brains, were collected at indicated time points using CTSD lysis buffer. Following centrifugation at 12,000 × g, for 10 min at 4⁰C, the supernatants were collected and quantified for protein levels. An equal amount of protein was used to measure CTSD activity. The reaction mixture contained a CTSD reaction buffer, cell lysates, and an internally quenched fluorescent substrate (included in the kit). Because other enzymatic activities might cleave the substrate, samples were also measured in the presence of pepstatin A (Sigma Aldrich, P5318) to inhibit CTSD activity. Activity in the presence of pepstatin A (non-CTSD activity in the samples) was subtracted from the total activity to obtain CTSD activity. CTSB activity was measured using the CTSB activity assay kit (Biovision, Inc., K140-100) using the method described above. Z-FA-FMK (Santa Cruz Biotechnology, sc-3131) was used as a CTSB inhibitor. All samples were assessed in duplicate from three biological replicates. Data is presented as the fold-change of control.

Real-time polymerase chain reaction

Total RNA extracted from DIV 11 primary cortical neurons was used for quantitative real-time polymerase chain reaction (qRT-PCR), using a commercially available kit (PureLink RNA Mini Kit, Thermo Fisher Scientific, 12183018A), following the manufacturer’s instructions. An Agilent 2100 Bioanalyzer (Agilent Technologies, CA, USA) was used to assess RNA integrity. Quantification of RNA was determined using a Nanodrop ND-2000 spectrophotometer (Thermo Fisher Scientific, CA, USA). Complementary DNA (cDNA) was prepared from 1 µg of RNA using the iScript cDNA Synthesis Kit (Bio-Rad Laboratories, 1708890) following the manufacturer’s instructions and was stored at – 20⁰C for subsequent analysis. Quantitative PCR was carried out using PowerUp SYBR Green Master Mix (Applied Biosystems, A25741) following the manufacturer’s instructions. Mouse primers used in the study were as follows Ctsd (Forward, 5ʹ-CGCAGTGTTTCACAGTCGT-3ʹ; Antisense, 5ʹ- TGAGCCGTAGTGGATGTCAA-3ʹ), Gapdh (Forward, 5ʹ- AACTTTGGCATTGTGGAAGG-3ʹ; Antisense, 5ʹ-ACACATTGGGGGTAGGAACA’). All primers were purchased from Thermo Fisher Scientific. Data were normalized to Gapdh, and mRNA abundance was calculated using the 2-ΔΔCT method. The expression levels of Ctsd mRNA are presented as fold increase of control.

Lysosomes staining with LysoTracker and LysoSensor

LysoTracker red DND-99 (Thermo Fisher Scientific, L7528) and pH-sensitive LysoSensor Blue DND-167 (Thermo Fisher Scientific, L7533) probes were used to visualize the lysosomes and determine lysosomal activity, respectively. LysoSensor Blue shows a pH-dependent increase in fluorescence intensity, while LysoTracker red exhibits fluorescence mostly independent of pH. To label the lysosomes, neurons were incubated with LysoTracker (1 µM) and LysoSensor (1 µM) in the Neurobasal medium for 1 h at 37⁰C. After incubation, the medium was aspirated and cells were washed twice in HBSS (Thermo Fisher Scientific, 14175–079) to remove the unbound probes. Live-imaging was captured using a Zeiss Axioobserver microscope (Wetzler, Hassen, Germany). Images were captured from 6 biological replicates for each group using the same setting parameters. Phase-contrast images were taken to visualize individual cells. A region of interest (ROI) for each cell with LysoTracker staining was defined using Zen software, and the values for six experiments were plotted in Excel.

NeuroTrace 435/455 Blue Fluorescent Nissl staining

Nissl staining in sham and MCAO brain sections was performed according to the instruction provided by the manufacturer. Briefly, brain sections (40 µm) fixed in paraformaldehyde (4% PFA) were washed in PBS (×3) and permeabilized with 0.3% (v:v) Triton X-100 in PBS for 1 h. Thereafter, the sections were incubated in NeuroTrace 435/455 Blue Fluorescent Nissl Stain solution (Thermo Fisher Scientific, N-21479) diluted in PBS for 30 min at room temperature. The sections were then washed with PBS (3 × 20 min) and mounted onto glass slides using Immuno-Mount (Thermo Fisher Scientific, 9990402). Images were captured using a Zeiss Axioobserver microscope with 10 × 1.35 NA objective (Wetzler, Hassen, Germany).

Proteolysis of long-lived protein

Proteolysis of long-lived proteins was performed based on the method described previously [30]. Briefly, neurons were incubated in complete Neurobasal medium supplemented with 1 μCi/ml [³ H]-leucine (Perkin Elmer, NET1166001MC) for 48 h at 37°C. Next, the neurons were incubated in fresh medium supplemented with 10 mM cold leucine (Sigma Aldrich, 61819) for 18 h. New media was added, and incubation was continued for an additional 2 h. Thereafter, the neurons were subjected to OGD and samples were collected at different time points for proteolysis analysis. The neuronal culture plates were briefly incubated on ice for 2 min followed by washing with PBS. Ice-cold PBS (50 µl) containing 2% bovine serum albumin (BSA, Sigma Aldrich, A9418) and ice-cold 25% TCA (Sigma Aldrich, T0699; 450 µl) was added to each well of the plates. The neuronal plates were then incubated at 4°C overnight with gentle shaking. The extract from each well was collected in 1.5 ml Eppendorf tubes and centrifuged at 6,500 × g for 15 min at 4°C. The TCA-soluble fractions (supernatant) were collected and transferred to scintillation tubes and mixed with 4 ml of EconoSafe (Research Product International, A4552) by vortexing. The TCA-insoluble fractions both in the Eppendorf tubes and the wells were dissolved in 500 µl of 0.2 M KOH and transferred to scintillation tubes and mixed with 4 ml Opti-Fluor (Perkin Elmer, 6013199) by vortexing. Radioactivity of TCA-soluble and -insoluble fractions were determined by liquid scintillation counting in a scintillation counter (Beckman, GA, USA, LS6500). Protein degradation rates for long-lived proteins was calculated as the percentage of radioactivity in the TCA-soluble fraction relative to the total radioactivity (i.e., TCA-soluble + insoluble fractions), divided by the incubation time.

Cell death assays

Cell viability was determined by Alamar Blue (Thermo Fisher Scientific, DAL1100) assay, and by Hoechst 33342 (Thermo Fisher Scientific, 62249) and propidium iodide (Thermo Fisher Scientific, P3566) staining, two complementary methods for assessing cell death. For Alamar Blue assay, the neurons were incubated in the Alamar Blue reagent 10% in the Neurobasal medium for 3 h. Following the 3 h incubation, 100 μL of the medium was collected from each well and transferred to a 96-well microplate. The fluorescence was measured at the excitation and emission wavelength of 540 and 595 nm, respectively, using Victor X5microplate reader (Perkin Elmer, MA, USA). For Hoechst and propidium iodide staining, neurons in culture medium were incubated for 5 min with 1 µM Hoechst 33342 and 5 µg/ml propidium iodide. Images were captured on an Axioobserver fluorescence microscope (Wetzler, Hassen, Germany). ImageJ software (NIH) was used for counting neurons stained with Hoechst and propidium iodide. Cell death was calculated by subtracting the dead neuron (propidium iodide positive) from total neurons (Hoechst positive).

Statistical analysis

For statistical analysis, GraphPad Prism version 6 (San Diego, CA, USA) was used. Western blot quantification was performed using ChemiDoc software. Data were analyzed with one-way or two-way analysis of variance (ANOVA) followed by Tukey’s post hoc test for multiple comparisons or with a Student’s t-test for two group comparisons. Data are expressed as means ± SEM. p < 0.05 was considered statistically significant.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Funding Statement

This work was supported by the National Institute of Neurological Disorders and Stroke [NS086953].

Disclosure statement

The authors declare no competing interests.

Supplementary material:

Supplemental data for this article can be accessed here