Mesenchymal stem cells enhance autophagy and increase β-amyloid clearance in Alzheimer disease models

Abstract

Current evidence suggests a central role for autophagy in Alzheimer disease (AD), and dysfunction in the autophagic system may lead to amyloid-β (Aβ) accumulation. Using in vitro and in vivo AD models, the present study investigated whether mesenchymal stem cells (MSCs) could enhance autophagy and thus exert a neuroprotective effect through modulation of Aβ clearance In Aβ-treated neuronal cells, MSCs increased cellular viability and enhanced LC3-II expression compared with cells treated with Aβ only. Immunofluorescence revealed that MSC coculture in Aβ-treated neuronal cells increased the number of LC3-II-positive autophagosomes that were colocalized with a lysosomal marker. Ultrastructural analysis revealed that most autophagic vacuoles (AVs) in Aβ-treated cells were not fused with lysosomes, whereas a large portion of autophagosomes were conjoined with lysosomes in MSCs cocultured with Aβ-treated neuronal cells. Furthermore, MSC coculture markedly increased Aβ immunoreactivity colocalized within lysosomes and decreased intracellular Aβ levels compared with Aβ-treated cells. In Aβ-treated animals, MSC administration significantly increased autophagosome induction, final maturation of late AVs, and fusion with lysosomes. Moreover, MSC administration significantly reduced the level of Aβ in the hippocampus, which was elevated in Aβ-treated mice, concomitant with increased survival of hippocampal neurons. Finally, MSC coculture upregulated BECN1/Beclin 1 expression in AD models. These results suggest that MSCs significantly enhance autolysosome formation and clearance of Aβ in AD models, which may lead to increased neuronal survival against Aβ toxicity. Modulation of the autophagy pathway to repair the damaged AD brain using MSCs would have a significant impact on future strategies for AD treatment.

Introduction

Alzheimer disease (AD), the most common neurodegenerative disorder, is characterized pathologically by synapse loss and the presence of amyloid-β (Aβ) plaques and tau tangles. ^{1 , 2} Aβ is a peptide consisting of 36 to 43 amino acids that is generated via sequential proteolysis of amyloid precursor protein (APP) by BACE1/β-secretase and γ-secretase. ³ Aβ plays an important role in AD pathogenesis. Aβ is specifically toxic to neurons through a series of downstream events that lead to increased intracellular calcium levels, impaired mitochondrial redox activity, and increased free radical generation, culminating in neuronal dysfunction and death. ⁴ With altered APP processing, failure of Aβ clearance leading to Aβ accumulation underlies AD pathogenesis.

The autophagic pathway delivers intracellular constituents to lysosomes by several processes including chaperone-mediated autophagy, microautophagy, and macroautophagy. Of these, macroautophagy, hereafter referred to as autophagy, is a major intracellular degradation process that acts as the principal mechanism involved in organelle turnover. Autophagy is also a vital pathway for degradation of abnormal and aggregated proteins, and thus acts as a cytoprotective response, particularly under stress or injury conditions. ^{5 - 8} There is ample evidence suggesting that the autophagy pathway is involved in a variety of neurodegenerative diseases, and data from several human pathological studies support the role of autophagy in AD. Lee et al. ⁹ have demonstrated that numerous autophagic vacuoles (AVs) accumulate within dystrophic neuritis in the brains of patients diagnosed with AD. Other pathological studies show that the expression of BECN1/Beclin 1, which plays a key role in autophagy, ^{10 - 14} is markedly reduced in the brains of AD patients. ¹⁵ BECN1 regulates the autophagy-promoting activity of PIK3C3/VPS34, ¹⁶ and is involved in the recruitment of membranes to form autophagosomes. Furthermore, Ma et al. ¹⁷ report that autophagy markers (i.e., ATG5, ATG12, and microtubule-associated protein 1 light chain 3 [LC3]) are associated with plaque and tangle pathologies in AD patients. Taken together, these data suggest that autophagy might be closely coupled with AD pathogenesis.

Mesenchymal stem cells (MSCs) are multipotent stem cells present in adult bone marrow and are capable of differentiating into various cell types under appropriate conditions. ^{18 , 19} Additionally, MSCs secrete various cytotropic factors that, in turn, exert neuroprotective effects. ²⁰ Our previous animal studies demonstrate that human MSCs have neuroprotective effects through complex mechanisms, such as modulation of neuroinflammation, enhancement of cell survival signals, increased neurogenesis, and modulation of ubiquitinated proteins. ^{21 , 22} To date, no studies have investigated the modulatory effect of MSCs on clearance of toxic proteins through stimulation of autophagy in neurodegenerative diseases. In the present study, we investigated whether MSCs would enhance autophagy and thus exert a neuroprotective effect through modulation of Aβ clearance using in vitro and in vivo AD models.

Results

Aβ treatment leads to induction of autophagosome formation in a dose- and time-dependent manner

LC3, the mammalian ortholog of yeast Atg8, is widely used to monitor autophagy. ^{23 - 25} During autophagy induction, the soluble form of LC3 (LC3-I) is converted to LC3-II, which is associated with autophagosomal membranes and is essential for autophagosome formation. Therefore, the amount of LC3-II is indicative of the number of autophagosomes and serves as a good marker of autophagosome formation. ²⁴ SH-SY5Y cells were treated with different concentrations to evaluate the direct effects of Aβ (Aβ_1–42). Aβ treatment resulted in cellular morphological changes and loss of cellular viability in a dose- and time-dependent manner (Fig. 1A and B). Additionally, Aβ treatment induced marked LC3-II expression in a dose- and time- dependent manner (Fig. 1C and D). However, when SH-SY5Y cells were treated with the reverse control peptide Aβ_42–1 (20 μM) for 24 and 72 h, there were no morphological changes or loss of cell viability. When a lysosomal inhibitor was added to Aβ-treated SH-SY5Y cells, the expression of LC3-II was markedly increased compared with only-Aβ-treated cells (Fig. S1). These results indicate that Aβ enhanced autophagic flux.

Figure 1. Aβ induces autophagosome formation in SH-SY5Y cells. SH-SY5Y cells were exposed to various concentration of Aβ (Aβ_1–42) for 24 or 72 h. Cell viability was decreased in a dose- and time-dependent manner (A and B). However, reverse peptide Aβ_42–1 treatment (20 μM for 24 and 72 h) did not lead to morphological changes or loss of cell viability. Aβ treatment enhanced the autophagy marker (LC3). Western blot analysis showed that Aβ treatment increased expression of LC3-II in a dose- and time-dependent manner (C and D). Values are expressed as mean ± SDs, n = 3. *P < 0.05 for Aβ-treated SH-SY5Y cells vs. untreated SH-SY5Y cells; ^# P < 0.05 for Aβ-treated SH-SY5Y cells at 24 h vs. Aβ-treated SH-SY5Y cells at 72 h.

MSCs increase cellular viability by enhancing autophagy induction in Aβ-treated SH-SY5Y cells

To evaluate the effects of MSCs on SH-SY5Y viability, Aβ-treated SH-SY5Y cells were cocultured with MSCs for 3, 6, 12, or 24 h. SH-SY5Y cell viability did not significantly differ from Aβ-treated SH-SY5Y cells after 3 or 6 h of MSC coculture. However, SH-SY5Y viability was significantly increased following 12 h of MSC coculture compared with baseline (Fig. 2A). Next, we evaluated whether MSCs increased the induction of autophagy in Aβ-treated SH-SY5Y cells. MSCs significantly increased LC3-II expression at 6 and 12 h after MSC coculture as compared with baseline (Fig. 2B and C).

Figure 2. MSCs increase cellular viability and enhance autophagy induction in Aβ-treated SH-SY5Y cells. Aβ-treated SH-SY5Y cells were cocultured with MSCs for 3, 6, 12, or 24 h. MSC coculture significantly increased the viability of Aβ-treated SH-SY5Y cells at 12 h compared with baseline (A). Additionally, MSCs enhanced LC3 expression in Aβ-treated cells compared with Aβ-treated cells. Bar charts illustrate quantification of LC3-II/ LC3-I (B and C). Values are expressed as mean ± SDs, n = 3. *P < 0.05 for Aβ-treated SH-SY5Y cells vs. untreated SH-SY5Y cells; ^# P < 0.05 for Aβ-treated SH-SY5Y cells vs. MSCs-cocultured SH-SY5Y cells.

MSCs enhance autolysosome formation in Aβ-treated SH-SY5Y cells

To determine whether increased autophagosome induction led to enhanced autolysosome formation, we analyzed autolysosomal maturation in SH-SY5Y cells cocultured with MSCs. Immunofluorescence analysis revealed that MSCs cocultured in Aβ-treated SH-SY5Y cells increased the number of LC3-positive autophagosomes (FITC; green) that were colocalized with a lysosomal marker (Cy3; red) at 6 and 12 h as compared with Aβ-treated SH-SY5Y cells (Fig. 3A), indicating that MSC coculture enhanced fusion of autophagosomes and lysosomes. A quantitative analysis of FITC and Cy3 double-positive puncta per cell in Aβ-treated SH-SY5Y cells showed that autolysosome formation was significantly greater in cells cocultured with MSCs compared with Aβ-treated cells (25.4 ± 3.7% vs. 1.2 ± 2.9% and 25.5 ± 1.7% vs. 2.5 ± 2.5%, P < 0.05; Fig. 3B). Additionally, expression of cathepsin B (CTSB) mRNA, the major protease in lysosomes, was significantly increased in cells cocultured with MSCs compared with Aβ-treated cells (Fig. 3C). To examine the modulatory effect of MSCs on lysosomal activity, we incubated Aβ-treated SH-SY5Y cells with chloroquine (CQ), a lysosomal inhibitor that prevents endosomal acidification. ²⁶ CQ treatment in Aβ-treated SH-SY5Y cells significantly attenuated increased cell survival in cells cocultured with MSCs (Fig. 3D). Additionally, we evaluated the effect of MSCs on autolysosome formation using Chinese hamster ovary (CHO) cells that naturally produce Aβ. When CHO cells were cocultured with MSCs for 6 h, the ratio of LC3-II/LC3-I (Fig. 4A and B) and expression of RAB7, which is required for the final maturation of late AVs and fusion with lysosomes, ^{27 , 28} significantly increased compared with cells without MSC coculture. (Fig. 4C and D).

Figure 3. MSCs enhance autolysosome formation. Immunofluorescence showed that MSCs cocultured in Aβ-treated SH-SY5Y cells increased the number of LC3-positive autophagosomes (FITC; green) that were colocalized with lysosomes (Cy3; red) at 6 and 12 h compared with Aβ-treated SH-SY5Y cells (A). A quantitative analysis of FITC and Cy3 double-positive puncta per cell in Aβ-treated SH-SY5Y cells showed that MSC coculture significantly enhanced autolysosome formation compared with Aβ-treated cells (B), data are the mean values of 3 independent experiments, each with no fewer than 100 cells). Additionally, MSC coculture increased lysosomal activity in Aβ-treated SH-SY5Y cells (C), and CQ treatment significantly attenuated cell survival, which was increased when cocultured with MSCs (D). Values are expressed as mean ± SDs, n = 3. *P < 0.05 for Aβ-treated SH-SY5Y cells vs. untreated SH-SY5Y cells; ^# P < 0.05 for Aβ-treated SH-SY5Y cells vs. MSC-cocultured SH-SY5Y cells; ^§ P < 0.05 for MSC-cocultured SH-SY5Y cells vs. CQ and MSC-cocultured SH-SY5Y cells. Scale bars: 5 μm.

Figure 4. MSCs enhance autophagy modulation in CHO cells, stably expressing human APP695. CHO cells were cocultured with MSCs for 6 h. Coculture with MSCs significantly increased expression of LC3 and RAB7 in CHO cells compared with cells without MSCs (A and C). Bar charts illustrate quantification of the LC3-I/II ratio and relative RAB7 expression (B and D). Additionally, coculturing CHO cells with MSCs tended to decrease in intracellular Aβ levels, however, the difference did not reach statistical significance (E). Values are expressed as mean ± SDs (SD), n = 3. *P < 0.05 for CHO cells vs. MSCs cocultured CHO cells.

Electron microscopy (EM) analyses reveal that MSCs upregulate autolysosome formation in Aβ-treated SH-SY5Y cells

We performed EM analyses to ultrastructurally evaluate whether MSC coculture had an effect on autophagy-related structures of AVs and autolysosomes. Morphometric ultrastructural analyses demonstrated that autophagosomes were rarely noted in SH-SY5Y cells without Aβ treatment (Fig. 5A), whereas AVs were abnormally accumulated in Aβ-treated SH-SY5Y cells (Fig. 5B and D). Most AVs observed in Aβ-treated SH-SY5Y cells, however, were not fused with lysosomes. In contrast, MSC coculture in Aβ-treated cells resulted in a large portion of autophagosomes being conjoined with lysosomes, indicating induction of autolysosome formation (Fig. 5C and E).

Figure 5. Ultrastructural analyses reveal that MSCs upregulate autolysosome formation in Aβ-treated SH-SY5Y cells. Autophagosomes were rarely noted in SH-SY5Y cells without Aβ treatment (A). Most autophagic vesicles observed in Aβ-treated SH-SY5Y cells did not fuse with lysosomes (B and C); however, MSC coculture led to a large portion of autophagosomes being conjoined with lysosomes (D and E). Yellow arrows indicate autophagosomes and red arrows indicate autolysosomes. Scale bars: 10,000 nm.

MSCs enhance Aβ clearance through the autophagy-lysosomal pathway

To evaluate whether autolysosomes induced by MSC coculture enhanced Aβ clearance in Aβ-treated SH-SY5Y cells, we assessed Aβ that was colocalized in lysosomes using the 6E10 antibody and intracellular Aβ concentration. Immunofluorescence analysis showed that 6 and 12 h Aβ treatment increased LC3 immunoreactivity (Cy3; red) that was colocalized with Aβ (green) as compared with controls (Fig. 6A). A quantitative analysis of Aβ and LC3 double-positive puncta per cell in Aβ-treated SH-SY5Y cells showed that Aβ within autophagosomes did not change in cells cocultured with MSCs compared with Aβ-treated cells (26.7 ± 2.6% vs. 27.4 ± 3.5% and 23.4 ± 1.7% vs. 23.1 ± 2.7%, respectively; Fig. 6B). However, immunofluorescence analysis showed that 6 and 12 h MSC coculture in Aβ-treated cells increased Aβ immunoreactivity (Aβ: green) that was colocalized in lysosomes (Cy3; red) compared with Aβ-treated-cells (Fig. 6C). A quantitative analysis of Aβ and lysosome double-positive puncta per cell in Aβ-treated SH-SY5Y cells showed that Aβ within lysosomes was significantly greater in cells cocultured with MSCs at 12 h compared with Aβ-treated cells (19.5 ± 3.0% vs. 1.4 ± 2.6%, P < 0.02; Fig. 6D). Because 6E10 recognizes the APP and C-terminal fragments, we further evaluated whether these results could be a consequence of increased APP production. However, the expression of APP in Aβ-treated- SH-SY5Y cells was not altered by MSC coculture (data not shown). The level of intracellular Aβ in Aβ-treated SH-SY5Y cells gradually increased in a time-dependent manner; however, when cocultured with MSCs, intracellular Aβ levels were significantly decreased at each time point compared with Aβ-treated cells (Fig. 6E). Additionally, coculturing CHO cells with MSCs tended to decrease intracellular Aβ levels with no statistically significant difference (Fig. 4E). When bafilomycin A₁ (Baf), a specific V-ATPase inhibitor, was applied in Aβ-treated- SH-SY5Y cells that were cocultured with MSCs, the levels of intracellular Aβ significantly increased and reached a level that was similar to that shown in only-Aβ-treated cells (Fig. S2). These results indicate that MSCs likely enhance Aβ clearance through the autophagy-lysosomal pathway. To determine whether neuronal cells other than MSCs also possess autophagy induction effects, Aβ-treated SH-SY5Y cells were cocultured with SH-SY5Y cells. SH-SY5Y cell coculture did not lead to a significant change in cell viability. Additionally, neither the expression of LC3-II and RAB7 in Aβ-treated SH-SY5Y cells nor intracellular Aβ levels were decreased after SH-SY5Y cell coculture (Fig. S3).

Figure 6. MSCs enhance Aβ clearance through the autophagy-lysosomal pathway. We assessed Aβ colocalized in lysosomes and its intracellular concentrations to determine whether MSCs-induced autophagy enhanced Aβ clearance. Immunofluorescence analysis of Aβ (green) and LC3 (red) double-positive puncta per cell showed that Aβ within autophagosomes was not changed in SH-SY5Y cells cocultured with MSCs compared with Aβ-treated cells (A and B). However, MSCs cocultured in Aβ-treated cells increased Aβ immunoreactivity (green) that was colocalized in lysosomes (red) compared with Aβ-treated cells (C and D). MSC coculture significantly decreased intracellular Aβ levels in Aβ-treated SH-SY5Y cells, where Aβ concentration was gradually increased in a time-dependent manner (E). Data are the mean values of 3 independent experiments, each with no fewer than 100 cells. Values are expressed as mean ± SDs, n = 3. *P < 0.05 for Aβ-treated SH-SY5Y cells vs. untreated SH-SY5Y cells; ^# P < 0.05 for Aβ-treated SH-SY5Y cells vs. MSCs-cocultured SH-SY5Y cells. Scale bars: 5 μm.

MSCs have neuroprotective effects on hippocampal neurons through enhancement of autolysosome formation in Aβ-treated animals

Using an AD animal model, we assessed RBFOX3/NeuN-positive hippocampal neurons in the CA1 subfield to investigate the potential neuroprotective effects of MSCs. Additionally, we attempted to identify transplanted MSCs in the brain using human-specific nuclear mitotic apparatus protein 1 (NUMA1) immunostaining. NUMA1- and human-specific NES/nestin-positive cells were recruited into hippocampal areas in MSCs-administrated mice; however, these cells did not react with the ELAVL-like 4 (ELAVL4) antibody suggesting that MSCs recruited into the brain would not transdifferentiate into neuronal cells (Fig. S4). Immunohistochemical analysis revealed that RBFOX3-positive and Nissl-stained cells in the hippocampus were prominently decreased in Aβ-treated mice compared with controls (Fig. 7A and B) and that MSC administration in Aβ-treated mice markedly increased the survival of hippocampal neurons (Fig. 7C). Stereological analysis revealed a decreased number of hippocampal neurons in Aβ-treated mice relative to controls and a much greater increase in the number of RBFOX3-positive cells in MSCs-administrated mice compared with Aβ-treated mice (Fig. 7D).

Figure 7. MSCs exert neuroprotective effects in an AD animal model by enhancing degradation of Aβ through autophagy. RBFOX3-positive cells and Nissl-stained cells in the hippocampus were markedly decreased in Aβ-treated mice compared with the controls (A and B), whereas MSC administration markedly increased the survival of RBFOX3-positive and Nissl-stained cells in the brain (C). On stereological analysis, RBFOX3-positive cells in the hippocampus were significantly decreased in in Aβ-treated mice compared with the controls (A and B), whereas MSC administration markedly increased the survival of hippocampal neurons in the AD animal brain (D). After MSC administration in AD animals significantly increased the ratio of LC3-II/LC3-I (E) and RAB7 expression (F). Additionally, MSC administration reduced Aβ immunoreactivity (G to I) and the level of Aβ level (J) compared with only-Aβ-treated mice. Ultrastructural analyses showed that autophagosomes were rarely noted in normal mice (K to M), whereas autophagic vesicles that were not fused with lysosomes were abnormally accumulated in Aβ-treated mice (N to P). In contrast, MSC administration induced a large portion of autophagosomes that were fused with lysosomes (Q to S). Yellow arrows indicate autophagosomes, and red arrows indicate autolysosomes observed in each group. Values are expressed as mean ± SDs, n = 4. *P < 0.05 for Aβ-treated mice vs. control mice; ^# P < 0.05 for Aβ-treated mice vs. MSCs-administrated mice. Scale bar: 50 µm (A–C); 10000 nm (H, K, and N); 5000 nm (I, L, and O); 2000 nm (J, M, and P).

To determine whether MSCs activated autophagosome formation in the Aβ-treated animals, the ratio of LC3-II/ LC3-I expression in the hippocampus was determined by western blotting. The LC3-II/LC3-I ratio, as well as the RAB7 expression, was significantly increased in MSCs-administrated mice compared with Aβ-treated mice (Fig. 7E and F). MSC administration significantly reduced Aβ immunoreactivity (Fig. 7G–I) and the level of Aβ (Fig. 7J) in the hippocampus, which was elevated in Aβ-treated mice, suggesting that MSCs could modulate Aβ levels in the hippocampus through autophagy induction. Additionally, we evaluated the effect of MSCs on autophagy modulation using transgenic AD mice. MSC treatment in AD mice significantly increased the expression of LC3-II and RAB7 and decreased the level of Aβ in the hippocampus compared with those in nontreated mice (Fig. S5). In ultrastructural analyses using EM, autophagosomes were rarely noted in normal mice (Fig. 7K–M), whereas AVs were abnormally accumulated in Aβ-treated mice (Fig. 7N–P). Most AVs observed in Aβ-treated mice, however, were not fused with lysosomes. In contrast, MSC administration in Aβ-treated mice led to a large portion of autophagosomes being conjoined with lysosomes, thus forming autolysosomes (Fig. 7Q–S).

MSCs regulate BECN1 expression to enhance autophagy in vitro and in vivo

BECN1 expression is a key determinant in the regulation of autophagic capacity, which promotes cell survival. ²⁹ To elucidate whether MSCs regulated autophagic activity through modulation of BECN1, we evaluated BECN1 expression following MSC application in both an in vitro and an in vivo model of AD. Aβ treatment led to a slight decrease in BECN1 expression in SH-SY5Y cells and animals compared with controls, although this difference was not statistically significant. However, coculture with MSCs in Aβ-treated SH-SY5Y cells significantly increased BECN1 expression at 6 and 12 h compared with Aβ-treated cells (Fig. 8A and B). Similarly, MSC administration in Aβ-treated mice led to a significant increase in BECN1 expression compared with that in Aβ-treated mice (Fig. 8C and D).

Figure 8. MSCs promote BECN1 expression in vitro and in vivo. Coculture with MSCs in Aβ-treated SH-SY5Y cells exhibited significantly increased BECN1 expression at 6 and 12 h compared with Aβ-treated cells (A and B). Additionally, MSC administration in Aβ-treated mice led to a significant increase in BECN1 expression compared with Aβ-treated mice (C and D). Values are expressed as mean ± SDs, n = 3. *P < 0.05 for Aβ-treated SH-SY5Y cells or mice vs. SH-SY5Y cells or control mice; ^# P < 0.05 for Aβ-treated SH-SY5Y cells or mice vs. MSCs-treated SH-SY5Y cells or mice.

Discussion

In the present study, we demonstrated for the first time that MSC treatment significantly enhances autolysosome formation and clearance of Aβ in Aβ-treated cellular and animal models, which might lead to increased neuronal survival against Aβ toxicity. Furthermore, upregulation of BECN1 appears to be one of the underlying contributors to cell survival associated with autolysosome formation by MSCs. Our data suggest that modulation of the autophagy pathway to repair the damaged AD brain using MSCs would have a significant impact on future strategies for AD treatment.

Several lines of evidence have suggested that Aβ is implicated in modulating autophagy. Cells treated with Aβ or overexpressing mutant APP increase autophagy induction with a large number of AVs, ^{30 - 32} where increased intracellular Aβ appears to regulate autophagy through an Akt-dependent downsignaling pathway or mitochondrial damage. In healthy neurons, AV clearance is efficiently mediated through fusion with lysosomes, and the process of autolysosome formation leads to increased neuronal viability by preventing neuronal accumulation of intracellular Aβ. ³³ However, excessive intracellular Aβ load impairs the process of autolysosome formation, leading to significant accumulation of autophagosomes. ³⁴ Consistent with these data, the present study showed that Aβ-treated cell cultures and animal models exhibited an increased induction of autophagosomes that contained Aβ with decreased cellular viability. Immunofluorescent and ultrastructural analyses revealed that most autophagosomes did not fuse to lysosomes in Aβ-treated cells and animals, suggesting that Aβ-induced neuronal death is associated with autophagy-dependent clearance of Aβ secondary to autolysosome formation failure.

Autophagy is essential for the clearance of detrimental Aβ aggregates and thus plays a critical role in maintaining Aβ homeostasis in the AD-related microenvironment. In AD transgenic mice, Aβ is colocalized with autophagosomes and thus could be a substrate for autophagy-mediated clearance. Additionally, a recent study demonstrated that PSEN1/presenilin 1, a subunit of the γ-secretase enzyme complex, is essential for lysosomal acidification, which is necessary to activate cathepsins and other hydroxylases that carry out digestion during autophagy. ³⁵ Moreover, several studies indicate that small molecular compounds that can activate autophagy or lysosomal proteolysis can also markedly decrease Aβ load in AD. ^{36 , 37} Therefore, a therapeutic strategy to enhance autolysosome induction may be an important pharmacological target in AD. Interestingly, immunofluorescent analysis in the present study showed that MSC treatment led to increased fusion of Aβ-containing autophagosomes (LC3-II) and lysosomes (LAMP2), as well as activity of lysosomal proteolytic enzymes. These data indicate that MSC treatment in Aβ-treated cells remarkably enhanced the process of autolysosome formation and autolysosomal catabolic function, which may be associated with increased neuronal survival against Aβ toxicity. Specifically, the neuroprotective effect of MSCs seems to be dependent on lysosomal activity mediated through autolysosome formation because a lysosomal inhibitor that prevents acidification counteracts the prosurvival effects of MSCs. Furthermore, MSC coculture in Aβ-treated cells showed that compared with only-Aβ-treated cells, Aβ immunoreactivity that was colocalized within lysosomes was markedly greater and intracellular Aβ levels were significantly decreased. These in vitro data suggest that MSCs enhance clearance of intracellular Aβ through the autophagy-lysosome pathway. In an Aβ-treated animal model of AD, autophagic regulation of MSCs was more evident, demonstrating that MSC administration significantly augmented autophagosome induction and final maturation of late AVs and fusion with lysosomes. As a consequence, MSC administration significantly reduced Aβ levels in the hippocampus, which were elevated in Aβ-treated mice, and concomitantly increased the survival of hippocampal neurons. Accordingly, the present data indicate that MSCs may play a critical role in neuronal homeostasis as an autophagy modulator that enhances clearance of toxic proteins in an AD-related microenvironment.

BECN1, a BCL2-homolog (BH)-3 domain-only protein, is localized primarily within cytoplasmic structures, including the endoplasmic reticulum, mitochondria, and the perinuclear membrane. ³⁸ BECN1 is a component of the class III PtdIns3K complex that participates in autophagosome formation and mediates the localization of other autophagy-related proteins to the phagophore membrane ³⁹ and thus has a central role in autophagy. Increasing evidence has suggested that several factors, including NFKB/ nuclear factor of kappa light polypeptide gene enhancer in B-cells, E2F transcription factors, and microRNAs, regulate BECN1 expression in the autophagic pathway. ^{40 , 41} Additionally, Wang et al. ⁴² demonstrated that the MAP2K1/2-MAPK1/3 signaling pathway participates in autophagy induction via modulation of BECN1 activity in several cell lines. In the present study, we demonstrated for the first time that MSCs can enhance autophagy through positive regulation of BECN1 expression in AD models, even though we did not formally prove a causal relationship between BECN1 expression and the autophagy signaling pathway. In accordance with our data, Sanchez et al. ⁴³ have shown that the stromal cell property of MSCs in breast cancers is associated with upregulation of autophagy. They report that cancer cells placed in serum-deprived conditions have significantly decreased cell viability, but in vitro coculture assays demonstrate that MSCs helped to maintain cancer cell survival under serum-deprived conditions by increasing expression of autophagy key regulators including BECN1. Accordingly, upregulation of BECN1 appears to be one of the underlying contributors to cell survival associated with autolysosome formation by MSCs.

Several studies demonstrate that MSC-derived soluble factors can modulate neuroprotective properties in AD models. Kim et al. ⁴⁴ report that human umbilical cord blood-derived MSCs have a neuroprotective effect in vitro against Aβ toxicity via gatectin-3 secretion. Additionally, they have found that transplantation of human umbilical cord blood-derived MSCs in AD transgenic mice induces the expression of MME/neprilysin in microglia and enhanced Aβ clearance, which is mediated though secretion of soluble ICAM1/intercellular adhesion molecule-1 from umbilical cord blood-derived MSCs. Recently, Lee et al. ⁴⁵ have described cellular and molecular mechanisms underlying neuroprotective effects of bone marrow-derived MSCs in AD models. They show that CCL5, a chemoattractive factor secreted from blood-derived MSCs, may play a critical role in recruitment of alternative microglia into the AD brain, and MME and IL4/interleukin 4 derived from the alternative microglia may lead to a reduction in Aβ deposition and memory impairment in AD mice. Along with Aβ clearance through modulation of neuroinflammation and growth factors, our study indicates that MSCs can regulate Aβ clearance through modulation of the autophagy pathway. Future studies focused on identifying small molecules that are relevant to MSC properties for autophagy modulation are needed to exploit therapeutic targets.

Regarding the functional impact of autophagy, several lines of evidence suggest that autophagy may have a protective role against the development of various neurodegenerative diseases. ^{46 , 47} In particular, pharmacological stimulation of autophagy can increase life span in yeast, ⁴⁸ C. elegans, ⁴⁹ and mice, ⁵⁰ and is beneficial for Aβ-induced toxicity in animal models of AD. ^{51 , 52} Therefore, a therapeutic perspective suggests that in addition to advantages in clinical applications, the use of MSCs as pharmacological boosters of autophagy to enhance the degradation of Aβ could represent an effective future therapeutic approach for AD.

In summary, the present study demonstrated that MSCs exert neuroprotective effects through enhancement of autolysosome formation and autophagy pathway-dependent Aβ clearance in Aβ-treated AD models. Additionally, MSCs regulate neuronal survival against Aβ toxicity by regulating BECN1 expression. Future studies to uncover soluble factors that are responsible for autophagy regulation are warranted, as these factors may be applicable to clinical strategies for future ad treatments.

Materials and Methods

Aβ preparation

Aβ was prepared as described previously. ⁵³ In details, Aβ peptide (Millipore, AG968) was suspended in 1,1,1,3,3,3 hexafluoro-2-propanol (HFIP; Sigma, 105228) to 1 mM. Aβ peptides were stored desiccated at −20 °C until further processing. To form Aβ oligomers, the peptide was resuspended to 200 µM in DMSO, bath-sonicated for 10 min and vortexed for 30 s. Aggregation was allowed to proceed for 72 h at 37 °C and incubated for 2 weeks at 4 °C to facilitate higher-order aggregation. The Aβ peptide stock was stored at −70 °C until further processing. Before use, the Aβ was diluted in Dulbecco’s Modified Eagle Medium (DMEM; HyClone) or phosphate buffered saline (PBS; HyClone, SH30243, 30021) to the indicated concentration. Most of the Aβ forms are oligomers but some monomers exist in the mixture. ⁵⁴

Cell culture

Frozen vials of characterized human MSCs at passages two to four were obtained from the Severance Hospital Cell Therapy Center. The human neuroblastoma cell line, SH-SY5Y, and CHO cells stably expressing human APP 695, were used in these experiments. SH-SY5Y cells, CHO cells, and MSCs were maintained in DMEM supplemented with 10% fetal bovine serum (FBS; HyClone) and an antibiotic mixture of penicillin (100 U/mL) and streptomycin (100 μg/mL). When these cells reached 70% to 80% confluence, the cells were trypsinized and subcultured. To test the effects of cocultures of Aβ-treated SH-SY5Y or CHO cells and MSCs without cell contact, SH-SY5Y and CHO cells were plated on transwell bottoms at a density of 2.0 × 10⁴/cm² (Corning, 3413, 3412) and MSCs were plated in transwell inserts at a density of 1.0 × 10⁴/cm². To identify a concentration and incubation time for cytotoxicity, SH-SY5Y cells were treated with varying concentrations of Aβ and 20 µM of Aβ_42–1. Additionally, SH-SY5Y cells were treated with Aβ or cotreated Aβ with 50 nM of Baf (Sigma, B1793). The effects of MSCs on autophagy modulation were tested in SH-SY5Y cells treated with 20 μM Aβ for 24 h. After 24 h, Aβ-containing medium was changed and then Aβ-treated SH-SY5Ycells were cocultured with MSCs or SH-SY5Y cells in a humidified incubator at 37 °C and 5% CO₂ for 3, 6, and 12 h. Additionally, we incubated Aβ-treated SH-SY5Y cells with Baf or CQ (10 μM, Sigma, C6628). The SH-SY5Y cells were then collected for assay. All experiments were replicated 3 times.

Cell viability analysis

The viability of cocultured SH-SY5Y cells was assessed by trypan blue dye exclusion. The proportion of viable cells was determined as the ratio between dead (stained) and viable (unstained). Cells were counted using a hemocytometer.

Aβ and MSC administration in animals

All animal procedures were performed in accordance with the Laboratory Animals Welfare Act, the Guide for the Care and Use of Laboratory Animals and the Guidelines and Policies for Rodent Experiment provided by the Institutional Animal Care and Use Committee (IACUC) at the Yonsei University Health System. In this study, male ICR mice (5-wk-old) and APP/PSEN1 double transgenic mice (6- to 7 mo old) generated by mating single transgenic mice expressing human mutant APP and mutant PSEN1 were used. Animals were acclimated to a climate-controlled room with a constant 12 h light/dark cycle (12 h on, 12 h off) for a week prior to the initiation of drug administration. At 6 weeks of age, the ICR mice received an intracerebroventricular injection of Aβ (400 ρmol/mouse) or control solution (PBS) as described previously, ^{55 , 56} with minor modifications. Briefly, mice were anesthetized with isoflurane gas, and Aβ (400 ρmol/2 uL) or control solution (PBS) was slowly injected (0.2 uL/min) bilaterally into the lateral ventricles (0.2 mm posterior to bregma, 1.0 mm lateral to midline, and 3.1 mm ventral to the brain surface) using a stainless-steel injection needle (26 gauge) connected to a 10 μL Hamilton micro-syringe (Hamilton). The needle was left in place for 10 min before it was slowly withdrawn. The ICR mice were randomly divided into 3 groups (n = 4 to 5 per group): (1) Sham; (2) PBS; (3) MSC. Mice in the MSC group were subjected to MSC (1.0 × 10⁶) delivery on post-operative day1. MSCs were suspended in 100 μL PBS and carefully injected in the tail vein using a 26-gauge syringe. All mice were sacrificed on post-operative day 4.

Brain sample preparation

For western blotting and enzyme-linked immunosorbent assay (ELISA), the isolated mouse hippocampus was rapidly frozen and stored at −70 °C until analyzed. For immunochemical analysis, all mice were deeply anesthetized with chloral hydrate (0.4 g/kg, i.p.) and then perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (PB; pH 7.4). The brains were then sectioned on a sliding microtome to obtain 30-µm thick coronal sections. All sections were stored in tissue stock solution (30% glycerol, 30% ethylene glycol, 30% 3rd D.W., 10% 0.2 M PB; pH 7.2; Sigma) at 4 °C until use.

Immunocytochemistry and immunohistochemistry

Brain sections and cocultured SH-SY5Y cells were rinsed twice in PBS and stained using specific antibodies. The primary antibodies used were as follows: rabbit anti-LC3 (1: 500 for immunocytochemistry; Sigma, L7543), mouse anti-amyloid β (6E10; 1: 200 for immunocyto- and immunohistochemistry; Covance, SIG39320), rabbit and mouse anti-LAMP2 (1: 200 for immunocytochemistry; Abcam, ab25631, ab37024), rabbit anti-NES (1:200 for immunohistochemistry, Millipore, ABD69), mouse anti-NUMA1 (1:200, for immunohistochemistry, Calbiochem, NA09L), rabbit anti-ELAVL4 (1:200 for immunohistochemistry, Thermo, PA5-26478), and mouse anti-RBFOX3/NeuN (1: 200 for immunohistochemistry, Chemicon, MAB377). Immunofluorescence labeling was performed by incubating the cells in goat anti-mouse Ig G and goat anti-rabbit Ig G (both FITC, AP181F, AP307F, green and Cy-3, red, AP124C, AP132C, Millipore) secondary antibodies. The cell nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; 1: 2,000, Invitrogen, D1306). The RBFOX3 antibodies were detected with 0.05% diaminobenzidine (DAB, Vector Laboratories, SK4100) and 0.03% H₂O₂ and the sections were treated with 1% cresyl violet (Sigma, C5042) solution for 3 min followed by differentiation in acetic acid in 100% ethanol (at 1:50,000 dilution) for 5 s. The immunostained cells were analyzed using bright-field microscopy and viewed under a confocal laser-scanning microscope (ZEISS, LSM700). Sections used for cell counting included the hippocampus. This generally yielded 8 to 12 sections in a series. Sampling was performed with the Olympus CAST-Grid system (Olympus Denmark A/S), using an Olympus BX51 microscope (Olympus), connected to the stage and feeding the computer with the distance information in the z-axis. A counting frame (55%, 53, 1281 µm²) was randomly placed on the first counting area and systematically moved through all counting areas until the entire delineated area were sampled. Actual counting was performed using a 100 × oil objective. Guard volumes (4 µm from the top and 46 µm from the bottom of the section) were excluded from both surfaces to avoid the problem of lost cap, and only the profiles that came into focus within the counting volume (with a depth of 10 µm) were counted. The total number of stained cells was calculated according to the optical fractionator formula. ⁵⁷

Total RNA extraction and reverse transcription polymerase chain reaction (RT-PCR)

Total RNA was extracted from the SH-SY5Y cells using TRIzol^® reagent (Invitrogen, 15596018) in accordance with the manufacturer's protocol. An equal amount of RNA (approximately 1 µg) in each experiment was reverse transcribed using an amfiRivert cDNA Synthesis Premix (GenDEPOT, R5101). Subsequently, 2 µL of cDNA was used as a template for RT- PCR analysis in an amfiRivert 1-Step RT-PCR Kit (GenDEPOT, P0316). The PCR reaction was performed using 10 ρmol each of the primers for CTSB (NM 001908; forward 5′-TGTGGTGGTC CTTGATCCTT-3′, reverse 5′-CCAGCCTGCC ATGTTGTATT-3′). After an initial denaturation at 94 °C for 2 min, 25 cycles of PCR were performed, consisting of denaturation (2 min, 94 °C), annealing (1 min, 55 °C), and extension (1 min, 72 °C), followed by a final extension (5 min, 72 °C). The PCR products were separated by electrophoresis on a 2% agarose gel and stained with ethidium bromide. Gels were examined under UV illumination.

Western blot analysis

For western blotting, cocultured SH-SY5Y cells or CHO cells were harvested and extracted using lysis buffer containing protease inhibitor (iNtRON Biotechnology, 17081). Briefly, 50 µg of protein was separated using sodium dodecyl sulfate-gel electrophoresis and transferred onto a nitrocellulose membrane (Amersham, RPN2032D). The membranes were blocked in nonfat milk (BD). Membranes were probed with 1:1,000 dilutions of the following primary antibodies: rabbit anti-LC3 (Sigma), rabbit anti-BECN1 (Santa Cruz Biotechnology, sc11427), and rabbit anti-RAB7 (Cell Signaling, sc11427). As a secondary antibody, a 1:2,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit antibody (Zymed Laboratories, sc11427) was used. Antigen-antibody complexes were visualized with a chemiluminescence system (Amersham, sc11427), followed by exposure to X-ray film (Kodak X-OMAT, sc11427). For semiquantitative analysis, immunoblot band densities were measured by computer imaging (Image J; NIH). All proteins were normalized against with ACTB expression (Santa Cruz Biotechnology, sc-8432, data not shown).

Aβ ELISA

Aβ was measured using a High Sensitivity Human Amyloid β ELISA kit (Millipore, EZHS42). The Aβ concentration in samples was measured as follows. Cocultured SH-SY5Y cells, CHO cells, and animal brains, 50 μL of each diluted samples, and standards included in the kit were applied to microtiter plates precoated with antibody that specifically recognized the N-terminal of Aβ peptides because its epitope is localized at amino acid positions 4 to 10. Therefore, the antibody reacts with both Aβ₄₀ and Aβ₄₂ peptides as well as other Aβ species. Following an overnight incubation at 4 °C and washing steps, a detection antibody indirectly linked to an enzyme was applied. After incubation and washing, substrate was added and incubated for 15 to 20 min, and the reaction was stopped with sulfuric acid. The color reaction was measured with an automatic ELISA reader (BIOTECH, ELX 800) with the wavelength set at 450 nm and a reference wavelength of 590 nm. The software (Microplate Manager, Bio Rad) was used to create standard curves and to calculate the concentration of the samples.

EM analysis

Cocultured SH-SY5Y cells and brain tissues were fixed with 2% glutaraldehyde - paraformaldehyde in 0.1 M PB (pH 7.4) for 2 h, and then washed 3 times for 30 min in 0.1 M PB. They were post-fixed with 1% OsO₄ dissolved in 0.1 M PB for 2 h and dehydrated in an ascending gradual series (50 to 100%) of ethanol and infiltrated with propylene oxide. Specimens were embedded using a Poly/Bed 812 kit (Polysciences). After pure fresh resin embedding and polymerization at 60 °C in an electron microscope oven (TD-700, DOSAKA) for 24 h 350-nm- thick sections were initially cut and stained with toluidine blue for light microscopy. Thin sections (70 nm) were double stained with 7% uranyl acetate and lead citrate for contrast staining. The sections were cut using a Leica Ultracut UCT Ultra-microtome (Leica Microsystems). All of the thin sections were observed by transmission electron microscopy (JEM-1011, JEOL) at the acceleration voltage of 80 kV.

Statistical analysis

the group means were compared using the Mann-Whitney u-test for pairs and kruskal-wallis analysis for multiple groups. P-values less than 0.05 were considered statistically significant. statistical analyses were performed using commercially available software (version 12.0; SPSS Inc.).

Glossary

Abbreviations:

AD

Alzheimer disease

Aβ

amyloid β

APP

amyloid precursor protein

AVs

autophagic vacuoles

LC3

microtubule-associated protein 1 light chain 3

MSCs

mesenchymal stem cells

CTSB

cathepsin B

CQ

chloroquine

CHO

Chinese hamster ovary cells

EM

electron microscopy

Baf

bafilomycin A₁

NUMA1

nuclear mitotic apparatus protein 1

ELAVL4

ELAV (embryonic lethal abnormal vision, Drosophila)-like 4

PSEN1

presenilin 1

10.4161/auto.26508

Disclosure of Potential Conflicts of Interest

All authors have no conflict of interest to declare.

Supplemental Materials

Supplemental materials may be found here: www.landesbioscience.com/journals/autophagy/article/26508