The Role of Chaperone-Mediated Autophagy in Huntingtin Degradation

Huntington Disease (HD) is caused by an abnormal expansion of polyQ tract in the protein named huntingtin (Htt). HD pathology is featured by accumulation and aggregation of mutant Htt in striatal and cortical neurons. Aberrant Htt degradation is implicated in HD pathogenesis. The aim of this study was to investigate the regulatory role of chaperone-mediated autophagy (CMA) components, heat shock protein cognate 70 (Hsc70) and lysosome-associated protein 2A (LAMP-2A) in degradation of Htt fragment 1-552aa (Htt-552). A cell model of HD was produced by overexpression of Htt-552 with adenovirus. The involvement of CMA components in degradation of Htt-552 was determined with over-expression or silencing of Hsc70 and LAMP-2A. The results confirmed previous reports that both macroautophagy and CMA were involved in degradation of Htt-552. Changing the levels of CMA-related proteins affected the accumulation of Htt-552. The lysosomal binding and luminal transport of Htt-552 was demonstrated by incubation of Htt-552 with isolated lysosomes. Expansion of the polyQ tract in Htt-552 impaired its uptake and degradation by lysosomes. Mutation of putative KFERQ motif in wild-type Htt-552 interfered with interactions between Htt-552 and Hsc70. Endogenous Hsc70 and LAMP-2A interacted with exogenously expressed Htt-552. Modulating the levels of CMA related proteins degraded endogenous full-length Htt. These studies suggest that Hsc70 and LAMP-2A through CMA play a role in the clearance of Htt and suggest a novel strategy to target the degradation of mutant Htt.


Introduction
Abnormal accumulation of misfolded and aggregated proteins in neurons is a hallmark of several neurodegenerative diseases. The mutant proteins of neurodegenerative diseases can cause dysfunction and death of neurons. Huntington's disease (HD) is an autosomal dominant neurodegenerative disease caused by an abnormal expansion of polyQ tract in the N-terminal huntingtin (Htt). HD is characterized by the accumulation and aggregation of mutant N-terminal Htt proteins in diseased neurons [1]. When the polyQ repeat expands above 35, disease will manifest, typically striking in the late 40 s [2]. The N-terminal fragment of Htt containing 552 amino acids (Htt-552) is a caspase 2/3 cleavage product and can be found in normal and HD patient brains. Htt-552 with an expanded polyQ repeats causes an aggressive HD-like disease in animal and cell models [3]. N-terminal Htt was reported to be degraded by macroautophagy and ubiquitin-proteasome pathways [4,5]. Although the chaperone-mediated autophagy (CMA) has been proposed to be involved in degradation of Htt, the molecular process and regulatory mechanisms have not yet been fully characterized.
The major pathways of mutant protein clearance are the proteasomal and the lysosomal systems in cells. These protein degradation pathways are compromised with aging [6][7][8]. In several types of neurodegenerative diseases, aggregated mutant proteins impair function of the proteasome, leading to accumulation of these diseases causing proteins in affected neurons. In this circumstance, the inducible autophagic pathway is likely to be the primary mechanism involved in the degradation of misfolded or aggregated proteins [9][10][11]. Autophagy is a degradation pathway for long-lived cytoplasmic components using lysosomes [12][13][14][15][16]. Based on the way substrates are transported into lysosomes, autophagy is classified into three types: macroautophagy, microautophagy and CMA. Among them, CMA has received particular attention because of its selectivity in degrading substrates compared to other forms of autophagy. In contrast to macroautophagy, a highly specific subset of cytosolic proteins with a motif recognized by the heat-shock cognate protein of 70 kDa chaperone (Hsc70) is selectively degraded in lysosomes via CMA [17,18]. Following binding of the chaperone-substrate complex to a lysosomal membrane receptor, lysosome-associated membrane protein 2A (LAMP-2A) [19], CMA substrate proteins are translocated into the lysosomal lumen for degradation by hydrolases [20].
The involvement of CMA in neurodegeneration has been illustrated in Parkinson's disease. A mutation of Parkin-1 affected degradation of a-synuclein through CMA, leading to accumulation and aggregation of a-synuclein and degeneration of nigral dopominergic neurons [21]. Furthermore, mutant Tau protein involved in Alzheimer's disease (AD) is also targeted to lysosomes via CMA [22]. The advantage of CMA for protein degradation is its relative substrate selectivity compared to macroautophagy.
Up to now, the degradation of Htt by macroautophagy has been investigated by several investigators [4,17,23], whereas the involvement of the chaperone-mediated autophagy has received less attention. The vacuole targeting (Cvt) pathway in yeast which is similar to the selective autophagy pathway may be involved in degradation of polyQ repeat Htt [24]. Some groups have reported the association of CMA and HD [25,26]. It has been proposed that N-terminal Htt fragment may contain KFERQ-like motifs following phosphorylation and be degraded via CMA. However, the supporting evidence on interactions of Htt with CMA related proteins and uptake and degradation of Htt by lysosomes is still missing. In this paper, we used rat neuronal like cells and human cells that were infected with adenovirus containing human Htt-552 with 18 or 100 glutamines (Htt-552-18Q or Htt-552-100Q) and evaluated whether Htt-552 interacted with LAMP-2A and Hsc70 and was transported into lysosomes for degradation. The results suggest that Htt-552 can be recognized by Hsc70 through KFERQ-like pentapeptide motif and is transported into lysosomes via LAMP-2A.

Materials and Methods
Cell culture and transfection by plasmids PC12 and Hela cell lines were purchased from Shanghai Institute of Biochemistry and Cell Biology (Shanghai, China), and were grown at 37uC in 5% CO 2 in DMEM medium supplemented with 2 mmol/L L-glutamine and 10% fetal bovine serum (FBS). To remove the serum, cells were washed twice with phosphate buffered saline (PBS) before replacing the complete medium with serum-free medium. We observed no changes in cell viability up to 72 h after serum removal. To activate CMA, cells were transfected with pcDNA4-LAMP-2A and pcDNA4-Hsp70 (human, kindly provided by Dr. GH. Wang, Soochow University School of Pharmaceutical Science, China). pcDNA4-Hsc70-His was generated by PCR from the I.M.A.G.E. clone for Hsc70 (Clone ID: NM_153201) with: ATTGAATTCATGTCCAAGG-GACCTGCAG (forward) and GCGCTCGAGTCAACCTCTT-CAATGGTGG (reverse), and subsequently digested with BamH I and Xho II and cloned in to the BamH I and Xho I sites of pcDNA4/HisA (V862-20, Invitrogen Carlsbad, CA, USA). Transfection was performed with Lipofectamine 2000 according to the manufacturer's instructions (11668-019, Invitrogen, Carlsbad, CA, USA).
Ad-Htt-18Q-552 stop and Ad-Htt-100Q-552 stop, were obtained through co-transfecting T293A cells with the backbone plasmid pBHG10 and the shuttle plasmids: pDC316-Htt-552-18Q stop and pDC316-Htt-552-100Q stop. Cytopathic effects happened at the 7th day. The cells were collected at the 10th day to obtain the first generation adenovirus. Then the first generation of adenoviruses was proliferated in 293T cells. The viruses are named Ad-Htt-552-18Qaa and Ad-Htt-552-100Qaa, respectively. The fourth generation adenovirus was used in later experiments. The titers were 5610 9 and 7610 9 , respectively. In addition, Adnull-GFP adenoviral vector was generated and proliferated following the same protocol. PC12 and Hela cells were incubated in 1640 medium with 5% FBS containing adenoviral vectors. The virus-containing medium was removed 5 h later and replaced with fresh medium with 10% FBS.

Immunofluorescence
For immunofluorescence microscopic examination, cells were plated onto 12-mm polylysine-coated coverslips and cultured for 24 h, cells were treated with siRNA or drugs. Cells were washed in PBS, fixed with 4% paraformaldehyde in PBS at 4uC for 10 min, and then washed again with PBS. The cells were permeabilized with 0.25% Triton X-100, and were then blocked with 10% normal goat serum (NGS) for 15 min. Primary antibodies, diluted in PBS with 1% bovine serum albumin were added to the cells and incubated for overnight at 4uC. The coverslips were washed three times before incubation with fluorescence-conjugated secondary antibodies using the same procedure as for the primary antibodies. The coverslips were mounted on slides with mounting medium (F4680, Sigma-Aldrich, Saint Louis, MO, USA) and were examined with a laser scanning confocal microscope. (C1S1, Nikon, Tokyo, Japan). The following primary antibodies were used: rabbit polyclonal anti-LAMP-2A antibody (ab18528, Abcam, Cambridge, MA, USA); goat polyclonal antibody against cathepsin D (sc-6488, Santa Cruz, Santa Cruz, CA, USA), rabbit polyclonal anti-Hsc70 antibody (ab51052, Abcam, Cambridge, MA, USA) and mouse monoclonal antibody against Htt (MAB2166, Chemicon (Millipore), Billerica, MA, USA).

Immunoprecipitation
For immunoprecipitation studies, cells in dishes were transfected using adenovirus Htt-552 and Lipofectamine 2000 with plasmids of LAMP-2A and Hsc70 for 6 h and then maintained in complete media for 48 h and 72 h. Lysates were centriffuged at 16,000 g at 4uC for 15 min, and the supernatant was collected in immunoprecipitation buffer (50 mM Tris, 274 mM NaCl, 5 mM KCl, 5 mM EDTA, 1% Triton X-100, 1 mM PMSF and a protease and phosphatase inhibitor cocktail), Supernatants were precleared with protein G and subsequently incubated with anti-Htt antibody, anti-LAMP-2A antibody or anti-Hsc70 antibody overnight, and then 20 mL protein G were added and incubated for 2 h at 4uC. Protein G beads were washed three times in immunoprecipitation buffer, resuspended in sample buffer with 5% beta-mercaptoethanol and heated at 95uC for 5 min. Then the samples were run on SDS-PAGE and blotted to nitrocellulose for standard Western blot analysis.

Western blot analysis
Western blot analysis was performed as described previously [32]. Cells were harvested and rinsed twice with ice-cooled PBS and homogenized in buffer containing 10 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 5 mol/L edetic acid, 1 mmol/L PMSF, 0.28 U/L aprotinin, 50 mg/L leupeptin, 1 mmol/L benzamidine, 7 mg/L pepstain A. Protein concentrations were determined using the BCA kit (Pierce, USA). Thirty micrograms of proteins from each sample were subjected to electrophoresis on 10-12% SDS-PAGE gel using a constant current. Proteins were transferred to nitrocellulose membranes and incubated with the Tris-buffered saline containing 0.2% Tween-20 (TBST) and 3% non-fat dry milk for 3 h in the presence of one of the following antibodies: mouse monoclonal antibody against Htt (MAB2166, Chemicon

Isolation of lysosomes
Subcellular fractionation was performed with a self-forming Percoll gradient. All steps were done at 4uC. Mouse liver was minced and washed in PBS, liver tissue was suspended in homogenization buffer (0.25 M sucrose, 2 mM EDTA, and 10 mM Hepes [pH 7.4]) and homogenized with a tight-fitting glass homogenizer. The homogenate was centrifuged (800 g, 10 min) to remove unbroken cells and the nuclei. The supernatant was centrifuged (6800 g, 10 min) to remove the large organelles including heavy mitochondria. The supernatant was centrifuged (25000 g, 10 min) to obtain the light organelles including lysosomes. The precipitation layered over 10 ml of a 27% Percoll (17-0891-01, Pharmacia Inc, Piscataway, NJ, USA) solution in homogenization buffer, under layered with 0.5 ml of a 2.5 M sucrose solution. Centrifugation was done in a rotor (Optima L90K, Beckman, Brea, CA, USA) for 1.5 h at 35000 g. The layer of crude lysosomes of about 1.5 ml was collected at the bottom and then was centrifuged (100000 g, 60 min) to remove the other light organelles including light mitochondria at the bottom of the tube. The lysosomal solution was washed with PBS (18000 g, 25 min) to remove the sucrose at the bottom of the tube. Following isolation, the purity of lysosomes was detected by Hsp60, b-actin, LAMP-2A and Hsc70. Lysosomal integrity was verified by measuring the activity of ß-hexosaminidase, a lysosomal enzyme, in the incubation medium. Preparations with more than 10% broken lysosomes after isolation or more than 20% at the end of the incubation were discarded [33]. Lysosomal solution was divided equally to 4 portions. Each portion was incubated with the cell lysates containing none, null, Htt-552-18Q and Htt-552-100Q for 3 h.

Uptake and degradation of Htt-552 by lysosomes
Uptake of Htt-552 by isolated lysosomes was analyzed as described previously [20]. Briefly, freshly isolated lysosomes from mice liver were incubated with Htt-552 in MOPS buffer at 37uC for 5 and 10 min. Where indicated, lysosomes were pre-incubated with a cocktail of protease inhibitors for 10 min at 0uC. Lysosomes were collected by centrifugation, washed with PBS buffer, and subjected to SDS-PAGE and immunoblotting for Htt-552. Uptake was calculated from densitometric analysis by subtracting the amount of Htt-552 associated with lysosomes in the presence (protein bound to the lysosomal membranes and taken by lysosomes) and absence (protein bound to the lysosomal membranes) of protease inhibitors. To separate lysosomal membranes and matrix, lysosomes were disrupted by hypotonic shock where indicated. Briefly, collect the isolated lysosomes by centrifugation (25,000 g for 10 min), resuspend the pellet fraction in a hypotonic buffer (0.025 M sucrose), and after 30-min incubation on ice, spin the samples at 150,000 g for 30 min to recover the membrane fraction in the pellet and the lysosomal matrix in the supernatant fraction [34].

FACScan Flow Cytometric Analysis
Cell viablility analysis was performed as described previously [35,36]. For flow cytometric analysis, Hela cells transfected with empty vector, Ad-Htt-552-18Qaa and Ad-Htt-552-100Qaa for 120 h were trypsinized, washed in PBS, and resuspended in icecold 80% ethanol. Briefly, 2.5610 5 fixed cells were incubated in 250 mL propidium iodide solution (500 mg/mL propidium iodide in 3.8 mol/L sodium citrate at pH 7.0) and 250 mL RNase A (10 mg/mL prepared in 10 mmol/L Tris-HCl at pH 7.5) for 30 min at 37uC in the dark. The stained cells were filtered through the cell strainer caps of Falcon polystyrene round-bottomed tubes. DNA content was analyzed on a Becton Dickinson FACScan (Becton Dickinson, San Jose, CA, USA). The population of sub-G1 was determined using Cell Fit software (Becton Dickinson, San Jose, CA, USA). Data were collected from at least 20,000 cells.

Statistical analysis
Statistical analysis was carried out by one-way analysis of variance (ANOVA) followed by Dunnett t-test. Differences were considered significant when p,0.05.

CMA is involved in degradation of Htt-552
The estimated rate of infection of adenovirus Htt-552 in PC12 or HeLa cells was over 95% (Figure S1 A). The expression of Htt-552-18Q and Htt-552-100Q in PC12 and HeLa cells was verified with Western blot analysis with anti-Htt antibodies 2166 and Ab1. Similar levels of wt and mutant Htt-552 were expressed in cells and the expression did not cause apparent cellular toxicity 48 h after infection ( Figure S2 A and B). However, cell toxicity and apoptosis by mutant Htt was seen 120 h after infection and was exacerbated when CMA was compromised ( Figure S2 C and D).
To determine the involvement of lysosomes in degradation of Htt-552, PC12 cells were treated with ammonium chloride (NH 4 Cl, 15 mM), which inhibits lysosomal proteolysis. The results showed that ammonium chloride significantly inhibited the degradation of Htt-552. It was noted that the increase in accumulation of Htt-552-18Q was more robust than that of Htt-552-100Q ( Figure 1A) in the presence of NH 4 Cl. To determine the contribution of macroautophagy to lysosomal degradation of Htt-552, PC12 cells were treated with 3-methyladenine (3-MA, 10 mM), an inhibitor of macroautophagy. The results demonstrated that 3-MA also reduced the degradation of Htt-552, but the effect was not as effective as ammonium chloride. It thus appeared that lysosomal degradation of exogenous Htt-552 could involve other autophagic mechanisms in addition to macroautophagy.
Starvation is a classical method to activate autophagy, including macroautophagy and chaperone-mediated autophagy (CMA). Serum removal markedly enhanced levels of Beclin 1, which is a macroautophagy regulatory protein, and LAMP-2A, which is a receptor protein of CMA ( Figure 1B and 1C). The time-course of elevation of Beclin 1 was faster than that of LAMP-2A. When the levels of Beclin 1 returned to about basal levels, LAMP-2A reached its peak induction. The distance of the lysosomal vesicles to the nucleus is a useful indicator of activation of CMA [37]. An average of 4 different visual fields per group and 15 cells per visual field were evaluated. After starvation for 12 h, the distance of CMA-active LAMP-2A labeled lysosomes to the nucleus was reduced, indicating increased activity of CMA ( Figure 1D).
Next, we examined the degradation of Htt-552 expressed in PC12 cells, at 6 h and 12 h after serum removal, the times at which there is maximal activation of macroautophagy and CMA, respectively. Removing serum for either 6 h or 12 h markedly reduced accumulation of Htt-552. However, Htt-552 levels appeared higher at 12 h than at 6 h ( Figure 1E). These results suggested that wt and mutant Htt-552 could be degraded by both macroautophagy and CMA, but macroautophagy may contribute more.

Association of Htt-552 and exogenously and endogenously expressed LAMP-2A and Hsc70
To characterize the association of exogenous Htt-552 clearance and CMA, the present study analyzed the co-localization of exogenously and endogenously expressed CMA component

Uptake of Htt-552 by lysosomes
The most direct evidence for a protein as a CMA substrate probably is to determine its binding, uptake, and degradation in isolated intact lysosomes [17]. The purity of the lysosome preparation was detected by the enrichment of the lysosomal markers LAMP-2A and Hsc70 and a marked decrease in levels of Hsp60, which is a marker of mitochondria, and in b-actin, which is a marker of cytoplasm. Based on these assessments, the present study had achieved isolation of lysosomes from mouse liver with a relatively high purity ( Figure 5A). b-hexosaminidase latency was measured as an index of the integrity of lysosomal membranes after isolation [38]. Results indicated that the percent of broken lysosome content was 6.1+0.3% of total at 0 min, 7.8+0.4% of total at 10 min, 9.2+0.3% of total at 20 min and 9.4+0.4% of total at 30 min ( Figure 5B), indicating that the lysosomes in the preparation remained mostly intact.
Disease-causing mutant proteins had been suggested to destabilize the membranes of lysosomes [39]. Therefore we first confirmed that isolated lysosomes were not disrupted by either wt or mutant Htt-552 proteins by detecting the leakage of bhexosaminidase from lysosomes after incubation with cell lysates containing exogenously expressed Htt-552-18Q or Htt-552-100Q ( Figure 5C). Incubation of cell lysates with lysosomal preparations significantly reduced the protein levels of Htt-552, reflecting degradation of Htt-552 by lysosomes ( Figure 5D). Next, we studied the association of Htt-552 to lysosomal membranes and uptake of Htt-552 into lysosomal lumen. Htt-552 was recovered in the absence of protease inhibitors, indicating its presence in the lysosomal membranes; Htt-552 was also recovered in the presence of protease inhibitors, indicating its localization to the lysosomal lumen and to membranes ( Figure 5E). Detection of Htt-552 binding to lysosomal membranes (no protease inhibitor) increased with incubation time. The level of lysosomal membrane bound Htt-552 was higher in the presence of the protease inhibitor cocktail. Moreover, the magnitude of increases in Htt-552-18Q The Role of CMA in Degradation of Htt-552 PLOS ONE | www.plosone.org was more robust than that of Htt-552-100Q ( Figure 5F). The difference of lysosomal Htt-552 in the presence and absence of protease inhibitors represents the amount of Htt-552 taken up by lysosomes.
After disrupting the lysosomes and separating lysosomal membranes and matrix fractions [34], Htt-552 was detected in both membrane and matrix fractions, suggesting that Htt-552   [40]. Inhibition of lysosomal proteases increased the accumulation of Htt-552-100Q in lysosomal membranes, however the membrane Htt-552-18Q levels were not significantly changed ( Figure 5G). In contrast, the accumulation of Htt-552-18Q but not Htt-552-100Q in the lysosomal matrix was observed in the presence of protease inhibitors, indicating that Htt-552-18Q is more readily taken into lysosomes for degradation than is Htt-552-100Q ( Figure 5F).
As CMA substrates use LAMP-2A and Hsc70 for recognition and membrane transport, there is a competition between CMA substrates for these mediators. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is a known CMA substrate. This study evaluated the effects of Htt-552 degradation by GADPH. The results showed that pre-incubation of isolated lysosomes with GAPDH for 30 min reduced degradation of Htt-552 by lysosomes ( Figure 6A). Furthermore, pre-incubation of isolated lysosomes with neutralizing antibodies against LAMP-2A and Hsc70 also decreased the degradation of Htt-552 by lysosomes ( Figure 6B and 6C).

Disruption of interactions between Htt-552 and Hsc70 by mutation of CMA recognition motifs
We mutated the sequence 99 KDRVN 103 and 248 NEIKV 252 in a putative CMA recognition motif in Htt-552-18Q by replacing hydrophilic amino acid asparagine (N) with hydrophobic alanine (A), replacing acidic aspartate (D) with basic arginine (R) of 99 KDRVN 103 (Htt-552 of 18Qm) and replacing glutamine (E) with basic lysine (K) of 248 NEIKV 252 (Htt-552 of 18Q&) ( Figure 7A). The results showed that although the transfection efficiency of adenovirus Htt-552-18Q with the mutant recognition motif (Htt-552 of 18Qm) in PC12 cells was similar to wild-type Htt-552-18Q, the co-precipitation of mutant Htt-552-18Qm protein with Hsc70 was reduced compared with the wt Htt-552-18Q ( Figure 7B). However the co-precipitation of mutant Htt-552 -18Q& protein with Hsc70 was not dramatically changed compared with that of Htt-552-18Q. Overexpression of human Hsc70 markedly decreased accumulation of mutant Htt-552. However, the overexpression of Hsc70 was less effective in reducing Htt-552-18Qm in comparison with Htt-552, suggesting that the clearance of Htt-552-18Qm may be inhibited by the mutation of the CMA recognition motif ( Figure 7C). The data suggest that 99 KDRVN 103 of Htt-552 is a functional site for substrate recognition by CMA.

CMA affects degradation of endogenous Htt
To verify the role of CMA in degradation of endogenous Htt, the effects of reducing and increasing LAMP-2A on endogenous Htt were determined. Endogenous Htt was significantly reduced in PC12 cells that overexpressed LAMP-2A or Hsc70 for two days and was slightly increased (though not statistically significant) in PC12 cells treated with LAMP-2A or Hsc70 siRNA for two days ( Figure 8A). The reduction in endogenous Htt after expression of LAMP-2A or Hsc70 was not due to general depression of protein synthesis as no change was detected in the levels of b-actin and p62. The later is a substrate of macroautophagy but not CMA ( Figure 8A). The decrease in endogenous Htt was also not due to loss of cell viability as the level of MTT transformation was not different between control untreated cultures and those exposed to cDNAs encoding LAMP-2A or Hsc70 ( Figure 8B).

Discussion
HD is a family of polyQ repeat expansion diseases characterized by the accumulation of mutant Htt protein in diseased neurons [1]. The present study used the first 552 amino acids of human Htt with an expanded polyglutamine tract to simulate cellular pathologic conditions of the disease because a native Htt fragment of this size has been detected in vivo in human brain and in HD animal models [41,42]. Moreover this N-terminal region of Htt contains domains and modifications that regulate the targeting and turnover of the full-length protein [26]. Expressed Htt-552 was detected by a well-described anti-Htt antibody, MAB2166 which recognizes Htt443-457. There are reports that MAB2166 may not interact well with Htt species that have been modified by phosphorylation or acetylation which may be important for its degradation [26]. To assure the accuracy of Htt-552 detection, we also used an antibody recognizing Htt1-17 and obtained similar results related to CMA dependent clearance (Figure S2 A).
Autophagy is an important intra-cellular metabolic pathway that includes macroautophagy, microautophagy and CMA. All of the substrates of autophagy are transported to lysosomes where the substrates are degraded by hydrolytic enzymes. Blockade of lysosomes means blockade of all three types of autophagic degradation. The best characterized autophagy is macroautophagy which probably contributes to the highest percentage of lysosomal degradation in cells [43]. There are reports that the accumulation of Htt is decreased through the activation of lysosomes [4,7] whereas Htt accumulates with the inhibition of autophagy [5,25]. The results of previous studies indicate that Htt-552 is not only degraded by macroautophagy but also by CMA. Although involvement of CMA in Htt metabolism has been suggested by other investigators, the essential evidence of CMA metabolism of Htt, including interactions of Htt with LAMP-2A, Hsc70, and uptake and degradation of Htt by lysosomes remains to be revealed. Thus, this work presented direct evidence for huntingtin clearance by CMA.
It is known that after serum removal, activation of macroautophagy occurs before the activation of CMA, reaches maximal activity 4-6 h after starvation, and then gradually declines to basal levels. In most cells continuous starvation beyond 6 h, will increase CMA activity to maximal levels at about 12 h [44] and continue as long as starvation persists. The time-course of induction of Beclin 1 and LAMP-2A was consistent with the activation of macroautophagy and CMA with different time frames. The activation of CMA with serum removal was supported by the presence of increased localization of lysosomes to peri-nuclear regions [37]. We found decreased levels of exogenous Htt-552 at 6 h  **P,0.01 (compared with Htt-552-18Q without treatment); ##P,0.01 (compared with Htt-552-100Q without treatment). (G) Htt-552 in lysosomal membranes (MEMB) and matrices (MTX). Isolated lysosomes were incubated cell lysates containing expressed Htt-552 for 20 min, lysosomes were recovered by centrifugation. Lysosomal membranes and matrix were separated after hypotonic shock and centrifugation, and were processed for Western blot analysis. doi:10.1371/journal.pone.0046834.g005 Figure 6. Inhibition of Htt-552 degradation by blocking CMA pathway. (A) The effects of GAPDH on degradation of Htt-552. The isolated lysosomes were incubated with 10 mg GAPDH and Htt-552 for 30 min at 37uC. At the end of the incubation, lysosomes were collected by centrifugation and the levels Htt-552 were determined with immunoblotting. Values are the mean6SE of three experiments. **P,0.01, *P,0.05 (compared with Htt-552-18Q without lysosomes); ## P,0.01, # P,0.05 (compared with Htt-552-100Q without lysosomes); $$ P,0.01 (compared with Htt-552-18Q without GAPDH); && P,0.01 (compared with Htt-552-100Q without GAPDH). (B) The effects of LAMP-2A neutralizing antibody on degradation of Htt-552 by isolated lysosomes. The incubation of lysosomal preparation with Htt-552 was described as above. Values are the mean6SE of three independent experiments **P,0.01 (compared with Htt-552-18Q without lysosomes); # P,0.05, ## P,0.01 (compared with Htt-552-100Q without lysosomes); $$ P,0.01 (compared with Htt-552-18Q without LAMP-2A antibody); && P,0.01 (compared with Htt-552-100Q without LAMP-2A antibody). (C) The incubation of lysosomal preparation with Htt-552 was described as above. The incubation of lysosomal preparation with Htt-552 was described as above. The effects of Hsc70 neutralizing antibody on degradation of Htt-552 by isolated lysosomes. Values are the mean6SE of three independent experiments. **P,0.01 (compared with Htt-552-18Q without lysosomes); ## P,0.01, # P,0.05 (compared with Htt-552-100Q without lysosomes); $$ P,0.01 (compared with Htt-552-18Q without Hsc70 antibody); & P,0.05 (compared with Htt-552-100Q without Hsc70 antibody). doi:10.1371/journal.pone.0046834.g006 (macroautophagy activation reached peak level) and 12 h (macroautophagy returned to basal level and CMA activation reached peak level) after starvation. These results suggest that both macroautophagy and CMA are involved in degradation of Htt-552. We also observed more Htt-552 clearance by macroautophagy than by CMA. However, due to its greater selectivity for substrates CMA may be more important than macroautophagy in degrading Htt.  To further assess the association between Htt-552 and CMA, this study observed the interactions of exogenous expressed Htt-552 with the critical components of CMA, LAMP-2A and Hsc70. LAMP-2A is the integral membrane receptor protein that can directly import CMA substrate proteins across the lysosomal membranes. Hsc70 is a CMA chaperone which preferentially interacts with CMA substrate proteins and assists proteins moving into lysosomal lumens [19]. Immunofluorescence data indicated that Htt-552 co-localized with LAMP-2A and Hsc70. Immunoprecipitation results confirmed that Htt-552 could interact with Hsc70 and LAMP-2A. However, co-immunoprecipitation of Htt-552 with LAMP-2A was not efficient, especially for wt. The possible reason could be that LAMP-2A is a membrane protein and is not easy to be precipitated with current protocol. It was noticed that mutant Htt-552 had stronger interactions with the critical components of CMA than did wt Htt-552. To assess the role of Hsc70 and LAMP-2A in mediating Htt-552 degradation, alterations in LAMP-2A and Hsc70 protein levels were achieved with adenovirus mediated overexpression or siRNA technology. Regulation of LAMP-2A and Hsc70 levels appeared to affect Htt-552 degradation in cells, presumably by changing CMA activity. Although Hsp70 is highly homologous to Hsc70 and almost indistinguishable from it in chaperone activity [43], the manipulation of Hsp70 levels had no significant effect on Htt-552 accumulation. These data suggest that Htt is degraded by a LAMP-2A and Hsc70-dependent mechanism through CMA.
In vivo, both macroautophagy and CMA use lysosomes for substrate degradation and differ in the way substrates are delivered into lysosomes. Macroautophagy contributes to the highest percentage of lysosomal degradation inside cells. The cytosolic components must be sequestered inside the autophagosomes to be degraded upon fusion of the membrane of autophagosomes with lysosomes [7,8]. Microautophagy is well-studied in yeast, but the understanding of microautophagy in mammals is limited because mammalian homologs of the yeast genes have not been identified [16,18,21]. For CMA degradation, proteins are directly taken up by lysosomal membranes through LAMP-2A. Determining a protein's binding, uptake, and degradation in isolated intact lysosomes is the most direct test whether a protein is a CMA substrate [19,20,45,46]. Thus, it was critical to demonstrate in our study that Htt-552 is taken up by lysosomes. We successfully isolated intact lysosomes from mouse liver that were suitable for studying lysosomal uptake of Htt-552 [37]. Because the lysosomal stabilization may be affected by some proteins [39], we first confirmed that isolated lysosomes were not disrupted by either wt Htt-552 or mutant Htt-552. The translocation of Htt-552 into lysosomal lumen and degradation by lysosomal enzymes were confirmed by the following evidence: 1 addition of isolated lysosomes to cell lysates containing exogenously expressed Htt-552 robustly reduced Htt-552 levels; 2 in the above assay system, LAMP-2A and Hsc70 neutralizing antibodies inhibited the lysosome-mediated Htt-552 degradation; 3 CMA substrate GAPDH [20,45] reduced the lysosome-mediated Htt-552 degradation by competing for LAMP-2A; 4 sub-lysosomal fractionation [37] revealed Htt-552 in the lysosomal membranes and matrix. In addition, this study found that after inhibiting lysosomal degradation, more mutant Htt-552 was accumulated in the lysosomal membranes but little change occurred in the matrix. In contrast, more wt Htt-552 was accumulated in the matrix but little change happened in the lysosomal membranes. In this study, the differences in effects in CMA activity on Htt-552-18Q and Htt-552-100Q were noticeable. The change in Htt-552-18Q accumulation was bigger than that of Htt-552-100Q when CMA was inhibited by inhibitors or knockdown of LAMP-2A and Hsc70; or when CMA was activated by starvation or overexpression of functional proteins of CMA. In isolated lysosomes, mutant Htt-552 was less able than wt Htt-552 to cross lysosomal membranes and be degraded by lysosomal enzymes. These data suggest that expansion of a polyQ tract impairs Htt clearance through CMA.
All substrate proteins of CMA contain a motif consisting of the pentapeptide KFERQ that targets the proteins to lysosomes [47]. This motif is recognized by Hsc70, which interacts with the substrate proteins in the cytosol [33,48]. The putative KFERQ motif in Htt-552 was analyzed as described by Liang et al [49]. There are two KFERQ-like motifs found between amino acid 99-103 ( 99 KDRVN 103 ) and 248-252 ( 248 NEIKV 252 ) [47]. The amino acid sequence 14 LKSFQ 18 is also considered to be a KFERQ-like motif when it is phosphorylated [26]. Mutation of CMA recognition motif was achieved by changing hydrophilic and hydrophobic amino acids in KFERQ-like motif located between aa99-103 and 248-252. The results showed that mutation of amino acids in 99-103 but not in 248-252 reduced Htt-552 interaction with Hsc70 and also reduced clearance of Htt-552. These data suggest that a KFERQ-like motif located between aa99-103 is a functional site for recognition by Hsc70. Due to technical difficulty, mutating these sites in Htt-552-100Q was unsuccessful. Some studies have suggested degradation of Htt by CMA. For example, Thompson et al verified that phosphorylated Htt at S13 could form a KFERQ-like motif [26]. Furthermore, Bauer et al found that the degradation of Htt fragment could be increased by Hsc70 [25]. The present study identified a new functional putative KFERQ motif in Htt-552 that bound Htt-552 to Hsc70 and allowed uptake by lysosomes via LAMP-2A.
In summary, this study provides evidence elucidating a role of CMA in degradation of Htt-552. Expansion of the polyQ tract in Htt may slow its transport across lysosomal membranes and thus increase its accumulation in the cytosol. The confirmation of the involvement of the CMA pathway in Htt clearance is important, as this pathway is relatively selective for its substrates. Thus developing therapies that target an enhancement of CMAmediated clearance of mutant Htt may have great value.