The Wnt Signaling Antagonist Dapper1 Accelerates Dishevelled2 Degradation via Promoting Its Ubiquitination and Aggregate-induced Autophagy*

Background: Protein aggregates could be degraded through autophagy. Results: Dpr1 promotes pVHL-induced Dvl2 ubiquitination and mediates the Vps34-Beclin1 complex formation induced by protein aggregates. Conclusion: Protein aggregates stimulate autophagy initiation in a Dpr1-dependent manner. Significance: This study shows that protein aggregates can induce autophagy to facilitate their clearance. Autophagy is a regulated process that sequesters and transports cytoplasmic materials such as protein aggregates via autophagosomes to lysosomes for degradation. Dapper1 (Dpr1), an interacting protein of Dishevelled (Dvl), antagonizes Wnt signaling by promoting Dishevelled degradation via lysosomes. However, the mechanism is unclear. Here, we show that Dpr1 promotes the von Hippel-Lindau tumor suppressor (VHL)-mediated ubiquitination of Dvl2 and its autophagic degradation. Knockdown of Dpr1 decreases the interaction between Dvl2 and pVHL, resulting in reduced ubiquitination of Dvl2. Dpr1-mediated autophagic degradation of Dvl2 depends on Dvl2 aggregation. Moreover, the aggregate-prone proteins Dvl2, p62, and the huntingtin mutant Htt103Q promote autophagy in a Dpr1-dependent manner. These protein aggregates enhance the Beclin1-Vps34 interaction and Atg14L puncta formation, indicating that aggregated proteins stimulate autophagy initiation. Ubiquitination is not essential for the aggregate-induced autophagy initiation as inhibition of the ubiquitin-activation E1 enzyme activity did not block the aggregate-induced Atg14L puncta formation. Our findings suggest that Dpr1 promotes the ubiquitination of Dvl2 by pVHL and mediates the protein aggregate-elicited autophagy initiation.

and thus inhibit both Dvl-mediated canonical and noncanonical Wnt signaling (17)(18)(19). Recently, we found that Dpr1 can promote autophagy initiation by enhancing the Vps34-Beclin1-Atg14L complex formation, and loss of Dpr1 in the central nervous system results in motor coordination deficiency and accumulation of ubiquitinated proteins (20). However, it is unclear how Dpr1 accelerates Dvl2 degradation. In this study, we report that Dpr1 enhances the interaction between Dvl2 and pVHL, thereby promoting the Dvl2 ubiquitination. Moreover, we observed that aggregated proteins could induce autophagy to accelerate their degradation, and Dpr1 promoted this process.

EXPERIMENTAL PROCEDURES
Cell Culture-HEK293T, NRK, NRK GFP-DFCP, and NRK CFP-LC3 stable cell lines were cultured in DMEM containing 10% fetal bovine serum (Hyclone) supplemented with 4 mM L-glutamine in 5% CO 2 at 37°C, and the transfection was conducted with Vigofect (Vigorous Biotechnology, Beijing, China) according to the manufacturer's recommendations. Primary mouse embryonic fibroblast (MEFs) were generated from 13.5 day embryos from Dpr1 Ϫ/Ϫ mattings (18) and were maintained in DMEM with 10% FBS added. For starvation, cells were first washed three times with PBS and cultured in Hanks' balanced salt solution (Invitrogen) for the indicated time.
Immunoblotting and Immunoprecipitation-For immunoblotting, cells were lysed at 4°C for 15 min in lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 2 mM EDTA, 25 mM NaF, 1% Triton X-100) plus complete protease inhibitor mixture (Roche Applied Science). After 12,000 rpm centrifugation for 15 min at 4°C, the supernatants were resolved by SDS-PAGE and trans-ferred to NC membranes. The membranes were blocked with TBST containing 5% nonfat dry milk and incubated overnight at 4°C with primary antibodies. Membranes were washed with TBST three times, incubated for 1 h at room temperature with HRP-conjugated secondary antibodies (Amersham Biosciences), and then washed. Immunoreactive bands were visualized by chemiluminescence (ECL Plus from Thermo Scientific).
For immunoprecipitation, cells were lysed with lysis buffer containing complete protease inhibitor mixture (Roche Applied Science) for 30 min at 4°C. Lysates were cleared by centrifugation at 12,000 rpm for 15 min at 4°C. The supernatants were immunoprecipitated with specific antibodies and protein A-Sepharose (Zymed Laboratories Inc.). Immunocomplexes were washed three times with washing buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate) and analyzed by immunoblotting.
Adenoviral Expression-HA-Dvl2 was subcloned into the KpnI and XhoI sites of pShuttle-CMV vector (Invitrogen) and transferred into pAdEasy-1 vector (Invitrogen) by performing the Clonase LR recombination reaction (Invitrogen). The production of adenovirus and adenovirus infection was performed according to the manufacturer's protocols.
Immunofluorescence Microscopy-Immunofluorescence analysis was performed as described previously (15). Cells grown on glass coverslips in six-well plates were transfected with the indicated plasmids for 24 h. After washed twice with PBS, the cells were fixed with 4% paraformaldehyde for 20 min at room temperature, permeabilized for 10 min with 0.2% Triton X-100 (Sigma), and blocked with 5% BSA in PBST (phosphate-buffered saline plus Tween 20) for 1 h. Before subjecting to secondary antibodies (Jackson ImmunoResearch) for 1 h at room temperature, the cells were incubated with the primary antibodies overnight at 4°C and washed with PBST three times. Confocal laser scanning of fixed cells was detected using a Zeiss LSM 710 laser scanning microscope.
Quantitation of Protein Puncta-The number of CFP-LC3 and Atg14L-mCherry puncta was measured by Image-ProPlus software. CFP-LC3 and Atg14L-mCherry puncta were quantified from at least 20 different cells in three separate experiments. These measurements were done on randomly selected fields of view. In addition, all data reported that show differences in puncta formation were verified qualitatively in blind fashion by an independent observer.
Statistical Analysis-Statistical analyses were performed using a two-tailed unpaired t test. p values of Ͻ0.05 were considered statistically significant. n represents the number of independent experiments used for statistical analysis.

Dpr1
Facilitates Dvl2 Degradation via Autophagy-We have previously shown that Dpr1 could promote Dvl2 degradation in a lysosomal inhibitor-sensitive manner (19), and Dvl could be degraded via autophagy (15). To test whether Dvl degradation induced by Dpr1 is dependent on autophagy, we examined the effect of the class III phosphatidylinositol 3-kinase inhibitor 3-methyladenine (3-MA), which is commonly used to block autophagy (21,22), on the Dpr1-mediated Dvl turnover. As shown in Fig. 1A, co-expression of Dpr1 accelerated Dvl2 deg-radation, and this process was attenuated by 3-MA, but not by the proteasome inhibitor MG132, suggesting that Dpr1 induced Dvl2 degradation via the autophagy pathway. As a control, the GFP expression level was not affected by Dpr1, indicating that the reduced Dvl2 level was not due to the expression competition between proteins. The starvation-induced Dvl2 turnover was attenuated by Dpr1 knock-out, consistent with reduced autophagic flux shown by p62 in Dpr1 Ϫ/Ϫ cells (Fig.  1B). Furthermore, overexpression of Dpr1 dramatically increased the degradation rate of Dvl2 (Fig. 1C). In contrast, Dpr2, a homolog of Dpr1, could also interact with Dvl2 ( Fig.  1D), but it had no effect on Dvl2 stability (Fig. 1E). Together, these data indicate that Dpr1 facilitates Dvl2 degradation via autophagy.
Dpr1 Promotes the pVHL-mediated Dvl2 Ubiquitination-Our previous data indicate that Dvl2 ubiquitination is critical for its autophagy-mediated degradation (15). Then we tested whether Dpr1 affected Dvl2 ubiquitination. As shown in Fig.  2A, Dpr1 dramatically increased Dvl2 ubiquitination. In agreement with our previous report (15), when the cells were treated with the lysosomal inhibitor bafilomycin A1 (BFA1), nutrient starvation caused increased Dvl2 ubiquitination, and Dpr1 could further enhance the level ( Fig. 2A). The promoting effect of Dpr1 on Dvl2 ubiquitination was confirmed by knockdown of Dpr1 under both starvation and normal conditions (Fig. 2B). As pVHL, a component of the SCF-like ubiquitin E3 ligase complex, can induce Dvl2 ubiquitination (15), we examined whether Dpr1-enhanced Dvl2 ubiquitination is mediated by pVHL. Knockdown of VHL decreased the ubiquitinated Dvl2 level (Fig. 2C). Furthermore, VHL depletion rescued the Dpr1induced Dvl2 degradation (Fig. 2D). These data together indicate that Dpr1-induced ubiquitination and degradation of Dvl2 depend on pVHL.
As Dpr1 interacts with Dvl proteins, it may function as a scaffold to strengthen the association between Dvl2 and pVHL. To test this possibility, we first examined whether Dpr1 could interact with pVHL. As shown in Fig. 2E, pVHL was found in the Dpr1 immunoprecipitant, indicating that these two proteins interact with each other. pVHL contains two major domains as follows: the ␣ domain (153-204 amino acids) that binds to the elongin/Cullin2/Rbx1, and the ␤ domain (63-152 amino acids) that binds directly to the substrates such as HIF␣ (23). We found that pVHL interacted with Dpr1 via its ␣ domain (Fig. 2E). Domain mapping also revealed that the N terminus of Dpr1 was needed to interact with pVHL ( Fig. 2F), which is different for Dpr1 interaction with Dvl2 via its C terminus (19). To further confirm the scaffold function of Dpr1, we knocked down Dpr1 expression and found that reduction of HEK293T cells transfected with the indicated plasmids were treated with the indicated inhibitors for 6 h before being harvested for immunoblotting. GFP was used as a transfection control (Con). B, half-life of Dvl2 was prolonged in Dpr1 Ϫ/Ϫ MEFs. Wild-type or Dpr1 Ϫ/Ϫ MEFs were cultured in Hanks' balanced salt solution medium (starvation) for the indicated times before being harvested for immunoblotting with the indicated antibodies. p62 was detected as a autophagy marker. C, Dpr1 accelerated Dvl2 degradation in HEK293T cells. The cells transfected with the indicated plasmids were treated with the translation inhibitor cycloheximide (10 M) for various times before being harvested for immunoblotting. The relative band intensity of Dvl2 was quantified and is shown at the right. D, pVHL interacted with Dpr1 and its homolog Dpr2. HEK293T cells were transfected with the indicated constructs and harvested for anti-FLAG immunoprecipitation (IP) and then anti-Myc immunoblotting (IB). E, Dpr1 but not Dpr2 promoted Dvl2 degradation.
Dpr1 decreased the interaction between the Dvl2 and pVHL (Fig. 2G). These data indicate that Dpr1 facilitates the interaction between Dvl2 and pVHL, therefore promoting Dvl2 ubiquitination.
Dpr1 Mediates Dvl2 Aggregate-induced Autophagy-Our previous work has shown that Dpr1 induces Dvl2 degradation through the lysosome (19), and Dvl2 aggregates could be degraded by the autophagy-lysosome pathway (Fig. 1). Dvl proteins undergo aggregation in the cytoplasm through its DIX domain (24,25), and the aggregation-deficient mutants of Dvl2 are resistant to the autophagy-mediated degradation (15). Therefore, to access whether Dpr1-mediated Dvl2 degradation is dependent on its aggregation, we examined the protein levels of wild-type Dvl2 and aggregation-deficient mutants upon Dpr1 expression. Although Dpr1 overexpression accelerated the degradation of wild-type Dvl2, it had minimal effect on Dvl2 mutants M1 (F43S) and M2 (V67A/K68A) (Fig. 3A), indicating that Dvl2 aggregation is required for its Dpr1-induced clearance. Then we tested whether the aggregation of Dvl2 could stimulate the autophagy, and we found that Dvl2 overexpression could increase the puncta formation of LC3 in NRK cells (Fig. 3B). Furthermore, Dvl2 overexpression reduced p62 protein level in Dpr1 ϩ/ϩ but not in Dpr1 Ϫ/Ϫ MEFs (Fig. 3C), indicating that Dvl2 could induce the autophagy dependent on Dpr1.
Dpr1 Has a More General Role on Protein Aggregate-induced Autophagy-Similar to Dvl aggregates, some other aggregateprone proteins, such as p62 and Htt103Q (huntingtin fragment carrying an expanded polyglutamine), have been shown to undergo autophagy-mediated degradation (15,26,27). We thus assessed whether these aggregates promote autophagy under nutrient-rich conditions. Overexpression of p62-GFP or Htt103Q-GFP increased the number of the Atg14L-mCherry dots (Fig. 4A), whereas the p62 D69A mutant, which is aggregation-deficient (28), failed to do so (Fig. 4B). Similarly, ectopically expressed Htt103Q and wild-type p62, but not p62 D69A FIGURE 2. Dpr1 promotes pVHL-mediated Dvl2 ubiquitination. A, Dvl2 ubiquitination (Ub) was enhanced by Dpr1 overexpression under both nutrient-rich (N) and starvation (S) conditions. HEK293T cells transfected with the indicated constructs were treated with 100 nM BFA1 for 6 h. Ubiquitinated proteins were precipitated with nickel-nitrilotriacetate (Ni-NTA) beads, followed by anti-FLAG immunoblotting. B, Dpr1 knockdown decreased Dvl2 ubiquitination under both starvation (S) and nutrient-rich (N) conditions. HEK293T cells transfected with indicated constructs were treated with 100 nM BFA1 for 6 h. Ubiquitinated proteins were precipitated with anti-HA antibody, followed by anti-Myc immunoblotting (IB). The cell lysates were adjusted to equal Dvl2 protein levels to have a fair comparison with the Dvl2 ubiquitination levels. C, Dvl2 ubiquitination elevated by Dpr1 was decreased by VHL knockdown. D, Dpr1-promoted Dvl2 degradation was attenuated by VHL knockdown under starvation condition. E, pVHL interacted with Dpr1 via its ␣ domain. HEK293T cells transfected with the indicated constructs were harvested for immunoprecipitation (IP) by anti-HA antibody, followed by anti-Myc immunoblotting. F, Dpr1 interacted with pVHL through its N terminus. G, Dpr1 knockdown reduced the pVHL-Dvl2 interaction. MAY 8, 2015 • VOLUME 290 • NUMBER 19 mutant, led to LC3 puncta formation (Fig. 4C). These Atg14L-mCherry and CFP-LC3 dots were well co-localized with p62-GFP or Htt103Q-GFP under nutrient-rich conditions. These data clearly support the notion that protein aggregates can induce autophagy. Moreover, p62 increased Atg14L-mCherry dot formation in wild-type MEFs under nutrient-rich conditions, but this function was dramatically reduced by Dpr1 deletion (Fig. 4D), indicating that Dpr1 is critical for p62-induced autophagy initiation.

Dpr1 Mediates Protein Aggregate-initiated Autophagy
Although overexpression of p62 and Htt103Q did not influence the activity of AMP-activated protein kinase and mammalian target of rapamycin by examining the phosphorylation levels of their respective substrates ULK1 and S6K (Fig. 5A), they efficiently enhanced the interaction between Vps34 and Beclin1 (Fig. 5B), which was reduced by Dpr1 depletion (Fig.  5C). Dpr1 overexpression also accelerated the turnover of p62 and Htt103Q (Fig. 5, D-G). These data together suggest that protein aggregates can induce autophagy by increasing the Beclin1-Vps34 interaction, and this induction is dependent on Dpr1.
Ubiquitination Is Not Essential for the Aggregate-induced Autophagy Initiation-To explore the mechanism underlying the aggregate-inducted autophagy initiation, we tested whether the autophagy initiation effectors could preferentially recognize the aggregates but not their nonaggregates mutants. As shown in Fig. 6A, Atg14L bound to wild-type Dvl2 and the Dvl2 M1 mutant equally. Similarly, Vps15, a common regulator of the Vps34 complex, interacted with Dvl2 wild-type and the M1 mutant with a similar affinity (Fig. 6B). Interestingly, overexpression of Dvl2 and p62 induced Vps15 from a diffusing distribution to a puncta formation (Fig. 6C), consistent with protein aggregate-induced autophagy initiation. Because the FIGURE 3. Dvl2 stimulates autophagy initiation, which is dependent on its aggregation. A, Dpr1-enhanced degradation of wild-type Dvl2, but not its aggregation-deficient mutants. HEK293T cells transfected with the indicated constructs were harvested for immunoblotting with the indicated antibodies. B, wild-type Dvl2 induced the LC3 puncta formation in NRK cells. NRK cells transfected with the indicated plasmids were fixed with 4% paraformaldehyde for immunofluorescence. The puncta number per cell was quantified (n ϭ 3). Scale bars, 10 m. **, p Ͻ 0.01. Data were represented as mean Ϯ S.D. C, Dvl2 promoted p62 degradation in a Dpr1-dependent manner. Immunoblotting is shown of p62 levels in Dpr1 ϩ/ϩ and Dpr1 Ϫ/Ϫ MEFs infected with or without Dvl2-expressing adenovirus. D, wild-type Dvl2, but not its M1 mutant, facilitated the Atg14L-mCherry puncta formation. The puncta number per cell was quantified (n ϭ 3). Scale bars, 10 m. ***, p Ͻ 0.001. Data were represented as mean Ϯ S.D. E, wild-type Dvl2, but not its M1 mutant, enhanced the Dpr1mediated GFP-DFCP1 puncta formation. The puncta number per cell was quantified (n ϭ 3). Scale bars, 10 m. **, p Ͻ 0.01. Data were represented as mean Ϯ S.D. F, ectopic expression of wild-type Dvl2, but not M1 and M2 mutants, enhanced the Beclin1-Vps34 interaction. IP, immunoprecipitation; IB, immunoblot; WCL, whole cell lysate. G, Dvl2 further increased the Dpr1-promoted Beclin1-Vps34 interaction in HEK293T cells.

Dpr1 Mediates Protein Aggregate-initiated Autophagy
WD40 domain of Vps15 could recognize and bind to ubiquitin (29), we tested whether this domain is important for the aggregate-induced autophagy initiation, and we found that deletion of the WD40 domain had no effect on the puncta formation of Vps15 upon Dvl2 or p62 overexpression (Fig. 6C). Atg16L, another WD40 domain-containing protein, is required for the recognition of ubiquitinated endosomes upon Salmonella infection (30). Surprisingly, we observed a strong interaction between Dvl2 M1 and M2 mutants and Atg16L, which is important for LC3-phosphatidylethanolamine conjugation, compared with wild-type Dvl2 (Fig. 6D). However, no association was observed between Atg16L with either wild-type p62 or its D69A mutant. Taken together, these data indicate that none of the Vps15, Atg14L, and Atg16L preferentially recognizes the aggregates over their aggregation-deficient mutants.
Misfolded protein aggregates can be recognized by heat shock proteins (Hsp), which can help with refolding or mediating degradation of the aggregates through the proteasome pathway (31). It was recently reported that acetylated Hsp70 is required for the Beclin1-Vps34 complex formation and autophagosome formation (32). Therefore, we examined whether Hsp70 is involved in the autophagy formation stimulated by Dvl2 and p62 aggregates. However, we found that Dvl2 and p62 aggregates could not induce the puncta formation of Hsp70 (Fig. 6E). Similar results were observed with Hsp90. These data suggest that Hsp70 and Hsp90 are not involved in the aggregate-induced autophagy initiation. Ubiquitination and aggregation have been shown to have a mutually promoting effect, and ubiquitination plays an impor-tant role in protein aggregate degradation via autophagy (33,34). Ubiquitination also plays a vital role in the recruitment of the autophagic machinery to endosomes during Salmonella infection (30). Therefore, we wanted to explore whether ubiquitination is critical for the aggregate-induced autophagy. To this end, we examined whether p62 aggregates are co-localized with endogenous ubiquitin. As shown in Fig. 6F, p62 aggregates were nicely co-localized with LC3, consistent with the stimulatory effect of p62 aggregates in autophagosome formation. Although most of the p62-LC3 co-staining puncta were co-localized with endogenous ubiquitin, some of them were not. Similarly, some LC3 puncta induced by Dvl2 aggregates were not co-localized with endogenous ubiquitin (Fig. 6G). Taken together, these data indicate that ubiquitin is not essential for LC3 puncta formation induced by protein aggregates.
To further address the function of ubiquitin in aggregateinduced autophagy, we employed the ubiquitin-activation E1 enzyme inhibitor PYR-41 to block the ubiquitin-conjugation process (35). Although the ubiquitin-conjugation process was efficiently blocked by PYR-41 (data not shown), PYR-41 could not affect the Atg14L puncta formation induced by p62 and Dvl2 aggregates (Fig. 6H). These data further demonstrate that ubiquitination is not essential for the autophagy initiation induced by protein aggregates.

DISCUSSION
Dvl proteins are key mediators of canonical and noncanonical Wnt signaling. The activity and stability of Dvl proteins are tightly controlled (17, 18, 36 -38). We have reported that pVHL  MAY 8, 2015 • VOLUME 290 • NUMBER 19 mediates Dvl2 ubiquitination, which is associated with Dvl2 aggregation and enhances its degradation via autophagy (15). In this study, we found that Dpr1 promoted the interaction between Dvl2 and pVHL and therefore accelerated the autophagy-mediated degradation of Dvl2. As Dpr1 can also facilitate autophagy initiation (20), it seems that Dpr1 promotes the autophagy-mediated degradation of Dvl2 in two independent ways: on the one hand, Dpr1 enhances the autophagy initiation process by activating the Vps34-Beclin1-Atg14L complex; on the other hand, Dpr1 acts as an adaptor to increase the ubiquitination of Dvl2 mediated by the E3 ligase pVHL. Both Dpr1 functions lead to the accelerated degradation of Dvl2 via autophagy.

Dpr1 Mediates Protein Aggregate-initiated Autophagy
Although it has been known for a long time that aggregated proteins can be degraded in lysosomes through autophagy (8, 39 -41), it is unclear whether aggregated proteins could stimulate autophagy and enhance their degradation in a feedback manner. Here, we have demonstrated that the aggregationprone proteins such as Dvl2, p62, and Htt103Q can increase the Vps34-Beclin1 interaction and Atg14L puncta formation, therefore facilitating the initiation of autophagy. We have also demonstrated that Dpr1 is critical in this process as p62-induced Atg14L puncta formation was much reduced in Dpr1 knock-out cells.
How the protein aggregates activate the autophagy initiating machinery is still an open question. Although wild-type Dvl2 interacts with LC3 stronger than its aggregation-deficient M1 and M2 mutants (15), none of the autophagy initiation machinery components, Vps15, Atg14L, and Atg16L, showed a better binding ability with wild-type Dvl2 than with the M1 mutant, indicating that these machinery components may not specifically recognize the aggregated substrates. This notion was supported by the data that Atg16L did not apparently interact with either wild-type p62 or its aggregation-deficient D69A mutant.
Ubiquitination plays an important role in the degradation of protein aggregates via autophagy (33,34). Ubiquitin has been shown to be critical for the recruitment of the autophagic machinery to endosomes to mediate pathogen clearance during Salmonella infection (30). However, aggregate-induced autophagy may not necessarily require ubiquitination as some of the LC3 puncta induced by aggregates were stained negatively for endogenous ubiquitin. In addition, the inhibition of ubiquitinconjugated E1 activity by PYR-41 did not block the aggregateinduced Atg14L puncta formation. However, our immunofluorescence analysis revealed that many p62 and Dvl2 aggregate-induced LC3 puncta were positively stained with ubiquitin. Therefore, we could not exclude the possible function of ubiquitination in selective aggrephagy. Other proteins could bridge the aggregates directly to the autophagy machin-ery. For instance, optineurin has been shown to interact with protein aggregates associated with neurodegenerative diseases via its C-terminal coiled-coil domain (42). This interaction is ubiquitin-independent. As optineurin contains an LC3-interacting motif, it promotes aggrephagy by recruiting aggregates to autophagosomes (42). It will be interesting to test whether optineurin can mediate aggregate-induced autophagy initiation by promoting Beclin1-Vps34 complex formation. Nonetheless, this study provides a new for Dpr1 in aggrephagy.