Sorting nexin 17 regulates ApoER2 recycling and reelin signaling

Abstract

ApoER2 is a member of the low density-lipoprotein receptor (LDL-R) family. As a receptor for reelin, ApoER2 participates in neuronal migration during development as well as synaptic plasticity and survival in the adult brain. A previous yeast two-hybrid screen showed that ApoER2 is a binding partner of sorting nexin 17 (SNX17) - a cytosolic adaptor protein that regulates the trafficking of several membrane proteins in the endosomal pathway, including LRP1, P-selectin and integrins. However, no further studies have been performed to investigate the role of SNX17 in ApoER2 trafficking and function. In this study, we present evidence based on GST pull-down and inmunoprecipitation assays that the cytoplasmic NPxY endocytosis motif of ApoER2 interacts with the FERM domain of SNX17. SNX17 stimulates ApoER2 recycling in different cell lines including neurons without affecting its endocytic rate and also facilitates the transport of ApoER2 from the early endosomes to the recycling endosomes. The reduction of SNX17 was associated with accumulation of an ApoER2 carboxy-terminal fragment (CTF). In addition, in SNX17 knockdown cells, constitutive ApoER2 degradation was not modified, whereas reelin-induced ApoER2 degradation was increased, implying that SNX17 is a regulator of the receptor's half-life. Finally, in SNX17 silenced hippocampal and cortical neurons, we underscored a positive role of this endosomal protein in the development of the dendritic tree and reelin signaling. Overall, these results establish the role of SNX17 in ApoER2 trafficking and function and aid in identifying new links between endocytic trafficking and receptor signaling.

Figure 1. SNX17 interacts with the NPxY motif of ApoER2, and both proteins colocalize after receptor endocytosis.

(A) HEK 293 cells were transfected with different myc-tagged SNX17 constructs, and their lysates were used for GST pull-down assays using GST or GST-ApoER2. The presence of SNX17 was evaluated by western blot with an anti-myc antibody. GST fusion proteins were detected by western blot using anti-GST antibody. F1, F2 and F3 indicates the three subdomains or modules of the FERM domain .(B) HEK 293 cells were transfected with HA-ApoER2 wild-type or mutated (NPxY/A) constructs. Cell lysates were used for a pull-down assay using GST or GST-SNX17. The receptor was evaluated by detecting the HA epitope. In both cases, Lys corresponds to 10% of the cell lysate used for the pull-down assay. (C) Cell extracts obtained from cells transiently transfected with myc-SNX17 and HA-ApoER2 were immunoprecipitated with anti-myc and probed for ApoER2 with the anti-HA antibody. Lys corresponds to 2% of the cell lysate used for the coinmunoprecipitation. (D) HeLa cells were transfected with HA-ApoER2, RAP, and myc-SNX17. Cells were incubated with anti-HA antibody for 1 h at 4°C, and receptor internalization was allowed for 10 min at 37°C. Cells were fixed and analyzed by immunofluorescence. Bar, 10 μm.

Figure 2. SNX17 knockdown diminishes surface levels of ApoER2 by decreasing its recycling.

(A) HEK 293 cells or (B) N2a cells infected with a lentiviral vectors expressing shRNA against human or mouse SNX17 or empty pLKO vector were transfected with HA-ApoER2. Cells were lysed with 1% Triton X-100 in PBS and analyzed by western blot. (C) HEK 293 clones transfected with HA-ApoER2 and RAP were used to analyze the ratio of cell surface to total ApoER2 by FACS, as described in the Methods section. The graphic shows the ratio of the values of non-permeabilized versus permeabilized cells considering the control condition as 100%. (D) N2a clones expressing ApoER2 and SNX17 silenced or control were treated as in C. (E) Control or SNX17 knockdown HEK293 clones were transfected with a plasmid for mMeg4, a construct of megalin carrying the fourth ligand binding domain, the transmembrane domain, and the cytosolic domain. The receptor was determined in these cells by FACS. (F) SNX17 knockdown HEK293 clones expressing HA-ApoER2 were transfected with a shRNA-resistant mouse myc-SNX17. The presence of ApoER2 and SNX17 proteins were analyzed using a chicken anti-HA antibody and a mouse anti-Myc antibody respectively, in permeabilized and non-permeabilized conditions. (G) Control or SNX17 knockdown HEK293 clones expressing HA-ApoER2 were incubated with an Alexa 488-conjugated anti-HA antibody for 1 h at 4°C. The temperature was increased to 37°C for the indicated time period, and theremaining surface bound antibody was removed with an acid wash. The intracellular antibody was detected in the cells by flow cytometry. The total antibody bound to the cell was also determined. The endocytic rate was calculated by subtracting the value of the cells exposed to the acid wash at time 0 (A₀) from each time point, and dividing by A_0. (H) The same experimental procedure described in G was performed with control or SNX17 knockdown N2a cells expressing HA-ApoER2. (I) The same HEK293T cells used in G were utilized in a recycling experiment. The cells were labeled with an Alexa 488-conjugated anti-HA antibody for 30 min at 37°C. The recycled receptor was then chased at the surface with a quenching anti-Alexa 488 antibody. The cells were analyzed by flow cytometry, and the percentage of the initial fluorescence remaining at each time point was calculated as the difference in the fluorescence between time 0 and each chased time point. Every time (including time 0) was normalized to the non-chased value. The percentage of recycling efficiency was calculated by subtracting the percentage of internal fluorescence from 100. (J) Control or SNX17 knockdown N2a clones expressing HA-ApoER2 were treated as in C. *p<0.05, **p<0.01, ***p<0.001.

Figure 3. SNX17 knockdown induces the retention of ApoER2 in early/recycling endosomes.

(A) Control (pLKO) or SNX17 knockdown N2a cells expressing ApoER2 were subjected to subcellular fractionation to isolate endosomal compartments by a discontinuous sucrose gradient. Samples of the different fractions were resolved in Tris/Tricine gels and analyzed by western blot. The input corresponds to 5% of the lysate. EE/RE: early endosome/recycling endosome; LE: late endosome; fractions were free of endoplasmic reticulum (RAP) or Golgi (γ-Adaptin) contamination. (B) Quantification of ApoER2 in different compartments normalized by the Input. *p<0.05.

Figure 4. SNX17 facilitates ApoER2 trafficking from the early endosome to the recycling endosome.

HeLa cells silenced for SNX17 or control pLKO cells were transfected with HA-ApoER2, RAP, and GFP-Rab5 (A); GFP-Rab11 (B); or GFP-Rab7 (C). Two days after transfection, cells were incubated with anti-HA antibody for 1 h at 4°C, incubated for 40 min at 37°C to allow for receptor internalization, and the remaining surface receptor was removed by acid wash. Cells were washed, permeabilized, and incubated with Alexa 594-conjugated goat anti-mouse IgG. Images were captured by confocal microscopy, and Mander's colocalization index and Pearson's coefficient were calculated in 10 cells for each condition. **p<0.01, ***p<0.001. Bars, 10 μm.

Figure 5. SNX17 does not regulate the half-life of ApoER2 under basal conditions.

(A) HEK 293 cells treated with shRNA against SNX17 and pLKO control clones were transfected with HA-ApoER2 and RAP, incubated with medium containing [35S]Met/Cys for 90 min, and chased with medium without [35S]Met/Cys for 1, 3, 6, 9, and 20 h, followed by immunoprecipitation with anti-HA antibody. The samples were separated by SDS-PAGE and subjected to autoradiography. (B) The radioactive intensity was plotted against time. In both conditions, ApoER2 is degraded with similar kinetics. (C) Western blot of equal amounts of cell lysate of N2a cells (pLKO and SNX17 KD) expressing ApoER2, which was detected with anti-HA antibody. The ER precursor and the mature fully glycosylated forms are shown. (D) There is no difference in the amount of mature form with respect to the precursor in SNX17 knockdown cells indicating that ApoER2 degradation was not affected by the lack of SNX17 (n = 3).

Figure 6. SNX17 knockdown increases the ApoER2 CTF level.

Control or SNX17 knockdown (A) HEK293 cells and (B) N2a cells expressing HA-ApoER2 were lysed in PBS-T. The samples were separated on Tris-Tricine gels and analyzed by western blot. (C) DIV 4 cortical mouse neurons were infected with a lentiviral vector expressing SNX17 shRNA or empty pLKO at an MOI of 1. Three days after the infection, the cells were lysed and the lysates were resolved on Tris-Tricine gels. (D) Control or SNX17 knockdown HEK293 cell clones were transfected with the ApoER2 plasmid and either the mouse myc-SNX17 plasmid or pcDNA3 as a control. The resultant cell lysates were analyzed by western blot. (E) The band intensity was quantified to determine the ApoER2 CTF level normalized to actin. In all conditions, the bands were normalized to the control condition (empty pLKO and pcDNA3). The figure shows the average of three independent experiments. *p<0.05, **p<0.01, ns p>0.05. (F) Control or SNX17 knockdown N2a cells expressing ApoER2 were treated with DAPT or DMSO for 16 h, and then the presence of ApoER2 and its CTF were determined by western blot. (G) Quantification was performed as previously described, using pLKO cells treated with DMSO as the control condition. **p<0.001 and ns p>0.05.

Figure 7. SNX17 regulates the degradation of ApoER2 induced by reelin.

(A) N2a-ApoER2 cells silenced or not for SNX17 were incubated with 100 μM cycloheximide for 6 h. After 1 h of serum depletion, cells were incubated with the respective conditioned media for 5 h. ApoER2 was detected by western blot. (B) Band intensities were quantified to determine the ratio of ApoER2 versus actin as the loading control. The graphic indicates the average of three independent experiments. *p<0.05; **p<0.01.

Figure 8. SNX17 knockdown diminishes the surface level of ApoER2 and reelin-induced dendritic development in neurons.

(A, B) The cell surface level of ApoER2 was determined in DIV 5 mouse cortical neurons co-transfected with HA-ApoER2 and either SNX17 shRNA or pLKO. The cell surface receptor was labeled 48 h after transfection with a mouse anti-HA antibody. To control for the absence of permeabilization, cells were simultaneously incubated with an antibody against the cytoplasmic tail of ApoER2. The intracellular ApoER2 was detected thereafter in the fixed and permeabilized neurons with a chicken anti-HA antibody. Images of individual cells (n = 10 for each condition) were captured by confocal microscopy and analyzed using ImageJ software, selecting the threshold for each channel to avoid background. Total fluorescence was calculated by adding the fluorescence of the permeabilized and non-permeabilized channels. (C, D) Mouse dissociated hippocampal neurons were transfected with GFP and the corresponding shRNA, treated with reelin for 3 days, fixed and analyzed by immunofluorescence. Images were captured by confocal microscopy and used for Sholl analysis (n = 20 cells per condition). (E) The length of dendrites of reelin treated cells was significantly reduced in SNX17 knockdown neurons *p<0.05; **p<0.01; ***p<0.001. Bars, 20 μm.

Figure 9. SNX17 has a positive role in the reelin signaling pathway.

(A) Mouse dissociated cortical neurons were transfected at DIV 4 with a GFP expression plasmid and the corresponding shRNA plasmid. At DIV 7, cells were treated with reelin, fixed, and analyzed by immunofluorescence using anti-phospho-Dab1 and anti-βIII-tubulin antibodies. Images of individual cells were captured, and the integrated fluorescence intensity of the soma was calculated using ImageJ software. Phosphorylation of Dab1 was quantified in cells under each condition, and the intensity of βIII-tubulin was used for normalization. In (B), the quantification of the fluorescence intensity is shown, of both cells transfected with pLKO plasmid (control), and with shSNX17 in the presence of reelin (Reelin) or DMSO (Mock). In (C), the same quantitative analysis of non-transfected cells present at each experimental condition is shown. **p<0.01, ***p<0.001.

Figure 10. The reelin signaling pathway is impaired in SNX17 knockdown cells.

DIV 4 cortical mouse neurons were infected with the a lentiviral system expressing the parental plasmid pLKO (Control) or SNX17 shRNA (shSNX17) at MOI 1; three days after infection, the cells were incubated with reelin-containing medium. Next, cells were lysed and analyzed by western blot. Dab1 was immunoprecipitated (IP) with an anti-Dab1 antibody and analyzed by western blot using an anti-phosphotyrosine antibody. The Dab1 downstream targets AKT and GSK3β were also analyzed by western blot. For all of the proteins analyzed, the reelin-induced phosphorylation was reduced in the SNX17 knockdown cells.

Figure 11. SNX17 knockdown decreases n-cofilin phosphorylation induced by reelin.

(A) Dissociated mouse cortical neurons were transfected at DIV 5 with a GFP expression plasmid and the corresponding shRNA plasmid. At DIV 7, cells were treated with reelin, fixed, and analyzed by immunofluorescence to detect phospho-cofilin and βIII-tubulin. Images were captured, and the integrated fluorescence intensity of the soma was calculated using ImageJ software. Phosphorylation of n-cofilin was quantified in cells at each condition, and the intensity of βIII-tubulin was used for normalization. In (B), the quantification of the fluorescence intensity is shown of both cells transfected with pLKO plasmid (Control), and with shSNX17 in the presence of reelin (Reelin) or with DMSO (Mock). In (C), the same quantitative analysis of non-transfected cells present in each experimental condition is shown. **p<0.01; ***p<0.001.