Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Sep 3;6:8200.
doi: 10.1038/ncomms9200.

Activity-regulated Trafficking of the Palmitoyl-Acyl Transferase DHHC5

Affiliations
Free PMC article

Activity-regulated Trafficking of the Palmitoyl-Acyl Transferase DHHC5

G Stefano Brigidi et al. Nat Commun. .
Free PMC article

Abstract

Synaptic plasticity is mediated by the dynamic localization of proteins to and from synapses. This is controlled, in part, through activity-induced palmitoylation of synaptic proteins. Here we report that the ability of the palmitoyl-acyl transferase, DHHC5, to palmitoylate substrates in an activity-dependent manner is dependent on changes in its subcellular localization. Under basal conditions, DHHC5 is bound to PSD-95 and Fyn kinase, and is stabilized at the synaptic membrane through Fyn-mediated phosphorylation of a tyrosine residue within the endocytic motif of DHHC5. In contrast, DHHC5's substrate, δ-catenin, is highly localized to dendritic shafts, resulting in the segregation of the enzyme/substrate pair. Neuronal activity disrupts DHHC5/PSD-95/Fyn kinase complexes, enhancing DHHC5 endocytosis, its translocation to dendritic shafts and its association with δ-catenin. Following DHHC5-mediated palmitoylation of δ-catenin, DHHC5 and δ-catenin are trafficked together back into spines where δ-catenin increases cadherin stabilization and recruitment of AMPA receptors to the synaptic membrane.

Figures

Figure 1
Figure 1. Changes in DHHC5 palmitoylation and localization following increased neuronal activity.
(a,b) Autopalmitoylation of DHHC5 is not altered following cLTP (glycine) stimulation (P=0.107, F3,8=2.83, n=3 blots from 3 cultures). Exclusion of NH2OH was used as a control for the specificity of biotin labelling. (c) Confocal image of 14 DIV neurons demonstrating co-localization of DHHC5 and PSD-95. (d,e) The IntDen of PSD-95 puncta is altered 40 min after treatment with cLTP and cLTD relative to control cells (ctrl), (P<0.001, F2,84=15.28, n=38, 35, 14), whereas the IntDen of gephyrin is not (P=0.354, F2,93=1.05, n=37, 39, 20). Confocal images of 14 DIV neurons (f,h) demonstrating increased co-localization of DHHC5 with PSD-95 (P<0.001, F2,84=22.98) (f,g), but no change in co-localization of DHHC5 and gephyrin (P=0.114, F2,93=2.21) (h,i) 40 min after cLTP. Co-localized puncta are denoted by white arrowheads. (j) Confocal images of 14 DIV neurons transfected with GFP–δ-catenin and immunostained for PSD-95 and DHHC5 (n=45 cells). Scale bars, 20 μm (c) and 5 μm (f,h,j). n=cells from three separate cultures. All graphs display mean±s.e.m. *P<0.05, ***P<0.001; one-way analysis of variance; Tukey's post-hoc test. (a) Five per cent of whole-cell lysates was loaded as input. Full-length blots of a presented in Supplementary Fig. 5.
Figure 2
Figure 2. Activity enhances DHHC5 trafficking from spines.
(a) Confocal images of 14 DIV neurons demonstrating partial co-localization of GFP–DHHC5 and PSD-95. (b) High-magnification confocal images of GFP–DHHC5 fluorescence (lower panels pseudocolored as a heat map) and DsRed before and after glycine stimulation. (c) GFP–DHHC5 fluorescence decreases transiently within spines after glycine stimulation (P<0.001, F9,220=42.45; n=221 spines, 8 cells). Treatment with AP5, DL-2-Amino-5-phosphonopentanoic acid, abolishes this (P=0.46, F9,141=0.974, n=142 spines, 6 cells). (d) Representative image of GFP–DHHC5 within masks made of spines (dashed white line) or dendritic shaft (dashed yellow line). High-magnification confocal images of GFP–DHHC5 and RFP–δ-catenin WT (e) or C960-1S (g; lower panels pseudocoloured as a heat map) within a single spine and region of dendrite shaft (traced with white and yellow dashed lines, respectively) before and after stimulation. (f) GFP–DHHC5 co-localized with RFP–δ-catenin WT decreases in spines (P<0.001, F9,6=91.14, n=7 cells) and increases in shafts (P<0.001, F9,6=58.24, n=7 cells) transiently following cLTP. (h) RFP–δ-catenin WT is recruited to spines (P<0.001, F9,6=7.549, n=7 cells) and is depleted from shafts (P<0.001, F9,6=11.83, n=7 cells) following activity, whereas RFP–δ-catenin C960-1S is unchanged in both spines (P=0.43, F9,6=1.025, n=7 cells) and shafts (P=0.64, F9,6=0.783, n=7 cells). n=number of cells or spines from three to five separate cultures. Scale bars, 20 μm (a) and 1 μm (b,d,e,g). All graphs show mean±s.e.m. (c,f,h). Asterisks and cross-hatches (f,h) above data points indicate significance relative to before stimulation within spines or dendrites, respectively. *P<0.05, **/##P<0.01, ***/###P<0.001; repeated-measures one-way analysis of variance, Tukey's post-hoc test.
Figure 3
Figure 3. Activity-induced DHHC5 endocytosis and trafficking on recycling endosomes.
(a,b) Fourteen DIV hippocampal neurons were stimulated with cLTP and then biotinylated at the indicated time points. Lysates were immunoprecipitated with neutravidin-coated beads to isolate all surface proteins and blots probed with the indicated antibodies. (b) Intensity of protein levels in the surface fraction, normalized to whole-cell input levels. DHHC5: P<0.001, F4,10=24.35; δ-catenin: P=0.016, F4,10=5.16; GluA1: P=0.036, F4,10=3.933; N-cadherin: P=0.724, F4,10=0.52). n=3 blots, 3 cultures. (c) Confocal image of 14 DIV hippocampal neurons demonstrating co-localization between TfR and DHHC5 (n=38 cells). (d) High-magnification confocal images of GFP–DHHC5 and TfR-mCh fluorescence in a single spine (white dashed line) over time (n=6 cells, 3 cultures). (e) IntDen of GFP–DHHC5 (P<0.001, F9,45=7.21) and TfR-mCh (P=0.002, F9,45=4.71) within spines. (f) Per cent GFP–DHHC5 co-localized with TfR-mCh in spines (P<0.001, F9,45=7.21). (g) Confocal image of hippocampal neurons demonstrating co-localization between VPS-35 and DHHC5 at 14 DIV (n=22 cells). Scale bars, 5 μm (c,g) and 1 μm (d). Graphs show mean±s.e.m. (b) *P<0.05, **P<0.01; one-way analysis of variance (ANOVA), Tukey's post-hoc test, (e,f) */#P<0.05, **/##P<0.01, ***P<0.001; repeated-measures one-way ANOVA, Tukey's post-hoc test. Asterisks and cross-hatches (e) above data points indicate significance relative to before stimulation for GFP–DHHC5 and TfR-mCherry, respectively. (a) Fifty per cent of whole-cell lysates were loaded as inputs. Full-length blots are presented in Supplementary Fig. 5.
Figure 4
Figure 4. Activity regulates the association of DHHC5 with δ-catenin and PSD-95.
(ad) Fourteen to 16 DIV hippocampal neurons were stimulated, lysed at the indicated time points, lysates immunoprecipitated and blots probed with the indicated antibodies. (a,d) DHHC5–δ-catenin interactions increase transiently following cLTP (P=0.0006, F6,14=8.172). (b,d) Reverse co-IPs show the same (P=0.004, F6,14=4.98). (c,d) PSD-95 /DHHC5 interactions decrease transiently following cLTP (P<0.001, F6,14=32.89). n=3 blots from 3 separate cultures. (eh) HEK293T cells were transfected with HA-DHHC5, GFP–δ-catenin or PSD-95-GFP constructs, for 36 h. Lysates were immunoprecipitated using anti-GFP and blots probed with the indicated antibodies. (e,f) DHHC5 WT associated with δ-catenin WT or C960-1S (P=0.0011, F3,8=7.46). (g,h) DHHC5 WT and ΔPDZb associated with PSD-95 or δ-catenin (P=0.019, F3,8=6.65). n=3 blots from 3 separate cultures. (i) Confocal image of 14 DIV neurons transfected with GFP–δ-catenin and HA-DHHC5 or DHHS5, and immunostained for PSD-95. Scale bar, 5 μm. (j) Per cent GFP–δ-catenin co-localized with PSD-95 (P<0.001, F2,59=8.54; n=24 (vector), 20 (HA-DHHC5) and 18 (HA-DHHS5) cells, 3 cultures). All graphs display mean±s.e.m. */#P<0.05, **/##P<0.01, ***P<0.001; one-way analysis of variance; Tukey's post-hoc test. Asterisks and cross-hatches (d) above data points indicate significance relative to unstimulated controls for DHHC5/δ-catenin or DHHC5/PSD-95 and δ-catenin/DHHC5, respectively. (a,b,c,e,g) Five percent of whole-cell lysate were loaded as inputs, and (a,b,c) are from the same blots but with different exposure times. Full-length blots are presented in Supplementary Fig. 5.
Figure 5
Figure 5. PSD-95 enhances the binding and tyrosine phosphorylation of DHHC5 by Fyn kinase.
(ad) Fourteen DIV hippocampal neurons were stimulated, lysed at the indicated time points, immunoprecipitated with anti- (a) DHHC5 or (c) Fyn and blots probed with the indicated antibodies. (a,b) DHHC5 tyrosine phosphorylation (phY; P=0.007, F6,14=7.749) and DHHC5/Fyn interactions (P<0.001, F6,14=11.06) are decreased transiently, whereas DHHC5/AP2μ interactions are increased transiently following cLTP (P=0.0076, F6,14=7.78). (c,d) Non-phosphoryated tyrosine 420 Fyn levels (P=0.0001, F6,14=11.24) and STEP61/Fyn interactions (P=0.004, F6,14=5.549) are transiently increased following cLTP. (ei) HEK293T cells were transfected with the indicated HA-DHHC5, PSD-95-GFP or Fyn constructs for 36 h. (e,f) PSD-95 enhances Fyn/DHHC5 WT but not Fyn/DHHC5 ΔPDZb interactions (P=0.0014, F2,6=24.11). (e,g) PSD-95 enhances Fyn-mediated phosphorylation of DHHC5 WT but not DHHC5 ΔPDZb (P=0.0104, F2,6=10.75). (h,i) Lysates were biotinylated and immunoprecipitated with neutravidin-coated beads and blots probed with the indicated antibodies. (i) Normalized intensity of DHHC5 in the surface fraction (P<0.001, F7,16=43.23). (b,d,f,g,i) n=3 blots from 3 separate cultures. All graphs display mean±s.e.m. */#P<0.05, **/##P<0.01, ***P<0.001; one-way analysis of variance; Tukey's post-hoc test. Asterisks and cross-hatches above data points indicate significance relative to unstimulated controls for (b) DHHC5 phY or AP2μ, and Fyn or (d) Y420-Fyn and STEP61, respectively. (a,c,e,h) Five per cent of whole-cell lysates were loaded as inputs and (a,c) are from the same blots but with different exposure times. Full-length blots are presented in Supplementary Fig. 5.
Figure 6
Figure 6. Phosphorylation of DHHC5 regulates its association with endocytic proteins and its subcellular localization.
(a) Schematic depiction of DHHC5 constructs N-terminally tagged with GFP or HA (not shown here) and illustrating the approximate localization of transmembrane domains (grey boxes), the DHHC motif, a putative Fyn-binding site (dashed line; RLLPTGP), a putative AP2μ-binding site (dashed line; YDNL) and the PDZ-binding motif required for binding PSD-95 (solid line; EISV). (be) HEK293T cells were transfected with the indicated HA-DHHC5 and Fyn constructs for 36 h, lysates immunoprecipitated with an HA antibody and blots probed with the indicated antibodies. (b,c) Fyn binding of the DHHC5 P520,3A mutant is reduced, but not for for the Y533E mutant (P=0.0153, F2,6=9.07). (b,d) Fyn-mediated tyrosine phosphorylation is attenuated in DHHC5 P520,3A and Y533E mutants (P<0.001, F5,12=20.04). (b,e) Fyn decreases AP2μ association with DHHC5 WT, but not P520,3A or Y533E mutants (P=0.018, F5,12=6.29). n=3 blots from 3 separate cultures. (f) Confocal images of 14 DIV neurons transfected with the indicated GFP–DHHC5 construct and immunostained for TfR and VGluT1. Scale bar, 5 μm. (g) DHHC5 P520,3A increases and Y533E decreases co-localization with TfR (P<0.001, F2,58=24.05), and (h) decreases and increases co-localization with VGlut1, respectively (P<0.001, F2,58=23.5). n=22 (WT), 18 (P520,3A) and 21 (Y533E) cells from 3 cultures. Co-localized puncta are denoted by white arrowheads. All graphs display mean±s.e.m. *P<0.05, **P<0.01, ***P<0.001; one-way analysis of variance; Tukey's post-hoc test. (b) Five per cent of whole-cell lysates were loaded as inputs. Full-length blots are presented in Supplementary Fig. 5.
Figure 7
Figure 7. PSD-95 and Fyn control DHHC5 turnover in spine heads.
(a,c,e) High-magnification confocal images of 14–16 DIV hippocampal neurons transfected with the indicated GFP–DHHC5, DsRed, PSD-95–RFP or Fyn constructs. GFP–DHHC5 fluorescence within a photobleached ROI (red circles) was analysed over 300 s (cells were initially photobleached at 0 s, white asterisks, within a 1-μm diameter ROI). Scale bar, 1 μm. (b,d,f) Relative fluorescence recovery of GFP–DHHC5. Solid lines represent single exponential fit. Points with error bars represent the mean±s.e.m. Statistical tests compare the plateau values from exponential fits±s.e.m. Neurons were obtained from three to five separate cultures. (a,b) Overexpression of PSD-95 significantly reduces the mobility of DHHC5 WT, but not ΔPDZb (P<0.001, F3,77=257.8; n=29 (DHHC5 WT+DsRed), 17 (WT+PSD-95-RFP), 20 (ΔPDZb+DsRed), 15 (ΔPDZb+PSD-95-RFP)). (c,d) Co-expression of Fyn and PSD-95 further decreases the mobile fraction of DHHC5 WT (P<0.001, F4,94=753.1; n=29 (WT+DsRed), 16 (WT+Fyn+DsRed), 19 (WT+Fyn+PSD-95), 16 (ΔPDZb+Fyn+DsRed), 19 (ΔPDZb+Fyn+PSD-95)). (e,f) Fyn does not impact the mobility of DHHC5 P520,3A nor Y533E (P<0.001, F4,98=1318; n=29 (WT+DsRed), 22 (P520,3A+DsRed), 17 (P520,3A+Fyn+DsRed), 18 (Y533E+DsRed), 17 (Y533E+Fyn+DsRed)). (g) The mobile fraction of GFP-DHHC5 (relative fluorescence fraction within the ROI at the 5-min time point normalized for photobleaching; mean±s.e.m.; P<0.001, F11,213=29.02; n values indicated above). *P<0.05, **P<0.01, ***P<0.001; one-way analysis of variance; Tukey's post-hoc test relative to control cells expressing WT+DsRed. n.s., not significant.
Figure 8
Figure 8. DHHC5 internalization is required for activity-induced δ-catenin trafficking.
(a,b) HEK293T cells were transfected with control shRNA (shRNA-c) or shRNA against DHHC5 (shRNA) plus the indicated HA-DHHC5 constructs (*shRNA resistance) and blots probed with the indicated antibodies (P=0.024, F3,8=5.484, n=3 blots from 3 separate cultures; one-way analysis of variance (ANOVA)). (c) Confocal images of 14 DIV hippocampal neurons transfected with GFP–δ-catenin and the indicated shRNA and HA-DHHC5* constructs. Cells were stimulated with cLTP (+Gly) or control buffer lacking glycine (–Gly), fixed 20 min after stimulation and immunostained with the indicated antibodies. Scale bar, 5 μm. (d) The cLTP-induced increase in δ-catenin/PSD-95 co-localization is abolished in DHHC5 knockdown cells or those expressing the DHHC5 Y533E mutant (P<0.001, F7,169=17.44). The n-values, indicating neuron numbers from –Gly and +Gly cells, respectively, among three separate cultures are as follows: shRNA-c (23, 28), shRNA (17, 16), shRNA+WT* (22, 23) and shRNA+Y533E* (28, 22). Scale bar, 5 μm. Asterisks denote significance (b,d) among all groups relative to (b) shRNA-c+WT cells or (d) to –Gly shRNA-c cells. All graphs display mean±s.e.m. *P<0.05, ***P<0.001, one-way ANOVA, Tukey's post-hoc test. Full-length blots of a are presented in Supplementary Fig. 5.
Figure 9
Figure 9. Activity-induced AMPAR surface insertion requires DHHC5 endocytic cycling.
(a) Confocal images of 14–16 DIV neurons transfected with SEP-GluA1, DsRed and the indicated shRNA and HA-DHHC5* constructs (*shRNA-resistance). SEP-fluorescent puncta are pseudocoloured as heat maps. Cells were imaged before and 20 min after glycine stimulation. HA-DHHC5* expression was confirmed by post-hoc immunostaining for HA. Scale bar, 5 μm. (b) IntDen of SEP-GluA1 puncta in hippocampal neurons normalized to the same puncta before glycine treatment. The n-values, indicating neuron numbers from three separate cultures, and the P-values from paired t-tests are as follows: shRNA-c (22, P<0.001), shRNA (18, P=0.769), shRNA+WT* (18, P=0.0009) and shRNA+Y533E* (18, P=0.468). Knockdown of DHHC5 decreased basal SEP-GluA1 IntDen (P=0.0016, F3,72=5.633, one-way analysis of variance (ANOVA)). Asterisks denote significance between groups before cLTP and relative to shRNA-c, and cross-hatches denote significance within groups before and after cLTP. Graph displays mean±s.e.m. **P<0.01, one-way ANOVA, Tukey's post hoc test; ###P<0.001, paired t-test.
Figure 10
Figure 10. Model of activity-regulated trafficking of DHHC5.
(1 and inset) Under basal conditions, DHHC5 is localized to the postsynaptic membrane in complex with PSD-95 and Fyn. PSD-95 binds DHHC5 through PDZ-dependent mechanisms. Fyn binds to the C-terminal poly-proline motif (PTGP) of DHHC5 through its SH3 domain and phosphorylates DHHC5 Y533 within the DHHC5 endocytic motif (YDNL). This inhibits the binding of the endocytic adaptor protein, AP2, to DHHC5. (2) Increased neuronal activity enhances STEP61-mediated dephosphorylation of Fyn and reduces its kinase activity (not shown), as well as decreases DHHC5 association with PSD-95 and Fyn, thereby enhancing DHHC5 internalization and trafficking to dendritic shafts on REs (2–3 min post stimulation). (3) DHHC5 associates with δ-catenin and palmitoylates it (3–10 min post stimulation). (4) DHHC5 and palmitoylated δ-catenin traffic back into spines (3–20 min post stimulation). (5) DHHC5 dissociates from δ-catenin and is reinserted into the synaptic membrane (10–20 min post stimulation). Significantly more DHHC5 is recruited to the synaptic membrane 20 min post stimulation compared with basal levels. δ-Catenin binds and stabilizes N-cadherin, leading to enhancements in synapse structure and increased membrane stabilization of AMPARs 20–60 min post stimulation. The association of δ-catenin with PSD-95 and Fyn, and δ-catenin tyrosine phosphorylation are also increased 20 min post stimulation (not shown), possibly linking cadherin–adhesion complexes with receptor-scaffold assemblies to coordinate synapse strengthening.

Similar articles

See all similar articles

Cited by 26 articles

See all "Cited by" articles

References

    1. Park M. et al. . Plasticity-induced growth of dendritic spines by exocytic trafficking from recycling endosomes. Neuron 52, 817–830 (2006). - PMC - PubMed
    1. Lu W. et al. . Activation of synaptic NMDA receptors induces membrane insertion of new AMPA receptors and LTP in cultured hippocampal neurons. Neuron 29, 243–254 (2001). - PubMed
    1. Bosch M. et al. . Structural and molecular remodeling of dendritic spine substructures during long-term potentiation. Neuron 82, 444–459 (2014). - PMC - PubMed
    1. Fu M., Yu X., Lu J. & Zuo Y. Repetitive motor learning induces coordinated formation of clustered dendritic spines in vivo. Nature 483, 92–95 (2012). - PMC - PubMed
    1. Lai C. S., Franke T. F. & Gan W. B. Opposite effects of fear conditioning and extinction on dendritic spine remodelling. Nature 483, 87–91 (2012). - PubMed

Publication types

MeSH terms

Feedback