Lipid-mediated activation of plasma membrane-localized deubiquitylating enzymes modulate endosomal trafficking

Abstract

The abundance of plasma membrane-resident receptors and transporters has to be tightly regulated by ubiquitin-mediated endosomal degradation for the proper coordination of environmental stimuli and intracellular signaling. Arabidopsis OVARIAN TUMOR PROTEASE (OTU) 11 and OTU12 are plasma membrane-localized deubiquitylating enzymes (DUBs) that bind to phospholipids through a polybasic motif in the OTU domain. Here we show that the DUB activity of OTU11 and OTU12 towards K63-linked ubiquitin is stimulated by binding to lipid membranes containing anionic lipids. In addition, we show that the DUB activity of OTU11 against K6- and K11-linkages is also stimulated by anionic lipids, and that OTU11 and OTU12 can modulate the endosomal degradation of a model cargo and the auxin efflux transporter PIN2-GFP in vivo. Our results suggest that the catalytic activity of OTU11 and OTU12 is tightly connected to their ability to bind membranes and that OTU11 and OTU12 are involved in the fine-tuning of plasma membrane proteins in Arabidopsis.

Fig. 1. GFP-OTU11 and GFP-OTU12 localize to the PM in Arabidopsis root cells.

a Representative confocal images of 35S promoter-driven YFP-fusions of UBP3, UBP18, UBP22, UBP23, UBP25, UBP27, OTU9, OTU10, OTU11, and OTU12. YFP-fusion proteins were transiently expressed in Arabidopsis root cell-derived protoplasts. For each construct, protoplast transformation was performed at least twice. Scale bars: 10 µm. b Representative confocal images of protoplasts expressing 35pro:GFP-OTU11 or 35pro:GFP-OTU12 together with 35Spro:RFP-SYP121. The signal intensity profile along the dotted line (merged image) is shown on the right. Scale bars: 10 µm. Protoplast transformations were performed at least twice. c Representative confocal images of wild-type (Col-0) Arabidopsis seedlings harboring OTU11pro:GFP-OTU11 or OTU12pro:OTU12-GFP. Both lines were co-stained with FM4-64 for 5 min before imaging. The confocal analysis was performed at least twice. The signal intensity profile along the dotted line (merged image) is shown on the right of the panels. Scale bars: OTU11 10 µm, OTU12 20 µm. d GFP-OTU11 is present in the enriched PM fraction. Total proteins were prepared from wild-type (Col-0) seedlings containing 35Spro:GFP-OTU11. The extracts were centrifuged at 8000 ×g, and the supernatant (S8) was further centrifuged at 100,000 ×g to yield supernatant (S100) and microsome (P100) fractions. P100 samples were subsequently treated with Brij-58 for the enriched PM fraction (ePM). H⁺-ATPase, UGPase, and Sec21 were used as markers for the PM, soluble fraction, and microsomal fractions, respectively. Three independent experiments were carried out, and a representative result is shown. Source data are provided as a Source Data file. e, f 35Spro:GFP-OTU11- and 35Spro:GFP-OTU12-expressing wild-type (Col-0) seedlings were stained with FM4-64 and treated with 50 µM brefeldin A (e) or 33 µM Wortmannin (f). The experiments were repeated at least three times; one representative image is shown. Scale bars: 10 µm. g 35Spro:GFP-OTU11-, 35Spro:GFP-OTU12-expressing wild-type (Col-0) seedlings, and the PI(4)P-sensor (P5R)-containing seedlings were treated with 60 µM phenylarsine oxide (PAO). Whereas PAO treatment causes dissociation of P5R from the PM, localization of GFP-OTU11 and GFP-OTU12 remains unaffected. The experiment was repeated at least three times; one representative image is shown. Scale bars: 10 µm.

Fig. 2. Polybasic motifs in OTU11 and OTU12 are required for the PM localization.

a, b OTU11(a) and OTU12(b) constructs used for this study. c, d BH-scores (window = 10) of OTU11(WT) (thin black line), OTU11(6A1) (dotted red line) and OTU11(6A2) (dotted blue line) (c) and OTU12(WT) (thin black line), OTU12(6A1) (dotted red line), and OTU12(6A2) (dotted blue line) (d). The thresholds of 0.6 are indicated with dotted lines. Wild-type OTU11 and OTU12 have two large peaks corresponding to PBM1 and PBM2. The 6A1 mutation leads to the loss of the PBM1-peak, whereas the 6A2 mutation causes the loss of the PBM2 peak. e, f PBM1 and PBM2 of OTU11 and OTU12 are important for the PM localization in cellula. Arabidopsis root cell culture-derived protoplasts were transformed with 35Spro:RFP-OTU11(WT) (n = 22 cells), 35Spro:Y/RFP-OTU11(6A1) (n = 32 cells), or 35Spro:YFP-OTU11(6A2) (n = 20 cells) constructs (e) or with 35Spro:YFP-OTU12(WT) (n = 77 cells), 35Spro:YFP-OTU12 (6A1) (n = 56 cells), or 35Spro:YFP-OTU12(6A2) (n = 15 cells) constructs (f). The distribution of fluorescent signals in the cells was analyzed by confocal microscopy and categorized in +PM (signals at PM) and −PM (signals not at PM). At least two independent transformations were performed for each set of constructs, and the percentage of cells with PM signals were calculated. g, h PBM1 is required for the PM localization of OTU11 and OTU12 in planta. Seven-day-old Arabidopsis seedlings harboring 35Spro:GFP-OTU11(WT), 35Spro:GFP-OTU11(6A1) (g), 35Spro:GFP-OTU12(WT) and 35Spro:GFP-OTU12(6A1) (h) were analyzed under the confocal microscope. Representative images are shown. Scale bars: 10 µm. i Boxplots of the quantification of the confocal microscopy analysis in (g) and (h). The number of cells with PM localization as well as the total number of cells (n) were counted and the percentage of cells with PM localization was calculated in 10 seedlings for each genotype [35Spro:GFP-OTU11(WT) (n = 163 cells, 61% with PM localization), 35Spro:GFP-OTU11(6A1) (n = 179 cells, 0%), 35Spro:GFP-OTU12(WT) (n = 169 cells, 83%), and 35Spro:GFP-OTU12 (n = 126 cells, 0%)]. Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers. The seedlings expressing wild-type OTU11 and OTU12 had significantly more cells with PM localization [two-tailed t test, no equal variance, OTU11: P = 0.00046 (***P < 0.001), OTU12 P = 2.75 × 10⁻⁵ (***P < 0.001)]. Source data are provided as a Source Data file.

Fig. 3. Phenotypes of otu11otu12 knock-out and OTU11- and OTU12 overexpressing seedlings.

a Photographs of 7-day-old wild-type, otu11, otu12, and otu11otu12 seedlings. Scale bars: 1 cm. b Boxplots of primary root length of 7-day-old wild-type (n = 158 seedlings), otu11 (n = 159 seedlings), otu12 (n = 160 seedlings), and otu11otu12 (n = 160 seedlings). Center line, median; box limits, first and third quartiles; whiskers, 1.5x interquartile range; points, outliers. Wild-type/otu11 P = 0.0723 (ns: not significant), wild-type/otu12 P = 0.00417 (**0.001 < P < 0.01), wild-type/otu11otu12 P = 3.28 × 10⁻⁷ (***P < 0.001), two-tailed t test, no equal variance. The experiment was conducted three times, and one representative result is shown. c Photographs of 7-day-old wild-type, otu11otu12, and otu11otu12 containing OTU12pro:OTU12-GFP. Scale bars: 1 cm. d Box plot of primary root length of 7-day-old wild-type (n = 280 seedlings), otu11otu12 (n = 256 seedlings), and otu11otu12 with OTU12pro:GFP-OTU12 (n = 256 seedlings). Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers. The experiment was conducted three times and one representative result is shown. Wild-type/otu11otu12 P = 1.18 × 10⁻⁸ (***P < 0.001), wild-type/otu11otu12 OTU12pro:GFP-OTU12 P = 0.258 (ns: not significant), two-tailed t test, no equal variance. e Photographs of 7-day-old wild-type and 35Spro:GFP-OTU11 (GFP-OTU11 overexpressor o/e) and 35Spro:GFP-OTU12 (GFP-OTU12 o/e #2) seedlings. Scale bars: 1 cm. f The primary root length of 7-day-old wild-type (n = 196 seedlings), GFP-OTU11 o/e (n = 171 seedlings), GFP-OTU12 o/e #1 (n = 197 seedlings), and GFP-OTU12 o/e #2 (n = 213 seedlings) is shown as a box plot. Center line, median; box limits, first and third quartiles; whiskers, 1.5x interquartile range; points, outliers. Wild-type/GFP-OTU11 o/e P = 8.08 × 10⁻⁴⁹ (***P < 0.001), wild-type/GFP-OTU12 o/e #1 P = 3.07 × 10⁻¹⁷ (***P < 0.001), wild-type/GFP-OTU12 o/e #2 P = 5.46 × 10⁻¹⁶ (***P < 0.001), two-tailed t test, no equal variance. The experiments were conducted three times and one representative result is shown. g The primary root length of 7-day-old wild-type (n = 160 seedlings), GFP-OTU11(WT)- (n = 116 seedlings) and GFP-OTU11(6A1) overexpressing seedlings (n = 169 seedlings) is shown as a box plot. Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers. Wild-type/OTU11(WT) o/e P = 8.31 × 10⁻³³ (***P < 0.001), wild-type/OTU11(6A1) o/e P = 1.71 × 10⁻¹¹ (***P < 0.001), GFP-OTU11(WT) o/e /GFP-OTU11(6A1) o/e P = 4.56 × 10⁻¹² (a: P < 0.001), two-tailed t test, no equal variance. The experiments were repeated twice and one representative result is shown. Source data are provided as a Source Data file.

Fig. 4. OTU11 and OTU12 modulate the endosomal transport of PMA-GFP-UB.

a Representative confocal images of protoplasts transformed with PMA-GFP-UB. The experiment was repeated at least three times. Scale bars: 10 µm. b Schematic representation of the CRISPR target- (CRISPR^OTU12) and mutated target- (CRISPR^OTU12m) sequences. c Protein extracts of protoplasts transformed with 35Spro:RFP-OTU12 alone, and with 35Spro:RFP-OTU12 and CRISPR^OTU12/3xFLAG-Cas9 or CRISPR^OTU12m/3xFLAG-Cas9 were subjected to anti-FLAG and anti-RFP immunoblots. An anti-ACTIN antibody was used as loading control on the anti-FLAG-treated membrane. d Protoplasts expressing PMA-GFP-UB alone (n = 225 cells), PMA-GFP-UB with CRISPR^OTU12−3xFLAG-Cas9 (n = 148 cells), and PMA-GFP-UB with CRISPR^OTU12m−3xFLAG-Cas9 (n = 125 cells) were analyzed and cells were categorized according to the localization of PMA-GFP-UB as in (a). Three independent transformations were performed and the results of all experiments are shown as a box plot. Center line, median; box limits, first and third quartiles; whiskers, 1.5x interquartile range; points, outliers. P values for the PM+ endosome/vacuole (PM+ end/vac) localization, without CRISPR/CRISPR^OTU12 P = 0.0453 (*0.01 < P < 0.05), without CRISPR/CRISPR^OTU12m P = 0.350 (ns: not significant); endosome (end), without CRISPR/CRISPR^OTU12 P = 0.112 (ns), without CRISPR/CRISPR^OTU12m P = 0.410 (ns); endosome and vacuole (end+vac), without CRISPR/CRISPR^OTU12 P = 0.356 (ns), without CRISPR/CRISPR^OTU12m P = 0.666 (ns); vacuole (vac), without CRISPR /CRISPR^OTU12 P = 0.0379 (*0.01 < P < 0.05), without CRISPR/CRISPR^OTU12m P = 0.420 (ns), two-tailed t-tests with no equal variance. e The expression of 3xFLAG-CAS9 in (d) was verified with an anti-FLAG immunoblot. An anti-ACTIN antibody was used as processing control on a separate gel. f Protoplasts were transformed with PMA-GFP-UB alone (n = 115 cells), PMA-GFP-UB with 35Spro:RFP-OTU12(WT) (n = 78 cells) or with 35Spro:RFP-OTU12(6A1) (n = 101 cells). Cells with both RFP and GFP signals were analyzed as in (d). P values PM+ end/vac, without OTU12 o/e/OTU12(WT) o/e P = 9.45 × 10⁻⁴ (***P < 0.001), without OTU12 o/e/OTU12(6A1) o/e P = 0.206 (ns); end, without OTU12 o/e/OTU12(WT) o/e P = 0.918 (ns), without OTU12 o/e/OTU12(6A1) o/e P = 0.238 (ns); end+vac, without OTU12 o/e/OTU12(WT) o/e P = 0.00688 (**0.001 < P < 0.01), without OTU12 o/e/OTU12(6A1) o/e P = 0.0138 (*0.01 < P < 0.05); vac, without OTU12 o/e/OTU12(WT) o/e P = 0.108 (ns), without OTU12 o/e/OTU12(6A1) o/e P = 0.671 (ns), two-tailed t tests with no equal variance. g The expression of RFP-OTU12 variants in (f) was verified by an anti-RFP immunoblot and anti-ACTIN antibody as loading control on the same membrane. Source data are provided as a Source Data file.

Fig. 5. OTU11 and OTU12 affect the endosomal transport of PIN2-GFP.

a–c Representative confocal images of epidermis cells of 7-day-old PIN2pro:PIN2-GFP expressing wild-type or otu11otu12 seedlings treated with 50 µM BFA (a), 33 µM Wortmannin (b), or in the dark (c) for the indicated time. Scale bars: 10 µm. d Quantification of the BFA treatment in (a). The experiment was conducted twice. For each seedling, cells with BFA bodies were counted and the percentage of cells with BFA bodies is shown as a box plot [wild-type 0 min (4 seedlings, 126 cells), wild-type 30 min (23 seedlings, 795 cells), wild-type 60 min (21 seedlings, 897 cells), otu11otu12 0 min (4 seedlings, 130 cells), otu11otu12 30 min (31 seedlings, 1312 cells), otu11otu12 60 min (22 seedlings, 877 cells)]. Center line, median; box limits, first and third quartiles; whiskers, 1.5x interquartile range; points, outliers. P values: wild-type 30 min/otu11otu12 30 min P = 4.95 × 10⁻⁶ (***P < 0.001), wild-type 60 min/otu11otu12 60 min P = 0.0377 (*0.01 < P < 0.05), two-tailed t-test with no equal variance. e Quantification of the Wortmannin treatment in (b). The experiment was conducted three times. For each seedling, cells with Wortmannin compartments were counted, and the percentage of cells with Wortmannin compartments is shown as a box plot as described in (d) [wild-type 0 min (8 seedlings, 322 cells), wild-type 45 min (23 seedlings, 855 cells), wild-type 90 min (21 seedlings, 731 cells), otu11otu12 0 min (5 seedlings, 364 cells), otu11otu12 45 min (25 seedlings, 1078 cells), otu11otu12 90 min (14 seedlings, 648 cells)], P values: wild-type 45 min/otu11otu12 45 min P = 2.97 × 10⁻⁸ (***P < 0.001), wild-type 90 min/otu11otu12 90 min P = 1.84 × 10⁻⁵ (***P < 0.001), two-tailed t test with no equal variance. f Quantification of the dark treatment in (c). The experiment was conducted three times. For each seedling, cells with vacuolar GFP signals were counted, and the percentage of cells with vacuolar signals is shown as a box plot. Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers. [wild-type (15 seedlings, 655 cells), otu11otu12 (14 seedlings, 631 cells)]. P value: wild-type/otu11otu12 P = 0.0121 (*0.01 < P < 0.05), two-tailed t test with no equal variance. Source data are provided as a Source Data file.

Fig. 6. PBMs in OTU11 and OTU12 are required for their interaction with anionic lipids in vitro.

a Lipid species spotted on the membrane for the lipid overlay assay. The solid dark gray and light gray spots indicate lipids that showed interactions in the lipid-overlay assays. b–d Lipid overlay assays with GST, GST-OTU11(WT), GST-OTU11(6A1), GST-OTU11(6A2), and GST-OTU11(6A1 + 6A2) (b), GST, GST-OTU12(WT), GST-OTU12(6A1), and GST-OTU12(6A2) (c) and GST, GST-OTU11(OTU), GST-OTU11(OTU-6A1), and GST-OTU11(N) (d). Bound proteins were detected with an anti-GST antibody. e Representative gel images of liposome sedimentation assays using GST-OTU11(WT) and GST-OTU12(WT). GST-fusion proteins were incubated with the liposome buffer alone or with liposomes (PC/PE) containing no PIPs or 5% of one of PI(3)P, PI(4)P, or PI(4,5)P₂. SDS-PAGE gels were stained with TCE. M molecular mass marker. f Quantification of the result in (e). The signal intensity of the protein band in the pellet fraction was divided by the signal intensity of the protein band in the supernatant fraction for each lane to calculate the intensity ratio of pellet/supernatant (P/S). Box plot shows the results of the quantification of at least three independent experiments. Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers, n = 3 experiments. g Representative gel images of liposome sedimentation assays of GST-OTU11(WT) and GST-OTU11(6A1). GST fusion proteins were incubated with the liposome buffer alone or with liposomes (PC/PE) containing 5% of PI(3)P, PI(4)P, or PI(4,5)P₂. M molecular mass marker. h Quantification of the result in (g). Box plot shows the results of the quantification of three independent experiments. Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers, n = 3 experiments. i A representative gel image of liposome sedimentation assays of GST-OTU11(WT) or GST-OTU11(N). GST-fusion proteins were incubated with liposomes (PC/PE) containing 5% of PI(4,5)P₂ or with the liposome buffer alone. M molecular mass marker. j Quantification of the result in (i). Box plot shows the results of the quantification of three independent experiments. Center line, median; box limits, first and third quartiles; whiskers, 1.5× interquartile range; points, outliers, n = 3 experiments. Source data are provided as a Source Data file.

Fig. 7. PBM1 is involved in lipid binding and could influence the catalytic site in the OTU domain.

a Modeling of OTU11 (left) and OTU12 (right) using alpha fold. PBM1: blue, residues of the active site: red, stick mode. The panels in (a), (b) and (c) were prepared using PyMol. b Structural alignment of OTU11 (light gray) and OTU12 (dark gray). PBM1: light blue (OTU11), blue (OTU12), active site: salmon (OTU11), red (OTU12), stick mode. c Modeling of the complex formed between OTU11 and ubiquitin. Color coding of OTU11 appears as in (a). Ubiquitin is colored in light orange. d, e Computational simulation of interactions between the OTU11 variants and the lipid bilayer. Representative snapshots of wild-type OTU11(Δ21) (d) and OTU11(OTU) (e) after 1000 ns are shown. The protein is represented as a new ribbon and colored according to the secondary structure. Side chains are represented as lines and colored according to the residue type (blue: basic, red: acidic, green: polar, white: hydrophobic). The atoms of the catalytic center and the PBM1 motif are highlighted as ball and sticks. PC and PE are depicted as lines and colored according to the atom type. PI(4,5)P₂ lipids are highlighted in licorice representation. Ions are shown as transparent spheres. f, g Propensities of explicit salt bridges between lysine and arginine residues and PI(4,5)P₂ during the simulations for wild-type and 6A1 variants of OTU11(∆21)(f) and OTU11(OTU) (g). For all variants, two setups with different initial distances of the PBM1 motif to the membrane were simulated (d: larger initial distance). The regions highlighted with a light gray frame indicate PBM1 and PBM2. The 6A1 mutation leads to a reduction in salt bridge contacts in PBM1. h Scatterplot and distributions of the distances between two pairs of not neighboring amino acids from the catalytic center, CYS112/HIS218 and ASP109/HIS218, for the OTU11(Δ21) simulations. i Snapshots from the final structures for OTU11(Δ21) simulations are compared with the endpoint of the bulk equilibration which is the starting point of the membrane simulations. Pink: WT, red: 6A1, gray: bulk. Highlighted are the three amino acids in the catalytic center (ASP109, CYS112, HIS218). Heteroatoms are colored according to atom type.

Fig. 8. The in vitro DUB activity of OTU11 and OTU12 is stimulated by anionic lipids.

a, b DUB assays with GST-OTU11 (a) and GST-OTU12 (b). 7.5 pmol of K63-linked tetra-UB was incubated with 7.5 pmol (substrate: enzyme ratio 1:1), 25 pmol (1:3), 50 pmol (1:6), 100 pmol (1:12), or 250 pmol (1:30) of GST-OTU11 or GST-OTU12 for 1 h at 21 °C. The experiments were repeated at least three times, and one representative image is shown. c DUB assay with phosphatase treated GFP-OTU11 purified from 35Spro:GFP-OTU11 expressing Arabidopsis seedlings and 15 pmol of K63-linked di-UB. active: active phosphatase, inact: heat-inactivated phosphatase. The experiment was repeated at least three times; one representative image is shown. d, e DUB assays with GST-OTU11 (WT) and GST-OTU11(6A1) pre-incubated with liposomes generated with PC and PE (PC/PE) alone (−PIP) or with liposomes (PC/PE) containing 5% PI(4,5)P₂ (+PIP). In (d), 25 pmol of GST-OTU11(WT) or GST-OTU11(6A1) was incubated for the indicated time with 25 pmol of K63-linked tetra-UB (negative controls: tetra-UB with +PIP or −PIP liposomes, incubated for 4 h). In (e), 25 pmol of GST-OTU11, pre-incubated with +PIP and -PIP liposomes, was mixed with 25 pmol of linear, K6-, K11-, K27-, K29-, K33-, or K48-linked tetra-UB for 4 h at 21 °C. The experiments were repeated at least three times; one representative image is shown. f–k Effect of liposomes on the DUB activity of OTU11 (f–h) and OTU12 (i–k). 3.75 pmol of Recombinant OTU11(WT), OTU11(6A1), OTU12(WT), and OTU12(6A1) were pre-incubated with either liposome with or without PI(4,5)P₂ [lipo(+PIP) and lipo(–PIP), respectively], or the liposome buffer alone before the addition of 3.75 pmol of di-UB FRET TAMRA K48 pos1 (f, i), K48 pos2 (g, j), and K63 pos1 (h, k). The assays were conducted at least three times. The result of one representative measurement is shown. Error bars: standard deviation of a technical quadruplicate, center of the error bars: mean of the quadruplicates. Source data are provided as a Source Data file.

Fig. 9. Model for OTU11 and OTU12 function.

In addition to DUBs such as AMSHs, UBP12, and UBP13, OTU11 and OTU12 could fine-tune the endocytic degradation pathway. OTU11 and OTU12 are found at the PM. When bound to the anionic lipid-containing PM, the DUB activity of OTU11 and OTU12 can be stimulated, and OTU11 and OTU12 can deubiquitylate ubiquitin-modified proteins at the PM. Deubiquitylation of PM proteins could affect their affinity to ubiquitin-binding domains of ESCRT and ESCRT-accessory proteins. Thus, for the UB-dependent and ESCRT-mediated endosomal degradation, the balance of the activity of ubiquitylating- and deubiquitylating enzymes can determine protein stability. OTU11 and OTU12 can be part of the molecular mechanisms modulating the endosomal transport and subsequent vacuolar degradation of PM proteins.