N-ethylmaleimide-sensitive factor attachment protein receptors (SNARE) family members drive membrane fusion by the formation of a trans-SNARE complex consisting of specific v- and t-SNAREs present at vesicle (v) and target (t) membranes. Most SNAREs have autonomously folding N-terminal domains, along with SNARE (coiled-coil) motifs and membrane anchors. The SNARE motifs are 60–70 amino acid residues long [1
] and contain repeated heptad patterns of hydrophobic residues. They assemble into parallel four-helix bundles stabilized by a hydrophobic helix that faces the bundle’s core. Within the hydrophobic core of the bundle X-ray crystallography [2
] revealed an unusual central hydrophilic layer composed of three glutamines (Q) and one arginine (R) residue, which led to the classification of Q- and R-SNAREs, respectively [4
Ykt6 is an unusual SNARE as it lacks a transmembrane domain and therefore can cycle between cytosol and membranes. Membrane localization depends on the intramolecular interaction of the N-terminal Longin and C-terminal SNARE domains and the presence of a farnesylation and reversible palmitoylation within a CCAIM/CAAX motif at the C-terminus [5
]. This interaction is exemplified in yeast, where the release of Ykt6 from endosomal membranes into the cytoplasm depends on a functional Longin domain and an intramolecular interaction with its SNARE domain to fold into a soluble, closed conformation [6
]. Additionally, a recent study identified a new geranylgeranyl transferase that plays an essential role for membrane-anchored Ykt6 in proper Golgi function [7
Albeit sometimes redundant, specific sets of SNAREs mediate distinct steps in intracellular protein trafficking [8
]. Correspondingly, Ykt6 interacts with different SNARE partners in vitro [9
] and is proposed to function as a membrane stress sensor within the secretory pathway of yeast [10
]. In addition to its functions in the homotypic fusion of Endoplasmic Reticulum (ER) and vacuolar membranes [11
], ER-Golgi trafficking [13
] and retrograde Golgi trafficking in yeast [14
], it was also described to function in in autophagy [16
] and lysosomal stress [17
]. Recent studies revealed that Ykt6 mediates several steps of autophagosome formation in human cells [18
], the Drosophila
fat body [19
], and yeast [15
]. Nevertheless, how Ykt6 is recruited to these different membranes remains unclear. Here, we investigate how the functional regulation of the SNARE domain can mechanistically regulate Ykt6 membrane recruitment and activity in mammalian cells and in Drosophila
2. Material and Methods
Ykt6 coding region was amplified and the PCR product was recombined into the pDONRTM
221 vector using the Gateway BP Clonase II Enzyme mix (Life Technologies, Carlsbad, CA, USA). Point mutations of potential phosphorylation sites (S175, S182, T188, T192) were introduced by site-directed mutagenesis. For the generation of transgenic flies, constructs were subcloned into expression vectors pUASt-attB-rfA-mCherry and pUASt-attB-mCherry-rfA (kind gift from Sven Bogdan) by LR recombination (Life Technologies, Carlsbad, CA, USA). Human Ykt6 was amplified from hYkt6-Myc (C-Terminal myc-destination plasmids (DKFZ—Genomics and Proteomics Core Facility)) and the PCR product was inserted into pcDNA3.1MycBioID (Addgene #35700). Point mutations for Ykt6-3A (S174A, T181A, S187A), Ykt6-3E (S174E, T187E, S181E), F42A, C194A, C195A, and relevant combinations were introduced by site-directed mutagenesis. MycBioID tag was removed via Nhe1/Xho1 to obtain untagged constructs in pcDNA3.1. RUSH-EGFP-Wnt3A was constructed by amplifying the core protein sequence of Wnt3A and integrating it by Gibson cloning [21
] with the Wnt3A signal peptide and the streptavidin-binding peptide sequence into the ER-hook containing the pCMV-KDEL-IVS-IRES-reporter plasmid backbone [22
]. The following expression constructs were used: pCMV-Wnt3A [23
], DsRed-Rab5-QL (E. De Robertis, Addgene #29688) (Table 1
Antibodies were used against Calnexin, 1:1000, (WB, rabbit), Dallas, Texas, US; CD81 1:1000 (188.8.131.52, WB, mouse (DLN-09707), Dianova, Hamburg, Germany; EEA1, 1:300 (IF, mouse (610456), BD, New Jersey, NJ, USA; GAPDH (6C5), 1:5000 [WB; mouse (AM4300)], Ambion, Austin, TX, USA; Hsc70 1:2000 (WB; mouse (sc-7298), Santa Cruz, CA, CA, USA; TSG101, 1:1000, (WB, rabbit (HPA006161), Sigma, MO, USA; Wnt3A, 1:500 (WB, rabbit), Abcam, Cambridge, UK, and Ykt6 WB and IF; mouse (sc-365732), Santa Cruz, CA, USA. Antibodies against Ykt6 were generated by immunizing two guinea pigs with the peptides KVSADQWPNGTEATI (aa 105–119, within Longin domain) and YQNPVEADPLTKMQN (aa 131–145, covers part of the SNARE domain). Final bleeds were pooled and affinity purified against the original peptides (Eurogentec). Secondary antibodies directed against the species of interest were coupled to Alexa Fluor 488, 568, 594 and 647 IF, 1:500, Invitrogen, Carlsbad, CA, USA and 680RD and 800CW WB, 1:20,000, LiCor, Lincoln, NE, USA.
2.2. Drosophila Stocks and Genetics
The following Drosophila stocks were used in this study: en-GAL4, UAS-GFP
(chr. II, a gift from J. Grosshans). The following stocks were obtained from the Bloomington Drosophila stock center: UAS-Dcr; enGAL4, UAS-GFP
(#7108), and vas-PhiC31; attP.ZH-86Fb
(#24749). The Ykt6 (KK105648) UAS-RNAi stock was obtained from the Vienna Drosophila RNAi Center. UAS-Ykt6 transgenic lines were generated according to standard protocols by φC31 integrase-mediated site-specific insertion in the attP landing site at ZH-86Fb [24
]. Fly stocks were kept on standard medium containing agar, yeast, and corn flour. Crosses were performed at 25 °C or RT.
2.3. Kinase Screen
Biotinylated Ykt6-WT peptide (GEKLDDLVSKSEVLGTQSKAFYKTARKQN) was tested in one concentration against 245 Ser/Thr kinases in a radiometric, FlashPlate PlusTM-based assay. Ten out of 18 kinase hits identified among the 245 kinases were subsequently confirmed in a hit confirmation experiment. For this, three peptide concentrations in triplicate for each of the 10 tested kinase hits were used for N-terminally biotinylated peptides of Ykt6-WT and Ykt6-3E (GEKLDDLVSKE
ARKQN). The Kinase screen was performed by Reaction Biology, available online: https://www.proqinase.com/products-service-biochemical-assay-services/kinasefinder
, received on 15th October 2018).
2.4. Cell Culture and Transfection
Hek293T and HCT116 cells were maintained in DMEM (Gibco) supplemented with 10% fetal calf serum (Biochrom) at 37 °C in a humidified atmosphere with 5% CO2. Cells were transiently transfected with Screenfect siRNA for siRNA and Screenfect A (Screenfect) for plasmids according to the manufacturer’s instructions. Cells were identified and checked regularly for mycoplasma contamination.
2.5. Blue Sepharose Precipitation
The relative amount of Wnts secreted into cell culture supernatant was analyzed using Blue Sepharose precipitation as described [23
]. Shortly, HEK293T cells were transiently transfected in 6-well plates with 1 µg of Wnt3A plasmids. Then, 72 h after transfection, the supernatant was collected and centrifuged at 4000× g
rpm to remove cell debris, transferred to a fresh tube, and rotated at 4 °C for 1 h with 1% Triton X-100 and 40 µL of Blue Sepharose beads. The samples were washed and eluted from the beads using 2X SDS buffer with β-mercaptoethanol and analyzed by immunoblotting.
2.6. Extracellular Vesicle purification
Extracellular vesicles were purified by differential centrifugation as described previously [26
]. In short, supernatants from mammalian cells were subjected to sequential centrifugation steps of 750× g
, 1.5 × 103
g and 1.4 × 104
g, before pelleting exosomes at 1× 105
g in a SW41Ti swinging bucket rotor for 2 h (Beckman). The supernatant was discarded, and exosomes were taken up in 1/100 of their original volume in H2
2.7. Immunostainings, Microscopy, and Image Analysis
For immunofluorescence staining, cells were reverse transfected with siRNAs, seeded in 6 well dishes or 8-well microscopic coverslips, 24 h later transfected with indicated plasmids, and 48–72 h later fixed with 4% paraformaldehyde. Cells were permeabilized with 0.1% Triton X-100 and blocked in 10% BSA/PBS. Primary antibodies in PBS were incubated for 1 h at room temperature and antibody binding was visualized by fluorochrome-conjugated secondary antibodies. Confocal images were processed with Zen lite (Zeiss, Oberkochen, Germany), Fiji/ImageJ (NIH, Rockville, Maryland, MA, USA) [28
] and Affinity Designer (Affinity Serif, San Francisco, CA, USA).
2.8. Rab5QL Assay and Quantification
Hek293T cells were co-transfected with plasmids for Rab5Q79L-DsRed and either control, Ykt6-WT or Ykt6-3E and analyzed by immunofluorescence microscopy, and the size of enlarged Rab5Q79L-positive endosomes was measured in different biological replicates.
2.9. Membrane Fractionation
As previously described [31
], HEK293T cells were seeded and transfected with Ykt6-WT plasmid. Then, 48 h post transfection, cells were lysed on ice with 1 mL of Lysis buffer A (150 mM NaCl, 50 mM HEPES, 0.1% Saponin, 1 M Glycerol, and 1% PIC) and then centrifuged at 2000× g
for 10 min at 4 °C; then, the supernatant (cytosolic fraction) was transferred to a new tube. The pellet was lysed in 1 mL of Lysis Buffer B (150 mM NaCl, 50 mM Hepes, 1% Igepal, 1 M Glycerol and 1% PIC) and incubated rotating for 30 min at 4 °C. Then, after being centrifuged at 7000× g
for 10 min at 4 °C, the supernatant was transferred to a new tube (membrane fraction).
2.10. BioID Pull Down and Mass Spectrometry
For large-scale BioID pull down, cells were seeded and 24 h later transfected with BioID-WT or mock constructs. Then, 36 h post transfection, 50 μM biotin was added over night. Cells were washed with PBS twice, cell fractionated, and then boiled 5 min in non-reducing SDS sample buffer (300 mM Tris-HCl pH 6.8, 12% SDS, 0.05% Bromphenolblue, 60% Glycerol, 12 mM EDTA), run a short-distance (1.5 cm) on a 4–12% NuPAGE Novex Bis-Tris Minigel (Invitrogen). Gels were stained with Coomassie Blue for visualization purposes. Full lanes were sliced into 23 equidistant slices regardless of staining, short runs cut out as a whole and diced. After washing, gel slices were reduced with dithiothreitol (DTT), alkylated with 2-iodoacetamide, and digested with trypsin overnight. Then, the resulting peptide mixtures were extracted, dried in a SpeedVac, reconstituted in 2% acetonitrile/0.1% formic acid/(v
), and prepared for nanoLC-MS/MS as described previously [32
For the generation of a peptide library for SWATH-MS, equal amount aliquots from each sample were pooled to a total amount of 80 μg and separated into eight fractions using a reversed phase spin column (Pierce High pH Reversed-Phase Peptide Fractionation Kit, Thermo Fisher Scientific, Waltham, Massachusetts, United States. MS analysis Protein digests were separated by nanoflow chromatography. Then, 25% of gel slices or 1 μg aliquots of digested protein were enriched on a self-packed precolumn (0.15 mm ID × 20 mm, Reprosil-Pur120 C18-AQ 5 μm, Dr. Maisch, Ammerbuch-Entringen, Germany) and separated on an analytical RP-C18 column (0.075 mm ID × 250 mm, Reprosil-Pur 120 C18-AQ, 3 μm, Dr. Maisch) using a 30 to 90 min linear gradient of 5–35% acetonitrile/0.1% formic acid (v:v) at 300 nl/ min.
SWATH-MS library generation was performed on a hybrid triple quadrupole-TOF mass spectrometer (TripleTOF 5600+) equipped with a Nanospray III ion source (Ionspray Voltage 2400 V, Interface Heater Temperature 150 °C, Sheath Gas Setting 12) and controlled by Analyst TF 1.7.1 software (SCIEX, Framingham, Massachusetts, MA, USA)build 1163 (all AB Sciex), using a Top30 data-dependent acquisition method with an MS survey scan of m/z 380–1250 accumulated for 250 ms at a resolution of 3.5 × 104 full width at half maximum (FWHM). MS/MS scans of m/z 180–1500 were accumulated for 100 ms at a resolution of 17,500 FWHM and a precursor isolation width of 0.7 FWHM, resulting in a total cycle time of 3.4 s. Precursors above a threshold MS intensity of 200 cps with charge states 2+, 3+, and 4+ were selected for MS/MS, and the dynamic exclusion time was set to 15 s. MS/MS activation was achieved by CID using nitrogen as a collision gas and the manufacturer’s default rolling collision energy settings. Two technical replicates per reversed phase fraction were analyzed to construct a spectral library.
For quantitative SWATH analysis, MS/MS data were acquired using 100 variable size windows [33
] across the 400–1200 m
range. Fragments were produced using rolling collision energy settings for charge state 2+, and fragments acquired over an m
range of 180–1500 for 40 ms per segment. Including a 250 ms survey scan, this resulted in an overall cycle time of 4.3 s. Two replicate injections were acquired for each biological sample.
2.11. Mass Spectrometry Data Processing
For SWATH-MS analysis, protein identification was achieved using ProteinPilot Software version 5.0 (SCIEX, Framingham, Massachusetts, MA, USA) build 4769 (AB Sciex) at “thorough” settings. MS/MS spectra from the combined qualitative analyses were searched against the UniProtKB Homo sapiens reference proteome (revision February 2017. 92,928 entries) augmented with a set of 51 known common laboratory contaminants to identify 597 proteins at a False Discovery Rate (FDR) of 1%. Spectral library generation and SWATH peak extraction were achieved in PeakView Software version 2.1 (SCIEX, Framingham, Massachusetts, MA, USA) build 11041 (AB Sciex) using the SWATH quantitation microApp version 2.0 SCIEX, Framingham, Massachusetts, MA, USA) build 2003. Following retention time correction on endogenous peptides spanning the entire retention time range, peak areas were extracted using information from the MS/MS library at an FDR of 1% [34
]. The 26 resulting peak areas were summed to peptide and protein area values, which were used for further statistical analysis. Reactome Functional Network analysis [35
] was performed with Cytoscape [www.cytoscape.org
(Accessed on 15th June 2020)] and Kegg pathway analysis was performed with David [36
All experiments were carried out at least in biological triplicates. Error bars indicate s.d. Statistical significance was calculated by carrying out Student’s t-test where appropriate or one-way ANOVA with Dunnett’s multiple comparison test to compare a control mean with the other means.
The fusion of eukaryotic transport vesicles with target organelles requires membrane-bridging complexes of membrane anchored SNAREs. Ykt6 is lacking a transmembrane domain, and thus, its site of action is determined by changing from a soluble to a membrane-bound conformation. How Ykt6 is recruited to membranes remains unclear. Here, we investigate how the functional regulation of its SNARE domain can mechanistically regulate Ykt6 membrane attachment and its activity in mammalian cells and in Drosophila
. We found that Ykt6 membrane attachment is regulated by modifications in its SNARE domain, and that this regulation affects cell survival in vivo. Ykt6 phosphorylation and attachment further affect the Ykt6 proximity profile and the cellular processes in which it takes part. Specific secretory processes, such as Wnt and EV secretion, furthermore depend on a functional Ykt6 SNARE domain and could thus be regulated by Ykt6 phosphorylation. Under normal growth conditions, these processes seem to be more important for cellular function, than its role in autophagosome formation. As it has also been reported that Ykt6 is involved in autophagosome–lysosome fusion in human cells and Drosophila
fat body under starvation-induced conditions [15
], it is possible that Ykt6 is allocated to alternate compartments, especially under nutrient stress condition [51
Non-neuronal SNAREs possess evolutionarily conserved phosphorylation sites [37
], which seem to prevent membrane fusion when SNAREs are phosphorylated. Based on our data, Ykt6 functions in a similar manner, as Ykt6-3E seems to prevent fusion as well. Yet, dynamic membrane recruitment is specific to SNAREs lacking a transmembrane domain such as Ykt6. Indeed, self-labeling and membrane attachment of different Ykt6 mutants varies and in agreement with previous work [53
] hints toward a two-step process of activation that is specific to Ykt6 and involves (1) a change from closed to open confirmation and (2) membrane recruitment. Post-membrane recruitment, true for all SNAREs with phosphorylation sites in the SNARE domain [37
], phosphorylation would prevent the fusion of different SNAREs to mediate membrane fusion.
Protein acylation is a regulatory post-translational modification regulating membrane association and the dissociation of its target proteins. The F42 position has been previously shown to participate in intramolecular interactions between the Ykt6 Longin domain and the SNARE motif. The hydrophobic face at position F42 within the Longin domain seems to accommodate the farnesyl anchor, as seen in a crystal structure with dodecylphosphocholine (DPC) [7
]. This intramolecular binding explains the membrane association of Ykt6 when F42 is mutated (Figure 3
A, right panel, lane13). Similarly, the phosphorylation sites we identified (S174/S181/T187) are directly facing this interface and therefore likely impact the tight folding of Ykt6 into a closed conformation and thus membrane attachment (Figure 3
E). Interestingly, the non-phosphorylatable mutant Ykt6-3A does not attach to membranes by itself, yet it is able to rescue the growth defects in vivo (Figure 1
G). Taken together, this suggests that the phosphorylation of Ykt6 prevents the conformational switch to the inactive state, which leads to membrane stabilization. Likely, an additional step is required to render Ykt6 fusion-competent, as Ykt6-3E did not rescue Ykt6 KD in vivo,
but it had no dominant-negative effects in the presence of endogenous Ykt6. This last step toward fusion-competent Ykt6 could be achieved by dephosphorylation [52
], for example by calcineurin, as suggested by a recent study [54
Which signaling pathways could stimulate the phosphorylation of Ykt6 and therefore its membrane recruitment? As PKC gets activated in the context of endocytosis to recruit adaptor complexes to endosomes [55
] local PKC-dependent conformation changes and subsequent palmitoylation could stimulate Ykt6 association with endosomes. Similarly, PDK1 is a growth factor-dependent kinase that could phosphorylate Ykt6 in the presence of mitogenic signals that require Wnt secretion and/or autophagy. Moreover, the Netphos prediction of Ykt6 S174 to be a CDK1 target and the fact that proteins proximal of Ykt6 are involved in cell cycle regulation opens interesting questions of whether Ykt6 is involved in cell growth-dependent regulations of organelle fusion. Furthermore, the identification of three phosphorylation sites bears the possibility that the three sites are individually targeted by different signaling pathways and differ in their role in recruiting and stabilizing Ykt6 to and at membranes under permissive circumstances. Yet, the physiological relevance of Ykt6 phosphorylation remains to be demonstrated, and more detailed analysis is necessary to understand the interconnection and hierarchy of different signaling pathways targeting Ykt6.
Interestingly, other SNAREs such as Syntaxin-5 are modified by monoubiquitination that regulates Golgi integrity during cell cycle progression [57
]. In addition to position S174 in Ykt6 being identified in Phosphoproteomic studies, its Lysine 182 and 186 (K182,K186) were found to be ubiquitinated in different human cell lines and the mouse liver [58
]. These positions lie close by the analyzed phosphorylation sites (S181, T187). Thus, Ykt6 function and its membrane recruitment could be the target of different posttranscriptional modifications at the crossroad of cellular trafficking events in cellular growth.