- Email: [email protected]
Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion
Biochemical and Biophysical Research Communications xxx (2017) 1e7
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Green fluorescence protein-based content-mixing assay of SNAREdriven membrane fusion Paul Heo, Byoungjae Kong, Young-Hun Jung, Joon-Bum Park, Jonghyeok Shin, Myungseo Park, Dae-Hyuk Kweon* Department of Genetic Engineering, College of Biotechnology and Bioengineering, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, South Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 27 April 2017 Accepted 1 May 2017 Available online xxx
Soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins mediate intracellular membrane fusion by forming a ternary SNARE complex. A minimalist approach utilizing proteoliposomes with reconstituted SNARE proteins yielded a wealth of information pinpointing the molecular mechanism of SNARE-mediated fusion and its regulation by accessory proteins. Two important attributes of a membrane fusion are lipid-mixing and the formation of an aqueous passage between apposing membranes. These two attributes are typically observed by using various fluorescent dyes. Currently available in vitro assay systems for observing fusion pore opening have several weaknesses such as cargo-bleeding, incomplete removal of unencapsulated dyes, and inadequate information regarding the size of the fusion pore, limiting measurements of the final stage of membrane fusion. In the present study, we used a biotinylated green fluorescence protein and streptavidin conjugated with € ster resonance energy transfer (FRET) donor and acceptor, respectively. This Dylight 594 (DyStrp) as a Fo FRET pair encapsulated in each v-vesicle containing synaptobrevin and t-vesicle containing a binary acceptor complex of syntaxin 1a and synaptosomal-associated protein 25 revealed the opening of a large fusion pore of more than 5 nm, without the unwanted signals from unencapsulated dyes or leakage. This system enabled determination of the stoichiometry of the merging vesicles because the FRET efficiency of the FRET pair depended on the molar ratio between dyes. Here, we report a robust and informative assay for SNARE-mediated fusion pore opening. © 2017 Elsevier Inc. All rights reserved.
Keywords: SNARE Membrane fusion Content-mixing assay Green fluorescence protein Fusion pore
1. Introduction Membrane fusion is an important cellular process by which two lipid bilayers merge and form a pore through which cargos diffuse. All intracellular membrane fusions are mediated by soluble-Nethylmaleimide-sensitive-factor attachment protein receptor (SNARE)-family members. The membrane fusion is catalyzed by assembly of 3 Q-SNAREs (glutamine residue at the zeroth layer) and 1 R-SNARE (arginine residue at the zeroth layer), with Q-SNAREs typically targeting the plasma membrane (t-SNARE) while RSNAREs are typically vesicle-associated (v-SNARE). SNARE complex formation, which involves the directional assembly of the SNARE motifs to form a parallel 4 helical bundle, is thought to provide the free energy required to promote lipid bilayer fusion [1,2]. An approach using reconstituted proteoliposomes with SNARE
* Corresponding author. E-mail address: [email protected] (D.-H. Kweon).
proteins was first used to examine the ‘SNAREpin’ as the minimal fusion machinery [3]. Since this discovery, the fluorescence €rster) resonance energy transfer (FRET)-based lipid mixing assay (Fo has been the gold standard in vitro tool for studying SNAREmediated membrane fusion [4e6]. In the most common scheme, t- and v-SNARE proteins are reconstituted into vesicles separately (the resulting vesicles are known as t- and v-vesicles, respectively), where one vesicle is fluorophore-free and the other contains both acceptor and donor fluorescence dyes. Upon fusion by SNARE complex formation, the average distance between the donors and acceptors in a vesicle increases following lipid mixing, dequenching donor fluorescence emission. The critical roles of SNARE proteins and regulatory proteins were revealed using this assay. However, the lipid mixing assay was not consistently directly correlated with fusion pore opening, which is the final step in the fusion pathway. A previous study showed that a high level of lipid mixing occurred with a limited degree of content mixing [7]. In viral fusion and vacuolar fusion, content mixing occurred seconds
http://dx.doi.org/10.1016/j.bbrc.2017.05.006 0006-291X/© 2017 Elsevier Inc. All rights reserved.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
2
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
to minutes later than initial lipid mixing [8,9]. Neuronal SNAREs also exhibited significant lipid mixing seconds before content mixing [10]. Thus, lipid mixing alone is not sufficient to clearly detect fusion pore opening, and thus an efficient content-mixing reporter system is needed [11]. Small molecule content-mixing reporters always suffer from leakage problems. The content mixing reporter pair 8aminonaphthalene-1,3,6-trisulfonic acid and p-xylene-bis-pyridinium bromide severely leak from closed systems [12]. Other content mixing studies using carboxyfluorescein showed membrane leakiness after SNARE protein reconstitution at a physiological protein-to-lipid ratio (>1:500) [13,14]. Kyung et al. performed a lipid- and content-mixing assay concomitantly in one system [15]. Lipid exchange and content mixing were monitored by two fluorescence dyes, 1,10 -dioctadecyl-3,3,30 ,30 -tetramethylindodicarbocyanine and sulforhodamine B. The system was elegant for observing whole membrane fusion, but could not avoid the leakiness of content dyes [16]. Leakage mainly results from the intrinsic membrane permeability of fluorescence dyes and transient destabilization of the membranes associated with protein reconstitution. This leakage problem makes the data analysis of ensemble experiments challenging, while single-vesicle experiments show some tolerance to mild content leakage [15,17]. To avoid the leakage problem, several DNA-based contentmixing assays have been developed. In the original DNA-based content-mixing assay which used 33P-labeled oligonucleotides, the content mixing signal was indirectly measured after the vesicles were lysed with detergent. Subsequently, DNA-based contentmixing was quantified using fluorescence. In this setting, a FRET pair present on a hairpin-like DNA in a vesicle was linearized by hybridization to a complementary DNA from the other vesicle, with contents mixing reported as increased FRET efficiency [18,19]. This approach was utilized to demonstrate that the yeast SNARE complex alone could expand the fusion pore large enough to transmit ~11 kDa cargoes, and that Ca2þ, Syt1, and complexin 1 stimulated pore expansion [20]. However, because of low encapsulation efficiency of unlabeled target DNA, it is necessary to conjugate the DNA to cholesterol. For this assay, some concerns remain regarding FRET because of lipid flip-flops [19]. Furthermore, unincorporated fluorescent DNA, which occurs for small molecule fluorescence dyes, is difficult to remove. Thus, the requirements for an efficient contentmixing assay are as follows: (1) there should be no leakage even after SNARE protein reconstitution, (2) reporters should be encapsulated inside of a vesicle in a purely soluble manner, and (3) complete removal of unincorporated dyes outside the liposome can be achieved. Here, we developed a robust content-mixing assay using a FRET pair, biotinylated-green fluorescence protein (bGFP) and Dylight 594-conjugated streptavidin (DyStrp). 2. Materials and methods 2.1. Materials and equipment 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1,2dioleoyl-sn-glycero-3-phospho-L-serine (DOPS) and cholesterol were purchased from Avanti Polar Lipids (Alabaster, AL, USA). Dylight 594-conjugated streptavidin (DyStrp) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). All other chemicals were obtained from Sigma Aldrich (St. Louis, MO, USA). Size distribution was determined by dynamic laser light scattering spectroscopy using DynaPro NanoStar (Wyatt Technology, Santa Barbara, CA, USA). Fluorescence signals were measured using a Spectramax M2 plate reader (Molecular Devices, Sunnyvale, CA, USA) or Synergy H1 Hybrid Reader (Bio-Tek Instruments, Winooski, VT, USA).
2.2. Protein expression and purification Protein expression and purification were performed as previously described [5]. Briefly, full-length VAMP2 (Vp2, amino acids 1e116), soluble SNARE motif of VAMP2 (VpS, amino acids 1e96), full-length syntaxin 1a (SynF, amino acids 1e288), and SNAP-25 (amino acids 1e206) were expressed as fusion proteins with glutathione S-transferase (GST). Recombinant GST fusion proteins were expressed in Escherichia coli CodonPlus RIL. 2.3. Conjugation of biotin to GFP Purified GFP was biotinylated using sulfo-N-hydroxysuccinimidyl-biotin, which reacts with primary amino groups to form stable amide bonds. Next, 2 mg of GFP dissolved in 0.5 mL of sodium phosphate buffer containing 150 mM sodium chloride (pH 7.2) was treated with a 20-fold molar excess of biotin reagent and incubated at 25 C for 0.5 h or at 4 C for 2 h. Unlabeled free biotin reagents were removed using a Zeba spin desalting column (7 K MWCO, Thermo Fisher Scientific). The incorporation rate of biotin was measured using 40 -hydroxyazobenzene-2-carboxylic acid assay kit following the manufacturer's instruction (Thermo Fisher Scientific). 2.4. Reconstitution of SNAREs in liposomes enclosing reporter proteins Liposomes were prepared from a mixture of POPC:DOPS:cholesterol (65:15:30). Lipid mixture in chloroform was dried using nitrogen gas and overnight desiccation. Lipid films were dissolved in PBS containing bGFP or DyStrp at a final lipid concentration of 100 mM. The hydrated lipid-reporter mixture was frozen at 80 C, and thawed by quickly placing the test tube in a water bath at 46 C. The volume of mixture was 0.1 mL, freezing time was 10 min, and thawing time was 50 s. After two freeze/thaw cycles, proteoliposomes were extruded through 100-nm polycarbonate filters (Avanti Polar Lipids). These processes cause the content reporters to become equally distributed inside and outside of large unilamellar vesicles (LUVs). The SNAREs were reconstituted in the reporter protein-containing LUVs. Each binary t-SNARE complex and Vp2 were mixed with donor- or acceptor-containing LUVs at a desired lipid-to-protein (L:P) ratio of up to 100:1. The concentration of OG was maintained at 0.8%, which is slightly above its critical micelle concentration. After incubating the mixture at room temperature for 30 min, dialysis was performed against 2 L dialysis buffer (25 mM HEPES, 100 mM KCl, pH 7.4) with SM2 Bio-beads (Bio-Rad, Hercules, CA, USA) at 4 C overnight to remove the detergent. Additionally, to remove residual OG, SM2 Bio-beads were added directly to the sample. Un-encapsulated protein reporters outside of liposomes were eliminated using 0.1 mL Ni-NTA agarose beads (~5 mg/mL binding capacity, Elpis-biotech, Daejeon, Korea) or biotin-agarose beads (~2 mg/mL binding capacity, Thermo Fischer Scientific) for bGFP or DyStrp, respectively, at room temperature for 30 min and at 4 C for 2 h. The concentration of encapsulated reporters and reconstituted SNAREs were quantified by fluorescence intensity measurement and SDS-PAGE, respectively. 2.5. Content-mixing assay To monitor content mixing, v-vesicles (0.1 mM lipids) containing the donor reporter (bGFP) were added to a microplate. After stabilization of both the donor (excitation/emission: 488 nm/ 535 nm, cutoff 530 nm) and FRET (ex./em.: 488 nm/618 nm, cutoff 610 nm) signals at 37 C in the fluorescence reader, t-vesicles (0.1e0.9 mM lipids) containing acceptor (DyStrp) were added to the
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
3
plate and the fluorescence signal was monitored. After reading the fluorescence at 37 C for 90 min, 0.1% Triton X-100 (TX100) was added and fluorescence intensity of complete mixing was measured. FRET efficiency (E) was calculated as E ¼ IA/(IA þ ID), where IA and ID are the intensities of the FRET and donor signals, respectively.
than 5 nm would allow the FRET pair pass through. Content-mixing reporters must be larger than 2 nm to precisely report fusion pore expansion [21].
3. Results
For the content-mixing assay of SNARE-driven membrane fusion, reporters must be efficiently encapsulated in the vesicle. Additionally, SNARE proteins should be functionally reconstituted on the vesicle. Four experimental schemes were tested and the resulting vesicles were characterized in terms of SNARE reconstitution and reporter protein encapsulation (Fig. 2 and Table 1). In scheme A (Fig. 2A) showing the co-micellization method for SNARE reconstitution, SNARE proteins and reporter proteins were initially co-solubilized together with phospholipids in a detergent OG solution. Next, the mixture was dialyzed to remove the detergent, and reporter proteins outside the vesicles were removed using biotinagarose or Ni-NTA agarose (Fig. 3). SNARE proteins showed good reconstitution into the vesicle, but the resulting vesicles did not contain detectable fluorescence of the reporter proteins (Table 1). Approximately 15% of initial reporter proteins are expected to be retained in the resulting vesicles when the ratio of entrapped volume to extravesicular volume was considered. However, nearly no reporter proteins were encapsulated because reporter proteins were leaked from the vesicles by OG during dialysis. Because simultaneous encapsulation of reporter proteins and reconstitution of SNARE proteins was not achievable, these processes were performed consecutively. Reconstitution of SNARE
3.1. GFP-based content-mixing assay A FRET pair, GFP and Dylight 594, was prepared to specifically interact with each other by conjugating these molecules to biotin and streptavidin, respectively (Fig. 1A). Thus, the donor bGFP transfers its emission energy to the acceptor DyStrp upon the biotin-streptavidin interaction with a dissociation constant of ~1015 M. FRET between GFP and Dylight 594 effectively occurred only when GFP was biotinylated with a long-chain alkyl spacer (Fig. 1B). The bGFP-DyStrp interaction caused an increase in the acceptor's FRET intensity, unlike the standard lipid mixing assay or DNA-based content-mixing assay [18,19], which measure the increase in donor signal resulting from dequenching. This difference in read-out method enabled measurement of the stoichiometry of the interaction between bGFP and DyStrp (Fig. 1B and C). An equimolar ratio of acceptor and donor showed a FRET efficiency of 0.5. A FRET efficiency of more than 0.75 was obtained when [acceptor]/[donor] was above 2. The hydrodynamic diameters of bGFP and DyStrp with molecular weights of ~28 and ~60 kDa were ~5.06 and 5.86 nm, respectively (Fig. 1D). Thus, a fusion pore larger
3.2. Experimental schemes for efficient encapsulation of reporter proteins and reconstitution of SNARE proteins
Fig. 1. FRET pair for content-mixing assay. (A) Schematic of GFP-based content-mixing assay. Biotinylated GFP (bGFP) and streptavidin conjugated with Dylight 594 (DyStrp) constitute a FRET pair. Donor bGFP (ex./em.: 488/535 nm, cutoff 530 nm) and acceptor DyStrp (ex./em.: 488/618 nm, cutoff 610 nm) are encapsulated in separated vesicles for the content-mixing assay. (B) FRET efficiency depending on GFP-biotin conjugation. Indicated concentrations of DyStrp were added to 50 nM GFP (triangle), biotin-short chain-GFP (biotin-GFP), or biotin-long chain-GFP (biotin-LC-GFP, bGFP). FRET efficiency is defined as Ia/(Ia þ Id), where Id and Ia are donor and FRET fluorescence intensities, respectively (t ¼ 15 s). (C) Time-dependent changes in FRET efficiency. FRET efficiency increases upon addition of DyStrp to bGFP. (D) Size distribution of bGFP and DyStrp.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
4
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
Fig. 2. Schematic diagram of various reconstitution schemes containing both SNARE proteins and content-mixing reporters.
Table 1 Schemes for content encapsulation and SNARE reconstitution into liposomes. Step (1) (2) (3)
(4)
a b c d
Procedures a
SNAREs during lipid hydration No. of freeze/thaw Extrusion (times) Removal of unencapsulated contents [OG] for SNARE reconstitution Dialysis (h)b Removal of outside contents (times) SNAREs reconstituted in liposomec Encapsulation of reportersd Leakage
Scheme A
Scheme B
Scheme C
Scheme D
þ 0 0 e 0.8% 20 2 þ e ND
e 2 10 þ 0.8% 20 2 þ e ND
e 2 10 e 0.3% 20 2 e þ e
e 2 10 e 0.8% 20 2 þ þ e
Contents (bGFP or DyStrp) were supplemented during lipid hydration step in all schemes. Dialysis was performed in a 10,000 MWCO dialysis tube against a buffer containing Bio-bead SM2. Determined by SDS-PAGE. Determined by measuring fluorescence intensity.
proteins into the liposome was performed after encapsulating the reporter proteins inside vesicles (Scheme BeD). First, after the lipids were hydrated in the presence of reporter proteins but without SNARE proteins, the hydrated lipids were subjected to 2 freeze/thaw cycles. While the freeze/thaw cycle is known to enhance liposome formation during subsequent extrusion steps, this treatment can also destroy protein structure. We found that vesicles were formed in the extrusion step without a significant loss in reporter fluorescence after 2 freeze/thaw cycles. Extrusion was performed using polycarbonate filters with 100-nm pore sizes. Unencapsulated reporter proteins were removed using biotin-agarose or Ni-NTA agarose. Next, SNARE proteins were reconstituted into vesicles containing reporter proteins. For efficient reconstitution, SNARE proteins solubilized in OG were mixed with extruded liposomes [5] at final concentrations of 0.8e1.0%. The mixture was
diluted by 2e3-fold to lower the concentration of OG below its critical micelle concentration (0.67e0.73%), and then further dialyzed at 4 C for ~20 h. This procedure efficiently delivered SNARE proteins from the detergent to membranes with preferred extravesicular orientation. SNARE proteins were reconstituted into vesicles containing reporter proteins using this procedure. However, the encapsulated reporter proteins disappeared during the dialysis process (scheme B). We assumed that the encapsulated reporter proteins were leaked from the vesicles during prolonged dialysis because there was 0.2e0.4% OG in the 2-fold diluted solution. To prevent leakage of the reporter proteins due to OG, which is necessary for SNARE reconstitution, reporter-containing vesicles were directly subjected to SNARE reconstitution before removing unencapsulated reporter proteins (Scheme C and D). OG-solubilized SNARE proteins were
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
5
Fig. 3. Removal of unencapsulated dyes and robustness of encapsulation. (A, B) DyStrp and bGFP not encapsulated in the vesicles can be removed by biotin-agarose and Ni-NTA beads, respectively. Approximately 20% for bGFP and 10% for DyStrp fluorescence are not removed by the beads because the indicators are inside the vesicles. They can be removed only after 0.1% TX100 treatment. (C) An additional round of bead treatment ensures removal of free dyes. The colors for bGFP (green) and DyStrp (pink) are not observed in the second bead treatment. (D) When t- and v-vesicles were mixed after removing unencapsulated dyes, FRET did not increase at 4 C. FRET efficiency increased only when TX100 was added. (E, F) Robust vesicles. Encapsulated DyStrp and bGFP are not removed by biotin-agarose or Ni-NTA beads even at 37 C for 10 h, indicating that the content-mixing reporters are not leaked from of vesicles during incubation. The reporters were maintained inside the vesicles even at a lipid-to-protein ratio of 100:1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
mixed with encapsulation solution so that the inward and outward flux of reporters were the same (Scheme C and D). Next, different concentrations of OG were treated for SNARE reconstitution (~0.3% in scheme C and 0.8% in scheme D). At low OG concentration (scheme C), the SNARE reconstitution was extremely inefficient and the protein contents were kept inside the vesicles. Scheme D enabled successful reconstitution of SNARE proteins and prevented the loss of encapsulated reporter proteins during the SNARE reconstitution step. The overall yield of reporter protein encapsulation following scheme D (Fig. 2D) was 10e20%. Unencapsulated reporter proteins were easily removed using Ni-NTA agarose and biotin-agarose in 5 min (Fig. 3AeC). When both vesicles containing bGFP and DyStrp were mixed, there was no FRET increase, indicating that the protein contents outside the vesicles were effectively removed (Fig. 3D).
When the vesicle mixture was treated with TX100, FRET between bGFP and DyStrp was increased. FRET efficiency of the bGFP-DyStrp pair measured after membrane disruption with TX100 depended on the [acceptor]/[donor] ratio consistently with the mixing of reporter proteins in solution. Leakage of reporter proteins from the vesicles containing reconstituted SNARE proteins was analyzed (Fig. 3E and F). Vesicles with different lipid-to-protein (L:P) ratios from 1000:1 to 100:1 did not leak encapsulated reporter proteins even at 37 C for 10 h. Reporter proteins bound to agarose beads only when the vesicles encapsulating the reporter proteins were disrupted by detergent, suggesting that there was virtually no leakage of protein contents from the vesicles.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
6
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
3.3. Large fusion pore is generated during SNARE-mediated membrane fusion
vesicle containing bGFP. 4. Discussion
Both t- and v-vesicles were mixed and FRET change was monitored as a function of time at 37 C (Fig. 4A). FRET between DyStrp and bGFP increased over time, indicating that fusion pores were formed through SNARE-driven fusion. When the soluble fragment of Vp2 (VpS) was added to the reaction mixture, content mixing was completely inhibited, excluding the possibility that FRET increased because of reporter protein leakage (Fig. 4B). Furthermore, up to a 4-M excess of DyStrp mixed with v-vesicle containing bGFP did not induce a FRET increase, excluding the possibility that FRET changed because of reporter protein leakage. Thus, SNARE proteins opened a fusion pore that was large enough to allow the passage of bGFP, which has a molecular weight larger than 28 kDa. Because the hydrodynamic diameter of bGFP was 5.06 nm (Fig. 1D), the fusion pore opened by SNARE proteins must be larger than 5 nm. Synaptotagmin 1 (Syt1) is a Ca2þ-sensor protein responsible for synchronous and/or fast membrane fusion. C2AB, which lacks the transmembrane domain of Syt1, facilitates multiple fusions between vesicles, although whether this occurs in the neuron or by full-length Syt1 in the reconstituted system remains unclear [4,17,22]. We made use of this feature to test whether the FRET can report stoichiometry of fusion. When Ca2þ was added to the vesicle mixture FRET efficiency was increased and reached ~0.6. Because this FRET value is achieved when DyStrp is >1.5 times above bGFP, we assume that a few t-vesicles containing DyStrp fused with a v-
Content-mixing between two merging vesicles is the final stage of membrane fusion. However, recent studies suggest that lipidmixing assays do not reflect the opening of fusion pores. While several content-mixing reporter systems based on small fluorescence dyes are currently utilized [12e15,23], they all exhibit severe leakage problems. To overcome this limitation, a hybridizing DNA pair probe was developed and utilized to analyze fusion pore opening [18]. However, the efficiency of encapsulation of DNAs into vesicle is extremely low. Encapsulation efficiency can be increased by attaching DNA to cholesterol, which can be easily incorporated into the vesicle. DNAs protruding towards the outside of the vesicle can be removed by DNase treatment, leaving the DNA inside vesicles intact. However, the reporter pair is a hybrid reporter of lipidmixing and content-mixing because one of the DNA pairs is anchored to the membrane, raising concerns regarding lipid flipflop. Although GFP variants are rich sources of FRET pairs that can be used as content-mixing reporters, these proteins have not been used for this purpose. This may be because of a low encapsulation efficiency similar to DNA reporters. Thus, one of the goals of this study was to develop an efficient method for encapsulating cargos in the vesicle during functional reconstitution of SNARE proteins into the vesicle. Cargo encapsulation and SNARE reconstitution were attempted simultaneously and sequentially (Fig. 2).
Fig. 4. Content-mixing assay of SNARE-driven membrane fusion using bGFP and DyStrp as a FRET pair. (A) Real-time trace of content-mixing assay. (B) FRET efficiency increases only by SNARE-driven fusion. Inhibition of membrane fusion by soluble domain of Vp2 (VpS) or addition of DyStrp dyes to t-vesicle did not induce FRET increase, indicating the robustness of the reconstituted system. (C, D) FRET efficiency depended on the stoichiometry of bGFP and DyStrp, indicating multiple rounds of fusion induced by Ca2þ and C2AB domain.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
Simultaneous encapsulation and reconstitution (scheme A) did not maintain GFP inside the vesicles, although SNAREs were reconstituted well. Both cargo encapsulation and SNARE reconstitution were successful only when the experimental procedures were as follows, in which cargo encapsulation and SNARE reconstitution were performed sequentially. First, the lipid film was hydrated in the presence of cargo. Second, the lipid solution was frozen and thawed twice to enhance multi-lamellar vesicle formation. Cargo molecules are evenly distributed outside and inside multi-lamellar structures through this treatment. However, the number of cycles should be optimized because freeze/thaw treatment can inactivate proteins. Two freeze/thaw cycles were experimentally determined to be optimal. Third, LUVs were formed by extruding the multilamellar vesicle through a polycarbonate membrane filter with 100-nm pores. Fourth, cargo molecules outside the LUVs should not be removed before the SNARE reconstitution step. When cargoes outside vesicles were removed at this time (scheme B), the final proteoliposome after SNARE reconstitution did not contain encapsulated cargos. Because the membrane was permeabilized by OG in the next step, the in-and-out flux of content-proteins was balanced only when there were equivalent concentrations of content-proteins inside and outside the vesicles. Fifth, binary tSNARE complexes or Vp2 were added to the mixture prepared above. A final OG concentration of 0.8% enabled efficient SNARE reconstitution (scheme D), while lower OG concentrations did not allow SNARE reconstitution (scheme C). Sixth, the vesicles containing cargo and SNAREs were anchored to the membrane and finally rigidified by removing OG through dialysis. Finally, cargo outside the vesicles was easily removed by treatment with biotinagarose beads and Ni-NTA-agarose for DyStrp and bGFP, respectively (Fig. 3). Vesicles formed using this procedure did not leak cargo molecules, even at a 100:1 LP ratio at 37 C for 10 h. Acknowledgements This research was supported by grants from the Korea Research Foundation (NRF-2017R1A2B2008211), and The Advanced Biomass R&D Center (ABC) of Korea (2011-0031359) funded by the Ministry of Science, ICT and Future Planning, South Korea. Appendix A. Supplementary data
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12] [13]
[14]
[15]
[16] [17]
[18]
[19]
[20]
Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.bbrc.2017.05.006.
[21]
References
[22]
[1] A.T. Brunger, K. Weninger, M. Bowen, S. Chu, Single-molecule studies of the neuronal SNARE fusion machinery, Annu. Rev. Biochem. 78 (2009) 903e928. [2] D.H. Kweon, B. Kong, Y.K. Shin, Hemifusion in synaptic vesicle cycle, Front. Mol. Neurosci. 10 (2017) 65. [3] T. Weber, B.V. Zemelman, J.A. McNew, B. Westermann, M. Gmachl, F. Parlati,
[23]
7
T.H. Sollner, J.E. Rothman, SNAREpins: minimal machinery for membrane fusion, Cell 92 (1998) 759e772. A. Stein, A. Radhakrishnan, D. Riedel, D. Fasshauer, R. Jahn, Synaptotagmin activates membrane fusion through a Ca2þ-dependent trans interaction with phospholipids, Nat. Struct. Mol. Biol. 14 (2007) 904e911. Y. Yang, J.Y. Shin, J.M. Oh, C.H. Jung, Y. Hwang, S. Kim, J.S. Kim, K.J. Yoon, J.Y. Ryu, J. Shin, J.S. Hwang, T.Y. Yoon, Y.K. Shin, D.H. Kweon, Dissection of SNARE-driven membrane fusion and neuroexocytosis by wedging small hydrophobic molecules into the SNARE zipper, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 22145e22150. P. Heo, J.B. Park, Y.K. Shin, D.H. Kweon, Visualization of SNARE-mediated hemifusion between giant unilamellar vesicles arrested by myricetin, Front. Mol. Neurosci. 10 (2017) 93. Y.H. Chan, B. van Lengerich, S.G. Boxer, Effects of linker sequences on vesicle fusion mediated by lipid-anchored DNA oligonucleotides, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 979e984. S.M. Dennison, M.E. Bowen, A.T. Brunger, B.R. Lentz, Neuronal SNAREs do not trigger fusion between synthetic membranes but do promote PEG-mediated membrane fusion, Biophys. J. 90 (2006) 1661e1675. D.L. Floyd, J.R. Ragains, J.J. Skehel, S.C. Harrison, A.M. van Oijen, Single-particle kinetics of influenza virus membrane fusion, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 15382e15387. J. Diao, P. Grob, D.J. Cipriano, M. Kyoung, Y. Zhang, S. Shah, A. Nguyen, M. Padolina, A. Srivastava, M. Vrljic, A. Shah, E. Nogales, S. Chu, A.T. Brunger, Synaptic proteins promote calcium-triggered fast transition from point contact to full fusion, Elife 1 (2012) e00109. J. Diao, M. Zhao, Y. Zhang, M. Kyoung, A.T. Brunger, Studying proteinreconstituted proteoliposome fusion with content indicators in vitro, BioEssays 35 (2013) 658e665. N. Duzgunes, J. Wilschut, Fusion assays monitoring intermixing of aqueous contents, Methods Enzymol. 220 (1993) 3e14. G. van den Bogaart, M.G. Holt, G. Bunt, D. Riedel, F.S. Wouters, R. Jahn, One SNARE complex is sufficient for membrane fusion, Nat. Struct. Mol. Biol. 17 (2010) 358e364. J. Zimmerberg, F.S. Cohen, A. Finkelstein, Fusion of phospholipid vesicles with planar phospholipid bilayer membranes. I. Discharge of vesicular contents across the planar membrane, J. Gen. Physiol. 75 (1980) 241e250. M. Kyoung, A. Srivastava, Y. Zhang, J. Diao, M. Vrljic, P. Grob, E. Nogales, S. Chu, A.T. Brunger, In vitro system capable of differentiating fast Ca2þ-triggered content mixing from lipid exchange for mechanistic studies of neurotransmitter release, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 304e313. G. van den Bogaart, R. Jahn, Inside insight to membrane fusion, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 11729e11730. P. Heo, Y. Yang, K.Y. Han, B. Kong, J.H. Shin, Y. Jung, C. Jeong, J. Shin, Y.K. Shin, T. Ha, D.H. Kweon, A chemical controller of SNARE-driven membrane fusion that primes vesicles for Ca2þ-triggered millisecond exocytosis, J. Am. Chem. Soc. 138 (2016) 4512e4521. J. Diao, Z. Su, Y. Ishitsuka, B. Lu, K.S. Lee, Y. Lai, Y.K. Shin, T. Ha, A single-vesicle content mixing assay for SNARE-mediated membrane fusion, Nat. Commun. 1 (2010) 54. J. Diao, Y. Ishitsuka, H. Lee, C. Joo, Z. Su, S. Syed, Y.K. Shin, T.Y. Yoon, T. Ha, A single vesicle-vesicle fusion assay for in vitro studies of SNAREs and accessory proteins, Nat. Protoc. 7 (2012) 921e934. Y. Lai, J. Diao, Y. Liu, Y. Ishitsuka, Z. Su, K. Schulten, T. Ha, Y.K. Shin, Fusion pore formation and expansion induced by Ca2þ and synaptotagmin 1, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 1333e1338. L. Shi, Q.T. Shen, A. Kiel, J. Wang, H.W. Wang, T.J. Melia, J.E. Rothman, F. Pincet, SNARE proteins: one to fuse and three to keep the nascent fusion pore open, Science 335 (2012) 1355e1359. M.C. Chicka, E. Hui, H. Liu, E.R. Chapman, Synaptotagmin arrests the SNARE complex before triggering fast, efficient membrane fusion in response to Ca2þ, Nat. Struct. Mol. Biol. 15 (2008) 827e835. J.M. Johnson, T. Ha, S. Chu, S.G. Boxer, Early steps of supported bilayer formation probed by single vesicle fluorescence assays, Biophys. J. 83 (2002) 3371e3379.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Green fluorescence protein-based content-mixing assay of SNAREdriven membrane fusion Paul Heo, Byoungjae Kong, Young-Hun Jung, Joon-Bum Park, Jonghyeok Shin, Myungseo Park, Dae-Hyuk Kweon* Department of Genetic Engineering, College of Biotechnology and Bioengineering, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, South Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 27 April 2017 Accepted 1 May 2017 Available online xxx
Soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins mediate intracellular membrane fusion by forming a ternary SNARE complex. A minimalist approach utilizing proteoliposomes with reconstituted SNARE proteins yielded a wealth of information pinpointing the molecular mechanism of SNARE-mediated fusion and its regulation by accessory proteins. Two important attributes of a membrane fusion are lipid-mixing and the formation of an aqueous passage between apposing membranes. These two attributes are typically observed by using various fluorescent dyes. Currently available in vitro assay systems for observing fusion pore opening have several weaknesses such as cargo-bleeding, incomplete removal of unencapsulated dyes, and inadequate information regarding the size of the fusion pore, limiting measurements of the final stage of membrane fusion. In the present study, we used a biotinylated green fluorescence protein and streptavidin conjugated with € ster resonance energy transfer (FRET) donor and acceptor, respectively. This Dylight 594 (DyStrp) as a Fo FRET pair encapsulated in each v-vesicle containing synaptobrevin and t-vesicle containing a binary acceptor complex of syntaxin 1a and synaptosomal-associated protein 25 revealed the opening of a large fusion pore of more than 5 nm, without the unwanted signals from unencapsulated dyes or leakage. This system enabled determination of the stoichiometry of the merging vesicles because the FRET efficiency of the FRET pair depended on the molar ratio between dyes. Here, we report a robust and informative assay for SNARE-mediated fusion pore opening. © 2017 Elsevier Inc. All rights reserved.
Keywords: SNARE Membrane fusion Content-mixing assay Green fluorescence protein Fusion pore
1. Introduction Membrane fusion is an important cellular process by which two lipid bilayers merge and form a pore through which cargos diffuse. All intracellular membrane fusions are mediated by soluble-Nethylmaleimide-sensitive-factor attachment protein receptor (SNARE)-family members. The membrane fusion is catalyzed by assembly of 3 Q-SNAREs (glutamine residue at the zeroth layer) and 1 R-SNARE (arginine residue at the zeroth layer), with Q-SNAREs typically targeting the plasma membrane (t-SNARE) while RSNAREs are typically vesicle-associated (v-SNARE). SNARE complex formation, which involves the directional assembly of the SNARE motifs to form a parallel 4 helical bundle, is thought to provide the free energy required to promote lipid bilayer fusion [1,2]. An approach using reconstituted proteoliposomes with SNARE
* Corresponding author. E-mail address: [email protected] (D.-H. Kweon).
proteins was first used to examine the ‘SNAREpin’ as the minimal fusion machinery [3]. Since this discovery, the fluorescence €rster) resonance energy transfer (FRET)-based lipid mixing assay (Fo has been the gold standard in vitro tool for studying SNAREmediated membrane fusion [4e6]. In the most common scheme, t- and v-SNARE proteins are reconstituted into vesicles separately (the resulting vesicles are known as t- and v-vesicles, respectively), where one vesicle is fluorophore-free and the other contains both acceptor and donor fluorescence dyes. Upon fusion by SNARE complex formation, the average distance between the donors and acceptors in a vesicle increases following lipid mixing, dequenching donor fluorescence emission. The critical roles of SNARE proteins and regulatory proteins were revealed using this assay. However, the lipid mixing assay was not consistently directly correlated with fusion pore opening, which is the final step in the fusion pathway. A previous study showed that a high level of lipid mixing occurred with a limited degree of content mixing [7]. In viral fusion and vacuolar fusion, content mixing occurred seconds
http://dx.doi.org/10.1016/j.bbrc.2017.05.006 0006-291X/© 2017 Elsevier Inc. All rights reserved.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
2
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
to minutes later than initial lipid mixing [8,9]. Neuronal SNAREs also exhibited significant lipid mixing seconds before content mixing [10]. Thus, lipid mixing alone is not sufficient to clearly detect fusion pore opening, and thus an efficient content-mixing reporter system is needed [11]. Small molecule content-mixing reporters always suffer from leakage problems. The content mixing reporter pair 8aminonaphthalene-1,3,6-trisulfonic acid and p-xylene-bis-pyridinium bromide severely leak from closed systems [12]. Other content mixing studies using carboxyfluorescein showed membrane leakiness after SNARE protein reconstitution at a physiological protein-to-lipid ratio (>1:500) [13,14]. Kyung et al. performed a lipid- and content-mixing assay concomitantly in one system [15]. Lipid exchange and content mixing were monitored by two fluorescence dyes, 1,10 -dioctadecyl-3,3,30 ,30 -tetramethylindodicarbocyanine and sulforhodamine B. The system was elegant for observing whole membrane fusion, but could not avoid the leakiness of content dyes [16]. Leakage mainly results from the intrinsic membrane permeability of fluorescence dyes and transient destabilization of the membranes associated with protein reconstitution. This leakage problem makes the data analysis of ensemble experiments challenging, while single-vesicle experiments show some tolerance to mild content leakage [15,17]. To avoid the leakage problem, several DNA-based contentmixing assays have been developed. In the original DNA-based content-mixing assay which used 33P-labeled oligonucleotides, the content mixing signal was indirectly measured after the vesicles were lysed with detergent. Subsequently, DNA-based contentmixing was quantified using fluorescence. In this setting, a FRET pair present on a hairpin-like DNA in a vesicle was linearized by hybridization to a complementary DNA from the other vesicle, with contents mixing reported as increased FRET efficiency [18,19]. This approach was utilized to demonstrate that the yeast SNARE complex alone could expand the fusion pore large enough to transmit ~11 kDa cargoes, and that Ca2þ, Syt1, and complexin 1 stimulated pore expansion [20]. However, because of low encapsulation efficiency of unlabeled target DNA, it is necessary to conjugate the DNA to cholesterol. For this assay, some concerns remain regarding FRET because of lipid flip-flops [19]. Furthermore, unincorporated fluorescent DNA, which occurs for small molecule fluorescence dyes, is difficult to remove. Thus, the requirements for an efficient contentmixing assay are as follows: (1) there should be no leakage even after SNARE protein reconstitution, (2) reporters should be encapsulated inside of a vesicle in a purely soluble manner, and (3) complete removal of unincorporated dyes outside the liposome can be achieved. Here, we developed a robust content-mixing assay using a FRET pair, biotinylated-green fluorescence protein (bGFP) and Dylight 594-conjugated streptavidin (DyStrp). 2. Materials and methods 2.1. Materials and equipment 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1,2dioleoyl-sn-glycero-3-phospho-L-serine (DOPS) and cholesterol were purchased from Avanti Polar Lipids (Alabaster, AL, USA). Dylight 594-conjugated streptavidin (DyStrp) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). All other chemicals were obtained from Sigma Aldrich (St. Louis, MO, USA). Size distribution was determined by dynamic laser light scattering spectroscopy using DynaPro NanoStar (Wyatt Technology, Santa Barbara, CA, USA). Fluorescence signals were measured using a Spectramax M2 plate reader (Molecular Devices, Sunnyvale, CA, USA) or Synergy H1 Hybrid Reader (Bio-Tek Instruments, Winooski, VT, USA).
2.2. Protein expression and purification Protein expression and purification were performed as previously described [5]. Briefly, full-length VAMP2 (Vp2, amino acids 1e116), soluble SNARE motif of VAMP2 (VpS, amino acids 1e96), full-length syntaxin 1a (SynF, amino acids 1e288), and SNAP-25 (amino acids 1e206) were expressed as fusion proteins with glutathione S-transferase (GST). Recombinant GST fusion proteins were expressed in Escherichia coli CodonPlus RIL. 2.3. Conjugation of biotin to GFP Purified GFP was biotinylated using sulfo-N-hydroxysuccinimidyl-biotin, which reacts with primary amino groups to form stable amide bonds. Next, 2 mg of GFP dissolved in 0.5 mL of sodium phosphate buffer containing 150 mM sodium chloride (pH 7.2) was treated with a 20-fold molar excess of biotin reagent and incubated at 25 C for 0.5 h or at 4 C for 2 h. Unlabeled free biotin reagents were removed using a Zeba spin desalting column (7 K MWCO, Thermo Fisher Scientific). The incorporation rate of biotin was measured using 40 -hydroxyazobenzene-2-carboxylic acid assay kit following the manufacturer's instruction (Thermo Fisher Scientific). 2.4. Reconstitution of SNAREs in liposomes enclosing reporter proteins Liposomes were prepared from a mixture of POPC:DOPS:cholesterol (65:15:30). Lipid mixture in chloroform was dried using nitrogen gas and overnight desiccation. Lipid films were dissolved in PBS containing bGFP or DyStrp at a final lipid concentration of 100 mM. The hydrated lipid-reporter mixture was frozen at 80 C, and thawed by quickly placing the test tube in a water bath at 46 C. The volume of mixture was 0.1 mL, freezing time was 10 min, and thawing time was 50 s. After two freeze/thaw cycles, proteoliposomes were extruded through 100-nm polycarbonate filters (Avanti Polar Lipids). These processes cause the content reporters to become equally distributed inside and outside of large unilamellar vesicles (LUVs). The SNAREs were reconstituted in the reporter protein-containing LUVs. Each binary t-SNARE complex and Vp2 were mixed with donor- or acceptor-containing LUVs at a desired lipid-to-protein (L:P) ratio of up to 100:1. The concentration of OG was maintained at 0.8%, which is slightly above its critical micelle concentration. After incubating the mixture at room temperature for 30 min, dialysis was performed against 2 L dialysis buffer (25 mM HEPES, 100 mM KCl, pH 7.4) with SM2 Bio-beads (Bio-Rad, Hercules, CA, USA) at 4 C overnight to remove the detergent. Additionally, to remove residual OG, SM2 Bio-beads were added directly to the sample. Un-encapsulated protein reporters outside of liposomes were eliminated using 0.1 mL Ni-NTA agarose beads (~5 mg/mL binding capacity, Elpis-biotech, Daejeon, Korea) or biotin-agarose beads (~2 mg/mL binding capacity, Thermo Fischer Scientific) for bGFP or DyStrp, respectively, at room temperature for 30 min and at 4 C for 2 h. The concentration of encapsulated reporters and reconstituted SNAREs were quantified by fluorescence intensity measurement and SDS-PAGE, respectively. 2.5. Content-mixing assay To monitor content mixing, v-vesicles (0.1 mM lipids) containing the donor reporter (bGFP) were added to a microplate. After stabilization of both the donor (excitation/emission: 488 nm/ 535 nm, cutoff 530 nm) and FRET (ex./em.: 488 nm/618 nm, cutoff 610 nm) signals at 37 C in the fluorescence reader, t-vesicles (0.1e0.9 mM lipids) containing acceptor (DyStrp) were added to the
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
3
plate and the fluorescence signal was monitored. After reading the fluorescence at 37 C for 90 min, 0.1% Triton X-100 (TX100) was added and fluorescence intensity of complete mixing was measured. FRET efficiency (E) was calculated as E ¼ IA/(IA þ ID), where IA and ID are the intensities of the FRET and donor signals, respectively.
than 5 nm would allow the FRET pair pass through. Content-mixing reporters must be larger than 2 nm to precisely report fusion pore expansion [21].
3. Results
For the content-mixing assay of SNARE-driven membrane fusion, reporters must be efficiently encapsulated in the vesicle. Additionally, SNARE proteins should be functionally reconstituted on the vesicle. Four experimental schemes were tested and the resulting vesicles were characterized in terms of SNARE reconstitution and reporter protein encapsulation (Fig. 2 and Table 1). In scheme A (Fig. 2A) showing the co-micellization method for SNARE reconstitution, SNARE proteins and reporter proteins were initially co-solubilized together with phospholipids in a detergent OG solution. Next, the mixture was dialyzed to remove the detergent, and reporter proteins outside the vesicles were removed using biotinagarose or Ni-NTA agarose (Fig. 3). SNARE proteins showed good reconstitution into the vesicle, but the resulting vesicles did not contain detectable fluorescence of the reporter proteins (Table 1). Approximately 15% of initial reporter proteins are expected to be retained in the resulting vesicles when the ratio of entrapped volume to extravesicular volume was considered. However, nearly no reporter proteins were encapsulated because reporter proteins were leaked from the vesicles by OG during dialysis. Because simultaneous encapsulation of reporter proteins and reconstitution of SNARE proteins was not achievable, these processes were performed consecutively. Reconstitution of SNARE
3.1. GFP-based content-mixing assay A FRET pair, GFP and Dylight 594, was prepared to specifically interact with each other by conjugating these molecules to biotin and streptavidin, respectively (Fig. 1A). Thus, the donor bGFP transfers its emission energy to the acceptor DyStrp upon the biotin-streptavidin interaction with a dissociation constant of ~1015 M. FRET between GFP and Dylight 594 effectively occurred only when GFP was biotinylated with a long-chain alkyl spacer (Fig. 1B). The bGFP-DyStrp interaction caused an increase in the acceptor's FRET intensity, unlike the standard lipid mixing assay or DNA-based content-mixing assay [18,19], which measure the increase in donor signal resulting from dequenching. This difference in read-out method enabled measurement of the stoichiometry of the interaction between bGFP and DyStrp (Fig. 1B and C). An equimolar ratio of acceptor and donor showed a FRET efficiency of 0.5. A FRET efficiency of more than 0.75 was obtained when [acceptor]/[donor] was above 2. The hydrodynamic diameters of bGFP and DyStrp with molecular weights of ~28 and ~60 kDa were ~5.06 and 5.86 nm, respectively (Fig. 1D). Thus, a fusion pore larger
3.2. Experimental schemes for efficient encapsulation of reporter proteins and reconstitution of SNARE proteins
Fig. 1. FRET pair for content-mixing assay. (A) Schematic of GFP-based content-mixing assay. Biotinylated GFP (bGFP) and streptavidin conjugated with Dylight 594 (DyStrp) constitute a FRET pair. Donor bGFP (ex./em.: 488/535 nm, cutoff 530 nm) and acceptor DyStrp (ex./em.: 488/618 nm, cutoff 610 nm) are encapsulated in separated vesicles for the content-mixing assay. (B) FRET efficiency depending on GFP-biotin conjugation. Indicated concentrations of DyStrp were added to 50 nM GFP (triangle), biotin-short chain-GFP (biotin-GFP), or biotin-long chain-GFP (biotin-LC-GFP, bGFP). FRET efficiency is defined as Ia/(Ia þ Id), where Id and Ia are donor and FRET fluorescence intensities, respectively (t ¼ 15 s). (C) Time-dependent changes in FRET efficiency. FRET efficiency increases upon addition of DyStrp to bGFP. (D) Size distribution of bGFP and DyStrp.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
4
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
Fig. 2. Schematic diagram of various reconstitution schemes containing both SNARE proteins and content-mixing reporters.
Table 1 Schemes for content encapsulation and SNARE reconstitution into liposomes. Step (1) (2) (3)
(4)
a b c d
Procedures a
SNAREs during lipid hydration No. of freeze/thaw Extrusion (times) Removal of unencapsulated contents [OG] for SNARE reconstitution Dialysis (h)b Removal of outside contents (times) SNAREs reconstituted in liposomec Encapsulation of reportersd Leakage
Scheme A
Scheme B
Scheme C
Scheme D
þ 0 0 e 0.8% 20 2 þ e ND
e 2 10 þ 0.8% 20 2 þ e ND
e 2 10 e 0.3% 20 2 e þ e
e 2 10 e 0.8% 20 2 þ þ e
Contents (bGFP or DyStrp) were supplemented during lipid hydration step in all schemes. Dialysis was performed in a 10,000 MWCO dialysis tube against a buffer containing Bio-bead SM2. Determined by SDS-PAGE. Determined by measuring fluorescence intensity.
proteins into the liposome was performed after encapsulating the reporter proteins inside vesicles (Scheme BeD). First, after the lipids were hydrated in the presence of reporter proteins but without SNARE proteins, the hydrated lipids were subjected to 2 freeze/thaw cycles. While the freeze/thaw cycle is known to enhance liposome formation during subsequent extrusion steps, this treatment can also destroy protein structure. We found that vesicles were formed in the extrusion step without a significant loss in reporter fluorescence after 2 freeze/thaw cycles. Extrusion was performed using polycarbonate filters with 100-nm pore sizes. Unencapsulated reporter proteins were removed using biotin-agarose or Ni-NTA agarose. Next, SNARE proteins were reconstituted into vesicles containing reporter proteins. For efficient reconstitution, SNARE proteins solubilized in OG were mixed with extruded liposomes [5] at final concentrations of 0.8e1.0%. The mixture was
diluted by 2e3-fold to lower the concentration of OG below its critical micelle concentration (0.67e0.73%), and then further dialyzed at 4 C for ~20 h. This procedure efficiently delivered SNARE proteins from the detergent to membranes with preferred extravesicular orientation. SNARE proteins were reconstituted into vesicles containing reporter proteins using this procedure. However, the encapsulated reporter proteins disappeared during the dialysis process (scheme B). We assumed that the encapsulated reporter proteins were leaked from the vesicles during prolonged dialysis because there was 0.2e0.4% OG in the 2-fold diluted solution. To prevent leakage of the reporter proteins due to OG, which is necessary for SNARE reconstitution, reporter-containing vesicles were directly subjected to SNARE reconstitution before removing unencapsulated reporter proteins (Scheme C and D). OG-solubilized SNARE proteins were
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
5
Fig. 3. Removal of unencapsulated dyes and robustness of encapsulation. (A, B) DyStrp and bGFP not encapsulated in the vesicles can be removed by biotin-agarose and Ni-NTA beads, respectively. Approximately 20% for bGFP and 10% for DyStrp fluorescence are not removed by the beads because the indicators are inside the vesicles. They can be removed only after 0.1% TX100 treatment. (C) An additional round of bead treatment ensures removal of free dyes. The colors for bGFP (green) and DyStrp (pink) are not observed in the second bead treatment. (D) When t- and v-vesicles were mixed after removing unencapsulated dyes, FRET did not increase at 4 C. FRET efficiency increased only when TX100 was added. (E, F) Robust vesicles. Encapsulated DyStrp and bGFP are not removed by biotin-agarose or Ni-NTA beads even at 37 C for 10 h, indicating that the content-mixing reporters are not leaked from of vesicles during incubation. The reporters were maintained inside the vesicles even at a lipid-to-protein ratio of 100:1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
mixed with encapsulation solution so that the inward and outward flux of reporters were the same (Scheme C and D). Next, different concentrations of OG were treated for SNARE reconstitution (~0.3% in scheme C and 0.8% in scheme D). At low OG concentration (scheme C), the SNARE reconstitution was extremely inefficient and the protein contents were kept inside the vesicles. Scheme D enabled successful reconstitution of SNARE proteins and prevented the loss of encapsulated reporter proteins during the SNARE reconstitution step. The overall yield of reporter protein encapsulation following scheme D (Fig. 2D) was 10e20%. Unencapsulated reporter proteins were easily removed using Ni-NTA agarose and biotin-agarose in 5 min (Fig. 3AeC). When both vesicles containing bGFP and DyStrp were mixed, there was no FRET increase, indicating that the protein contents outside the vesicles were effectively removed (Fig. 3D).
When the vesicle mixture was treated with TX100, FRET between bGFP and DyStrp was increased. FRET efficiency of the bGFP-DyStrp pair measured after membrane disruption with TX100 depended on the [acceptor]/[donor] ratio consistently with the mixing of reporter proteins in solution. Leakage of reporter proteins from the vesicles containing reconstituted SNARE proteins was analyzed (Fig. 3E and F). Vesicles with different lipid-to-protein (L:P) ratios from 1000:1 to 100:1 did not leak encapsulated reporter proteins even at 37 C for 10 h. Reporter proteins bound to agarose beads only when the vesicles encapsulating the reporter proteins were disrupted by detergent, suggesting that there was virtually no leakage of protein contents from the vesicles.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
6
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
3.3. Large fusion pore is generated during SNARE-mediated membrane fusion
vesicle containing bGFP. 4. Discussion
Both t- and v-vesicles were mixed and FRET change was monitored as a function of time at 37 C (Fig. 4A). FRET between DyStrp and bGFP increased over time, indicating that fusion pores were formed through SNARE-driven fusion. When the soluble fragment of Vp2 (VpS) was added to the reaction mixture, content mixing was completely inhibited, excluding the possibility that FRET increased because of reporter protein leakage (Fig. 4B). Furthermore, up to a 4-M excess of DyStrp mixed with v-vesicle containing bGFP did not induce a FRET increase, excluding the possibility that FRET changed because of reporter protein leakage. Thus, SNARE proteins opened a fusion pore that was large enough to allow the passage of bGFP, which has a molecular weight larger than 28 kDa. Because the hydrodynamic diameter of bGFP was 5.06 nm (Fig. 1D), the fusion pore opened by SNARE proteins must be larger than 5 nm. Synaptotagmin 1 (Syt1) is a Ca2þ-sensor protein responsible for synchronous and/or fast membrane fusion. C2AB, which lacks the transmembrane domain of Syt1, facilitates multiple fusions between vesicles, although whether this occurs in the neuron or by full-length Syt1 in the reconstituted system remains unclear [4,17,22]. We made use of this feature to test whether the FRET can report stoichiometry of fusion. When Ca2þ was added to the vesicle mixture FRET efficiency was increased and reached ~0.6. Because this FRET value is achieved when DyStrp is >1.5 times above bGFP, we assume that a few t-vesicles containing DyStrp fused with a v-
Content-mixing between two merging vesicles is the final stage of membrane fusion. However, recent studies suggest that lipidmixing assays do not reflect the opening of fusion pores. While several content-mixing reporter systems based on small fluorescence dyes are currently utilized [12e15,23], they all exhibit severe leakage problems. To overcome this limitation, a hybridizing DNA pair probe was developed and utilized to analyze fusion pore opening [18]. However, the efficiency of encapsulation of DNAs into vesicle is extremely low. Encapsulation efficiency can be increased by attaching DNA to cholesterol, which can be easily incorporated into the vesicle. DNAs protruding towards the outside of the vesicle can be removed by DNase treatment, leaving the DNA inside vesicles intact. However, the reporter pair is a hybrid reporter of lipidmixing and content-mixing because one of the DNA pairs is anchored to the membrane, raising concerns regarding lipid flipflop. Although GFP variants are rich sources of FRET pairs that can be used as content-mixing reporters, these proteins have not been used for this purpose. This may be because of a low encapsulation efficiency similar to DNA reporters. Thus, one of the goals of this study was to develop an efficient method for encapsulating cargos in the vesicle during functional reconstitution of SNARE proteins into the vesicle. Cargo encapsulation and SNARE reconstitution were attempted simultaneously and sequentially (Fig. 2).
Fig. 4. Content-mixing assay of SNARE-driven membrane fusion using bGFP and DyStrp as a FRET pair. (A) Real-time trace of content-mixing assay. (B) FRET efficiency increases only by SNARE-driven fusion. Inhibition of membrane fusion by soluble domain of Vp2 (VpS) or addition of DyStrp dyes to t-vesicle did not induce FRET increase, indicating the robustness of the reconstituted system. (C, D) FRET efficiency depended on the stoichiometry of bGFP and DyStrp, indicating multiple rounds of fusion induced by Ca2þ and C2AB domain.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006
P. Heo et al. / Biochemical and Biophysical Research Communications xxx (2017) 1e7
Simultaneous encapsulation and reconstitution (scheme A) did not maintain GFP inside the vesicles, although SNAREs were reconstituted well. Both cargo encapsulation and SNARE reconstitution were successful only when the experimental procedures were as follows, in which cargo encapsulation and SNARE reconstitution were performed sequentially. First, the lipid film was hydrated in the presence of cargo. Second, the lipid solution was frozen and thawed twice to enhance multi-lamellar vesicle formation. Cargo molecules are evenly distributed outside and inside multi-lamellar structures through this treatment. However, the number of cycles should be optimized because freeze/thaw treatment can inactivate proteins. Two freeze/thaw cycles were experimentally determined to be optimal. Third, LUVs were formed by extruding the multilamellar vesicle through a polycarbonate membrane filter with 100-nm pores. Fourth, cargo molecules outside the LUVs should not be removed before the SNARE reconstitution step. When cargoes outside vesicles were removed at this time (scheme B), the final proteoliposome after SNARE reconstitution did not contain encapsulated cargos. Because the membrane was permeabilized by OG in the next step, the in-and-out flux of content-proteins was balanced only when there were equivalent concentrations of content-proteins inside and outside the vesicles. Fifth, binary tSNARE complexes or Vp2 were added to the mixture prepared above. A final OG concentration of 0.8% enabled efficient SNARE reconstitution (scheme D), while lower OG concentrations did not allow SNARE reconstitution (scheme C). Sixth, the vesicles containing cargo and SNAREs were anchored to the membrane and finally rigidified by removing OG through dialysis. Finally, cargo outside the vesicles was easily removed by treatment with biotinagarose beads and Ni-NTA-agarose for DyStrp and bGFP, respectively (Fig. 3). Vesicles formed using this procedure did not leak cargo molecules, even at a 100:1 LP ratio at 37 C for 10 h. Acknowledgements This research was supported by grants from the Korea Research Foundation (NRF-2017R1A2B2008211), and The Advanced Biomass R&D Center (ABC) of Korea (2011-0031359) funded by the Ministry of Science, ICT and Future Planning, South Korea. Appendix A. Supplementary data
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12] [13]
[14]
[15]
[16] [17]
[18]
[19]
[20]
Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.bbrc.2017.05.006.
[21]
References
[22]
[1] A.T. Brunger, K. Weninger, M. Bowen, S. Chu, Single-molecule studies of the neuronal SNARE fusion machinery, Annu. Rev. Biochem. 78 (2009) 903e928. [2] D.H. Kweon, B. Kong, Y.K. Shin, Hemifusion in synaptic vesicle cycle, Front. Mol. Neurosci. 10 (2017) 65. [3] T. Weber, B.V. Zemelman, J.A. McNew, B. Westermann, M. Gmachl, F. Parlati,
[23]
7
T.H. Sollner, J.E. Rothman, SNAREpins: minimal machinery for membrane fusion, Cell 92 (1998) 759e772. A. Stein, A. Radhakrishnan, D. Riedel, D. Fasshauer, R. Jahn, Synaptotagmin activates membrane fusion through a Ca2þ-dependent trans interaction with phospholipids, Nat. Struct. Mol. Biol. 14 (2007) 904e911. Y. Yang, J.Y. Shin, J.M. Oh, C.H. Jung, Y. Hwang, S. Kim, J.S. Kim, K.J. Yoon, J.Y. Ryu, J. Shin, J.S. Hwang, T.Y. Yoon, Y.K. Shin, D.H. Kweon, Dissection of SNARE-driven membrane fusion and neuroexocytosis by wedging small hydrophobic molecules into the SNARE zipper, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 22145e22150. P. Heo, J.B. Park, Y.K. Shin, D.H. Kweon, Visualization of SNARE-mediated hemifusion between giant unilamellar vesicles arrested by myricetin, Front. Mol. Neurosci. 10 (2017) 93. Y.H. Chan, B. van Lengerich, S.G. Boxer, Effects of linker sequences on vesicle fusion mediated by lipid-anchored DNA oligonucleotides, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 979e984. S.M. Dennison, M.E. Bowen, A.T. Brunger, B.R. Lentz, Neuronal SNAREs do not trigger fusion between synthetic membranes but do promote PEG-mediated membrane fusion, Biophys. J. 90 (2006) 1661e1675. D.L. Floyd, J.R. Ragains, J.J. Skehel, S.C. Harrison, A.M. van Oijen, Single-particle kinetics of influenza virus membrane fusion, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 15382e15387. J. Diao, P. Grob, D.J. Cipriano, M. Kyoung, Y. Zhang, S. Shah, A. Nguyen, M. Padolina, A. Srivastava, M. Vrljic, A. Shah, E. Nogales, S. Chu, A.T. Brunger, Synaptic proteins promote calcium-triggered fast transition from point contact to full fusion, Elife 1 (2012) e00109. J. Diao, M. Zhao, Y. Zhang, M. Kyoung, A.T. Brunger, Studying proteinreconstituted proteoliposome fusion with content indicators in vitro, BioEssays 35 (2013) 658e665. N. Duzgunes, J. Wilschut, Fusion assays monitoring intermixing of aqueous contents, Methods Enzymol. 220 (1993) 3e14. G. van den Bogaart, M.G. Holt, G. Bunt, D. Riedel, F.S. Wouters, R. Jahn, One SNARE complex is sufficient for membrane fusion, Nat. Struct. Mol. Biol. 17 (2010) 358e364. J. Zimmerberg, F.S. Cohen, A. Finkelstein, Fusion of phospholipid vesicles with planar phospholipid bilayer membranes. I. Discharge of vesicular contents across the planar membrane, J. Gen. Physiol. 75 (1980) 241e250. M. Kyoung, A. Srivastava, Y. Zhang, J. Diao, M. Vrljic, P. Grob, E. Nogales, S. Chu, A.T. Brunger, In vitro system capable of differentiating fast Ca2þ-triggered content mixing from lipid exchange for mechanistic studies of neurotransmitter release, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 304e313. G. van den Bogaart, R. Jahn, Inside insight to membrane fusion, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 11729e11730. P. Heo, Y. Yang, K.Y. Han, B. Kong, J.H. Shin, Y. Jung, C. Jeong, J. Shin, Y.K. Shin, T. Ha, D.H. Kweon, A chemical controller of SNARE-driven membrane fusion that primes vesicles for Ca2þ-triggered millisecond exocytosis, J. Am. Chem. Soc. 138 (2016) 4512e4521. J. Diao, Z. Su, Y. Ishitsuka, B. Lu, K.S. Lee, Y. Lai, Y.K. Shin, T. Ha, A single-vesicle content mixing assay for SNARE-mediated membrane fusion, Nat. Commun. 1 (2010) 54. J. Diao, Y. Ishitsuka, H. Lee, C. Joo, Z. Su, S. Syed, Y.K. Shin, T.Y. Yoon, T. Ha, A single vesicle-vesicle fusion assay for in vitro studies of SNAREs and accessory proteins, Nat. Protoc. 7 (2012) 921e934. Y. Lai, J. Diao, Y. Liu, Y. Ishitsuka, Z. Su, K. Schulten, T. Ha, Y.K. Shin, Fusion pore formation and expansion induced by Ca2þ and synaptotagmin 1, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 1333e1338. L. Shi, Q.T. Shen, A. Kiel, J. Wang, H.W. Wang, T.J. Melia, J.E. Rothman, F. Pincet, SNARE proteins: one to fuse and three to keep the nascent fusion pore open, Science 335 (2012) 1355e1359. M.C. Chicka, E. Hui, H. Liu, E.R. Chapman, Synaptotagmin arrests the SNARE complex before triggering fast, efficient membrane fusion in response to Ca2þ, Nat. Struct. Mol. Biol. 15 (2008) 827e835. J.M. Johnson, T. Ha, S. Chu, S.G. Boxer, Early steps of supported bilayer formation probed by single vesicle fluorescence assays, Biophys. J. 83 (2002) 3371e3379.
Please cite this article in press as: P. Heo, et al., Green fluorescence protein-based content-mixing assay of SNARE-driven membrane fusion, Biochemical and Biophysical Research Communications (2017), http://dx.doi.org/10.1016/j.bbrc.2017.05.006