SUPPLEMENTARY INFORMATION

Transcription

1 DOI: 1.138/NNANO Mixing sub-attolitre volumes in a quantitative and highly parallel manner with soft matter nanofluidics Sune M. Christensen, Pierre-Yves Bolinger,Nikos S. Hatzakis, Michael W.Mortensen and Dimitrios Stamou* Equal contribution. * Corresponding author. Supplementary Methods Surface preparation. Glass slides were cleaned using Helmanex, rinsed by copious wash/sonication cycles in water and stored in methanol. Before use the glass slides were dried under nitrogen flow and etched in oxygen plasma for 6 seconds. Immediately after plasma etching the glass slides were fixed in a microscope chamber and incubated with 1 g/l PLL(2)-g[3.5]-PEG(2) and PLL(2)-g[3.5]-PEG(2)/PEG(3.4)-Biotin (18%) (15mM HEPES, ph 5-6) at a 199:1 ratio for 3 minutes. The poly (L-lysine) grafted PEGs were purchased from Susos (Dübendorf, Switzerland). The glass slides were then rinsed in buffer to remove unbound material and hereafter immersed in.1 g/l Neutravidin for 1 minutes. Finally, the surfaces were copiously washed to remove excess protein. Vesicle immobilisation and microscopy. Target SUV reactors were immobilised at the functionalised surface by adding 5 µl of the SUV suspension to the microscope chamber (total chamber volume: 8 µl) and incubating for approximately 1 minute. Unbound material was removed by gentle wash. The microscope was then focused at the glass surface and time series acquisition was initiated upon addition of cargo SUV reactors to the microscope chamber. Typically, 5 µl of vesicles at a lipid concentration of 1 mg/l was added to the chamber. The Fluorescence measurements were performed using a Leica TCS SP5 inverted confocal microscope equipped with an oil immersion objective HCX PL APO, x1 magnification and 1.4 NA. Fluorescence emission was collected using either photo-multiplier tubes or avalanche photodiodes. The TLL experiments were performed using a Leica DMI 6 B TIRF setup equipped with an oil immersion objective HCX PL APO x1 with 1.46 NA and an Andor ixonem+ 897 camera. TLL: Purified TLL was a kind gift from Novozymes A/S, Bagsvaerd,Denmark. The purification has been described elsewhere (Svendsen et al., in Lipases, Part A, Methods in Enzymology, Vol. 284, p. 317, 1997 and Skjot et al., ChemBioChem, vol. 1, p. 52, 29.) NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

2 DOI: 1.138/NNANO Image acquisition. Lipid mixing assay: Fusion movies were acquired exiting at 458 nm and integrating DiO (48-53 nm) and DiI (59-65 nm) in two parallel channels. The microscope was operated at a scan speed of 14 Hz. The image resolution was 128x128 pixels with a bit depth of 8. The time resolution was 65 ms and typically an area of 2x2 µm 2 was imaged. Content mixing assay: The three labels of the system were imaged sequentially. The product signal (Fluorescein) was acquired with excitation at 488 nm and Fluorescence emission integrated in the interval nm. The target reactor signal (Texas red- DHPE) was acquired with excitation at 564 nm and integrating emission in the interval nm. The cargo reactor signal (DiD) was acquired with excitation at 633 nm and integrating emission in the interval nm. To reduce noise the micrographs were smoothed before analysis using the build in 3x3 gauss filter in Igor Pro version 6.12 (Wavemetrics, OR, USA). Leakage assay: Fusion movies were acquired by exciting simultaneously with the 476 nm and the 543 nm laser line. Fluorescence emission was collected in three parallel channels to monitor Alexa 488 Hydrazide ( nm), DiI (56-61 nm) and DiD (65-75 nm) emission. The microscope was operated at a scan speed of 4 Hz. The image resolution was 512x512 pixels with a bit depth of 8. The time resolution was 2.6 s and typically an area of 6x6 µm 2 was imaged. Data treatment. All data treatment was performed using customised routines written in Igor Pro version 6.12 (Wavemetrics, OR, USA). The analysis package overall consisted of two units: (i) A routine for extracting fluorescence traces of single target vesicles in parallel channels from the time series based on region of interest (ROI) integration of intensity. In a subsequent step the software cropped single events in the time-traces, performed cross talk correction and (for the lipid mixing experiments) calculated the apparent (E=I acc /(I acc +I donor )) as a function of time for each event. (ii) A routine for extracting single vesicle intensities and performing background correction and co-localisation of the different labels. To accurately determine the integrated intensity of the imaged vesicles each diffraction limited spot on the micrographs was fitted with a two-dimensional gaussian function and the single vesicle intensity was calculated as the volume under the gaussian bell. 2 NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

3 DOI: 1.138/NNANO SUPPLEMENTARY INFORMATION Quantification of vesicle size. We converted the fluorescence intensity of single immobilised vesicles into physical size by a technique recently developed in our lab (Kunding et al., Biophysical Journal, Vol. 95, p. 1176, 28 and Lohr et al., Meth. Enzymol., Vol. 465, p. 143, 29). Briefly, the size-calibration was performed by examining an extruded vesicle sample using both dynamic light scattering (DLS) and confocal fluorescence microscopy, Supplementary Methods - Figure 1. The size distribution of the vesicle population was constructed by measuring the integrated intensity of single surface immobilised vesicles. Dynamic light scattering was performed using an ALV-5 Correlator equipped with a 633 nm laser line. The vesicle size distribution was measured at 9 and the scattered intensity was measured for 1 x 3 seconds. All data were collected at 2 C. The vesicles were diluted to a final concentration of.1% (w/v) in.2 µm filtered buffer. The correlation function was translated into the intensity based size distribution using the dls 2g(t) regularised fit routine implemented in the ALV correlator software. The formfactor of vesicles with a membrane thickness of 4.5 nm was applied to all data. Using DLS the calibration sample was measured to have a mean radius of 35 nm. The distribution of single vesicle intensities collected from the micrographs of the calibration sample was then aligned to fit the mean vesicle size from the DLS measurement, yielding the conversion factor between integrated fluorescence intensity and physical size. a 3 Mean = b 25 2 Mean = 35 nm Particles sampled = Vesicle intensity ( I ) / A.U Vesicle radius / nm 3 Supplementary Methods - Figure 1: Size correlation of vesicles extruded through polycarbonate membranes with 5 nm pore diameter. (a) Distribution of integrated single vesicle intensities as measured with confocal fluorescence microscopy. (b) DLS-calibrated size distribution. NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

4 DOI: 1.138/NNANO Supplementary Information Figure 1 Premixed sample E_avg=.51 E_min= Supplementary Information Figure 1: Histogram of apparent FRET efficiencies recorded by imaging an immobilised vesicle sample premixed with 1 mol% DiO (donor) and 1% DiI (acceptor), thus simulating the product of a full fusion event. A threshold for categorisation of lipid mixing in the fusion experiments was fixed at E= NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

5 DOI: 1.138/NNANO SUPPLEMENTARY INFORMATION Supplementary Information Figure 2 a b c Intensity x1 3 A.U Donor Acc. FRET Intensity x1 3 A.U Intensity x1 3 A.U d e f Intensity x1 3 A.U Intensity x1 3 A.U Intensity x1 3 A.U Supplementary Information Figure 2: Additional lipid mixing events. The threshold in the apparent used to classify lipid mixing is indicated by the dashed lines. The time axis for each event was adjusted relatively to the onset of donor fluorescence. NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

6 DOI: 1.138/NNANO Supplementary Information Figure 3 a ± 3 % fusion Before After.8 1. c Fusion efficiency / % Ensemble average I Fluorescein /I TxR Sampled SUVs b ± 3 % fusion Before After 8 1 d Deviation from ensemble / % Deviation in fusion eff. from enesmble average I Fluorescein /I TxR Sampled SUVs Supplementary Information Figure 3: In order to count reactors with successful product formation we evaluated the distribution of Fluorescein intensities recorded on single vesicles before and after incubation with cargo reactors. The Fluorescein intensity was weighted with the intensity of the target reactor membrane label (Texas red-dhpe) to account for cross talk from Texas red (TxR) detected in the Fluorescein channel. The fluorescence of non-processed FDP and cross talk from the cargo reactor membrane label was insignificant, as determined by immobilisation of FDP loaded cargo reactors on a separate surface. We used the distribution recorded before addition of cargo reactors to assign a threshold for the classification of successful delivery events (dashed line on the graphs). a and b show results from two independent experiments. Product could be detected in respectively 88±3 % and 83±3 % of the imaged reactors. The reported error corresponds to the tail of the intensity histogram acquired before incubation situated above the threshold used for classification of successful product formation. The difference in the absolute x-values between a and b is due to differences in imaging conditions. c-d To test the confidence in a measured DE as a function of the number of sampled target reactors data subsetswith N data points were randomly chosen from the dataset in a using a random number generator (enoise function in IgorPro). For each data subset DE was evaluated and plotted versus N (c). The black line in c indicates the ensemble average. d shows the deviation in % units from the ensemble average in the measured DE as a function of N calculated from the data in c. As expected, significant variations from the ensemble average is seen for low number of containers. When more than 2 target reactors are analysed the deviation is always less than 5 % units. 6 6 NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

7 DOI: 1.138/NNANO SUPPLEMENTARY INFORMATION Supplementary Information Figure 4 c a α-hemolysin b DiI A488 A488 Integrated Time / s t= s t=29 s t=36 s t=44 s 44 1 Reactor lumen - Al488 ROI Reactor membrane - DiI Supplementary Information Figure 4: a Vesicles composed of POPC:PO(3-ethyl)PC: POPG:DOPE-biotin: DiI, 16:4:4:2:2 molar ratio (simulating the resulting lipid composition after fusion of a cargo and a target reactor) were loaded with Alexa Hydrazide 488 (A488) (4.5 mm) and immobilised. At time zero the pore forming toxin α-hemolysin was added to the chamber at a concentration of 1.5 µm and imaging was initiated. Upon insertion of α-hemolysin into the vesicle bilayer the water soluble A488 was emptied through the 1 nm wide pore. b Sample time trace of A488 intensity during insertion and resulting emptying of a single reactor. The temporal resolution was 65 ms. c Sample micrographs of three A488 loaded vesicles during the experiment. Reactors were emptied successively, indicating that each leakage event was caused by insertion of a single pore in each vesicle. Bar: 2 µm. 7 NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

8 DOI: 1.138/NNANO Supplementary Information Figure 5 Premixed sample Counts E_avg=.59 E_min= Supplementary Information Figure 5: Histogram of apparent FRET efficiencies recorded by imaging an immobilised vesicle sample premixed with 1.5 mol% DiI and DiD, thus simulating the product of a fusion event in the leakage assay. The threshold for lipid mixing was fixed to E= NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.

9 DOI: 1.138/NNANO SUPPLEMENTARY INFORMATION Supplementary Information Figure 6 a Alexa 488 DiI DiD b c d Supplementary Information Figure 6: Additional traces from the leakage assay with corresponding traces. The shaded region indicates the threshold for lipid mixing according to Supplementary Information Fig. 5. The time axis for each event was adjusted relatively to the onset of acceptor fluorescence. NATURE NANOTECHNOLOGY Macmillan Publishers Limited. All rights reserved.