Haemagglutination tests (HA tests) were performed in U-bottom 96-well microtiter plates in 2-fold or 10-fold serial dilutions of the viral stocks from X-31 or Udorn strains. Viral dilutions were done in PBS and no virus was added to the negative controls. To perform the test, 50 μl of a previously prepared red blood cell (RBC) suspension was added to every well and plates were incubated for one hour at room temperature (RT) before the assay results were checked. The RBC suspension was prepared from sheep whole blood cells in Alsever’s solution (Håtunalab, Bro, Sweden) in the following manner. To separate the different cell populations and plasma from the blood sample, 4 mL of the whole blood cells were added to 8 mL PBS and centrifuged at 800 x g for 10 minutes at RT. The supernatant (corresponding to plasma) and the buffy layer (where leukocytes and platelets are found), were discarded and the remaining layer containing the RBCs was washed by adding PBS and centrifuging at 800 x g for 10 minutes at RT. This washing step was repeated three times. Finally, the RBC suspension was diluted to make a 1% RBC suspension and it was tested for agglutination at RT for one hour. The 125X dilution was the chosen one to perform the HA tests.
Protein gel electrophoresis was performed in denaturing conditions using SDS-PAGE (sodium dodecyl sulphate – polyacrylamide gel electrophoresis) for protein separation. Commercial ready-made gels were used for this (NuPAGE™ 4-12% Bis-Tris Gel; Invitrogen, Carlsbad, CA). After protein separation, the proteins were transferred to a nitrocellulose blotting membrane (Amersham™ Protran™ 0.45 μm NC; GE Healthcare Life Sciences) using Trans-Blot Turbo Transfer System (Bio-Rad Laboratories, Inc., Hercules, CA). When the transfer was completed, the membranes were incubated in 3% skimmed milk blocking solution for one hour prior antibody treatment. Rabbit α-HA (CHA1) was used as a primary antibody (1:1000 dilution) and a goat α-rabbit IgG HRP-conjugated was utilized as the secondary antibody (1:10,000 dilution). After each antibody treatment, the membranes were washed with PBST (PBS with 0.1% Tween-20) while shaking for 15 minutes. The washing steps were repeated three times. Once the membranes were incubated with both antibodies and rinsed, they were treated with 500 μl of a 1:1 mixed solution of luminol and peroxide buffer. These reagents were used as part of the Immun-Star™ HRP Chemisluminiscent Kit (Bio-Rad Laboratories, Inc., Hercules, CA) to trigger the luminescence reaction. Western blots were visualized with Image Lab software using ChemiDoc™ MP Imaging System (Bio-Rad Laboratories, Inc., Hercules, CA). Thermo Scientific™ PageRuler™ Plus Prestained Protein Ladder (10 to 250 kDa) was the reference molecular weight marker used for immunoblotting experiments.
Fluorescent Viral Labelling with Lipophilic Dyes
Influenza A X-31 virions were labelled using TR-DHPE at self-quenching concentration, whilst Udorn virions were labelled with R18 or TR-DHPE in the following manner. The dye solution was prepared by mixing Texas Red (TR) or R18 at 0.75 mg/mL and 1.25 mg/mL respectively in ethanol with HB buffer at a ratio of 1:40. For Udorn viral particles different ratios were tested (1:40, 1:20, 3:40 and 3:20). 36 μl of the dye solution was mixed with 9 μl of viral stock (a Udorn-labelled batch was also made adding 18 μl of the viral stock instead) and incubated at 600 rpm for two hours at RT. 1.3 mL of HB buffer was added after incubation and tubes were centrifuged at 21,000 x g for 50 minutes at 4°C to pellet the labelled virus. Pellet resuspension was done in 10 μl HB buffer and the tubes were left on ice at 4°C overnight.
Liposome Suspension Preparation
The liposome suspension used in the experiments was prepared by extruding a lipid mixture (66.5% POPC, 20% DOPE, 10% cholesterol, 2% GD1a, 1% DPPE-biotin and 0.5% OG-DHPE) through a nitrocellulose membrane (pore size = 100 nm) an uneven number of times. The extruded liposomes were then stored at 4°C until further use.
Single Virus Lipid Mixing Assay
Single virus lipid mixing assays were performed in polydimethylsiloxane (PDMS) microfluidic flow cells (Fig. 2B) that were built up as previously described26. Briefly, PDMS was prepared by mixing silicone elastomer base and curing agent (SYLGARD™ 184 Silicone Elastomer Kit, Ellsworth Adhesives, Germantown, WI) in a 1:10 ratio. The mix was vacuumed for 45 minutes to remove air bubbles. Flow cell molds were filled in with PDMS and incubated at 70°C for two hours until solidified. Once solidified, flow cells were cut (each flow cell contains two channels), inlet and outlet holes were made and the PDMS surface was attached to clean coverslips by 25 seconds glow discharge at 20 mA (PELCO easiGlow™, TED Pella Inc.). Coating of the channels’ surface with polylysin (PLL), polyethylene glycol (PEG) and biotin was done by pipetting 5 μl/channel of previously mixed PLL-PEG-biotin and PLL-PEG in a 1:20 ratio and incubating the flow cells at RT for two hours. After incubation, the channels were rinsed with 1 mL deionized water and 1 mL Rxn buffer prior functionalization with avidin. Flow cells were incubated at RT for 40 minutes and then washed with 2 mL Rxn buffer. 10 μl of fluorescently labelled liposome suspension was added into each channel, recirculated several times and let overnight at 4°C. A buffer well was made in the inlet hole by cutting the end of a 1000 μL plastic pipette tip and gluing it with five-minute epoxy. The channels were washed again with 2 mL Rxn buffer prior addition of the labelled virions (5 μl/channel), which were recirculated in the channel to favour receptor binding. Labelled X-31 viral particles were diluted five times in Rxn buffer before they were added to the channels. To measure fusion events, each channel was connected to a suction system that enabled buffer exchange. Fusion was triggered by infusing fusion buffer (pH 5.0) and single virus-liposome lipid mixing was monitored by fluorescence microscopy.
Antibody Treatment in Flow Cells
For the antibody treatment, flow cells were prepared as mentioned before, except the liposome suspension used was not fluorescently labelled (Fig. 3). For the primary antibody treatment, 10 μl of a 50X diluted solution of mouse α-H3 HA (HC3) in Rxn buffer was infused in each channel and flow cells were incubated at RT for two hours. Subsequently, 2 mL Rxn buffer was used to wash the channels and the secondary antibody was added. 5 μl of a 1000X diluted solution of α-mouse IgG FITC in PBS was used. Incubation of the flow cells was done at RT for 1.5 hours and flow cells were rinsed with 2 mL Rxn buffer.
Single Virus-Liposome Lipid Mixing Assays with Spherical Virions
Hemifusion between the viral membrane of X-31 virions and the liposomal membrane was measured by monitoring lipid mixing in microfluidic devices. Viral particles were labelled with a lipophilic dye at a self-quenching concentration so that when fusion is triggered by lowering the pH, the fluorescent dye dilutes into the target lipid membrane and fluorescence increases consequently. By measuring the fluorescent signal during the process, we are able to determine lipid mixing corresponding to membrane hemifusion and characterize the kinetics underlying membrane fusion events. To do this, fluorescent liposomes were immobilized in microfluidic flow cells via biotin-avidin interactions and labelled viruses were added subsequently. The HA viral glycoprotein recognises GD1a molecules present in the liposomes’ membrane and will bind to these. Particles will not fuse until fusion is triggered by a pH drop. This was accomplished by performing a buffer exchange (from pH 7.4 to pH 5.0) within the flow cells. Waiting times, i.e., the lapse between the pH drop and a sudden increase in fluorescence which reports the lipid mixing, were plotted as a cumulative distribution function (CDF) (Fig. 4) after analyzing the recorded videos. Most fusion events occur within 50 seconds after the pH drop. A minor part of the events that were observed took place at later times. The efficiency percentage of the process was ~17%, which was calculated as the percentage of viral particles that fused relative to the total number of viral particles detected in the field.
Single Virus-Liposome Lipid Mixing Assays with Filamentous Virions
The same strategy was followed to measure fusion events with filamentous virions, using in this case, the Udorn strain. When using the same viral labelling conditions as used for the X-31 virions (dye concentration of 10 μM), no fusion events could be observed when performing the lipid mixing assays. Thus, some modifications were made regarding the viral labelling step in order to accomplish a self-quenching concentration of the dye used. Despite different dye concentrations (20 μM, 30 μM and 60 μM) being tested, none of them yielded fusion when performing the fluorescence dequenching assays using TR. Thus, R18 was used as an alternative fluorescent dye to TR and several dye concentrations (33 μM, and 95 μM) were tested again resulting in no observed fusion events for any of them. The next step that was performed was to increase the amount of virus used for labelling. The previous trials along with the viral labelling of the X-31 strain were made using a virus:dye ratio of 1:5. We decided to double the amount of virus used maintaining constant the quantity of dye solution, meaning that a ratio of 1:3 was tested instead. This time, a single fusion event could be measured when monitoring lipid mixing in one of the flow cells using R18 at a concentration of 41 μM. The fusion event was clearly seen by eye when watching the recording (Fig. 5A) and it was later confirmed by video software analysis. Fluorescence intensity of each particle is measured as a function of time during the analysis. Hence, monitoring the sudden changes on the intensity signal enables the determination of the timepoints when a fusion event occurred (Fig. 5B), in this case approximately 13 s after the pH drop. The software produces individual plots of fluorescence intensity against time for each particle and classifies them into particles that fused, particles that did not fuse and particles whose increase in fluorescence intensity happened too slow for it to be considered a fusion event. To check how well the coating of the flow cells and the liposome immobilization was, snapshots of the fluorescently labelled liposomes were taken prior fusion triggering in all of the experiments performed. As shown in Figure 5C with two snapshot examples, the flow cells were nicely coated with liposomes in all experiments that were made, both with X-31 and Udorn virions. Barely any spaces were found throughout the entire channel indicating a successful coating of the microfluidic flow cells.
Table of contents :
Materials and methods
Cell culture and infection
Fluorescent viral labelling with lipophilic dyes
Liposome suspension preparation
Single virus lipid mixing assay
Antibody treatment in flow cells
Single virus-liposome lipid mixing assays with spherical virions
Single virus-liposome lipid mixing assays with filamentous virions
Antibody labelling of viral particles in flow cells
Haemagglutination test results
Viral propagation of Udorn virus in MDCK-2 cells