Host Cell Mimic Polymersomes for Rapid Detection of Highly Pathogenic Influenza Virus via a Viral Fusion and Cell Entry Mechanism.

Highly pathogenic avian influenza virus (HPAIV) infections have occurred continuously and crossed the species barrier to humans, leading to fatalities. A polymerase chain reaction based molecular test is currently the most sensitive diagnostic tool for HPAIV; however, the results must be analyzed in centralized diagnosis systems by a trained individual. This requirement leads to delays in quarantine and isolation. To control the spread of HPAIV, rapid and accurate diagnostics suitable for field testing are needed, and the tests must facilitate a differential diagnosis between HPAIV and low pathogenic avian influenza virus (LPAIV), which undergo cleavage specifically by trypsin- or furin-like proteases, respectively. In this study, a differential avian influenza virus rapid test kit is developed and evaluated in vitro and using clinical specimens from HPAIV H5N1-infected animals. It is demonstrated that this rapid test kit provides highly sensitive and specific detection of HPAIV and LPAIV and is thus a useful field diagnostic tool for H5N1 HPAIV outbreaks and for rapid quarantine control of the disease.

Introduction

Rapid diagnosis and containment of highly pathogenic avian influenza viruses (HPAIVs), such as H5N1 and H5N8, are vital for the successful control of influenza outbreaks.1, 2, 3 A current state‐of‐the‐art point‐of‐care rapid kit with a lateral flow immunoassay is capable of rapid one‐step diagnosis, but the detection accuracy is much lower than that of the rigorous laboratory diagnosis systems based on the polymerase chain reaction (PCR) test.4, 5, 6, 7 Additionally, the one‐step kit cannot differentiate HPAIV form low pathogenic avian influenza viruses (LPAIVs). However, the intrinsically slow PCR test requires additional sequencing procedures to differentiate viral phenotypes, which may squander the time needed for quarantine and isolation. Centralized diagnosis systems and expertise are also needed to obtain results from the PCR test. In this study, we report a simple rapid diagnostic kit with an accuracy of >99% and an operation time of <30 min that can differentiate HPAIV from LPAIV and could potentially revolutionize control practices for influenza outbreaks.

Proteolytic activation of the hemagglutinin (HA) is indispensable for viral infectivity, and the responsible cellular proteases influence pathogenicity. Our rapid and sensitive diagnostic test for the differentiation of HPAIV and LPAIV is based on the fundamental difference in the membrane fusion mechanisms of HPAIV and LPAIV.8, 9, 10 HPAIV and LPAIV show a distinct structural difference in the basic cleavage site of HA, which is a surface glycoprotein of the influenza virus that plays a crucial role in receptor binding and membrane fusion.11, 12, 13 The inactive form of HA (HA0), expressed as an uncleaved homotrimer, can be cleaved by proteases to generate HA1 and HA2, thus exposing the fusion peptide of the conserved cleavage sequence of the stalk domain. This cleavage is essential for HA activation, which eventually leads to membrane fusion under low pH conditions. We recognized that LPAIV contains a monobasic cleavage site that can be cleaved by trypsin‐like serine proteases, whereas HPAIV typically possesses a polybasic cleavage site that can be activated not only by trypsin‐like serine proteases but also by furin‐like proteases. To utilize between LPAIV and HPAIV in the enzyme activities at the cleavage sites, we fabricated a cell‐mimetic polymersome (called “FluSome”) containing fluorescence resonance energy transfer (FRET) pair fluorophores (3,3′‐dioctadecyloxacarbocyanine perchlorate (DiO) and DiI) that can be fused with the virus membrane (Figure a). Upon fusion, FRET FluSome changes its fluorescence emission from orange (565 nm) to green (504 nm), which enables visual identification of HPAIV or LPAIV.

Figure 1

Rapid and sensitive diagnostic test for differentiation between HPAIV and LPAIV. a) A cell‐mimetic polymersome (FRET FluSome) containing FRET pair fluorophores (DiO and DiI) that can be fused with the virus membrane. b) Structural arrangement of HA trimers in the prefusion state and the structure of the protease‐treated HA fusion domain. c) CD spectra of the influenza HA domain treated with trypsin (solid line) or furin (dashed line) for LPAIV (black) and HPAIV (red). The spectra were recorded at 37 ± 2 °C.

Results and Discussion

Differences in Virus Membrane Fusion to the Host Cell Membrane Depending on the Enzyme Activity of Furin and Trypsin

Trimeric HA consists of HA1 and HA2 subunits linked by a single disulfide bond that can be cleaved by the activity of furin or trypsin.14, 15 HA1 binds to terminal sialic acid receptors expressed on the host cell surface, and the virus is then internalized by the host cell. The HA protein undergoes an irreversible conformational change in the acidic conditions of the endosome, followed by a loop‐to‐helix transition of HA2, which contains the fusion peptide at the N terminus. HA2 pulls the viral envelope and endosomal membrane together via energy conversion of HA from a metastable high‐energy prefusion state to a low‐energy membrane fusion state, which mediates the release of viral ribonucleoprotein particles into the cytoplasm (Figure 1b).

To examine the cleavage‐activation and membrane fusion properties of HA proteins of each viral subtype, we prepared influenza A viruses that were representative of HPAIVs and LPAIVs (Table ). HPAIV can be defined as having multiple basic amino acids at the HA cleavage site, whereas LPAIV has mono basic amino acids. In HPAIV, the number of basic amino acids (arginine and lysine) can be increased by specific mutations or insertions of amino acids at the cleavage site.16 The HA of HPAIV can be cleaved by ubiquitous furin‐like endoproteases, which increases the ability of the virus to be infective in extra intestinal and extra respiratory tissues and promotes systemic viral replication.17 By contrast, LPAIV has a restricted tissue tropism because its HA can only be cleaved by trypsin‐like proteases, limiting viral replication to the respiratory system.

Table 1

Origin of the examined HA and NA proteins and HA cleavage site sequences

Pathogenicity	Subtype	HA and NA donor	HA Cleavage site sequence
			P4	P3	P2	P1	P1′	P2′	P3′	P4′
HPAIV (Highly Pathogenic Avian Influenza Virus)	H5N1	A/chicken/VN/KienGiang/P140082/2014	R	R	K	R	G	L	F	G
	H5N1	A/duck/VN/QuangTri/P140164/2014	R	K	K	R	G	L	F	G
	H5N6	A/chicken/VN/LangSon/P140450/2014	R	R	K	R	G	L	F	G
LPAIV (Low Pathogenic Avian Influenza Virus)	H1N1	A/California/04/2009	I	Q	S	R	G	L	F	G
	H2N1	A/wild bird feces/Korea/KRIBB‐KU/2014	I	E	S	R	G	L	F	G
	H2N4	A/wild bird feces/Korea/KRIBB‐KU/2014	I	E	S	R	G	L	F	G
	H3N2	A/canine/Korea/01/2007	R	Q	T	R	G	L	F	G
	H3N8	A/wild bird feces/Korea/KRIBB‐KU/2014	K	Q	T	R	G	L	F	G
	H5N2	A/wild bird feces/Korea/KRIBB‐KU/2014	R	E	T	R	G	L	F	G
	H5N3	A/wild bird feces/Korea/KRIBB‐KU/2014	R	E	T	R	G	L	F	G
	H9N2	A/chicken/Korea/S1/2003	A	S	G	R	G	L	F	G

We were intrigued by the different enzyme activities of LPAIVs and HPAIVs and therefore investigated the effect of enzyme interactions on the secondary structure of the fusion peptides of HPAIV and LPAIV by using circular dichroism (CD) spectroscopy. The fusion peptide of influenza HA adopts an α‐helical conformation in the bilayer of FluSome and folds as a random coil in solution. HPAIV fusion peptides that had α‐helices were observed in the presence of furin or trypsin, while the α‐helix content of LPAIV was observed only in the addition of trypsin (Figure 1c). Thus, we determined the structures of the perdeuterated fusion peptides of HPAIV and LPAIV in the presence of furin or trypsin at pH 5.0 using 1H‐NMR spectroscopy.

Mechanism of FluSome

FluSome was prepared with a thin‐film hydration method as previously described.18, 19 mPEG‐b‐pLeu copolymers with hydrophilic and hydrophobic segments were assembled into membranes (see Supporting Information for experimental details). Methoxy‐poly (ethylene glycol) (mPEG) has been chosen as the hydrophilic moiety here because of its physicochemical characteristics. And, pLeu, well‐known as a neutral hydrophobic polypeptide, improve the solubility. The TEM images of the formulated FluSomes typically showed circular rings consisting of entangled and interdigitated hydrophobic shells with a hydrophilic interior. The mean size was 112.49 nm, as estimated by laser scattering. The core/shell structured FluSomes could efficiently load FRET pair dyes (DiO and DiI) in the shell of pLeu through a hydrophobic interaction (Figure a).

Figure 2

Study of the stable incorporation of FRET pair dyes (DiO and DiI) in FluSome using a FRET assay. a) TEM image of FluSomes. b) Fluorescence spectra of furin, trypsin, and FluSome. c) Fluorescence spectra of FluSome treated with furin, trypsin, pH 7.4 solution, pH 5.5 solution, and Triton X‐100. The red line shows the spectrum of the FRET system after the FluSome was disrupted with Triton X‐100. d) Evolution of IA/I0 + IA over time for c). e) Kinetics of DIO release from FluSome treated with furin, trypsin, pH 7.4 solution, pH 5.0 solution, and Triton X‐100. f) Kinetics of DiI release from FluSome treated with furin, trypsin, pH 7.4 solution, pH 5.0 solution, and Triton X‐100. g) Time‐resolved intensity curves of the FRET pair dyes (DiI and DiO, hydrophobic) and calcein‐encapsulated FluSome (hydrophilic). Docking of a FluSome, a hemifusion event, and pore formation in the fusion state. The dotted vertical lines separate the docking time (tdock), hemifusion time (themi), and pore formation time (tpore). Right: schematics of the events at the end of the time series.

We measured the fluorescence of furin, trypsin, and FRET FluSome. Furin and trypsin had minimal emission, whereas FRET FluSome emitted fluorescence at 565 nm due to the occurrence of FRET (Figure 2b). DiO has an emission maximum at 504 nm for 484 nm excitation. Therefore, 484 nm excitation results in FRET between the DiO and DiI molecules. To investigate the stability of dye‐incorporated FluSome, we measured the fluorescence of FRET FluSome + furin at pH 5.0, FRET FluSome + trypsin at pH 5.0, FRET FluSome at pH 7.4, and FRET FluSome at pH 5.0 (Figure 2c). For a stable FluSome structure, total quenching of DiO fluorescence and a strong emission with a maximum at 565 nm were observed. The FRET of FluSome was maintained in the presence of furin at pH 5.0, trypsin at pH 5.0, pH 5.0, and pH 7.4. However, the addition of Triton X‐100 to this FRET system resulted in complete disruption of FluSome, which switched off the FRET.

Figure 2d shows that the acceptor/donor maximum intensity ratio was essentially invariable between FRET FluSome + furin at pH 5.0, FRET FluSome + trypsin at pH 5.0, FRET FluSome at pH 7.4, and FRET FluSome at pH 5.0 and remained similar to the initial ratio for FRET FluSome. The success of FluSome‐based virus detection critically depends on the release rate of encapsulated FRET dyes (DiO and DiI) from FluSome. To evaluate the FRET dye release profiles, temporal FRET intensity changes were examined. As shown in Figure 2e,f, no significant release of the FRET dyes was observed for FluSome, which was incubated for 30 min. These results confirmed that FluSome was stable for at least 30 min, which enabled FRET FluSome to be used to detect HPAIV and LPAIV.

Fusion with Influenza Virus

The fusion between FRET FluSome and activated influenza virus was monitored continuously via a FRET‐based fluorescence assay using fluorescence microscopy.20 The time‐dependent fluorescence intensity changes of DiO (red) and calcein (green) upon the docking and fusing of the activated influenza virus to the membrane of FRET FluSome were monitored around the docked vesicle. The fluorescence intensities were integrated as a function of time. The obtained time traces of the two fluorophores, including all characteristic features of the FRET FluSome event, are depicted in Figure 2g. Influenza virus binds to host cell through sialic acid in natural environment. In this FluSome assay, however, fusion of FluSome and the influenza virus occurs by the fusion peptide within the virus, not by sialic acid.21 Prior to FRET FluSome docking, low and insignificant fluorescence was observed for DiO and calcein. When FRET FluSome approached the surface and docked, a sudden increase in DiO fluorescence occurred, whereas the calcein fluorescence remained unaffected. At the onset of fusion, the calcein intensity transiently increased in conjunction with a small transient decrease in the DiO fluorescence intensity. These changes are caused by the FRET that occurs between the DiO and calcein dyes when the outer leaflet of FRET FluSome starts mixing with the outer leaflet of the pore‐spanning membrane (DiO). After this first transient change in DiO and calcein fluorescence, the fluorescence intensity remained at the same level. These fluorescence time traces can be attributed to an intermediate hemifused state. In the hemifused state, the outer leaflets merged, whereas the inner leaflet of the calcein‐modified vesicle remained unaffected. Therefore, after the first merging of the outer leaflets, the DiO fluorescence intensity decreased to an intermediate intensity level as the DiO located in the outer leaflet of the vesicle diffused into the pore‐spanning membrane. Simultaneously, calcein diffused into the outer leaflet of the vesicle and moved away from the porous surface, which resulted in an increase in the calcein fluorescence intensity (Figure 2g).

The disparity in the enzyme activity at the cleavage sites of LPAIV and HPAIV was used to evaluate nasal swab specimens and stool specimens, which were called positive and negative samples, respectively. We examined the fluorescence intensity of FRET FluSome at 504 nm upon the addition of the HPAIV or LPAIV from a nasal swab in the presence of furin or trypsin at pH 5.0 (Figure a) In the absence of activated influenza virus, little variation in FRET efficiency was observed, which demonstrated the stability of FluSome. When FRET FluSome was mixed with activated influenza virus, a rapid increase in DiO fluorescence intensity was observed, as shown in Figure 3b, which indicates the fast release of DiO from FRET FluSome. We tested whether FRET FluSome can differentiate HPAIV from LPAIV. DiO emission appeared when HPAIV was treated with furin or with trypsin, whereas it was only observed for LPAIV with the trypsin treatment. We further incubated HPAIV and LPAIV with furin at pH 5.0, trypsin at pH 5.0, furin at pH 7.4, trypsin at pH 7.4, furin + trypsin at pH 5.0, furin + trypsin at pH 7.4, furin + fecal sample at pH 5.0, trypsin at pH 7.4, fecal sample at pH 5.0, and fecal sample at pH 7.4 in 96‐well plates, and we determined the fluorescence intensity 30 min postactivation. As reflected by the measured photon flux, FRET FluSome could be completely disrupted by activated influenza virus (Figure 3c). These results demonstrated the potential of using FRET FluSome to detect HPAIV and LPAIV.

Figure 3

FluSome detection of HPAIV and LPAIV in nasal swab specimens and stool specimens, which are labeled as positive and negative samples, respectively. a) Fluorescence spectra of FluSome treated with HPAIV H5N1 and H5N6, and LPAIV H1N1, H2N1, H2N4, H2N4, H3N2, H3N8, H5N2, H5N3, H9N2, Control, and Blank. Lines: Furin at pH 5.0 (orange), furin at pH 7.4 (blue), trypsin at pH 5.0 (green), and trypsin at pH 7.4 (purple). b) Fluorescence intensity based on a). c) Fluorescence images of FluSome treated with HPAIV and LPAIV by IVIS. * indicates shed virus from experimentally infected animal by each virus.

To optimize the fusion efficiency of FRET FluSome with activated influenza virus, we performed experiments at various temperatures, FRET FluSome concentrations, and pH values. The fusion of activated influenza virus with FRET FluSome was measured by the degree of fluorescence dequenching of DiO. The fluorescence intensity after the addition of Triton X‐100 was used to define 100% fusion. FRET FluSome had a mean apparent fusion efficiency of 15% with total FRET FluSome. We first performed experiments to determine the fusion efficiency of i) HPAIV + furin with the addition of FluSome, ii) LPAIV + furin with the addition of FluSome, iii) HPAIV + trypsin with the addition of FluSome, and iv) LPAIV + trypsin with the addition of FluSome at different temperatures and at pH 5.0 and 7.4 (Figure a). The fusion efficiency at 37 °C was nearly twofold higher than at 25 °C. The fusion efficiency at 42 °C was not significantly different from the efficiency at 37 °C (Figure 4a,i,iii,iv). There was no significant fusion activity for LPAIV + furin with the addition of FluSome at pH 5.0 and 7.4 at any of the tested temperatures (Figure 4a,ii). We next performed experiments on the fusion efficiency of i) HPAIV + furin with the addition of FluSome, ii) LPAIV + furin with the addition of FluSome, iii) HPAIV + trypsin with the addition of FluSome, and iv) LPAIV + trypsin with the addition of FluSome for different concentrations of FluSome at pH 5.0 and 7.4 ( 4b). The results indicate that both concentrations of FRET FluSome resulted in increasing fusion efficiency until 10 mg mL−1. Concentrations greater than 10 mg mL−1 did not increase the fusion efficiency. Therefore, the optimum concentration of FRET FluSome was 10 mg mL−1.

Figure 4

Fusion efficiency of FluSome. Fusion efficiency of FluSome at various a) temperatures and b) concentrations of FluSome. i) HPAIV + furin with the addition of FluSome, ii) LPAIV + furin with the addition of FluSome, iii) HPAIV + trypsin with the addition of FluSome, and iv) LPAIV + trypsin with the addition of FluSome at different temperatures and at pH 5.0 and 7.4.

To compare the qPCR, rapid kit, and FluSome assays, we tested nasal swab and stool samples which were called positive and negative samples for influenza viruses, respectively. Out of the 51 specimens, 40 were influenza virus positive and 11 were negative by the M gene real‐time PCR (RT‐PCR), whereas 37 were positive and 14 were negative by FluSome, which remarks the relative sensitivity and specificity of the FluSome compared with those of the RT‐PCR were 92.5% and 100%, respectively (Table ). Furthermore, we determined their detection limits by serially diluting the HPAIV and LPAIV, and FluSome presented response with tenfold higher diluted each pathogenic types of virus. The results indicated that the FluSome assay is more sensitive than the rapid kit for three HPAIVs (H5N1, H5N1, and H5N6) and five LPAIVs (H1N1, H3N2, H5N2, H5N3, and H9N2) (Table S1, Supporting Information). In addition, we confirmed a lack of cross‐reactivity in the FluSome assay when testing avian infectious bronchitis (IBV), canine distemper virus (CDV), infectious bursal disease virus (IBDV), and Newcastle disease virus (NDV) virus (Table S2, Supporting Information). This study is based on the fusion with endosome of host cell. Because FluSome shows response to fusion peptide of activated influenza virus, it provided high specificity to influenza virus. We conducted experiments at the optimal conditions for the activation of the influenza virus. Therefore, other noninfluenza viruses did not show any FRET‐ON reactivity with FluSome.

Table 2

Comparison of Rapid‐kit, qPCR, and FluSome assays for detecting viruses from nasal swab and stool specimens

Sample No.	Specimen type	Subtype	Cq of qPCR	Rapid kit	FluSome assay
Pig #1		Seasonal H3N2	27.16	+	+
Pig #2		Pandemic H1N1	25.86	+	+
Dog #1		A/canine/Korea/01/2007(H3N2)	20.48	+	+
Dog #2		A/canine/Korea/01/2007(H3N2)	20.74	+	+
Dog #3	Nasal swab	A/brisbane/10/2007(H3N2)	34.13	−	+
Dog #4		A/california/04/2009(H1N1)	30.43	+	+
Dog #5		A/chicken/VN/KienGiang/P140082/2014(H5N1)	29.12	+	+
Dog #6		A/duck/VN/QuangTri/P140164/2014(H5N1)	27.60	+	+
Dog #7		A/chicken/VN/LangSon/P140450/2014(H5N6)	30.00	+	+
0401	Stool	H1	34.57	−	+
0405		H10	33.96	−	+
0427		H10	35.73	−	+
0445		H2	29.68	−	+
0582		H6	28.41	−	+
0675		H6	29.57	−	+
0685		H6	27.31	−	+
0696		H9	28.9	−	+
1108		H6	26.26	−	+
1120		H6	32.46	−	+
1133		H11	29.32	−	+
1138		H11	31.94	−	+
1178		H6	29.17	−	+
1194		H1	28.71	−	+
1204		H1	26.05	−	+
1233		H7	33.09	−	+
1234		H11	32.24	−	+
1235		H1	28.16	−	+
1323		H7	34.57	−	+
1333		H1	31.75	−	+
1335		H5	28.66	−	+
1336		H5	29.69	−	+
1391		H11	26.6	−	+
1458		H1	32.78	−	+
1462		H3	33.49	−	+
1466		H2	27.36	−	+
1497		H4	29.73	−	+
1537		H5	26.31	−	+
0436		H4	33.3	−	−
1176		H1	35.07	−	−
1177		H1	28.11	−	−
1341			37.51 (Negative)	−	−
1415			37.96 (Negative)	−	−
1474			37.65 (Negative)	−	−
1482			37.13 (Negative)	−	−
1490			36.31 (Negative)	−	−
1520			36.12 (Negative)	−	−
1528			38.92 (Negative)	−	−
1536			38.41 (Negative)	−	−
1174			42.27 (Negative)	−	−
1327			38.65 (Negative)	−	−
1330			37.32 (Negative)	−	−

Conclusion

This study is the first to show distinct FluSome‐based diagnoses between HPAIV and LPAIV that underwent membrane fusion by acidification. To evaluate the possibility of interference from other avian pathogenic viruses that use receptor‐mediated membrane fusion, we assessed the cross‐reactivity of FluSome by applying it to IBV, IBDV, NDV, and parainfluenza virus. In addition, we checked viruses with nonavian origins, including alpha corona viruses (porcine epidemic diarrhea virus, feline coronavirus, and canine coronavirus) and morbillivirus (CDV; Table S2, Supporting Information) However, these noninfluenza viruses did not elicit FRET‐ON reactivity, which showed that FluSome has the needed target specificity for avian flu virus targets.

FluSome recognizes the PBCS cleaved by furin‐like proteases, and docking occurred between FluSome and the fusion peptide that had α‐helical content in the activated influenza HA. This docking allowed for rapid distinction between HPAIV and LPAIV. As mentioned above, HPAIV can be activated both trypsin‐like serine proteases and furin‐like proteases. However, activated LPAIV only needs trypsin‐like serine proteases. Regardless of whether the virus is activated, we can identify the virus type by the presence of α‐helices, which were sufficient for crossing cell membranes and delivering the virus to the cytoplasm of cells.22, 23 We fabricated a diagnostic tool called FluSome that included a cell‐mimetic structure to mimic the α‐helical conformation of the activated fusion peptide to monitor the fusion of activated influenza viruses. This tool is a single‐step detection test using FluSome and direct visualization. Therefore, FluSome is a novel and time‐saving diagnostic tool improving the sensitivity comparable to RT‐PCR, and this facilitates early preventative procedures for controlling HPAIV outbreaks, which currently depends on time‐consuming RT‐PCR and genomic sequencing to confirm HPAIV.

Experimental Section

Materials: mPEG with a molecular weight of 2000 Da was purchased from Fluka. d,l‐leucine (dl‐Leu) was obtained from Sigma‐Aldrich. Influenza A H5N1 (A/chicken/Vietnam/NCVD‐016/2008) and H1N1 (A/California/04/2009) HAs were obtained from Sino Biological Inc. Furin and trypsin were purchased from Merck Millipore. 1,1′‐Dioctadecyl‐3,3,3′,3′‐tetramethylindocarbocyanine perchlorate (DiI) and DiO were purchased from Invitrogen. All other chemicals and reagents were analytical grade.

Block Co‐Polymer‐Peptide Synthesis and Preparation of DiO and DiI‐Loaded Polymersomes (FRET FluSome): A series of mPEG‐b‐pLeu variants with different mass fractions of mPEG (0.25, 0.30, 0.35, and 0.40) was synthesized by ring‐opening polymerization of Leu‐NCA with amine‐terminated mPEG as a macroinitiator. mPEG (0.5 mmol) was introduced to a 250 mL three‐necked flask along with 30 mL of anhydrous N,N‐dimethylformamide (DMF). Various concentrations of Leu‐NCA (10.5, 13.0, 16.3, and 20.0 mmol) were then injected into the flask, and the reaction mixture was heated with reflux at 40 °C for 24 h under a nitrogen atmosphere. After the reaction was completed, the reaction mixture was precipitated in excess cold diethyl ether to produce mPEG‐b‐pLeu. The copolymer products were dried at room temperature for 1 d and stored under vacuum for later use.

To prepare the FRET FluSome (fmPEG = 0.30), 10 mg of mPEG‐b‐pLeu copolymer was dissolved in 1 mL of chloroform. The chloroform was then removed using a rotary vacuum evaporator to form a thin film of the mPEG‐b‐pLeu copolymer on the round‐bottom flask, which was placed under high vacuum for an additional 12 h to remove any residual chloroform. The dried film was then directly hydrated in a solution of 2 mL of Dulbecco's phosphate‐buffer saline (DPBS) for 6 h at 37 °C and then magnetically stirred for an additional 6 h at room temperature to produce FluSomes. The DiO and DiI dissolved in DMSO were then mixed with 2.0 mL of FluSome solution under vigorous stirring for 4 h. Finally, the mixture was dialyzed against DPBS using a system with a molecular weight cut‐off of 1000 Da (Tube‐ODIALYZERTM, G‐Biosciences, USA) for 24 h to produce the FRET FluSomes.

CD Spectroscopy: CD spectroscopy analysis was performed to study the secondary structure of the fusion peptides. CD spectra were acquired on a Jasco J‐815 spectropolarimeter (Jasco, Tokyo, Japan) using a 1 nm bandwidth with a 1 nm step resolution from 200 to 240 nm at room temperature. Spectra were corrected by subtraction of the respective solvent signal. The α‐helical content was estimated from the ellipticity value at 222 nm.

Fusion Assay: To monitor the fusion events of FRET FluSome with influenza virus, the fluorescence intensity was recorded at a time resolution of approximately 120 ms/frame over a period of 300 s after the FRET FluSome fusion events, and the threshold was subtracted from the green and red channel intensities to locate the FRET FluSomes that were diffusing near the FRET FluSome membrane.

FRET FluSome and Influenza Virus Fusion Detection: Fusion of FRET FluSome and the influenza virus was monitored as fluorescence dequenching at excitation and emission wavelengths of 484 nm using a hybrid multimode microplate reader (SynergyTM H4, BioTek, USA). Experiments involving a 30 min preincubation of HPAIV and LPAIV at pH 5.0 with furin or trypsin were performed.

FluSome Application to Animals: In total, stool samples were collected from a stopover site of wild migratory aquatic birds in South Korea and mixed with DPBS by vortexing and then centrifuged for 5 min at 2000 g of which supernatant was stored at −80 °C until use. Seven dogs, 7 weeks old, were used for an experimental infection with HPAIV [A/chicken/VN/KienGiang/P140082/2014 (H5N1), A/duck/VN/QuangTri/P140164/2014 (H5N1), A/chicken/VN/LangSon/P140450/2014 (H5N6)] and LPAIV [A/California/04/2009 (H1N1), A/Brisbane/10/2007 (H3N2) and A/canine/Korea/01/2007 (H3N2)] by nasal inoculation with 500 µL of 106.5 EID50/mL. Two pigs, 6 weeks old, were inoculated by A/California/04/2009 (H1N1) and A/Brisbane/10/2007 (H3N2) with 500 µL of 106.5 EID50/mL. Nasal swabs were collected at 3 d postinfection, and viral loads of the swabs were quantified by RT‐PCR using a commercial one‐step RT‐PCR kit according to the instruction of the manufacturer as described previously.24 The log EID50/mL was calculated from the RT‐PCR results, using the regression curve of Ct values in serially diluted viruses. General animal care was provided as required by the Institutional Animal Care and Use Committee of the National Center for Veterinary Diagnosis in Vietnam.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary

Click here for additional data file.