Immunology and Cell Biology {\992) 70, 181-191
Western blot analysis of antibody responses to influenza virion proteins DIWEN QiU, GREGORY A. TANNOCK, RICHARD D. BARRY and *DAVID C.JACKSON Faculty of Medicine, The University of Newcastle, Callaghan, New South Wales, and *Department ofMicrobiology, The University of Melbourne, Melbourne, Victoria, Australia
Summary An immunoblotting procedure was developed to detect antibody responses in mice and humans to influenza virion proteins. The technique was capable of detecting 1.5 jig of haemagglutinin (HA) on nitrocellulose strips at a 1 : 5000 dilution of a mouse serum with an initial haemagglutination inhibition titre of 20. The effects of the use of the blocking agent Tween-20 on virion proteins were also studied. The commonly used concentration of 0.05% (v/v) Tween-20, when included in blocking and incubation buffers, greatly reduced the amount of detectable matrix protein but caused no detectable loss of HA and neuraminidase/nucleoprotein proteins. If virion proteins were separated by polyacrylamide gel electrophoresis under reducing conditions, antibody bound to HA2 more strongly than to HAl. Under non-reducing conditions, more antibody bound to the uncleaved HA protein than to other proteins. IgGl and IgG2a antibody responses in mice to each protein were stronger than IgG2b and IgG3 responses. Introduction The influenza virus comprises two morphologically distinct spike-like glycoprotein surtace projections, haemagglutinin (HA) and neuraminidase (NA). HA is tbe major antigen concerned witb protection, wbile NA has a lesser role witb a subsidiary function in virus release. Both antigens project radially from the surface through a cell-derived lipid bilayer. Internal to tbe bilayer but associated witb it is tbe matrix (M) protein, wbicb is tbe major structural protein. Within both structures are tbe nucleoprotein and three large polymerase (P) proteins that are responsible for RNA replication. Since first described by Towbin et al. (1979)/ immunoblotting techniques have been widely used for the detction of antibody responses to antigens. These techniques allow access of antibody probes to virion proteins
after electrophoretic separation and transfer to nitrocellulose paper or other immobilizing matrices. The antigens of a number of viruses, including influenza viruses,^ have been examined by tbe technique. Most reports of tbe blotting technique witb influenza antigens have, however, been concerned witb studies of individual viral antigens using monoclonal antibodies or monospecific antibodies as probes,""^ and tbere have been surprisingly few reports of its use in tbe study of influenza virus infections. Foy et al. (1987) reported the use of immunoblotting as an accessory test to detect antibody responses to influenza B viral antigens using human post-infection sera, but few details were given/ In tbe present study, the immunoblotting technique was chosen to detect antibody responses to influenza virus antigens in both human and mouse sera after infection or parenteral vaccination. Tbe antibody responses of mice to vaccination by
Correspondence: Gregory A. Tannock, Faculty of Medicine, The University of Newcastle, Callaghan, NSW 2308. Australia. Accepted for publication 30 April 1992.
182
D. Qiu et al.
ditferent routes were also studied by examining the IgG subclass response.
Materials and methods Viruses Allantoie preparations of influenza strain lVAnn Arbor/1/86 (B/AA/1/86) and rt-combinant virus Memn-13el,s,(H3Nl) with the HA of A/Mcm/1/71 (H3) and the NA of A/Beilamy/42 (Nl)^ were used in the study. Virus stocks were diluted to 1 : 1000 in phosphate-buffered saline (PBS) and inoculated into the allantoie cavity of 9-10 day old embryonated eggs (0.1 mL/egg). Infected eggs were incubated for 2-3 days at 34 "C before the allantoie fluids were harvested. Viruses were assayed by a plaque technique in cells of the Madin Darby canine kidney (MDCK) line/' and titres were expressed as plaque forming units (PFU)/mL. Purification of virus Allantoie fluids from eggs infeeted with B/AA/1/86 were harvested and clarified by centrifugation at 1500^^ for 5 min at 4''C. Solid polyethylene glycol (PEG 6000; BDH Chemieals Australia Pty Ltd, Victoria, Australia) was added to the virus-containing supernatant to a final concentration of 8% (w/v) and stirred at 4''C for 1 h. The precipitate was eolleeted by centrifugation at 13 000 u for 20 min in a Beckman SW28 rotor and resuspended in a small volume of PBS. This eoneentrated virus preparation was stored frozen at - 8 0 ° C . For eggs infeeted with MemnBel^, the PEG step was omitted and clarified fluids were eentrifuged at 96 500^ for 1 h in a Beekman SW28 rotor at 4°C in order to pellet the virus. Concentrated vims was centrifuged through 30% sucrose onto a 60% sucrose cushion at 160 000 i; for 1.5 h m a Beckman SW41 rotor at 4°C. Virus was removed from the interface and diluted 1 : 3 in NTE buffer (100 mmol/L NaCl, 1 nimoI/L ethylenediamine tetra-acetie acid [EDTA], 10 mmol/L Tns HCl [pH 7.4]). The diluted virus was layered over an 8 mL 15-60% sucrose gradient and centrituged at 49 000^ for 1 h in a Beckman SW41 rotor at
4°C, and the virus band was collected. The virus was concentrated by dilution to 1 : 5 in NTE buffer and by sedimenting at 150 000^ for 1 h in a Beckman SW41 rotor at 4°C. The pellet was suspended in a small volume of NTE buffer, assayed for protein concentration and stored at - 8O''C. Polyacrylamidegel electrophoresis and semidry transfer h l i One hundred microlitres of purified virus containing 1-2 mg/niL protein, as determined by the method of Bradford (1976),^ were mixed with 100|J,L electrophoresis sample buffer {2% sodium dodecylsulfate [SDS], 20% sucrose, 0.01% [v/v] bromophenol blue, 0.125 mol/L Tris HCl [pH 6.8] with or without 5% 2-mereaptoethanoI [2-ME]) for electrophoresis under reducing or non-reducing conditions, and boiled for 2 min. Disrupted virus preparations were electrophoresed in a 14% slab polyacrylamide gel separated by a 4% stacking gel at 60 V for 18 h at 22 °C according to the method of Laemli (1970)." Following electrophoresis, the gel was equilibrated in transfer buffer (39 mmol/L glycine, 20% [v/v] methanol, 0.0375% [w/v] SDS 48 mmol/L Tris HCl [pH 9.2]) for 15-30 min. Nitrocellulose (NC) membranes and filter papers (MM P K G / 5 0 0 ; Pharmacia LKB, Bronmia, Sweden) were cut to the same size of the gel. Each NC membrane was soaked in transfer buffer for 15 min. Nine layers of wet filter paper were placed on the platinum anode, and air bubbles were removed. The NC mcmlirane was then placed on top of the filter paper and the gel on top of the NC membrane. Air bubbles were removed and nine layers of wet filter paper were placed on top of the gel. After electrophorctic transfer for 55 min at 20 V at ambient temperature (23°C), NC membranes were cut into 4 mm strips and dried at room temperature before storing at 4°C. Immunodetection The NC strips were incubated with 5 mL of blocking buffer (0.15 mol/L NaCl, 0.3% [w/v] bovine serum albumin [BSA] 0.05% [v/v] Tween-20, 10 mmol/L Tris HCl [pH 7.4]) for 30 min on a rotating platform at
Western blot analysis of antihody rootn temperature. The NC strips were then incubated in 5 mL of: blocking buffer containing a 1 : 100 dilution of serum, and incubated overnight on the rotating platform at room temperature. Control NC strips were incubated in blocking buffer in the absence of serutn or in blockitig buffer containing a similar dilution of negative serum. The strips were then washed exhaustively in washing buffer and incubated with biotinylated second antibody (Biotinylation Kit; Amersham International PIC, Amersham, UK) diluted 1 :400 in blocking buffer. Following a 2 h incubation, strips were washed a further three times and then incubated for 10 min with horse radish per oxidase-conjugated streptavidin (Amersham) diluted 1 : 1000. After washing, the strips were incubated in developing buffer {0.06% [w/v] 4-chloro-l-naphthol, 19.9% [v/v] methaiiol, 0.01% [v/v] hydrogen peroxide, 80% [v/v] 0.1 moI/L Tris HCl [pH 7.4]) for 3-5 min and the reaction stopped by rinsing with distilled water. Strips were photographed after development. Human sera Paired acute and convalescent sera were obtained from patients experiencing natural infections by B/AA/1/86. Haemagglutination inhibition (HI) titres for these sera were determined by a standard procedure.'^ Mouse sera Inbred BALB/c male mice (6-7 weeks of age) were infected according to previously described methods.^ Briefly, each mouse was anaesthetized by intraperitoneal (i.p) itijection of a 3.33% (w/v) solution of 2,2,2-tribromoethanol in 3.33% (v/v) tertiary amyl alcohol (0.01 mL/g bodyweight). Various doses of vims were instilled by the intranasal (i.n.) route in a total volume of 50 [iL using a calibrated 25 ^L dropping pipette. For i.p. inoculation, similar doses were administered in a total volume of 59 |aL. Control mice received the same volume of PBS by each route. After 3 weeks, booster inoculations wereadtninisteredatadoseof3.16 x lO'^PFU/ 50 |XL by the appropriate route. Mice were sacrificed after 3 weeks and their sera collected and tested by an HI test. The dose of virus that
183
induced the highest geometric mean titre of antibody within the group was chosen as the inoculum for further experiments. Monoclonal antibodies Monoclonal antibodies A2, A3, Bl and B2 were obtained from Dr N. J. Cox, Influenza Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control, Atlanta, GA, USA.
Results Identification ofvirion proteins of B/AA/1/86 and Mem,.rBel^. with monoclonal and polyclonal antibody Purified B/AA/1/86 and MemH-Bel^, viruses were separated under non-reducing conditions and transferred to NC tnembranes. Monoclonal and polyclonal specific antibodies were used to confirm the location of the proteins of each vims (Fig. 1). Monoclonal antibodies Bl and B2 are specific for the HA and NP of influenza B (lanes 3 and 4), while A2 and A3 are specific for the M and NP proteins ot influenza A (lanes 7 and 8), respectively.'" The HA of MetHn-Belj-^ was detected by polyclonal antibodies specific for the HA of MemH-Bel^. and a small amount of HA2specific activity was also present (Fig. 1, lane 9, arrowed). Detection of responses to influenza virus proteins after PAGE tinder reducing^ conditions usin^ human or mouse antisera Antibodies to the HA, HA 1, HA2 and NA/NP antigens of influenza Memi_i-Bel[sj and B/AA/1/86 viruses are present in both mouse sera prepared after i.n. inoculation and also in human convalescent sera (Figs 2, 3). Antibody binding to HAl, especially to that of B/AA/ 1/86, was very weak (Fig. 2, lanes 10-12; Fig. 3, lanes 4-7). On the other hand, strong antibody binding was detected to the HA2 proteins of both Mem,[-Belr^ and B/AA/ 1/86 viruses. Antibody to the M protein was only detected in human convalescent sera as a faint band (Fig. 3, lanes 6 and 7 [arrowed]). Responses to HA, HAl and HA2 were elicited
184
D. Qiu et al. 5
6
7
8
9
1
1! .1 4
5
6
94 0 0 0 * WA/NP
67 000-
43
000-
30
000-
•NA/NP
HAl
M HA,
-
-HA2
20
100-
Fig.l . Detection of antibody to influenza virus proteins after PAGE under non-reducing conditions with monoclonal or polyelonal nionospeciflc antihodies. Purified B/AA/1/86 viruses and Mcm,.,Belp..j viruses were submitted to PAGE anti the separated proteins transferred to NC strips. Strips in lanes 1 and 5 contained protein standards. Strips in lanes 1, 2, 5 and 6 were stained with amido black. Strips in lanes 3. 4, 7 and 8 were probed with biotinylated monoclonal antibodies Bl, B2, A2 and A3, specific for the HA and NP proteins of influenza B and the M and NP proteins of influenza A. respectively. The strip in lane 9 was incubated with polyclonal antibodies raised against the HA of Menin-Belr^, using HRP-conjugated goat antimouse serum as the secondary antibody.
Fig- 2. Detection of antibody to influenza virus proteins after PAGE under reducing conditions with antisera prepared in mice. Mem,|-Bel^j and B/AA/1/86 proteins were separated by PAGE under reducing conditions and transferred to the NC membrane. Strips in lanes 1 and 7 were stained with amido black. Strips in lanes 2, 3, 8 and 9 were probed with biotinylated monoclonal antibodies A3. A2. Bl and B2, which are specific for the NA and M proteins of influenza A and the HA and NP proteins of influenza B, respectively. The strip in lane 4 was reacted with polyclonal antibodies raised to the HA of Menin-Bel,^, Strips in lanes 5 and 6 were incubated with sera prepared after two i.n. inoculations of Mem,,-Bel^ (HI titres 80 and 40). Strips in lanes 10-12 were treated with sera similarly prepared against B/AA/1/86 (HI titres 40, 20 and 20, respectively).
Detection of responses to influenza virus proteins ajter PAGE under non-reducing conditions using human or mouse antisera by mouse polyclonal antibodies raised to the HA of MeniH-Bel^ (Fig. 2, lane 4 [arrowed]). Also present is a faint band of antibody to M protein, which was probably induced in response to small contaminating amounts of M protein. It is difficult to prepare purified HA without trace contamination with the more abundant matrix protein. A band of monoclonal antibody specific for the M protein is also present in lane 3, which was also observed under non-reducing conditions (Fig. 1, lane 7). Monoclonal antibody Bl, which is specific for the HA of influenza B,"^ was also able to bind to the HAl protein (Fig. 2, lane 8 [arrowed]).
Figures 4 and 5 show detection of the protein antigens of Meniii-Bel^ and B/AA/1/86 vims after separation by PAGE under nonreducing conditions with mouse sera prepared after i.n. inoculation or with human convalescent sera. Antibody to the HA and NA/NP of both vimses can be detected together with the polymerase (P) proteins of B/AA/1 /86. Again, no significant response to the M proteins of either Meni||-BelM and B/AA/1/86 viruses can be detected with mouse antisera, and there was only a very faint band of M protein with human convalescent sera (Fig. 5, lanes 2-4 [arrowed]). The human convalescent sera used in Fig. 5 (lanes 2 and 4) were
185
Western blot analysis of antibody 1
2
3
•)
5
G
HA
P
NA/NP•
HA NA/NP
• Fig. 3. Detection of antibody to H/AA/1/86 virus proteins after PAGE uniier reducing conditions using human convalescent sera. Viral proteins of B/AA/1/86 were separated under non-reducing conditioiLS by PAGE and transferred to an NC membrane. Strips in lanes 4-7 were incubated with human convalescent sera (HI titres 20, 40, 80 and 320, respectively) at a dilution of 1 : 100. Biotinylated goat anti-humaii serum was used as the secondary antibody. Tlie strip in lane 3 was detected by biotinylated monoclonal antibody B2 (Specific for the NP of influenza B) and chat in lane 1 was incubated in the absence of serum as a control. The strip in lane 2 was stained with amido black after transfer. Strips in lanes 8-11 were incubated with the corresponding acute sera (HI titre of each
Western blot analysis of antibody responses to influenza virion proteins DIWEN QiU, GREGORY A. TANNOCK, RICHARD D. BARRY and *DAVID C.JACKSON Faculty of Medicine, The University of Newcastle, Callaghan, New South Wales, and *Department ofMicrobiology, The University of Melbourne, Melbourne, Victoria, Australia
Summary An immunoblotting procedure was developed to detect antibody responses in mice and humans to influenza virion proteins. The technique was capable of detecting 1.5 jig of haemagglutinin (HA) on nitrocellulose strips at a 1 : 5000 dilution of a mouse serum with an initial haemagglutination inhibition titre of 20. The effects of the use of the blocking agent Tween-20 on virion proteins were also studied. The commonly used concentration of 0.05% (v/v) Tween-20, when included in blocking and incubation buffers, greatly reduced the amount of detectable matrix protein but caused no detectable loss of HA and neuraminidase/nucleoprotein proteins. If virion proteins were separated by polyacrylamide gel electrophoresis under reducing conditions, antibody bound to HA2 more strongly than to HAl. Under non-reducing conditions, more antibody bound to the uncleaved HA protein than to other proteins. IgGl and IgG2a antibody responses in mice to each protein were stronger than IgG2b and IgG3 responses. Introduction The influenza virus comprises two morphologically distinct spike-like glycoprotein surtace projections, haemagglutinin (HA) and neuraminidase (NA). HA is tbe major antigen concerned witb protection, wbile NA has a lesser role witb a subsidiary function in virus release. Both antigens project radially from the surface through a cell-derived lipid bilayer. Internal to tbe bilayer but associated witb it is tbe matrix (M) protein, wbicb is tbe major structural protein. Within both structures are tbe nucleoprotein and three large polymerase (P) proteins that are responsible for RNA replication. Since first described by Towbin et al. (1979)/ immunoblotting techniques have been widely used for the detction of antibody responses to antigens. These techniques allow access of antibody probes to virion proteins
after electrophoretic separation and transfer to nitrocellulose paper or other immobilizing matrices. The antigens of a number of viruses, including influenza viruses,^ have been examined by tbe technique. Most reports of tbe blotting technique witb influenza antigens have, however, been concerned witb studies of individual viral antigens using monoclonal antibodies or monospecific antibodies as probes,""^ and tbere have been surprisingly few reports of its use in tbe study of influenza virus infections. Foy et al. (1987) reported the use of immunoblotting as an accessory test to detect antibody responses to influenza B viral antigens using human post-infection sera, but few details were given/ In tbe present study, the immunoblotting technique was chosen to detect antibody responses to influenza virus antigens in both human and mouse sera after infection or parenteral vaccination. Tbe antibody responses of mice to vaccination by
Correspondence: Gregory A. Tannock, Faculty of Medicine, The University of Newcastle, Callaghan, NSW 2308. Australia. Accepted for publication 30 April 1992.
182
D. Qiu et al.
ditferent routes were also studied by examining the IgG subclass response.
Materials and methods Viruses Allantoie preparations of influenza strain lVAnn Arbor/1/86 (B/AA/1/86) and rt-combinant virus Memn-13el,s,(H3Nl) with the HA of A/Mcm/1/71 (H3) and the NA of A/Beilamy/42 (Nl)^ were used in the study. Virus stocks were diluted to 1 : 1000 in phosphate-buffered saline (PBS) and inoculated into the allantoie cavity of 9-10 day old embryonated eggs (0.1 mL/egg). Infected eggs were incubated for 2-3 days at 34 "C before the allantoie fluids were harvested. Viruses were assayed by a plaque technique in cells of the Madin Darby canine kidney (MDCK) line/' and titres were expressed as plaque forming units (PFU)/mL. Purification of virus Allantoie fluids from eggs infeeted with B/AA/1/86 were harvested and clarified by centrifugation at 1500^^ for 5 min at 4''C. Solid polyethylene glycol (PEG 6000; BDH Chemieals Australia Pty Ltd, Victoria, Australia) was added to the virus-containing supernatant to a final concentration of 8% (w/v) and stirred at 4''C for 1 h. The precipitate was eolleeted by centrifugation at 13 000 u for 20 min in a Beckman SW28 rotor and resuspended in a small volume of PBS. This eoneentrated virus preparation was stored frozen at - 8 0 ° C . For eggs infeeted with MemnBel^, the PEG step was omitted and clarified fluids were eentrifuged at 96 500^ for 1 h in a Beekman SW28 rotor at 4°C in order to pellet the virus. Concentrated vims was centrifuged through 30% sucrose onto a 60% sucrose cushion at 160 000 i; for 1.5 h m a Beckman SW41 rotor at 4°C. Virus was removed from the interface and diluted 1 : 3 in NTE buffer (100 mmol/L NaCl, 1 nimoI/L ethylenediamine tetra-acetie acid [EDTA], 10 mmol/L Tns HCl [pH 7.4]). The diluted virus was layered over an 8 mL 15-60% sucrose gradient and centrituged at 49 000^ for 1 h in a Beckman SW41 rotor at
4°C, and the virus band was collected. The virus was concentrated by dilution to 1 : 5 in NTE buffer and by sedimenting at 150 000^ for 1 h in a Beckman SW41 rotor at 4°C. The pellet was suspended in a small volume of NTE buffer, assayed for protein concentration and stored at - 8O''C. Polyacrylamidegel electrophoresis and semidry transfer h l i One hundred microlitres of purified virus containing 1-2 mg/niL protein, as determined by the method of Bradford (1976),^ were mixed with 100|J,L electrophoresis sample buffer {2% sodium dodecylsulfate [SDS], 20% sucrose, 0.01% [v/v] bromophenol blue, 0.125 mol/L Tris HCl [pH 6.8] with or without 5% 2-mereaptoethanoI [2-ME]) for electrophoresis under reducing or non-reducing conditions, and boiled for 2 min. Disrupted virus preparations were electrophoresed in a 14% slab polyacrylamide gel separated by a 4% stacking gel at 60 V for 18 h at 22 °C according to the method of Laemli (1970)." Following electrophoresis, the gel was equilibrated in transfer buffer (39 mmol/L glycine, 20% [v/v] methanol, 0.0375% [w/v] SDS 48 mmol/L Tris HCl [pH 9.2]) for 15-30 min. Nitrocellulose (NC) membranes and filter papers (MM P K G / 5 0 0 ; Pharmacia LKB, Bronmia, Sweden) were cut to the same size of the gel. Each NC membrane was soaked in transfer buffer for 15 min. Nine layers of wet filter paper were placed on the platinum anode, and air bubbles were removed. The NC mcmlirane was then placed on top of the filter paper and the gel on top of the NC membrane. Air bubbles were removed and nine layers of wet filter paper were placed on top of the gel. After electrophorctic transfer for 55 min at 20 V at ambient temperature (23°C), NC membranes were cut into 4 mm strips and dried at room temperature before storing at 4°C. Immunodetection The NC strips were incubated with 5 mL of blocking buffer (0.15 mol/L NaCl, 0.3% [w/v] bovine serum albumin [BSA] 0.05% [v/v] Tween-20, 10 mmol/L Tris HCl [pH 7.4]) for 30 min on a rotating platform at
Western blot analysis of antihody rootn temperature. The NC strips were then incubated in 5 mL of: blocking buffer containing a 1 : 100 dilution of serum, and incubated overnight on the rotating platform at room temperature. Control NC strips were incubated in blocking buffer in the absence of serutn or in blockitig buffer containing a similar dilution of negative serum. The strips were then washed exhaustively in washing buffer and incubated with biotinylated second antibody (Biotinylation Kit; Amersham International PIC, Amersham, UK) diluted 1 :400 in blocking buffer. Following a 2 h incubation, strips were washed a further three times and then incubated for 10 min with horse radish per oxidase-conjugated streptavidin (Amersham) diluted 1 : 1000. After washing, the strips were incubated in developing buffer {0.06% [w/v] 4-chloro-l-naphthol, 19.9% [v/v] methaiiol, 0.01% [v/v] hydrogen peroxide, 80% [v/v] 0.1 moI/L Tris HCl [pH 7.4]) for 3-5 min and the reaction stopped by rinsing with distilled water. Strips were photographed after development. Human sera Paired acute and convalescent sera were obtained from patients experiencing natural infections by B/AA/1/86. Haemagglutination inhibition (HI) titres for these sera were determined by a standard procedure.'^ Mouse sera Inbred BALB/c male mice (6-7 weeks of age) were infected according to previously described methods.^ Briefly, each mouse was anaesthetized by intraperitoneal (i.p) itijection of a 3.33% (w/v) solution of 2,2,2-tribromoethanol in 3.33% (v/v) tertiary amyl alcohol (0.01 mL/g bodyweight). Various doses of vims were instilled by the intranasal (i.n.) route in a total volume of 50 [iL using a calibrated 25 ^L dropping pipette. For i.p. inoculation, similar doses were administered in a total volume of 59 |aL. Control mice received the same volume of PBS by each route. After 3 weeks, booster inoculations wereadtninisteredatadoseof3.16 x lO'^PFU/ 50 |XL by the appropriate route. Mice were sacrificed after 3 weeks and their sera collected and tested by an HI test. The dose of virus that
183
induced the highest geometric mean titre of antibody within the group was chosen as the inoculum for further experiments. Monoclonal antibodies Monoclonal antibodies A2, A3, Bl and B2 were obtained from Dr N. J. Cox, Influenza Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control, Atlanta, GA, USA.
Results Identification ofvirion proteins of B/AA/1/86 and Mem,.rBel^. with monoclonal and polyclonal antibody Purified B/AA/1/86 and MemH-Bel^, viruses were separated under non-reducing conditions and transferred to NC tnembranes. Monoclonal and polyclonal specific antibodies were used to confirm the location of the proteins of each vims (Fig. 1). Monoclonal antibodies Bl and B2 are specific for the HA and NP of influenza B (lanes 3 and 4), while A2 and A3 are specific for the M and NP proteins ot influenza A (lanes 7 and 8), respectively.'" The HA of MetHn-Belj-^ was detected by polyclonal antibodies specific for the HA of MemH-Bel^. and a small amount of HA2specific activity was also present (Fig. 1, lane 9, arrowed). Detection of responses to influenza virus proteins after PAGE tinder reducing^ conditions usin^ human or mouse antisera Antibodies to the HA, HA 1, HA2 and NA/NP antigens of influenza Memi_i-Bel[sj and B/AA/1/86 viruses are present in both mouse sera prepared after i.n. inoculation and also in human convalescent sera (Figs 2, 3). Antibody binding to HAl, especially to that of B/AA/ 1/86, was very weak (Fig. 2, lanes 10-12; Fig. 3, lanes 4-7). On the other hand, strong antibody binding was detected to the HA2 proteins of both Mem,[-Belr^ and B/AA/ 1/86 viruses. Antibody to the M protein was only detected in human convalescent sera as a faint band (Fig. 3, lanes 6 and 7 [arrowed]). Responses to HA, HAl and HA2 were elicited
184
D. Qiu et al. 5
6
7
8
9
1
1! .1 4
5
6
94 0 0 0 * WA/NP
67 000-
43
000-
30
000-
•NA/NP
HAl
M HA,
-
-HA2
20
100-
Fig.l . Detection of antibody to influenza virus proteins after PAGE under non-reducing conditions with monoclonal or polyelonal nionospeciflc antihodies. Purified B/AA/1/86 viruses and Mcm,.,Belp..j viruses were submitted to PAGE anti the separated proteins transferred to NC strips. Strips in lanes 1 and 5 contained protein standards. Strips in lanes 1, 2, 5 and 6 were stained with amido black. Strips in lanes 3. 4, 7 and 8 were probed with biotinylated monoclonal antibodies Bl, B2, A2 and A3, specific for the HA and NP proteins of influenza B and the M and NP proteins of influenza A. respectively. The strip in lane 9 was incubated with polyclonal antibodies raised against the HA of Menin-Belr^, using HRP-conjugated goat antimouse serum as the secondary antibody.
Fig- 2. Detection of antibody to influenza virus proteins after PAGE under reducing conditions with antisera prepared in mice. Mem,|-Bel^j and B/AA/1/86 proteins were separated by PAGE under reducing conditions and transferred to the NC membrane. Strips in lanes 1 and 7 were stained with amido black. Strips in lanes 2, 3, 8 and 9 were probed with biotinylated monoclonal antibodies A3. A2. Bl and B2, which are specific for the NA and M proteins of influenza A and the HA and NP proteins of influenza B, respectively. The strip in lane 4 was reacted with polyclonal antibodies raised to the HA of Menin-Bel,^, Strips in lanes 5 and 6 were incubated with sera prepared after two i.n. inoculations of Mem,,-Bel^ (HI titres 80 and 40). Strips in lanes 10-12 were treated with sera similarly prepared against B/AA/1/86 (HI titres 40, 20 and 20, respectively).
Detection of responses to influenza virus proteins ajter PAGE under non-reducing conditions using human or mouse antisera by mouse polyclonal antibodies raised to the HA of MeniH-Bel^ (Fig. 2, lane 4 [arrowed]). Also present is a faint band of antibody to M protein, which was probably induced in response to small contaminating amounts of M protein. It is difficult to prepare purified HA without trace contamination with the more abundant matrix protein. A band of monoclonal antibody specific for the M protein is also present in lane 3, which was also observed under non-reducing conditions (Fig. 1, lane 7). Monoclonal antibody Bl, which is specific for the HA of influenza B,"^ was also able to bind to the HAl protein (Fig. 2, lane 8 [arrowed]).
Figures 4 and 5 show detection of the protein antigens of Meniii-Bel^ and B/AA/1/86 vims after separation by PAGE under nonreducing conditions with mouse sera prepared after i.n. inoculation or with human convalescent sera. Antibody to the HA and NA/NP of both vimses can be detected together with the polymerase (P) proteins of B/AA/1 /86. Again, no significant response to the M proteins of either Meni||-BelM and B/AA/1/86 viruses can be detected with mouse antisera, and there was only a very faint band of M protein with human convalescent sera (Fig. 5, lanes 2-4 [arrowed]). The human convalescent sera used in Fig. 5 (lanes 2 and 4) were
185
Western blot analysis of antibody 1
2
3
•)
5
G
HA
P
NA/NP•
HA NA/NP
• Fig. 3. Detection of antibody to H/AA/1/86 virus proteins after PAGE uniier reducing conditions using human convalescent sera. Viral proteins of B/AA/1/86 were separated under non-reducing conditioiLS by PAGE and transferred to an NC membrane. Strips in lanes 4-7 were incubated with human convalescent sera (HI titres 20, 40, 80 and 320, respectively) at a dilution of 1 : 100. Biotinylated goat anti-humaii serum was used as the secondary antibody. Tlie strip in lane 3 was detected by biotinylated monoclonal antibody B2 (Specific for the NP of influenza B) and chat in lane 1 was incubated in the absence of serum as a control. The strip in lane 2 was stained with amido black after transfer. Strips in lanes 8-11 were incubated with the corresponding acute sera (HI titre of each
