2. gen. ViroL (I977), 38, I87-I93
I87
Printed in Great Britain
Quantification of Influenza Virus Structural Proteins
using Rocket Immunoelectrophoresis
(Accepted I6 August I977)
SUMMARY
The major influenza A virus structural antigens, matrix protein, nucleoprotein,
haemagglutinin and neuraminidase were measured rapidly and accurately using a
rocket immunoelectrophoresis technique. Virus was disrupted with 1% (w/v)
sodium sarcosyl and electrophoresed into agarose containing specific antiserum
to the individual virus structural proteins in 0"05 M-barbitone buffer, pH 8-6, for
I to 4 h. For haemagglutinin antigen assays statistical analysis indicated that the
coefficient of variation within an immunoelectrophoresis plate was 8.0% for
antigen concentrations in the range I5 to 40/~g/ml protein. For haemagglutinin
and matrix protein the method was sufficiently sensitive to measme concentrations
of antigens as low as I-5 and 2.o/,g/ml respectively. By incorporation in the agarose
of mixtures of antisera against specific antigens of the virus, haemagglutinin,
matrix or nucleoprotein could be assayed simultaneously.
The electrophoresis of antigens into an agarose gel containing specific antisera is a
simple, rapid and reproducible procedure for antigen quantification (Laurell, I965;
Verbruggen, I975; R. Mayner & D. Barry, unpublished data). The present paper describes
the development of the method for the quantitative assay of each of the major antigens of
the influenza virus particle; matrix protein (M), nucleoprotein (NP), neuraminidase (NA)
and the haemagglutinin (HA).
Influenza viruses were grown and purified as described by Skehel & Schild (I970, and
concentrated virus suspensions containing approx. I5 mg/ml protein were stored at
- 7 0 °C.
Influenza virus HA was released from X-53 [with surface glycoprotein antigens identical
to A/New Jersey/76 (HswIN0] and X-47 [surface antigens of A/Victoria/75 (H3N2) virus]
recombinants by digestion with bromelain at 37 °C and further purified by rate zonal
centrifugation in 5 to 2o % sucrose gradients (Brand & Skehel, I972). Neuraminidase was
released by disruption of X-42 recombinant virus concentrates with 1% (w/v) SS (sodium
sarcosyl, NL97 Ciba-Geigy) followed by separation by electrophoresis on cellulose acetate
strips at pH 6.6. The recombinant X-42 has the neuraminidase antigen derived from
A/Port Chatmers/I/73 (H3N2) virus and haemagglutinin derived from A/Equine/Prague/56
(HeqINeql) virus. M antigen was separated by cellulose acetate electrophoresis at pH 6.6
from the avian influenza A virus A/Chicken/Germany 'N'/49 (Hav2Neq0 and NP was
separated similarly from X-42 recombinant virus as described previously (Oxford & Schild,
I975, I976), except that the viruses were disrupted with 1% (w/v) SS. Virus structural
proteins thus separated were analysed for contaminating polypeptides by polyacrylamide
gel electrophoresis using gels with discontinuous buffers (Lamb, 1975) and virus antigen
preparations with nc, contaminating virus polypeptides were used to immunize rabbits,
goats or sheep to obtain monospecific antisera. Animals were given two injections of approx.
40/~g virus protein in Freund's complete adjuvant at two weekly intervals and the serum
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33
I88
Short communications
collected z weeks later. Sera were demonstrated to be monospecific f o r the respective
influenza virus antigen by immunodouble diffusion analysis as described by Schild, Oxford
& Virelizier (I976).
For quantitative analysis by rocket immunoelectrophoresis a preparation of purified
virus to be used as a standard and unknown preparations to be quantified were disrupted
by incubation at r o o m temperature for 15 min with ~ % (w/v) SS detergent and serial twofold dilutions prepared in o'o25 M-barbitone acetate buffer, p H 8.6 (Oxoid Ltd) containing
1% (w/v) SS. Thirty-five ml of 1% agarose A37 (Indubiose, l'Industrie Biologique Frangais)
dissolved in o'o5 M-barbitone buffer at p H 8.6 and containing monospecific antiserum to the
influenza antigen to be quantified was poured on to a glass plate approx. 23 cm x 8.5 cm.
The antiserum concentration in the agarose was determined by experiment. Optimum concentrations of antiserum resulted in rocket heights between 4 and 3o m m in length and
varied f r o m I #l/ml agarose for rabbit, goat and sheep anti-HA sera to 3o/~l/ml for rabbit
anti-M sera. A row of forty 4 m m wells 1.5 m m apart were made in the agar0se approx.
1.5 cm from the longer edge of the plate and 2o/~1 samples of the dilutions of detergent disrupted viruses were added to each well. Wells were not made within 1.5 cm of the shorter
side of the plate because of variations in potential difference at the edge. Electrophoresis was
carried out in an L K B 21 I7 Multiphor apparatus cooled with circulated water (4 °C) at
8V/cm of agarose for I to 4 h. Under these alkaline p H conditions I g G antibody molecules have been shown to remain stationary in the agarose gel or to migrate slowly towards
the two electrodes (Laurell, I965; GraubaUe et al. I977). Influenza virus antigens under the
same conditions in the presence of SS detergent migrated rapidly towards the anode and
formed a rocket-shaped immunoprecipitate. In some experiments electrophoresis was carried
out at 2-5 V/cm of agarose for I7 h. After electrophoresis the gel was pressed (io g/cm ~) for
I5 min under several layers of soft blotting paper, washed in phosphate buffered saline for
2 h, dried at 6o °C for 3o min and stained in o'3 % (v/v) Coomassie blue dissolved in 5o %
methanol and lo % acetic acid. Plates were destained in the same solvent. The heights of
the stained rocket shaped immunoprecipitates were measured to an accuracy of o-I m m
with a calibrating viewer (Transidyne GeneIal Corporation). In preliminary experiments
other buffer systems were investigated including o'o75 M-calcium-lactate buffer, p H 8.6,
o.i M- or o'o25 M-barbitone buffer, p H 8 . 6 , and o.o8 M-barbital-glycine-tris buffer at
p H 8-8. These buffers were less satisfactory either because of non-specific precipitation
reactions with sodium sarcosyl detergent or because diffuse rocket immunoprecipitates were
obtained.
To allow migration of M protein, sodium sarcosyl detergent was required to disrupt
Fig. I. Rocket-shaped immunoprecipitates of influenza virus structural antigens (a) M protein:
serial twofold dilutions (r to 5) of the standard reference virus and ~thethree virus samples to be
quantified (VI, V2 and V3) were disrupted with I ~ (w/v) SS detergent and electrophoresed
into agarose gel containing 30/~l/ml of specific anti-M serum for 45 win at 8 V/cm of agarose.
MI and M2 are samples of purified M antigen of known protein concentration. The standard
reference virus contained 30/zg/ml of M protein. (b) Haemagglutinin: (i) the standard reference
A/New Jersey/8/76 virus was diluted in four twofold steps (I to 4). VI to V8 are virus preparations
to be assayed for HA content containing various high and low concentrations of HA. The standard
reference virus contained 38/~g/ml of HA activitY. (ii) Within:plate variation of a single dilution of
standard HA antigen. Electrophoresis at 8 V/cm for 4 h into agarose containing I ,ug/ml of specific
anti-HA serum. (c) Nucleoprotein: electrophoresis was for 4 h at 8 V/cm and:3/zl of specific antiNP serum was incorporated in the agarose gel. S is the sample of standard reference A/NJ/8/76
(HswIND virus diluted in twofold dilution steps (t to 4). VI, V2, V3, V4 and V5 are a series of
virus preparations to be assayed for NP content diluted in twofold steps. The standard reference
virus contained 30 #g/ml of NP antigen.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33
~o
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33
I9o
Short communications
influenza virus. No rockets were obtained when virus was disrupted at room temperature
with non-ionic detergents such as 1% (v/v) Triton X-Ioo. Heating virus which has been
disrupted in 1% (w/v) SS detergent at Ioo °C for 2 min inactivates other virus coded
antigens but had no effect on the rocket size or shape of M protein and confirmed previous
studies of the heat stability of the M protein type-specific determinants (Oxford & Schild,
I976).
Influenza M protein migrated more rapidly than the other virus antigens when electrophoresed and readily measurable rocket-shaped precipitates with diffuse edges were produced after 45 rain electrophoresis (Fig. 0. Similarly to the single radial diffusion method
for M protein, the area within the immunoprecipitate was inversely proportional to the
concentration of antibody in the agarose (Oxford & Schild, 1976). Quantities of M protein
as low as 2 #g/ml could be detected and thus the lower limits of sensitivity of the method
were of the same order as the previously reported values of 0"5/zg/ml for human placental
lactogen hormone and 2"7 #g/ml for IgA (Bock, I973; Norgaard-Pedersen & Gaede, I973).
Matrix protein concentrations in samples for analysis were calculated by comparison
to a reference standard influenza A virus preparation with a total protein content of
Io mg/ml and M protein content of 22.2 % determined from densitometry tracings of
analytical polyacrylamide gels of the virus (Skehel & Schild, I97I). The reference preparation itself was further calibrated initially by comparison with purified M protein
samples of known protein concentration when both samples were electrophoresed on the
same plate. In subsequent experiments precise M protein concentrations of test samples were
calculated by comparison of their rocket heights with that of the whole virus standard.
Similar methods were used for calculation of HA and NP antigen concentrations. Fig. 2
illustrates the quantitative relationship between rocket immunoprecipitate height and
antigen concentration for the M and HA antigens. A straight line relationship was obtained
when the logarithm of the rocket height was plotted against the logarithm of the M or HA
antigen concentrations. Deviation from linearity was noted when concentrations of antigen
were very high (rockets greater than 30 mm) or low (rockets less than 4 mm). The concentration of M or HA antigen in an unknown sample was calculated using standard parallel
line assay methods in comparison with the standard virus of known M and HA antigen
concentration (Finney, 1970.
Well defined rocket shaped precipitates were obtained when HA from influenza A/New
Jersey/76 (HswiNi) or A/Victoria/75 (H3Nz) viruses were quantified using specific antisera
against these antigens (Fig. I b) and minimal concentrations of 1.5/~g/ml HA antigen could
be detected. Statistical analysis of the data indicated that the coefficient of variation for
within-plate rocket heights was 8.o % of the mean rocket height for antigen concentrations
in the range I5 to 4o/~g/ml protein. For the lowest antigen concentrations (2 #g/ml) the
coefficient of variation increased to I5 %.
Antiserum prepared against the pmified HA of a particular influenza A virus subtype
could be used to measure the HA content of other related viruses of the same subtype,
although such viruses showed considerable antigenic change in the strain specific antigenic
determinants. Thus, a monospecific antiserum prepared against A/Port Chalmers/r/73
(H3N2) virus was incorporated in the agarose and the heights of the rockets obtained with
the homologous virus were compared with those obtained with A/Victoria/75 (H3Nz) and
Fig. z. Quantitative relationship between antigen concentration and height of rocket-shaped
immuno precipitates for influenza M and H A proteins. Twofold dilutions of the standard
reference A/NJ/8/76 (HswlNi) virus (A) and the virus samples to be quantified (B and C) were
disrupted and electrophoresed as described in Fig. I.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33
1:4
1:8
1:16
~,'~.
O
p~
1:1
1:2
1:4
1:8
1:1
"~ 1:2
>
1:1
1:2
1:4
1:8
1:16
1:32 - -
I
bO
u
°
°
I
I
•
I
I
!
I
Rocket height (ram)
I
I
II
I92
Short communications
A/Hong Kong/i[68 (H3N2) viruses electrophoresed in the same plate. Relatively short,
dense rockets were detected with the homologous virus whereas longer, less densely staining
rockets were obtained with the two related viruses A/Victoria/75 and A/Hong Kong/i/68.
A straight line dose response was obtained for each virus and the single antiserum. Thus
antiserum prepared against the HA of a single influenza A subtype may be used to quantify
the HA antigen of the other members of the subtype providing that the serum concentration
in the gel is increased and that an appropriate homologous virus of defined HA content is
used as a standard. Similar results with a single antiserum for the assay of viruses of the
A[Hong Kong/I/68 subtype have been obtained using the single radial diffusion test
(Wood et al. I977).
Essentially similar quantitative results were obtained for determinations of the typespecific nucleoprotein antigen except that the rocket immunoprecipitates had more sharply
defined precipitate margins similar to those obtained with HA antigen (Fig. I e). In contrast
to M protein, 1% (v/v) Triton X-Ioo could be used as an alternative detergent to sodium
sarcosyl to disrupt virus particles before electrophoresis for determinations of NP antigen.
Rockets were obtained when a specific antiserum to virus NA was incorporated in the
agarose at a concentration of 3 #l/ml agarose and electrophoresis carried out at 8 V/cm for
4 h. A linear dose response was obtained when the logarithm of the rocket height was
plotted against the logarithm of the NA antigen dilution and concentrations of NA could
be calculated by comparison with the standard reference virus as described above for the
HA antigen. However, the rockets were faint and poorly defined compared to rocket
immunoprecipitates obtained with the HA, NP and M antigens. Further work is in progress
to provide more potent antisera to virus NA antigen.
When concentrations of monospecific antisera in the agarose were suitably adjusted the
method was used to assay two influenza virus antigens simultaneously. Thus, with relatively
low concentrations of NP antibody in the agarose long rockets were obtained, whereas the
incorporation of higher concentrations of anti-HA antibody resulted in shorter, easily
distinguishable dense rockets within the longer NP immunoprecipitates. No interference
was detected between the two immunoprecipitates when two viruses were tested in parallel.
Thus, samples containing equivalent protein concentrations of the avian virus ' N ' and
A/Victoria/75 (H3N2) virus were tested in an immunoelectrophoresis plate containing a
mixture of anti-NP antiserum and antiserum prepared against the H3 subtype of haemagglutinin. A single NP antigen rocket was obtained with ' N ' virus, whereas two rocket
(NP and HA) immunoprecipitates were obtained with the A/Victoria]75 virus. Rocket
heights for NP activity were comparable for the two viruses and were also identical when
the NP activity of both viruses was assayed in plates containing only antiserum to NP
antigen. The method was also used to assay simultaneously M and HA antigens in a virus
preparation.
Influenza antigens, particularly HA, can be quantified by single radial diffusion, where the
virus antigens move from the well into the antibody-containing agarose gel by simple
diffusion over a period of ~5 to 24 h (Schild et al. 2976; Schild, Wood & Newman, 1975).
However, rocket immunoelectrophoresis allows much more rapid quantification in I to
4 h than SRD. The within-plate reproducibility of the two methods for the detection of
concentrations of HA antigen in the range I5 to 4o/zg/ml protein are comparable and
approximate to 5 to 8 ~o coefficient of variation. Both SRD and rocket immunoelectrophoresis allow the detection of approx. 2/zg/ml of influenza antigens. Immunoelectrophoresis may be a useful method for the accurate and rapid assay of influenza vaccines
(R. Mayner, personal communication) and also other virus vaccines where specific antisera
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33
Short communications
I93
against virus subcomponents can be prepared, such as rabies (Crick & Brown, I97O),
measles (Hall & Martin, 1974) and adenovirus (Mautner & Willcox, 1974). In preliminary
experiments we have used the method for the assay of the M, NP, HA and NA contents
of a number of fl-propiolactone inactivated bivalent and subunit influenza vaccines
(Oxford et al. 1977) and also for the analysis of sucrose gradients containing influenza
antigens.
We would like to acknowledge the collaboration of R. Mayner, Bureau of Biologics,
Washington, who indicated the usefulness of the technique for the assay of influenza
vaccines. We thank Valerie Seagroatt for the statistical analysis of the data, R. Newman
for the preparation of specific antisera and T. Corcoran for technical assistance. B. Watts
provided extensive photographic assistance.
Division of Virology
National Institute for Biological Standards
and Control
Holly Hill, Hampstead
London NW3
J.S. OXFORD
G . C . SCmLD
REFERENCES
BOCK, E. (I973). Q u a n t i t a t i o n o f p l a s m a proteins in cerebrospinal fluid. Scandinavian Journal of Immunology
2 (Supplement I), 111-117.
BRAND, C. M. & SKEHEL, J. J. (1972). Crystalline antigen f r o m the influenza virus envelope. Nature New Biology
238, 145-147.
CRICK, J. & BROWN, F. (I970). Small i m m u n i z i n g s u b u n i t s in rabies virus. I n The Biology of Large RNA
Viruses, pp. 133-139. Edited by R. D. Barry a n d B. W. J. Malay. L o n d o n : A c a d e m i c Press.
FINNEY, D. S. (1971). In Statistical Method in Biological Assay, pp. 99-136. L o n d o n : Charles Griffin.
GRAUBALLE, P. C., GENNER, J., MEYLING, A. & HORNSLETH, A. (1977)- R a p i d diagnosis o f rotavirus infections:
c o m p a r i s o n o f electron m i c r o s c o p y a n d i m m u n o e l e c t r o - o s m o p h o r e s i s for the detection of rotavirus in
h u m a n infantile gastroenteritis. Journal qfGeneral Virology 35, 203--218HALL, W. W. & MARTIN, S. J. (1974). T h e biochemical a n d biological characteristics of the surface c o m p o n e n t s
of measles virus. Journal of General Virology 22, 363-374.
LAMB, R. A. (1975). T h e p h o s p h o r y l a t i o n of Sendal virus proteins by a virus particle associated protein kinase.
Journal of General Virology 26, 249-263.
LAURELL, C.-B. (1965). A n t i g e n - a n t i b o d y crossed electrophoresis. Analytical Biochemistry xo, 358-361.
MAUTNER, V. & WILLCOX, I-I. N. A. (1974). A d e n o v i r u s antigens: a m o d e l system in mice for s u b u n i t vaccination. Journal of General Virology 25, 325-336.
NOR,GAARD-PEDERSEN, B. & GAEDE, P. (1973)- I m m u n o e l e c t r o p h o r e t i c a l q u a n t i t a t i o n o f h u m a n placental
lactogen h o r m o n e (HPL). Scandinavian Journal of Immunology 2 (Supplement I), 129-131.
OXFORD, J. S. & SCHILD, G. C. (I975). I m m u n o l o g i c a l studies with influenza virus m a t r i x protein. In Negative
Strand Viruses, pp. 611-62o. Edited by B. W. J. M a h y a n d R. D . Barry. L o n d o n : A c a d e m i c Press.
OXFORD, J. S. & SCHILD, G. C. (I976). I m m u n o l o g i c a l a n d physicochemical studies o f influenza m a t r i x (M) polypeptides. Virology 74, 394-4o2.
OXFORD, J. S., SCHILD, G. C., WOOD, J. M., NEWMAN, R. W. & SEAGROATT,V. (1977). T h e assay of influenza virus
structural antigens in vaccines by rocket immunoelectrophoresis. Developments in Biological Standardisation (in the press).
SCHILD, G. C., OXFORD, J. S. & VIRELIZIER, J. L. (I 976). T h e i m m u n e response to influenza - m e t h o d s o f analysis
for antibodies a n d antigens. In The Role of Immunological Factors in Infections, Allergic and Autoimmune Processes. Edited by R. F. Beers a n d E. G. Bassett. N e w Y o r k : R a v e n Press.
SCHILD, G. C., WOOD, J. M. & NEWMAN, R. W. (I975). A single radial i m m u n o d i f f u s i o n technique for the assay
o f influenza h a e m a g g l u t i n i n antigen. Bulletin of the Worm Health Organization 52, 223-231.
SKEHEL, J. J. & SCHILD, G. C. (1971). T h e polypeptide c o m p o s i t i o n of influenza A viruses. Virology 44, 396-4o8.
VERBRUGGEN, R. (1975). Quantitative i m m u n o e l e c t r o p h o r e t i c m e t h o d s : a literature survey. Clinical Chemistry
2x, 5-43.
WOOD, s. M., SCHILD, G. C., NEWMAN, R. & SEAGROATT,V. (1977). Application of a n i m p r o v e d single radial imm u n o d i f f u s i o n technique for the assay of h a e m a g g l u t i n i n antigen content of whole virus a n d s u b u n i t
influenza vaccines. Developments in Biological Standardization (in the press).
(Received 4 July 1977)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 16:38:33