A single-radial-immunodiffusion technique for the assay of influenza haemagglutinin antigen

INFLUENZA A single-radial-immunodiffusion technique for the assay of influenza haemagglutinin antigen Proposals for an assay method for the haemagglutinin content of influenza vaccines G. C. SCHILD,1 J. M. WOOD,2 & R. W. NEWMAN 3 Single-radial-diffusion techniques are proposed as possible alternatives to tests based on agglutination of erythrocytes for the assay of the haemagglutinin content of influenza vaccines. Two test procedures (microtest and macrotest) and the use of reference reagents to assay vaccines using these tests are described. The two tests are ofsimilar reproducibility and accuracy but the macrotest is technically simpler to perform and results ofassays may be obtained more rapidly. Assay systems for influenza haemagglutinin that are based on the agglutination of erythrocytes have a number of practical and theoretical disadvan- tages (1). Such assays are subject to a high degree of intra- and inter-laboratory variation. Also it must be accepted that the ability of a haemagglutinin pre- paration to agglutinate erythrocytes depends not only on the quantity of haemagglutinin present but also on its physical state. Haemagglutinin may be present on the intact virus particle, which may be spherical or filamentous in morphology or may exist as subviral structures or as isolated haemagglutinin in the form of polyvalent or monovalent structures. So-called 'subunit ' influenza vaccines may contain a wide variety of subviral structures in various degrees of purity and may differ considerably in their ability to agglutinate erythrocytes. In addition, the ability of the haemagglutinin to react with erythro- cytes does not establish its antigenicity. Because of this attempts have been made to devise alternative methods applicable to the standardization of influ- enza vaccines and independent of haemagglutina- tion. Schild et al. (2) proposed a method for the assay of influenza haemagglutinin antigen based on the principles of single-radial-immunodiffusion (3). 1 Director, WHO Collaborating Centre for Virus Refer- ence and Research, National Institute for Medical Research, Mill Hill, London, NW7 1AA, England. Present address: Virology Division, National Institute for Biological Stan- dards and Control, Holly Hill, Hampstead, London, NW3 6RB, England. 2Virology Division, National Institute for Biological Standards and Control. 'Technical Officer, National Institute for Medical Re- search. In this test antibody to purified haemagglutinin is incorporated in agarose gels and the haemagglutinin antigen for assay is added to wells in the gel after treatment with detergent to ensure its diffusibility in agarose. Haemagglutinin antigen diffuses radially from the wells and reacts with specific antibody, producing a zone of opalescence in the gel. The area of the reaction zone surrounding antigen-containing wells is an estimate of the quantity of antigen added to the well. The present paper describes two alternative test procedures for the assay of influenza haemagglutinin antigen, a microtest and a macrotest. Standard reagents for the assay of the haemagglutinin of A/Port Chalmers/1/73, the currently prevalent influ- enza A virus, are also described. Both tests enabled assays to be performed with a high degree of accu- racy and reproducibility. The macrotest is techni- cally simpler than the microtest and the results of assays may be obtained more rapidly (within 24 hours) since, unlike the microtest, it is not necessary to stain the reaction zones. The microtest is, how- ever, more economical of reagents. PREPARATION OF SINGLE-RADIAL-DIFFUSION IMMUNOPLATES Microtest 10-mi aliquots of molten A37 agarose are cooled to 56°C and 125 I-l of reconstituted (i.e., 1: 10) antihaemagglutinin serum (74/524-see Annex 1) are added. This is equivalent to a concentration of 1.25 ml of undiluted serum per litre of gel. 3-ml 3342 - 223- BULL. WORLD HEALTH ORGAN., Vol. 52, 1975 5 224 G. C. SCHILD ET AL. volumes of gel are added to empty Hyland immuno- plates a placed on a horizontal surface. After the gel has set (approx. 20 min), a series of circular wells each 3 mm in diameter is cut in the gel using a punch and a plastic template (2). The plates are left on the bench with the gel surface exposed for about 20 min to allow excess moisture to evaporate before use. Prepared immunoplates should be stored at 4°C if not used immediately. Provided that drying and shrinkage of the gel is avoided by the use of an airtight lid or by placing the plates in a moist chamber, the immunoplates may be stored for 3-4 weeks without deleterious effects. Macrotest Immunoplates containing approximately twice the concentration of antiserum are prepared in 9-cm petri dishes in an essentially similar manner. 24-ml aliquots of 1.5% agarose (56°C) are dispensed and the reconstituted contents of one ampoule of anti- serum (0.5 ml of 1: 10 dilution) are added. This is equivalent to a concentration of 2.08 ml of undiluted antiserum per litre of gel. Twelve-millilitre volumes are then added to petri dishes and after allowing time for the gels to set, 4-mm circular wells are cut in the gel. The immunoplates are then ready for use. TREATMENT OF TEST ANTIGEN AND SETTING UP THE TEST Microtest Dilutions of reference antigen (74/560-see Annex 1) and also test antigens are prepared in tubes, or in wells in plastic haemagglutination trays, using 0.5-ml volumes in a doubling dilution series, i.e. 1: 1, 1: 2, 1: 4, 1: 8, with phosphate buffered saline as the diluent. It is recommended that a fresh pipette be used for each step in the dilution series. To each 0.5-ml volume of antigen dilution, 0.2 ml of a 50-g/litre solution of sodium sarcosyl sulfate b is added. After standing for 30 min at room tempera- ture 10-pl volumes of detergent-treated antigen are added to wells in the prepared immunoplate, using at least two wells for each sample. The standard antigen (74/560) should be tested on each of the immunoplates used for the assay of the unknown antigens. a Microscope-slide-sized plastic chambers obtained from Hyland Division, Travenol Laboratories Inc., CA 92629, USA. b Geigy NL 97, from Ciba Geigy Corporation, Ardsley, NY 10502, USA. Macrotest In this test treatment of the virus antigen with detergent is carried out in the wells of the immuno- plates. Dilutions (1: 1, 1: 2, 1: 4, and 1: 8) of the reference antigen and test antigen are prepared and 20-,ud volumes of antigen are introduced into the 4- mm wells in the immunoplates. Time is allowed for the wells to empty (about 30 min) and 10-pl volumes of sodium sarcosyl sulfate are added to the wells. READING OF RESULTS To allow the reaction zones to develop, immuno- plates fitted with air-tight lids or placed in a moist container to prevent drying are incubated at room temperature (18-24°C) for 24 hours. To obtain provisional data, the diameters of the opalescent zones of reaction are measured without staining the plates using an eyepiece micrometer scale a cali- brated in 0.1-mm divisions. However, in the micro- test unstained zones are faint and to obtain accurate readings it is desirable 'to stain the zones with a protein stain. Unstained zones in the macrotest are more easily measured, especially after immersing the plates in phosphate buffered saline for several hours. After provisional measurements have been taken, unprecipitated antigen and serum components are eluted from the agarose by washing in phosphate buffered saline for 24 hours. The plates are then immersed in Coomassie brilliant blue (5 g/litre) for 15-30 minutes. Destaining is carried out using several changes of tapwater until the background agarose is colourless or light blue; this takes 24-36 hours. The diameter of each zone is measured in two directions at right angles and the mean value re- corded to allow for a lack of circularity since zones are occasionally formed which are elliptical in shape. The mean diameter for each antigen dilution is obtained from the results obtained by testing the antigen in at least two wells. The zone annulus area-i.e., the zone area omitting the area of the well, corresponding to each zone diameter may be read from a standard table (see Annex 2). INTERPRETATION OF DATA Microtest A typical microtest immunoplate is illustrated in Fig. la. A plot of zone annulus area against antigen a Matchless Machines Ltd, Horsham, Sussex, England. SINGLE-RADIAL-IMMUNODIFFUSION ASSAY OF INFLUENZA HAEMAGGLUTININ Fig. 1a. A microtest. The gel contains 1.25 ml/I of antiserum G8. The wells in rows T and R contain dilutions of the test and reference antigen (74/560), respectively. 30- 25 + 56 20 4 74, 560E 1l5 + 0 0 ~~~~~~~~~~~~~~~~2010 ± ~~~~~~~VIl 5 V2 0 0< 1:8 1:4 1:2 Dilution of antigenWH73J Fig. 1 b. Dose response curves for reference antigen (74/560) and two test vaccines (Vi and V2) in the microtest when dilution of antigen is plotted against zone annulus area. dilution should approximate to linearity for the standard and test antigens. Fig. lb shows the dose response curves for the standard antigen 74/560 and two test antigens, VI and V2. The dose responses for the standard and test antigens are seen to be linear when zone annulus area is plotted against haemag- glutinin antigen concentration. However, the slopes of the graphs were greater for antigens of higher haemagglutinin antigen content. Fig. 2 shows the dose response reactions of known concentrations of purified haemagglutinin or purified virus particles after disruption with sodium sarcosyl sulfate. In the latter case the haemagglutinin content was estimated at 33% of the total weight of the virus particle. For both types of preparation, purified haemagglutinin or purified virus, the dose response curves had similar slopes and the actual zone annulus areas for both types of preparation when containing equi- valent haemagglutinin protein concentrations were similar, differing by no more than 10%. Since the standard antiserum 74/524 has been calibrated against purified haemagglutinin or purified virus of known haemagglutinin protein concentration an estimate of the haemagglutinin content of the refer- ence antigen 74/560 can be made (Fig. Ib). In the 75 T .- 25 IC 0 10 20 30 Zone annulus area (mm2) A 0 Purified A/PC/1/73 HA Purified AIPC/1/73 virus L/~~~~~~~ .// Fig. 2. Dose response curves for purified haemag- glutinin and purified virus in the microtest when con- centration of haemagglutinin protein is plotted against zone annulus area. 225 40 50 WHO 738 I4 G. C. SCHILD ET AL. Fig. 3a. A macrotest performed in a 9-cm petri dish. The gel contains 2.08 mI/I of antiserum G8. The wells in rows Ta and Tb contain dilutions of two test antigens whereas those in row R contain dilutions of reference antigen (74/560). E E_ c e c N Fig. 3b. Dose response curve for reference antigen (74/560) and two test vaccines (Vl and V2) in the macrotest when antigen dilution is plotted against zone annulus area. Fig. 4. Dose response curve for purified virus in the macrotest when concentration of haemagglutinin pro- tein is plotted against zone annulus area. 100 0 -Z75- E r 50 / 2 co/ ~0/ I I I I I 0 10 20 30 40 50 60 70 Zone annulus area (rmm2)I~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 226 Dilution of antigen *WO 7581s SINGLE-RADIAL-IMMUNODIFFUSION ASSAY OF INFLUENZA HAEMAGGLUTININ 227 Table 1. Standard deviations of annulus area measurements from macromethod Mean Standard Coefficient Dilution HA content annulus deviation of variation(mg/I) area (mm2) (SD) (SD/mean) Reference 1 :1 42.22 3.2141 7.61 % antigen 74/560 2:3 31.32 1.9645 6.27 % 1 :2 26.45 1.0184 3.85 % 1:4 18.50 2.1293 11.51 % 1 :8 14.07 1.0620 7.55% 1 :32 10.01 0.6221 6.21 % Purified A/ 100 65.41 2.6187 4.00 % Port Chalmers/ 1/73 (H3N2) 66.6 49.85 3.5916 7.20 % virus 50 42.06 3.5593 8.46 % 25 25.91 1.9824 7.65 % 12.5 17.29 1.2574 7.27 % 6.25 12.84 1.5765 12.28 % 3.125 11.52 1.1107 9.64% microtest, preparation 74/560 was estimated to con- tain 54.6 mg/l of haemagglutinin protein. For estimations of the haemagglutinin antigen content in preparations it is recommended that a series of dilutions (1: 1 to 1: 8) of the reference antigen be tested on the same immunoplate as dilutions of the test preparations. A plot of the zone annulus area responses against antigen concentra- tion is made for the reference and test antigens. Since the haemagglutinin antigen content of the reference preparation has been established, estimates of the haemagglutinin antigen content of test pre- parations may be made as follows when zone annu- lus areas of their responses are plotted against antigen dilution: HAa content of test _ DRSb of test antigen 54.6 antigen (mg/litre) DRS of reference antigen This estimate assumes a linear relationship be- tween zone area and antigen concentration and that the slopes of the dose responses curves for different antigens are proportional to antigen concentration. Alternatively, unknown antigen preparations may be assessed for haemagglutinin content simply by a HA = haemagglutinin. b DRS = dose response slope. determining the ratio of their potency to that of the working reference 74/560 based on measurements of the relative zone annulus areas of their reactions when tested on the same immunoplate. Macrotest A typical macrotest immunoplate is illustrated in Fig. 3a. A plot of zone annulus area against dilution of reference antigen 74/560 was also shown to be approximately linear over an eight-fold range of antigen concentrations when the macrotest was used (Fig. 3b). The dose response curve for a preparation of highly purified A/Port Chalmers/1/73 (H3N2) influenza virus of known haemagglutinin content is shown in Fig. 4. Test antigens may be assessed for haemagglutinin content in a similar way to that described for the microtest, when reaction zones are compared with those of the reference antigen. The reference antigen is estimated to contain 50.5 mg/ litre of haemagglutinin protein by the macrotest using the same comparisons as described above for the microtest. STATISTICAL EVALUATION OF THE METHOD Schild et al. (2) carried out a statistical evaluation of the accuracy and reproducibility of the single- G. C. SCHILD ET AL. radial-immunodiffusion microtest for haemagglu- tinin antigen assays. In these studies multiple tests were carried out by different operators in the same laboratory on a series of dilutions of two antigens. The coefficient of variation in the annulus area with the same antigen sample did not exceed 10% even when the tests were performed by different operators who prepared their own immunoplates. The accu- racy and reproducibility of the microtest has recently been examined more extensively in a collaborative study. Three test antigens were assayed against a reference antigen in 9 different laboratories. There was in general excellent agreement between labora- tories for each antigen assay. The results of this study will be described in full elsewhere. The accuracy and reproducibility of the macrotest was also assessed statistically. Table 1 shows an analysis of the data from stained plates which was used for Fig. 3 and 4 in which 8 tests were performed in the same laboratory. The coefficients of variation in the annulus area with either reference antigen or purified virus did not exceed 12 %. Similar values for the reproducibility of the macrotest were also ob- tained when unstained plates were examined. SPECIFICITY OF THE TEST It should be noted that the reagents 74/524 and 74/560 are intended for the assay of the haemagglu- tinin of A/Port Chalmers/1/73 virus (subtype H3) or very closely related strains. In the event of further antigenic ' drift ' in current influenza A viruses new reference reagents will be required. In addition anti- gen preparations containing influenza A viruses of haemagglutinin subtypes other than H3, or influ- enza B antigens, will not react in single-radial- diffusion immunoplates containing 74/524 anti- serum. A positive reaction is evidence that the antigen under test contains haemagglutinin of the H3 subtype; thus the test is of value in control procedures for the antigenic composition of influ- enza vaccines as well as of their potency. QUALITATIVE TESTS FOR HAEMAGGLUTININ For rapid qualitative tests for A/Port Chal- mers/l/73 haemagglutinin using the microtest, immunoplates may be prepared that contain higher concentrations of 75/524 serum than those used in the quantitative test described above. A concentra- tion of 0.3 ml of reconstituted serum 74/524 per 10 ml of gel is appropriate. Under such test condi- tions qualitative estimates of haemagglutinin may be made by the appearance of intense zones of opales- cence which do not require staining. Alternatively the macrotest may be used for both qualitative and quantitative tests of antigens, without the need for staining the plates. These procedures may be of value in a variety of situations where exact quantita- tion is not required. Examples of such situations might be in the routine process control during the production of influenza vaccines or for the detection of haemagglutinin in sample fractions collected dur- ing biochemical purifications of virus or haemagglu- tinin antigen-i.e., on sucrose density gradients. DISCUSSION The current recommendations for the standardiza- tion of influenza vaccines state that the amount of virus antigen shall be determined by haemagglutina- tion methods (4). In view of the acknowledged technical and theoretical disadvantages of tests based on haemagglutination, assay methods for haemagglutinin based on single-radial-diffusion are proposed as a possible alternative. When a reference antigen of known haemagglutinin protein content is assayed by single-radial-diffusion in parallel with test antigens, the haemagglutinin content of the test antigens can be determined accurately and repro- ducibly in different laboratories. The ability of vac- cines to form antigen-antibody complexes, as mea- sured in single-radial-diffusion tests, may be more relevant to the control of influenza vaccines than is the ability to agglutinate erythrocytes. Since single- radial-diffusion is dependent on the antigenic char- acter of the haemagglutinin the test gives additional control of the antigenic composition of vaccines. The preparation and standardization of influenza B haemagglutinin antigen and specific antiserum is in progress. With the appropriate reagents it will be possible to assay the haemagglutinin content of both influenza A and B components of a bivalent vaccine, independently of each other. The macrotest was developed to provide rapid, accurate, and reproducible assays of haemagglutinin antigen. The main features of the macrotest are that approximately twice the concentration of antiserum and three times the volume of antigen are used compared with those used in the microtest. The reaction zones are consequently more intense in the macrotest and results of assays can be read after 24 228 SINGLE-RADIAL-IMMUNODIFFUSION ASSAY OF INFLUENZA HAEMAGGLUTININ hours. This time compares favourably with the 3 days required before readings can be made in the microtest. The ability of a preparation of influenza virus or haemagglutinin to react in the single-radial-diffusion test is indicative of its antigenic activity but is not proof of the ability of the preparation to elicit an antibody response. Evidence for the immunogenicity of test antigen preparations should be sought by examination of sera from immunized laboratory animals. Studies in this laboratory are in progress to establish the dose relationship between the single- radial-diffusion response of influenza haemaggluti- nin preparations to their immunogenic activity in experimental animals and in man. Assays of the quantity of the other major antigens of the influenza virus particle, the neuraminidase, the nucleoprotein, and the M (matrix) protein, may be performed using the same principles of single-radial- diffusion in tests employing specific antisera to these components. Studies are currently in progress in our laboratories to assess the sensitivity and reproduci- bility of these methods and to develop appropriate reference standards and reagents. RI-SUMI TECHNIQUE D'IMMUNODIFFUSION RADIALE SIMPLE POUR LE TITRAGE DE L'HbMAGGLUTININE GRIPPALE Il est propos6 une microm6thode et une macromethode d'epreuve d'immunodiffusion radiale simple pour rem- placer dans certains cas les epreuves fond&es sur l'agglu- tination des 6rythrocytes pour le dosage de l'h6magglu- tinine grippale. Les deux methodes ont des qualit6s comparables de reproductibilit6 et de pr6cision, mais le macrotest est plus simple du point de vue technique et permet d'obtenir plus rapidement les resultats des titrages (en 24 heures). Dans les deux tests, I'anticorps a l'6gard de l'h6mag- glutinine purifi6e est incorpor6 A des gels d'agarose. L'h6magglutinine a titrer, qui est trait6e par un d6tergent afin que les sous-unites soient diffusibles, est vers&e dans des godets m6nages dans l'agarose. L'hemagglu- tinine diffuse radialement a partir des godets et sa reaction avec l'anticorps specifique produit une zone d'opalescence dans le gel. La surface de la zone de reaction est proportionnelle a la quantite d'antigene contenue dans le godet. Dans le microtest, le detergent utilise pour traiter l'antigene de r6f6rence et l'antigene & expertiser est le sulfate de sodium-sarcosyle (0,5% p/v) (a raison de 0,5 ml d'antigene pour 0,2 ml de d6tergent). Au bout de 30 minutes, on verse des volumes de 10 pl dans des godets de 3 mm de diametre pratiques dans un gel d'agarose a 1,5y% sur des immunoplaques vides de Hyland. Le gel contient 1,25 ,l de serum par ml de gel. On attend 24 heures pour permettre la diffusion des antigenes; les gels sont ensuite laves pendant 24 heures dans une eau physiologique tamponn6e au phosphate pour eliminer la prot6ine qui n'a pas reagi, puis color6s au bleu brillant de Coomassie (0,5% dans l'eau) et decolores par lavage a l'eau du robinet. Les zones peuvent alors etre mesurees avec un oculaire micrometrique et la surface de l'anneau peut etre lue sur un tableau preetabli. Dans le macrotest, le traitement au d6tergent des antigenes est fait dans les godets des immunoplaques. On verse des volumes de 20 pl de dilution d'antigene dans des godets de 4 mm de diametre pratiqu6s dans des gels d'agarose en boites de Petri de 9 cm. Le gel contient 2,08 yA d'antiserum. On ajoute des volumes de 10 IA de d6tergent et 24 heures plus tard, on peut mesurer les zones de reaction, mais cette mesure est plus facile si les gels ont et6 laves pendant plusieurs heures dans une eau physiologique tamponn6e au phosphate. Lorsque les valeurs de la surface de l'anneau sont port6es en fonction des concentrations d'h6maggluti- nine, les courbes dose-reponse pour l'antigene de ref6rence et l'antigene a titrer sont lin6aires. La pente de la courbe est une mesure directe de la quantit6 d'anti- gene. La concentration d'h6magglutinine pour les antigenes a expertiser peut etre obtenue en comparant les pentes des courbes dose-r6ponse avec celles de l'antigane de r6f6rence de teneur connue en hemagglu- tinine (54,6 lj/ml). Les courbes dose-r6ponse d'h6mag- glutinine purifi6e ou de virus purifi6 de teneur connue en h6magglutinine sont 6galement lin6aires lorsque les valeurs de la surface de l'anneau sont port6es en fonction des valeurs de la teneur en hemagglutinine. Les caract6ristiques quantitatives du macrotest et celles du microtest sont similaires. La concentration de l'antigene de ref6rence en hemagglutinine a ete estim6e a 50,5 ,ug/ml dans le macrotest. Les qualites de pr6cision et de reproductibilite des deux tests ont ete evaluees par des titrages multiples de dilutions d'antigenes pratiques dans le meme labora- toire. Dans le microtest, le coefficient de variation de la surface de l'anneau ne depassait pas 10% et dans le macrotest le coefficient de variation ne d6passait pas 12%. Les titrages d'antigenes par le microtest pratiques dans 9 laboratoires ont donne des resultats concordants 229 230 G. C. SCHILD ET AL. REFERENCES 1. PERKINs, F. T. & REGAMEY, R. H. (EDrTORS). Proceed- ings of an International Symposium on Influenza Vaccines for Men and Horses, London, November, 1972. Symposia series in immunobiological standardi- zation. Basel, S. Karger, 1973, Vol. 20 2. SCHILD, G. C. ET AL. In: Proceedings, International Conference on Standardization of Diagnostic Mate- rials, Atlanta, Georgia, USA, June, 1973. Atlanta, Center for Disease Control, 1974, pp. 243-258 3. MANCiNi, G. ET AL. Immunochemistry, 2: 235-254 (1965) 4. WHO Technical Report Series, No. 384, 1968 5. BRAND, C. M. & SKEHEL, J. J. Nature new biology, 238: 145-147 (1972) 6. AYMARD-HENRY, M. ET AL. Bulletin of the World Health Organization, 48: 199-202 (1973) 7. DULBECCO, R. Journal of experimental medicine, 99: 167-199 (1954) Annex I WORKING REAGENTS FOR SINGLE-RADIAL-DIFFUSION OF HAEMAGGLUTININ Reference antiserum 74/524 to purified A/Port Chal- mers/l/73 haemagglutinin This antiserum was prepared in goats to the purified haemagglutinin of A/Port Chalmers/1/73 (H3N2) virus, MRC11 strain. MRC11 is a recombinant between A/PR8/34 (HON1) and A/Port Chalmers/I/73 (H3N2). It contains haemagglutinin and neuraminidase antigens, H3N2, derived from A/Port Chalmers/l/73 and possesses the high yielding growth characteristics of A/PR8/34. The haemagglutinin antigen was extracted from purified MRC11 virus by treatment with bromelain (5) and was purified by rate zonal centrifugation on sucrose density gradients. Polyacrylamide gel analysis indicated that the purified HA preparation contained two glycopeptides with molecular weights of 58 000 and 28 000, respectively, which have previously been identified as structural com- ponents of the haemagglutinin (i.e., HAI and HA2), and no contaminating proteins. Two doses of approximately 300 ,xg of HA with Freund's complete adjuvant were given intramuscularly at 3-week intervals and serum was collected 10 days after the second immunization. Each ampoule of lyophilized serum (74/524) contains 0.5 ml of a 1: 10 dilution of goat serum and should be reconsti- tuted by the addition of 0.5 ml of sterile distilled water. Reference antigen 74/560 Each ampoule of lyophilized antigen contains 1.0 ml of a preparation of partially purified A/Port Chalners/1/73 (H3N2) virus (MRC11 strain) obtained by centrifugation of infected allantoic fluid on a continuous flow, zonal ultracentrifuge. Each ampoule should be reconstituted by adding 1.0 ml of sterile distilled water. HA content: 1 962 IU per ml Neuraminidase activity: 3 275 units per ml (ref. 6) Agar or agarose gels Agarose (A-37 Indubiose) from l'Industrie biologique francaise SA, Gennevilliers, France, is recommended for the single-radial-diffusion tests described. Gels are pre- pared that contain 1.5% agarose in phosphate buffered saline (7) at a final pH of 7.0-7.1; 0.1 % sodium azide is added to prevent microbial growth. The hot agar or agarose is filtered through a Green's clarifying filter after preparation and is stored at 4°C before use. SINGLE-RADIAL-IMMUNODIFFUSION ASSAY OF INFLUENZA HAEMAGGLUTININ Annex 2 RELATION BETWEEN ZONE DIAMETER AND ANNULUS AREA Annulus area (mm') 3-mm well a 4-mm wellb 0.479 0.974 1.48 2.01 2.55 3.11 3.68 4.27 4.88 5.50 6.13 0.63 6.79 1.29 7.45 1.95 8.14 2.64 8.84 3.34 9.55 4.05 10.28 4.78 11.03 5.53 11.79 6.29 12.57 7.07 13.36 7.86 14.17 8.67 14.99 9.49 15.83 10.33 16.69 11.19 17.56 12.06 18.45 12.95 19.35 13.85 20.27 14.77 21.21 15.71 22.16 16.66 23.12 17.62 24.10 18.62 25.10 19.60 26.11 20.61 27.14 21.64 28.19 22.69 29.28 23.78 30.32 24.82 31.42 25.92 32.52 27.02 33.65 28.15 34.79 29.29 35.94 30.44 37.11 31.61 38.30 32.80 39.50 34.00 40.72 35.22 41.95 36.45 Zone diameter (mm) 8.0 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9.0 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 10.0 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 11.0 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.9 Annulus area (mm") 3-mm well a 4-mm well b 43.20 44.46 45.74 47.04 48.35 49.68 51.02 52.38 53.75 55.14 56.55 57.97 59.41 60.86 62.33 63.81 65.31 66.83 68.36 69.91 71.47 73.05 74.64 76.25 77.88 79.52 81.18 82.85 84.54 86.24 87.96 89.70 91.45 93.32 95.00 96.80 98.61 100.44 102.29 104.15 37.70 38.96 40.24 41.54 42.85 44.18 45.52 46.88 48.25 49.64 51.05 52.47 53.91 55.36 56.83 58.31 59.81 61.33 62.86 64.41 65.97 67.55 69.14 70.75 72.38 74.02 75.68 77.35 79.04 80.74 82.46 84.20 85.95 87.82 89.50 91.30 93.11 94.94 96.79 98.65 a Microtest. b Macrotest. Zone diameter (mm) 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.0 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5.0 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6.0 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 231