Induction of an Inhibitor of Influenza Virus Hemagglutination

APPLIED MICROBIOLOGY, Apr. 1968, p. 563-568 Copyright @ 1968 American Society for Microbiology Vol. 16, No. 4 Printed in U.S.A. Induction of an Inhibitor of Influenza Virus Hemagglutination by Treatment of Serum with Periodate W. B. BEARDMORE, K. V. JONES, T. D. CLARK, AND E. K. HEBEKA Research Laboratories, Parke, Davis & Co., Detroit, Michigan 48232 Received for publication 3 January 1968 Serum is commonly treated with potassium periodate to destroy nonspecific inhibitors of influenza virus hemagglutination. We have observed, however, that such treatment of serum without pre-existing inhibitor produced high titers of inhibitor against certain strains of influenza virus. Inhibitor was induced in the serum from several different animal species but not in hamster or mouse serum. Periodate treatment of serum albumin, fraction V, from several animals, including man, creates this inhibitor. Our data indicate that the inhibitor induced in the serum of various animal species differs in its mechanism of induction and in its resistance to receptor-destroying enzyme and trypsin. Hemagglutination by B/Singapore/3/64, B/Colorado/2/ 65, B/Georgia/1 /65, and B/Massachusetts/3/66 strains of influenza virus is inhibited by periodate-treated human serum at dilutions as high as 1:5,120. The routine use of periodate treatment of serum in diagnostic and surveillance studies of influenza virus infections is not recommended. Certain strains of influenza virus are extremely sensitive to nonspecific,inhibitors in serum. These inhibitors cannot be removed by the use of heat, trypsin, or potassium periodate, individually or in combination, but they are reduced by receptordestroying enzyme (RDE) and are removed completely by treatment with dry ice (7). When we were preparing and testing standardized lots of B/Singapore/3/64 influenza virus seed and specific rooster antiserum, we observed the apparent induction of nonspecific inhibitor by treatment of the serum with potassium periodate. While this study was in progress, Mascoli et al. (6) reported a similar phenomenon with chicken and human serum with the 1965 influenza B viruses. The finding that periodate induces inhibitor is contrary to all other published studies on the elimination of nonspecific serum inhibitors. Potassium periodate has generally been considered to be an accepted method of removing nonspecific inhibitors (5). This report characterizes the induction and certain properties of this inhibitor. MATERIALS AND METHODS Viruses. All influenza virus strains used in this study were obtained from the Virus and Rickettsia Section of the National Communicable Disease Center, Atlanta, Ga. These seed viruses were passed in 10-day-old embryonated chicken eggs until satisfactory hemagglutination and infectivity titers were produced. The viruses either were stored frozen at -56.6 C or lyophilized and stored at 4 C. Treatment of serum: Potassium periodate (). We mixed one volume of phosphate-buffered saline (PBS) and two volumes of serum and maintained these mixtures at 56 C for 30 min. Then we added six volumes of 0.011 M and the mixture was held at room temperature for 15 min. Six volumes of 1% glycerol in distilled water was added, and the mixture was maintained at room temperature for an additional 15 min. The required dilution was then prepared with PBS. RDE (cholera filtrate). RDE (four volumes) containing 100 units/ml and one volume of serum were mixed and incubated overnight in a water bath at 37 C. Then, three volumes of 2.5% sodium citrate was added, and the mixture was held at 56 C for 30 min. The desired final dilution was prepared by the addition of PBS. Trypsin. We used a modification of the method described by Lennette and Schmidt (5). An amount of 0.5 ml of 0.1 M PBS, ph 7.2, containing 4 mg of tryp- 563 sin (Difco, 1:250), was added to 1.0 ml of serum at room temperature. This mixture was immediately placed in a 56 C water bath for 30 min, and PBS was then added to produce the desired final dilution. A fresh trypsin solution was prepared daily. Dry ice (CO2). Distilled water (0.8 ml) was mixed with 0.1 ml of serum previously heated at 56 C for 30 min. A small pellet of dry ice was then added to this serum. After activity ceased, the resulting precipitate was removed by centrifugation. An amount of 0.1 ml of 8.5% NaCl was added to the supernatant fluid; this made the 1:10 dilution from which the desired dilutions were prepared.

564 BEARDMORE ET AL. APPL. MICROBIOL. Kaolin. A 0.9-ml amount of a 20% suspension of acid-washed kaolin was added to 0.3 ml of serum, and the mixture was shaken every 5 min for 20 min. The kaolin was removed by centrifugation at 1,500 rev/min for 10 min. Hemagglutination-inhibition (HI) test. Twofold dilutions of serum, with PBS as diluent, were prepared; 0.25 ml was used in each tube. To each tube, 0.25 ml of virus containing four hemagglutination units was added, and the tubes were incubated at 37 C for 1 hr. A 0.5-ml amount of 5% human "O" erythrocytes was then added to each tube. The results were read after incubation at room temperature for 1 hr. RESULTS Table 1 summarizes our initial observations on the induction of inhibitor of B/Singapore/3/64 hemagglutination by treatment of heterologous influenza rooster antisera. Treatment with resulted in a 30- to 60-fold increase in hemagglutination titer over simple heat treatment of the serum. Since treatment utilizes KI04, glycerol, and heat, we attempted to determine which of these factors was responsible for the induction of the inhibitor. The results (Table 2) revealed that, of all the various procedures listed, heating the serum for 1 hr at 56 C followed by treatment produced the greatest titer of the inhibitor of B/Singapore/3/64 hemagglutination. The KI04 treatment had no detectable influence on homologous HI titer against A/Equine-2/Miami antigen. Glycerol alone, alone, and plus glycerol had no detectable effect on hemagglutination of B/Singapore/3/64 virus in the absence of serum. When the same A/Equine-2/Miami antiserum pool was treated by other common procedures (heating plus dry ice, heating plus RDE, and trypsin), no inhibitor was produced against B/ Singapore hemagglutination. Three different lots of KI04 were tested for their capacity to induce the inhibitor in normal TABLE 1. Induction of nonspecific inthibitor oj B/Singapore/3/64 influenza virus hemagglutination of human "0" erythrocytes by KI04 treatment ofheterologous rooster antiseruma Hemagglutina- Antiserum Serum treatment tion-inhibition treatment ~~ti ter A/WS/33... 56 C, 1 hr 1:80 A/WS/33... 1:2,560 A/Equine-2/Miami.. 56 C, 1 hr 1:40 A/Equine-2/Miami... 1:2,560 a With all antisera used the antigen was B/Singapore. TABLE 2. Influence of heat, KI04, and glycerol on nonspecific inhibitor of B/Singapore influenza virus in A/Equine-2/Miami rooster antiserum Serum treatment Hemagglutination-inhibition titer A/Equine-2/ Miami antigen B/Singapore antigen Unheated...... 1:320 < 1: 20 0.5 hr at 56C. 1:320 1.0 hr at 56 C... 1:320 1.0 hr at 56 C + (no glycerol). 1:320 1: 1,280 1.0 hr at 56 + glycerol (no ). 1:320 1.0 hr at 56 C + + glycerol..... 1:320 1:1,280 Unheated + KIO4. 1:320 1:320 Unheated + KIO4 + glycerol... 1:640 1:320 Unheated + glycerol. 1:320 chicken serum. All three lots gave identical results (Table 3). The influence of various concentrations of in the induction of inhibitor in normal chicken serum is also presented in Table 3. The concentrations that produced the highest titers were 0.033 and 0.011 M. We carried out a study to determine the effects of other procedures commonly used to destroy nonspecific inhibitors on normal chicken serum before and after treatment of the serum. The influence of such treatment on the final resulting titer of the inhibitor with B/Singapore as challenge virus is given in Table 4. Our salient observations showed that trypsin and RDE treatment of the serum before treatment prevented the induction of the inhibitor by, however, when applied after treatment, destruction of the inhibitor which was induced occurred. The site of action of the -induced inhibitor is unknown. We attempted to determine whether the inhibitor blocks hemagglutination by adsorption onto the virus particles or by adsorption onto the receptor sites on the surface of the erythrocytes. B/Singapore/3/64 influenza virus and a 1:160 dilution of -treated normal chicken serum were mixed together and incubated at 37 C for 30 min; the virus was removed by centrifugation in a Spinco ultracentrifuge. The HI activity of the adsorbed serum against B/ Singapore/3/64 virus was then determined. The results indicated that the B/Singapore/3/64 virus removed the -induced inhibitor from the serum (Table 5). To determine whether the inhibitor was adsorbed onto erythrocytes, 0.2 ml of normal

VOL. 16, 1968 INHIBITOR OF INFLUENZA VIRUS HEMAGGLUTINATION 565 TABLE 3. Induction of B/Singapore inhibitor with different lots and concentrations of Hemagglu- K1I4 concn (M) Lot no. and supplier tinationtiter 0.011 26408, J. T. Baker Chemi- 1:640 cal Co., North Phillipsburg, N.J. 0.011 27120,J. T. Baker Chemi- 1:640 cal Co. 0.011 109065, G. Frederick 1:640 Smith Chem, Columbus, Ohio 0.033 26408, J. T. Baker Chem- 1:80 0.017 26408,J. T. Baker Chem- 1:640 0.056 26408,J. T. Baker Chem- 1:320 0.018 26408, J. T. Baker Chem- 1:40 None (heat 1:40 1 hr) Untreated TABLE 4. Effect of various treatments on periodateinduced inhibitor in normal chicken serum First Heat KIO4 Trypsin RDE Dry ice KI04 Kaolin KI04 Serum treatment Second Heat Trypsin KIO4 RDE Dry ice KI04 Kaolin Hemagglutinationinhibition titer 1:640 1:640 1:160 1:320 1:640 1:40 TABLE 5. Adsorption of KI04-induced inhibitor with B/Singapore/3/64 influenza virus chicken serum, treated with, was diluted to with PBS. A 10% suspension (2 ml) of human "O" cells was added to the serum and was incubated at 37 C for 1 hr. The cell-serum mixture was then centrifuged at 2,000 rev/min for 10 min, and the supernatant fluid was drawn off and discarded. A 0.5% suspension of the erythrocytes from the pellet was made and used in hemagglutination test with B/Singapore/3/64 influenza virus. A parallel test was run with untreated cells as a control. There was no evidence of adsorption of this inhibitor onto human "O" erythrocytes. A virus preparation had a hemagglutination titer of 1: 320 when tested with normal erythrocytes and eryth- ~~~Hemagglutina- ~~~~~~~~titer Normal chicken serum treatment Centrifugation tion-iniibition + virus... 1 hr at 30,000 rev/ <1:160 min Untreated + virus... 1 hr at 30,000 rev/ <1:160 (no min virus)... 1 hr at 30,000 rev/ 1:640 min (no virus)... None 1:640 Untreated... None < 1:160 rocytes previously treated with -processed serum. To establish whether the -induced inhibitor had any effect on the infectivity of a virus, doubling dilutions (1 :10 to 1:2,560) of periodatetreated normal chicken serum (which had an inhibitor titer against B/Singapore hemagglutination of 1:640) were incubated with an equal volume of B/Singapore virus (560 EIDo) for 1 hr at 37 C. Each of six 10-day-old embryonated chicken eggs was then inoculated in the allantoic sac with 0.2 ml of each dilution of the serum-virus mixture. The eggs were incubated at 35 C for 48 hr and refrigerated overnight; the allantoic fluid was harvested and tested for hemagglutination. All eggs were positive; this indicated that there was no inhibition of infectivity by the chicken serum before or after treatment even at a 1:10 dilution. Serum from eight adults and six children were tested to determine whether KI04 treatment induced inhibitor in human serum, as well as in chicken serum. The untreated sera had HI titers against B/Singapore virus of to 1:80. After treatment, the HI titer of all of the sera had increased to 1:1,280 to 1:2,560. Other strains of influenza virus reported to be very sensitive to nonspecific inhibitors (M. T. Coleman, Respiratory Virus Infections Unit, National Communicable Disease Center, personal communication) were also tested with - treated normal chicken and human sera (Table 6). treatment of all human and chicken serum induced an inhibitor against B/Colorado/2/65 and B/Georgia/i/65. An increase in the inhibitor of the B/Massachusetts/3/66 strain occurred in the treated human serum but not in the treated chicken serum. Normal chicken serum treated with did not inhibit hemagglutination by the following strains of influenza viruses: A/PR-8/34, Ai/

566 BEARDMORE ET AL. APPL. MICROBIOL. TABLE 6. Inhibition of different strains of influenza virus by KI04-treated chicken and human sera Serum no. Serum treatment Influenza virus strain Hemagglutination inhibition titer KI04-treated chicken serum B/Colorado/2/65 1:640 RDE-treated chicken serum B/Colorado/2/65 -treated chicken serum B/Georgia/1/65 1:320 RDE-treated chicken serum B/Georgia/l/65 < KI04-treated chicken serum B/Massachusetts/3/66 1:40 RDE-treated chicken serum B/Massachusetts/3/66 254 C -treated children's serum B/Massachusetts/3/66 1:320 254 C Untreated children's serum B/Massachusetts/3/66 254 C -treated children's serum B/Colorado/2/65 1:5,120 254 C Untreated children's serum B/Colorado/2/65 1:160 254 C -treated children's serum B/Georgia/1/65 1:28560 254 C Untreated children's serum B/Georgia/l/65 1:80 44 A -treated adult human serum B/Massachusetts/3/66 1:640 44 A Untreated adult human serum B/Massachusetts/3/66 1:40 44 A KI04-treated adult human serum B/Colorado/2/65 1:5,120 44 A Untreated adult human serum B/Colorado/2/65 1:160 44 A -treated adult human serum B/Georgia/1/65 1:5,120 44 A Untreated adult human serum B/Georgia/l/65 1:160 TABLE 7. Influence of receptor-destroying enzyme (RDE) and trypsin treatment of human serum, before and after KI04 treatment, on the inhibition of B/Singapore/3/64 influenza virus hemagglutination Serum Untreated KI04 + RDE RDE + KIO KI04 + trypsin Trypsin + K14 Adult... 1:40 1:2,560 1:80 1:640 1:640 Adult... 1:80 1:1,280 1:80 1:1,280 1:5,120 1:2,560 Child... 1:2,560 1:40 1:80 1:640 1: 1,200 Child... 1:2,560 1:40 1:320 1:640 1:1,280 Child... 1:2,560 1:40 1:80 1:12,80 1:1,280 TABLE 8. Induction of inhibitor i) n serum from human adult and children's sera before and after various animal species treatment was examined. A summary of the influence of these treatmeats on inhibition of B/ Hemagglutination-i nhibition titer Singapore virus is given in Table 7. Apparently, Animal Human... Rabbit... Horse... Guinea pig... Rat... Chicken... Monkey. Fetal calf... Calf... Goat... Hamster... Mouse... Untreated 1:160 1:40 -treated RDE treatment before treatment of the serum permitted the induction of only low levels 1:2,560 of the inhibitor. RDE destroys almost all of the 1:1,280 KI04-induced inhibitor. In contrast, trypsin treat- 1:640 ment before or after K04 treatment has relatively 1:640 little effect. 1:640 Since both chicken and human serum developed 1:320 high titers of inhibitor to B/Singapore hemagglu- 1:160 tination after treatment, we tested serum 1:160 samples from several other animal species to 1:160 determine whether this same phenomenon would 1:80 occur. There was great variation among these different sera in their capacity to form inhibitor <* 20 as a result of KI04 treatment. The spectrum ranged from human serum, which consistently FM-1 /47, A/Swine/31, A2/Japan /170/62, A2/ showed the greatest increase in inhibitor, to Japan/305/57, B/Maryland/i /59, B/Taiwan/2/ hamster and mouse serum, neither of which 62, and B/Lee/40. demonstrated a detectable increase in inhibitor The effect of RDE and trypsin treatment of (Table 8).

VOL. 16, 1968 INHIBITOR OF INFLUENZA VIRUS HEMAGGLUTINATION 567 In an effort to determine which human serum substrate the periodate converted to inhibitor, preparations of serum albumin (1:10) and -yglobulin (1:100) were titrated for inhibitor after periodate treatment and after RDE treatment. There was no detectable inhibitor induced in the albumin or y-globulin after RDE treatment. However, after periodate treatment, the -y-globulin was not inhibitory, but the serum albumin inhibited B/Singapore virus hemagglutination at 1:640. This suggested that the -converted serum albumin into a hemagglutinationinhibitor. The concentration of serum albumin normally found in human serum is 35 mg/ml; therefore, to determine the influence of the concentration on the titer of the -induced inhibitor, we treated a series of dilutions of serum albumin with (Table 9). The results demonstrated that the treatment of higher concentrations of serum albumin with produced higher titers of the inhibitor; but, at normal serum levels, an inhibitor with a titer of only 1:80 was induced. These data suggest that another component or components of serum may be acted upon by to induce the usually observed titer of 1:1,280. The serum albumin from representative animal species whose serum showed high or poor capacity to produce inhibitor by treatment was tested at a concentration of 35 mg/ml before and after treatment. The titers of B/Singapore TABLE 9. Induction of inhibitor by treatment of various concentrations of human serum albumin, fraction V Concn of albumin Hemagglutination-inhibition titer mg/ml 20 1:40 35 1:80 70 1:320 140 1:1,280 TABLE 10. inhibitor that were induced are summarized in Table 10. Human and chicken serum albumin showed an increase in inhibitor, and the sera from both of these species demonstrated a considerable increase in inhibitor titer after treatment (Table 8). Mouse serum albumin did not have any detectable inhibition of hemagglutination before or after treatment. Apparently treatment of mouse serum did not induce any detectable inhibitor (Table 8). The correlation between the induction of inhibitor in serum and serum albumin nevertheless breaks down; in the case of the horse and bovine serum albumin, however, in which no inhibitor was induced, both sera showed a great increase in inhibitor after treatment. These data suggest that more than one component of serum may be converted into inhibitor, and this may vary from one species to another. DISCUSSION The induction of an inhibitor of influenza virus hemagglutination by treatment of mucoids with has been recognized for a long time. In 1948, Burnet (1, 2) reported that purified cyst mucoid, which in the unmodified state has very little inhibiting effect on the A/W SE strain of influenza virus, could be converted into a strong inhibitor of the virus by treatment with periodate. Other similar investigations on other purified or semipurified mucoids were subsequently published (3, 4). The inhibitor induced by periodate treatment of serum differs from the inhibitor described by Burnet. He found that a sharply defined level of periodate was required to convert cyst mucoid into an active inhibitor of virus hemagglutination and that when this critical concentration was doubled it destroyed the inhibitor. The concentration of periodate required for the induction of inhibitor in serum is not critical, and doubling the optimum concentration does not destroy the Induction of inhibitor by KI04 treatment of serum albumin, fraction V, from various animal species Animal Serum supplier Jiemagglutination- inhibition titer Untreated K IO-treated Human Nutritional Biochemicals Corp., Cleve- 1:80 land, Ohio Chicken Nutritional Biochemicals Corp. < 1:80 Bovine Armour Pharmaceut, Kankakee, Ill. Horse... Nutritional Biochemicals Corp. < Mouse... Pentex, Inc., Kankakee, 111. < <

568 BEARDMORE ET AL. APPL. MICROBIOL. inhibitor (Table 3). Burnet also showed that the inhibitor created from mucoproteins prevented infection by influenza virus, as well as inhibiting hemagglutination of erythrocytes. The inhibitor created by treatment of serum with periodate does not have any detectable influence on the infectivity of B/Singapore/3/64 influenza virus, although it strongly inhibits viral hemagglutination. Possibly, the inhibitor-coated virus either enters into chicken embryo cells where the complex is broken, and the virus particles initiate infection or, perhaps, dissociation of the inhibitor occurs at the cell membrane. Aside from the unusual method of creation of this inhibitor, it has many characteristics in common with other well-known serum inhibitors. This inhibitor occurs in the serum of a variety of animal species and is destroyed by conventional methods used to remove nonspecific inhibitors, e.g., RDE and trypsin. The action of the inhibitor appears to be on the virus rather than on the erythrocyte surface. The components of serum which are converted by periodate into this inhibitor are unknown, although our data suggest that serum albumin of humans and certain animals plays some role. The mechanism of inhibitor creation may differ from one animal species to another for the following reasons. (i) RDE treatment of both normal chicken and human sera before treatment blocks the formation of the inhibitor. RDE also destroys the induced inhibitor. Trypsin has the same effect on chicken serum but has relatively little effect on human serum. (ii) Periodate treatment of normal chicken serum does not induce inhibitor against B/Massachusetts/3/66 influenza virus, although the same treatment of human serum does induce a high titer of inhibitor against this virus. (iii) Serum from a variety of relatively unrelated animal species, such as man, rabbits, horse, guinea pig, rat, and chicken, produces high titers of an inhibitor to B/Singapore influenza virus hemagglutination following KI04 treatment. No inhibitor is produced by the same treatment of the sera from the hamster and mouse. (iv) Periodate treatment of serum albumin, fraction V, from humans and chickens (both of which show induction of serum inhibitor by ) induces inhibitor, but periodate treatment of serum albumin, fraction V, of cows and horses (which also show induction of serum inhibitor by ) does not induce inhibitor. An immediate practical application of these findings would appear to be the invalidation of the routine periodate treatment of serum to remove nonspecific inhibitors of influenza virus in surveillance and diagnostic studies. One strain of influenza virus isolated in 1964, two strains isolated in 1965, and one strain isolated in 1966 were found to be strongly inhibited by this periodate-induced inhibitor. All of the viruses are B strains, but the sensitivity of all future isolates to this inhibitor would have to be determined before periodate heat treatment of human or animal serum could be utilized. ACKNOWLEDGMENT This study was conducted under contract PH43-62-841 with the Research Reference Reagents Branch of the National Institute of Allergy and Infectious Diseases. LITERATURE CITED 1. BURNET, F. M. 1948. Mucins and mucoids in relation to influenza virus action. Inhibition by purified mucoid of infection and haemagglutination with the virus strain WSE. Australian J. Exptl. Biol. Med. Sci. 26:381-387. 2. BURNET, F. M. 1948. Mucins and mucoids in relation to influenza virus action. The destruction of "Francis Inhibitor" activity in a purified mucoid by virus action. Australian J. Exptl. Biol. Med. Sci. 26:389-402. 3. BURNET, F. M. 1949. The effect of periodate on the virus-inhibiting qualities of mucoids in solution. Australian J. Exptl. Biol. Med. Sci. 27:361-374. 4. BURNET, F. M. 1951. Mucoproteins in relation to virus action. Phys. Rev. 31:131-150. 5. LENNEFFE, E. H., AND N. J. SCHMIDT. 1964. Diagnostic procedures for viral and rickettsial diseases, 3rd ed., p. 38. American Public Health Association, Inc., New York. 6. MASCOLI, C. C., M. B. LEoGus, AND M. R. HILLE- MAN. 1966. Influenza B in the Spring of 1965. Proc. Soc. Exptl. Biol. Med. 123:952-960. 7. ROBINSON, R. Q. 1965. Communicable Disease Center Influenza-Respiratory Disease Surveillance Report No. 81 IV, Laboratory Rept. 21. National Communicable Disease Center, Atlanta, Ga.