Thomas P. Magill

A Virus from Cases of Influenza-like Upper-respiratory Infection

During the past few years at least two groups of investigators 1 , 2 have reported cases of upper-respiratory infection of unknown etiology, that resembled true “epidemic influenza” in clinical respects but which differed in that the convalescent serums failed to show any increase in capacity to neutralize standard strains of the influenza virus. Two similar cases of influenzalike infection occurred among the workers in this laboratory in February, 1940. The serums obtained from these cases 4 weeks after the infection and those obtained either at the time of or 3 weeks before the onset of illness were tested against the PR8 strain of the virus of epidemic influenza. Neither of the convalescent serums showed any detectable increase in capacity to neutralize the usual 1000 lethal doses of the virus; and neither of them fixed complement in tests against antigens prepared in the usual manner 3 from mouse lung suspensions. This apparent lack of development of antibodies reactive against this standard strain of influenza virus seemed to indicate that the influenza-like infection of neither of the cases had been due to the virus of epidemic influenza. The throat washings from both patients when inoculated intranasally in ferrets under ether anesthesia evoked a fever of the type considered characteristic of influenza. The infectious agent from one of the cases was transmitted serially in ferrets, but even after 10 passages the lungs of infected animals showed no consolidation. The infection was transmitted by intranasal inoculation from ferrets to Swiss mice; in that species blue-gray areas of pulmonary consolidation were produced on the first passage. After 9 passages the virus increased sufficiently in virulence to kill some of the mice but although it has now been passed over 20 times it has not yet acquired the capacity of regularly killing all of the mice that are inoculated.

Antigenic Differences in Strains of Human Influenza Virus

On the basis of earlier studies with mouse and ferret passage virus, 1 it was concluded that the Puerto Rico (PR8) and the Philadelphia (Phila) strains of human influenza virus were immunologically identical, while the swine influenza virus was serologically distinct. The 2 strains of human virus were also indistinguishable from the WS strain of the English workers. 1 2 After repeated inoculation of ferrets with human influenza virus, however, it was noted that the serum of an animal so treated developed the capacity of neutralizing the swine virus as well. 3 Moreover, the serum of rabbits (a non-susceptible animal species) vaccinated with ferret-passage human influenza virus developed antibodies against both the human and swine viruses, whereas rabbits vaccinated with swine influenza virus produced antibodies which neutralized only the swine virus. 3 Identical results were obtained with horse sera prepared by Laidlaw, Smith, Andrewes and Dunkin. 3 4 It was suggested, therefore, that the human and swine viruses, while immunologically distinct contained common antigens and that the swine antigenic components were present in the human virus as secondary antigens, 3 5 . Since it seemed likely that the antibody response of an insusceptible animal might reflect the secondary antigens of the virus more completely than that of the susceptible animal in which antibodies to the primary antigen rather than to the secondary antigen would probably constitute the initial response, further studies were carried out in rabbits using the tissue culture virus as the source of the various strains. Rabbits were inoculated intraperitoneally with 2 cc. of culture fluid containing the PR8, Phila, and the Swine-2 strains, respectively. Blood was taken from the marginal ear vein 9 to 15 days later.

Direct Transmission of Human Influenza Virus to Mice

Repeated attempts have been made to isolate human influenza virus by the direct inoculation of mice with material obtained from patients acutely ill with the disease. Up to the present, however, the establishment of experimental influenza in mice has been successful only when the virus was first subjected to primary passage in ferrets. In December, 1936, strains of human influenza virus were isolated from fresh cases of influenza occurring in New York. Following one ferret passage, the strains were transferred to mice and became so rapidly adapted to this species of animal that deaths occurred among the mice of the second passage. The ease with which these results were obtained heightened the possibility of transmitting the virus infection directly from human throat washings to mice without intermediate ferret passage. With plain meat infusion broth as the medium, nasal and pharyngeal washings were collected from 2 patients acutely ill with influenza. The washings were pooled. One portion was used for ferret inoculation and yielded a strain of virus which was readily transferred from ferret to mice. The remainder was stored in the frozen state at —78°C. for 19 days. It was then removed from storage, allowed to thaw at room temperature and centrifugalized at 2500 revolutions per minute for 30 minutes. The supernatant fluid was then centrifugalized in the air-driven centrifuge 1 at 14,000 r.p.m. for 3 hours. The supernatant fluid was removed, the sediment was resuspended in the supernatant to 1/6 the original volume and anesthetized mice were then given 0.05–0.08 cc. of the material intranasally. Serial passages were made in mice at 4 to .i-day intervals. A heavy suspension (50 to 7070) of the finely ground lung tissue from inoculated mice was given intranasally to normal mice. During the first 3 passages no significant pulmonary lesions were observed. It was shown, however. that virus n-as present in the lungs of the third passage mice, since a ferret inoculated intranasally ivith a suspension of these lungs dei-eloped the characteristic signs of experimental influenza and subsequently developed neutralizing antibodies against human influenza virus.

Direct Isolation of Human Influenza Virus in Tissue Culture Medium and on Egg Membrane

Throat washings in Tyrodes solution were obtained on the second day of illness from a patient acutely ill with influenza. The material was centrifugalized at 2500 revolutions per minute for 30 minutes. The supernatant fluid was then filtered through a graded collodion membrane of 500 mμ average pore size. 1 Flasks containing 4.5 cc. of chick embryo Tyrode medium 2 were inoculated with 2.5 cc. amounts of the filtrate (shown by ferret inoculation to contain active virus). Transfers of 0.5 cc. amounts of the culture material to 4.5 cc of fresh medium were made at 48-hour intervals. Mice were inoculated intranasally with culture fluid of the 5th transfer. Mouse passages were made at 4-day intervals. No significant lesions were seen in the lungs of mice of the first 3 passage groups. In mice of the 4th passage, however, sugestive lesions were seen; these were more definite in the 5th passage, and in the 6th passage death of the mice with extensive pulmonary involvement occurred. The virus was identified as human influenza virus by means of neutralization tests with known immune serum. These results indicate that virus was recovered directly from the throat of a human influenza patient by the introduction of the filtered throat washings of the patient into tissue culture medium. The virus not only survived but probably multiplied, since the final dilution of the original material was approximately 1:300,000 at the time the culture material was first given to mice. The behavior of the virus after its inoculation into mice was very similar to that of virus established directly in mice from human throat washings. 3 Burnet has reported in detail the behavior of human influenza virus introduced after passage through ferrets, onto the chorioallantoic membrane of the developing chick. The direct cultivation of the virus of common cold on the chorio-allantoic membrane has also been repeated. Attempts were therefore made to utilize this procedure in the isolztion of virus directly from the human patient.

A Flocculation Phenomenon with Human Sera and Suspensions of the Virus of Epidemic Influenza

Neutralization and complement-fixation tests with the virus of epidemic influenza have been described and studied rather extensively but no agglutination, precipitation, or flocculation reaction has as yet been reported. Theoretically such reactions should be demonstrable, provided a sufficient quantity of virus is contained in the antigen. 1 Attempts to purify and concentrate the virus by ultracentrifugation in this laboratory have not in general yielded suitably reactive antigens. During the course of investigation it was observed, however, that a flocculative phenomenon could be obtained consistently with certain human sera and crude suspensions of lungs of mice infected with influenza virus, the reaction being especially marked with sera of patients convalescent from epidemic influenza. It is the purpose of this report to describe the conditions under which the reaction occurs and to discuss briefly some of the results. Methods. Mice of the Swiss strain are inoculated intranasally with 1 to 2% suspensions of lungs of mice infected with PRS strain of influenza virus. At the time of appearance of visible areas of pulmonary consolidation—approximately 40 to 50 hours after inoculation—the mice are sacrificed and the lungs ground with alundum in meat-infusion broth (pH 7.8) to form a 10% suspension by weight. After light centrifugation equal volumes of supernatant fluid and serum are mixed in small tubes and incubated at 37°C. After 1 hours incubation the tubes are shaken lightly and the degree of flocculation is recorded in the usual manner. Frequently flocculation does not occur until the tubes are shaken but once the reaction takes place a precipitate settles to the bottom of the tubes in large coarse floccules. Meat-infusion broth is not essential in the preparation of the antigen and can be replaced by M/150 phosphate buffer pH 7.2 or 0.4% NaCl solutions.

Two Outbreaks of Influenza Caused by Antigenically Different Viruses

Early studies on influenza suggested that the virus agents of the disease, although not antigenically identical were closely enough related for the infection to evoke a demonstrable rise in antibodies reactive against the recognized strains of the influenza virus. The first convincing indication that influenza might be due to antigenically unrelated viruses was presented by Stuart-Harris, Smith and Andrewes, 1 who found that a rise in antibodies against the usual strains of influenza virus occurred in only 33% of the cases they studied in England in 1939; although they failed to isolate the actual virus, they concluded that a number of their cases must have been caused by an agent other than the usual influenza virus. More direct evidence has recently been obtained independently by Francis and by ourselves. In February, 1940, we 2 isolated a virus (termed TM) from 2 cases of influenzå and showed that in each instance the infection evoked an increase in antibodies against the homologous (TM) virus but caused no detectable increase in antibodies reactive against the PR8 3 strain of influenza virus. At about the same time Francis 4 isolated a strain (termed “Lee”) which was also distinct from the PR8 and from other previously recognized strains. He showed that the convalescent serums from cases of influenza which had occurred in various places during the early part of 1940 and also serums from 2 cases that had occurred in 1936 had increased titers of antibodies against his Lee strain but not against the PR8 strain. Francis concluded that his Lee strain was a new type of virus and proposed that it be termed influenza B to distinguish it from the older strains (PR8, WS, etc.) for which the term influenza A had already been suggested.