COMPARISONS OF INFLUENZA VIRUS STRAINS FROM THREE EPIDEMICS

Some of the peculiarities of strains of influenza A and B virus from two epidemics have been described. The influenza B virus of 1945–46, when compared with influenza A virus, proved to be much more difficult to isolate from human sources by any known means. Its adaptation to the chick embryo (by any route) or to mice was much slower than that of A virus. It did not keep nearly as well on storage at –72°C. either in throat garglings or as passage material. Its adaptation to amniotic growth was usually much better than to allantoic growth even after repeated allantoic passages. It failed to show primary evidence of occurring in the O form, although many of the secondary O characteristics were present and persisted. Its titer in throat washings was not demonstrably high as compared with certain strains of A virus, which were demonstrated in garglings at dilutions of 10^–5 and 10^–6. The antigenic patterns of influenza A strains from two epidemics were compared. No antigenic differences of significant degree were found among the strains of either epidemic and the difference between the strains of the two epidemics was very slight. A similar study was made of the influenza B strains of the epidemic of 1945–46. This also showed complete lack of significant strain differences. The implications of these findings for influenza prophylaxis are discussed.     

Influenza virus infection in the hamster; a study of inapparent virus infection and virus adaptation

A STUDY OF INFLUENZA VIRUS INFECTION IN THE HAMSTER HAS YIELDED THE FOLLOWING RESULTS: 1. Two influenza A strains (Ga. 47 and PR8) multiplied readily in the hamster lung, although no lung lesions were produced during six serial passages. On further passage both viruses abruptly acquired the capacity to produce pulmonary consolidation and death of the animals. 2. Extracts of the lungs during the early passages contained complement-fixing antigen and infectious virus, as revealed by titration in mice and embryonated eggs. Agglutinins for chicken, human, and guinea pig red cells, however, were not demonstrable at this time. With further passage a close correlation was observed between the capacity of the virus to produce lung lesions in the hamster and to agglutinate mammalian types of red cells. In addition, quantitative changes in the virus population were demonstrated in the lung extracts by complement fixation tests and titrations in mice and eggs. 3. Incubation at 37 degrees C. was effective in bringing out agglutinins in high titer for chicken red cells in lung extracts, which originally failed to agglutinate chicken cells but agglutinated mammalian red cells. This method did not increase the titers of mammalian cell agglutinins. 4. The body temperature of the hamster was found to decrease within 1 to 4 days after inoculation of influenza virus. In the early passages the temperature returned to normal within 24 hours, but with the development of the pathogenic strain of virus the temperature remained at subnormal levels until death. 5. The Lee strain of influenza B virus produced pulmonary lesions in the hamster on the first passage and no increase in pathogenicity of the virus occurred during eleven serial passages. Virus was demonstrable in extracts of the lungs by all the methods used and no difference was observed in its capacity to agglutinate fowl and mammalian types of red cells. The implications of these findings are considered briefly in the discussion.     

THE INFECTION OF MICE WITH SWINE INFLUENZA VIRUS

The experiments confirm the earlier observation of Andrewes, Laidlaw and Smith that the swine influenza virus is pathogenic for white mice when administered intranasally. Two field strains of the swine influenza virus were found to differ in their initial pathogenicity for mice. One strain was apparently fully pathogenic even in its 1st mouse passage while the other required 2 or 3 mouse passages to acquire full virulence for this species. Both strains, however, were initially infectious for mice, without the necessity of intervening ferret passages. There is no evidence that bacteria play any significant rôle in the mouse disease though essential in that of swine, and fatal pneumonias can be produced in mice by pure virus infections. Mice surviving the virus disease are immune to reinfection for at least a month. In mice the disease is not contagious though it is notably so in swine. The virus, while regularly producing fatal pneumonias when administered intranasally to mice, appears to be completely innocuous when given subcutaneously or intraperitoneally. Prolonged serial passage of the virus in mice does not influence its infectivity or virulence for swine or ferrets. It is a stable virus so far as its infectivity is concerned, and can be transferred at will from any one of its three known susceptible hosts to any other. In discussing these facts the stability of the swine influenza virus has been contrasted with the apparent instability of freshly isolated strains of the human influenza virus. Though the mouse is an un-natural host for the virus it is, nevertheless, useful for the study of those aspects of swine influenza which have to do with the virus only.     

Formation of non-infectious influenza virus in mouse lungs: its dependence upon extensive pulmonary consolidation initiated by the viral inoculum

Formation of non-infectious virus—particles which hemagglutinate red blood cells and react with antibody to fix complement but do not infect the chick embryo or mouse—occurred when large quantities of certain strains of influenza viruses were inoculated intranasally into mice. Dependent upon the agent employed, 10^6.5 to 10^8.5 E.I.D. was essential to elicit this phenomenon. To accomplish this unusual multiplication it was essential to use a strain of virus which effected extensive pulmonary consolidation; strains of virus which did not produce marked lung lesions, even when as much as 10^8.5 E.I.D. was inoculated, did not form non-infectious virus. The development of this viral form was directly dependent upon the extent of cell damage obtained: consolidation of more than 50 per cent of the lung volume was required. The majority of non-infectious particles developed during the initial cycle of viral multiplication, and concurrently with the formation of non-infectious virus there was a corresponding decrease in the number of infectious viral particles. Non-infectious virus could not be propagated on serial passage in mouse lungs: on second lung passage only fully infectious virus was detectable. The formation of the non-infectious viral form was not the result of interference with synthesis of infectious virus by inactivated virus in the inoculum; for inoculation of heated infected allantoic fluid which contained more than 99 per cent of non-infectious virus did not result in the development of new non-infectious virus. Although inoculation of a large quantity of virus resulted in infection which yielded a relatively low titer of infectious and high titer of non-infectious virus, inoculation of a small quantity of the agent resulted in a high yield of infectious virus and no non-infectious that was detectable. In both instances the total quantity of antigenic viral material synthesized in the mouse lungs was the same. These data do not support the hypothesis that the non-infectious virus formed consisted of immature or incomplete viral particles, but suggest instead that non-infectious virus is inactivated virus or some aberrant form of the agent.     

THE QUANTITATIVE DETERMINATION OF INFLUENZA VIRUS AND ANTIBODIES BY MEANS OF RED CELL AGGLUTINATION

1. The agglutination titer for chicken red cells of freshly prepared or carefully stored suspensions of PR8 influenza virus, that is to say virus of maximum pathogenicity, was found to be proportional to the mouse lethal titer of the same preparations. 2. The agglutination titer of infected allantoic fluid procured in a standard way is relatively constant, regardless of the influenza strain used and its pathogenicity for mice. 3. Virus preparations inactivated by heat or storage may retain their agglutinating power. 4. Certain animal sera contain a partially heat-labile factor which, in low dilution, inhibits the agglutination of chicken red cells by influenza A and influenza B viruses. 5. The agglutination inhibition test, using ferret and human sera, gives qualitative data regarding influenza antibodies which are similar to the information obtained on the same sera by means of the virus neutralization test. 6. There is a definite relationship between the agglutination inhibition titer and the virus neutralization titer of a serum. On a logarithmic scale of both variables, this relationship is essentially linear within the range investigated. 7. The agglutination inhibition titer of immune ferret serum is inversely proportional to the amount of virus used in the test.     

Evaluation of the anti-influenza virus activities of 1,3,4-thiadiazol-2-ylcyanamide (LY217896) and its sodium salt.

1,3,4-Thiadiazol-2-ylcyanamide (LY217896) and its sodium salt were shown to be effective against influenza A and B viruses in vitro and in the mouse model. In nondividing confluent MDCK cells, the 50% inhibitory concentration of LY217896 ranged from 0.37 to 1.19 micrograms/ml against various strains of influenza A virus and from 0.75 to 1.54 micrograms/ml against various strains of influenza B virus, with no apparent cytotoxicity. However, at a concentration of 0.31 microgram/ml, LY217896 inhibited the replication of dividing MDCK cells. LY217896 (9 mg/m2 of body surface area per day) administered in the diet, in the drinking water, by oral gavage, by intraperitoneal injection, or by aerosolization was well tolerated and protected CD-1 mice infected with a lethal dose of influenza A or B virus. Effective administration of the compound could be delayed for up to 96 h postinfection. Virus titer was reduced by 1 to 2 log10 units in lungs of mice given LY217896 in the drinking water. Mice treated initially with protective levels of LY217896 were resistant to a subsequent challenge of influenza virus in the absence of the compound, indicating that the animals were able to develop immunity to the initial infection. Administration of LY217896 to uninfected mice did not induce interferon-like activity or interfere with natural killer cell function. In the ferret, LY217896 was effective in preventing fever induced by influenza virus.     

GENETIC STUDIES OF INFLUENZA VIRUSES : I. VIRAL MORPHOLOGY AND GROWTH CAPACITY AS EXCHANGEABLE GENETIC TRAITS. RAPID IN OVO ADAPTATION OF EARLY PASSAGE ASIAN STRAIN ISOLATES BY COMBINATION WITH PR8

The passage of newly isolated, filamentous Asian (A2) influenza viruses in the presence of non-infective PR8 (A) virus results in the rapid emergence of virus of Asian (A2) antigenicity but PRS-like growth capacity and spherical morphology. Evidence is presented that this effect results from genetic interaction of the infective Asian and non-infective PR8 viruses rather than from spontaneous change of the Asian strain. It is concluded that influenza viral morphology, growth rate and growth capacity are associated genetic traits which distinguish unadapted from adapted strains, and which are transferable by recombination. A pragmatic consequence of these experiments is the fact that conditions have been defined for the rapid adaptation of early passage influenza virus isolates to the chick embryo allantoic sac. Such adaptation is attended by an increase in viral yield which has obvious implications for vaccine production during future epidemics with new antigenic types.     

THE ANTIBODY RESPONSE OF HUMAN SUBJECTS VACCINATED WITH THE VIRUS OF HUMAN INFLUENZA

Human influenza virus cultivated in tissue culture medium may be administered subcutaneously or intradermally to human individuals without causing evidence of infection. Subjects so treated develop a good titer of circulating antibodies effective against mouse passage virus and, if antibodies were previously present, vaccination stimulates the production of more antibody. The antibodies so induced persist for at least 5 months, although in this period of time some decline in titer may have begun. The antibody response to vaccination parallels both in extent and persistence that occurring as a result of the naturally acquired disease. The available data do not enable one to evaluate the effect of vaccination in preventing human infection with influenza. It seems not unlikely that the increase in circulating antibody will be accompanied by an increased ability to combat the natural infection.     

EXPERIMENTAL INFECTION WITH INFLUENZA A VIRUS IN MICE : THE INCREASE IN INTRAPULMONARY VIRUS AFTER INOCULATION AND THE INFLUENCE OF VARIOUS FACTORS THEREON

Following intranasal inoculation of influenza A virus (strain PR8) there is a rapid increase of the virus in the lungs which with large doses reaches a maximum within 24 hours. With smaller doses, although the proportional increase is greater, the maximum concentration is not reached until 48 hours following inoculation. If a lethal dose is administered, the ultimate concentration of the virus in the lungs is the same, irrespective of the size of the dose. If a sublethal dose is given, the titer of the virus in the lungs does not achieve the titer reached in mice receiving a lethal dose. Within 48 hours following inoculation of a sublethal dose the lungs of a mouse may contain at least 76,000 M.L.D., yet the mouse survives. The intranasal instillation of sterile fluid (distilled water, varying concentrations of NaCl, broth, or 10 per cent normal serum) into a mouse sublethally infected produces a sharp rise in the virus content of the lung usually followed by death within 3 to 8 days. If, however, the instillate consists of 10 per cent immune serum, there is no rise in the virus titer, and no apparent harm results from the instillation. The implications of these phenomena are discussed and an hypothesis presented to explain their occurrence.     

THE PROPAGATION OF THE VIRUS OF EPIZOOTIC HEMORRHAGIC DISEASE OF DEER IN NEWBORN MICE AND HELA CELLS

The New Jersey strain of EHD virus has been propagated in newborn Swiss mice by the intracerebral route and is regularly lethal beyond the first serial mouse passage. A complement-fixing antigen prepared from the brains of infected mice reacts positively with the sera of deer recovered from infection with either the New Jersey or South Dakota strain of virus, but not with the serum of normal deer. The mouse-passaged virus induced an inapparent infection in an experimental deer. The virus can also be grown serially in HeLa cell culture and induces a characteristic cytopathic effect. It is neutralizable in such cultures to high titer by the sera of deer recovered from EHD (New Jersey strain) and to lower titer by the serum of a deer recovered from EHD (South Dakota strain). Normal deer serum does not neutralize the virus in tissue culture. The HeLa cell-passaged virus induced typical lethal EHD in an experimental deer and virus could be recovered from most of the tissues of this animal in HeLa cell culture. An unexplained prozone of inhibition of cytopathogenicity at low dilutions was observed in cultures of some of the organs. The fact that EHD virus exhibited a limited sensitivity to sodium desoxycholate suggests that it may belong in the arbor virus group.     

STUDIES OF ANTIGENIC DIFFERENCES AMONG STRAINS OF INFLUENZA A BY MEANS OF RED CELL AGGLUTINATION

A study of cross inhibition tests among strains of influenza A virus and their antisera showed that the results obtained were subject to a certain amount of variation due to the red cells, the virus suspensions, and the ferret antisera employed. Methods have been demonstrated for handling the data obtained from such tests, so that these variables were corrected or avoided, making it possible to use the agglutination technique for antigenic comparisons. The antigenic pattern of eighteen strains of influenza A virus, obtained from the 1940–41 epidemic in the United States, has been compared by means of agglutination inhibition tests with ferret antisera. No significant antigenic differences were found among sixteen of these strains (all isolated from throat washings by the inoculation of chick embryos) although they were obtained from individuals in widely separated regions of the country. Two strains, from cases occurring early in the epidemic and isolated from throat washings by ferret and mouse passage, showed a slight but significant strain difference from the other strains and from each other. One of the 1940–41 strains on cross test resembled the PR8 strain more closely than any other stock strain tested.     

INDICATIONS OF HEREDITARY, SPATIAL REARRANGEMENT OF ANTIGEN COMPLEXES, IN THE INFLUENZA VIRUS

During passage in mice which had been vaccinated with the homologous, and with closely related strains of influenza virus, the passage strain developed a lessened susceptibility to the deleterious effects of the "immune" environment, concommittant with which was a developed capacity to evoke antibodies which reacted with earlier strains of virus—a capacity which was inapparent in the parent strain. However, the parent strain exhibited a relatively broad range of surface reactivity which was not apparent in the derived strain. The data are interpreted to mean that the hereditary change resulted from spatial rearrangement and quantitative redistribution of antigens in the virus particle (in which the surface is viewed as being distinct from the inner bulk), and are viewed as enhancing the idea that influenza virus variation (i.e., "mutation") may result from a rearrangement of existing hereditary elements.     

STUDIES ON THE TOXICITY OF INFLUENZA VIRUSES : I. THE EFFECT OF INTRACEREBRAL INJECTION OF INFLUENZA VIRUSES

Intracerebral injection of preparations of influenza viruses into mice led to tonic and clonic convulsions and death in tetanus, usually within 24 to 72 hours. Histological examination revealed the destruction of the ependymal lining of the ventricles as the dominant finding. These reactions were obtained in four different strains of mice as well as in rats, guinea pigs, and hamsters. They were observed in mice after injection of mouse and egg-adapted virus, and all strains of virus tested gave these responses as long as sufficient quantities of the agents were injected. Control materials, such as normal allantoic fluids, particulate components of normal chorio-allantoic membranes, and suspensions of normal murine lungs, gave uniformly negative results. Activation of a latent virus as well as the inadvertent admixture of a neurotropic agent to the influenza cultures was excluded. Propagation of the influenza viruses in the central nervous system could not be demonstrated and, in fact, the agents were no longer demonstrable in 4 days. It was concluded that the observed neurological reactions were the result of toxic activities rather than of virus propagation. Comparison between the infectivity, hemagglutinating capacity, and the toxic activity showed that preparations with the higher titers of virus and hemagglutinin were also the more toxic ones. The toxic agents could not be separated from the infectious particles by the use of such technics as differential high speed centrifugation, adsorption onto and elution from chicken red cells, and precipitation by protamine. The toxic effect of influenza A virus preparations was specifically neutralized by anti-influenza A and not by anti-influenza B serum, and conversely. In addition, antigenic differences were noted between two strains of influenza A virus by this method of testing.     

SEROLOGICAL STUDIES OF SWINE INFLUENZA VIRUSES

1. Cross-neutralization tests with sera from swine recovered from infection with swine influenza indicated the serological identity of 7 strains of swine influenza virus obtained from different sources. 2. Cross-neutralization tests with sera from rabbits, immunized to swine influenza virus, exposed serological differences among the same 7 swine influenza virus strains. Two strains appeared to be serologically similar and were characterized by the ability to produce effective homologous virus-neutralizing sera which were, however, poor or ineffective against the heterologous virus strains. Two other strains were also serologically similar but produced antibodies effective not only against themselves, but against all heterologous strains as well. The remaining 3 strains were intermediate in their ability to produce heterologous virus-neutralizing antibodies. 3. The human influenza viruses included, especially strains WS and Oakham, were most effectively differentiated serologically from the swine influenza viruses by rabbit antisera. 4. The suggestion is advanced that swine antisera express the antigenic composition of the swine influenza viruses, while rabbit antisera reflect either their antigenic arrangement or the arrangement of the components responsible for their mouse pathogenicity. On this interpretation the 7 strains of swine influenza virus studied would be considered to have similar antigenic compositions but differing antigenic structures. 5. The serological differences among strains of the swine influenza virus, detectible by rabbit antisera, are probably of no practical significance so far as the natural disease, swine influenza, is concerned.     

Morphological and quantitative comparison between infectious and non-infectious forms of influenza virus

Electron microscopic study has revealed the morphological entity responsible for the rise in viral hemagglutinin observed in brains of mice after intracerebral inoculation of non-neurotropic strains of influenza virus. This rise in hemagglutinin, although dependent on inoculation of fully infectious virus, is not associated with an increase in infectious titer. The hemagglutinating principle is functionally similar to the "incomplete" influenza virus which can be obtained from chick embryos by serial egg-to-egg transfer of undiluted, infected allantoic fluid according to the method of von Magnus. A method has been described which facilitates selective adsorption of viral particles recovered from organ extracts on saponine-lysed ghosts of fowl erythrocytes. This procedure has been utilized in studying the morphology of non-infectious, hemagglutinating virus from chorio-allantoic membranes or mouse brains and in comparing these two forms with each other and with ordinary, infectious (standard) influenza virus. Standard virus isolated from allantoic fluids or membranes of infected eggs was found to contain uniform particles of predominantly spherical shape with smooth surface and even density, resembling those described by others. The appearance of such particles was not affected by the procedure of extraction and concentration used. In contrast, non-infectious, hemagglutinating virus obtained either from allantoic sacs ("undiluted passages") or from mouse brain was pleomorphic and seemed to consist of disintegrating particles. The majority appeared flattened and bag-like and had a rough, granular surface and reduced, uneven density. 37 per cent of the non-infectious particles isolated from mouse brain infected with the non-neurotropic strain WS had diameters in excess of 170 mµ, as compared with only 2 per cent of the particles of the parent strain itself. Regardless of whether or not the contrast in appearance of standard and of non-infectious particles was due to differing resistance to the preparatory treatment, it indicated the existence of basic structural differences between the two types of virus. Correlation of particle counts with hemagglutinin titers has shown that the non-infectious virus obtained from mouse brain is, unit for unit, an equivalent counterpart of standard virus derived from infected eggs. The end-point of hemagglutination in a pattern test corresponds for both forms to that dilution at which the ratio virus particles/red cells approaches one. The quantitative data based on particle counts support the assumption that non-infectious virus arises in mouse brain as a product of viral multiplication.     

ADSORPTION OF INFLUENZA VIRUS ON CELLS OF THE RESPIRATORY TRACT

A study of the reaction between influenza virus and the cells of the excised and perfused ferret lung has yielded the following results: (1) The cells of the lung rapidly adsorbed large amounts of intratracheally inoculated virus. (2) After a short interval the pulmonary cells began spontaneously to release the adsorbed virus, and in the case of influenza B the release was 75 per cent complete after 5 hours. (3) The Lee strain was more completely released from pulmonary cells after 5 hours than was the PR8 strain. (4) After the cells released the adsorbed virus they appeared incapable of adsorbing virus as before. (5) When the mouse-infecting capacity of the virus had been done away with by heat or formalin, the virus was adsorbed by the pulmonary cells but was not released. In all except the last of the characteristics listed the interaction between influenza virus and the pulmonary cells closely resembles that between influenza virus and avian red blood cells. In the living ferret inhaled influenza virus was also rapidly adsorbed by the lung, but in a very short time the adsorbed virus which at first could be readily eluted (after perfusion and excision of the lung) became so much more firmly fixed as not to be released by this method. Free virus could not be demonstrated in the living ferret until 24 hours after the animal had been exposed to the inoculum. On the basis of these and previous experiments it is postulated that the destruction of a specific receptor substance,—which may involve an enzymatic reaction,—may be a necessary preliminary event in the parasitism of susceptible cells by influenza virus.     

STUDIES ON THE IN VIVO AND IN VITRO MULTIPLICATION OF THE LDH VIRUS OF MICE

In vivo analysis of the virus titer in various loci, 24 hr after infection, showed that a titer similar to that in the blood plasma was found in the ascitic fluid of Erlich ascites cancer-bearing mice, and in lymph nodes, spleen, and thymus, i.e. loci which contain macrophages as a common cell type. However, only in the lymph nodes and in the ascitic fluid did the increase in virus titer precede or parallel the increase in the plasma. The LDH virus titer in the plasma of X-irradiated mice was similar to that of control mice, eliminating radiation-sensitive cells but not macrophages as target cells of the virus. Electron microscopic observation of infected lymph node cells revealed the presence of two types of particles: one consisting of small densely stained annuli, about 25 mµ in diameter and one of similar dense annuli with a halo extending the diameter to about 50 mµ. Such particles were repeatedly observed within single or double membraned vesicles. In vitro, the LDH virus multiplied only in cultures of mouse peritoneal macrophages, maintained in medium 199 with 10% FBS. The virus titer could be maintained for at least 33 days, during eleven serial passages, involving an overall dilution factor of 10¹¹. These results corroborate the findings of Evans and Salaman, who used peritoneal macrophages maintained in Eagle's medium and 5 to 10% lamb serum. However, in the serial passage experiments reported here, the virus titer could only be maintained following trypsinization of each successive inoculum. The role of macrophages as the target cell for LDH virus multiplication in vivo is discussed.     

STUDIES ON PNEUMONIA VIRUS OF MICE (PVM) IN CELL CULTURE : I. REPLICATION IN BABY HAMSTER KIDNEY CELLS AND PROPERTIES OF THE VIRUS

Pneumonia virus of mice (PVM) has been serially propagated in a line of baby hamster kidney (BHK21) cells. A maximum titer of 6.3 x 10⁶ TCID₅₀ per ml was obtained, and there was little variation in yield on serial passage. PVM grown in BHK21 cells was antigenically similar to virus obtained from the mouse lung, but was somewhat less virulent for the mouse after 10 serial passages in these cells. Virus produced by BHK21 cells agglutinated mouse erythrocytes without prior heating or other treatment. Sedimentation of PVM in the ultracentrifuge or precipitation by ammonium sulfate resulted in a loss in infectivity but an increase in hemagglutinating activity, presumably due to disruption of the virus particle. In a potassium tartrate density gradient, the major portion of infective virus sedimented at a density of approximately 1.15, and noninfective hemagglutinin, at a density of approximately 1.13. Stock virus preparations appear to contain a large amount of noninfective hemagglutinin. The replication of PVM was not inhibited by 5-fluoro-2'-deoxyuridine, 5-bromo-2'-deoxyuridine, or 5-iodo-2'-deoxyuridine. Infected cells contained eosinophilic cytoplasmic inclusions which showed the acridine orange staining characteristic of single-stranded RNA. Foci of viral antigen were observed in the cytoplasm of infected cells by fluorescent antibody staining. The results suggest that PVM is an RNA virus that replicates in the cytoplasm.     

THE IMMUNOLOGICAL RESPONSE TO INFLUENZA VIRUS INFECTION AS MEASURED BY THE COMPLEMENT FIXATION TEST : RELATION OF THE COMPLEMENT-FIXING ANTIGEN TO THE VIRUS PARTICLE

A quantitative complement fixation test with influenza immune sera and virus antigens obtained from allantoic fluid is described. The method utilizes a photoelectric densitometer which provides a simple, objective, and accurate determination of the hemolytic reaction. The enhancement of the hemolytic activity of complement in the presence of serum or allantoic fluid necessitates a preliminary titration of complement in the presence of these agents. An accurate appraisal of the activity of the complement under the conditions of the actual test permits the selection of an optimal amount of complement and greatly increases the sensitivity of the test. The substance (or substances) responsible for the enhanced hemolytic activity of complement has been found in human and many animal sera and in allantoic fluids obtained from the developing chick embryo. It requires the presence of both complement and hemolysin, resists heating at 100°C. for 2 hours, and is dialyzable. Allantoic fluid or mouse lung preparations of influenza virus contain a complement-fixing antigen which is intimately associated with the virus particle. It sediments in the high speed centrifuge at the same rate as the hemagglutinin and infective particle and, like the latter, is adsorbed by fowl red blood cells and eluted from the cells on standing at room temperature or 37°C. It cannot be separated from the virus particle by repeated washings in the centrifuge or repeated adsorptions with red blood cells; the hemagglutinin and complement-fixing antigen titers remain roughly proportional. This antigen shows a high degree of strain specificity in cross complement fixation tests with PR8, W.S., and swine ferret antisera, and, as found with the neutralization test, it shows little or no strain specificity with human sera. A soluble antigen is also present in influenza virus preparations which can be readily separated from the virus particle by centrifugation. It is not adsorbed by red blood cells. Furthermore, it reacts in lower titer with ferret antisera and usually shows less strain specificity in cross complement fixation tests. In general, allantoic fluid virus preparations contain much less of the soluble antigen than mouse lung extracts.     

THE INFLUENCE OF INFLUENZA VIRUS INFECTION ON EXOGENOUS STAPHYLOCOCCAL AND ENDOGENOUS MURINE BACTERIAL INFECTION OF THE BRONCHOPULMONARY TISSUES OF MICE

Mice infected with a non-mouse-adapted Asian strain of influenza A virus suffered an impaired capacity to destroy or remove staphylococci introduced by the respiratory route. This temporary inhibition of local defense mechanisms was of 7 to 10 days' duration. The persistence of staphylococci in the lung following influenza did not appear to alter the nature of the pathologic reaction to influenza virus. The presence of influenza virus infection in the respiratory tract of the mouse did not alter the fate of intravenous staphylococci in the lung or other organs. In 40 to 50 per cent of mice with influenza, purulent bronchopneumonia and infection with Pasteurella and Hemophilus of murine origin were noted. A minority of control animals evidenced such infection. The administration of antimicrobials to which the murine bacteria were susceptible prevented both the appearance of the endogenous infection with Pasteurella or Hemophilus and the purulent sequelae to influenza virus infection. The true picture of uncomplicated bronchopulmonary influenza virus infection was thus separated from the combined virus-bacteria effect otherwise encountered.