Streptococci From Outdoor Air in Relation to the Seasonal Occurrence of Infections Involving the Respiratory Tract and Nervous System Respectively

The importance of spread by air of streptococci and other inciting agents of disease has been emphasized and greatly extended in recent bacteriological studies.^1,3,4,14 Interest of the writer in the importance of airborne streptococcal infections dates back forty-five years, when streptococci identical to those causing postoperative and epidemic respiratory infections including tonsillitis were isolated from the air of operating rooms.⁹ The probability that the streptococci associated with epidemic influenza, poliomyelitis and encephalitis may in part be airborne was indicated (1) by the isolation of a pandemic strain of streptococcus during the pandemic of influenza in 1918,¹⁰ (2) by the demonstration by a precipitin reaction of the poliomyelitic type of streptococcus in the throats of persons ill and well in high incidence during epidemics of poliomyelitis,⁷ (3) by the cutaneous reaction of persons ill and well to intradermal injections of “encephalitic” streptococcal antiserum, and (4) by the isolation of the “encephalitic” type of streptococcus from outdoor air, mosquitoes, from persons and from the brain of seven species of animals that were ill with or that had died of encephalitis during the epidemic of encephalitis in Minnesota and the Dakotas in 1941.⁶ Control samplings of air made in parallel far remote from the epidemic failed to yield the encephalitic type of streptococcus. Airplane samplings at an altitude of about 1,000 feet during this epidemic yielded the encephalitic type of streptococcus, while corresponding samplings made six weeks later when respiratory infections were generally prevalent yielded the respiratory or influenzal type of streptococcus.

The isolation of specific types of streptococci and the demonstration of specific streptococcal precipitinogen in indoor and outdoor air, in relation to epidemic diseases involving the central nervous system in summer and the respiratory tract in winter months, also suggest the probable importance of aerial spread of the streptococci.⁸ It is the purpose of this report to describe the special methods used and record the results obtained in a further bacteriologic study of outdoor air to determine, perhaps more precisely than heretofore, whether the streptococci so generally prevalent during respective epidemics may be airborne or not and whether, in fact, the streptococci isolated from outdoor air and from persons while well remote from epidemics or ill during epidemics are alike in infecting and serologic properties.

Methods of Study

The methods employed were essentially like those previously described.⁸ Chemically inert materials presenting a large surface, such as pebbles and glass balls of varying size, drinking straws and wood applicators, coated with liquid petrolatum, and sticks of corn pith, glass and sheep’s wool and “fiberglas” within test tubes, 1.5 by 10 cm., taped together in bundles of seven, or pairs of light-weight metal tubes, 3.5 by 20 cm., with open ends screened with a double layer of surgical gauze, were exposed to currents of air after sterilization in the autoclave. Mobile samplings were made on the front end of the hood of automobiles, atop Diesel locomotives and outside of nonworked vestibule doors of Pullman cars.a Stationary samplings were made atop 12-foot poles in the ground or atop tall buildings. Thirty-four mobile samplings were made on automobiles or on trains during transit, totalling 18,200 miles and comprising eighteen states. Twenty-eight stationary samplings were made in five states during exposures altogether for 245 days, ranging from one to twenty-three days for the different samplings. The samplings were made in relation to epidemics of respiratory infection and poliomyelitis, respectively. A very wide range of types of weather occurred from dry, hot sunshine in summer to fog, slight and pouring rains and falling snow during spring, autumn and winter. In some instances, sterile funnels in test tubes were exposed to first rains and newly fallen snow was collected in sterilized jars, and cultures were made from the water and the melted snow thus obtained. The exposed materials, before and after washing in NaCl solution, and part of the NaCl solution washings were inoculated into tall tubes of freshly prepared dextrose-brain broth onto blood-agar plates and into mice.¹ Gram-safranin stained films of the sediment of centrifuged washing were examined for bacteria. Precautions were taken to prevent extraneous contamination in making and collecting the samplings and in making the cultures. In this as in previous studies, the mediums dextrose-brain broth or soft dextrose-brain agar which afford a gradient of oxygen tension and other conditions highly favorable for growth of specific types of streptococci were used for the primary isolation. Pure cultures in dextrose-brain broth of the primary, secondary and usually not higher than the third subculture of rapidly repeated transfer cultures were centrifuged, and antigenic specificity of the sedimented streptococci was preserved by partial dehydration in very dense suspensions of glycerin, 2 parts, and saturated NaCl solution, 1 part, at 10°C. Antiserums, thermal antibody and suspensions for agglutination tests were prepared from appropriate dilutions of the streptococci from these dense suspensions and not from strains long cultivated on artificial mediums meanwhile. To determine the infectivity of the streptococci in NaCl solution washings of exposed material and of unconcentrated or 10:1 suspension in NaO solution of dextrose-brain broth cultures, four mice were inoculated routinely-two cerebrally and two cerebrally and nasally under deep ether anesthesia with .03 ml. and .06 ml. respectively of freshly isolated strains consisting of young primary pure cultures in tall columns of dextrose-brain broth or as obtained in serial dilution cultures in this medium and not from cultures on blood agar. Blood-agar platings were made of apparently pure cultures in the dextrose brain broth to check the purity and to determine the type of streptococci isolated.

Results of Examination of Stained Films and of Cultures

Microscopic examination of stained films of the sediment of NaCl solution washings of exposed material uniformly revealed moderate or small numbers of Gram-positive bacilli resembling B. subtilis and sometimes a few Gram-negative bacilli resembling B. aerogenes. Diplococci singly or in short chains were found always only in small numbers in the sixty-two samplings, and micrococci in but a few samplings. There was a fair correlation between the results of examination of stained films and the numbers of respective organisms that grew on blood-agar plates. The streptococcal colonies on blood-agar plates were often extremely small, gray in color, nonadherent, nonhemolyzing and sometimes green-producing. Beta streptococci were isolated in but two instances. Growth of streptococci uniformly occurred in serial dilution cultures in dextrose-brain broth of NaCl solution washings at dilutions indicating the presence of streptococci or of submicroscopic phases or units 10 to 1,000 fold greater than indicated by microscopic examination of stained films and by the number of colonies that grew on blood-agar plates.

Results in Mice

The mortality and the incidence of pneumonitis and polio-encephalitis in mice following inoculation of pure cultures of the streptococci, as isolated according to season and years, are shown graphically in Figure 1. Great variations in virulence or localizing properties, as indicated by the incidence of pneumonitis and polio-encephalitis, occurred according to season following inoculation of altogether 684 mice. The percentage incidence of pneumonitis was very high during the winters of 1947-1948, moderately high during spring, 1948, and almost nil in August, 1948. The incidence of pneumonitis then rose slightly during September and moderately during October, 1948, and was maximal in winter during December-January, 1948-1949. It dropped somewhat during March-April, and as warm weather returned, it dropped to a low level during May-June, 1949, and was again lowest during August-September, 1949, and rose moderately during September-October, 1949, and abruptly in December, 1949, as respiratory infections of influenzal type occurred.

Fig. 1. Incidence of pneumonitis and polio-encephalitis in mice following cerebral and/or nasal inoculation of nonhemolytic streptococci, as isolated from outdoor air, in relation to the seasonal occurrence of epidemics of influenza and other respiratory infections and poliomyelitis.

In sharp contrast, the incidence of polio-encephalitis in mice similarly inoculated was minimal during the winter and spring of 1947-1948 and 1948-1949 and was maximal in summer, 1948 and 1949, and again minimal in December, 1949. It was minimal in winter and remained low during March-April and May-June, increased sharply during August-September and dropped sharply during October-November and December, 1949. Spasms occurred in low incidence regardless of season and of the incidence of experimental pneumonitis or polio-encephalitis. The mortality of inoculated mice, except in August of 1948 and 1949 when the incidence of poliomyelitis was maximal, paralleled roughly the incidence of pneumonitis.

The lesions of lungs following inoculation of the streptococci were similar to the early lesions in mice that follow inoculation of influenza “virus”.^2,12

In previous studies, it was found that the nonhemolytic streptococci and related filtrable agents in lungs and brains of mice kept in 50 per cent glycerin at 10°C. live and retain characteristic virulence for a long time. The results of inoculation of a composite emulsion and corresponding filtrate of pneumonic lungs, that had been thus preserved for about one year, in the initial mouse passage and of five successive passages of dextrose-brain broth cultures of the streptococcus, of emulsions and filtrates of emulsions of pneumonic lungs, are shown graphically in Figure 2.

Fig. 2. Results of cerebral and/or nasal inoculation in successive mouse passages of a streptococcus, originally isolated from outdoor air during an epidemic of influenza, and emulsions or filtrates of emulsions of pneumonic lungs of mice stored in 50 per cent glycerin at 10° C. for one year.

The streptococcus with which the pneumonic lungs were originally produced was isolated in January, 1948, from outdoor air atop a four-story building far removed from human habitation during a severe epidemic of influenza in Southern California. It will be seen that the incidence of pneumonitis following inoculation of the composite emulsion and filtrate of emulsion from the preserved pneumonic lungs m the initial passage was very high while the incidence of polio-encephalitis was low, which was also true of successive mouse passages of the streptococcus of emulsions and filtrates of emulsions of pneumonic lungs. The mortality was relatively low as is usual in such experiments. The streptococcus was isolated from pneumonic lungs (not shown in Figure 2) in twenty-two of thirty cultured. Isolations of the streptococcus from the pneumonic lung were highest following inoculation of cultures of the streptococcus, next highest in mice receiving emulsions, and least often in mice inoculated with filtrates of emulsions on successive passage. Cultures from filtrates on blood-agar were always negative, and streptococci often grew in dextrose-brain broth.

In order to determine whether the great difference in the incidence of pneumonitis and polio-encephalitis in mice inoculated with streptococci isolated from outdoor air in winter and summer respectively, and whether perhaps a similar reason for such difference in the incidence in temperate climates of spontaneously occurring respiratory infections in winter and of poliomyelitis or encephalitis in summer in human beings exists, or whether due to a possible increased seasonal susceptibility of respiratory tract to streptococci in winter and of the nervous system in summer, mice were inoculated in winter with “neurotropic” streptococci isolated in summer and that had produced polio-encephalitis in high incidence in mice on isolation and in summer with “pneumotropic” streptococci that were isolated in winter and that had produced pneumonitis in high incidence in mice on isolation. The characteristic pneumotropic and neurotropic virulence of both groups of strains was preserved meanwhile at 10°C., either in dehydrated form in dense suspension in glycerin, 2 parts, and saturated NaCl solution, 1 part, or in pneumonic lungs or “polio-encephalitic” brains in 50 per cent glycerin. The results of these experiments were as follows:

Of seventy-five mice inoculated in summer with twenty-four strains originally isolated in winter and whose “pneumotropic” virulence was preserved, pneumonitis developed in forty-six (61 per cent); polio-encephalitis in only six (8 per cent), and gastroenteritis in but one mouse. The streptococcus was isolated from the pneumonic lungs in thirty-five (85 per cent) of forty-one cultured. In sharp contrast, of fifty-six mice inoculated in winter with twenty strains originally isolated in summer and whose neurotropic virulence was preserved, polio-encephalitis developed in thirty (54 per cent); pneumonitis in nine (16 per cent), and gastroenteritis in one. The streptococcus was isolated fromi the brain in twenty-seven (60 per cent) of forty-five cultured.

The gross and microscopic lesions of the lungs resembled those following inoculation of streptococci isolated from persons ill with respiratory infection during epidemics.¹⁰ Abscess formation at the site of injection in the brain and suppurative meningitis were not observed. Hemorrhage, edema, degeneration of ganglion cells, mild perivascular and diffuse infiltration by polymorphonuclear leukocytes and round cells of the meninges especially of sulci of the brain and mild round cell infiltration characterized the microscopic lesions of the brain, the medulla and spinal cord in mice in which symptoms of polio-encephalitis developed following inoculation of the streptococcus.

Methods and Results of Agglutination Tests

The antiserums and thermal antibodies used in agglutination experiments were prepared not from streptococci grown indefinitely on artificial mediums, but from appropriate dilutions of the dense glycerin-NaCl solution suspensions of streptococci freshly isolated in dextrose-brain broth from persons suffering from respiratory infection and poliomyelitis respectively. The antiserums, thermal antibodies and convalescent serums were diluted with NaCl solution containing .2 per cent phenol at four fivefold dilutions of 1:20 to 1:2,500. Two-tenths ml. of these dilutions and .2 ml. of the respective dense suspensions of streptococci diluted to 6,000,000,000 per ml. with NaCl solution containing .2 per cent phenol were added to and mixed in test tubes, 1 by 7 cm. The mixtures were incubated at 48°C. overnight when readings were made. The degree of agglutination was recorded according to the arbitrary scale of 0 to 4 plus. In order to facilitate comparisons of agglutinative values of the very large number of agglutination tests made with widely different materials containing antibodies, the percentage of the total possible agglutination for the different suspensions was determined. A four plus agglutination for each of the four dilutions or 16/16= 100 per cent. A total of nine plus agglutination or 9/16 = 56 per cent, etcetera.

The percentages of total possible agglutination by thermal antibody and antiserums prepared with streptococci isolated in studies of respiratory infections and poliomyelitis and by convalescent poliomyelitis serum in a large number of agglutination tests, representing altogether sixty-two samplings or strains and 135 suspensions of streptococci isolated from outdoor air, are shown graphically according to season in Figure 3.

Fig. 3. Percentage incidence of total possible agglutination at four five-fold dilutions (1:20 to 1:2500) of thermal antibody and antiserum prepared respectively from streptococci isolated in studies of respiratory infection and poliomyelitis and of “poliomyelitis” convalescent serum for streptococci as isolated seasonally from outdoor air.

It will be seen that the percentage incidence of agglutination by thermal antibody prepared in vitro and antiserum prepared in horses with streptococci isolated from persons having influenza or other respiratory infections for streptococci isolated from outdoor air in winter was uniformly significantly and usually far higher than for streptococci isolated in summer. The percentages in winter, 1947, were 95 and 49; in spring, 1948, 52 and 29; in summer, 1948, 30 and 19; in winter, 1948-1949, 58 and 40; in late spring, 1949, 28 and 20; in summer-autumn, 1949, 41 and 26; and in autumn-winter, 1949-1950, 65 and 80, respectively. The percentage incidence of agglutination of streptococci isolated from outdoor air in summer by thermal antibody and antiserums prepared with streptococci isolated from persons having poliomyelitis were in reverse, being low for streptococci isolated from outdoor air in winter and spring and high for strains isolated in summer. Thus the percentages in winter, 1947, were 2 and 7; in spring, 1948, 12 and 2; in summer, 1948, 66 and 55; in winter, 1948-1949, 14 and 8; in spring and early summer, 1949, 40 and 31; in summer, 1949, 72 and 38 and in late autumn, 1949, 35 and 25, respectively. The agglutinative titer of poliomyelitis convalescent serum was likewise significantly higher for streptococci isolated from outdoor air in summer than for strains isolated in winter.

In addition to the striking seasonal correlation of agglutinative titers, the respective specific agglutinations by thermal antibody and anti-streptococcic serum paralleled closely the degree of prevalence of spontaneously occurring respiratory infections and of poliomyelitis.

Agglutinations (not shown in Figure 3) by control thermal antibody and antiserums prepared with streptococci isolated in studies of arthritis were uniformly negligible or far less than by the respiratory and poliomyelitis thermal antibody and antiserum and there was no seasonal variation.

Comments and Summary

The isolation of streptococci from outdoor air and a study of their infective and serologic properties at different seasons over a period of three years are reported. The incidence of isolations were somewhat greater during spring, summer and autumn than during winter especially when the ground was covered with snow, and were uniformly high from water collected in first rains and snows. The streptococci isolated were of the nonhemolytic or green-producing type. Beta hemolytic types and Types I, II and III pneumococci were isolated in only a few instances. The morphology, staining reactions and growth requirements of the streptococci were similar without regard to season but their infecting and serologic properties were very different according to season.

Their virulence, with few exceptions, was less than corresponding streptococci isolated from persons ill during current epidemics but was roughly proportional to the severity of current infections especially infections of the respiratory tract. Those isolated in winter months during the sporadic, general or epidemic prevalence of influenza and milder types of respiratory infection produced a high incidence and proportionally severe or milder pneumonitis in mice and a low incidence of polio-encephalitis and were agglutinated specifically in high incidence by thermal antibody11 and antiserums prepared with streptococci isolated from persons ill with epidemic influenza or other respiratory infections. In sharp contrast, those isolated in summer during current epidemics of poliomyelitis produced a much higher incidence of polio-encephalitis and a greatly lower incidence of pneumonitis in mice on comparable inoculations. The striking difference in the incidence of pneumonitis and polio-encephalitis on injection of the respective streptococci was found to be due to seasonal differences in infectivity or localizing properties of the streptococci and not due to seasonal changes in susceptibility of lungs and brain respectively in mice.

The results in this study are in accord with those obtained by means of erythematous reactions to intradermal injections of thermal antibody and of antigen prepared with the respective streptococci. Presumptive evidence thus obtained which indicated the presence of “pneumotropic” streptococci in moderate incidence in well persons and in very high incidence in persons having respiratory infections in winter months during such epidemics5 and of “neurotropic” or “poliomyelitic” streptococci in moderate incidence in well persons during epidemics of poliomyelitis and in very high incidence in persons having poliomyelitis in summer.13 The parallelism in cultural characteristics, type of virulence or elective localizing and agglutinative properties of the streptococci isolated seasonally from outdoor air and those isolated from persons ill during respective epidemics was most striking.

The data obtained indicate (1) that the streptococci isolated from outdoor air were in fact similar or identical to those isolated seasonally from well persons in moderate incidence and from persons ill in high incidence during epidemics of respiratory infections and poliomyelitis, respectively; (2) that the respective streptococci in outdoor air may in part be derived from “human” sources, and (3) that the presence in high incidence seasonally of the respective streptococci in outdoor air may through respired air be responsible in part for the widespread occurrence of epidemics of respiratory infections in winter and of encephalitis and poliomyelitis in summer and in studies of which the respective streptococci are uniformly isolable by adequate methods. The questions regarding the presence also of filtrable agents or viruses in outdoor air and the relation of these to the streptococci will be considered in separate reports.

^aGrateful acknowledgement is hereby made for the co-operation of Mr. Gordon Strachan of Public Relations of the Atchison, Topeka and Santa Fe Railway Co. for permitting samplings on Diesel locomotives in transit from Los Angeles to Chicago, and for the privilege of making samplings on Pullman cars granted by Mr. W. P. Bristow, superintendent of the Cincinnati Division of The Pullman Co.

References Cited:

1. Cruickshank, R., and Godber, G. E.: The aerial spread of streptococcal infections. Lancet, 1:741-746, 1939.
2. Loosli, C. G.: The pathogenesis and pathology of experimental air-borne influenza virus A infection in mice. J. Infect. Dis., 84:156-168, 1949.
3. Loosli, H. M.; Lemon, H. Wise, and Robertson, O. H.: Studies on the transmission and control of respiratory disease within army barracks. J. Infect. Dis., 82:58-71, 1948.
4. Mudd, S.: Confined air as vehicle of infection; Current progress in air sanitation. M. Clin. North America, 28: 1293-1310, 1944.
5. Rosenow, E. C.: Diagnostic cutaneous reactions to intradermal injection of natural and artificial antibody and of antigen prepared from streptococci isolated in studies of diverse diseases. Ann. Allergy, 6:485-496, 1948.
6. Rosenow, E. C.: Epidemic encephalitis in North Dakota and Minnesota, 1941: studies on etiology, epidemiology, and serum treatment. Journal-Lancet, 63:247-257, 1943.
7. Rosenow, E. C.: A specific precipitin reaction in epidemic poliomyelitis. J.A.M.A., 84:429-432, 1925.
8. Rosenow, E. C.: Specific types of alpha streptococci and streptococcal precipitinogen in air in relation to epidemic infections of the respiratory tract and nervous system. Journal-Lancet, 65:108-122, 1945.
9. Rosenow, E. C.: Streptococci in air of operating room and wards during an epidemic of tonsillitis. Am. J. Obst. & Gynec., 50:762-768, 1904.
10. Rosenow, E. C.: Studies in influenza and pneumonia. III. The occurrence of a pandemic strain of streptococcus during the pandemic of influenza. J.A.M.A., 72:1604-1609, 1919.
11. Rosenow, E. C.: Studies on the nature of antibodies produced in vitro from bacteria with hydrogen peroxide and heat. J. Immunol., 55:219-232, 1947.
12. Rosenow, E. C.: Studies on the relation of pneumotropic streptococci to influenza virus. Am. J. Clin. Path., 15:362-380, 1945.
13. Rosenow, E. C.: A study of the 1946 poliomyelitis epidemic by new bacteriologic methods. Journal-Lancet, 68:265-277; 1948.
14. Wells, W. F.: Measurement of airborne infection by disinfection of air. Am. J. M. Sc., 209:177-180, 1945.