RELATIONSHIP TO RESISTANCE IN KLEBSIELLA PNEUMONIAE

THE SIGNIFICANCE OF LACTOSE FERMENTATION AND ITS RELATIONSHIP TO RESISTANCE IN KLEBSIELLA PNEUMONIAE VICTOR J. CABELLI' AND M. J. PICKETT Department of Bacteriology, University of California, Los Angeles, California In experiments similar to those of Humphries (1944), in which we used brilliant green to obtain smooth and rough phase variants from mucoid strains of Friedlander's bacillus, it was found that in nearly all of the field strains a correlation existed between resistance to the dye and ability to ferment lactose. It was found also that this correlation extended to growth rate and to growth in extreme ranges of ph. The correlation was such that lactose positive strains of Friedlander's bacillus were more resistant to brilliant green, had a faster growth rate, and were better able to grow in extreme ph ranges than were the lactose negative strains which were studied, In view of this correlation and with an eye toward its practical application, the lactose positive and negative strains were compared in respect to their sensitivity to various antibiotics. A limited investigation of the comparative pathogenicity of the strains also was undertaken. It was found also that when serologically typeable lactose negative strains of KlebsieUa pneumoniae were grown in the presence of increasing concentrations of brilliant green, two patterns of resistance appeared. In the first, which occurred quite commonly, the resistance increased without any increased growth rate or increased tolerance to ph changes; in fact, in some cases a decrease in growth rate was observed. There was also no change toward an ability to ferment lactose. The second pattern occurred quite rarely; indeed, it was observed only five times in the course of these studies. In this pattern, the resistance increased quite rapidly, and with this increase there was also a gain in the ability to ferment lactose, an increased growth rate, and a greater tolerance to growth in extreme ph ranges. 1 Present address: Department of Bacteriology and Preventive Medicine, School of Medicine, University of Missouri, Columbia, Missouri. Received for publication April 2, 1953 MATERIALS AND METHODS The strains were grown in a medium of 1 per cent peptone (Difco) in distilled water. After twenty-four hours' incubation at 37 C, the tubes were standardized turbidimetrically so that there was 80 per cent light transmission, using an uninoculated tube as a standard. All turbidimetric readings were performed on a Coleman Junior spectrophotometer using light at a wavelength of 450 m,u. Three-tenths ml amounts, which contained about 109 cells, were inoculated into tubes of peptone broth to which varying dilutions of brilliant green were added. Lactose fermentation tubes were inoculated from the standardized cultures using a 4 mm loop. The highest concentration of brilliant green in which there was visible growth was noted as was the time of fermentation, that is, the time the indicator (brom cresol purple) turned a definite yellow. Tubes of the peptone medium at varying initial ph values were inoculated similarly in experiments designed to show the growth response of the strains to varying hydrogen ion concentrations. The ph was adjusted by the use of a Beckman ph meter. In studying the growth curves, a similar standardization procedure was followed with the exception that the turbidity was standardized to a reading of 86 per cent transmission and that 0.1 ml inocula were used. Turbidimetric readings were taken hourly during the early part of the run and every two hours thereafter. The study of the sensitivity of the strains to antibiotics was carried out using penicillin, streptomycin, aureomycin, chloramphenicol, and terramycin. The inocula were standardized by the usual procedure, and serial dilutions of the antibiotics in peptone broth were inoculated. Approximately that concentration of the antibiotic which was found to be inhibitory for the most sensitive strain of K. pneumoniae used by Frank et al. (1950) was taken as the highest dilution. 443

444 VICTOR J. CABELLI AND M. J. PICKETT [VOL. 66 The tubes were read at twenty-four and fortyeight hour intervals. The comparison of pathogenicity was carried out by intraperitoneal injection and intranasal infection. White mice weighing approximately 25 g each were used. For each strain three mice were infected by depositing 0.1 ml of a suspension of cells on their nostrils while they were in a lightly anesthetized state, and one mouse was RESULTS The data for the correlation of lactose fermentation with resistance to brilliant green and with growth in extreme ph ranges are summarized in table 1. The results show that lactose negative strains are much more sensitive to both conditions and that in the lactose positive strains a semiquantitative relationship exists between time of fermentation and sensitivity to the dye. The TABLE 1 Time of fermentation, sensitivity to brilliant green, and sensitiv?ty to initial hydrogen ion concentrations in the 8tock strains STRIN CONCENTRAnON O1 DYE, 1: TIaM OF ntal ph 01ME FEREN- 10 mil 5 mil 2.5 mil I mil 800 th 500 th 2S0 th 100 th 5o th TATION 4.0 5.0 8.0 9.0 10.0 SAl + _ - >10 dy - 4 - SA2 4+ 4+ 13 hr 4+ SA3 + - _. >10 dy - 4 - SA4 + -. _ - >lo dy - - a - SA5 4+ - 14 hr 4+ 4+ SA6 + -. >10 dy - i - SA6-L 4+ 4+ 15 hr 4+ SB1 4+ 4+ _ 13 hr 4+ 4+ SB2 4+. _ 38 hr + SB3 4+ 15 hr 4+ 4+ SB4 - - - _ 14 hr 4+ SB5 8 hr 4+ 4+ SC1 4+ - - - _ 15 hr 4+ 4+ + SC2 4+ 10 hr 4+ 4+ SC3 4+ 10 hr 4+ UBI 4+ 4+ - 13 hr UCi 4+ 8 hr 4+ mil-million; th-thousand; dy-days; hr-hours. injected intraperitoneally with the same amount. The suspensions contained about 108 cells per ml. Sixteen strains, of which four were lactose negative, were grown in increasing concentrations of the dye. In the course of four months, lactose negative strains initially inhibited by a 1:5,000,- 000 dilution of brilliant green were able to tolerate a dilution of 1:150,000. The cultures were transferred every other day using an inoculum of 0.1 ml; if there was no growth after forty-eight hours, a larger inoculum was used. The concentration of the dye was increased when good growth was obtained from a 0.1 ml inoculum in the previous culture. After four months, when the resistance to brilliant green had been increased greatly, growth studies were performed according to the procedure outlined previously. growth curves presented in figure 1 demonstrate the relationship of growth to the preceding characteristics. The results set forth in figure 2 indicate that although there was a considerable increase in resistance when lactose negative strains were passed through increasing concentrations of brilliant green, there was no increase in growth rate; in fact, in some cases there was a decrease. Neither was there an increase in ph tolerance nor an increase in lactose fermentation. This is the first pattern of increased resistance to the dye. It should be pointed out that in subculturing lactose negative cells to a higher concentration of brilliant green, it was necessary to use heavy inocula and to subculture in replicate so as to obtain growth in at least one of the tubes.

LACTOSE FERMENTATION IN K. PNEUMONIAE When growth was obtained, it was usually very meager. However, on very rare occasions it was observed that there was luxuriant growth in one of the replicate tubes which contained a concentration of the dye higher than that of the inoculum. Upon testing these variants, it was found that there was a great increase in resistance to the dye, that they had gained the AA45 pathogenic for mice by both the intraperitoneal and the intranasal routes; the lactose positive strain, on the other hand, was found to be pathogenic by the intraperitoneal route only (table 2). Both of the strains used in the study of pathogenicity were type A and were strongly capsulated. The dissfimilarity in the sensitivity of the lactose negative and positive strains to various 0I 70 so 90 901 Figure 1. Growth curves of lactose positive and lactose negative strains. ability to ferment lactose, that they were able to grow in media of increased acidity and alkalinity (table 1, SA6-L), and that they had increased growth rates (figure 2, SA6-lac). In effect, they behaved as typical lactose positive strains. This is the second pattern of resistance to the dye, and it was observed at least once in three of the four lactose negative strains which were used in these experiments. The lactose negative strain was found to be i00 Figure B. Growth curves of lactose negative strains before and after passage through brilliant green. antibiotics can be seen from table 3. Strains SAl, 3, and 4 were lactose negative; strain SB2 was a late lactose fermenter, and strains SB5 and SC3 fermented the sugar within twentyfour hours. The fact that SA4 cells grew in concentrations of aureomycin, chloramphenicol, and terramycin which were expected to be inhibitory was extremely disquieting until it was observed that cells from the tubes which showed growth fermented lactose within seventy-two hours. It should be noted also that the lactose negative strains were much more difficult to keep in t0 to

446 VICTOR J. CABELLI AND M. J. PICKETT [VOL. 66 TABLE 2 Comparative pathogenicity of lactose positive and lactose negative strains STRAN ROUT Ol INFCTON T=M 0P DEAT hr SA2 IP 72 SA3 IP 12 SA3 IN 18 SA3 IN 20 SA3 IN 24 IP-intraperitoneally; IN-intranasally. * Alive after eight days. TABLE 3 Sensitivity of lactose positive and lactose negative strains to various antibiotics Penicillin ANTIBIOTIC Streptomycin Aureomycin Chloramphenicol Terramycin CONC IN AG/ML 400 100 20 105 25.6 6.4 1.6 0.8 SA1'SA3 SA41SB2!SB5ISC3 I I 4+14+14+ 4+ 4+14+ - 1 _ 4+ 4+ - + 4+ 4+ + DISCUSSION Since the genetic and taxonomic aspects of these studies are still under investigation, further discussion along these lines will be presented at a later date when more data are available. However, the basic diimilarity between the lactose negative and positive strains and the ability of the former to mutate to the latter (Henriksen, 1950) appears to be well established. This difference exists in serologically similar if not identical strains (Cabelli and Pickett, 1953). Studies on the lactose fermentation of this organism are now in progress in order to ascertain the mechanism of this linkage. Although at the present time the consensus of opinion is that both the lactose negative and positive strains are the causative agents of Friedlander's pneumonia, there is some evidence that the lactose negative organism may be the sole etiological agent (Perkins, 1904; Coulter, 1917). Furthermore, in the only studies in which we could find mention that the intranasal route of infection was used, the organisms were definitely lactose negative in one case and probably so in the other (Branch, 1927; Friedlander, 1882). It is also of interest to note that "A" is the most common serological type isolated in cases of Friedlander's pneumonia and that all our lactose negative strains of K. pneumoniae were of that serological type. SUMMARY - A correlation was shown between the ability - to ferment lactose and resistance to brilliant - green, to extreme ph ranges, and to antibiotics. - The correlation also extended to growth rate and to a lesser extent to pathogenicity. Two patterns of resistance -to the dye in 4+ lactose negative organisms were described. In 4+ one the resistance was found to be specific, and 4+ in the other the ability to ferment lactose as 4+ well as other correlated characteristics was gained concurrently with the resistance to the dye. - The results as well as future lines of investiga- - tion were discussed briefly. culture and needed transplanting at more frequent intervals. REFERENCES BRANCH, A. 1927 Spontaneous infections of guinea pigs. J. Infectious Diseases, 40, 533-548. CABELLI, VICTOR, J., AND PICKRETT, M. J. 1953

1953] LACTOSE FERMENTATION IN K. PNEUMONIAE 447 The precipitating antigens of Friedlander's bacillus. J. Gen. Microbiol., in press. COULTER, C. B. 1917 The biological entity of the Friedlander bacillus. J. Exptl. Med., 26, 763-768. FRANK, P. F., WILcox, C., AND FINLAND, M. 1950 In vitro sensitivity of coliform bacilli to seven antibiotics (penicillin, streptomycin, bacitracin, polymyxin, aerosporin, aureomycin, and chloromycetin). J. Lab. Clin. Med., 35, 188-205. FRIEDLANDER, C. 1882 Quoted from Edwards 1929 Relationships of the encapsulated bacilli with special reference to Bact aerogenes. J. Bact., 17, 339-353. HENRIKSEN, S. D. 1950 Mutative lactose fermentation in Klebsiella. Acta Pathol. Microbiol. Scand., 27, 35-40. HUMPHRIES, J. C. 1944 Bacterial variation: The influence of environment upon the dissociation pattern of Klebsiella pneumoniae. Yale J. Biol. and Med., 16, 639-658. PERKINS, R. G. 1904 Bacillus mucosu8 capsulatus: a study of the group and an attempt at classification of the varieties described. J. Infectious Diseases, 1, 241-266.