BREWING INDUSTRY RESEARCH FOUNDATION THE SIGNIFICANCE OF THE USE OF HOPS IN REGARD TO THE LOGICAL STABILITY OF BEER. II. THE DEVELOPMENT OF RESISTANCE TO HOP RESINS BY STRAINS OF LACTOBACILLI By M. Richards, M.Sc, and R. M. Macrae, B.Sc. Ph.D., M.I.Biol. (Brewing Industry Research Foundation, Nutfield, Surrey) Received \Oth June, 1904 Resistance to humulone develops rapidly in a number of strains of lactobacilli, 2-4 subcultures being required to increase resistance 8-20 fold to a value of greater than loojtg./ml. humulone, depending upon the strain. Increased resistance towards /sohumulone A develops similarly. Resistant populations retain a degree of stability upon subculture in the absence of humulone. This suggests that selection of resistant mutants contributes to the development of humuloneresistance. The humulone-resistant strains possess increased powers of infection of beer over the parents from which they were derived. In these laboratory trials it was necessary to increase the hop rate by not less than 50% in order to augment the antibacterial potency of the resulting beers. Introduction as Acetobacter spp., and the comparatively Following the demonstration10 that there sensitive Gram-positive bacteria such as was a correlation between the Gram-staining Laclobacillus spp. Shimwell8'0 recognized, reaction of a bacterium and its sensitivity to however, that varying degrees of sensitivity hop substances, it became clear that beer- to the hop resins existed amongst the Gramcontaminating bacteria could be divided into positive beer-contaminating bacteria. Such the insensitive Gram-negative organisms such variations in the degree of sensitivity amongst
Vol. 70, 1964] RICHARDS AND MACRAE: LOGICAL STABILITY OF BEER 485 different strains of lactobacilli have also been demonstrated by later investigators (Harris & Watson8; Hough, Howard & Slater8). Shimwell0' had demonstrated previously that the resistance of lactobacilli to the hop anti bacterial substances in beer could be reduced by prolonged serial subculture in unhopped beer. Macrae5 noted the emergence of resist ance in lactobacilli contaminating hopped beer, but very few quantitative experiments into the acquisition of resistance to the hop antibacterial agents in beer have been docu mented. The present work was therefore undertaken to investigate the ease with which resistance to hop resins is acquired by a number of different strains of lactobacilli, and to assess the significance of this pheno menon with regard to brewery and beer infection. Experimental Owing to its stability and ease of prepara tion, humulone was used as the main test compound, but isohumulone A has also been employed. Humulone was prepared from "Lupulin" (E. Clemens Horst Co., San Francisco) via the o-phenylene-diamine com plex (minimum m.p. 115 C.) and was stored as the crystalline solid under gaseous nitrogen at 3 C. /sohumulone A was prepared from the same starting material following the method of Howard,4 but was dissolved in M/100 Na2HPO4-KH8PO4 buffer ph 6-5 through which nitrogen had previously been bubbled vigorously for about 15 min. The solution was then stored under gaseous nitrogen at 3 C. No tsohumulone solution was used more than 5 days after it had been prepared. Solutions of humulone were made up freshly as required in the same buffer. Both the humulone and tsohumulone solu tions were assayed spectrophotometrically.7.11 The sensitivity of the lactobacilli to humu lone and tsohumulone A was estimated by preparing five two-fold dilutions from the original solution (i.e., original concentration to 1/32) in M/100 buffer ph 6-5. Portions (5 ml.) of each dilution were then pipetted aseptically into sterile double-strength M.Y.G.P..1* (5 ml.) ph 6-5. A tube con taining the buffer (5 ml.) and double-strength M.Y.G.P. (5 ml.) was included as a control. All the lactobacilli tested had been main tained for considerable period's on medium L.u These strains were of diverse origin, some having been isolated from infected beers, whereas others were "non-brewing" strains obtained from culture collections. All the strains are referred to by their laboratory reference number. Approximately 18 hr. M.Y.G.P. cultures of the lactobacilli, washed three times in sterile isotonic saline, were used to determine their sensitivities. The final saline suspension was diluted with saline to a standard opacity. A standard volume (0-2 ml.) of suspension was then inoculated into each of the tubes of the series, corresponding to an infection rate of approxi mately 3-5 x 10s cells per ml. of medium. All the tubes were incubated at 30 C. for 43 hr., after which the optical densities of the growths were read on a Hilger Spekker. The end-point was taken as that concentration of resin which still allowed trace growth, corres ponding to 6-11 X 10 cells per ml., of the organism. An attempt was made to increase resistance to humulone and tsohumulone A by serial subculture through M.Y.G.P. containing gradually increasing concentrations of the corresponding resin. In general the concen tration in each step of the series was twice the minimum inhibitory concentration as determined by the sensitivity test. The sensitivity of the bacteria after subculture into the higher concentration of resin, and therefore the rate at which resistance was attained, was then assessed. Eventually all strains would grow in the presence of 100 /Kg./ml. humulone, which is the maximum concentration attainable using the above technique. In a similar manner, the stability of the humulone-resistant strains was investi gated by determining the sensitivity of the culture resulting from subculture in M.Y.G.P. without added humulone or tsohumulone. The significance of the humulone-resistant strains of lactobacilli as beer-contaminants was determined by comparing the survival curves of this strain with that of the parent in a number of beers. These were brewed from amounts {21.) of wort (O.G. 10) prepared in a similar manner to that described by Macrae,8 and pitched with 5 g. (moist weight) of Sacch. cerevisiae N.C.Y.C. 10. The fermentations were carried out in tall tubes (5 ft. x 2 in. internal diameter) at 18 C, to a final gravity of 1010. Most of the yeast in suspension was then removed by centrifugation, and the supernatant beer was filtered sterile through an Oxoid membrane filter. Two beers were prepared in such a manner,
486 RICHARDS AND MACRAE: LOGICAL STABILITY OF BEER [J. Inst. Brew. an unhopped beer and a beer hopped at a rate equivalent to 1$ lb./brl. The corres ponding worts were boiled under reflux for 1 hr., and to prepare the hopped beer 9 g. of minced 1962 Petham Golding hops (a-acid content 4-93%, 6 months after picking) were added. Beers of hop rates equivalent to lb. and 1 lb./brl. were obtained by suitable dilution of the beer hopped at 1J lb./brl., with the unhopped beer. This particular experimental design was adopted in order to overcome any different extraction rates of hop substances which might occur if the beers of the lower hop rates had been pre pared by conventional methods. Such a procedure results in slightly weaker beers for the lower hop rates than if the beers are prepared conventionally. Beers of lower hop rates prepared by the dilution method were 5-15% less bitter estimated by the technique of Brenner, Vigilante & Owades1 than if prepared by boiling hops with the wort. Assuming the differences in antibacterial potency to be of a similar order, they could make little difference to the biological studies here reported. The beers were then distributed asepticauy into sterile McCartney bottles in 25-ml. amounts. Two series were prepared, con taining the three hopped and one unhopped beer, each sample being duplicated. One series was then inoculated with sensitive LactobaciUus and the other with resistant, in both cases to give a final concentration of 10 viable cells per ml. of beer. All beer samples were then forced at 25 C. under an increased CO2 tension obtained by partially evacuating the container to atm. three times, and replacing the evacuated air with CO2 on each occasion. Viable plate counts were performed using the Miles & Misra8 technique, on medium L of Williamson.13 Results and Discussion As is implied by the presence of lactobacilli capable of infecting hopped beers, strains of these organisms may possess increased resistance to hop resins. Figs. l(a) and (b) show the manner in which resistance to humulone develops for six different strains of lactobacilli. Resistance in all cases in creased very readily, and in all cases a final degree of resistance equivalent to a value of greater than loo/xg./ml. of humulone was obtained. The number of subcultures re quired to increase resistance to this level. 0 10 -W 60 Cultural concentration of humulone (ug. per ml.) to eo ioo Cultural concentration of humulone (ug. per ml.) Fig. 1 (a) and (6). Development of humuloneresistance in 6 strains of lactobacilli. however, varied from strain to strain, only two subcultures being required by Lactobacillus B28, while four subcultures were required by LactobaciUus B26. The final degree of resistance represents an eight to twenty-fold increase in resistance over the initial value. Humulone, being completely isomerized to the antibacterial isohumulone during the ml.) 60 :f.c 2» J, 8. / Bo/ 20» 60 Cultural cone, of /johumulone (ug. per ml.) Fig. 2. Development of tjohumulone-rcsistance in 2 strains of lactobacilli.
Vol. 70, 1064] RICHARDS AND MACRAE! LOGICAL STABILITY OF BEER 487 boiling of hops with wort, and during fermentation, cannot normally account for any of the antibacterial characteristics of beer. As may be seen from Fig. 2, relatively brief exposure to tsohumulone A raises the resistance of the two strains of lactobacilli tested to this compound. Although the degree of resistance was not increased to the same extent as for humulone, two subcultures increased resistance to tsohumulone in each strain sufficiently to suggest that the resist ance can generally be increased quite readily. This finding supports the assumption that results obtained for humulone may equally apply to tsohumulone. The demonstration of the ease with which lactobacilli develop resistance to humulone and tsohumulone A emphasizes the need for careful control measures in the brewery. Any source of Ladobacillus infection inside the brewery (e.g., contaminated pitching yeast), is likely to contain strains which will have become resistant to the hop antibacterial substances in beer. Lactobacilli gaining entry from outside the brewery, however, may well have had no previous contact with the hop resins, in which case the growth of such strains may be restricted by the hop antibacterial substances in the beer. As may be seen from Table I, subculture of resistant populations in the absence of humulone results in a slow decrease in the degree of resistance. It is interesting to note that Shimwell8 succeeded in reducing the resistance of lactobacilli to the antibacterial substances in beer, only after extensive subculture in unhopped beer over a period of a year. The comparative stability of this character suggests that selection of resistant mutants contributes to the over-all humuloneresistance of these strains of lactobacilli. Infection of a series of beers, hopped at various rates, with the humulone-sensitive parent organism,, and with the humuloneresistant strain derived therefrom, B10(R), reveals, as shown in Fig. 3, that the resistant I (a) Unhopped (c) Hopped I Ib./bri. I: ^ Days incubation 25 C. Days Incubation 25*C. Ib./brl. e (d) Hopped 4 Ib./brl. = 7 [j t.. S I I, Days incubation 25'C. k ' iv- Days incubation 25 C. To" Fig. 3 (a), (6), (c) and (d). Growth of humulone-scnsitive and humulone-resistant lactobacilli in beers hopped at different rates.
488 RICHARDS AND MACRAE: LOGICAL STABILITY OF BEER [J. Inst. Brew. TABLE I Stability of Humulone-Resistance Upon Subculture of Resistant Lactobacilli in the Absence op Humulone Sensitivity (jig./ml.) No. of subculture Strain Initially 1 2 3 4 BO Bl 60 00 47 35 35 37 42 23 23 strain possesses increased powers as a beer contaminant. Thus, comparing the survival curves of both strains in the hopped beers with those obtained in the unhopped beer, it is evident that the resistant organisms are inhibited to a lesser degree and that a recovery in viability occurs much more quickly and to a greater extent than with the sensitive parent strain. A further point of resemblance between this artificially derived humulone-resistant strain and lactobacilli which contaminate commercial beers, is that both strains may fail to grow on the surface of solidified media in the absence of an increased CO2 tension. The sensitive parent from which the humu lone-resistant strain was derived, however, gave perfectly satisfactory surface growth on solidified media without an increased CO8 tension. The results presented in Fig. 3 for the sensitive parent (B10) show that significant differences existed in preservative potency between the beers hopped at rates equivalent to lb. and 1 Ib./brl. The lower rate allowed recovery in the numbers of viable organisms, whereas the higher rate repressed growth during the period of incubation. No such differences were detected between beers hopped at rates equivalent to 1 lb. and 1 lb./brl. Thus the practice of increasing the hop rate, at this level, by a factor equal to or less than 50%, does not demonstrably increase the antibacterial potency of the resulting beers. Acknowledgement. The authors acknow ledge the encouragement of Dr. A. H. Cook, F.R.S., Director of this Foundation. References 1. Brenner, M., Vigilante, C, & Owades, J. L., Proc. A. M. Amer. Soc. Brew. Chem., 1950, 48. 2. Harris, J. O., & Watson, \V., this Journal, 1900, 151. 3. Hough, J. S., Howard, G. A., & Slater, C. A., this Journal, 1957, 331. 4. Howard, G. A., this Journal. 1959, 417. 5. Macrae, R. M., this Journal, in the press. 0. MUes, A. A., & Misra, S. S., /. Hyg. (Lond.), 1938 38 732 7. Rudin.A. D., this Journal, 1900, 18. 8. Shimwell, J. L., this Journal, 1930, 127. 9. Shimwell, J. L., this Journal, 1930, 2. 10. Shimwell, J. L., this Journal, 1937, 111. 11. Stevens, R., & Wright, D., this Journal, 1901, 490. 12. Wickerham, L. J., U.S. Dept. Agric. Tech. Bull., 1951, No. 1029. 13. Williamson, D. H., this Journal, 1959, 154.