Scientific Publications - Work Done by Microbiology Reader Bioscreen C

Journal of Applied Microbiology 2001 Jul;91(1)147-53

Production of growth-inhibiting factors by Lactobacillus delbrueckii

van de Guchte M, Ehrlich SD, Maguin E.

ABSTRACT

AIMS: The detection of growth-inhibiting factors produced by Lactobacillus delbrueckii. METHODS AND RESULTS: A bioscreen assay was developed to study the effect of Lact. delbrueckii culture supernatant fluids on the growth of phylogenically or functionally related bacteria in broth cultures. Several growth-inhibiting factors could be distinguished based on differential effects on different test strains, separation by ultrafiltration and sensitivity to heat, proteinase treatment or catalase addition. CONCLUSION: Lactobacillus delbrueckii strain VI1007 was found to produce at least three growth-inhibiting factors, other than lactic acid, when grown under microaerobic conditions in MRS broth. These included H2O2 and a bacteriocin-like, heat- and proteinase-sensitive bactericidal molecule or complex with a molecular weight greater than 50 kDa. A third factor inhibited the growth of Streptococcus thermophilus. SIGNIFICANCE AND IMPACT OF THE STUDY: The assay system used allows the detection of subtle interactions between strains, that are likely to be of ecological importance in mixed cultures but would go unnoticed in classical agar diffusion tests.

INTRODUCTION

Lactobacillus delbrueckii is an industrially important organ­ism, belonging to the group of lactic acid bacteria that are widely used in food fermentation. These bacteria produce lactic acid as one of their metabolic end-products, the accumulation of which results in a rapid acidification of the culture medium, ultimately causing the growth arrest of spoilage bacteria and of the lactic acid bacteria themselves. Lactic acid production thus aids the conservation of the food. Apart from lactic acid, these bacteria may produce a number of other substances with the potential to suppress the growth of spoilage bacteria or of other strains of the same species. Notably, an ever-increasing number of bacteriocins are being isolated from these bacteria, which have been divided into three classes: (I) lantibiotics, (II) small heat­stable non-lantibiotics and (III) large heat-labile proteins (Klaenhammer 1993). Representatives of classes I and II have been characterized at the molecular level and recent years have seen important progress in the understanding of the mechanism of action of these proteinaceous compounds (for reviews see Nes et al. 1996; Moll et al. 1999). In contrast, the structural gene has only been cloned for one of the class III bacteriocins, helveticin J (Joerger and Klaenhammer 1990). In addition to these substances, lactic acid bacteria may produce metabolites such as H₂0₂, organic acids other than lactic acid or ethanol, that may cause growth inhibition.

In spite of its major importance in dairy fermentation, where this bacterium is often used in coculture with Streptococcus thermophilus, Lact. delbrueckii has been strongly under-represented in studies of growth-inhibiting factors (Miteva et al. 1998a). Only recently, the screening of a large number of strains indicated that bacteriocins were also widespread among Lact. delbrueckii (Miteva et al. 1998b). The detection methods commonly applied to reveal the presence of inhibitory substances rely on the diffusion of these substances through an agarose-solidified growth medium. Here we demonstrate the presence of inhibitory substances in the culture supernatant fluid of Lact. delbrueckii VI1007 that could not be detected on solidified media and would, therefore, go unnoticed in the classical inhibition tests.

MATERIALS AND METHODS Strains and sample preparation

Lactobacillus delbrueckii subsp. bulgaricus strains VI1005, VI1015 and ATCC 11842, Lact. delbrueckii subsp. lactis VI1007 and Strep. thermophilus CNRZ 302 were grown at 42°C under microaerobic conditions in MRS broth (Difco Laboratories) with an initial pH of 6^.4. The cells from culture samples were removed by centrifugation. The pH of the supernatant fluids was adjusted to 6^.4 with NaOH, after which the supernatant fluids were sterilized by filtration and stored in aliquots at -20°C until further use.

Bioscreen assay

Test strains were grown to an optical density at 600 nm (O.D.₆₀₀) of 0^.1-0^.2 (path length 1 cm), after which the culture was diluted in fresh MRS medium to an OD₆₀₀ of 0^.011. A volume (50 µl) of the diluted test strain culture and 300 µl of the supernatant fluid to be tested were mixed in a microwell plate and incubated at 42°C in a Bioscreen C plate reader (Labsystems, Helsinki, Finland). Every 15 min, the plate was shaken briefly and the OD₆₀₀ measured. When indicated, incubations were performed in the presence of catalase (from bovine liver; Sigma), which was added to a final concentration of 5 µg ml^-1.

Plate count assay

Test strains were prepared and mixed with the VI1007 culture supernatant fluids as described for the Bioscreen assay. The volumes were scaled up to obtain a final mix of approximately 1 ml, which was incubated in an Eppendorf tube at 42°C. Immediately after mixing cells and super­natant fluid, and at consecutive 2-h intervals, 100 µl of the mix and convenient dilutions were plated on MRS agar to determine the number of colony-forming units (cfu).

Ultrafiltration

A VI1007 culture supernatant fluid was filtered through ultrafiltration membranes with different cut-off values using Ultrafree centrifugal filter devices (Millipore, Bedford, MA, USA) according to the supplier's instructions. The concen­trate was resuspended in the same volume of MRS. The antimicrobial activity in the filtrate and resuspended concentrate was evaluated in a Bioscreen assay in the presence of catalase.

Proteinase sensitivity assays

Culture supernatant fluids were incubated for 30 min at 37°C with pronase (4 mg ml^-1), proteinase K (0^.5 mg ml^-1), trypsin (1 mg ml^-1) or a-chymotrypsin (1 mg ml^-1), after which the antimicrobial activity was evaluated in a Bioscreen assay or in a plate count assay (a-chymotrypsin treatment only).

a-chymotrypsin treatment of test strain V11005

VI1005 was grown to an O.D.₆₀₀ of 0^.1-0^.2, after which a-chymotrypsin (1 mg ml^-) was added. The mixture was incubated for 30 min at 37°C, after which the cells were washed and resuspended in Ringer solution and used in a plate count assay as described above.

RESULTS

Inhibition of Lactobacillus delbrueckii

When searching for the presence of growth-promoting or -inhibiting factors in the supernatant fluids of Lact. delbrueckii subsp. bulgaricus VI1005 and Lact. delbrueckii subsp. lactis VI1007 MRS-grown cultures, we observed that the supernatant fluids of strain VI1007 affected the growth of VI1005 and VI1007 differently. In order to analyse these effects in more detail, samples were withdrawn from a VI1007 culture at different time points (Table 1) and processed as described in Materials and Methods. Since no antagonistic effects could be observed in classical agar diffusion tests, a Bioscreen assay (Materials and Methods) was developed to analyse the effect of the culture superna­tant fluids (sl-s6) on the growth of test strains VI1007 and VI1005. In this assay the amounts of test cells and culture supernatant fluid used proved to be critical parameters. When too many cells were used or when the supernatant fluids were diluted, the growth-inhibiting effects were rapidly lost.

The two test strains appeared to be differently affected by the VI1007 supernatant fluids. Whereas the addition of the supernatant fluids caused a reduction in the growth rate of VI1007 (Fig. 1), the principal effect on VI1005 (Fig. 2a) seemed to be a delay of the onset of growth, leaving the growth rate largely unchanged. For both VI1007 and VI1005 the final density of the cultures was reduced compared with that of an MRS-grown culture. In both cases, the effector molecules appear to accumulate in the growth medium, supernatant fluids of older cultures generally exerting a more pronounced effect than the supernatant fluids of younger cultures.

Control experiments using MRS medium, the pH of which had been lowered to 4^.7 by the addition of lactic acid and subsequently readjusted to pH 6^.4 with NaOH, showed that the effects on the two test strains were not due to the accumulation of lactate in the VI1007 culture medium (results not shown). Growth of the test strains after the

Table 1 Characteristics of Lactobacillus delbrueckii VI1007 cultures at time of sampling

Supernatant fluids of three independent MRS-grown cultures of VI1007 were sampled at the times indicated to test for the presence of growth-inhibiting factors. The O.D.₆₀₀ and the pH of the cultures at the time of sampling were determined. Samples sl-s6 were taken from one culture.
T, Time after inoculation of the culture (O.D.₆₀₀ at T = 0 between 0^.02 and 0^.05); o/n, overnight.
addition of H₂O instead of the culture supernatant fluids excluded nutrient exhaustion as a possible cause of the effects of the VI1007 supernatant fluids (results not shown).

Growth of V11007 is affected by H2O2

As it is known that Lact. delbrueckii may produce ^H2^O2 when grown in a (micro) aerobic environment (Marty-Teysset et al. 2000), the Bioscreen assays were repeated after the addition of catalase to the VI1007 supernatant fluids. The addition of catalase abolished the growth rate-reducing effect on VI1007 (Fig. 1) while only in supernatant fluid s6 did a slightly extended lag phase persisted, indicating that

this effect was due to the presence of _H2O2 in the spent supernatant fluids. Moreover, the addition of catalase allowed a higher final density to be reached. The same effects of catalase addition, i.e. a higher growth rate and final density, were observed in the MRS-grown control culture indicating that, in microaerobically-grown cultures of Lact. delbrueckii, the production of _H2O2 contributes to growth arrest. The effect of the supernatant fluids on the growth of VI1007 was not affected by heating (15 min at 100°C) of the supernatant fluids (results not shown).

Growth of V11005 is affected by a thermolabile bactericidal factor in addition to H₂O₂

Catalase treatment of the VI1007 culture supernatant fluids also improved the growth rate and the final density of VI1005, but an apparently prolonged lag phase in the growth curves persisted (Fig. 2b). The prolonged lag phase could be due to the presence of a bacteriostatic or bactericidal factor in the VI1007 supernatant fluids. In order to discriminate between these two possibilities, a plate count assay was performed which showed that the effect on VI1005 was bactericidal. During incubation of VI1005 cells in the VI1007 supernatant fluids, the number of cfu initially decreased (1000-fold during the first 4 h of incubation in supernatant fluid s6, 100-fold in supernatant fluid s5) to rise afterwards or to fall to 0 when the supernatant fluid of a 24-h culture of VI1007 was used (results not shown). In the control experiment (VI1005 incubated in MRS), the number of cfu had increased 40-fold after 4 h. Therefore, the seemingly delayed onset of growth of VI1005 in the VI1007 supernatant fluids can be explained by a reduction in the initial number of viable cells in the culture.

Fig. 1 Growth of Lactobacillus delbrueckii VI1007 in MRS medium (q) and in VI1007 culture supernatant fluids (Table 1) s2 (∆), s4 (à) and s6 (O). Closed symbols, cultures in the presence of catalase

Fig. 2 (a) Growth of Lactobacillus delbrueckii VI1005 in MRS medium and in VI1007 culture supernatant fluids sl-s6 (Table 1). q, MRS; +, sl; ∆, s2; x, s3; à, s4; *, s5; O, s6. (b) Growth of Lact. delbrueckii VI1005 in MRS medium and in VI1007 culture super­natant fluid s4 (Table 1). q, MRS; à, s4; ∆, MRS heated to 100°C for 15 min; O, s4 heated to 100°C for 15 min. Closed symbols, cultures in the presence of catalase

 These results show that VI1007, in addition to H₂O_2i produces a factor reminiscent of a bacteriocin with a bactericidal effect on VI1005, to which VI1007 itself is resistant. This factor proved to be thermolabile. Heating of the VI1007 supernatant fluids for 15 min to a temperature between 70 and 100°C completely eliminated the growth lag of VI1005 (Fig. 2b), while heating to 50°C resulted in a reduction of the lag phase (results not shown). In the heat­treated supernatant fluids, VI1005 exhibited a slower growth than in MRS, reminiscent of the effect of the supernatant fluids on the growth of VI1007. This difference disappeared when, in addition to the heat treatment, catalase was added to the supernatant fluids (Fig. 2b), thus confirming the presence of two distinct growth-affecting factors in the VI1007 supernatant fluids.

The bactericidal factor produced by V11007 is a proteinaceous factor

Ultrafiltration of the VI1007 supernatant fluids, followed by testing of the filtrates and concentrates resuspended in MRS, indicated that the bactericidal activity was associated with a molecule or complex with a molecular weight in the 50-100 kDa range. While the active molecule was clearly retained by filters with a cut-off value of 50 kDa, part of the activity was found in the filtrate after filtration through a membrane with a cut-off value of 100 kDa (Fig. 3).

Bioscreen assays aimed at testing the proteinase sensitivity of the bactericidal factor did not yield conclusive results. Although a partial inactivation of the bactericidal factor seemed to occur, the prolonged lag phase in the growth of VI1005 could not be completely eliminated by the treatment of the VI1007 supernatant fluids with any of four proteinases tested (pronase, proteinase K, trypsin or a-chymotrypsin; results not shown). A plate count assay with a-chymotryp­sin-treated supernatant fluids, however, revealed a clear effect of proteinase treatment (Table 2). Whereas the growth of Lact. delbrueckii VI1005 in MRS was not affected by a-chymotrypsin treatment, the bactericidal effect of the VI1007 supernatant fluid was nearly annihilated by the proteinase treatment. In contrast, an immediate growth (increase in number of cfu) was not restored, explaining the

persistence of a lag phase in the Bioscreen assays. These results strongly suggested that the bactericidal factor is a protein, although the possibility remained that it is the target on the VI1005 cell surface rather than the bactericidal factor that is a protein. To rule out this possibility, VI1005 cells were treated with a-chymotrypsin and washed in Ringer solution before use in a plate count assay. In this experiment, similar results were obtained with the a-chymotrypsin-treated cells and cells that had not been treated with the proteinase (results not shown). This result supports our conclusion on the proteinaceous character of the bactericidal factor.

At present it is not clear whether the lag phase in the a-chymotrypsin-treated supernatant fluid is due to an incom­plete inactivation of the bactericidal factor or to the activity of a degradation product. Doubling the amount of a-chymo­trypsin did not change the result (not shown). a-Chymotryp­sin treatment allowed VI1005 to grow in VI1007 supernatant fluids that otherwise completely inhibited growth. However, while growth and cell separation in MRS medium did not seem to be affected by the addition of a-chymotrypsin, VI1005 systematically formed very long filaments in these proteinase-treated VI1007 supernatant fluids.

Fig. 3 Results of ultrafiltration assays. Growth of Lactobacillus delbrueckii VI1005 in MRS medium (n), in VI1007 culture supernatant fluid sA (*; Table 1), in filtrates of sA after ultrafiltration through a filter with a cut-off value of 50 kDa (O) or 100 kDa (∆) and in the corresponding concentrates (● and ▲) resuspended in one volume of MRS. Catalase was present in all samples

Table 2 Results of plate count assays of Lactobacillus delbrueckii VII005

Cells were incubated in MRS or in VI1007 culture supernatant fluid sB (Table 1), in the same media treated with a-chymotrypsin (30 min at 37°C) or in sB previously heated to 37°C (30 min).

To, T_4i Time after mixing of cells and MRS or supernatant fluid (0 and 4 h, respectively). Numbers represent cfu 100 µl^-1.

Of four different Lact. delbrueckii strains tested, including VI1005, VI1007, VI1015 and the Lact. delbrueckii subsp. bulgaricus type strain ATCC 11842, only strain VI1007 produced this bactericidal factor, and all but VI1007 were sensitive to the factor produced by VI1007 (results not shown). All these strains produced H₂O₂, as judged by the observation that the growth-limiting effect of the culture supernatant fluids of these strains on VI1007 could be eliminated by the addition of catalase (results not shown).

Inhibition of Streptococcus thermophilus

In dairy industry applications, Lact. delbrueckii is often found in conjunction with Strep. thermophilus. Therefore, we examined the effect of the Lact. delbrueckii VI1007 culture supernatant fluids on Strep. thermophilus. The growth of Strep. thermophilus CNRZ 302 appeared to be inhibited by these supernatant fluids and also in this case, the supernatant fluids of older cultures exerted a more pronounced effect than the supernatant fluids of younger cultures, indicative of the accumulation of a growth-inhibiting factor in the culture medium (Fig. 4). However, the growth inhibition could not be alleviated by the addition of catalase nor by heat treatment of the supernatant fluids (results not shown), suggesting that another factor than those described above is responsible for the inhibition of Strep. thermophilus. The bactericidal effect of this factor was established in a plate count assay where the number of cfu of Strep. thermophilus CNRZ 302 diminished 100-fold after 6 h of incubation in VI1007 supernatant fluid sB (Table 1).

DISCUSSION

During the growth of Lact. delbrueckii VI1007 under microaerobic conditions in MRS broth various growth­inhibiting factors appeared to accumulate in the culture medium. One of these factors affected the growth of two test strains, VI1007 and VI1005. The molecule involved in this

inhibition appears to be H₂O2_i since the growth rate­reducing effect was eliminated by the addition of catalase. This result corroborates earlier observations of ^H2^O2 production by lactobacilli under (micro) aerobic conditions (Murphy and Condon 1984; Marty-Teysset et al. 2000). In spite of the reported presence of antioxidative and ^H2^O2 scavenging activities in Lact. delbrueckii cell extracts (Lin and Yen 1999), _H2O2 production appears to contribute significantly to the growth arrest of these bacteria under microaerobic conditions, as judged from the beneficial effect of catalase addition to MRS cultures.

A second factor affected the growth of test strain VI1005 and of two other strains tested, but not of the producing strain VI1007, a property reminiscent of bacteriocins. Its bactericidal effect on VI1005 was annulled by proteinase treatment, demonstrating the proteinaceous character of this factor. After the proteinase treatment a bacteriostatic effect remained, however, and growth of the test strain VI1005 only resumed after a prolonged lag phase. This result suggests that a degradation product of the factor can still bind to the test cells and impair their growth. The bactericidal molecule or complex appears to be large, between 50 and 100 kDa, and thermolabile. These proper­ties seem to fit the criteria of a class III bacteriocin. Only very few bacteriocins belonging to this class of large heat­labile bacteriocins are known to date (Dodd and Gasson 1994) and only one gene involved in the production of such a protein, helveticin J, has been characterized at the molecular level (Joerger and Klaenhammer 1986; Joerger and Klaenhammer 1990). Interestingly, helveticin J and the other large heat-labile bacteriocins (lacticin A, lacticin B, acidophilucin A, caseicin 80 (Dodd and Gasson 1994 and references therein) and helveticin V-1829 (Vaughan et al. 1992)) all originate from Lactobacillus strains.

The thermolabile character of the factor produced by VI1007, with a considerable reduction in its effect already visible after incubation for 15 min at 50°C, seems to point to an enzymatic activity as the causative agent. In addition, it may explain why the maximal activity observed in the supernatant fluids of VI1007 cultures, grown at 42°C, is very low. This low activity necessitated the use of a detection method with a higher sensitivity than the classically used agar diffusion tests. Whereas the latter did not reveal the inhibitory effects described in this study, they were clearly exposed in the Bioscreen test, where the ratio of inhibiting substance and number of test cells can be more easily adjusted. The results of this test show that the inhibiting effects of VI1007 are important enough to be of ecological importance in the competition between strains, and especially to prevent the growth of a second strain once the producing strain has established itself.

Lactobacillus delbrueckii strains VI1007 and VI1005 have been classified as members of the subspecies bulgaricus on the basis of their sugar-fermenting capabilities (L. Benbadis, personal communication). Information based on more recently developed criteria such as random amplified polymorphic DNA profiles and restriction analysis of polymerase chain reaction-amplified rDNA (Giraffa et al. 1998) suggests that strain VI1007 belongs to the subspecies lactis rather than bulgaricus (E.M. Lim and P. Tailliez, personal communication; L. Benbadis and T. Smokvina, personal communication). For both subspecies it has been reported that Lact. delbrueckii and Strep. thermophilus mutually benefit from each other's presence in a mixed culture in milk (Accolas et al. 1977). In this medium, Lact. delbrueckii would stimulate Strep. thermophilus through its proteolytic activity which liberates peptides and amino acids that can be used by Strep. thermophilus. The bactericidal effect of the VI1007 supernatant fluids on Strep. thermophilus observed in this study, which is in apparent contradiction to this stimulating effect, may therefore be strain specific. Alternatively, production of the bactericidal factor may depend on the growth medium or the inhibiting effect may be masked by the stimulating effect in milk.

In conclusion, the results of this study indicate that the ecological relationships between lactic acid bacteria may involve rather subtle interactions that would go unnoticed in classical agar diffusion tests.

Fig. 4 Growth of Streptococcus thermophilus CNRZ 302 in MRS medium (q) and in VI1007 culture supernatant fluids (Table 1) s2 (∆), s4 (à) and s6 (O)

ACKNOWLEDGEMENTS

This work was financed in part by Danone and Rhodia Food. The authors thank Danone and Rhodia Food for providing strains VI1005, VI1007 and VI1015.

REFERENCES

Accolas, J.P., Bloquel, R., Didienne, R. and Regnier, J. (1977) Proprietes acidifiantes des bacteries lactiques thermophiles en relation avec la fabrication du yoghourt. Le Lait 57, 1-23.

Dodd, H.M. and Gasson, M.J. (1994) Bacteriocins of lactic acid bacteria. In Genetics and Biotechnology of Lactic Acid Bacteria eds Gasson, M.J. and De Vos, W.M. pp. 211-251. London: Blackie Academic & Professional.

Giraffa, G., De Vecchi, P. and Rossetti, L. (1998) Identification of Lactobacillus delbrueckii subspecies bulgaricus and subspecies lactis dairy isolates by amplified rDNA restriction analysis. Journal of Applied Microbiology 85, 918-924.

Joerger, M.C. and Klaenhammer, T.R. (1986) Characterization and purification of helveticin J and evidence for a chromosomally determined bacteriocin produced by Lactobacillus helveticus 481. Journal of Bacteriology 167, 439-446.

Joerger, M.C. and Klaenhammer, T.R. (1990) Cloning, expression, and nucleotide sequence of the Lactobacillus helveticus 481 gene encoding the bacteriocin helveticin. ,journal of Bacteriology 172, 6339-6347.

Klaenhammer, T.R. (1993) Genetics of bacteriocins produced by lactic acid bacteria. FEMS Microbiology Reviews 12, 39-86.

Lin, M.Y. and Yen, C.L. (1999) Reactive oxygen species and lipid peroxidation product-scavenging ability of yogurt organisms. ,journal of Dairy Science 82, 1629-1634.

Marty-Teysset, C., de la Torre, F. and Garel, J.-R. (2000) Increased production of hydrogen peroxide by Lactobacillus delbrueckii subsp. bulgaricus upon aeration: involvement of an NADH oxidase in oxidative stress. Applied and Environmental Microbiology 66, 262-267.

Miteva, V., Ivanova, I., Budakov, I., Pantev, A., Stefanova, T., Danova, S., Moncheva, P., Mitev, V., Dousset, X. and Boyaval, P. (1998a) Detection and characterization of a novel antibacterial substance produced by a Lactobacillus delbrueckii strain 1043. journal of Applied Microbiology 85, 603-614.

Miteva, V., Stefanova, T., Budakov, I., Ivanova, I., Mitev, V., Gancheva, A. and Ljubenov, M. (1998b) Characterization of bacteriocins produced by strains from traditional Bulgarian dairy products. Systematic and Applied Microbiology 21, 151-161.

Moll, G.N., Konings, W.N. and Driessen, A.J. (1999) Bacteriocins: mechanism of membrane insertion and pore formation. Antonie Van Leeumenhoek 76, 185-198.

Murphy, M.G. and Condon, S. (1984) Correlation of oxygen utilization and hydrogen peroxide accumulation with oxygen induced enzymes in Lactobacillus plantarum cultures. Archives of Microbiology 138, 44-48.

Nes, I.F., Diep, D.B., Havarstein, L.S., Brurberg, M.B., Eijsink, V. and Helgo, H. (1996) Biosynthesis of bacteriocins in lactic acid bacteria. Antonie Van Leeumenhoek 70, 113-128.

Vaughan, E.E., Daly, C. and Fitzgerald, G.F. (1992) Identification and characterization of helveticin V-1829, a bacteriocin produced by Lactobacillus helveticus 1829. ,journal of Applied Bacteriology 73, 299-308.

(order Full Text from publisher)

© 2005 Transgalactic Ltd (manufacturer of Bioscreen C software) | Privacy Statement | P.O. Box 1393, 00101 Helsinki, Finland, phone: +358 9 85172920, fax: +358 9 8749481, e-mail: microbiology@bionewsonline.com
 

Last modified: May 25, 2005