Comparative genomics enables a genetic barcode to discriminate and score beer-spoiling and non-spoiling Lactobacillus brevis

Technical Session 18: Microbiology II Session
Rudi F Vogel, Technische Universität München, Freising, Germany
Co-author(s): Patrick Preissler, Angel Angelov, and Wolfgang Liebl, Technische Universität München, Freising, Germany

ABSTRACT: Beer is an uncomfortable environment for many bacteria. Nevertheless, specific bacteria, mainly lactic acid bacteria, are able to grow in beer and spoil it. In this group of bacteria, Lactobacillus brevis is the most common beer spoiler found in breweries. Within the large biodiversity of this genus, different ecotypes exist, some of which exhibit stress responses enabling survival under the antimicrobial conditions in beer. Thereby, the tolerance to hop compounds, which are mainly responsible for inhibition of growth in beer, is a multifactorial process. For this reason, any approaches to predicting the physiological differences between beer-spoiling and non-spoiling strains on the basis of a single marker gene are limited. In addition, most known genetic determinants that are potentially useful for PCR detection of beer-spoilage bacteria are widely spread in strains with no reference to high hop tolerance. Comprehensive and strain specific information about the ecotype beer spoiler compared with non-spoiler strains reside in their genomes. In this study we aimed to identify genes related to the ability to grow in stronger hopped beers (e.g., pilsner beer) via comparative genomics of four different strains of L. brevis. The genomes of two beer isolates (L. brevis TMW 1.313 and 1.465) and one strain isolated from feces (L. brevis TMW 1.6T) were determined by next generation pyrosequencing. A fourth sequence of a published genome (L. brevis ATCC 376, silage isolate) was included in the genome comparison. Redundant information, which resides in the core genome of all L. brevis ecotypes or strain-specific sequences were removed, and gene fragments exclusively occurring in beer-spoiling strains were identified, as well as ecotype-specific DNA sequences of non-spoiling strains. Subsequently, targeted arrays derived from these sequences were established and hybridized with DNA from a bigger set of different L. brevis strains to identify discriminative marker sequences for the ecotype “beer-spoiler” or “non-spoiler.” As a result, 34 oligonucleotides could be identified that are able to differentiate the ecotype “beer-spoiler” and are useful for predicting beer-spoiling potential. Furthermore, four oligonucleotides specific for the ecotype “non-spoiler” were found that can be used as negative markers for beer-spoiling strains. The cumulative detection of more than one of these marker sequences to a score enables the establishment of a genetic barcode that can be used by brewers to predict the beer-spoiling potential of L. brevis isolates. For practical applications, a multiplex PCR targeted toward a further reduced set of selected marker sequences proved effective.

Rudi F. Vogel is a biochemist interested in food microbiology and biotechnology. As head of Technische Mikrobiologie at the Technische Universität München he conducts research on starter culture development, high pressure in food, and biosciences, as well as control of unwanted microbes in food. A clear focus is on lactic acid bacteria, their metabolism and genetics, pre- and probiotic functionality, and mechanisms of stress response and adaptation. In this context beer-spoiling lactobacilli are used as models to understand molecular mechanisms of hop resistance.