Susceptibility of Pediococcus isolates to antimicrobial compounds in relation to hop-resistance and beer-spoilage

Isolates from the genus Pediococcus are particularly problematic for the brewing industry where hop-compounds are used to provide flavour to beer. Hop-compounds are antimicrobial in that they dissipate the trans-membrane pH gradient of microbes, thereby inhibiting growth and potential spoilage of product [1]. As pediococci are also used as beneficial microbes in the context of food microbiology and animal husbandry (e.g., wine, cheese, and yogurt industries as well as for the production of silage), the emergence of hop-resistant Pediococcus isolates in the brewing industry is of broader interest. These isolates frequently harbour one or more ATP-binding cassette type multidrug resistance (ABC MDR) genes, suggesting that resistance to hop-compounds may also confer resistance to other antimicrobial compounds [2]. We have previously shown that several genes can be correlated with ability of Pediococcus isolates to grow in beer and to resist the antimicrobial activity of hop-compounds [3–5]. These are the ABC MDR genes ABC2, bsrA, bsrB, [6] and horA [2], a putative divalent cation transporter known as hitA [7], and horC which codes for a protein possessing little homology to any known protein [8, 9].

Because, many pediococci possess special growth requirements, conventional antimicrobial-sensitivity testing media have been demonstrated to be unsuitable for testing of Pediococcus isolates for antimicrobial resistance [10–12]. However, enriched media that permits growth of pediococci may inhibit the antimicrobial activity of some compounds under investigation. Previously, antimicrobial susceptibility testing of Pediococcus isolates has been attempted by several methods, many of which are performed using some variety of agar diffusion [10, 11, 13, 14]. However, more recently, dilution methods have been preferred over diffusion tests as the former allow for determination of minimum inhibitory concentrations (MICs), which is a more reliable and reproducible indicator of resistance [10, 11]. For these reasons, lactic acid bacteria susceptibility test broth medium (LSM), which was recently developed by Klare et al. [11], should be considered the new testing standard for assessing the antimicrobial resistance spectra of lactic acid bacteria. Despite this medium being shown to be very effective for establishing antimicrobial susceptibilities of two species of Pediococcus, namely, P. acidilactici, and P. pentosaceus [10], it previously has not been used to study the prevalence, and spectrum, of antimicrobial resistance among other members of the genus.

Overall, the use of antimicrobial compounds by industries such as animal husbandry, brewing, and fuel ethanol to combat Pediococcus contaminants (e.g., hop-compounds, Penicillin, and Virginiamycin which is structurally similar to Synercid) is long-standing. However, knowledge about the resistance of pediococci to antimicrobial agents is minimal [12]. As such, the focus of this research was to determine whether the use of antimicrobial hop-compounds in the brewing industry is associated with an increase in the overall antimicrobial resistance of Pediococcus isolates. Here we report on the testing of isolates from six species of the genus Pediococcus against 17 antimicrobial compounds using LSM broth in commercially available Sensititre GPN3F Gram-positive MIC plates (TREK Diagnostic Systems, Cleveland OH).

The finding of Pediococcus isolates that showed only moderate resistance to Vancomycin is discordant with other studies to date which have consistently reported the genus Pediococcus to be intrinsically Vancomycin-resistant [10, 12–14]. The isolates that were not resistant to all concentrations of Vancomycin tested were from the species P. acidilactici (N = 1), P. claussenii (Ropy, N = 1; Non-ropy, N = 3), P. damnosus (N = 1), and P. parvulus (Non-ropy, N = 2), suggesting that the phenomenon is not the product of a clonal event. It has previously been shown that intrinsic Vancomycin resistance in P. pentosaceus is due to a modified peptidoglycan precursor ending in D-Ala-D-lactate [15]. While this may also be the mechanism used by other Vancomycin-resistant pediococci, it is likely that the eight susceptible isolates do not possess this mechanism. Because media previously used for Pediococcus antimicrobial susceptibility testing have since been shown to be inappropriate for such testing (11), it is possible that the earlier finding of intrinsic Pediococcus Vancomycin-resistance was an artifact of the testing medium used, rather than reflective of pediococci genetic content.

The ropy phenotype did not associate with resistance to any of the antimicrobial compounds tested. This was an unexpected result as the ropy phenotype acts to create a biofilm which is expected to act as a physical barrier for the bacteria, putatively protecting them from the antimicrobial compounds. Why no associations were found is unclear. It may be that the type of exopolysaccharide matrix produced by these isolates did not result in a sufficiently dense matrix so as to inhibit the passage of antimicrobial compounds. Alternatively, the amount of energy expended on the production of exopolysaccharide may have caused a decreased ability to grow in the presence of the antimicrobial compounds, despite the partial antimicrobial barrier created by the exopolysaccharide.

Of particular interest to the brewing industry is the presence in pediococci of hop-resistance or beer-spoilage correlated genes (ABC2, bsrA, bsrB, hitA, horA, and horC). Of these six genes, only horA has been conclusively shown to function as a multidrug transporter, however, the ABC2, bsrA, and bsrB genes are highly similar to known ABC MDR genes, and the hitA gene is similar to divalent cation transporters. As such, all six of these beer-spoilage or hop-resistance correlated genes were assessed for associations with antimicrobial resistance. The genes hitA, horC, and ABC2 did not occur with sufficient frequency to determine statistical correlation [Additional file 2]. It is important to note that, as was found for ability to grow in beer, the bsrA, bsrB, and horA genes did not demonstrate significant associations with resistance to any of the antibiotics tested, but rather with susceptibility.

When MIC was compared to ability of isolates to grow in beer, eight of the 17 antibiotics that we tested surprisingly demonstrated a significantly lower MIC in isolates that could grow in beer. The eight antibiotics included Synercid, Ampicillin, Levofloxacin, Penicillin, Ciprofloxacin, Sulfamethoxazole/Trimethoprim, Gatifloxacin, and Oxacillin + 2% NaCl. This suggests that, despite repeated exposure to antimicrobial hop-compounds in the brewery setting, Pediococcus isolates capable of growing in the beer tend to be more susceptible to commonly used antimicrobial compounds than are isolates which cannot grow in beer. It is possible that this association may actually be independent of the presence of hop-compounds, instead being an indication of the environment encountered within the brewery environment by the beer-spoilage isolates. Although beer-spoilage bacteria must originate from outside the brewery, isolates capable of growing in beer have presumably become highly acclimatized or especially adapted to grow in the beer environment. Ideally, beer will not contain any wild yeasts or bacteria and, as such, contaminating pediococci are growing in an environment that does not contain a plethora of antimicrobial compounds naturally created by other organisms living in the same environment. Based on this scenario, Pediococcus isolates entering the brewery environment from outside sources (e.g., plant materials such as hop cones or barley) would possess mechanisms of resistance against multiple antimicrobial compounds. However, upon entering the brewery environment which should be free of other competing microbes, the pediococci would encounter no selective pressures other than hop-compounds and thus fail to maintain the genetic mechanisms for antimicrobial resistance.

It is curious to note that the bsrA and bsrB genes, hop-resistance, and beer-spoilage are all significantly negatively-associated with resistance to Ciprofloxacin. Moreover, although horA is strongly correlated to ability to grow in beer, this gene does not show any association (negative or otherwise) with Ciprofloxacin resistance. While the underlying mechanism for this association with lowered resistance to Ciprofloxacin is unknown, it strongly suggests that hop-resistance, and in turn beer-spoilage, is linked to the presence of the bsrA and bsrB genes, while the horA gene may simply be correlated by chance to ability of Pediococcus isolates to spoil beer. That is to say, because the bsrA and bsrB genes (like the beer-spoilage phenotype) are negatively correlated to ciprofloxacin resistance, while the horA gene is not, the bsrA and bsrB genes are likely more closely associated with beer-spoilage than is the horA gene.

Testing the susceptibility of Pediococcus isolates to antimicrobial compounds was effective using LSM in GPN3F antimicrobial susceptibility testing plates. In contrast with previous studies, we found Pediococcus isolates that are not intrinsically resistant to Vancomycin. A significant negative association was identified between resistance to Ciprofloxacin and the presence of the bsrA and bsrB genes as well as the hop-resistant and beer-spoilage phenotypes. Significantly lower MICs to antimicrobial compounds were found in isolates that were hop-resistant and/or capable of growing in beer. Similarly, the presence of genes previously correlated with beer-spoilage (i.e., bsrA, bsrB, and horA) was also found to be associated with significantly lower MICs to several of the antimicrobial compounds tested. These results suggest that the ongoing use of the antimicrobial hop-compounds in the brewing industry and the phenomenon of hop-resistance mediated by ATP-binding cassette type multi-drug transporters is not associated with the emergence of greater antimicrobial resistance in beer-spoilage pediococci.