Resistance to the main antibiotics registered for the control of SD complicates the management of this disease on swine farms and makes antimicrobial susceptibility testing an essential tool for rational antimicrobial treatment [5]. For this purpose, a commercial plate (VetMIC-Brachy, SVA) which includes six antibiotics, tiamulin, valnemulin, tylosin, tylvalosin, doxycycline and lincomycin, is commonly used for resistance determination in Brachyspira species in different countries [5, 14]. Although gentamicin is registered for the treatment of SD in several countries such as Spain (2,000 international units (IU) per kg of body weight during 3 days in drinking water according to the Spanish technical sheet), its use for this purpose is still limited and the information regarding resistance in field isolates is scarce.
In our investigation, a collection of fifty-six Spanish field isolates of B. hyodysenteriae showed low MIC values for gentamicin, with a MIC50 of 4 µg/mL, a MIC90 of 16 µg/mL and a MIC mode of 2 µg/mL. These values were similar to those reported in 2011 for American (MIC50 = 4 µg/mL, MIC90 = 8 µg/mL, MIC mode = 4 µg/mL) [8] or 2018 for Taiwanese isolates of this bacteria (MIC50 = 2 µg/mL, MIC90 = 4 µg/mL, MIC mode = 2 µg/mL) [15] in previous studies using a similar approach. Whilst the activity of gentamicin and other aminoglycosides decreases significantly in the absence of oxygen due to a reduced uptake by bacteria [16, 17], the low MIC values obtained in our study and in previous ones for B. hyodysenteriae isolates suggest that even a low uptake could be enough to make it effective against this bacterial species in the strict anaerobic environment of the large intestine. Assessing the evolution of these values over time is essential for the surveillance of resistance development. In this sense, neither the temporal distribution of MICs nor the survival analysis showed any trend towards the development of resistance to gentamicin in field isolates of B. hyodysenteriae.
The interpretation of gentamicin MICs for Brachyspira species is complicated due to the scarce information on clinical breakpoints. These clinical breakpoints must consider the availability of the molecule in its active form in the lumen of the large intestine. There are studies relating the administered dosages in feed or water of some antibiotics used in the treatment of SD with the concentrations reached in the large intestine [18, 19] although there are no similar studies for gentamicin. Nevertheless, it is well known that absorption and degradation of gentamicin is minimal in the small intestine [10, 20] and hence, a high concentration is expected in the lumen of the pig colon. Duhamel et al. proposed the following MIC breakpoints for the evaluation of gentamicin activity in Brachyspira spp. by the agar dilution method: susceptible for MIC ≤ 1 µg/mL, intermediate for MIC = 5 µg/mL and resistant for MIC ≥ 10 µg/mL [21]. CLSI guideline for susceptibility testing of infrequently isolated or fastidious bacteria isolated from animals proposed a breakpoint of ≤ 8 µg/mL to classify B. hyodysenteriae isolates as susceptible to gentamicin using the agar dilution method [22]. Taking into account that MICs obtained in broth microdilution assays are usually one dilution step lower than the corresponding values from agar dilution [23, 24], we have classified isolates as gentamicin susceptible if MIC ≤ 2 µg/mL, resistant if MIC ≥ 16 µg/mL and intermediate for MIC values of 4 and 8 µg/mL. Therefore, 44.6% (25/56) of the B. hyodysenteriae isolates were identified as susceptible to gentamicin, 33.9% (19/56) had reduced susceptibility and 21.4% (12/56) were resistant.
We further explored the genetic determinants of resistance to gentamicin by the study of two extreme populations. Two nucleotide variations were observed in the 16S rRNA gene for all sequenced isolates classified as resistant to gentamicin. These changes were not detected in any of the low MIC isolates. Within the genus Brachyspira, point mutations in genes encoding ribosomal RNA and ribosomal proteins have been associated with an increased MIC for drugs that act by inhibiting protein synthesis such as macrolides, lincosamides [6, 25], pleuromutilins [26–28] and tetracyclines [29, 30]. However, this is the first time to the authors knowledge that point mutations in the 16S rRNA gene are proposed as a plausible cause of resistance to gentamicin in Brachyspira spp. Genetic changes in the 16S rRNA sequence have been widely described to confer resistance to aminoglycosides including gentamicin in other bacterial species like Escherichia coli or Mycobacterium tuberculosis [31–33]. Typically, these changes affect the aminoacyl-tRNA recognition site (A-site), where aminoglycosides bind in a pocket formed by the A1408 · A1493 base pair and the bulged nucleotide A1492 leading to inhibition of protein synthesis [31, 34]. However, no differences were found at, or nearby, these positions in our B. hyodysenteriae isolates. In the present research, mutations were located at positions 990 (C◊T) and 1185 (A◊G), more than 400 and 200 nucleotides away from the A-site, respectively. The role of these mutations could therefore involve conformational changes that spatially impair the binding of gentamicin, as has also been suggested elsewhere for mutations in distal regions of the 16S rRNA that affect the decoding process in the A-site [35, 36]. Additionally, two of the gentamicin resistant isolates shared a mutation at 1058 (C◊G), a position in which nucleotide transversions have been associated with resistance to doxycycline in B. hyodysenteriae [26, 29, 37] and B. intermedia [30].
As diseases caused by Brachyspira rely heavily on the use of mechanistically analogous antimicrobials, mainly pleuromutilins, macrolides and lincosamides, which act by binding to the 50S ribosomal subunit, there is a high selection pressure for the development of resistance to these drugs [17]. Accordingly, a high proportion of resistance was reported against the main antimicrobials used for the control of SD among the tested B. hyodysenteriae isolates, particularly against the pleuromutilin family. These results are in concordance with a number of longitudinal studies from North America, Europe and Japan, reporting the rise of pleuromutilin resistance in this bacterial species as a major thread to the effective control of SD [6, 38–41]. It is worthy note that, in our research, resistance to pleuromutilins was associated to a high susceptibility to gentamicin. Although further studies are required to confirm this relationship, gentamicin could presumably be a suitable alternative for the treatment of SD in outbreaks caused by tiamulin or valnemulin resistant B. hyodysenteriae isolates.