Antimicrobial susceptibility of U.S. porcine Brachyspira isolates and genetic diversity of B. hyodysenteriae by multilocus sequence typing

Abstract

Swine dysentery, caused by Brachyspira hyodysenteriae and the newly recognized Brachyspira hampsonii in grower-finisher pigs, is a substantial economic burden in many swine-rearing countries. Antimicrobial therapy is the only commercially available measure to control and prevent Brachyspira-related colitis. However, data on antimicrobial susceptibility trends and genetic diversity of Brachyspira species from North America is limited. We evaluated the antimicrobial susceptibility profiles of U.S. Brachyspira isolates recovered between 2013 and 2022 to tiamulin, tylvalosin, lincomycin, doxycycline, bacitracin, and tylosin. In addition, we performed multilocus sequence typing (MLST) on 64 B. hyodysenteriae isolates. Overall, no distinct alterations in the susceptibility patterns over time were observed among Brachyspira species. However, resistance to the commonly used antimicrobials was seen sporadically with a higher resistance frequency to tylosin compared to other tested drugs. B. hampsonii was more susceptible to the tested drugs than B. hyodysenteriae and B. pilosicoli. MLST revealed 16 different sequence types (STs) among the 64 B. hyodysenteriae isolates tested, of which 5 STs were previously known, whereas 11 were novel. Most isolates belonged to the known STs: ST93 (n = 32) and ST107 (n = 13). Our findings indicate an overall low prevalence of resistance to clinically important antimicrobials other than tylosin and bacitracin, and high genetic diversity among the clinical Brachyspira isolates from pigs in the United States during the past decade. Further molecular, epidemiologic, and surveillance studies are needed to better understand the infection dynamics of Brachyspira on swine farms and to help develop effective control measures.

Brachyspira spp. are gram-negative, anaerobic spirochetes that cause diarrhea in pigs and other animals. ⁸ The genus contains 9 officially recognized species, most of which are associated with diarrheal conditions (https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=29521).^1,10 The economically important pathogenic species in pigs include B. hyodysenteriae, B. suanatina, B. hampsonii, and B. pilosicoli.^8,10 B. hyodysenteriae is the classical etiologic agent of swine dysentery (SD), a disease associated with severe mucohemorrhagic diarrhea commonly affecting the cecum and colon of grower-finisher and adult pigs. ⁸ However, the newly recognized species B. suanatina, which is mainly restricted to Scandinavia, causes a disease consistent with SD, and B. hampsonii, which is mainly reported in Europe and North America, can cause disease that is indistinguishable from classical SD. ⁹ As well, B. pilosicoli is associated with intestinal spirochetosis and milder colitis in pigs and other host species. ¹¹ Diseases caused by the pathogenic Brachyspira species in pigs are common globally, contributing to a substantial economic burden worldwide. ⁹

Because there are no commercial vaccines for prevention of SD in pigs, both control and treatment of Brachyspira-associated diseases in pigs primarily depend on antimicrobial agents. ¹⁰ The commonly used antimicrobials for SD in the United States and other swine-rearing countries include lincosamides (lincomycin), macrolides (tylosin, tylvalosin), and pleuromutilins (tiamulin, valnemulin). Other antimicrobials labeled for treatment and control of SD include bacitracin, carbadox, doxycycline, gentamicin, and virginiamycin.^8,28 Carbadox has been banned from swine production in the European Union and Canada because of its potential carcinogenic residues in pork products. ⁹

Two methods are used for determining the susceptibility of Brachyspira species to the commonly used drugs: the agar dilution method, ¹⁸ and a more widely used broth microdilution method. ²¹ In the United States, the Clinical and Laboratory Standards Institute (CLSI) provides information on the antimicrobial susceptibility testing procedures and guidelines of B. hyodysenteriae. ⁴ However, no universally accepted, standardized methods exist for antimicrobial susceptibility testing of this highly fastidious spirochete. Although a ring trial was performed in 2020 to establish standardized methods, the investigation only included B. hyodysenteriae and B. pilosicoli ³³ with further research required to validate this method for other economically important Brachyspira spp., such as B. hampsonii and B. suanatina. CLSI ⁴ also provides interpretive breakpoints for the antimicrobial classes used to treat diseases associated with Brachyspira spp., derived from various published literature on the antimicrobial susceptibility of Brachyspira spp. However, CLSI ⁴ states that the breakpoints are derived from published studies on the antimicrobial susceptibility of Brachyspira spp., and for which no quality control testing was performed, and the information provided is not intended for making diagnostic decisions.

SD essentially disappeared from the United States during the 1990s, at least in part as a result of changes in management practices and biosecurity, but outbreaks of SD were reported starting in 2007. ²⁶ Since its reemergence, there has been an increase in the isolation of Brachyspira spp. overall by veterinary diagnostic laboratories in the United States. ⁵ Reduced susceptibility or in vitro resistance to commonly used antimicrobials has been reported among B. hyodysenteriae and B. pilosicoli isolates in many European countries,^12,17 Australia, ²¹ and the United States. ²⁵ Hence, reports of resistance create a substantial threat to the effective control of SD and colitis in pigs and further highlight a need for temporal monitoring of antimicrobial susceptibility of B. hyodysenteriae and other Brachyspira spp.

Combination of antimicrobial susceptibility data with information on the genotype of resistant strains and their origin can provide insight into the factors contributing to the development of resistance. ³¹ In addition, analyzing the genetic diversity and relatedness of different sequence types can enable us to track the source of SD agents, allowing targeted control measures to prevent disease outbreaks. ⁹ However, data are limited on antimicrobial susceptibility trends and the molecular epidemiology of Brachyspira species from the United States.^9,26 Although the microevolution of B. hyodysenteriae in the United States has been studied since its reemergence, ²⁶ data comparing the genotypic information on B. hyodysenteriae determined by sequence types (STs) originating from diverse sources are scarce, with only 2 studies to date.^25,26

We evaluated the antimicrobial susceptibility of isolates of 3 Brachyspira spp. (B. hyodysenteriae, B. hampsonii, B. pilosicoli) collected over time and originating from different United States regions and swine production systems. Given that B. hyodysenteriae is the most economically important agent of SD, ¹⁰ we used a B. hyodysenteriae species–specific multilocus sequence typing (MLST) scheme ²⁴ to characterize and evaluate the genetic diversity of B. hyodysenteriae isolates.

Materials and methods

Bacterial isolates and culture methods

We obtained 207 Brachyspira isolates recovered from swine diagnostic submissions during 2013–2022 from the Iowa State University–Veterinary Diagnostic Laboratory (ISU-VDL) culture collection. These included 107 B. hyodysenteriae, 60 B. hampsonii, and 40 B. pilosicoli isolates. The isolates originated from various swine farms across 16 states (Colorado, Hawaii, Iowa, Illinois, Indiana, Kansas, Maryland, Missouri, Montana, North Carolina, Nebraska, New York, Oklahoma, Pennsylvania, South Dakota, Utah) within the 4 major regions of the United States (Northeast, Midwest, South, West), and from 94 different production systems. An alphabetic code was assigned for each production system because of client confidentiality reasons. The isolates were recovered from −80°C frozen stocks on trypticase soy agar (TSA; BD Difco) supplemented with 5% bovine blood following anaerobic incubation at 42°C for 48 h. All isolates were confirmed for identification to species level by MALDI-TOF MS (Brucker) following the manufacturer instructions and ISU-VDL standard operating procedures.

Antimicrobial agents and custom Sensititre panel

A 96-well minimum inhibitory concentration (MIC) dry custom panel (Sensititre; Thermo Fisher) was designed such that it contained 6 antimicrobial agents (bacitracin, doxycycline, lincomycin, tiamulin, tylosin, tylvalosin) in duplicate. The antimicrobials evaluated in the Sensititre panel were selected because they are commonly used to treat or control diseases caused by Brachyspira spp. in pigs in the United States. The concentrations of antimicrobials tested were as follows: bacitracin (2–256 μg/mL), doxycycline (0.12–16 μg/mL), lincomycin (0.5–64 μg/mL), tiamulin (0.6–8 μg/mL), tylosin (2–128 μg/mL), and tylvalosin (0.25–32 μg/mL). Quality control was performed on the Sensititre panels using the type strain B. hyodysenteriae ATCC 27164^T as the control organism following the manufacturer guidelines (Thermo Fisher) and the CLSI guidelines ⁴ before the study.

In vitro antimicrobial susceptibility testing

Antimicrobial susceptibility testing of Brachyspira spp. was performed using the broth microdilution method with the custom Sensititre panel. Briefly, 2 well-filled 1-μL plastic loops (~2 μL) of Brachyspira culture were harvested from 3-d-old fastidious anaerobic agar (FAA; Neogen) plates supplemented with 5% defibrinated horse blood (Remel, Thermo Fisher) and suspended in 2 mL of brain heart infusion (BHI) broth (Oxoid) supplemented with 10% fetal calf serum (FCS; Hyclone) to obtain a suspension of ~1 × 10⁷ cfu/mL. The suspension was further diluted at 1:10 in BHI + 10% FCS to obtain a final inoculum concentration of ~1 × 10⁶ cfu/mL, as recommended by CLSI. ⁴ To determine the actual inoculum density, 1 μL from the final inoculum was transferred into 10 mL of BHI + 10% FCS with subsequent plating of 100 μL onto TSA supplemented with 5% bovine blood. The plates were incubated in anaerobic boxes using anaerobic gas packs (GasPak EZ; BD) for 2 d at 42°C. Each well of the Sensititre plate was filled with 50 μL of the final inoculum and incubated anaerobically at 37°C with agitation (90 rpm) for 4 d. All isolates were tested in duplicate in the same Sensititre plate.

MICs were determined as the lowest concentration of drugs that inhibited visible growth (VIZION Sensititre; Thermo Fisher). With each batch of the isolates evaluated for susceptibility, the B. hyodysenteriae type strain ATCC 27164^T was used as the control organism.^27,33 For interpretation of the susceptibility data, proposed clinical breakpoints from published studies were used to categorize isolates as resistant or susceptible.^28–30 The proposed clinical breakpoints for resistance to the following drugs were as follows: doxycycline > 4 μg/mL, lincomycin > 16 μg/mL, tiamulin > 2 μg/mL, tylosin > 16 μg/mL, and tylvalosin > 16 μg/mL. Given that there were no proposed clinical breakpoints for bacitracin, the results for this antibiotic were presented simply as MIC values. The results were considered valid based on the following criteria: purity check plates were deemed pure, growth was observed in positive control wells, and the estimated cfus of the inoculum were within the recommended range of ~1 × 10⁶ cfu/mL. ⁴

Multilocus sequence typing

For MLST, 64 B. hyodysenteriae representative isolates, including the well-characterized reference strain B204 (ATCC 31212),^2,37 were chosen after confirmation of identification with MALDI-TOF MS. Pure cultures were suspended in 700 μL of sterile PBS and frozen at −20°C before DNA extraction. Genomic DNA was extracted (MagMAX pathogen RNA/DNA kit; Applied Biosystems). MLST was performed on B. hyodysenteriae isolates using the published scheme for B. hyodysenteriae. ²⁴ MLST analysis was performed on 7 reference genes coding for alcohol dehydrogenase (adh), alkaline phosphatase (alp), esterase (est), glutamate dehydrogenase (gdh), glucose kinase (glpK), phosphoglucomutase (pgm), and acetyl-coA acetyltransferase (thi). Conventional PCR assays were performed to amplify the 7 reference genes of the selected isolates (Multiplex PCR kit; Qiagen) with added DNA polymerase (HotStarTaq; Qiagen). The PCR products were purified (ExoSAP-IT; Thermo Fisher) per manufacturer instructions. The purified products were sequenced at the ISU DNA facility (Ames, IA, USA). Sanger sequencing results for both forward and reverse sequences were trimmed and assembled using Geneious Prime v.2022.2.2 (Dotmatics). Once the consensus sequence of each gene was obtained, the allelic number and ST were identified by querying the PubMLST database ¹⁵ for Brachyspira (https://pubmlst.org/organisms/brachyspira-spp). Any novel alleles or STs were submitted to the PubMLST database to be curated for allelic number and ST assignment. ¹⁵ The novel allele and STs were then deposited to the PubMLST database for Brachyspira.

Data analysis

The proposed clinical breakpoints were used to interpret the susceptibility results of Brachyspira spp.^28–30 The MIC data obtained at each Brachyspira species level were divided into two 5-y periods to evaluate the antimicrobial susceptibility trends over time. The first group consisted of isolates recovered in 2013–2017; the second group consisted of isolates recovered in 2018–2022.

Logistic regression and chi-squared tests were performed for each Brachyspira species to investigate whether there was a statistically significant difference in resistance or susceptibility to one or more antibiotics among different geographic regions or production systems. For the statistical analysis, production systems containing < 3 isolates were excluded. Therefore, 113 Brachyspira isolates originating from 13 different production systems within the 3 major regions (Midwest, South, West) were included. The same statistical model and test were used to determine whether a correlation existed between certain STs of B. hyodysenteriae and resistance or susceptibility to a specific antibiotic. STs comprising < 3 isolates were excluded from the analysis. Thus, 45 B. hyodysenteriae isolates belonging to 2 STs (ST93, ST107) were included in the analysis. R v.4.2.1 (https://www.r-project.org/) was used for developing the statistical model. P-values ≤ 0.05 were considered statistically significant.

Results

Antimicrobial susceptibility of Brachyspira spp.

At the genus level, the MIC ranges for each antibiotic were as follows: bacitracin, 32 to > 256 μg/mL; doxycycline, ≤ 0.12 to 16 μg/mL; lincomycin, ≤ 0.5 to 64 μg/mL; tiamulin, ≤ 0.06 to > 8 μg/mL; tylosin, ≤ 2 to > 128 μg/mL; and tylvalosin, ≤ 0.25 to 32 μg/mL (Table 1). Overall, based on the proposed breakpoints, a larger percentage of all 3 Brachyspira spp. were susceptible to tylvalosin and doxycycline than the other agents. Among the 3 species investigated, only B. pilosicoli had resistance to tylvalosin (2.5%), whereas 8.4% of B. hyodysenteriae isolates and 1.7% of B. hampsonii isolates had doxycycline resistance. Overall, B. hampsonii had a higher percent of susceptible isolates to most of the tested drugs than the other 2 species. All 3 Brachyspira spp. had a low level of susceptibility to tylosin and relatively moderate susceptibility to lincomycin. Nonetheless, B. hyodysenteriae had the highest percent resistance to tylosin and lincomycin compared with the other Brachyspira spp. All Brachyspira isolates had a high MIC to bacitracin (Table 1).

Table 1.

Susceptibility profiles of Brachyspira spp. to 6 antimicrobial agents.

BAC = bacitracin; DOX = doxycycline; LIN = lincomycin; NA = not applicable; %R = percent resistance; TIA = tiamulin; TVN = tylvalosin; TYL = tylosin. Unshaded cells indicate the number of isolates at each MIC for the drugs tested. Shaded cells indicate concentrations that were not tested for the respective drugs. The proposed clinical breakpoints^28–30 are indicated by the ¶ symbol.

^* Should be read as ≤ to the corresponding concentration.

^† Should be read as > than the corresponding concentration.

For most of the antimicrobial agents, no noticeable changes in antimicrobial susceptibility were observed over time (Fig. 1; Suppl. Figs. 1, 2). Nonetheless, there was an increase in tiamulin MICs for both B. hyodysenteriae isolates and B. pilosicoli isolates from the first to the second time period (Fig 1; Suppl. Fig 2). In addition, an increase in lincomycin and tylvalosin MICs was observed among B. pilosicoli isolates during the second time period (Suppl. Fig. 2). No statistically significant differences in the percent of resistance or susceptibility to one or more antibiotics were detected among the different regions of the United States or different production systems.

Figure 1.

Distribution of MIC values for Brachyspira hyodysenteriae isolates to 6 antimicrobial agents (bacitracin, doxycycline, lincomycin, tiamulin, tylosin, and tylvalosin). A. 2013–2017. B. 2018–2022. MIC values are color coded by the isolation year, and the number of isolates tested per year is shown in parentheses in the key for each panel.

Multilocus sequence typing

All 7 reference genes were amplified and sequenced successfully from the 64 selected B. hyodysenteriae isolates. Sequence typing of the B. hyodysenteriae isolates revealed 16 different STs based on analysis of the 7 loci (Fig. 2). The phylogenetic tree constructed using the concatenated sequences of 7 MLST loci shows the genetic relatedness of the 16 STs (Fig. 2). Five of these were previously known STs (STs 54, 93, 106, 107, 109); 11 were novel STs (ST 298, 300–309). The reference strain B204 was allocated to ST54. Most of the isolates belonged to a known sequence type, ST93 (n = 32), followed by ST107 (n = 13); 2 isolates were allocated to each of ST298, ST300, ST303, ST305; detected only once were ST307, ST54, ST106, ST109, ST301, ST302, ST304, ST306, ST308, and ST309. Five of the 11 newly described STs each contained one newly identified allele; the remainder of the new STs were the result of novel allelic profiles.

Figure 2.

Phylogenetic tree based on the concatenated sequences of 7 MLST genes constructed by the neighbor-joining method using MEGA11 shows the genetic relatedness of 16 sequence types (STs) detected among the 64 representative Brachyspira hyodysenteriae isolates derived from swine diagnostic submissions to ISU-VDL between 2013 and 2022. Red triangles denote novel STs. Bar = 0.0010 shows the base substitutions per site.

Statistically significant results were observed between certain STs and susceptibility or resistance phenotype to tiamulin. ST107 had a higher likelihood of being resistant to tiamulin than ST93 (p < 0.001). More specifically, the estimated probability of ST93 being resistant to tiamulin is 0, whereas the estimated probability of ST107 being resistant to tiamulin is 0.54.

The B. hyodysenteriae ST93 isolates originated from the South (n = 18), Midwest (n = 13), and Northeast (n = 1) within 9 different production systems (Suppl. Table 1). In comparison, ST107 originated from the Midwest (n = 11) and South (n = 2) within 4 different production systems.

There were only a few instances in which both STs were found in the same production systems (i.e., G and I). Furthermore, there were instances where the same ST (ST93 or ST107) was found within 1 production system (G and I) but in 2 different regions (Suppl. Table 1).

Discussion

When comparing the antimicrobial susceptibility profiles of U.S B. hyodysenteriae isolates to those reported internationally, our findings suggest that, overall, U.S B. hyodysenteriae isolates are more susceptible to antimicrobials. Mainly, the U.S. B. hyodysenteriae isolates are more susceptible to tiamulin than the isolates from the United Kingdom, ¹⁷ Italy, ³¹ Poland, ³⁸ Germany,^13,17 the Czech Republic, ³² Brazil, ⁶ Japan, ¹⁶ and Spain. ¹² However, it is worth noting that, in Europe, there are limited antimicrobials available for the treatment of SD. ¹⁰ Therefore, high resistance to tiamulin may be attributed to the selective pressure applied by tiamulin use, which may lead to higher levels of resistance in European B. hyodysenteriae isolates. In addition, although SD remained prevalent in other swine-rearing countries, it had mainly disappeared from the United States during the 1990s and then reemerged during the late 2000s. ¹ The reduced occurrence of SD for over a decade may also explain the high susceptibility of U.S. B. hyodysenteriae isolates to tiamulin compared to the isolates from Europe. However, 2 published studies investigating the antimicrobial susceptibility of Swiss B. hyodysenteriae isolates reported high susceptibility to pleuromutilins. Swiss B. hyodysenteriae exhibited an MIC range of < 0.063–0.125 μg/mL and < 0.063–0.25 μg/mL for tiamulin and valnemulin, respectively.^7,22 In our study, we found a tiamulin MIC range of U.S. B. hyodysenteriae isolates of ≤ 0.06 to > 8 μg/mL. Interestingly, tiamulin is 1 of the 3 antimicrobials approved for treating SD in pigs in Switzerland. The findings from Swiss studies indicate that strains with increased resistance to pleuromutilins have not emerged in Switzerland,^7,22 which could be attributed to the prudent use of antimicrobials in Switzerland.

We did not observe a temporal change in susceptibility to most antimicrobials tested among Brachyspira species in our study. However, we observed an increase in tiamulin MICs among both B. hyodysenteriae and B. pilosicoli isolates during the second period (2018–2022). Interestingly, the tva(A) gene associated with tiamulin resistance was first identified in 2018 from resistant strains of B. hyodysenteriae through genome-wide studies following whole-genome sequencing (WGS). ³ Approximately 10% of B. hyodysenteriae isolates from our study had a MIC of 8 μg/mL to tiamulin in the second period.

Tylvalosin is a derivative of tylosin, and cross-resistance between the 2 closely related macrolide antimicrobials has been identified. ²⁰ However, our findings did not support cross-resistance between the 2 drugs. Most Brachyspira isolates in our study were susceptible to tylvalosin, whereas resistance to tylosin, particularly in B. hyodysenteriae isolates (60% resistance), was observed. Similarly, we detected resistance to lincomycin among Brachyspira isolates, especially a higher percentage resistance (15%) among B. hyodysenteriae isolates. Further investigations using tools such as WGS are warranted to determine if previously described mutations^19,20 or antimicrobial resistance (AMR) genes ³ are responsible for the macrolide, lincosamide, and pleuromutilin resistance that we observed.

We observed high percent susceptibility to doxycycline among the Brachyspira spp. tested. Our findings concur with previous data from the United States, in which was observed relatively high susceptibility among U.S. B. hyodysenteriae, B. hampsonii, and B. pilosicoli isolates to doxycycline. ²⁵ Nevertheless, B. hyodysenteriae isolates from Poland, ³⁸ Germany, ¹⁷ and the United Kingdom ¹⁷ appear less susceptible to doxycycline.

All Brachyspira spp. from our study had a high MIC to bacitracin (32 to > 256 μg/mL). Unfortunately, there are no recent reports regarding the effectiveness of bacitracin in treating SD or other Brachyspira-related diseases in pigs. There are historical reports, using the agar dilution method, on the antimicrobial susceptibility of B. hyodysenteriae to bacitracin with a wide MIC range (12.5–50 μg/mL and 838 to ≥ 1,675 μg/mL).^23,35 Given that bacitracin is labeled for SD associated with B. hyodysenteriae in pigs, ⁸ it may still be effective against Brachyspira infections by altering the colonic microbiota that generally forms a synergistic relationship with Brachyspira species^10,36 even if it has limited in vitro effect against B. hyodysenteriae itself. In a challenge study, chickens were inoculated with B. pilosicoli strain CPsp1, which is associated with diarrhea, reduced egg production, and fecal staining. ¹⁴ All 40 challenged birds were given a commercial diet during the study, and 10 of those were given 50 μg/g zinc bacitracin. Interestingly, in birds receiving the diet with zinc bacitracin, 7 of the 10 were colonized with B. pilosicoli, whereas in birds receiving a diet without zinc bacitracin, only 3 of the 30 were colonized with B. pilosicoli. ¹⁴ Furthermore, despite being colonized with B. pilosicoli, no significant changes in production were observed among the challenged birds; the study further highlights the need for understanding the relationship between Brachyspira and the gut microbiome, and how zinc bacitracin may alter the gut microbiota and directly or indirectly contribute to colonization with Brachyspira spp. ¹⁴

Overall, B. hampsonii had a higher susceptibility to antimicrobials than B. hyodysenteriae and B. pilosicoli. Given that B. hampsonii is a newly recognized species, the higher antimicrobial susceptibility of this species has been suggested as possibly being the result of a shorter exposure to antimicrobials used in swine production. ²⁵ In general, our findings concur with previous U.S. data on the antimicrobial susceptibility trend of U.S. Brachyspira species. ²⁵ However, the percent resistance among B. pilosicoli isolates was higher than B. hyodysenteriae in a previous study. ²⁵ At the same time, we observed that the percent resistance in B. hyodysenteriae was higher than in B. pilosicoli. Slight differences in MICs between our study and the published data ²⁵ may be the result of factors such as production systems or regions where the isolates originated.

Our identification of 16 different STs indicates high genetic diversity among these U.S. B. hyodysenteriae isolates. Similarly, a study using MLST on 59 B. hyodysenteriae isolates post-2010, originating from different farms in the United States, identified 13 STs, indicating high diversity among U.S B. hyodysenteriae isolates. ²⁶ Additionally, in that same study, ST93 was identified from multiple sites and production systems, ²⁶ which concurs with our findings that ST93 was widely distributed among different production systems. Overall, the high genetic diversity among B. hyodysenteriae isolates tends to be common; a study from the United Kingdom reported 32 STs from 115 isolates originating from different regions. ³⁴

To date, all of the known and novel STs that we reported are unique to the United States. ²⁶ However, the PubMLST database for B. hyodysenteriae and Brachyspira spp., in general, is not as extensive as other bacterial groups. Currently, only ~300 STs are available for B. hyodysenteriae. With the limited data available, it is not possible to conclude if the 16 STs found in our study have been only detected in the United States. However, it is evident that the predominant U.S clone (ST93) has not yet been reported from any other swine-rearing countries, which may be because of the limited pig trade between the United States and other countries. ³⁴

It is apparent from our findings that ST93 is widely distributed throughout the swine-rearing regions given that it was found in 9 production systems. In contrast, ST107 was restricted to the Midwest, originating from only a few production systems. The movement of pigs between different sites within a production system may be responsible for the instances in which STs were found in the same production system but in 2 regions. Furthermore, detecting a ST from the same production system over time indicates that STs can be sustained in a pig population for long periods if not eliminated.

Given that we detected most of the STs no more than twice, it is hard to conclude if a particular ST is associated with resistance or susceptibility to a particular antimicrobial. However, we found statistically significant differences between ST93 and ST107, the predominant STs. We found that ST107 has a significantly higher chance of being resistant to tiamulin than ST93 (p = 0.00002). Interestingly, a MLST study from the United States also found that ST107 had significantly lower susceptibility to pleuromutilins. ²⁶

With the exception of the identified point mutations and AMR genes in B. hyodysenteriae, data are lacking on a clear understanding of the association between the strains of multidrug-resistant B. hyodysenteriae isolates and their genetic basis for resistance. Further studies are warranted to determine the genetic basis for multidrug resistance, particularly with WGS, an emerging tool that allows researchers to understand the potential mechanisms of resistance to multiple drugs through genome-wide association studies. ³ Additionally, MLST data can easily be extracted to determine STs and clonal complexes, contributing to a better understanding of epidemiologic relationships among economically important Brachyspira spp.