Antimicrobial Resistance Profile, Whole-Genome Sequencing and Core Genome Multilocus Sequence Typing of B. anthracis Isolates in Croatia from 2001 to 2022

Abstract

Bacillus anthracis, the causative agent of anthrax disease, is a worldwide threat to livestock, wildlife and public health. It is also considered one of the most important pathogens of bioterrorism. Rapid and reliable diagnosis and administration of antimicrobials are essential for effective anthrax treatment. In this study, we determined the in vitro susceptibilities of 40 isolates of B. anthracis isolated in Croatia over the recent two decades to 18 antimicrobials. Whole-genome sequencing was performed, and bioinformatics tools were used to determine virulence factors and antimicrobial resistance genes. Core genome-based multilocus sequence typing was used for isolate comparison and phylogenetic analysis. All isolates were susceptible to all antimicrobials recommended for post-exposure prophylaxis or anthrax therapy. Susceptibility was found to all other tested antimicrobials that are an alternative for primary therapy. We found two beta-lactamase genes, but their expression is not sufficient to confer resistance. In all isolates used in this study, we found 21 virulence genes, 8 of which are responsible for toxin and capsule production. As far as phylogenetic analysis is concerned, the B. anthracis isolates from Croatia are categorised into two clades. The first is clade A, subclade Trans Eurasia, and the other is clade B, subclade B2.

1. Introduction

Bacillus (B.) anthracis, the causative agent of anthrax, belongs to the Bacillus cereus group together with at least six other species (B. cereus, B. thuringiensis, B. mycoides, B. pseudomycoides, B. weihenstephanensis and B. cytotoxicus). They are Gram-positive, spore-forming, aerobic, facultatively anaerobic, rod-shaped bacteria [1]. In contrast to other Bacillus species, it is non-motile, non-haemolytic on sheep blood agar, grows at a temperature of 37 °C, and forms large colonies with irregularly tapered outgrowths. In culture, it tends to form long chains of bacteria. Within a host, however, it appears either as single organisms or as chains consisting of two or three bacilli [2]. Anthrax is primarily a disease of herbivores, although it has also been reported to occur in omnivores, carnivores, and other vertebrates. Herbivorous animals, such as cattle, sheep, and goats, ingest spores in the soil and become infected. Humans can become infected through direct contact with infected animals or tissue from infected animals or through direct exposure to B. anthracis [3]. Based on the initial route of infection, three primary forms of clinical infection in humans are described: inhalational anthrax, gastrointestinal anthrax, and cutaneous anthrax [4] A new form of soft tissue infection associated with injection drug use, termed injectional anthrax, has also been described [5].

B. anthracis is also considered one of the most important pathogens of bioterrorism. Due to the often-fatal outcome of human cases, the rapid administration of clearly effective antimicrobials is critical, whether for prophylaxis, following suspected exposure, or for the treatment of clinical cases. B. anthracis is a bacterium that is sensitive to most antibiotics. However, early treatment is critical to eliminate the bacterium before it releases toxins into the bloodstream [6,7].

When conditions are not favourable for the growth and multiplication of vegetative forms of B. anthracis, they begin to form spores. Spores are highly resistant to biological extremes such as heat, cold, pH, desiccation, chemicals, irradiation, and other unfavourable conditions. The organism can remain in the spore stage for long periods of time, waiting for the moment when conditions are favourable for germination and multiplication [8]. Jacotot and Virat [9] and Wilson and Russell [10] succeeded in reviving spores after more than 60 years, while de Vos [11] recovered anthrax spores from bones after 200 ± 50 years. From the earliest historical records until the development of an effective veterinary vaccine in the middle of the 20th century and the subsequent advent of antibiotics, the disease was a major cause of uncontrolled mortality in cattle, sheep, goats, horses, and pigs worldwide [8]. Anthrax is found all over the world, on every continent except Antarctica. There are endemic areas with more frequent outbreaks, while sporadic outbreaks occur in other areas [8,12,13]. B. anthracis harbours two plasmids, pXO1 (182 kb) and pXO2 (95 kb). The most important virulence factors, the tripartite anthrax toxin genes cya, lef, and pag, are located on the plasmid pXO1 [14,15,16], while the poly-γ-D-glutamic acid capsule genes capABCDE are localised on the plasmid pXO2 [17,18]. Previously, the presence of these two virulence plasmids was considered the main distinguishing feature between B. anthracis and closely related species, particularly Bacillus cereus and Bacillus mycoides [19]. In the meantime, the virulence plasmids pXO1 and pXO2 have also been found in other isolates of the B. cereus group [20,21,22]. B. anthracis is an evolutionarily young pathogen characterised by a lack of molecular variation. Isolates from all over the world show an extremely high genetic homology [23,24]. It is assumed that the dormant state of its spores probably greatly reduces the rate of evolutionary change of B. anthracis and contributes to its extreme homogeneity [25]. Due to this lack of diversity, only modern molecular characterisation techniques have proven effective in distinguishing strains within this highly clonal species. Standard genotyping methods to determine the phylogenetic relationships between B. anthracis isolates include canonical single-nucleotide polymorphisms (canSNPs) and multilocus variable-number tandem repeat analysis (MLVA) [26,27,28]. Recently, core-genome-based multilocus sequence typing (cgMLST) has been increasingly used as a comparative tool using WGS data [29]. Phylogenetic analysis has defined three major clades, with C being basal to clades A and B. Members of clade A are most commonly observed worldwide (~90%), while members of clades B (~10%) and C (<1%) are much less frequent [27]. Clade A is divided into seven subclades: Ancient A, Vollum, V770, Sterne, Ames, Australia 94, and Trans Eurasia [30].

Historical data on anthrax in humans and animals in Croatia can be found in the Supplementary Materials.

Anthrax has occurred only sporadically in Croatia, as immunoprophylaxis was systematically carried out in enzootic areas where anthrax outbreaks had occurred [31]. Since 1981, a total of 59 cases of anthrax in animals have been confirmed at the Croatian Veterinary Institute in Croatia. The course of the disease is characterised by the fact that in some periods the disease does not occur. Within the period of more than four decades, no disease was detected in 20 years (1981–1983, 1992–1996, 1998–2000, 2008–2010, 2013, 2015–2019 and 2021) [31,32]. Only a few major outbreaks have been described in recent decades: 2001/2002 (11 animals–5 isolates) [33], 2006/2007 (11 animals–6 isolates) [34], and the largest outbreak in 2022 (323 animals–25 isolates) [35]. Sporadic cases occur mainly in the southern part of Croatia, where sporadic cases have been recorded over several years (4 isolates tested in this study) [31].

The aim of this study was to describe the reliable and rapid diagnosis of epizootic 2022 and to determine the in vitro antimicrobial resistance (AMR) of 40 isolates of B. anthracis isolated in Croatia in the recent two decades to 18 different antibiotics. In addition, we performed whole-genome sequencing (WGS) on all isolates to determine AMR and virulence genes and used cgMLST as a tool for isolate comparison and phylogenetic analysis.

2. Results

2.1. Isolates

A total of 40 isolates were included in this study. In total, 25 isolates were from the 2022 outbreak, while 15 isolates were obtained from laboratory strain collections from outbreaks in the previous years (2001/2002-5 isolates, 2006/2007-6 isolates) and from areas with sporadic anthrax occurrence (ST) (4 isolates). As shown in Table 1, the presence of the virulent plasmids pXO1 and pXO2 was confirmed in all 40 isolates. Strain Sterne 34F2 lacked the pXO2 plasmid.

2.2. Antimicrobial Susceptibility

The results—expressed as MIC at which 50% and 90% of the tested isolates were inhibited (MIC₅₀, MIC₉₀)—the range (mg/L), the breakpoint, and the percentage of susceptible isolates are shown in Table 2.

All isolates showed susceptibility to quinupristin/dalfopristin, vancomycin, ampicillin, rifampin, levofloxacin, penicillin, oxacillin + 2%NaCl, clindamycin, linezolid, tetracycline, gentamicin, gatifloxacin, and ciprofloxacin. However, 40% of the isolates were intermediate susceptible to ceftriaxone, while 5% were intermediate susceptible to erythromycin. All isolates tested were resistant to daptomycin and trimethoprim/sulfamethoxazole.

2.3. WGS and Bioinformatic Analysis

The genes for antimicrobial resistance and virulence factors, protein function, percentage of overlap, and identities found in 40 B. anthracis strains isolated in Croatia are listed in Table 3.

We identified a total of four AMR genes in all B. anthracis isolates from the 2001/2002 outbreak and isolates from the southern part of Croatia (bla, blaII, fosB, satA), while isolates from the 2006/2007 and 2022 outbreaks lacked the fosB gene.

As far as the virulence genes are concerned, all genes listed in Table 3 were found in all tested isolates without exception.

2.4. Core-Genome-Based Multilocus Sequence Typing (cgMLST)

cgMLST analysis was performed for 3803 core loci and 1263 accessory genomic loci. Forty-seven and forty-five loci were missing in the 2001/2002 outbreak samples and in the samples from southern Croatia, respectively. Complex type 219 was defined for the samples from the 2001/2002 outbreak and 222 for the samples from the southern part of the country.

In total, 145 and 148 loci were missing in the samples from the 2006/2007 and 2022 outbreaks, respectively. Complex type 220 was defined for samples from 2006/2007 and 221 for samples from the 2022 outbreak.

The phylogenetic analysis of the strains using a neighbour-joining tree cluster analysis based on the cgMLST data (Figure 1) shows that the B. anthracis isolates from the outbreaks in Croatia were divided into two clades. Isolates from the 2001/2002 outbreak and isolates from southern part of country were categorised in clade A, subclade Trans Eurasia, while isolates from the 2006/2007 and 2022 outbreaks were categorised in clade B, subclade B2.

3. Discussion

In Europe, anthrax is considered a sporadically occurring disease that almost exclusively affects animals at pastures [36]. In Croatia, anthrax is also rare in animals and humans, but there are sporadic outbreaks, especially in the anthrax districts, which are mainly located in the Lonjsko Polje region and in the southern parts of Croatia [31]. The situation is similar in the neighbouring countries where the occurrence of anthrax has been described: Italy [37], Slovenia [38], Hungary [39], and Bosnia and Herzegovina [40].

Antimicrobial susceptibility testing plays a key role in the treatment of anthrax infections. Rapid diagnosis is also crucial for post-exposure prophylaxis (PEP) or the early treatment of anthrax to eliminate the bacterium before it releases toxins into the bloodstream [6]. Beta-lactam antibiotics such as penicillin are recommended by the WHO and CDC as prophylactic therapy for anthrax, while tetracyclines and fluoroquinolones are recommended by the CDC for PEP and for the treatment of anthrax [8,41].

As shown in this study, a reliable diagnosis of anthrax can be made within 24 h based on colony morphology, haemolysis on blood agar, and the detection of the plasmids pXO1 and pXO2. In this study, a total of 40 B. anthracis strains isolated in Croatia were analysed for their susceptibility to antimicrobial agents. All tested isolates were susceptible to penicillin, tetracycline, and ciprofloxacin, with MIC₉₀ of ≤0.06, ≤2, and ≤0.5, respectively. The first penicillin resistance in a clinical B. anthracis isolate was described in 1976 in a fatal case of anthrax in Northampton, England [42]. Although penicillin resistance of B. anthracis is thought to be at least 10% [43,44], the largest study which included 110 strains from outbreaks in Italy reported a 100% susceptibility to penicillin [45]. In this study, we found two beta-lactamase genes, bla and blaII. It has already been described that B. anthracis possesses these two beta-lactamase genes, bla1 (penicillinase) and bla2 (cephalosporinase), but that their transcription is very low and gene expression is not sufficient to confer resistance to beta-lactam antibiotics [46,47]. Although these genes are present, it has been hypothesised that they are not expressed due to a truncation of the positive regulatory gene, plcR [48]. The transcriptional activator PlcR is responsible for the expression of extracellular proteins such as haemolysins, enterotoxins, and proteases in B. cereus and B. thuringiensis [49,50]. The sequence of the plcR gene of B. anthracis strains indicates that the expression of the gene results in the synthesis of a truncated PlcR protein, in which codon 214 for glutamic acid (GAA or GAG) in B. cereus and B. thuringiensis is converted to a termination codon (TAA) in B. anthracis by a single nucleotide substitution (G→T) [51]. Although the plcR homologue of B. anthracis is truncated, there are 56 putative plcR binding motifs on the chromosome and 2 on pXO2. Extracellular protein genes include phospholipases, enterotoxins, and haemolysins, and the plcR mutation has been shown to be responsible for a drastic reduction in lecithinase, protease, and haemolysin production, giving B. anthracis a different secretome to B. cereus [48,52]. It has been shown that the transcription of beta-lactamases is controlled by an extracytoplasmic sigma factor (ECF) SigP and its anti-sigma factor RsiP [53]. Beta-lactamase expression in B. cereus and B. thuringiensis is inducible by beta-lactam antibiotics, leading to the proteolysis of RsiP, release of SigP, and transcription of the corresponding genes [54]. This beta-lactam feedback system is not functional in prototypic B. anthracis isolates. However, in a penicillin-resistant B. anthracis strains, the rsiP gene was nonsense mutated and the strain constitutively produced beta-lactamases [53].

All B. anthracis isolates in our study were also found to be susceptible to ampicillin and oxacillin +2%NaCl, with MIC₉₀ less or equal to the minimum concentration tested (≤0.12; ≤0.25). This is consistent with previous publications reporting MICs for the other penicillin that closely matched the penicillin MIC [44].

Regarding other fluoroquinolones used in this study, all isolates were found to be susceptible to levofloxacin and gatifloxacin, with MIC₉₀ values ≤0.25 and ≤1, respectively. Ciprofloxacin and tetracycline resistance of B. anthracis strains is very rare and is only described in a few publications [44]. The susceptibility data presented in this study support the current recommendations for the use of penicillin, tetracycline, and ciprofloxacin in PEP and anthrax threats.

Forty percent of the isolates tested in this study were intermediate susceptible to ceftriaxone, while sixty percent were sensitive. This is in partial agreement with Manzulli et al. [45] and Cavallo et al. [43], who categorised all isolates as intermediate susceptible to ceftriaxone. In contrast to these studies, Jones et al. [55] reported all isolates as sensitive to ceftriaxone. Due to this variable susceptibility, non-penicillin beta-lactams are not recommended for the treatment of anthrax.

All B. anthracis isolated in this study were also susceptible to linezolid, rifampin, gentamycin, vankomycin, quinupristin/dalfopristin, and clindamycin. This is consistent with previous published studies [45,56,57].

Linezolid is used for the treatment of a variety Gram-positive infections. It is a bacteriostatic oxazolidinone with a unique mode of action that inhibits bacterial protein synthesis by preventing the formation of the 70S ribosomal complex [58]. For this reason, cross-resistance is unlikely. It is therefore a good alternative for resistant B. anthracis strains. Although all isolates were susceptible to rifampin and clindamycin, they are not considered a good alternative for anthrax treatment as resistance can easily be generated in vitro [59,60]. Nevertheless, rifampin, clyndamicin, and vancomycin together with fluoroquinolones are considered good candidates for the treatment of inhalational anthrax and meningitis [61]. Athamna et al. [62] found that quinupristin/dalfopristin killed B. anthracis the fastest of all antimicrobial agents tested.

Resistance to trimethoprim/sulfamethoxazole was detected in all tested isolates. This result was to be expected as intrinsic resistance to this antimicrobial agent has already been described [63]. All isolates in this study were also resistant to daptomycin with MIC₅₀ and MIC₉₀ of 4 mg/L. This result is consistent with previous publications [56,64].

Another AMR gene found in all isolates was satA, while the fosB gene was only found in the B. anthracis isolates from the 2001/2002 outbreak and isolates from sporadic outbreaks isolated in southern Croatia. SatA belongs to the streptothricin acetyltransferase gene family and confers resistance to nucleoside antibiotics, which are produced by actinomycetes in soil. It has already been described in B. anthracis and B. subtilis [65]. The fosB gene is a chromosomally derived thiol transferase and confers resistance to fosfomycin [48]. In their study, Bruce et al. [66] assumed that the detection of the fosB gene can be a clear prediction for the affiliation of the isolates to clade C. Clade C is largely defined by the absence of the fosB gene, which is found across all other populations. However, this is in partial agreement with the results of our study, in which all isolates that possessed the fosB gene belonged only to clade A, a subclade of Trans Eurasia, but not to clade B. Similar results to ours were also reported by Chiaverini et al. [67].

Pathogenicity-related plasmids, pxO1 and pXO2, were found in all 40 B. anthracis isolates in this study. In addition, all isolates possessed the atxA gene, which is localised on the pathogenicity island pXO1 and is responsible for the expression of the three toxin genes [68,69].

Other virulence genes found in all B. anthracis strains in this study were inhA1, inhA2, entFM, nheABC operon, plcR, sph, plcA, and alo.

InhA1 cleaves multiple host proteins, including collagen, fibrinogen, plasminogen activator inhibitor, and prothrombin [70,71]. It also downregulates BslA, a B. anthracis cell surface protein associated with the adhesion of vegetative bacterial cells to the host endothelium [72]. The B. anthracis strains in this study also contained a gene encoding the metalloprotease InhA2, although it is not known whether this protease is expressed and secreted. This gene is an orthologue of InhA1 described above (68% amino acid identity) [73]. The metalloprotease (inhA) and the phospholipases sph genes are important virulence factors as they enable the bacilli to evade innate and adaptive immune responses during the infective phases [74,75].

The chromosomally encoded nheABC operon was previously described in all B. anthracis isolates [76]. It is responsible for the synthesis of termolabile non-haemolytic enterotoxin, which, together with enterotoxin FM and haemolysin HII, is involved in the diarrhoeic syndrome [77,78,79].

In this study, we used the cgMLST method to genotype B. anthracis isolates from the recent two decades in Croatia. The development of a cgMLST scheme for the entire B. cereus group is timely and ensures high-resolution strain typing of B. cereus group isolates in an open and fully publicly accessible online user interface for the future [80]. The first application of a whole-genome allele typing system was presented by Abdel-Glil et al. [29]. We used cgMLST-searching 3803 loci and accessory cgMLST-searching 1263 loci. After phylogenetic analysis based on the cgMLST data neighbour-joining tree was constructed (Figure 1). The colours represent the clade and subclade, which are also indicated by name, while the node labels correspond to the names of the strains and the country of origin. As shown in Figure 1, the Croatian B. anthracis isolates were categorised into two main clades, depending on the outbreaks: A and B. isolates from the 2006/2007 and 2022 outbreaks were categorised in clade B, subclade B2, while the isolates from the 2001/2002 outbreak and isolates from sporadic cases from the southern part of Croatia (ST2011) were categorised in clade A, subclade Trans Eurasia (TEA). The existence of two different genotypes of B. anthracis isolates in Croatia was also indicated by the finding that the fosB gene was discovered in all isolates from the 2001/2002 outbreak and in isolates from sporadic outbreaks from southern Croatia, while it was not present in isolates from the 2006/2007 and 2022 outbreaks.

It has already been reported that strains belonging to clade A are responsible for the majority of anthrax cases reported worldwide, while the TEA subclade is one of the most common. These isolates are widespread in Europe, the Middle East, Russia, and the Chinese province of Xinjang [27,81]. In Europe, it is the most dominant subclade in Albania, Bulgaria, Hungary, and Italy [27,82]. The MLVA analysis of 234 B. anthracis strains isolated in Italy between 1972 and 2018 revealed the distribution of 55 B. anthracis genotypes, while the CanSNPs analysis categorised 53 of them into a TEA subclade [83]. As can be seen in Figure 1, the Croatian isolates assigned to the TEA subclade showed the greatest homology with the Italian isolate (3IZSLT).

Clade B occurs mainly in South Africa and in the Central Europe, especially in France and Poland. The South African strains mainly belong to the B1 subclade (B.Br.Kruger and B.Br.001/002 groups) [84], while the European strains belong to the B2 subclade (B.Br.CNEVA) [85,86]. Examples of these subclades are rarely found outside these regions [27]. The B2 subclade is widespread in Europe in south-west France, north-east Poland, northern Italy, Switzerland, Germany, and Slovenia, and it is the most dominant subclade in France, Poland, and northern Italy [87,88,89]. The Croatian B. anthracis isolates belonging to subclade B2 showed the most similarities with the Austrian isolates (Tyrol 3520) and, as can be seen in Figure 1, the entire B2 subclade branches from isolate ‘RA3 France’. Similar results were also obtained by Chiaverini et al. in Italy [67].

The TEA subclade has been found to occur in countries where the B2 subclade is dominant and is associated with sporadic cases [90], which is in complete agreement with our study.

These data show that anthrax outbreaks in Croatia are limited to a relatively small area of the country and are caused by two genetically very different B. anthracis genotypes belonging to two different, phylogenetically unrelated clades, indicating a different origin of the isolates.

4. Materials and Methods

4.1. Epidemiological Investigation

At the end of June 2022, cattle breeders noticed an increase in cattle deaths at the Osekovo, Donja Gračenica, and Veliko Svinjičko pastures in the Lonjsko Polje Nature Park area in Croatia (Figure 2).

Considering the fact that it was very close to the border of the anthrax district, the Veterinary and Food Safety Authority authorised the dissection of the two carcasses. The first dissection of non-autolytic cattle was carried out on 15 July 2022 at the rendering plant under controlled conditions and the samples (blood sample, spleen) were delivered to the Laboratory of General Bacteriology and Mycology of the Croatian Veterinary Institute on the same day, as it was suspected that the animals had died of anthrax. Since the beginning of the outbreak of the disease until the end of 2022, 323 animals (246 cattle, 77 horses) have died in the Lonjsko Polje Nature Park area. This number is probably even higher, as we only have data on the number of dead animals brought to the rendering centre.

By the end of 2022, a total of 101 necropsies were performed, mainly on animals that died on the same day or at least the day before. The autopsy of the corpses revealed that there was no rigour mortis and that unclogged blood leaked from all natural orifices. Oedema was found in the subcutaneous tissue. Splenomegaly, haemorrhages in the epicardium and endocardium, and pulmonary oedema with haemorrhages in the parenchyma were noted. The lymph nodes were enlarged and oedematous, and haemorrhages were observed in the parenchyma. Ulcerations were found on the mucosa of the abomasum and small intestine. Petechial haemorrhages were found in the serosa of the intestine.

Anthrax in humans in the Republic of Croatia is characterised by the occurrence of isolated cases. According to data from the Croatian Agency for Agriculture and Food, 16 cases of the disease were recorded in the Republic of Croatia in the period from 2005 to 2016 [91], while, according to data from the Croatian Institute of Public Health [92], two cases of the disease were recorded in humans in the period from 2017 to 2021 (in 2018). Figure 3.

The anthrax epidemic in humans in the Lonjsko Polje Nature Park began on 13 July 2022 and ended on 19 August 2022. Seventeen people fell ill with the cutaneous form of anthrax (Figure 4). Among the infected were 3 children up to the age of 8, 4 aged 15–18, and 10 adults. All had contact with sick cattle and were treated with antibiotics [93].

4.2. Laboratory Diagnosis

Fresh samples (mainly blood, spleen, and ears, with parts of the musculature) were sent to the Laboratory for General Bacteriology and Mycology of the Croatian Veterinary Institute. The spleen samples and ears with muscles were delivered in sterile plastic containers with screw caps, with each container being wrapped in two plastic bags to prevent leakage. The blood samples were delivered in syringes, and each syringe was placed in a plastic container with a screw cap.

Thin smears were made from all samples supplied and a Giemsa stain was used. The samples were inoculated onto blood agar (OXOID blood agar base No. 2 + 5% defibrinated sheep blood). After incubation under aerobic conditions at a temperature of 37 °C for 18–24 h, the plates were examined to determine bacterial growth, morphology, and colour of the bacterial colonies. Bacterial colonies morphologically consistent with B. anthracis were resuspended in distilled water and heated to 95 °C for 20 min. After cooling and centrifugation, the supernatant was used for a PCR reaction to confirm the presence of plasmids pXO1 and pXO2, using the previously described method [94,95].

The presence of plasmids pX01 and pX02 was confirmed in all B. anthracis strains from this epizootic as well as in the other strains from previous outbreaks in Croatia.

4.3. Isolates Used in This Study

We analysed a total of 40 B. anthracis isolates: 5 isolates from the 2001/2002 outbreak, 6 isolates from the 2006/2007 outbreak, 4 isolates from a southern part of Croatia, and 25 isolates from the 2022 outbreak. The molecular part of this study included the B. anthracis Sterne 34F2 (pXO1+, pXO2−) strain.

4.4. Antimicrobial Susceptibility Testing

Antimicrobial susceptibility testing (AST) was performed using the broth microdilution method according to Clinical and Laboratory Standards Institute (CLSI) guidelines [96].

Briefly, isolates were recovered from freezer stocks (tryptic soya broth with 20% glycerol, −80 °C) and incubated overnight on blood agar supplemented with 5% sheep blood in an aerobic atmosphere. AST was performed on GPN3F microtiter plates (Sensititer, Trek Diagnostic Systems Ltd. East Grinstead, West Sussex, UK). Cation-adjusted Mueller–Hinton broth was used to prepare the 0.5 McFarland solution. The microtiter plates were incubated in sealed 96-well microtiter plates at 35 ± 1 °C for 18 ± 2 h.

The susceptibility to erythromycin (ERY; 0.25–4 mg/L), quinupristin/dalfopristin (SYN; 0.12–4 mg/L), vancomycin (VAN; 1–128 mg/L), ampicillin (0.12–16 mg/L), rifampin (0.5–4 mg/L), levofloxacin (0.25–8 mg/L), penicillin (0.06–8 mg/L), trimethoprim/sulfamethoxazole (0.5/9.5–4/76 mg/L), oxacillin + 2%NaCl (0.25–8), ceftriaxone (8–64 mg/L), clindamycin (0.12–2 mg/L), daptomycin (0.25–8 mg/L), linezolid (0.5–8 mg/L), tetracycline (TET 2–16 mg/L), gentamicin (2–16 mg/L), gatifloxacin (1–8 mg/L), ciprofloxacin (0.5–2 mg/L), and streptomycin (1000 mg/L) was tested.

The lowest concentration of an antibiotic that prevented bacterial growth was taken as the minimum inhibitory concentration (MIC).

For penicillin, ampicillin, tetracycline, ciprofloxacin, and levofloxacin, the reported CLSI breakpoints for B. anthracis were used; for ceftriaxone, vancomycin, gentamicin, erythromicin, clindamycin, trimethoprim/sulfamethoxazole, and rifampin, Bacillus spp. breakpoints were used [96]. Interpretation criteria for Staphylococcus sp. were used for quinupristin/dalfopristin, oxacillin + 2%NaCl, linezolid, and gatifloxacin [97].

Streptomycin was not considered further, as there was only one concentration of this antimicrobial agent on the microplate, which was not important for further consideration.

The reference strains Staphylococcus aureus ATCC 29213 and Escherichia coli ATCC 25922 were used for quality control.

4.5. Whole-Genome Sequencing (WGS) and De Novo Assembly

The B. anthracis strains were sequenced using Illumina MiSeq (Illumina, San Diego, CA, USA). DNA was extracted using the Nucleospin Microbial DNA Kit (Macherey-Nagel, Düren, Germany). Paired-end sequencing libraries were prepared using the DNA LP Tagmentation kit (Illumina, USA), and the sequencing was performed using the MiSeq Reagent Kit v2 (300 cycles) (Illumina, USA). The raw sequences were trimmed with Trimmomatic v. 0.38.1. (options SLIDINGWINDOW 4, quality 20; MINLEN 100; LEADING 15; TRAILING 15; HEADCROP 15; AVGQUAL 25) [98]. Genome assembly was performed with Shovill v.1.1.0 (DEPTH 100; GENOME SIZE 5,5 Mb; POSTASSEMBLY CORRECTION 500; MIN COVERAGE 5) [99] for paired-end Illumina data. The assembly statistics were obtained using the program Quast v.5.0.2 [100].

The genome assemblies were deposited in GenBank under the BioProject accession number PRJNA1129309 (SAMN42150189—2001/2002 outbreak; SAMN42150190—2006/2007 outbreak; SAMN42150191—isolates from southern Croatia; SAMN42150192—2022 outbreak).

4.6. Bioinformatic Analysis of Antimicrobial Resistance and Virulence Genes

Virulence was tested with the VFDB database [101]. NCBI AMRFinderPlus v.3.11.26 was used for the identification of the resistance and virulence genes using BLASTX search [102]. MobileElementFinder v.1.0.5 was used for the identification of Mobile Genetic Elements (MGE) [103].

4.7. Core-Genome-Based Multilocus Sequence Typing (cgMLST)

cgMLST typing was performed using Ridom SeqSphere+ v.10.0.0 [104]. Both cgMLST-searching 3803 loci and accessory cgMLST-searching 1263 loci were used. For all analysed genomes, the cgMLST allelic profiles were pairwise-compared (missing values were ignored), and the resultant calculated pairwise distances were used to generate a neighbour-joining tree using SeqSphere+.

5. Conclusions

Anthrax in animals and humans in Croatia is characterised by sporadic outbreaks, which are limited to anthrax districts located mainly in the Lonjsko Polje region and in the southern parts of Croatia.

Rapid and reliable diagnosis and administration of antimicrobials are essential for PEP or effective treatment of anthrax. Molecular diagnosis, based on the presence of pXO1 and pXO2 plasmids and the morphological appearance of colonies, enables reliable anthrax diagnosis within 24 h in material that is not heavily contaminated.

Isolates of B. anthracis isolated in Croatia in the recent two decades were susceptible to all antimicrobials recommended for PEP or anthrax therapy (penicillin, ampicillin, tetracycline, and ciprofloxacin). Susceptibility was observed for all other antimicrobials tested, which represent an alternative for primary therapy. Three or four AMR genes were found in our isolates, depending on the area and year of the outbreak, of which two beta-lactamase genes were not expressed. In all isolates used in this study, we found 21 virulence genes, 8 of which are responsible for toxin and capsule production.

As far as phylogenetic analysis is concerned, the B. anthracis isolates from the outbreaks in Croatia are divided into two clades. The first is clade A, subclade Trans Eurasia, which is predominant in the Mediterranean region and the eastern part of Europe and was responsible for the 2001/2002 outbreak and the outbreaks in the southern part of the country. Isolates from the 2006/2007 and 2022 outbreaks are categorised in clade B, subclade B2, which is dominant in Central European countries. These results indicate that two different B. anthracis clades are circulating in a very small area, which is probably due to the geographical location of Croatia.