European Journal of Clinical Microbiology & Infectious Diseases

, Volume 31, Issue 10, pp 2783–2790

Molecular epidemiology of the Bacillus anthracis isolates collected throughout Turkey from 1983 to 2011


    • Department of Clinical Microbiology, Medical FacultyKirikkale University
    • Molecular Microbiology Research and Application LaboratoryRefik Saydam National Public Health Agency
  • M. Doganay
    • Department of Infectious Disease, Medical FacultyErciyes University
  • M. Sahin
    • Department of Microbiology, Veterinary FacultyKafkas University
  • D. Percin
    • Department of Clinical Microbiology, Medical FacultyErciyes University
  • M. K. Karahocagil
    • Department of Infectious Disease, Medical Faculty100 Yil University
  • U. Kayabas
    • Department of Infectious Disease, Medical FacultyInonu University
  • B. Otlu
    • Department of Clinical Microbiology, Medical FacultyInonu University
  • A. Karagoz
    • Molecular Microbiology Research and Application LaboratoryRefik Saydam National Public Health Agency
  • F. Buyuk
    • Department of Microbiology, Veterinary FacultyKafkas University
  • O. Celebi
    • Department of Microbiology, Veterinary FacultyKafkas University
  • Z. Ozturk
    • Department of Infectious Disease, Medical FacultyAtatürk University
  • M. Ertek
    • Refik Saydam National Public Health Agency
  • Anthrax Study Group

DOI: 10.1007/s10096-012-1628-4

Cite this article as:
Durmaz, R., Doganay, M., Sahin, M. et al. Eur J Clin Microbiol Infect Dis (2012) 31: 2783. doi:10.1007/s10096-012-1628-4


The main perspective of this study was to determine cross-transmissions amongst anthrax cases and provide detailed information regarding the genotypes of Bacillus anthracis isolates circulating in Turkey. A total of 251 B. anthracis isolates were obtained from human (93 isolates), animal (155 isolates), and environmental (three isolates) samples in various provinces of Turkey. All isolates were susceptible to quinolones, vancomycin, tigecycline, and linezolid, but not to ceftriaxone. Excluding human isolates, one of the animal isolates was found to be resistant to penicillin, erythromycin, and doxycycline. Multiple-locus variable-number tandem repeats analysis including 8 loci (MLVA8) revealed 12 genotypes, in which genotype 43 was observed at the highest frequency (41.8 %), followed by genotype 35 (25.5 %) and genotype 27 (10.4 %). Major subtype A3.a was the predominant cluster, including 86.8 % of the isolates. The MLVA25 analysis for the 251 isolates yielded 62 different genotypes, 33 of which had only one isolate, while the remaining 29 genotypes had 2 to 43 isolates, with a total of 218 isolates (86.9 %). These findings indicate very high cross-transmission rates within anthrax cases in Turkey. The genotypes diagnosed in Turkey are populated in the A major cluster. Penicillin prescribed as the first-choice antibiotic for the treatment of anthrax is still effective.


Anthrax is a zoonotic disease caused by Bacillus anthracis, which is a Gram-positive, capsule-, and spore-forming rod-like bacterium. The spores of this bacterium are very resistant to external conditions and can survive for decades based on the pH of soil, chemicals, irradiation, humidity, and seasonal features [1, 2]. Soil is the natural reservoir of B. anthracis. Herbivores become infected by B. anthracis by contacting with spores while feeding on grass. Infection in animals results in gastro-intestinal anthrax and sepsis, ultimately leading to death [3]. Humans acquire anthrax via direct or indirect contact with the infected animals’ meat, wool, bone, hair, hide, and so on. Anthrax in humans can be of industrial, agricultural, or laboratory origin [4]. Depending on the method of entry, three clinical forms can be seen as cutaneous anthrax, which is the most common form including more than 95 % of all cases, gastro-intestinal anthrax, and respiratory tract anthrax [5].

Although anthrax decreased to a very low level in developed countries, it has been seen widely in several Asian and African countries where vaccination programs are inadequate [2]. It is hyperendemic in Guinea, Niger, Chad, Zambia, and Zimbabwe in Africa, and Myanmar and Tajikistan in Asia [1, 2, 4]. According to record of the Turkish Ministry of Health, a total of 6,873 human anthrax cases were reported between 1995 and 2005 in Turkey, during which the mortality rate varied from 0.38 to 0.77 per 100,000 people [6]. Looking at the distribution of human anthrax cases on a regional basis in Turkey, the Eastern Anatolia region possesses the highest rate of human anthrax cases, followed by Central Anatolia, which has three times fewer cases than Eastern Anatolia [79]. By increasing the rate of animal vaccination and awareness adapted to the control programs in Turkey, the number of human anthrax cases decreased from 396 in 2000 to 148 in 2009, and finally dropped to 93 cases in 2010 [10]. More than 95 % of the human cases in Turkey have been seen as diagnosed with cutaneous anthrax [5]. In terms of animal anthrax cases, a total of 1,304 cattle were influenced by anthrax between 1996 and 2004 in Turkey, and 882 of those died, while the number of influenced and death sheep/goats was 556 and 518, respectively [11].

Molecular tools have been widely used to determine cross-transmissions in B. anthracis isolates between or within different sources, such as soil, animals, and humans. Moreover, they provide useful data regarding the introduction and distribution of predominant genotyping patterns in a population, region, and country, or all over the world. The studies used for the molecular typing of anthrax bacilli are commonly based on variable-number tandem repeats located in multiple loci in chromosome and pXO1 and pXO2 plasmids [1214]. The application of multiple-locus variable-number of tandem repeats analysis (MLVA) for typing B. anthracis isolates from different countries revealed that MLVA genotypes possess a geographically specific distribution [1517]. There has been only two reports in literature the so far [12, 18] on the MLVA types of B. anthracis, which covers only a limited number of isolates in Turkey. In order to gain further information regarding cross-transmission rates and dynamics, as well as major genotypes and clones in human, animal, and environmental isolates, the MLVA25 analysis was conducted for typing the 251 B. anthracis isolates collected throughout Turkey over a 28-year period. In addition, it was aimed in this study to determine the antimicrobial resistance of the isolates and to address whether there is any relationship between drug resistance and MLVA genotypes.

Materials and methods

Bacillus anthracis isolates

A total of 251 B. anthracis isolates collected both prospectively and retrospectively were analyzed. The isolation and identification of B. anthracis were done following the protocol described by Turnbull et al. [2]. The isolates which were nonmotile, sensitive to penicillin and gamma phage, and have positive McFadyean stain were identified as B. anthracis. All available clinical, demographic, and epidemiological information of the study population were recorded. Of the 251 isolates, 93 (37.05 %) were cultured from humans, 155 (61.75 %) were from animals, and 3 (1.2 %) were from environmental samples. There was no information in regards to the type of clinical samples for the 21 human isolates. Sixty-nine isolates (95.8 %) out of the remaining 72 human isolates were obtained from the cutaneous anthrax, one isolate (1.4 %) from the blood, and one (1.4 %) from the cerebrospinal fluid. Of the 155 animal isolates, 129 isolates (83.2 %) were acquired from cattle, 23 (14.8 %) were from sheep, and three (1.9 %) were from goats. Detailed information regarding the distribution of the isolates according to the geographic regions, provinces, isolation source, and period is given in Table 1. B. anthracis Sterne strain was obtained from the Refik Saydam National Public Health Agency, Ankara, Turkey.
Table 1

Distribution of the 251 isolates according to geographic regions, provinces, isolation source, and period

Geographic regions


Human (n)

Animal (n)

Environment (n)

Total, n( %)

Isolation period

Black Sea (n = 10)



1 (0.4)




2 (0.8)

2004, 2006



1 (0.4)




1 (0.4)




1 (0.4)




4 (1.6)


Central Anatolia (n = 94)



13 (5.2)




1 (0.4)






22 (8.8)




3 (1.2)




8 (3.2)




1 (0.4)





38 (15.1)





7 (2.7)






1 (0.4)


East Anatolia (n = 143)




2 (0.8)

1997, 2003



3 (1.2)




10 (4.0)




9 (3.6)






98 (39)






18 (7.2)




3 (1.2)


Southeast Anatolia (n = 1)



1 (0.4)


Mediterranean Region (n = 2)




2 (0.8)

1995, 2006

Marmara Region (n = 1)



1 (0.4)


Total, n (%)

93 (37.05)

155 (61.75)

3 (1.2)



Drug susceptibility testing

All isolates were tested for susceptibility to penicillin G, ceftriaxone, ciprofloxacin, levofloxacin, doxycycline, tigecycline, linezolid, erythromycin, gentamicin, and vancomycin. The susceptibility tests were performed by the agar dilution method on Mueller–Hinton agar medium [19]. Briefly, a bacterial suspension of 2 × 106 CFU/ml was prepared for each isolate and 5 μl of each suspension was inoculated onto agar plates including appropriate concentrations of each antibiotic. The inoculated plates were incubated at 35 °C for 18 h. The minimum inhibitory concentration (MIC) was regarded as the lowest concentration of an antimicrobial agent at which complete inhibition of bacterial growth was observed.

Molecular typing

Multiple-locus variable-number of tandem repeats analysis including 25 loci (MLVA25) was used for the molecular typing of B. anthracis isolates following the protocol by Lista et al. [14]. Briefly, 2–3 colonies on a blood agar plate were suspended in 100 μl of TE buffer (10 mM Tris–HCl, 0.1 mM EDTA, pH 7.6). The bacterial suspension was heated at 100 °C for 15 min and then centrifuged at 10,000 × g for 1 min. The supernatant was used as the DNA source. Four multiplex PCR reactions were prepared to amplify 25 loci. The amplification mixture included 2 μl of 10 × PCR buffers, 0.2 mM of dNTP mix, primers with concentrations indicated by Lista et al. [14], 0.75 U of Taq DNA polymerase, and 5 μl of DNA suspension. The amplification conditions were as follows: an initial denaturation at 96 °C for 3 min, followed by 36 cycles at 95 °C for 20 s, 60 °C for 30 s, and 72 °C for 120 s. The amplification product was electrophoresed through 2 % agarose gel and amplification was controlled. Then, the PCR products were diluted by 1 to 5 times using distilled water. A quantity of 5 μl of the diluted products were mixed with 40 μl of sample loading solution (Beckman Coulter, Inc., Boston, MA, USA) and 0.5 μl of WellRED labeled size standard (MapMaker 1200; BioVentures, Inc., Murfreesboro, TN, USA). Then, capillary electrophoresis was implemented on a Beckman Coulter 8000 instrument (Beckman Coulter, Inc., Boston, MA, USA) using the following conditions: denaturation at 90 °C for 120 s, injection at 2.0 kV/20 s, and separation at 3.0 kV/3 h. The fragment size of each locus was estimated using the size standard, which included fragments from 75 to 1,200 bp.

Data analysis

The BioNumerics (Applied Maths, Sint-Martens-Latem, Belgium) version 6.1 software program was used to analyze the genetic relationship amongst the isolates analyzed. Dendrograms of the MLVA25 and MLVA8 typing results were generated using the unweighted pair group method with arithmetic mean (UPGMA) algorithm. MLVA types were compared to epidemiological data.


Drug susceptibility of the isolates

All of the 251 isolates analyzed were found to be susceptible to vancomycin, quinolones, tigecycline, and linezolid, whereas they were resistant to ceftriaxone. It was determined that only a limited number of animal isolates were resistant to penicillin G (one isolate; 0.63 %), doxycycline (one isolate; 0.63 %), erythromycin (one isolate; 0.63 %), and gentamicin (two isolates; 1.3 %). Fifty-eight isolates (23.1 %) were intermediately susceptible to erythromycin. The percentage rate of the isolates susceptible to erythromycin at intermediate levels was 29 % (27 isolates) for those collected from humans and 19 % (31 isolates) for those collected from animal and environmental samples.

MLVA typing results

MLVA analysis yielded complete typing data for the isolates tested. All isolates possessed the pX01 and pX02 plasmid markers. The MLVA25 profile of the ‘Sterne’ strain for CG3, bams44, bams3, vrrB2, bams5, bams15, bams1, vrrC1, bams13, vrrB1, bams28, vrrC2, bams53, bams31, vrrA, bams25, bams21, bams34, bams24, bams51, bams22, bams23, bams30, pXO1-aat, and pXO2-at was 158, 417, 579, 162, 385, 616, 485, 583, 886, 229, 493, 532, 236, 781, 313, 391, 676, 425, 595, 493, 735, 651, 691, 129, and 135 base pairs, respectively.

MLVA8 results

A total of 12 MLVA8 genotypes were determined amongst the 251 B. anthracis isolates. All 12 MLVA8 genotypes were found to correspond to 89 GK genotypes previously determined by Keim et al. [12]. The GK 43 was the most frequently encountered genotype, which comprised 105 isolates (41.8 %), followed by the GK 35 genotype with 64 isolates (25.5 %), the GK 27 genotype with 26 isolates (10.4 %), and the GK 44 genotype with 16 isolates (6.4 %). The remaining 40 isolates correspond to eight different GK genotypes of the least occurrence (Fig. 1). The GK 43 genotype was detected in both animal and human isolates obtained from provinces in the Eastern Anatolia, Central Anatolia, and Marmara regions. The other two dominant genotypes GK 35 and GK 27 were detected in isolates obtained from the provinces located in the Eastern Anatolia, Central Anatolia, and Black Sea regions.
Fig. 1

A dendrogram of the 12 multiple-locus variable-number tandem repeats analysis including 8 loci (MLVA8) genotypes found among the 251 Bacillus anthracis isolates which were typed. Three major clusters were detected; A3.a included 218 isolates (86.9 %), A1.b included 29 isolates (11.6 %), and A1.a included four isolates (1.6 %). G: genotype, n: no. of isolates, Geo region: geographic region: TURK: Turkey, NAM: North America, UK: United Kingdom, USA: United State of America, SAF: South Africa, ARG: Argentina, MOZ: Mozambique, ZIM: Zimbabwe, NOR: Norway

All 12 MLVA8 genotypes were in the A major cluster, in which the A3.a cluster comprised 86.8 % of the isolates, the A1.b cluster included 11.6 % of the isolates, and the remaining isolates (1.6 %) corresponded to the A1.a cluster. No isolate/genotype was found in the B major cluster. Except for two provinces, the subtype A3.a cluster was found to be in almost all provinces from where B. anthracis isolates were collected in this study. The rare subtype A1.b was found in nine provinces, whereas the second rare subtype A1.a was detected in three provinces (Fig. 2). The clusters A3.a and A1.b have existed in Turkey from 1983 to 2011 and from 1990 to 2011, respectively, while the cluster A1.a existed between 1994 and 2006.
Fig. 2

Distribution of MLVA8 clusters in provinces of Turkey. The predominant A3.a subtype was spread throughout the nation. The subtype A1.b was found to be in nine provinces; Igdir, Van, Malatya, Siirt, Sivas, Kayseri, Yozgat, Kirsehir, and Samsun, whereas the second rare subtype A1.a was detected in only three provinces; Yozgat, Icel, and Sivas

MLVA25 results

Sixty-two different genotypes (from GT1 to GT62) were determined amongst the 251 B. anthracis isolates. Thirty-three genotypes comprised only one isolate (one isolate/one genotype). Each of the remaining 29 genotypes comprised 2 to 43 isolates (see the additional file in the Supplementary Material), giving a total of 218 clonally related B. anthracis isolates (86.9 % of 251) in the study population. Amongst the 93 human isolates, 37 different MLVA25 genotypes were detected, 19 of the genotypes possessed one isolate each, while the remaining 18 genotypes included a total of 74 (79.6 %) clonally related isolates. Thirty-eight MLVA25 genotypes were determined amongst 158 isolates from animal and environmental samples. A total of 136 isolates (86.1 %) scattered amongst 16 genotypes were found to be clonally related. One environmental isolate acquired from the Central Anatolia region possessed the same genotype as the isolates acquired from the animals in the Eastern Anatolia region, while the other two environmental isolates exhibited unique genotypes.

No correlation was found between the epidemiological data and MLVA25 typing results obtained in this study. The isolates within the same MLVA25 genotypes did not have strict epidemiological correlations with their isolation period, geographic localization, or host distribution. For instance, the most common MLVA25 genotype (GT42) included 43 isolates, in which 27 isolates were obtained from cattle, 11 isolates from humans, and five isolates from sheep. This genotype existed between 1983 and 2011 in Turkey, which was discovered in four different provinces located in two geographic regions in Turkey. The second most prevailing genotype (GT13) included 24 isolates collected between 1993 and 2011, in which five isolates were obtained from humans, 15 isolates from cattle, three isolates from sheep, and one isolate from a goat. This genotype was seen in six different provinces located in three geographic regions (see the additional file in the Supplementary Material).

Since most of the isolates showed a homogenous drug susceptibility pattern, a correlation between the drug resistance and genotyping results could not be drawn.


The genetic diversity and drug susceptibility of B. anthracis isolates circulating in Turkey were investigated from of a total of 251 B. anthracis isolates collected from animal (155 isolates), human (93 isolates), and environmental samples over a 28-year period. To the best of our knowledge, this is the first multicentral genotyping study conducted on a large number of B. anthracis isolates in Turkey. A great majority of the isolates (62 %) tested in this study correspond to those obtained from animals, which is possibly responsible for a particularly significant number of human anthrax cases in the North-Eastern Anatolia region of Turkey. Nowadays, anthrax is controlled by the vaccination of healthy animals, while infected animals are culled, sacrificed, and burned in compliance with health rules [2]. The presence of B. anthracis in three environmental samples collected in three different provinces is a proof of failure in the decontamination procedure applied to died animals in those provinces.

Penicillin is still the first-choice antibiotic applied in the treatment of anthrax. Because B. anthracis is highly resistant to second- and third-generation cephalosporins, these antibiotics are not suggested in the treatment of anthrax. Nevertheless, erythromycin, tetracyclines, chloramphenicol, quinolones, and first-generation cephalosporins can be chosen as alternative antibiotics in patients who are allergic to penicillin. In severe infections, doxycycline or ciprofloxacin can be used intravenously [20]. Either no [21, 22] or very low resistance (2.8–11.5 %) to penicillin, especially in animal and environmental isolates, is reported in the literature [23, 24]. In addition, high resistance to second- and third-generation cephalosporins was reported [25]. In agreement with the literature findings, in the current study, all of the B. anthracis isolates were found to be susceptible to quinolones, vancomycin, tigecycline, and linezolid, but resistant to ceftriaxone. Although low rates of resistance to penicillin (0.6 %), doxycycline (0.6 %), and gentamicin (1.3 %) were detected amongst the isolates obtained from animals, no resistance was observed amongst the isolates obtained from humans. As indicated in the literature [19, 22], a considerable number of the isolates tested (58 isolates; 23 %) were found to be intermediately resistant to erythromycin; this ratio was 29 % (27 isolates) for those obtained from humans and 19 % (31 isolates) for those from animal samples.

MLVA8 distinguished 12 different GK genotypes and a considerable number of the isolates (105 isolates) was identified as the GK 43 genotype. The two other common types were the GK 35 genotype with 64 isolates and the GK 27 genotype with 26 isolates. The genotypes GK 43 and GK 35 were also observed in two previous studies in Turkey [12, 18]. The GK genotypes were reported in different countries were classified in branch A and in some studies also in branch B [12, 16, 26]. All of the 12 GK genotypes found in the current study were in the A major cluster, which is quite common throughout the world. In the A major cluster, A1.a was determined as being dominant in the North-West part of Bulgaria [27], France [14], Italy [28], and Kazakhstan [26]; A3 in Mongolia [29]; A3.a in Georgia [30], Turkey [12], Iran, and South and East parts of Bulgaria [27]; A4 in Switzerland [31] and Korea [16]. In our study, 86.8 % of the strains were in cluster A3.a, indicating that the B. anthracis isolates circulating in Turkey exhibit the ‘South Caucasian-Turkish’ regional strain pattern, which was also detected in the eastern (Iran) and western (Bulgaria) neighboring countries of Turkey. Using these data, it is proposed that B. anthracis isolates which circulated in Turkey descend from a common ancestor, and that minor mutations within few genomic loci lead to different MLVA genotypes amongst the isolates of the common ancestor. While A1.a and A1.b were detected in only a few provinces in Turkey, the predominant A3.a cluster is spread throughout the nation, indicating that the cluster A3.a has existed for a long time. Indeed, it was determined in our laboratory that the clusters A3.a, A1.a, and A1.b have been present in Turkey since 1983, 1990, and 1994, respectively. Furthermore, cluster B was not seen in any of the isolates studied in this laboratory, which seems to be consistent with other laboratory findings [12, 26] that reflected a narrow distribution for cluster B.

MLVA25 molecular typing provides very useful data regarding the origin and magnitude of cross-transmission within and between animals and human anthrax cases in a study population [14, 32]. MLVA25 yielded only one genotype and confirmed cross-transmission amongst 53 B. anthracis isolates cultured during a 41-day epidemic time interval in Italy [32]. In another study carried out on 160 B. anthracis strains isolated mainly from Italy and France, 67 different MLVA25 genotypes had been identified and the isolates within the same genotype were found to be epidemiologically related [14]. Here, in this study, 62 different MLVA25 genotypes were identified in 251 B. anthracis isolates and 86.9 % of the isolates tested were grouped in 29 genotypes, in which each group included 2 to 43 isolates. These findings indicate that the largest number of anthrax cases in Turkey exhibited cross-transmission. A detailed analysis of the grouped MLVA25 types showed that there was no strict epidemiological correlation with the isolates in the same genotype; cross-transmission was not only observed among the isolates cultured from only humans or animals in a specific province in a limited period, but it was also very common between animals and humans and among provinces of different geographic regions in Turkey. The most numerous MLVA25 type (GT42), which includes 43 isolates, has been existing in Turkey for at least 28 years, confirming ongoing cross-transmission amongst the anthrax cases in Turkey.

The main findings in this study are as follows: (1) more than 95 % of the human isolates originated from cutaneous anthrax, while about 83 % of the animal isolates originated from cattle; (2) except for ceftriaxone, there were no or rare rates of resistance to the antibiotics tested; (3) MLVA25 yielded 62 different genotypes in 251 isolates; the percentage rate of the isolates showing a related MLVA25 profile amongst the isolates obtained from humans and animals were 79.6 and 86.1 %, respectively; (4) MLVA8 resulted in 12 GK genotypes, in which the genotype GK 43 exhibited the highest percentage population, being 41.8 % out of 251 isolates; (5) the MLVA genotype A3.a was found to be the predominant cluster, including 86.8 % of the isolates, followed by clusters A1.b (11.6 %) and A1.a (1.6 %); (6) no isolate was identified in the B cluster; (7) a strict epidemiological correlation amongst the isolates within the same MLVA genotypes was not observed.

In conclusion, although almost half of the anthrax cases were observed in the Eastern Anatolia region of Turkey, these cases can still be seen all over the country. The high cross-contamination rate (86.9 %) among different hosts (human, animal, environment), provinces, and years (1983–2011) is most probably due to the relocation of animals throughout Turkey. The predominant cluster A3.a suggests that B. anthracis isolates circulating in Turkey have a common ancestor including the neighboring countries. The results of the antibiotic susceptibility testing confirmed the current effectiveness of penicillin in the treatment of anthrax cases. All strains tested in this study were delivered to the Bacteriology Laboratory of Refik Saydam National Public Health Agency, Ankara, Turkey, for future studies.


This study was supported by a grant (108 S164) from the Scientific and Technological Research Council of Turkey. The authors thank Dr. Florigio Lista and Dr. Andrea Ciammaruconi from the Histology and Molecular Biology Section, Army Medical and Veterinary Research Center, Via Santo Stefano Rotondo 4, 00184, Rome, Italy, for their help and encouraging suggestions on the MLVA typing protocol.

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

10096_2012_1628_MOESM1_ESM.docx (144 kb)
Additional File 1This file contains a dendrogram of the 62 different MLVA25 genotypes of the 251 Bacillus anthracis isolates from Turkey. (DOCX 144 kb)

Copyright information

© Springer-Verlag 2012