ObjectivesThe aim of this study was to characterize by classical biotyping and Multi-Locus variable number tandem repeats (VNTR) Analysis (MLVA) all Brucella spp. derived from human cases in Belgium from 1996 to 2015. Final goals were to determine the species and biovar, to trace-back on genetic grounds the origin of each strain when patient history and risk factors were missing, and to survey for particular trends at the national level.MethodsA total of 37 Brucella strains, isolated from 37 patients in Belgium, were analyzed by both classical biotyping and MLVA, and the genetic patterns compared to those of human strains isolated worldwide.ResultsClassical biotyping revealed that isolates were mainly Brucella melitensis. Most of them belonged to biovar 3, the most abundant biovar in the Mediterranean region. MLVA confirmed that Brucella melitensis is too diverse in VNTRs to be able to make clusters associated to each biovar, but it allowed retrieving precious epidemiological information. The analysis highlighted the imported nature of the strains from all over the world with a dominant part from the Mediterranean countries. Findings of the MLVA11 testing were in line with the travel history of patients coming from Italy, Turkey, Lebanon and Peru. The analysis was particularly useful because it suggested the geographical origin of the infection for 12/16 patients for whom no case history was available.ConclusionClassical biotyping and MLVA analysis are not exclusive but remain complementary tools for Brucella melitensis strain surveillance. MLVA11 is sufficient for Brucella-free countries such as Belgium to trace the geographical origin of infection, but complete MLVA16 is needed to search for links with endemic areas.

Q fever is a zoonotic disease caused by the bacteria Coxiella burnetii. Domestic ruminants are the primary source for human infection, and the identification of likely contamination routes from the reservoir animals the critical point to implement control programs. This study shows that Q fever is detected in Belgium in abortion of cattle, goat and sheep at a different degree of apparent prevalence (1.93%, 9.19%, and 5.50%, respectively). In addition, and for the first time, it is detected in abortion of alpaca (Vicugna pacos), raising questions on the role of these animals as reservoirs. To determine the relationship between animal and human strains, Multiple Locus Variable-number Tandem Repeat Analysis (MLVA) (n=146), Single-Nucleotide Polymorphism (SNP) (n=92) and Whole Genome Sequencing (WGS) (n=4) methods were used to characterize samples/strains during 2009-2019. Three MLVA clusters (A, B, C) subdivided in 23 subclusters (A1-A12, B1-B8, C1-C3) and 3 SNP types (SNP1, SNP2, SNP6) were identified. The SNP2 type/MLVA cluster A was the most abundant and dispersed genotype over the entire territory, but it seemed not responsible for human cases, as it was only present in animal samples. The SNP1/MLVA B and SNP6/MLVA C clusters were mostly found in small ruminant and human samples, with the rare possibility of spillovers in cattle. SNP1/MLVA B cluster was present in all Belgian areas, while the SNP6/MLVA C cluster appeared more concentrated in the Western provinces. A broad analysis of European MLVA profiles confirmed the host-species distribution described for Belgian samples. In silico genotyping (WGS) further identified the spacer types and the genomic groups of C. burnetii Belgian strains: cattle and goat SNP2/MLVA A isolates belonged to ST61 and genomic group III, while the goat SNP1/MLVA B strain was classified as ST33 and genomic group II. In conclusion, Q fever is widespread in all Belgian domestic ruminants and in alpaca. We determined that the public health risk in Belgium is likely linked to specific genomic groups (SNP1/MLVA B and SNP6/MLVA C) mostly found in small ruminant strains. Considering the concordance between Belgian and European results, these considerations could be extended to other European countries.

Ruminants are considered the commonest animal reservoir for human infection of Coxiella burnetii, the Q fever causative agent. Considering the recently described importance of human Q fever in Greece, we aimed at providing the first comprehensive direct evidence of C. burnetii in dairy cows in Greece, including the genetic characterization of strains. The 462 examined dairy farms represented all geographical areas of Greece. One bulk tank milk sample was collected from every farm and tested for the presence of C. burnetii. Molecular genotyping of strains, performed directly on samples, revealed the existence of two separate clades characterized by single nucleotide polymorphism (SNP) genotypes of type 1 and type 2. The two clades were clearly distinguished in multiple locus variable-number tandem repeat analysis (MLVA) by two discriminative loci: MS30 and MS28. Whereas MLVA profiles of SNP-type 2 clade were closely related to strains described in other European cattle populations, the MLVA profile observed within the SNP type 1 clade highlighted a peculiar genetic signature for Greece, related to genotypes found in sheep and goats in Europe. The shedding of C. burnetii bearing this genotype might have yet undefined human epidemiological consequences. Surveillance of the genetic distribution of C. burnetii from different sources is needed to fully understand the epidemiology of Q fever in Greece.