Brucellosis, a widespread bacterial zoonosis, affects livestock and wildlife, and increasingly pets. In dogs, *Brucella canis* is the most common cause, leading to reproductive failure and posing a zoonotic risk, although human infections are underreported. Diagnosis is challenging due to asymptomatic carriers and limitations of traditional methods. Serological tests, relying on *B. abortus* antigens, often fail to detect *B. canis* antibodies. Culture-based methods are time-consuming and hazardous. This study aimed to evaluate the effectiveness of MALDI-TOF MS and whole-genome sequencing for the rapid and accurate identification of *B. canis* in a suspected outbreak in a Brazilian dog breeding kennel, improving diagnostic tools and understanding transmission dynamics.

Canine brucellosis, primarily caused by *B. canis*, is a significant concern in breeding kennels worldwide. Transmission occurs through mating, ingestion of aborted material, or contact with reproductive secretions and urine. Control is difficult due to close dog contact and kennel management practices. *B. suis*, particularly biovar 1 (commonly found in feral pigs), has also been implicated in canine brucellosis, often in hounds exposed to wild boar. The zoonotic potential of *B. canis* and *B. suis* biovars 1 and 4 highlights the need for effective diagnostic tools. Current diagnostic limitations include the difficulty in distinguishing infected from uninfected dogs clinically due to high rates of asymptomatic infection and the limitations of serological tests that don't reliably detect *B. canis* antibodies due to the absence of the immunodominant O-polysaccharide in *B. canis*.

In 2013, a Brazilian breeding kennel experienced multiple canine abortions. Seventeen adult pugs (2 males, 15 females) were investigated. Clinical examination revealed brucellosis-like symptoms (vaginal discharge, lymphadenopathy, abortion/stillbirth, infertility) in 14 dogs. Serological tests (ICT, RSAT, *B. canis* IgG ELISA) were performed on serum samples collected at three time points. Blood cultures were conducted on samples taken at the same time points and plated on Brucella-selective agar after enrichment broth culture. Gram-negative, short rod-shaped, non-motile bacteria were isolated and identified by genus-specific PCR as *Brucella*. Phenotypic characterization was performed using standard microbiological and biochemical tests. MALDI-TOF MS analysis was conducted using an in-house reference database improved to better differentiate *B. canis* from related species. Whole-genome sequencing (NGS) using Illumina MiSeq was performed on selected isolates for phylogenetic characterization and SNP analysis to examine the clonality and evolutionary relationships. Bioinformatics tools including SPAdes, Quast, BowTie2, QualiMap, Gegenees, Bandage, and ParSNP were employed for genome assembly, quality assessment, read mapping, SNP calling and phylogenetic tree construction. BioNumerics and Mauve were used for independent SNP analysis.

Serological tests showed antibodies to rough *Brucella* antigens in all 17 dogs, but results were often inconsistent. *Brucella* was isolated from 15 out of 17 dogs at least once via blood culture, totaling 29 isolates. PCR detected *Brucella* DNA in three additional dogs where culture failed. Phenotypic characterization suggested *B. canis*. MALDI-TOF MS, using an optimized in-house database, accurately identified all isolates as *B. canis*, differentiating them from *B. suis* biovar 1 and even the closely related *B. suis* biovar 4 by identifying novel discriminatory biomarker peaks. Whole-genome sequencing confirmed all isolates as *B. canis* with high clonality (only seven SNPs). Phylogenetic analysis using ParSNP placed the isolates in a distinct clade with other South American *B. canis* strains, including human isolates, suggesting a shared origin and endemic circulation. Analysis of SNPs in coding regions revealed potential phenotypic variations that warrant further research.

This study demonstrates the superior performance of MALDI-TOF MS, especially with an optimized in-house database, compared to traditional methods for rapid and accurate *B. canis* identification. The high concordance between MALDI-TOF MS and whole-genome sequencing validates the use of MALDI-TOF MS as a primary diagnostic tool. The identification of a clonal *B. canis* outbreak underscores the importance of rapid diagnostics in controlling disease spread. The phylogenetic clustering with other South American strains highlights the endemic nature of canine brucellosis in the region and potential cross-border transmission risks. The discovery of SNPs affecting coding regions may explain diverse clinical manifestations observed in dogs with seemingly identical strains, and might provide insights into virulence factors. The high rate of asymptomatic infection further indicates the need for proactive surveillance programs in at-risk areas. This emphasizes the zoonotic potential of *B. canis* and necessitates effective prevention and control strategies.

This study demonstrates the combined power of MALDI-TOF MS and whole-genome sequencing for rapid diagnosis and epidemiological investigation of canine brucellosis outbreaks. MALDI-TOF MS provided quick and accurate *B. canis* identification, distinguishing it from *B. suis*. Genomic analysis revealed clonal spread within the kennel and linked the outbreak to a South American *B. canis* sub-lineage. Future studies should focus on the functional implications of identified SNPs and investigation into the potential virulence differences associated with these SNPs and the broader implications of this endemic situation for public health.

While this study provided valuable insights, the sample size was limited to one kennel. The study's findings might not be fully generalizable to other geographical regions or dog populations. Furthermore, the limited sequencing coverage (5x for certain isolates) may have affected the SNP calling accuracy. Though this was mitigated by using concatenated reads from four high-quality samples to create the consensus genome, future research using longer read sequencing technologies would benefit further analyses.