The samples used in this study were collected as part of the service provided by the Federal University of Minas Gerais (UFMG) to the Municipal Park and Zoobotanical Foundation of Belo Horizonte (FPMZB), under extension code Siex-UFMG 302557. All procedures related to animal handling, including capture and restraint, were conducted by veterinarians from the FPMZB staff, ensuring adherence to the protocols established for extension activities by UFMG.

The present study was conducted at the FPMZB, which covers a total area of 10.7 million square meters. This foundation hosts, within its Zoo, more than 3,500 individuals, representing over 235 species, with approximately 36 species of mammals totaling 117 individuals. Among these species, over 40 are at risk of extinction, spanning reptiles, birds, fish, amphibians, and mammals from all five continents. It's worth noting that the Zoo's infrastructure includes a veterinary hospital [14].

FPMZB has a well-established history of preserving biodiversity and managing environmental protection areas in Belo Horizonte (MG) (-19.857889841, -44.0075113912). Its foundation was established through Decree 16,684 on August 31, 2017, which merged the former Municipal Parks Foundations and the Zoo-Botanical Foundation. In the context of Belo Horizonte, FPMZB plays a significant role in the conservation of local ecosystems and the research of various mammal species present in its parks and botanical gardens. Remarkably, these spaces house a variety of over a thousand species, representing the Cerrado and Atlantic Forest, biomes found in different regions of Brazil [15].

The relevance of FPMZB as a research site lies in the diversity of mammal species found in its parks, including the marsh deer (Blastocerus dichotomus), the black tufted marmoset (Callithrix penicillata), and the maned wolf (Chrysocyon brachyurus), which may serve as important potential reservoirs for zoonotic pathogens, such as SARS-CoV-2 [16]. Given the global concern about interspecies transmission of the virus between humans and animals, the study investigated the role of these species in the transmission chain and sought a better understanding of the pandemic's dynamics to develop effective prevention strategies. FPMZB is the second-largest public green area in Belo Horizonte and receives over one million visitors annually.

Oropharyngeal, rectal, and nasal swabs were collected from 47 captive animals of the class Mammalia between November 2021 to March 2023, with the aim of investigating SARS-CoV-2 infection in captive animals with no prior knowledge of virus exposure.

Capture and containment of the animals were carried out by the veterinary team at the FPMZB Zoo, who possess the necessary experience and knowledge to ensure the well-being and safety of the animals during the collection process, which was done after the physical and/or chemical restraint of the animals. This was done in accordance with the zoo's management schedule, and whenever containment was necessary for obtaining biological samples.

Laboratories responsible for the analysis in this study provided 1.5 ml Eppendorf tubes containing DNA/RNA Shield™ (2X Concentrate) 125 ml (ZymoResearch) for the collection of animal samples. Each collected sample was properly identified with the animal's microchip number, ensuring data traceability. After collection, the samples were stored in transport containers and subsequently in -80 °C freezers at the Laboratory of Integrative Biology (ICB/UFMG) to maintain their integrity and viability.

Viral RNA was extracted from swab samples (rectal, oropharyngeal, and nasal) using the PureLink™ RNA Mini Kit (Invitrogen™, MA, USA), following the manufacturer's protocol.

RNA samples were tested for the presence of SARS-CoV-2 using the CDC 2019-Novel Coronavirus Real-Time RT-PCR Diagnostic Panel (N1, N2) [17]. These tests were performed using the iTaq™ Universal Probes One-Step (Bio-RAD™, CA, USA), following the manufacturer's instructions. The diagnosis was carried out by RT-PCR following the CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-PCR Diagnostic Panel protocol.

The samples underwent a concentration and purification process using the RNA Clean & Concentrator™-5 kit (Zymo Research™, CA, USA) to concentrate the amount of viral RNA and eliminate potential contaminants. The cDNA was synthesized using the SuperScript™ IV First-Strand Synthesis System (Thermo Fisher Scientific™, MA, USA).

Sequencing libraries were prepared using the xGen™ SARSCoV-2 Amplicon Panel xGen SARS-CoV-2 S Gene Amplicon Panel (Integrated DNA Technologies™, San Diego, CA, USA) following the manufacturer's instructions. Multiplex PCR was used to enrich the target sequences, as the samples had a Ct > 30.

After library preparation, samples were quantified using the Invitrogen™ Qubit™ 4 Fluorometer (Thermo Fisher Scientific™, MA, USA), and 10 samples were selected to continue the sequencing process. The genomes were sequenced on the MiSeq platform (Illumina™, San Diego, CA, USA) with v3 cartridges (600 cycles) following the manufacturer's instructions. After sequencing, the raw data were used to a custom pipeline optimized for viral genome assembly, following the same protocol described in Moreira et al., 2021 [18].

Briefly, the reads were filtered by quality Phred > 30, the remaining reads were aligned to the reference genome of SARS-CoV-2 (GenBank accession number NC_045512.2) using Bowtie2 [19], and consensus sequences were generated using BEDtools [20] and BCFtools [21]. Sequences that had the expected sequencing quality parameters (coverage greater than or equal to 60% of the genome) were used in our study.

The consensus genomes generated were classified using the Pangolin tools [22] and the NextClade web application v2.8.1 [23]. After classification, we constructed a dataset based on human sequences for each of the VOCs (Alpha—190, Beta—13, Zeta—150, Gamma—40, Delta—116, and Omicron—150) from Belo Horizonte (MG) and animal genomes (n = 220; identified all around the world) publicly available in public databases, plus the genomes generated in the study (n = 03). The dataset was aligned using the minimap2 program v.2.2.24 [24] and then used to infer a maximum likelihood phylogeny using the program IQ-Tree v2.0.3 and the evolutionary model GTR + F + I + G4 [25]. Shimodaira-Hasegawa-like approximate likelihood ratio test (SH-aLRT) was used to calculate the branch uncertainty in the phylogeny.