Bacterial communities associated to the urethra of healthy gilts and pregnant sows undergoing different reproductive protocols

Abstract

Nowadays, it is known that the urogenital microbiota plays a key role in the urinary health of mammalians. Despite the urinary infections affect the health and the welfare of breeding sows, the urethral microbiota of healthy sows remains unknown. Therefore, this work evaluates the urethral bacterial communities of healthy gilts and sows to determine the presence of Enterobacteriaceae populations, and the structure of this microbiota in gilts (G) and pregnant (P) sows. Samples were collected by scraping the urethral mucosa of G (n = 9) and P sows, which included natural mating (NM, n = 9) and artificial inseminated (AI, n = 7) sows. Samples were analyzed by culture-dependent techniques and 16S-rRNA gene high-throughput-sequencing. All females were positive for Enterobacteriaceae culture, without significant differences (Kruskal–Wallis) between G and P groups (median values: 2.78 and 3.09 log CFU/mL, respectively; P = 0.497). Also, the rate of Enterobacteriaceae/total mesophilic microorganisms was individually calculated, without significant differences between G and P sows (median values: 0.61 and 0.66, respectively; P = 0.497). When analyzing the bacterial communities, it was found similar richness in G, NM, and AI; however, diversity was lower in P sows than G (Mann Whitney/Kruskal–Wallis test, P < 0.01). The dominating phyla that constituted a “core microbiome” included Firmicutes, Proteobacteria, Cyanobacteria, Actinobacteria, and Bacteroidetes, which were common for all the studied females. The relative abundance for phyla, families, and genera was estimated, and Firmicutes was significantly higher in NM than AI sows (P = 0.02, Mann–Whitney/Kruskal Wallis test for univariate statistical comparisons); Pseudomonadaceae and Enterobacteriaceae were higher in AI than in NM (Mann–Whitney/Kruskal–Wallis, P < 0.05). Lactobacillus and Pseudomonas were among the dominant genera; however, only Pseudomonas sp. was significantly higher in AI than NM (Mann–Whitney/Kruskal–Wallis, P = 0.006). The results represent the first evidence about the existence of a urethral microbiota that includes Enterobacteriaceae, as well as the patterns of this microbiota in G and P sows. The knowledge of this urethral microbiota might allow for future research to develop innovative protocols to restore and/or preserve the healthy ecology of the urinary microbiome to prevent diseases ensuring the welfare of breeding sows.

Introduction

The microbiome is defined as the genome of all the microorganisms, symbiotic and pathogenic, living in and on the body of the vertebrates (Berg et al., 2020). The microbiota comprises all living members forming the microbiome, which means the living organisms of an ecosystem or a particular area (Berg et al., 2020). Thus, the mucosal surfaces of humans and animals are colonized by the communities of commensal, symbiotic, and pathogenic microorganisms (Proctor, 2019). The interactions between this commensal microbiota and the host influence their physiology, thereby regulating metabolism and immune function, as well as their complex behaviors (Lynch and Hsiao, 2019). Several studies concluded that the structure of the bacterial communities in the urinary tract (UT) could have an important effect on the host’s health (Horwitz et al., 2015; Whiteside et al., 2015; Thomas-White et al., 2016; Bao et al., 2017; Brubaker and Wolfe, 2017). However, the urinary microbiome of sows has been unexplored, and the patterns of their microbiota in gilts (G) and pregnant (P) sows remain unknown up to date.

The urogenital health of G and sows is determinant for the reproductive performance, which is a key factor for productivity in the pig farming (Koketsu et al., 2017). The urinary tract infections (UTI) are a common problem in breeding sows, which are responsible for reducing animal welfare, decreasing productivity, and resulting in a premature culling (Wanyoike and Bilkei, 2006; Stalder et al., 2012; Drolet, 2019). Escherichia coli and Proteus spp. belong to the Enterobacteriaceae family and are recognized pathogens of these urinary infections (Moreno et al., 2018; Drolet, 2019). Nevertheless, it not clear if Enterobacteriaceae are part of healthy urinary microbiota and, thus, some species could be potential pathogens. Therefore, the aim of this work was to evaluate the microbial communities of the urethral mucosa in healthy G and P sows and to determine if pregnancy drives changes in the autochthonous microbiota. Thus, the urethral microbial ecology of G and P sows was examined by high-throughput sequencing approach based on Illumina MiSeq sequencing of the V3-V4 16S rRNA and culture-dependent methods focusing the study on the Enterobacteriaceae population. The knowledge of the patterns of the urethral microbiota in pregnancy will allow for future research on the urinary microbiome and might promote the development of innovative therapeutic strategies to prevent diseases ensuring the welfare of breeding sows.

Materials and Methods

Animals and sampling

Twenty five contemporary healthy females (Duroc × [Landrace × Yorkshire]) were sampled: 9 G (body weight 121.4 ± 6.3 kg [average ± SD], age 7 ± 1 mo [average ± SD]) and 16 P sows (average body weight 223.4 ± 12.5 kg [average ± SD], age 18.6 ± 5.6 mo [average ± SD], gestation 60 ± 5 d). They were group-housed in pens (250 m² per female) according to the category (G or P sows). G and P sows had free access to water and received standard gestation feed: 74% corn, 23% soybean expeller, and 3% premix for gestation (Vetifarma S.A., Buenos Aires, Argentina). The G expressed two estrous cycles before sampling. Sows: pregnancy has been achieved by natural mating (NM; by hand-mating system; n = 9) in two nulliparous, three primiparous, and four multiparous sows (with three to four previous farrowing). Seven of the P sows were artificial inseminated (AI); all of them were multiparous sows (two to three previous farrowing).

The sampling was conducted at outdoor pig farm located in Leales, Tucumán, Argentina (27°12′54.1″S and 65°15′15.8″W) during autumn (May 2018, AI group) and winter (July 2018, G and NM groups).

For samples collection, perineum and vulvar areas were washed with sterilized water and dried by using paper towels. Then, stainless steel specula were placed to access the meatus, and cytobrushes were used to scrape the urethral wall approximately at the internal urethral orifice level. Finally, cytobrushes were put in 1 mL phosphate-buffered saline (PBS) solution-containing tubes, pH 7.0, and kept at 4 °C until processing. All procedures were conducted under the Argentinean Animal Welfare Legislation, Law N°14.346, SENASA-R70/2001, with the approval of the Institutional Committee for the Care and Use of Laboratory Animals of the National University of Tucumán (CICUAL–UNT, Research Protocol N° 030/2019).

Microbial populations: culture-dependent methods

Enterobacteriaceae population

The tubes with cytobrushes and PBS were vigorously agitated for 2 min to dislodge cells. Then, 50 µL of pure and 0.01 dilution of each sample were inoculated on plates containing LAPTg agar (in g/L: peptone, 15; tryptone, 10; yeast extract, 10; d-glucose, 10; agar, 15; Raibaud et al., 1963), Columbia agar supplemented with 5% sheep blood (Britania Laboratories, Buenos Aires, Argentina), and MacConkey agar (Britania Laboratories, Buenos Aires, Argentina). Plates were incubated for 24 to 48 h at 37 °C in aerobic conditions, with the exception of Columbia agar plates, which were incubated in microaerophilic conditions (5% CO₂-enriched chamber). After incubation, the colonies grown on MacConkey plates were evaluated by morphology and Gram staining. The number of viable microorganisms, expressed as colony-forming units per milliliter (CFU/mL), was determined to quantify the cultivable microbial populations of mesophilic microorganisms (LAPTg and Columbia plates) and Enterobacteriaceae (MacConkey plates).

Microbial populations: culture-independent techniques

Nucleic acid extraction and amplifications

The deoxyribonucleic acid (DNA) of the samples was extracted using QIAGEN kits (QIAamp DNA mini kit, Hilden, Germany) according to the manufacturer’s instructions. Quantification and integrity were checked before amplification reactions and stored at −20 °C. The bacterial V3-V4 16S rRNA region was amplified with the primer pairs 343F (5′-TACGGRAGGCAGCAG-3′) and 802R (5′-TACNVGGGTWTCTAATCC-3′) using Phusion Flash High-Fidelity MasterMix (Thermo Fisher Scientific, Inc. Waltham, MA, USA). A two-step nested polymerase chain reaction (PCR) was applied, and conditions used for reaction mix and amplification experiments were those described by Vasileiadis et al. (2015). In the second PCR, the 343F primer was labeled with a different “barcode” for each sample. The PCR products from all samples were joined in a single pool in equimolar concentrations based on the QuBit quantification data and were concomitantly purified by solid-phase reversible immobilization using the Agencourt AMPure XP kit (Beckman Coulter, Milano, Italy). The PCR product pool was sequenced by PTP—Science Park (Parco Tecnologico Padano, Lodi, Italy) using a MiSeq Illumina Reagent Kit v3 (Illumina Inc., San Diego, CA, USA), which generated 300 bp paired-end reads.

Data processing and bioinformatics analysis

Quality check from raw reads was performed using FastQC v0.11.2 (Babraham Bioinformatics, Cambridge, UK). Samples were demultiplexed using ea-utils v.1.1.2-537 fastq-multx (Aronesty, 2013) relying on a metadata file provided by the customer. Illumina raw sequences were trimmed using Trimmomatic v0.32 (Bolger et al., 2014). Minimum base quality 20 (Phred-scale) over a 4-bases sliding window was required. Only sequences above 36 nucleotides in length were included in downstream analysis. For original amplicon reconstruction, overlapping R1 and R2 paired reads were joined using ea-utils v.1.1.2-537 fastq-join tool (Aronesty, 2013). Nonoverlapping R1 and R2 paired reads were concatenated using one “N” base separator. Amplicons were dereplicated, sorted, and clustered at 97% identity using VSEARCH v1.1.3 (Rognes et al., 2016) following the standard Quantitative Insight into Microbial Ecology (QIIME; Caporaso et al., 2010) pipeline parameters. For taxonomy-based analyses, QIIME-formatted Greengenes v.13.8 database was used. Taxonomies were adapted to QIIME taxonomy standards uniforming to the seven main taxa ranks (superkingdom, phylum, class, order, family, genus, and species). The operational taxonomic units (OTU) were identified against reference databases (Greengenes v.13.8 database) using NCBI-Blast v2.2.27 (Basic Local Alignment Search Tool of National Center for Biotechnology Information online website). After counting the abundance of each OTU, a final OTU table output file was created using custom scripts.

Statistics

Data from the bacterial cultures (logarithmically transformed) were tested for normality and homoscedasticity. Then, a nonparametric test (Kruskal–Wallis test) was applied to compare G and P groups. Minitab Statistical Software version 15.1.20.0 (Minitab. LLC. State College, PA, USA) was used for this analysis.

To analyze DNA read mapping two indexes, Chao’s wealth and Shannon’s diversity and a principal coordinate analysis (PCoA) were performed using the QIIME package, version 1.5.0 in the pipeline Microbiome Analyst (http://microbiomeanalyst.ca/faces/home.xhtml). Past (Paleontology Statistics) software version 3.23 (Hammer et al., 2001) was used to perform the Mann–Whitney/Kruskal–Wallis test for the comparison of the relative abundance of OTU among the groups of females.

Results

Urethral microbiota: Enterobacteriaceae population (studies based on cultures)

The magnitude of the urethral colonization by Enterobacteriaceae was assessed by culture-dependent techniques using a selective medium. It is interesting to point out that all females were positive for this culture, and there were no significant differences (Kruskal–Wallis test) between G and P groups (median values: 2.78 and 3.09 log CFU/mL, respectively; H = 0.46; df = 1; P = 0.497). As an estimation of the overall colonization, the data were also analyzed taking into account the total mesophilic microorganisms detected in each sample; thus, the rate of Enterobacteriaceae/total mesophilic microorganisms (E/M) was individually calculated, and no differences were found between G and P groups (median values: 0.61 and 0.66, respectively; H = 0.46; df = 1; P = 0.497).

Urethral microbiota: 16S metagenomics approach

The microbial diversity was measured using the Shannon and Chao1 indices, which evaluate abundance (number of different species) and homogeneity (Shannon, 1997) and richness, respectively. When comparing the index values of G and P groups, significant differences were observed between the Shannon index, but not between the Chao1 index, estimated for each group (Mann–Whitney/Kruskal–Wallis test, P < 0.01; Figure 1A). Therefore, the urethral microbiota from P sows had lower diversity than G, but all urethral samples had a similar richness. Moreover, a PCoA based on the β-diversity/Bray–Curtis was performed to evaluate the differences between the bacterial communities associated to each group. Thus, there was no significant separation or distinct clustering (PERMANOVA, P > 0.01) in the taxonomic composition of the urethral microbiota in G and P sows (Figure 1B).

Figure 1.

Box-plot showing α-diversity in samples using Chao1 and Shannon index in samples from the urethra of G and P sows. Mean (♦), median, and quartile range are shown (A). *Indicates significant differences between gilts (G) and pregnant sows (P) (P < 0.01; Mann–Whitney/Kruskal–Wallis test). PCoA based on Bray–Curtis ß-diversity showed no clear distinct clustering of the G (pink) and P sows (blue) (B).

Considering only the P group and comparing the estimators for the urethral microbiota from AI and NM, no significant differences were detected (Shannon and Chao1 indexes, Mann–Whitney/Kruskal–Wallis test, P > 0.05; Figure 2A). However, when evaluating β-diversity based on Bray–Curtis, significant differences were observed (PERMANOVA, P < 0.02) between the microbial communities’ structures from AI and NM sows, although the data were partially overlaid (Figure 2B).

Figure 2.

Box-plot showing α-diversity in samples using Chao1 and Shannon index in samples from the urethra of pregnant (P) sows by NM or AI (A). Mean (♦), median, and quartile range are shown. PCoA based on Bray–Curtis ß-diversity showed partially overlaid distinct clustering of the pregnant sows: AI (red) and NM (blue) (PERMANOVA, P < 0.02) (B).

Structure of the microbial communities

Nineteen phyla were found in the porcine urethral microbiota. The bacterial taxa with the highest relative abundances were Firmicutes (37%), Proteobacteria (26%), Actinobacteria (12%), Cyanobacteria (8%), Fusobacteria (8%), Bacteroidetes (6%), Acidobacteria (1%), and Thermi (1%) (Figure 3A). The remaining 11 phyla were represented by less than 1% of the total sequence reads.

Figure 3.

Composition of the bacterial urethral microbiota of sows. Contribution of the most abundant phyla (A) and orders (B).

The urethral core microbiome defined as the group of phyla present in 90% of the samples (Lorenzen et al., 2015) was constituted by Firmicutes, Proteobacteria, Cyanobacteria, Actinobacteria, and Bacteroidetes. Moreover, sequences from Fusobacteria, Acidobacteria, and Thermi (Deinococcus–Thermus) were detected in 17, 16, and 14 from a total of 25 samples, respectively. The remaining 11 phyla defined in our pooled urethral sequence dataset were present in ≤10 samples; among them, Verrucomicrobia, Gemmatimonadetes, Spirochaetes, Nitrospirae, and Planctomycetes were detected in 10/25, 6/25, 6/25, 5/25, and 5/25 samples, respectively.

Taxonomical assignment at the bacterial order level resulted in 52 taxa; however, only 14 showed a relative abundance > 1%, with Clostridiales, Actinomycetales, Lactobacillales, Fusobacteriales, Enterobacteriales, and Pseudomonadales, being the most abundant (≥7%; Figure 3B). In total, 131 OTU families were identified; those with ≥0.5% relative abundance and present in half of the animals at least in one group (G or P) or in half of the total animals were included in Table 1. From this group, the most prevalent (>5%) were Tissierellaceae, Fusobacteriaceae, Clostridiaceae, Enterobacteriaceae, and Streptococcaceae. The sequences that could be assigned at the genus level and that were present in >90% of the samples were identified as Lactobacillus, Pseudomonas, Rhodoplanes, Enterococcus, and unclassified OTUs derived from Clostridiaceae, Micrococcaceae, and Bradyrhizobiaceae families (Supplementary Figure S1).

Table 1.

Distribution of most prevalent family OTUs in the urethral microbiota of sows

Family OTU1	Mean abundance, %	No. of samples/25
Tissierellaceae	10.13	22
Fusobacteriaceae	8.23	22
Clostridiaceae	7.83	24
Enterobacteriaceae	6.59	24
Streptococcaceae	5.33	20
Lachnospiraceae	4.05	24
Micrococcaceae	3.92	25
Ruminococcaceae	3.79	24
Pasteurellaceae	3.44	21
Moraxellaceae	3.39	22
Lactobacillaceae	2.36	24
Sphingomonadaceae	1.96	22
Pseudomonadaceae	1.87	24
Peptostreptococcaceae	1.72	18
Flavobacteriaceae	1.6	10
Campylobacteraceae	1.49	17
Bacillaceae	1.42	22
Jonesiaceae	1.36	11
Actinomycetaceae	1.31	18
Hyphomicrobiaceae	1.24	24
Bradyrhizobiaceae	1.14	23
Cellulomonadaceae	1.08	14
Porphyromonadaceae	0.98	22
Intrasporangiaceae	0.95	19
Veillonellaceae	0.94	20
Aeromonadaceae	0.94	14
Bacteroidales2	0.88	19
Aerococcaceae	0.79	15
Burkholderiaceae	0.79	17
Nocardioidaceae	0.74	21
Bacteroidaceae	0.67	15
Enterococcaceae	0.64	24
Deinococcaceae	0.61	19
Rhodobacteraceae	0.61	15
Caulobacteraceae	0.6	21
Rhizobiaceae	0.59	20
Rhodospirillaceae	0.58	17
Comamonadaceae	0.55	22
Staphylococcaceae	0.51	18
Gaiellaceae	0.5	21

¹OTUs with ≥0.5% relative abundance.

²Family not identified from order Bacteroidales.

Comparison of the urethral microbiota in G and P sows (AI and NM)

Overall, no significant differences were found for bacterial relative abundance with respect to the dominant phyla between G and P sows (Mann–Whitney/Kruskal–Wallis, P > 0.05), with the exception of Thermi, which was significantly higher in P sows than G (Mann–Whitney/Kruskal–Wallis, P < 0.05; Figure 4A). Considering the most abundant families (relative abundance ≥ 0.5%), no differences were found between G and P groups (Figure 4B).

Figure 4.

Urethral microbiota of healthy sows. (A) The relative abundance of major bacterial phyla found in the sequence pool of urethral samples from G and P sows. “Other” represents minor groups. (B) The relative abundance of the most prevalent families found in urethral samples from gilts (G) and pregnant (P) groups. *Indicates significant differences between G and P groups (P < 0.05; Mann–Whitney/Kruskal–Wallis test).

The relative abundances of phyla present in both AI and NM P sows are shown in Figure 5A. Proteobacteria and Firmicutes were the most abundant in both groups; the relative abundances of Proteobacteria were 51% and 25% for AI and NM, respectively; however, no significant differences were detected. Conversely, the relative abundance of Firmicutes was significantly higher in NM (39%) than AI (17%; P = 0.02, Mann–Whitney/Kruskal–Wallis test for univariate statistical comparisons). Among the remaining phyla, only Cyanobacteria showed a significant difference between both groups, being higher in AI sows (P = 0.04; Figure 5A).

Figure 5.

Urethral microbiota in pregnant (P) sows. Relative abundance profiles at phylum (A), family (B), and (C) genus (or the lowest common taxon) levels obtained from the sequence classification of the 16S rRNA gene. *Indicates significant differences between AI and NM groups (P < 0.05; Mann–Whitney/Kruskal–Wallis test).

At the family level, the relative abundance of Jonesiaceae, Streptococcaceae, Flavobacteriaceae, and Peptostreptococcaceae was significantly higher in NM than in AI sows, while Pseudomonadaceae and Enterobacteriaceae were significantly higher in AI than NM (Mann–Whitney/Kruskal–Wallis, P < 0.05; Figure 5B).

When analyzing the genera with at least 0.5% relative abundance, Pseudomonas was significantly (Mann–Whitney/Kruskal–Wallis, P = 0.006) higher in AI (0.176 ± 0.06) than NM sows (0.04 ± 0.04), while Streptococcus was most abundant (Mann–Whitney/Kruskal–Wallis, P = 0.02) in NM than AI sows (0.08 ± 0.02 and 0.007 ± 0.003, respectively; Figure 5C).

Discussion

For the first time, the present study provides evidences of the existence of a urethral microbiota in healthy G and P sows. Also, through culture-dependent techniques, it was possible to demonstrate that this microbiota includes Enterobacteriaceae populations. Moreover, next-generation sequencing allowed to describe the patterns of this microbiota for G and P sows, to define an urinary “core microbiome”, and to describe the bacterial communities at family/genus levels in NM and AI sows.

A careful sampling of the mucosal surfaces (by scraping), instead of a microbial recovery from urine (Gusmara et al., 2011; Moreno et al., 2018), ensured that the collected microorganisms were those colonizing the urethra and free of fecal contamination.

It is worth highlighting that all sows and gilts were positive for the Enterobacteriaceae culture. Several genera belonging to this taxon represent recognized pathogens for urinary infections in sows (Gusmara et al., 2011; Moreno et al., 2018; Drolet, 2019); therefore, their presence in the microbiota of healthy female could mean that some of them have a potential pathogenic role.

The metagenomic approaches allowed us to identify the complex microbial communities colonizing the urethral mucosa of G and P sows, the richness of these communities being similar in both groups, regardless of how the pregnancy was achieved. However, the Shannon predictor indicated a low diversity in P sows. Probably, physiological conditions induce a loss of diversity that naturally characterizes the porcine urethral microbiota as observed in the G group. The changes in the native urethral microbiota during pregnancy must be further studied to determine if they are responsible for dysbiosis and increased susceptibility to infections during the gestation and after farrowing (Fangman and Carlson Shannon, 2007; Baricco, 2011). In this sense, it was demonstrated that a loss of diversity in the urinary microbiota in women predisposed to developing UTI (Brubaker et al., 2014; Brubaker and Wolfe, 2017).

Likewise, we evaluated and compared the structure of urethral microbial communities by β-diversity analysis between G and P sows (NM and AI) groups. This analysis indicated compositional differences between AI and NM groups. Further research with a higher number of animals must be carried out to evaluate these differences and to identify if populations of potential urogenital pathogens are implicated.

The analysis of the taxonomic distribution of the bacterial communities indicated that Firmicutes, Proteobacteria, and Bacteroidetes were the most prevalent phyla associated to the porcine urethral microbiota. Regarding the porcine urogenital tract, only the vaginal microbiota has been previously described (Wang et al., 2017), and these same phyla were present as the main constituents in both healthy and endometritic adult sows. In our study, the occurrence and dominance of these three phyla together with Cyanobacteria and Fusobacterium could indicate that they constitute the microbial nucleus that colonizes the urethra of G and P sows (without significant differences between both groups); therefore, the dynamics of colonization under physiological conditions would probably take place among the populations of these taxonomic groups. These key issues were also addressed in the studies of the urinary microbiota in pregnant (Ollberding et al., 2016) and nonpregnant women (Siddiqui et al., 2011; Wolfe et al., 2012; Lewis et al., 2013).

In this work, the occurrence of Enterobacteriaceae as commensal microorganisms of the porcine urethra was detected by using both culture-dependent and culture-independent techniques. To our knowledge, there are few reports available describing Enterobacteriaceae as members of the urethral microbiota in healthy females; in this sense, Ollberding et al. (2016) reported Serratia sp. in urine from pregnant women.

Our analysis showed that the largest numbers of sequence reads belonged to the Enterobacteriaceae in AI sows. This subject should be further studied since the gestation implies high susceptibility to infections; particularly, the colonization by E. coli, Proteus spp., or Klebsiella spp. must be taken into consideration because they were related to urogenital tract infections in sows during the postpartum period (Fangman and Carlson Shannon, 2007; Baricco, 2011; Gusmara et al., 2011; Moreno et al., 2018; Drolet, 2019).

Among the predominant OTUs, Pseudomonas stands out as one of the most prevalent (>90% of the urethral samples) and associated to P group. Its presence should be troubling since this genus has been described as an agent of endometritic and vaginal discharges in sows (Torremorrell, 2007). If the urethral microbiota is part of a “urogenital microbiome” (Burton et al., 2017), we could hypothesize that the increase of urethral population of Pseudomonas among the P sows could represent a risk for the genital health.

The Firmicutes members, Lactobacillus and Streptococcus, detected in this study may have a putative antagonistic role in the UT. To our knowledge, this is the first report on the presence of urethral Lactobacillus in healthy sows; previously, they have been described in the porcine vagina (Lorenzen et al., 2015; Wang et al., 2017). This genus of lactic acid bacteria was widely studied because of its beneficial effects on the human UT (Siddiqui et al., 2011; Hilt et al., 2014; Jacobs et al., 2017). Therefore, we can assume that the presence of these bacterial populations in the anterior urethra of sows would play a protective role against potential pathogenic microorganisms, such as members of Enterobacteriaceae. Further research is necessary to determine the potential protective role of urethral Lactobacillus, especially after farrowing, a stress condition that increases the infection risks in sows (Fangman and Amass, 2007; Falceto et al., 2012).

In this study, Streptococcus appears mostly associated to the urethra of P sows. This finding should be considered in futures studies because they might enter the vagina from the urethra and colonize the birth canal; this would represent a risk since some Streptococcus species are pathogens in newborn piglets (Gottschalk and Segura, 2019).

The results presented in this work contribute to the knowledge regarding the urinary ecosystem and their bacterial communities in healthy G and P sows.

Supplementary Data

Supplementary data are available at Journal of Animal Science online.

Supplementary Figure S1. Stacked bar charts showing the relative abundance of microbial DNA detected via 16S rRNA amplicon sequencing and annotated to the genus or the lowest common taxon, in samples from healthy G and P sows.

Glossary

Abbreviations

AI

pregnant sows by artificial insemination (artificially inseminated)

DNA

deoxyribonucleic acid

OTU

operational taxonomic unit

P

pregnant

PBS

phosphate-buffered saline solution

PCoA

principal coordinate analysis

PCR

polymerase chain reaction

QIIME

Quantitative Insight into Microbial Ecology

RNA

ribonucleic acid

rRNA

ribosomal RNA

UT

urinary tract

UTI

urinary tract infection

Conflict of interest statement

All authors declare no conflict of interest.