Culturomic-, metagenomic-, and transcriptomic-based characterization of commensal lactic acid bacteria isolated from domestic dogs using Caenorhabditis elegans as a model for aging

Abstract

In tandem with the fast expansion of the pet-economy industry, the present aging research has been noticing the function of probiotics in extending the healthy lifetime of domestic animals. In this study, we aimed to understand the bacterial compositions of canine feces and isolating lactic acid bacteria (LAB) as commensal LAB as novel potential probiotics for the use of antiaging using Caenorhabditis elegans surrogate animal model. Under an anaerobic, culturomic, and metagenomic analysis, a total of 305 commensal LAB were isolated from diverse domestic dogs, and four strains, Lactobacillus amylolyticus, L. salivarius, Enterococcus hirae, and E. faecium, made prominence as commensal LAB by enhancing C. elegans life span and restored neuronal degeneration induced by aging by upregulating skn-1, ser-7, and odr-3, 7, 10. Importantly, whole transcriptome results and integrative network analysis revealed extensive mRNA encoding protein domains and functional pathways of naturally aging C. elegans were examined and we built the gene informatics basis. Taken together, our findings proposed that a specific gene network corresponding to the pathways differentially expressed during the aging and selected commensal LAB as potential probiotic strains could be provided beneficial effects in the aging of domestic animals by modulating the dynamics of gut microbiota.

The culturomic-, metagenomic-, and transcriptomic-based results suggested that a specific gene network corresponding to the pathways differentially expressed during the aging of Caenorhabditis elegans and selected commensal lactic acid bacteria as newly isolated potential probiotic strains could be provided health-promoting benefits in the aging of domestic animals by modulating dynamics of gut microbiome.

Introduction

Maintaining one’s health and avoiding sickness are obviously vital components of contemporary living. As the average longevity of humans has increased during the past centuries, so too have companion animals been confronted with many chronic morbidities. In contrast to experimental animal models such as mice and rats, companion animals tend to have a longer lifespan and share comparable living situations with people, making them susceptible to the same “naturally occurring” diseases as humans (Hoffman et al., 2018; Kim et al., 2021).

In recent years, the efficient functioning of the complex gastrointestinal system, taken most parts by gut microbiome, has been considered a critical element in sustaining well-being. In this response, the microbiome, compounding microbiota and genome, has been noticed as the modulator of the whole body network, not limited to humans but it has been expanded to companion animals, too (Adams and Gutierrez, 2018; Rinkinen and Beasley, 2019)^, (Chun et al., 2020). With the fast expansion of the pet-economy industry, the current field of aging research has highlighted the function of probiotics with commensal lactic acid bacteria (LAB) in extending the healthy lives of domestic animals. Recent efforts have been made to identify commensal LAB, such as the Lactobacilli and Enterococci genera, which are frequently used as probiotics for dogs and cats and have been shown to have increased health-promoting effects (Lee et al., 2022). But their functional mechanism in companion animals is still unclear.

The gastrointestinal tract is responsible for a major portion of animal health (Oh et al., 2021; Mun et al., 2021b). The living microbial ecosystem works on absorption as well as nutritional metabolism, and the trophic and defensive activities against exotic hazards make the use of probiotics in animal feeds rather intriguing to the scientific field and pet market. As the probiotics could benefit the animal’s gut microbiome and general health, now people desire the aging companion animals to be reduced in deterioration of physical strength and to achieve healthy aging. However, a limited number of culturomic and metagenomic studies have deciphered the microbiome of companion animals. As mentioned previously, companion animals would be an even better animal model for studying the human microbiome than pigs or mice because they not only share very similar habitat conditions with humans but also have a more similar gut microbiome. Nevertheless, the low number of research restricts the development and implementation of probiotics for companion animals with varying applications^, (Chun et al., 2020; Lee et al., 2022).

In this study, our research focused on two main objects for exploring the beneficial role of commensal LAB-isolated domestic dogs. The first is to explore the aging mechanism of organisms by using the mRNAs of aging Caenorhabditis elegans and developing protein and genetic networks that correspond to the pathways that changed throughout the aging process. The C. elegans has been recognized as the premier model organism in the study of aging. Aside from its similarity to the human genome, C. elegans exhibits evident age-dependent metabolic changes and a reduced life span as compared to other animal models, and its genetic tractability makes it an excellent model for aging research (Olsen et al., 2006; Son et al., 2019). The other intended object was to explore the potential probiotics that are not commercially available yet from young and senior ones using in vitro fermentation system. Fermentation of the Intestinal Microbiota Model (FIMM) is an in vivo intestine mimicking cultivation of the microbiome of living things such as domestic animals and humans. This system has been proven to be a very useful model since it stimulates the actual condition of the living intestine with the adjusted pH, temperature, and resistance time (Sivieri et al., 2013; Van de Wiele et al., 2015; Pérez-Burillo et al., 2021). Through this research, we propose elucidating the mechanism and functions and broadening the selection of commensal LAB that can be applied in the future companion animal market by clarifying their purpose and function.

Materials and Methods

Metagenomic analysis of canine fecal microbiota

The research protocol for the present experiment was approved by the Institutional Animal Care and the Use Committee (IACUC) at Chungnam National University (202109-CNU-104). In order to analyze the culture-independent microbial composition differences between the young (6–8 yr old; n = 6) and the senior (9–12 yr old; n = 6) domestic dogs including mongrel, Poodle, Maltese, Beagle, and Jindo dogs, the fecal samples were collected, aseptically homogenized, and gDNA was extracted with the DNeasy PowerSoil Pro Kit (Qiagen, Hilden, Germany). The V4 region of the 16S rRNA genes was amplified (V4 amplicon primer sequences: forward, 5’-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGGTGCCAGCMGCCGCGGTAA-3’; reverse, 5’-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGGACTACHVGGGTWTCTAAT-3’), and the amplified DNA was sequenced using Illumina® iSeq 100 (Illumina, Inc. San Diego, CA, USA) following the manufacturer’s manuscripts (Kazantseva et al., 2021).

Culturomic analysis and isolation of probiotic strains from canine feces

To analyze the culture-dependent microbial composition differences between the young and the senior canines (Mun et al., 2021a), 10 g of each fecal sample was aseptically put into a sample bag (3M, St. Paul, MN, USA) filled with 90 mL of 0.1% buffered peptone water (Oxoid, Hampshire, UK) and stomached for 2 min, at a speed level of 10. Then, the homogenized solution was serial diluted and plated on de Man, Rogosa & Sharpe (MRS; BD Difco, Franklin Lakes, NJ, USA), Phenylethyl Alcohol Agar (PEA; BD Difco), and Bifidobacterium Selective Agar (BS; BD Difco) supplemented with 7.5% Bacto^TM Agar medium (BD Difco) and incubated for 48 h at 37°C under anaerobic conditions. Isolated colonies were harvested and subcultured in MRS broth for 48 h at 37°C and identified using 16S rRNA sequencing for further experiment. The colonies with 16S rRNA identity of more than 97% were defined as operational taxonomic units (OTUs) and the others with an identity of less than 97% were classified as unknown (Edgar, 2018). LAB isolates were cultured and used for further study.

Evaluation of probiotic activity for canine-originated isolates

Acid and bile tolerance test

To assess the acid and bile tolerance of the four probiotic candidates, the modified method of Dianawati et al. was carried out (Dianawati et al., 2016). In order to prepare an acidic conditioned broth reflecting an artificial stomach, MRS broth with pH adjusted to 2.5 with 6 N HCl (Sigma-Aldrich) was made. The broth was autoclaved and added with 0.45 μm syringe-filtered (Sartorius Korea Biotech, Gyeong-gido, Republic of Korea) 1000 unit/mL concentration of pepsin from porcine gastric mucosa (Sigma-Aldrich). For the acid tolerance test, 100 μL of overnight-cultured bacteria were inoculated into 10 mL acidic broth and incubated at 37°C for 3 h, and plated on MRS agar for colony-forming units counting (CFU/mL). The artificial bile broth was prepared by suspending oxgall (Acumedia Manufacturers, Inc, Lansing, MI, USA) with MRS broth to a final concentration of 0.5%. Then 100 μL of overnight-cultured bacteria were inoculated into 10 mL artificial bile broth and incubated at 37°C for 24 h. The survival rate was calculated by counting the final CFU/mL compared to incubated counts at the starting time.

Mucin adhesion assay

In order to examine in vitro mucus surface adhesion ability of bacteria, artificial mucin broth was prepared by dissolving type III porcine gastric mucin (Sigma-Adrich) into distilled water to the final concentration of 1%. The mixture was filter-sterilized using a 0.45-μm syringe filter, then 100 μL of the mixture was added to the wells of 96-well plate for the triplicate trials for each bacterium. The plate was incubated at 4°C for 24 h. After the incubation, all the supernatants were removed and 100 μL of overnight-cultured bacteria were inoculated to each well and incubated at 37°C for 2 h. Subsequently, all the supernatants were removed and washed the plate five times with PBS solution (Gibco, NY, USA) and detached bacteria using 200 μL of Triton X-100 (Daejung Chemicals & Metals, Gyeonggi-do, Korea). Then, the bacterial solution was serial diluted, plated on MRS agar medium, incubated at 37°C for 48 h. The adhesion rate was calculated using CFU/mL before and after the adhesion assay of each bacterium (Tsilia et al., 2015).

Antibacterial activity

In this study, four representative pathogenic bacteria were used to determine Lactobacillus rhamnosus GG (LGG) and the potential probiotics’ antibacterial activity. The pathogenic bacteria were cultured in LB medium at 37°C for 24 h previously: Listeria monocytogenes EGD-e, Staphylococcus aureus MW2, Salmonella Typhimurium SL1344, and Escherichia coli O157:H7 EDL933. The broth-cultured pathogenic bacteria were plated on LB agar medium, then 5 μL of overnight-cultured of our bacteria were dropped on the plate, and the plates were incubated at 37°C for 48 h. The inhibition zone was classified according to the degree of diameters: ++++, >25 mm; +++, >20 mm; ++, >15 mm; +, >10 mm; -, <10 mm (Balouiri et al., 2016).

Antibiotic sensitivity

The disc diffusion method following the standard Kirby–Bauer (Hudzicki, 2009) method was applied to investigate the antibiotic susceptibility of LGG and potential probiotics. To begin with, 100 μL of overnight-cultured bacteria were plated on MRS at 37°C for 48 h and placed the discs with antibiotics on the surface of each plate, incubated at 37°C for 24 h subsequently. The antibiotic discs (Flinn Scientific, Batavia, IL, USA) used for this experiment were ampicillin (10 μg), chloramphenicol (30 μg), kanamycin (30 μg), penicillin (10 μg), tetracycline (30 μg), and vancomycin (30 μg), and each inhibition zone diameter was measured and categorized into three levels according to the degree of the zone: S—Susceptible: >20 (mm), I—Intermediate: 15–19 (mm), R—Resistant: ≤14 (mm).

Caenorhabditis elegans culture condition and behavior assays

Caenorhabditis elegans fer-15(b26)II;fem-1(hc17)IV was obtained from the Caenorhabditis Genetic Center (St. Paul, MN, USA) and maintained on NGM plates at 15°C. The reference feed of C. elegans, E. coli strain OP50 (OP50) was cultured in Luria–Bertani medium (LB Broth, Miller; BD Difco) at 37°C for 24 h with shaking at 225 rpm. For long-term storage, bacterial cultures were maintained at −80°C containing 15% glycerol in cryo-protectant. OP50 was subcultured twice prior to experimental analysis (Yoo et al., 2022).

For the preparation of live bacterial lawn for C. elegans feeding, bacterial pellet was collected by centrifugation at 13,000 rpm for 1 min, washed twice with sterile M9 buffer (3 g KH₂PO₄, 6 g Na₂HPO₄, and 5 g NaCl dissolved in 1 L distilled H₂O), autoclaved, and then 1 mL of 1 M MgSO₄ was added (Sigma-Aldrich, St. Louis, MO, USA). The purified bacterial pellet was collected by centrifugation at 13,000 rpm for 1 min to remove the supernatant, and the bacterial pellet was concentrated to the final concentration of 2.5 mg/μL (wet weight) in M9 buffer and suspended on nematode growth medium (NGM; 3.5 g Bacto^TM Peptone [BD Difco], 3 g NaCl [Sigma-Aldrich], and 20 g agar [BD Difco] dissolved in 1 L distilled H₂O) plates and dried subsequently.

Life span analysis

For the life span assay, eggs were obtained using sodium hypochlorite–sodium hydroxide solution (Sigma-Aldrich) from egg-bearing worms and synchronized to L1-staged worms on NGM plates at 25°C. The young adult L4-staged worms were plated on 35-mm-diameter NGM plates seeded with OP50, and microRNA expression was analyzed at 24 h, 6 and 12 d of adults. All worms were transferred to fresh OP50 lawns daily until the termination of all worms.

Thrashing assay

In order to assess the motor ability of C. elegans, the number of thrashes in the M9 buffer was observed. Twelve days of potential probiotics and OP50 exposed worms (n = 10 per group, 3 replicates) were moved to a sterile 35-mm-diameter NGM agar plate without bacteria and allowed to crawl for a minute to remove the aggregated bacteria. Then prepare another sterile 35-mm-diameter NGM agar plate filled with 1 mL of M9 buffer, put a worm into the buffer, and let it accustomed to the environment. Count the number of thrash for a minute; a single valid movement means the worm bends its head and tail to the same side (Koopman et al., 2020).

Chemotaxis assay

The chemotactic ability of C. elegans can be projected to olfactory behavior. In order to assess olfactory plasticity, the reaction degree of worms in response to attractant and repellent chemicals was measured. Twelve days of potential probiotics and OP50 exposed worms (n = 10 per group, 3 replicates) were moved to a sterile 35-mm-diameter NGM plate without bacteria and allowed to crawl for a minute to remove the aggregated bacteria. Then we prepared another sterile 35-mm-diameter NGM agar plate with a drop of 100 mM isopropyl alcohol diluted in 100% EtOH (Merk, Kenilworth, NJ, USA) as an attractant dried on the edge of the plate. The worms were moved to the middle of the plate, let freely crawl for 30 min, and measured the number and direction the worms turned to the attractant. The same procedure was repeated with the repellent, 30% octanol (v/v) in 100% EtOH (Mills et al., 2012).

RNA isolation and transcriptome analysis

L4-staged fer-15; fem-1 worms were randomly harvested at 24 h, 6 and 12 d after exposure to OP50 to examine gene expressions that differed during aging. Total RNA was isolated using TRIZOL reagent (Invitrogen, Carlsbad, CA, USA) and purified using RNeasy Mini Kit (Qiagen) with a mini bead beater (Biospec, Bartlesville, OK) according to the manufacturer’s instructions. Target cRNAs were synthesized, fragmented, and hybridized using Agilent’s Low RNA Input Linear Amplification kit (Agilent Technologies, Santa Clara, CA, USA) following the manufacturer’s instructions. The fragmented target cRNA probes were suspended in 2X hybridization buffer offered by the kit, and then applied directly onto C. elegans 4 X 44K microarray chip (Agilent Technologies). The microarray chips were hybridized in Agilent’s Hybridization Oven at 65°C for 17 h, and then washed according to the manufacturer’s instructions (Agilent Technologies).

The fluorescence intensity was analyzed using GenePix Pro 6.0 (Axon Instruments, Foster City, CA, USA), and the average fluorescence intensity for each well was measured with the local background subtracted. The data normalization and fold changes were analyzed using GeneSpring 7.3.1 (Agilent Technologies). The fluorescence intensity was normalized by LOWESS (locally weighted regression scatter plot smoothing), comprising the ratio reduced to the residual LOWESS fit of the intensity versus ratio curve. The average normalized ratios were calculated by the ratio of the average normalized signal channel intensity with the average normalized control channel intensity. Each gene was considered to be expressed differentially when the p-value comparing two chips was less than 0.05.

Integrative network analysis

Functional integrative annotation analysis was conducted from the Gene Ontology Consortium with GeneSpring GX 7.3. Genes based on searches of web gene ontology clustering tool DAVID (http://david.abcc.ncifcrf.gov/) and Medline (http://www.ncbi.nlm.nih.gov/) databases. The genes were annotated under Biological Process, Cellular Component, and Molecular function (functional false discovery rate [FDR] <0.05). Significant terms (P < 0.05) of functional pathway and KEGG were plotted using a plugin of Cytocape (http://cytoscape.org) with filters of ClueGO+CluPedia and BiNGO(Maere et al., 2005); the heat map of differentially expressed genes were retrieved using GraphPad Prism 7.0.4 (GraphPad Software, San Diego, CA, USA).

Quantitative reverse transcription PCR

qRT-PCR was practiced to determine the differential levels of gene expressions according to the aging. Twelve days of exposure to OP50, LGG, and potential probiotics, worms were harvested, washed twice using M9 buffer, and then total RNA was isolated using TRIZOL reagent (Invitrogen, Carlsbad, CA, USA) and purified using RNeasy Mini Kit (Qiagen, Hilden, Germany) with a mini bead beater (Biospec, Bartlesville, OK) according to the manufacturer’s instructions. Fifty nanograms of total RNA was reverse transcribed for complementary DNA using miScript II RT kit (Qiagen) (Park et al., 2018) and cDNA samples were diluted to 40 ng/μL using nuclease-free water (Gibco), and then qRT-PCR was practiced using CFX96 real-time system (Bio-rad) with miScript SYBR Green PCR kit. The primers used for this study are presented in Supplementary Table 1 (Wang et al., 2015; Qi et al., 2017; Li et al., 2018).

Fermentation of the Intestinal Microbiota Model (FIMM) system

In reference to the SHIME model (van de Wiele et al., 2015), the FIMM system was utilized as a canine in vitro colon atmosphere. Feces of young and aged canines were pooled by groups and incubated in mGAM media (HIMEDIA, DB Maarn, Netherlands), pH adjusted to 7.0 for 48 h under anaerobic conditions. The aged groups were cocultured with Lactobacillus salivarius (LS) and Enterococcus hirae (EH) separately. After 48 h, the cultivates were collected by centrifuge (7,000 rpm, 15 min) and stored at −80^°C until further metagenome analysis. Metagenomic analysis for determining the dynamics of gut microbiota were performed using Illumina® iSeq 100 platform as the described method above.

Statistical analysis

All data points used for this study were performed in triplicates and data were expressed as the mean ± standard deviation based on three repeated experiments, and the significant differences were determined according to the Student’s t-test and one-way ANOVA followed by Turkey’s post hoc test.

Availability of data and materials

The datasets generated and/or analyzed during this study are available and deposited in NCBI under the Bioproject accession PRJNA865497, PRJNA636267, and GEO accession GSE210389. Additional datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Results

The microbial composition was altered as domestic canines getting aged

In order to analyze the microbial transitions according to aging, we conducted metagenomic analysis of canine feces from the young and the senior ones. The hypervariable V4 region of the 16S rRNA was sequenced and a total 185,032 sequences were retrieved and classified to 1,320 OTUs; 209 of them were recognized with greater than 97% identity and assigned to bacteria. Then, the bacteria classified by genus were ordered by abundance, and the top 10 from each group were identified. Bacteroides were the most prevalent genus in both groups (young, 23.0%; aged, 21.2%), and there was a couple of genus that exhibited distinct differences between the young and the aged groups. Especially young fecal microbiota exhibited 15.3% of Lactobacillus and 3.8% of Enterococcus; however, they were lost as canines got old, both genera reduced to 0.7% in the aged fecal microbiota. Moreover, Clostridium (7.9%), Allobaculum (9.6%), and Staphylococcus (15.8%) genera were increased in the aged group (Figure 1A). These compositional changes led to the reduction of the relative abundance of LAB (Lactobacillus and Enterococcus) in the aged group; the relative abundance of LAB in the young group was 152.4 ± 54.0, while the old group had 47.6 ± 16.1 OTUs (Figure 1B). In addition, there were obvious diversity differences between the two groups according to the Chao and Shannon diversity index (young vs. aged: 84.9 ± 8.6 vs. 56.1 ± 13.9; 2.9 ± 0.2 vs. 2.4 ± 0.2, respectively) and distinct clusterings of weighted and unweighted unifrac, separated from each group, as well as alterations in microbial compositions in the aged ones (Figure 1D). This result implies the potent role of Lactobacillus and Enterococcus as a modulator of senior fecal microbiota.

Figure 1.

The diversity and richness of fecal microbiota were altered in the aged canine group. Metagenomic approach indicates relative abundance of bacteria isolated from fecal samples of young and aged canines (A) and the relative abundance of lactic acid bacteria in young and aged canine groups (B). The Chao and Shannon index values representing the alpha-diversity of microbiome from the feces of canines (C). All values are expressed as mean ± SD; significant differences were determined using Student’s t-test at **P < 0.01. PCoA plots based on weighted and unweighted UniFrac distances of fecal microbiome (D). Each plot represents each sample; the axes represent the two dimensions that account for the highest amount of variance in the communities.

Metagenomic and culturomic analyses of young fecal microbiota comprehensively present potential probiotics

The metagenomics analysis of young fecal microbiota revealed the substantial differences in compositions of Lactobacillus and Enterococcus when compared to the older fecal microbiota. Therefore, we sought to see if the same result might be obtained by cultural approaches. For practical discovery, three types of media were utilized to isolate as many distinct LAB as possible, and a total of 305 isolates from both groups were harvested and identified by 16s rRNA sequencing. Interestingly, the elderly group had a greater number of isolated species (10 vs. 19); however, the proportion of lactic acid bacillus among the isolates was much higher in the young group (94% vs. 87.5%) (Figure 2A and B). In accordance with the results of metagenomics, the culturomic study of young and elderly canine fecal microbiota revealed a high abundance of LAB as one of the major differences between the young and aged microbiota. Five species of Lactobacillus (L. salivarius, L. animalis, L. johnsonii, and L. reuteri) and three species of Enterococcus (E. faecium, E. hirae, and E. faecalis) were isolated from young canine culturomic isolates. The metagenomic LAB isolates of the young canine were also composed of Lactobacillus species (L. amylolyticus, L. salivarius, L. mucosae, L. gasseri, fermentum, L. plantarum, and L. murinus) and Enterococcus species (E. hirae, E. durans, E. faecium, and E. faecalis) (Figure 2C and D). Then, we tracked the number of overlapping LAB species between the culturomic and metagenomic approaches. Then, we tracked the number of overlapping LAB species between the culturomic and metagenomic approaches. Two species of Lactobacillus (L. salivarius and L. amylolyticus) and three species of Enterococcus (E. faecium, E. hirae, and E. faecium) were shared by both culturomic and metagenomic analyses of immature LAB, as seen by a Venn diagram (Figure 2D). Based on this outcome, we anticipated that the LAB shared by both strategies may be potent probiotics that are especially advantageous for the elderly.

Figure 2.

The comparison analysis of culturomics and metagenomics in young and aged canines. Culturomic approach indicates relative abundance of bacteria isolated from fecal samples of young and old canines. The abundant lactic acid bacteria of the young canines were presented using both culturomic (C) and metagenomics approach (D). Venn diagram describes the number of lactic acid bacterial species sorted by culturomics (yellow), metagenomics (blue), and the species detected by both analyses are described (green) (E).

Four lactic acid bacteria isolated from canine feces exhibited probiotic properties similar to LGG

Then we investigated the probiotic properties of the five bacterial species and sorted out the four most potential probiotics candidates: Lactobacillus amylolyticus (LA), LS, EH, and Enterococcus faecium (EF). The four potential probiotics were inspected of acid tolerance, bile tolerance, and mucin adhesion ability compared to control probiotics, Lactobacillus rhamnosus GG, the probiotic bacteria that have the most clinical recognition up to date (Gorbach, 2000).

Acid and bile tolerance tests are one of the universal methods to determine probiotic capability, since probiotics must survive physical and chemical constraints in the gastrointestinal tract, such as acid and bile (Dianawati et al., 2016). Compared with LGG, LA, LS, EH, and EF exhibited similar survival rates in artificial gastric acid without significant differences (Figure 3A). In the artificial bile environment, the potential probiotics demonstrated a significant increased survival rate compared with LGG; LA survived 120.2% ± 1.2%, LS survived 130.5% ± 5.1%, EH survived 115.9% ± 1.2%, and EF survived 108% ± 0.5% more (Figure 3B).

Figure 3.

The lactic acid bacteria isolated from diverse canine feces demonstrated probiotic properties. The potential probiotics are evaluated tolerance to artificial gastric acid (A), artificial bile (B) adhesion ability to in vitro mucus surface (EC as properties of probiotics. All values are expressed as mean ± SD; significant differences were determined using Student’s t-test at *P < 0.05, **P < 0.01 compared with LGG. Lactobacillus rhamnosus GG (LGG), Lactobacillus amylolyticus (LA), Lactobacillus salivarius (LS), Enterococcus hirae (EH), and Enterococcus faecium (EF).

After the probiotics bypass the harsh GI tract, they must survive in the gut to proliferate, and the capability is determined by the adhesion ability to the mucus layer (Tsilia et al., 2015). All four potential probiotics proved to have alike mucin adhesive ability with LGG (Figure 3C).

The potential probiotics were mostly susceptible to antibiotics and resistant to pathogenic bacteria

The gut microbiome stores antibiotic resistance genes, which could be transmitted horizontally to pathogens, contributing to the establishment of drug-resistant bacteria (Hudzicki, 2009). Naturally, LAB are resistant to certain antibiotics; nevertheless, the resistance is not transmissible and the bacteria are still susceptible to many clinically used antibiotics (Zhou et al., 2005). LGG was susceptible to ampicillin, chloramphenicol, penicillin, and tetracycline while resistant to kanamycin and vancomycin. LS and EF exhibited the identical antibiotic resistance pattern to LGG, while LA was resistance to tetracycline and susceptible to vancomycin. EH had the most resistance to antibiotics; it was resistant to kanamycin, tetracycline, and vancomycin, and was susceptible to ampicillin, chloramphenicol, and penicillin (Table 1).

Table 1.

Susceptibility of Lactobacilli and Enterococci isolated from feces against six antibiotics.

Strains	Inhibition zone of antibiotics*
	Ampicillin	Chloramphenicol	Kanamycin	Penicillin	Tetracycline	Vancomycin
LA	S	S	R	S	R	S
LS	S	S	R	S	S	R
EH	S	S	R	S	R	R
EF	S	S	R	S	S	R
LGG	S	S	R	S	S	R

*Antibiotic resistance was evaluated by disc diffusion (inhibition zone diameter): S—Susceptible: >20 (mm), IIntermediate: 15–19 (mm), R—Resistant: ≤14 (mm); Lactobacillus rhamnosus GG (LGG), Lactobacillus amylolyticus (LA), Lactobacillus salivarius (LS), Enterococcus hirae (EH), and Enterococcus faecium (EF).

The first condition of pathogenic bacteria inducing trouble in the host is the adhesion ability to mucosal surfaces, and the bacterial adhesions are responsible for pathogen adherence. Hence, inhibition of adhesion may prevent the pathogen from colonizing and infection (Balouiri et al., 2016). LGG was capable of inhibiting the four representative pathogenic bacteria: EGD-e, MW2, SL1344, and EDL933. Enterococcus were capable of inhibiting the most pathogenic bacteria; LA had antimicrobial activity against EGD-e, SL1344, and EDL933; and LS mildly inhibited all pathogenic bacteria (Table 2).

Table 2.

Antimicrobial activity of Lactobacilli and Enterococci isolated from canine feces against four pathogenic bacteria.

Strains	Inhibition zone of pathogenic bacteria*
	EDG-e	Newman	SL1344	EDL93
LA	+++	+	+++	++
LS	+	+	+	++
EH	++++	+	++++	+++
EF	+++	+	+++	++++
LGG	++++	+++	++++	++++

*Diameter of inhibition zone: ++++, >25 mm; +++, >20 mm; ++, >15 mm; +, >10 mm; -, <10 mm.

Lactobacillus rhamnosus GG (LGG), Lactobacillus amylolyticus (LA), Lactobacillus salivarius (LS), Enterococcus hirae (EH), and Enterococcus faecium (EF).

The potential probiotics upregulated C. elegans life span and relieved behavior degeneration caused by aging

Caenorhabditis elegans is a well-established genetic model organism that could mimic most human diseases, and it has been considered as valuable for studying in vivo at both the metabolic and genetic levels (Olsen et al., 2006; Son et al., 2019). In this study, the beneficial health effect of the potential probiotics was investigated using C. elegans life span assay. Interestingly, all bacteria upregulated lifespan compared to E. coli OP50 (OP50), the standard feed, and most potential probiotics were able to enhance C. elegans life span compared with LGG; LA-fed C. elegans survived 5.3%, LS 15.8%, EF 5.3% significantly longer than LGG (Figure 4A–D)

Figure 4.

The potential probiotics isolated from canine feces enhanced C. elegans longevity and behavior degeneration. The life span assay of C. elegans seeded with each potential probiotic is evaluated (A–D). The behavior degeneration level of aging worms exposed to each potential probiotic are evaluated as follows: motor behavior (E) and sensing behavior especially olfactory competence (F, G). All values are expressed as mean ± SD, normalized to the mean of OP50; significant differences were determined using Student’s t-test at *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 compared with E. coli OP50; #P < 0.05, ##P < 0.01, ###P < 0.001 compared with LGG. Lactobacillus rhamnosus GG (LGG), Lactobacillus amylolyticus (LA), Lactobacillus salivarius (LS), Enterococcus hirae (EH), and Enterococcus faecium (EF).

In addition, one of the typical behaviors of aged C. elegans is the motor and sensitive neuron degeneration (Van Pelt and Truttmann, 2020). In this study, body bending frequency, which represents motor behavior, was significantly sustained with all treatments (Figure 4E). Thus, sensing behavior represented by olfactory sensing reactivity measured by the attraction and repulsion to favorable chemical, isoamyl alcohol, and abhorrent chemical, octanol, was also significantly prolonged by probiotics treatment. There was no significant difference between LGG and potential probiotics (Figure 4F and G).

Microarray of mRNAs of C. elegans exhibited changes in reaction to aging

To discover the changes made in genetic levels according to aging of C. elegans, we carried out a microarray of mRNAs of C. elegans exposed to OP50. Subsequently, total RNA from worms was isolated, synthesized to cRNA, and hybridized for microarray chip assay. Following the alignment data, total of 27,423 mRNA expression changes were analyzed between two time points, 24 h and 12 d. The plot indicates the degree of correlation between two variables (Friendly and Denis, 2005). The positive correlation between each time group means that the gene expressions are actually changing in proportion to worms’ aging. The aging process in C. elegans usually accompanies cognitive decline and physical defects. Therefore, the changes in the worm’s sensing behavior confronting stimulus could indicate a positive or negative healthy lifespan, in which case, chemosensory neurons such as olfactory or gustatory neurons could promote or limit longevity by influencing Insulin/IGF-1 signaling (Jeong et al., 2012; Mutlu et al., 2020). In this study, as the worms were undergoing a transformational stage from young adult to elderly model, most enriched genetic ontology (GO) terms were related to collagen, membrane component, and cuticulin-based development, physiological changes were also noted, such as innate immune response and sensory perception to olfaction (Figure 5A-C).

Figure 5.

Microarray of mRNAs from C. elegans at different aged point seeded with E. coli OP50. During the experiment, E. coli OP50 exposed worms were randomly selected at 24 h and 12 d to assess the changes in microRNAs. The plotted spots are normalized to the signal ratio scale. Dot plot visualizes the significant pathways (FDR < 0.05) of biological process (G), cellular component (H) and molecular function (I) identified using DAVID to be enriched for GO terms at different time points; each data is normalized to the ratio of the number of related genes expressed.

GO ontology analysis indicated aged C. elegans had changes in functional genes

GO is the initiative bioinformatics designating nomenclatures of genes and gene products (Ashburner et al., 2000). In our study, GO enrichment analysis was performed to investigate the network related to C. elegans aging process. GO terms that were enriched >2.0-fold over the average with P < 0.05 were considered to be significant terms. The aging process induced changes in genes related to biological process, cellular components and molecular function networks (Figure 6A), and those networks were led to functional changes of mitogen-activated protein kinase (MAPK), calcium, Wnt signaling pathways in association with longevity regulating pathway. Moreover, fatty acid degradation and neuroactive ligand–receptor interacting pathways were also mapped (Figure 6B).

Figure 6.

GO ontology based functional analysis of differentially expressed genes. GO ontology analysis of differentially expressed genes in the form of biological process, cellular component and molecular function networks (A). Bigger sizes of the plot represent the higher level mechanisms and different color indicates the ratio of genes involved. Protein domains and related genes expressed differentially between the 12 d and 24 h E. coli OP50 exposed C. elegans (B). The significant pathways (Benjamini–Hochberg FDR <0.05) were analyzed using DAVID and each point was normalized to the ratio of the number of related genes expressed.

Elderly worms exhibited changes in metabolic pathways

The gene expression changes between young and elderly worms were assessed. As mentioned above, the aging and dauer-forming marker kinases-related genes, scd-2, kin-3, kin-5, were upregulated in elderly worms and 48 functional pathways including calcium signaling, longevity regulating pathways, and fatty acid degradation related to P450 were upregulated (Manning, 2005; Judy et al., 2013) (Figure 6A) On the contrary, component protein domain, cytosolic fatty acid binding genes (EEED8.2, lbp-3, lbp-4, lbp-5, lbp-7, lbp-8) were downregulated, and 97 including oxidative phosphorylation, MAPK, Wnt, FoxO, and sphingolipid metabolism were downregulated in elderly worms (Figure 6C-D).

Potential probiotics treatment enriched the relative quantitation of functional genes

To evaluate the effects of probiotics in mRNAs related to aging, quantification of genes using qRT-PCR was performed with 12-d-old C. elegans. Bacterial treatment significantly enriched survival mediators compared with OP50 (Figure 7A). Compared with LGG, LA significantly upregulated daf-2, age-1, and akt-2, which are related to insulin-like signaling pathway, moderate development, lifespan, and stress response (Larsen et al., 1995; Evans et al., 2008). Bacterial treatment led significant changes in egl-1 and ced-9, an antiapoptotic gene that encodes antiapoptotic BCL-2-like protein (Nehme and Conradt, 2008; Conradt et al., 2016) compared with OP50 (Figure 7B). In succession, elderly C. elegans typically suffer from phenotypic changes. Daf-16, which works with the forkhead transcription factor FOXO, antagonizes age-1 and hpk-1, the stress indicators, and let-363, which induces developmental arrest (Vellai et al., 2003; Hesp et al., 2015’ Berber et al., 2016) were significantly downregulated by LA, LS, and EF compared with OP50. In accordance with the previous results, skn-1, the stress responder (Blackwell et al., 2015) that could promote C. elegans longevity, was significantly enhanced with all bacterial treatments, especially ser-7, mammalian 5-HT₇-like receptor that influences motor behavior (Hobson et al., 2006). Furthermore, odr-3, 7, 10 were significantly upregulated by all bacterial treatments compared with OP50, which is noteworthy since those genes are part of the main chemosensory neuron components which regulates motivation to move and olfaction (Shen et al., 2010) (Figure 7C).

Figure 7.

Relative expressions of aging-related genes were modified by potential probiotics in C. elegans. The gene expression level related to ageing were categorized into three: survival mediators (C), apoptosis mediators (D), and senescence index (E) were quantified using qRT-PCR and ΔC(t) values were normalized to internal control act-1. All values are expressed as mean ± SD, normalized to the mean of OP50; significant differences were determined using one-way ANOVA followed by Turkey’s post hoc test at *P < 0.05, **P < 0.01, ****P < 0.0001 compared with OP50; ##P < 0.01 compared with LGG. Lactobacillus rhamnosus GG (LGG), Lactobacillus amylolyticus (LA), Lactobacillus salivarius (LS), Enterococcus hirae (EH), and Enterococcus faecium (EF).

Potential probiotics treatment altered the fecal microbiota of aged canines through FIMM analysis

Based on the previous results about the benefits of potential probiotics to eldered worms, two of the potential probiotics were investigated whether they could make actual changes to canines in vitro. Fecal samples of young canines, aged canines, and aged canine fecal samples cocultured with LS and EH were incubated using FIMM, the colon-mimicking cultivation system. Unfortunately, there were no significant changes between groups in Shannon index; EH supplement significantly diversified the fecal microbiota of aged canines by recovering Chao index to young level and there were distinct clusterings of fecal microbiota between the FIMM-aged group and FIMM-probiotics supplemented aged canine groups (Figure 8D and E). There were also compositional changes after FIMM incubation. In the order level, both LS and EH supplements increased Lactobacillales and Coriobacteriales, while reduced Actinomycetales (Figure 8A). More specifically, in the family level, LS and EH upregulated S24-7, Lactobacillaceae and Ruminococcaceae during FIMM, which resulted in the increase of the genera Ruminococcus, Blautia, Lactobacillus, Enterococcus, while reducing Clostridium. On the other hand, Bifidobacteriaceae from young canine groups was lost, and Erysipelotrichaceae family was sustained in all aged canine groups regardless of LS and EH supplement (Figure 8B and C).

Figure 8.

The diversity and richness of fecal microbiota were altered in the aged canine group with the treatment of potential probiotics. FIMM of young, aged, and aged fecal sample with potential probiotics altered the bacterial compositions. Order level (A) and family level (B). The Chao and Shannon index values representing the alpha-diversity of FIMM-microbiome (C). All values are expressed as mean ± SD; significant differences were determined using Student’s t-test at **** P < 0.0001. PCoA plots based on unweighted and weighted UniFrac distances of FIMM-microbiome (D). Each plot represents each sample and the axes represent the two dimensions that account for the highest amount of variance in the communities. Lactobacillus salivarius (LS), Enterococcus faecium (EF).

Discussion

Aging accompanies the accumulation of degenerative functional and structural changes that are considered a universal process among all living organisms, which is inexorable. Aged mammals exhibit reduced metabolic rates including motor and cognitive functions as well as increased DNA damage (Harman, 1981), and probiotics have been widely used in both human and household animals in health-promoting “functional foods,” especially the LAB (Sornplang and Piyadeatsoontorn, 2016; Yang et al., 2022). Therefore, in this research, we intended to explore potential probiotics from the feces of young canines that are capable of enhancing healthy longevity using C. elegans in vivo model.

To begin, the research of exploring LAB habituating in healthy canine gut microbiome was conducted. Among 305 anaerobic LAB isolated from canine feces, 4 potential probiotics were singled out depending on the fundamental probiotic properties: acid tolerance, bile tolerance, mucin adhesion ability, and resistance against antibiotics and pathogens. In fact, the potential probiotics we chose for further studies, LA, LS, EH, and EF, are yet entitled in the list of authorized microorganisms in commercial diet supplements (Irianto and Austin, 2002), though previous studies revealed the health-promoting effects of them. Lactobacillus amylolyticus has been called a potential probiotic previously; it was certified for its usage in fermented foods such as grain to feed growing pigs, tofu, and soymilk; no harm reported yet (Grosu-Tudor et al., 2014; Fei et al., 2017). Lactobacillus salivarius has been considered probiotic already in some studies since it has been proven to be effective in alleviating inflammation, such as bowel syndrome, ulcerative colitis (O’Mahony et al., 2005), and suppressing colonization of pathogens like Helicobacter pylori and Salmonella spp. (Aiba et al., 1998; Pascual et al., 1999). The genus Enterococcus, mostly originated from an animal source, has been controversial in the field of probiotics of safety because some species of them possess antibiotic resistance (Ogier and Serror, 2008; Hanchi et al., 2018); still, they are considered as prevailing potential probiotics. Enterococcus hirae was renowned for producing bacteriocin active against pathogens and was reported to dwindle the cancer cell lines (Gupta and Tiwari, 2015; Sharma et al., 2018). Enterococcus faecium recently established its role in pharmaceutical use by preventing vibriosis and pneumonia (Arun and Rahul, 2019), as well as an effectual immune modulator (Mansour et al., 2014).

To determine the appropriate time point for assessing different gene expressions related to aging process, we conducted a life span assay with C. elegans fer-15;fem-1 strain seeded with E. coli OP50. The worms are considered to be appropriate for life span assay since they lack progeny without phenotypic alterations at 25°C (Park et al., 2018), and OP50 has been widely used as a standard feed when maintaining C. elegans. Accordingly, we observed affirmative effects of the four potential probiotics; they significantly lengthened the life span of C. elegans along with the motor and sensing behavior compared with the aging-induced diminished group.

Studying for healthy longevity, the most important groundwork is to define the changes accompanied by functional and molecular changes due to the aging process. In the same context, age-related molecular signal changes are considered the most profound source to estimate the biomarkers for aging, and investigating changes in aging-associated gene expression would also be a valuable source to determine the decrease or increase in life span.

As noted above, to identify the aging process, examining the changes in molecular genes was necessary, and by analyzing C. elegans mRNA, the desirable outcomes in molecular, functional, and transcriptional regulators were achievable. The primary mRNAs were reported from C. elegans. Lin-4 was the first identified mRNA, which regulates the ovulation of C. elegans, and the second reported mRNA (Lee et al., 1993), let-7, regulates C. elegans development, and it has been reported to be present in other mammals, including humans (Reinhart et al., 2000). Respectively, lin-14 and lin-41, which were also discovered from C. elegans are involved in the development of larvae; mir-58 family, which targets daf-1 and daf-4 functions in dauer formation, and more have been discovered until recent days (Harman, 1981; Ambros and Ruvkun, 2018).

Although C. elegans has been a widely used genetic model, the related studies are still limited to certain genes focusing on functional metabolic analysis. Therefore, in this study, we built a naturally aging model focused on identifying and listing a wide range of genes affected by aging in C. elegans by analyzing mRNA expression changes at various time points to set up the foundation for further applicable aging studies.

The GO is the initiative of bioinformatics designating nomenclatures of genes and gene products (Ashburner et al., 2000). In our study, GO enrichment analysis was performed to investigate the network related to C. elegans aging process. GO terms that were enriched >2.0-fold over the average with P < 0.05 were considered to be significant terms. The aging process induced changes in genes related to biological process, cellular component, and molecular function networks, and those networks led to functional changes of MAPK, calcium, and Wnt signaling pathways in association with the longevity regulating pathway. Moreover, fatty acid degradation and neuroactive ligand–receptor interacting pathways were also mapped. Among them, protein kinase activity and MAPK signaling were the most significant and noticeable since there are previous studies indicating the interaction between the regulation of protein kinase, such as AAK-2(AMPK homolog), or PI3K is directly related to the aging process of C. elegans (Apfeld et al., 2004), (Curtis et al., 2006).

For further comparisons of differentially expressed genes throughout the aging, we examined detailed changes of protein domain expressions and functional pathways between 12-d OP50-treated worms to 24-h worms. Interestingly, some genes translating chitinase II and zinc/PHD-finger protein domains were shown to be downregulated, which are responsible for guarding germ cells conversing into neurons (Hajduskova et al., 2019) and zinc/PHD-finger translating gene, daf-2, the insulin receptor functions in DAF-16/FOXO pathways, which contribute to innate immunity, was also downregulated. In addition, old-1 and old-2, the aging marker genes; scd-1, suppressor of dauer gene; and trk-1, the pro-granulin receptor, which regulate stress resistance were increased (McKay et al., 2003; Manning, 2005; Reiner et al., 2008; Judy et al., 2013). As the worms got aged, the wide ranges of protein domain expressions were decreased and the number of downregulated KEGG functional pathway was significantly increased by 101. Universally, the aged C. elegans becomes less active due to muscle degeneration and shows decreased macromolecular signaling. In the same context, in our study, metabolic functional pathways were typically downregulated in 12-d worms, such as glycolysis/gluconeogenesis, ATP-synthase-related oxidative phosphorylation, and longevity regulating pathway. Molecular signaling pathways such as MAPK, mTOR, ribosome, and FoxO signaling pathways, which previously introduced to be closely related to healthy life span, were also downregulated (Harman, 1981; McKay et al., 2003; Son et al., 2019).

Next, the mRNA expression changes between middle-aged and elderly worms were assessed. As mentioned above, survival-related markers including dauer-forming and insulin-like signaling kinases-related genes (Manning, 2005; Judy et al., 2013), daf-2, age-1, and akt-2, were downregulated in elderly worms, while the potential probiotics significantly increased their expressions and the apoptotic protein-encoding genes such as egl-1 and ced-9 were upregulated in aged model, and the probiotic treatment significantly reduced their expressions. Moreover, the neuronal degradation was reflected by the decreased expression of ser-7, odr-3, 7, and 10, which were recovered by probiotic treatments.

Finally, to investigate whether the potential probiotics could benefit vertebrates, LS and EH were supplemented to the fecal sample of aged canines and incubated under FIMM. During FIMM incubation, LS and EH increased Lactobacillales and Coriobacteriales in the order level, which are the taxons of order composing probiotics (Chen et al., 2020), while decreasing the level of Actinomycetales, which has known to be increased as canines age and could induce allergic infections (Park et al., 2019). Even in the family-level taxons, both LS and EH upregulated S24-7, Lactobacillaceae and Ruminococcaceae, which have been reported to be abundant in healthy canines (Adams and Gutierrez, 2018; Tal et al., 2021). To conclude, we built a specific gene network corresponding to the pathways that differed during the aging process and the potentiality of the commercial value of the four probiotic candidates, and they proved their potentiality by diversifying and increasing the amount of beneficial commensal bacteria of canines during FIMM. Still, further studies are required for the numerous genes whose functions are not sufficiently studied and the application of the potential commensal LAB, where commercial probiotics are usually composed of multiple strains.

Conclusion

In our study, culturomic-, metagenomic-, and transcriptomic-based results suggested that a specific gene network corresponding to the pathways differentially expressed during the aging of C. elegans and selected commensal LAB as newly isolated potential probiotic strains could be provided health-promoting benefits in the aging of domestic animals by modulating dynamics of gut microbiota.

Glossary

Abbreviations

ARGs

antibiotic resistance genes

EF

Enterococcus faecium

EH

Enterococcus hirae

OP50

Escherichia coli OP50

FIMM

Fermentation of the Intestinal Microbiota Model

GO

genetic ontology

LAB

lactic acid bacteria

LA

Lactobacillus amylolyticus

LS

Lactobacillus salivarius

LGG

Lactobacillus rhamnosus GG

MAPK

mitogen-activated protein kinase

Conflict of Interest Statement

The authors declare no real or perceived conflicts of interest.

Author Contributions

ANK, DM, SR, JJL, SJO, MKK, MS, SO, and YK conceived and designed research. ANK, DM, SR, JJL, SJO, MKK, MS, SO, and YK conducted experiments. ANK, DM, SR, JJL, and YK conducted bioinformatics analyses. ANK, DM, SR, JJL, SJO, MKK, MS, SO, and YK analyzed the data. ANK, DM, SR, JJL, SJO, MKK, MS, SO, and YK prepared the manuscript. All authors read and approved the final manuscript.