Safety assessment of Streptococcus thermophilus IDCC 2201 used for product manufacturing in Korea

Abstract Safety evaluation of probiotics has become increasingly important for human consumption in food industry. The aims of this study were to assess safety of Streptococcus thermophilus IDCC 2201 through in vitro and in vivo tests. In results, this strain was found to be negative for hemolytic and β‐glucuronidase activity. In addition, thermophilus IDCC 2201 was susceptible to nine antibiotics suggested by EFSA. In accordance with MIC tests, whole‐genome analysis indicated that S. thermophilus IDCC 2201 neither harbors antibiotic resistance nor toxigenic genes. Furthermore, none of the biogenic amines including tyramine and histamine was produced and negligible amounts of D‐lactate were produced by S. thermophilus IDCC 2201. Finally, it was confirmed that there was no mortality and toxicity throughout single‐dose oral toxicity tests in rats. Therefore, we report that S. thermophilus IDCC 2201 is considered to be safe for human consumption as probiotics.

of probiotics have brought their safety aspects into focus (Sanders et al., 2010). Thus, the FAO/WHO guidelines recommend performing clinically standardized methods for assessing the safety of probiotics (Hill et al., 2014). Specifically, to ensure that probiotics are safe for human and animal consumption, the guidelines recommend establishing the antibiotic resistance patterns of probiotic strains and the absence of acquired or transferable resistance factors (Yahav et al., 2018). In accordance with this notion, the transmission of antibiotic resistance genes from probiotics to gut microbiota is a major health concern (Wang et al., 2006).
In this study, the safety of S. thermophilus IDCC 2201 used as a part of commercial probiotic products was evaluated by using phenotypic and genotypic methods. Firstly, in vitro tests were performed to analyze its hemolytic and enzymatic activities, and the minimal inhibitory concentrations (MICs) against a variety of antibiotics. Secondly, whole-genome analysis was performed to determine whether this strain harbors toxigenic and antibiotic resistance genes and whether these genes are transferable to commensal or pathogenic bacteria. Finally, a single-dose oral toxicity test was performed in rats. Therefore, this study verifies the safety of S. thermophilus IDCC 2201.

| Bacterial strains and growth conditions
Streptococcus thermophilus IDCC 2201 strain isolated from homemade yogurt was identified phenotypically and genotypically and have been included in products manufactured by Ildong Bioscience, Korea (Table S1 and Table S2). It was grown in MRS medium (BD Difco, Franklin Lakes, NJ, USA) at 37°C with 0.5% CO 2 in static incubator. Staphylococcus aureus ATCC 25923 was used as a positive strain for hemolysis assay, and it was incubated in brain heart infusion (BHI; BD Difco) medium at 37°C with vigorous shaking at 200 rpm.

| Hemolysis assay and enzymatic activities test
Hemolysis assay was performed by streaking bacterial cells on sheep blood agar plates (BBL Microbiology Systems, Cockeysville, MD, USA). Then, the plates were incubated at 37°C until the clear zones around the colony were observed.
Enzymatic activities were determined using an API-ZYM kit (BIOMÉRIUX, Marcy-l'Étoile, France) according to the manufacturer's instructions. Briefly, cell cultures were harvested and resuspended in sterile distilled water. Afterward, 65 μl suspension of McFarland standard was deposited into each well, and the plates were incubated at 37°C for 4 hr. Then, one drop of ZYM-A and ZYM-B reagent was added to each well, and enzyme activity was analyzed after 5 min.

| Determination of minimum inhibitory concentration and whole-genome analysis
The MICs of S. thermophilus IDCC 2201 were determined by Etest method (Mayrhofer et al., 2008). Briefly, 1% (v/v) of cells grown overnight was transferred into fresh MRS broth. When the cell density of the culture plates reached approx. at 1-2 × 10 8 CFU (colony forming units)/ml, cells were spread onto MRS agar plate prior to the overlay of antibiotic strips (Liofilchem, Roseto degli Abruzzi, Italy).
The antibiotics tested in the study were ampicillin, vancomycin, gentamycin, kanamycin, streptomycin, erythromycin, clindamycin, tetracycline, and chloramphenicol according to the technical guidelines of the EFSA (EFSA, 2012).

| Determination of biogenic amine concentrations
The biogenic amines (BAs) produced by S. thermophilus IDCC 2201 were determined according to the previous study (Deepika Priyadarshani & Rakshit, 2011) and EFSA standard protocol (EFSA, 2012) with minor modifications. Briefly, a single colony was cultured overnight in 10 ml of MRS broth at 37°C in static incubator.
The grown cells were transferred into 10 ml of fresh MRS broth with a dilution at 1:100 and incubated at 37°C for 24 hr. Then, the supernatant from the culture was collected by centrifugation Peaks were detected at 254 nm by using a UV detector (UV-2075 plus, Jasco) and quantified according to calibration curves of each BA such as tyramine, histamine, putrescine, 2-phenethylamine, and cadaverine.

| L/D-lactate formation
The quantification of L-lactate and D-lactate was performed by using an assay kit (Megazyme, Bray, Ireland). Briefly, cell-free supernatant from overnight culture of S. thermophilus IDCC 2201 was assayed with the following enzymes; L-/D-lactate dehydrogenase and glutamate-pyruvate transaminase. Then, the absorbance of diluted supernatant was measured at 340 nm, and L-/D-lactate concentrations were calculated according to the manufacturer's protocol.

| Acute oral toxicity test
Acute oral toxicity (AOT) test was performed by Korea Testing and Research Institute (KTR; Hwasun-gun, Jeollanam-do, Korea). The AOT test was performed according to OECD guidelines (2008) for testing of chemicals. Briefly, 12 Crl:CD(SD) female rats aged 9-10 weeks were divided into four groups of 3 rats each. Each group was orally dosed with 300 or 2,000 mg of freeze-dried Streptococcus powder in 10 ml sterilized water per kg body weight. Then, the mortality, signs of toxicity, and body weight changes were monitored for 14 days. Finally, 100 ml of isoflurane was injected to euthanize the rats, and an autopsy for the examination of organs was performed on the 14th day.

| Hemolytic property and enzymatic activities
A hemolytic activity is generally caused by hemolysin produced by bacteria and it induces the lysis of red blood cells that result in mild to severe infection by a variety of pathogens (Nodzo et al., 2014).
Thus, hemolytic activity is an important criterion of safety in selecting probiotics (Sorokulova et al., 2008). In this study, S. thermophilus IDCC 2201 were found to be negative for hemolytic activity. In contrast, Staphylococcus aureus ATCC 25923 as a positive control clearly induced a β-hemolysis on sheep blood agar ( Figure S1).
β-Glucosidase of lactic acid bacteria hydrolyzes glucose conjugates from plants, generating a variety of plant secondary metabolites in the colon. These resulting metabolites function as health-promoting substances (e.g., antioxidants) (Michlmayr et al., 2013). However, it was also reported to produce potential toxins (e.g., deoxynivalenol) or carcinogenic compounds in rare cases (Cole & Fuller, 1987). Streptococcus thermophilus IDCC 2201 tested in this study was found to have no activity of β-glucosidase (Table 1 and Figure S2). Another safety concern is that β-glucuronidase produced by microorganisms can develop toxic steroidal (e.g., estrogen) or carcinogenic compounds and thereby increase risk for colorectal cancer (Kim & Jin, 2001). Preferably, S. thermophilus IDCC 2201 in this study have no activity of β-glucuronidase (Table 1 and Figure S2). Thus, S. thermophilus IDCC 2201 is unlikely to produce toxic chemicals during the fermentation.

| Determination of MICs and wholegenome analysis
Streptococcus thermophilus IDCC 2201 were evaluated whether they are susceptible to a variety of antibiotics, which are typically used to treat enterococcal infections (EFSA, 2012). In this test, nine antibiotics were used as follows: ampicillin, vancomycin, gentamicin, kanamycin, streptomycin, erythromycin, clindamycin, tetracycline, and chloramphenicol. In results, S. thermophilus IDCC 2201 was susceptible to all the antibiotics tested (Table 2). In accordance with MIC tests, the whole-genome analysis of S. thermophilus IDCC 2201 (1.79 Mb) indicated that it has no gene or similar gene characterized as antibiotic resistance gene was found in this genome ( Figure S3).
In previous study (Rizzotti et al., 2009) genes tet(S), tet(M), and tet(L). Meanwhile, gene coding for hyaluronic acid capsule as a virulence factor was found with the BLASTP parameters in this genome. However, this factor is frequently found in the genomes of many other S. thermophilus strains (Wessels et al., 1991). Additionally, 81 mobile elements such as transposase were found in this genome.

| Biogenic amine production
Biogenic amines (BAs) are organic compounds with low molecular weight and can be produced by lactic acid bacteria harboring amino acid decarboxylase gene (e.g., tdc) (Barbieri et al., 2019).
They are present in various fermented foods and have a lot of biological activities such as essential psychoactive or vasoactive effects (Erdag et al., 2018). However, some of these amines are so bioactive that they can cause various adverse effects in human health (Spano et al., 2010). For example, they can affect the vascular system as well as the central nervous system, resulting in cardiovascular hypertension, vomiting, and headache. Furthermore, some BAs have the potential to be converted into powerful carcinogens (e.g., nitrosamine) (Lonvaur-Funel, 2001). Among BAs, histamine and tyramine are considered the most important in food safety, and they are responsible for scombroid fish poisoning (histamine intoxication), food-induced migraine, and hypertensive crisis. (Izquierdo-Pulido et al., 1996). Putrescine has been implicated in cell proliferation and has been linked to cancer. In this study, tyramine, histamine, putrescine, 2-phenethylamine, and cadaverine were not detected in S. thermophilus IDCC 2201 (Table 3).
Thus, it was concluded that S. thermophilus IDCC 2201 is considered safe in terms of BAs production.

| L/D-lactate formation
Lactate can be produced either via homofermentative or heterofermentative pathway by lactic acid bacteria (Drinan et al., 1976).
Although L-lactate is predominantly produced from pyruvate by L-lactate dehydrogenase, D-lactate can be produced by lactic acid bacteria, depending on the strains and environmental conditions (Zuniga et al., 1993). D-lactate is so difficult to be metabolized that it can be accumulated in humans, causing D-lactic acidosis (Petersen, 2005;Schiraldi et al., 2003). Therefore, D-lactate formation by lactic acid bacteria is also an important criterion for safety evaluations. In this study, 20.1 g/L (99.85%) of L-lactate and 0.03 g/L (0.15%) of D-lactate were produced by S. thermophilus IDCC 2201 (Table 4). Thus, the results indicate that the formation of D-lactate by this strain is negligible.

| Single-dose acute oral toxicity study
To evaluate the safety of S. thermophilus IDCC 2201 in vivo, a singledose acute oral toxicity tests were performed with 4 test groups (Table 5). The observation for 14 days indicated that a single oral dose of 2.3 × 10 11 -1.6 × 10 12 CFU and 2.2 × 10 11 -1.6 × 10 12 CFU of S. thermophilus IDCC 2201 did not cause mortality and toxicity in 9 week-aged rats and 10 week-aged rats, respectively. Additionally, there were no significant changes in the appearances (e.g., skin, fur,

| CON CLUS ION
The safety assessment of S. thermophilus IDCC 2201 used for probiotics manufacturing in Ildong Bioscience was performed through in vitro and in vivo tests. In results, this strain was found to be negative for hemolytic activity, and no endogenous enzymes which make toxic substances were found. In addition, the strain was susceptible to nine antibiotics by MICs test and did not have any antibiotic resistance gene as analyzed in the whole-genome sequencing. Additionally, biogenic amines were not produced and D-lactate formation was negligible after fermentation. In acute oral toxicity test, no hazardous phenomenon was observed in rats. Therefore, we report that S. thermophilus IDCC 2201 is considered to be safe for human consumption as probiotics. Finally, these findings contribute to screening for safe potential probiotics and for safe starters in the dairy industries. Additionally, we thank Jiah Yoo for design and quality improvement of graphical abstract.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

E TH I C A L S TATEM ENT
The animal experiments in this study were conducted by Korea