Safety Assessment of Indigenous Probiotic Strain Lactobacillus plantarum Mut-7 Using Sprague Dawley Rats as a Model

Corresponding Author: Endang S. Rahayu Department of Food and Agricultural Product Technology, Faculty of Agricultural Technology, Universitas Gadjah Mada, Yogyakarta, Indonesia Email: endangsrahayu@ugm.ac.id Abstract: Lactobacillus plantarum Mut-7 is a probiotic candidate isolated from gatot, traditional fermented cassava from Java, Indonesia. This study aimed to evaluate safety aspects of high dose consumption of L. plantarum Mut-7 (10 CFU/ml/day) on Sprague Dawley rats for 21 days. Twenty four female rats were randomly divided into 4 groups; initial condition group (P.0), control group (P.1), skim milk group (P.2) and probiotics group (P.3). All groups followed adaptation phase of 7 days, followed by treatment phase of 21 days for P.1, P.2 and P.3. The results showed that supplementation of high dose of L. plantarum Mut-7 did not have detrimental effects on general health, organ weight, hematology and histology parameters of treated rats. Feed intake and body weight showed no significant difference between groups. L. plantarum Mut-7 can survive in gastrointestinal tract of rats, resulting in an increased population of L. plantarum in the fecal matter and the digesta of treated rats. Bacterial translocation of L. plantarum Mut-7 was not detected in the blood and organ of treated rats as confirmed by rep-PCR with BOXAIR primer and further 16S RNA gene sequencing analysis. Twenty-six isolates from blood and organs of treated rats had low-level similarity (<75%) to that of L. plantarum Mut-7, with 10 isolates were further analyzed and found that none of them belong to L. plantarum. Although this study was limited to the use of animal study, the findings are useful to support the safety assessment of the use of L. plantarum Mut-7 as a probiotic according to the abovementioned parameters.


Introduction
Probiotics are live organisms that, when consumed in sufficient amounts, confer beneficial effects on the host (FAO/WHO, 2002). Most probiotics belong to Lactic Acid Bacteria group (LAB) such as Lactobacillus spp. and some other groups such as Bifidobacterium group, Saccharomyces, Streptococcus and Lactococcus. However, some species of Lactobacillus and Bifidobacterium are not natural microflora and can adapt to conditions in the gastrointestinal tract to withstand the acidic conditions of the stomach/digestive tract bile salts and to survive pathogens (Emmawati, 2014;Rahayu et al., 2015).
In recent years, Lactic Acid Bacteria (LAB) has received increasing attention due to their benefits for humans. LAB has long been used for traditional food fermentation . A study by Rahayu (2003) reported that 109 of 194 new LAB strains isolated from 21 different traditional Indonesian fermented food were classified as Lactobacillus, particularly dominated by Lactobacillus plantarumpentosus. L. plantarum Mut-7 is a potential local probiotic strain from gatot, a fermented cassava product that is widely consumed in Java, Indonesia. Subsequently, Rahayu et al. (2019a) also mentioned that Indonesian adults tend to have a very high prevalence of 95% of L. plantarum which could be linked to consumption of fermented foods as part of Indonesian diet. Therefore, L. plantarum can be considered as a potential probiotic for Indonesian population.
Multiple studies have tested favorable characteristics of L. plantarum Mut-7 for probiotic functionality such as it may provide health effects including lowering cholesterol body improve immune system through an increase in phagocytic activity of macrophages, is able to survive in a pH 2, has the ability to inhibit the growth of pathogenic bacteria such as Escherichia coli, Clostridia and Salmonella in the feces (Ngatirah et al., 2000;Lestari et al., 2013).
Lactobacillus spp. used as probiotics are generally classified as GRAS (Generally Recognized as Safe) (Gharaei and Eslamifar, 2011). However, probiotic effects are strain-specific which consequently needs a strain by strain basis for its safety assessment (Kemgag et al., 2014). Therefore, regardless of the long use of L. plantarum in food fermentation without causing adverse health effects, its safety aspects for its potential use as probiotic needs to be studied before further use in commercial scale. Several significant risks that could arise when probiotics are supplemented in humans are potential transmigration, bacteremia, endocarditis potential, gastrointestinal toxicity and antibiotic resistance gene transfer. In recent years, some species of genus Lactobacillus, Leuconostoc, Pediococcus, Enterococcus and Bifidobacterium are often isolated from different types of wound infection. Lactobacillus rhamnosus, L. acidophilus, L. plantarum, L. casei, L. paracasei, L. salivarus, L. lactis and Leuconostoc mesenteroides are some examples of Lactobacilli isolated from bacterial endocarditis. Additionally, Pediococcus acidilactici, Bifidobacterium eriksonii and Bifidobacterium adolescentis have been isolated from bloodstream infection and local infection (Ishibashi and Yamazaki, 2001).
Amplification of bacterial DNA with repetitive PCR (rep-PCR) has been recognized as one of a simple PCR technique with a discriminatory power, low cost, very suitable for the identification of bacterial strains and reliable method for classifying and typing gram-negative and gram-positive bacteria (Gevers, 2001). Berthier and Ehrlich (1998) stated that the intergenic spacer region (RISA/ITS-PCR) can be used to discriminate between species of Lactobacillus with close genetic relationship, namely L. paraplantarum, L. pentosus and L. plantarum using a primer with a target gene in 16S/23S rRNA spacer region. The principle of RISA-PCR technique is through amplification in the intergenic small (16S) and sub-units of large rRNA gene (23S) in the rRNA operon as a target region oligonucleotide primer (Yu and Mohn, 2001). Furthermore, 16S rRNA is a gene that is generally targeted for DNA sequencing analysis. This technique has a high discriminatory power of 100% type ability and good reproducibility. A similarity score of <97% is interpreted that an isolate represents a new species, while >97% similarity score can be either a new species or an alternative to a new cluster in the previous taxonomy (Janda and Abbott, 2007). Additionally, if an isolate has a minimum 99% similarity, it is identified as a similar strain or species to that of reference species of the gene bank.
The aim of this study, was to evaluate the safety of L. plantarum Mut-7, using Sprague Dawley rats as a model, by orally introducing high doses (10 11 CFU) of L. plantarum Mut-7 for 21 days. Some parameters were analyzed including body weight change, organ weight, hematological and histological parameters, such as SGOT, SGPT, morphology of gastrointestinal tract and bacterial analysis in the organs (blood, lungs, heart, spleen, liver and kidneys). The possibility of translocation of L. plantarum Mut-7 to the blood and organs of the rats was also investigated using Rep-PCR methods with BOXAIR primer and further 16S RNA gene sequencing analysis.

Bacterial Strain and Cell Production
L. plantarum Mut-7 was obtained from the Food and Nutrition Culture Collection (FNCC), Center for Food and Nutrition Studies, Universitas Gadjah Mada, Yogyakarta, Indonesia. Biomass production was carried out by inoculation of L. plantarum Mut-7 into Peptone Glucose Yeast extract broth and incubation for 48 h. The culture was then centrifuged and the pellet was resuspended in 10% skim milk solution at concentration of 10 11 CFU/ml.

Experimental Design
This research adhered to the Declaration of Helsinki guideline and had been approved by the Medical and Health Research Ethics Committee (MHREC) according to ethical clearance number 331/KEC-LPPT/X/2015. Twenty four female Sprague Dawley rats (albino Norway/Rattus norvegicus) aged 8 weeks with body weight 200+20 g were housed individually in stainless steel cage. A 12 h light-dark cycle and a temperature 25C were maintained in this study. Animals were fed with 15 g of AIN-93 M without tertbutylhydroquinone diet and water ad libitum (Reeves et al., 1993 with modification).
The rats were adapted for one week in experimental condition and subsequently assigned into four groups. First group, named as initial condition group (P.0) followed only adaptation phase of 7 days and were humanely euthanized on day 8. The other three groups were control group (P.1), skim milk group (P.2) and probiotics group (P.3) followed adaptation phase of 7 days and 21 days of treatment fed with standard based diet. Group P.1 did not receive any other treatment than standard feeding, while P.2 and P.3 were supplemented with skimmed milk 1ml 10% and 1ml cell suspension of L. plantarum Mut-7 10 11 CFU/rats/day in skimmed milk 10%, respectively. On day 29, the rats were euthanized Anatomy of organ from each rat was checked and recorded. Organ weight index was expressed as actual organ weight (mg) divided by body weight (g). Feed intake and body weight were measured weekly. Furthermore, activity and behavior of each rat were observed and recorded daily. Their blood and tissue (kidney, liver, lung, spleen and heart) samples were collected aseptically for further analysis.

Biochemical, Hematological and Gut Histological Analyses
Blood samples were analyzed for Serum Glutamic Oxaloacetic Transaminase (SGOT) and Serum Glutamic Pyruvic Transaminase (SGPT). Blood serum was obtained by centrifugation and reacted with enzyme reagents and reagent starter. Absorbance was measured using MicroLab 300 by spectrophotometric method at wavelength 340 nm (Steppe et al., 2014).
Blood samples were obtained by orbital sinus on the eyes and collected into an EDTA treated tube for further hematological analysis, including White Blood Cells (WBC) (leukocytes, lymphocytes and neutrophils), number of Red Blood Cells (RBC), Platelet Count (PLT), Hematocrit (HCT), Hemoglobin (HGB), Mean Corpuscular Volume average (MCV), number of mean corpuscular hemoglobin (MCH) and concentration of Mean Corpuscular Hemoglobin (MCHC). Hematological tests aimed to detect abnormal function of red blood cells (anemia and leukemia), to detect systemic disorders (liver and kidney) and the presence of infection. Hematology analyzer is based on the principle of flow cytometer by measuring the number and properties of the cells coated by the flow of liquid through the narrow opening. Thousands of cells flowed through the gap allowing the cells to pass one by one, then the number of cells and the size is calculated .
Ileum, caecum and colon morphologies were analyzed qualitatively using a microscope and by measuring epithelial height, villous height and mucosa thickness. The tissue samples were stained using Hematoxylin Eosin (HE) (Shu et al., 2000).

Microbial Analyses of Fecal Matter, Digesta, Blood and Organs in Rats
Analysis of LAB and L. plantarum population in the digesta and feces referred to the method by  with modification, Yakabe et al. (2009 and Rahayu et al. (2019b).
To evaluate possible bacterial translocation in blood and organs of the rats, 0.1 mL sample from the liver, kidneys, lungs, heart, spleen and blood of each rat was enriched into MRS broth medium for 1 h and 24 h. Growth with streak plate method on MRS agar and LPSM were incubated at 37C for 48 h. Colonies appearing on MRS were suspected as LAB, while those appeared on LPSM were suspected as L. plantarum (Zhou et al., 2000 with modification;Rahayu et al., 2019b).

Isolation of Bacterial Genome DNA
Isolation of bacterial DNA from blood and organs of rats and nine strains of L. plantarum was performed using DNA mini kit from GeneaidPrestoTM. The isolated DNA is stored in a 1.5 mL microtube at -20°C.
The process of DNA sequencing was done by 1st BASE DNA Sequencing Service, Malaysia. The results of subsequent determination of DNA base sequence were read using DNA baser program. Furthermore, the matching process was done by selecting menu Alignment with Bioedit. Matching results were used to search for DNA base sequence of genes that are similar or similar as DNA base sequence data for genes from an international gene bank database through the NCBI BLAST method. Based on phylogenetic trees, DNA sequence was used as data reference. DNA sequence data can be retrieved from international databases of NCBI. The selected bacterial DNA sequence was stored as reference isolates using MEGA7 format (version 7.0.14) (Tamura et al., 2011) and followed by neighbor joining method (Saitou and Nei, 1987).

Population of LAB in Digesta and Feces
After consumption of L. plantarum Mut-7 for 21 days, LAB population in P.3 (probiotic group) showed an increase to max 10 8 CFU/ml in both feces and digesta. These results are in agreement with those of Rahayu et al. (2019b), where the numbers of LAB in digesta and feces were 10 7 CFU/ml and 10 8 CFU/ml after administration of L. plantarum Dad-13. These findings indicated that L. plantarum survive gastrointestinal tract into feces. Ability to survive in gastrointestinal tract is an important characteristic of probiotic to be able to give beneficial effects to the host as defined by FAO/WHO (2002).
Survival rates of L. plantarum Mut-7 in gastrointestinal tract can be seen in Table 1. L. plantarum was only detected in probiotic group (P.3) with an average growth of 10 4 -10 5 CFU/ml and promotes the growth of LAB in the digestive tract of probiotic group. These results are consistent with studies by Rahayu et al. (2015) where, in in vitro testing of L. plantarum Mut-7, it was able to survive in pH 2 gastric acid, a 3% concentration of bile salts and to inhibit pathogenic bacteria.

General Health Indicator and Biochemical and Hematological Parameters
All rats in group P.1, P.2 and P.3 survived until day 28 of the experiment. Physiological conditions were consistent and showed no adverse effects during the consumption period of probiotics. This consistency in physiological conditions was used as an indicator of infection caused by microbial metabolite production and release of cytokines which may affect central nervous system (Kanra et al., 2006).
The results of body weight and feed intake showed no abnormal changes were recorded. The organ weight index ratio can be seen in Table 2. The mean index of organ weights between different groups did not experience a significant difference (P>0.05) and thus did not result a significant increase in organ weight ratio. It shows that the administration of probiotics for 28 days did not lead to a bacterial infection in rats' organs The results of hematological tests showed no significant increase (P>0,05) in any of the levels of hematologic parameters after ingestion of L. plantarum Mut-7 as shown in Table 3. SGOT and SGPT were two parameters to assess hematological condition of rats. SGOT and SGPT were higher than normal values in all groups, showing that elevated levels of SGOT and SGPT had allegedly caused hepatocyte damage which had occurred since initial conditions. Oxidative stress can also lead to high levels of SGOT and SGPT due to an accumulation of metabolites in the body disturbing the balance of free radicals and antioxidants. It can be expected that the increase in SGOT and SGPT levels above the normal values was not caused by the consumption of L. plantarum Mut-7 at a high dose.    0,56 a +0,03 0,50 a +0,11 0,77 a +0,14 2,52 a +0,04 0,79 a +0,10 Note: Different superscript notation in the same column indicate significant difference (P>0.05) + SEM = standard error  (Steppe et al., 2014)), WBC (leukocytes), LYM (lymphocytes), NEUT (neutrophil), RBC (red blood cells), PLT (platelet count), hematocrit (HCT), hemoglobin (HGB), the mean corpuscular volume average (MCV), the number of mean corpuscular hemoglobin (MCH) and the concentration of mean corpuscular hemoglobin (MCHC)

Gut Histology Analysis
The results showed that after consumption of L. plantarum Mut-7, there was no inflammation which subsequently triggered further infection. An inflammation is characterized by changes in the structure of ileum such as reddish spots, enlarged globules of mucosa, thinning of the mucosa, lengthened intestinal villi, thickening of bowel wall and widening of space between villi and crypts (Shackelford and Elwell, 1999). Figure 1 shows epithelial cells in healthy status. The mechanism of bacterial attachment (adhesion) first took place in the mucous layer as a bond between the structure of the bacteria and specific receptors on the surface of the epithelial cells in which glycoconjugates attach to the side of the oligosaccharide chains of membrane microvilli (Ouwehand and Salminen, 2003). The ability of a bacteria strain to attach to mucosal layer may vary, depends on the strain type. Pathogenic bacteria causing erosion and damage to the mucosal lining of the epithelial cell surface.

Bacterial Analyses of Blood and Organs of Rats
The results of microbiological analysis (Table 4) using MRS with the addition of CaCO3 detected bacterial colonies in the blood of both P.0 (initial condition group) and P.3 (probiotic group). Moreover, bacterial colonies were also detected in the heart, spleen, liver, lungs and kidneys in all groups. Bacterial colonies detected in MRS media were suspected as LAB. The highest bacterial translocation in organs occurred in liver with 79% occurrence in all rats. Rahayu et al. (2019b) conducted an evaluation for probiotic candidate, L. plantarum Dad-13, in which bacterial colony detected in MRS media were suspected as LAB. According Frizzo et al. (2010), an organ incubated on an agar medium is not free from microbes which renders a possibility for other bacteria to grow. A positive culture in organs and blood of rats treated with probiotic did not indicate that the probiotics are responsible for the translocation because translocation of LAB can also be found even in untreated animals (Trevisi et al., 2007).
Nine strains of L. plantarum from FNCC (Food and Nutrition Culture Collection UGM, Indonesia) were evaluated by RISA-PCR and did not produce a specific fingerprint patterns between the strains ( Fig. 2A.) Limited number of bands and separation of the bands did not clearly show the differences between strains of bacteria. Limited number of bands obtained from RISA-PCR amplification might be due to the long intragenic spacer region on each strain L. plantarum which is relatively similar so it does not form specific fingerprint patterns. Berthier and Ehrlich (1998), states that beside the 16S -23S spacer, intergenic spacer region of the t-RNA genes can be used to discriminate between species Lactobacillus based on the unique fingerprint patterns of t-DNA. However, there is a failure to distinguish very similar species. RISA-PCR method used in this study proved to be unable to distinguish between strains of L. plantarum. In addition, the composition of the target gene amplification (16S-23S ISR) among strains of L. plantarum has higher similarity.
The results of the genotyping of nine strains of L. plantarum using Rep-PCR method showed satisfactory results (Fig. 2B.). This is shown by the results of PAGE PCR in which fingerprint patterns ware different between strains. Differences of bands pattern (fingerprint) at each different strains of L. plantarum in the first group of species indicate that Rep-PCR has a high resolution in the analysis of closely related bacterial strains. This statement is supported by previous studies in distinguishing L. plantarum Dad-13 with other strains of L. plantarum by finger printing patterns using molecular analysis of rep-PCR with primers BOXA1R (Rahayu et al., 2019b). According Brusetti et al. (2008), BOX-PCR has the ability to differentiate strains of the same bacteria even better than ribosomal intergenic spacer analysis (RISA), restriction fragment length polymorphism (RFLP), amplified fragment length polymorphism (AFLP), random amplified polymorphic DNA (RAPD) and in some cases have a better discriminatory power of pulsed field gel electrophoresis (PFGE), although generally less good in terms of reproducibility.  Table 4 shows that in all groups there were no bacterial colonies detected on LPSM. These results indicated that LAB detected in MRS media did not belong to L. plantarum colony, particularly L. plantarum Mut-7 supplemented to probiotic group. These findings are consistent with previous finding in which Lactobacillus do not experience translocation by itself, but can occur when intestinal mucosa is interrupted or when the immune system is unable to control translocation of pathogenic microorganisms into the bloodstream, causing sepsis or local infection (Kemgag et al., 2014). Translocation is associated with immune system and damage to the mucosa. When damage occurs to mucosal epithelial cells, IgA secretion occurs in the immune system, which clinically affects probiotic strain ).

Identification of Bacterial Isolated in Blood and Organs of Rats
Total isolates identified in this study were 33 isolates, consisting of 9, 6, 7 and 11 isolates in the following groups P.0, P.1, P.2, P.3, respectively. However, from 33 isolates analyzed by BOX-PCR, only 26 were successfully amplified by 8% Polyacrylamide Gel Electrophoresis Analysis (PAGE) (Fig. 3). Seven isolates that did not show good results were isolates from group 4.0 (4 isolates) and P1 (3 isolates), due to the reproducibility ability of BOX-PCR method. According to Ranjbar et al. (2014), the reproducibility of Rep-PCR can be influenced by the PCR reagents, thermal cycle and gel electrophoresis conditions. Furthermore, PAGE and silver staining can provide better resolution fragment band than agarose gel but may cause low reproducibility due to gel staining procedure (Brusetti et al., 2008). The number of bands from BOX-PCR analysis of 26 bacterial isolates were 353 bands with an average of 13 bands/isolates. The observation of the pattern of bands (fingerprint) of each isolate was compared to those of L. plantarum Mut-7. None of the isolates in each group showed similarities to L. plantarum Mut-7. The pattern of bands produced by amplification with BOX-PCR produce are quite complex to discriminate on the level of subspecies, while the presence or absence of one or more bands in one cluster generates heterogeneity among strains clearer than by using another primer (Gevers, 2001). BOX-PCR is not only suitable for identification purposes but also for isolates classification, although only isolates with high similarity can be shown. As for the band patterns that are not identical to BOX-PCR molecular typing techniques require further analysis, such as 5 isolates of P.0 (I1BD, I1LR, I3LR, I4LR) and P.1 (II1LR, II3LR, II4LR).
Based on fingerprint patterns by BOX-PCR analysis, 26 isolates of L. plantarum and strain Mut-7 having a similarity level between 54%-90% (Fig. 4). Rep-PCR is used as a genotyping method when the similarity level reaches >95% or 93% (Nucera et al., 2013;Gevers, 2001). Therefore, it can be considered that among 26 isolates there was no resemblance of identical genes to bacterial strain L. plantarum Mut-7.

Identification of Bacteria based on 16S rRNA Gene Sequencing
Results of analysis of the 16S rRNA gene of the 10 isolates were uploaded into the database Genebank (www.ncbi.nlm.nih.gov/BLAST) and analyzed for sequence homology with species of reference. Based on identification of isolates with 16S rRNA gene analysis, 8 of 10 isolates belong to genus Lactobacillus (similarity level 87%-96%), while 2 were Enterococcus. Four isolates were classified as Lactobacillus murinus (similarity >97.5%) and identified as a strain. Other isolates (16) which were not sequenced were analyzed by amplification with BOX-PCR (Fig. 4).
The mechanism of bacterial translocation from genus Lactobacillus could not be determined because the histopathological analysis of intestine did not indicate any presence of inflammation, while interference of microvilli intestine was not analyzed. Therefore, possibility of microbial growth in the gut could not be certainly explained although the number of bacteria in the guts of rats treated with L. plantarum Mut-7 (log 8.16 CFU/g) increased compared to those of control group log 5.88 CFU/g). Previous study found that analysis of whole-genome sequence of L. plantarum JDM1 has 126 genes associated with virulence factors (Zhang et al., 2012). Additionally, Ishibashi and Yamazaki (2001) also stated that there were other virulence factors that are owned by a bacteria of genus Lactobacillus (L. rhamnosus, L. paracasei subsp. paracasei and other strains), that are glycosidase and protease enzyme activity (arylamide) which can cause damage to the human glycoprotein and the synthesis and fibrin clot lysis, wherein the bacterial strain that has the enzyme can be a cause of infection (endocarditis). Therefore, L. plantarum Mut-7 did not have the ability to translocate to organs and blood of rats due to the genes associated with virulence factors do not have the ability to produce these enzymes.

Conclusion
The results of this study showed that supplementation of high doses (10 11 CFU/ml) of L. plantarum Mut-7 for a period of 21 days in Sprague Dawley rats had no deleterious effect on the performance of rat (feed intake, body weight, hematological concentration, physiological and stress markers and gut morphology). L. plantarum Mut-7 can survive in gastrointestinal tracts of rats, resulting in a significant increase of the L. plantarum population in the fecal matter and the digest of treated rats. Additionally, consumption of high doses of L. plantarum Mut-7 for 21 days did not cause bacterial translocation in the organs or blood of the rats. Although this study was limited to the use of animal study, the findings are useful to support the safety assessment of the use of L. plantarum Mut-7 as a probiotic.