**2. Selection of probiotic bacteria**

Probiotics are living, health-promoting microorganisms that are incorporated into various kinds of foods. Although there has been a growing interest in using LAB isolated both from naturally fermented products and humans for health benefits (Lim & Im, 2009), the strains should preferably be of human origin and possess a Generally-Recognized-As-Safe status (Rönkä et al., 2003).

In order to exhibit their benecial effects, probiotic bacteria need to survive during the foodmanufacturing process and in human ecosystem conditions; therefore it is important to investigate bacterial behavior under conditions which mimic the GIT (Zago et al., 2011; Lo Curto et al., 2011). Stresses to microorganisms begin in the mouth, with the lysozymecontaining saliva; continue in the stomach, which has a pH between 1.5 and 3.0; and go on to the upper intestine, which contains bile (Corzo & Gilliland, 1999). Acid and bile tolerances are two fundamental properties that indicate the ability of a probiotic microorganism to survive the passage through the GIT, resisting the acidic conditions in the stomach and the bile acids at the beginning of the small intestine (Prasad et al., 1998; Park et al., 2002). To evaluate the probiotic survival in the GIT, several *in vitro* static models of digestion have been developed (Kitazawa et al., 1991; Charteris et al., 1998). One of them is the gastric–small intestinal system TIM-1 (Minekus et al., 1995), which consists of four serial compartments simulating the stomach and the three segments of the small intestine: the duodenum, jejunum, and ileum. Another one, the TIM-2 model, is a more sophisticated *in vitro* model of fermentation in the proximal large intestine. It consists of a series of linked glass vessels containing exible walls which allow simulation of peristalsis (De Preter et al., 2011). The simulator of the human intestinal microbial ecosystem (SHIME) was developed to simulate the entire human gastrointestinal system (Molly et al., 1993). SHIME consists of a series of ve temperature- and pH-controlled vessels that simulate the stomach; small intestine; and ascending, transverse and descending colon, respectively. The SHIME harbors a microbial community resembling that from the human colon both in fermentation activity and in composition (De Preter et al., 2011). Yet another model of the digestive system has been developed by such as TNO to mimic human physiological conditions in the stomach and small intestine (Blanquet et al., 2001). The major limitations of those systems is that digestion products are not removed during the incubation, and they may have a potential inhibitory effect on enzyme activities and on probiotic survival (Pitino et al., 2010). Furthermore, such systems ignore key GIT physical processes, including the temporal nature of gastric and duodenal processing, structure of food, pattern of mixing, particle size reduction and shear, which all affect the digestion rate (Shah 2000; Sumeri et al., 2008).

Effects of probiotics are strain specific. Strain identity is important in order to link a strain with a specific health effect, as well as to enable accurate surveillance and epidemiological studies (Ganguly et al., 2011). It is very important to be able to identify specically and unambiguously the particular probiotic LAB strains from clinical fecal and intestinal biopsy specimens and from food samples (Tilsala-Timisjärvi & Tapanialtossava, 1998). Identication of bacterial species and strains from commercialized probiotics has been conducted mostly using molecular methods (Holzapfel et al., 2001; Schillinger et al., 2003; Huys et al., 2006; Sheu et al., 2009).

244 Lactic Acid Bacteria – R & D for Food, Health and Livestock Purposes

probiotics, and health effects of probiotic bacteria.

**2. Selection of probiotic bacteria** 

(Rönkä et al., 2003).

Several aspects, including general, functional and technological characteristics, have to be taken into consideration while selecting probiotic strains (Sanders & Huis in't Veld 1999; Šušković et al., 2001). This chapter includes selection criteria of bacteria as probiotics, technological usage of probiotics, new approaches for enhancing the performance of

Probiotics are living, health-promoting microorganisms that are incorporated into various kinds of foods. Although there has been a growing interest in using LAB isolated both from naturally fermented products and humans for health benefits (Lim & Im, 2009), the strains should preferably be of human origin and possess a Generally-Recognized-As-Safe status

In order to exhibit their benecial effects, probiotic bacteria need to survive during the foodmanufacturing process and in human ecosystem conditions; therefore it is important to investigate bacterial behavior under conditions which mimic the GIT (Zago et al., 2011; Lo Curto et al., 2011). Stresses to microorganisms begin in the mouth, with the lysozymecontaining saliva; continue in the stomach, which has a pH between 1.5 and 3.0; and go on to the upper intestine, which contains bile (Corzo & Gilliland, 1999). Acid and bile tolerances are two fundamental properties that indicate the ability of a probiotic microorganism to survive the passage through the GIT, resisting the acidic conditions in the stomach and the bile acids at the beginning of the small intestine (Prasad et al., 1998; Park et al., 2002). To evaluate the probiotic survival in the GIT, several *in vitro* static models of digestion have been developed (Kitazawa et al., 1991; Charteris et al., 1998). One of them is the gastric–small intestinal system TIM-1 (Minekus et al., 1995), which consists of four serial compartments simulating the stomach and the three segments of the small intestine: the duodenum, jejunum, and ileum. Another one, the TIM-2 model, is a more sophisticated *in vitro* model of fermentation in the proximal large intestine. It consists of a series of linked glass vessels containing exible walls which allow simulation of peristalsis (De Preter et al., 2011). The simulator of the human intestinal microbial ecosystem (SHIME) was developed to simulate the entire human gastrointestinal system (Molly et al., 1993). SHIME consists of a series of ve temperature- and pH-controlled vessels that simulate the stomach; small intestine; and ascending, transverse and descending colon, respectively. The SHIME harbors a microbial community resembling that from the human colon both in fermentation activity and in composition (De Preter et al., 2011). Yet another model of the digestive system has been developed by such as TNO to mimic human physiological conditions in the stomach and small intestine (Blanquet et al., 2001). The major limitations of those systems is that digestion products are not removed during the incubation, and they may have a potential inhibitory effect on enzyme activities and on probiotic survival (Pitino et al., 2010). Furthermore, such systems ignore key GIT physical processes, including the temporal nature of gastric and duodenal processing, structure of food, pattern of mixing, particle size reduction and shear, which all affect the digestion rate (Shah 2000; Sumeri et al., 2008).

Verdenelli et al. (2009) investigated the probiotic potential of 11 *Lactobacillus* strains isolated from the faeces of elderly Italians. For this purpose, the researchers identied the *Lactobacillus* strains and examined them for resistance to gastric acidity and bile toxicity, adhesion to HT-29 cells, antimicrobial activities, antibiotic susceptibility and plasmid prole. They also examined the survival of the strains as they moved through the human intestine in a 3-month human feeding trial. According to the results, *L. rhamnosus* IMC 501 and *L. paracasei* IMC 502 present favourable strain-specic properties for their utilisation as probiotics in functional foods. Both *in vitro* and *in vivo* studies confirm the high adhesion ability of *L. rhamnosus* IMC 501 and *L. paracasei* IMC 502, used in combination, indicating that the two bacterial strains could be used as health-promoting bacteria.

Başyiğit Klç & Karahan (2010) isolated one hundred seven strains of human originated LAB identified by 16S rRNA analysis and examined them for resistance to acidic pH, bile salts and antibiotic susceptibility. They found that *L. plantarum* (AA1–2, AA17–73, AC18–88, AK4–11, and AK7–28), *L. fermentum* (AB5–18, BB16–75, and AK4–180), *Enterococcus faecium* (AB20–98 and BK11–50) and *E. durans* (AK4–14 and BK9–40) are potentially good probiotic candidates for use as health-promoting bacteria. In another study, the *L. plantarum* strains were examined for resistance to gastric acidity in simulated gastric juice at pH 2.0, 2.5, 3.0 and 3.5; 0.4% phenol; production of H2O2; adhesion to Caco-2 cell line; and antimicrobial activities. The researchers determined that the artificial gastric juice, even at pH 2.0, did not significantly change the viability of the cultures, and all *L. plantarum* strains showed good resistance to 0.4% phenol. They also reported antimicrobial activity and good adhesion of *L. plantarum* strains to Caco-2 cells. The researchers concluded that all of the strains showed probiotic properties, but *L. plantarum* AB6-25, AB7-35, AA13-59, AB16-65, BC18-81 and AK4- 11 were the best potential probiotic strains for human use, given their ability to survive in gastric conditions, strong resistance to phenol, and the ability to adhere to the Caco-2 cell line (Başyiğit Klç et al., 2011a).

Lo Curto et al. (2011) investigated the survival of three commercial probiotic strains (*L. casei* subsp. *shirota*, *L. casei* subsp. *immunitas*, *L. acidophilus* subsp. *johnsonii*) in the human upper GIT. They used a dynamic gastric model (DGM) of digestion followed by incubation under duodenal conditions. The DGM is a computer-controlled gastric model which incorporates the chemical, biochemical, physical environment and processes of the human stomach; the model is based on kinetic data derived from the Echo planar-MRI and data on the rates of GI digestion obtained from human studies (Marciani et al., 2001; 2003; 2005; 2006). The researchers used water and milk as food matrices, and survival was evaluated in both

logarithmic and stationary phases. The researchers found that the % of recovery in the logarithmic phase ranged from 1.0% to 43.8% in water for all tested strains, and from 80.5% to 197% in milk. They observed higher survival rates in the stationary phase for all strains. *L*. *acidophilus* subsp. *johnsonii* showed the highest survival rate in both water (93.9%) and milk (202.4%).

Highlights in Probiotic Research 247

colonization of the gut epithelium prolongs the time for microorganisms to influence the immune system and microbiota of the host (Forestier et al., 2001). HT-29 and Caco 2 cells, the two colonic adenocarcinomas, are derived from human intestinal epithelium. Because they have structural and functional features of normal human enterocytes, they have been extensively used as *in vitro* models in the study of human enterocytic function (Moussavi &

The ability of probiotic bacteria to adhere to Caco-2 cells can be determined by plate counting or real time PCR (Matijasic et al., 2003; Candela et al., 2005). Nawaz et al. (2011) used both of these methods and did not find a statistically significant difference. Gaudana et al. (2010) investigated the ability of four different isolates (*L. plantarum* CS23, *L. rhamnosus* CS25, *L. delbrueckii* M and *L. fermentum* ASt1) and two standard strains (*L. plantarum* ATCC 8014 and *L. rhamnosus* GG) to stimulate three types of cells (Caco-2 cells, human peripheral blood mononuclear cells [PBMC] and THP-1 cells). The researchers reported that child faecal isolate CS23 showed high binding ability, high tolerance to acidic pH and bile salts, and signicant immunomodulation; therefore they concluded that CS23 can be a good potential probiotic candidate. Duary et al. (2011) determined the colonization potentials of five human faecal *L. plantarum* isolates to the Caco-2 cells. Based on direct adhesion to epithelial cells, *L*. *plantarum* Lp91 was the most adhesive strain to the Caco-2 cell lines, with adhesion values of approximately 10.2%. They also mentioned that the percentage of adhesion to Caco-2 and HT-29 cell lines was higher among the strains isolated from the human faecal

The use of starter cultures in the production of fermented food is necessary for guaranteeing safety and standardizing properties. LAB functions primarily to drop the pH of the batter; lower pH a) promotes product safety by inactivating pathogens, b) creates the biochemical conditions to attain the nal sensory properties through modication of the raw materials, and c) improves the product stability and shelf life by inhibiting undesirable changes brought about by spoilage microorganisms or abiotic reactions (Ammor & Mayo, 2007).

Functional starter cultures are dened as microbes that possess at least one inherently functional property aimed at improving the quality of the end product (De Vuyst, 2000). The use of probiotics in food has reinforced the acclaimed healthy properties and given rise to an increased consumption of these products in Europe and the USA (Kristo et al., 2003). Probiotics have been evaluated as functional starter cultures in various types of fermented food products such as yoghurt, cheese, dry sausage, salami, and sourdough. They have also been studied in therapeutic preparations to assess their positive effects on physico-chemical properties of foods and their impact on the nutritional quality and functional performance

Fermented dairy products are widely-accepted, healthy food products and valued components of diets. The incorporation of probiotic bacteria as adjuncts in various fermented milk products is currently an important topic with industrial and commercial

samples and buffalo milk than that which had been isolated from cheese.

**3. Technological usage of probiotics** 

of the raw material (Knorr, 1998; Rodgers, 2008).

Adams, 2009).

The safety of probiotic bacteria must be carefully assessed, with particular attention to transferable antibiotic resistance (Mathur & Singh, 2005). In the last decade, increasing concern has arisen about the safe use of LAB cultures for food and feed applications, in light of the latest knowledge about their possible role as an antibiotic-resistant gene reservoir. Particular concern is due to evidence of widespread occurrence in this bacterial group of conjugative plasmids and transposons (Clementi & Aquilanti, 2011). It is known that lactobacilli have a high natural resistance to bacitracin, cefoxitin, ciprofloxacin, fusidic acid, kanamycin, gentamicin, metronidazole, nitrofurantoin, norfloxacin, streptomycin, sulphadiazine, teicoplanin, trimethoprim/sulphamethoxazole, and vancomycin (Danielsen & Wind, 2003).

One of the primary benets associated with probiotic bacterial cultures is that they can exclude pathogenic bacteria from the small and large intestine (Kos et al., 2008). Another benefit is that in food products, antimicrobial activity of probiotic bacteria may contribute to an improvement in the quality of fermented foods. This may result from control of spoilage and pathogenic bacteria, extension of shelf life, and improvement of sensory quality (Wei et al., 2006; Siripatrawan & Harte, 2007). Kos et al. (2008) used overnight cultures and cell-free supernatants of the three probiotic strains *L. acidophilus* M92, *L. plantarum* L4, and *E. faecium* L3 for determining the antagonistic effect against *Listeria monocytogenes*, *Salmonella typhimurium*, *Yersinia enterocolitica*, and *Acinetobacter calcoaceticus*. The researchers determined that probiotic strains *L. acidophilus* M92, *L. plantarum* L4, and *E. faecium* L3 demonstrated anti-*Salmonella* activity. *L. acidophilus* M92 was also shown to have antilisterial activity, as demonstrated by *in vitro* competition test.

Production of antimicrobial compounds, which may take part in the inhibition of intestinal pathogens, is another criterion for classifying a potentially probiotic bacteria **(**Hutt et al., 2006). The inhibition of pathogenic microorganisms by selected probiotic strains may occur via a) production of antibiotic-like substances, b) bacteriocins and bacteriocin-like inhibitory substances such as acidophilin and reuterin, c) lowering of pH by producing organic acids such as acetic, lactic and phenyllactic acid, d) production of hydrogen peroxide and short chain fatty acids, e) decreasing the redox potential, and f) consumption of available nutrients (Holzapfel et al., 1995; Ouwehand, 1998; Tharmaraj & Shah, 2009).

The ability of LAB to adhere to epithelial cells and mucosal surfaces is thought to be an important property of many bacterial strains used as probiotics (FAO/WHO, 2001). Cell adhesion is a complex process involving contact between the bacterial cell membrane and interacting surfaces. Difficulties experienced in studying bacterial adhesion *in vivo*, especially in humans, have stimulated interest in the development of *in vitro* models for preliminary screening of potentially adherent strains (Duary et al., 2011). Attachment and colonization of the gut epithelium prolongs the time for microorganisms to influence the immune system and microbiota of the host (Forestier et al., 2001). HT-29 and Caco 2 cells, the two colonic adenocarcinomas, are derived from human intestinal epithelium. Because they have structural and functional features of normal human enterocytes, they have been extensively used as *in vitro* models in the study of human enterocytic function (Moussavi & Adams, 2009).

The ability of probiotic bacteria to adhere to Caco-2 cells can be determined by plate counting or real time PCR (Matijasic et al., 2003; Candela et al., 2005). Nawaz et al. (2011) used both of these methods and did not find a statistically significant difference. Gaudana et al. (2010) investigated the ability of four different isolates (*L. plantarum* CS23, *L. rhamnosus* CS25, *L. delbrueckii* M and *L. fermentum* ASt1) and two standard strains (*L. plantarum* ATCC 8014 and *L. rhamnosus* GG) to stimulate three types of cells (Caco-2 cells, human peripheral blood mononuclear cells [PBMC] and THP-1 cells). The researchers reported that child faecal isolate CS23 showed high binding ability, high tolerance to acidic pH and bile salts, and signicant immunomodulation; therefore they concluded that CS23 can be a good potential probiotic candidate. Duary et al. (2011) determined the colonization potentials of five human faecal *L. plantarum* isolates to the Caco-2 cells. Based on direct adhesion to epithelial cells, *L*. *plantarum* Lp91 was the most adhesive strain to the Caco-2 cell lines, with adhesion values of approximately 10.2%. They also mentioned that the percentage of adhesion to Caco-2 and HT-29 cell lines was higher among the strains isolated from the human faecal samples and buffalo milk than that which had been isolated from cheese.
