inclusion of other cereals in our diets in order to enhance the overall nutritional status as well as address malnutrition across the world. Two of ICRISAT’s mandate crops, sorghum (Sorghum bicolor) and pearl millet (Pennisetum glaucum), can serve this purpose. Sorghum and pearl millet are an important food for millions of people inhabiting the semi-arid tropics in Africa and Asia and are a major source of calories and vital component of food security in the semi-arid areas in the developing world (Amadou et al., 2013). Sorghum and pearl millet are often ground into flour and consumed as roti and/or porridge with milk and sometimes prepared as beverages (Obilana and Manyasa, 2002; Amadou et al., 2011). These crops can serve as the source of prebiotics (total dietary fiber, resistant starch, total oligosaccharides and β-glucan) for functional food (Awika and Rooney, 2004). Functional foods are referred as food ingredients or the food as such that influence beneficial effect on the host and/or reduce the risk of chronic diseases (Huggett and Schliter, 1996; Charalampopoulos et al., 2002). Probiotic microorganisms have been used for preparation of dairy food for thousands of years. Non-dairy based probiotic drinks utilizing cereals, including pearl millet, are also reported (Mridula and Sharma, 2015). Probiotic foods contain a single or mixed culture of probiotic microbes that improves the health of the host by improving intestinal microbial balance (Fuller, 1989). Information on the probiotic microorganisms associated with sorghum and pearl millet cultivars are inadequate (Badau, 2006). Hence, the present investigation was aimed to isolate and identify bacteria from flours and batter of sorghum and pearl millet cultivars and to further characterize these bacteria for their probiotic traits, in order to understand the probiotic potential of sorghum and pearl millet which are grown extensively in the Semi-Arid Tropics of the world. 

Collection of sorghum and pearl millet grain samples

A total of eight sorghum grain samples including K648 stress, K648 control, 6040 stress, 6040 control, R16 stress, R 16 control, K359 stress and K 59 control and two pearl millet grain samples including dual purpose hybrid (DPH) and high Fe hybrid were grown at ICRISAT and used for this study.

 

Collection of human pathogenic bacteria

Human pathogenic bacteria such as Escherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 6538), and Salmonella typhi (ATCC 14028) were acquired from American Type Culture Collection (ATCC), USA.

 

Preparation of the grain samples

The grain samples of both sorghum and pearl millet were dried at room temperature (30 to 32°C) for 3 days and milled in  Cyclotech^TM- mill. The ground samples were sieved through 0.2 mm sieve, in order to get fine flour. The batter samples were prepared, by mixing 5 g of flour in 5 ml of sterilized distilled water in a sterilized beaker and incubated/fermented at 28°C for 12 h. At the end of fermentation, the batter samples were immediately used for isolating the bacteria.

 

Isolation and maintenance of bacterial isolates

Ten grams of flour/batter from each of sorghum/millet grain sample was separately suspended in 90 ml of sterilized physiological saline (0.85% of NaCl in water) in a flask and placed on an orbital incubator shaker (at 120 rpm) for 30 min. At the end of shaking, the flour/batter samples were serially diluted up to 10⁶ dilutions with physiological saline. Dilutions 10⁴ to 10⁶ were spread plated (0.1 ml) on five different selective media including plate count agar, yeast glucose chloramphenicol agar, Bifidobacterium agar, actinomycetes isolation agar and de ManRogosa and Sharpe (MRS) agar and incubated at 30°C for 24 h. Bacteria were enumerated and the most prominent colonies which were found abundantly in the plate isolated and maintained on the respective media slants at 4°C for further studies.

 

Morphological and biochemical characterization of bacterial isolates

The bacteria were streaked on respective media and incubated at 30°C for 24 h. At the end of incubation, the colonies were observed for its morphological traits including form, surface, texture, color, elevation and margin. Gram staining reaction of the bacteria was done as per the protocols of Pelczar et al. (2008). The bacteria were also characterized for their biochemical traits including hydrogen sulphide, urease, catalase, oxidase, nitrate reduction, gelatin liquefaction, starch hydrolysis and IMViC (Indole, Methyl red, Voges Proskauer and simmons Citrate) tests as per the methods of Holt (1984). Utilization of carbohydrates such as lactose and sucrose were determined as per Forouhandeh et al. (2010).

 

Antibiotic resistance pattern and antimicrobial activity of bacterial isolates

Antibiotic resistance pattern was conducted by disc diffusion method. The resistance or susceptibility to antibiotics of bacterial isolates to ampicillin (10 µg), chloramphenicol (30 µg), ciprofloxacin (10 µg), erythromycin (15 µg), kanamycin (30 µg), penicillin (10 µg), streptomycin (10 µg), tetracycline (30 µg) and vancomycin (10 µg) (HiMedia, Mumbai, India) were determined according to the guidelines of Clinical and Laboratory Standards Institute (CLSI) (Wikler, 2006). In brief, antibiotic discs were placed on the Muller-Hinton agar plates immediately after swab with actively grown bacterial cultures. The plates were incubated at 30°C for 24 h. At the end of incubation period, zone of inhibition was measured.

The antimicrobial activity of the bacterial isolates was done by ditch assay method as per the protocols of Aswathy et al. (2008) with slight modifications. In brief, the bacterial isolates were grown in Muller-Hinton broth at 30°C for three days. At the end of incubation period, the culture filtrates were collected by centrifugation at 10,000 × g for 20 min and concentrated on a rotary evaporator until one fifth of the original volume. The concentrated culture filtrate samples were filter-sterilized through 0.2 µm membrane filter. Human pathogens such as E. coli, S. aureus and S. typhi, grown separately on nutrient broth at 37°C for 24 h, were amended with sterilized nutrient agar (2.5%) at 45°C, poured on sterile Petri plates and allowed to solidify. Upon solidification, a ditch of 0.5 mm was cut in the Petri plate and filled with  membrane-filtered concentrated culture filtrate of bacterial isolate (0.2 ml). The plates were initially placed at 4°C for 1 h, for diffusion of metabolites present in the culture filtrate, and further incubated at 37°C for 18 h. At the end of incubation, zone of inhibition was noted.

 

Probiotic traits of bacterial isolates

The bacterial isolates were characterized for their probiotic traits including tolerance to acid (low pH), bile salt, phenol and NaCl as per the standardized protocols used at ICRISAT.

 

Acid tolerance

This test was done as per the protocols of Liu et al. (2007) for identifying the bacterial isolates which could tolerate simulated gut acidic conditions. In brief, the Muller-Hinton broth was adjusted to pH 2.0, 3.0 and 4.0, and inoculated with one ml of log phase bacterial isolate. The inoculated broth was adjusted to 0.6 OD at 620 nm in UV spectrophotometer and incubated at 37°C for 120 min. The samples were withdrawn at the interval of 30 min and absorbance was measured at 620 nm. 

 

Bile tolerance

Bile tolerance test was determined based on Aswathy et al. (2008). Brain heart infusion (BHI) broth was amended with various concentrations of bile salt (0.3, 0.5 and 0.8%) and inoculated with one ml of log phase bacterial isolate. The inoculated broth was adjusted to 0.6 OD at 620 nm in UV spectrophotometer and incubated at 37°C for 120 min. The samples were withdrawn at the interval of 30 min and absorbance was measured at 620 nm. 

 

NaCl tolerance

NaCl tolerance was done as per the methods of Graciela and Maria (2001). Muller-Hinton broth was adjusted to different concentration of NaCl (3, 6, 9, and 12%) and inoculated with 1 ml of log phase bacterial isolate. The inoculated broth was adjusted to 0.6 OD at 620 nm in UV spectrophotometer and incubated at 37°C for 120 min. The samples were withdrawn at the interval of 30 min and absorbance was measured at 620 nm.

 

Phenol tolerance

The phenol tolerance of the bacterial isolates was assessed using the protocols of Teply (1984) with slight modifications. Log phase of bacterial isolate was inoculated in Muller-Hinton broth containing 0.2 and 0.4% of phenol. Absorbance was measured at 620 nm using UV spectrophotometer initially at 0 h and after 24 h of incubation at 37°C.

 

Molecular identification of the bacterial isolates

Pure cultures of probiotic potential bacteria were grown until log phase and genomic DNA were isolated according to Bazzicalupo and Fani (1995). The amplification of 16S rDNA gene was done by using universal bacterial primer 1492R (5'-TACGGYTACCTTGTTACGACTT-3') and 27F (5'-AGAGTTTGATCMTGGCTC AG-3') according to the conditions by Pandey et al. (2005). The polymerase chain reaction (PCR) product was sequenced at Macrogen Inc. Seoul, Korea. The sequences were  compared  with   those   from   GenBank   using   the   BLAST program (Alschul et al., 1990), aligned using the ClustalW software (Thompson et al., 1997) and phylogenetic trees inferred using the neighbor-joining method (Saitou and Nei, 1987).

 

Submission of sequences to NCBI

The 16S rDNA gene sequences of potential probiotic isolates were submitted to NCBI and the GenBank accession numbers were obtained.

 

Compatibility of the nine probiotic potential bacteria

Compatibility study was done as per the published protocols of Gopalakrishnan et al. (2011). In brief, the nine probiotic potential bacteria were streaked on Luria Agar in 10 cm diameter plate in a specific pattern that was first drawn on a paper as template. The template was prepared as follows: A 5 cm long vertical line at a distance of 5 mm from the margin was drawn on one side of the circle. Five lines, each 5 mm away and as perpendicular to the verticalline, were drawn. Each perpendicular line is 5 cm long and 1 cm apart from each other. Keeping this as a template below the Petri dish, the nine test bacteria were streaked over the vertical line and over the perpendicular lines. Care was taken that the cultures did not touch wall of the Petri plate. The plates were incubated for 24 h and observed for compatibility. If zone of inhibition was found between the two bacteria both of them were considered as not compatible whereas if no zone was found, these were considered as compatible.

 

 

Compatibility of the selected nine bacterial isolates

In the present study, when the probiotic potential bacteria were characterized for their compatibility  there  was  no definite sign of suppression of four bacteria (S8SF-32, SKSF-55, SKSF-7 and SKSF-8) on any of the other bacteria while the remaining five bacteria showed inhibition over others, suggesting these were not compatible.

Inhibition of human pathogenic microbes and resistance to antibiotics by probiotic microbial strains are advantageous traits. Hence, in the present study, the nine selected bacterial isolates were found to inhibit all the three tested pathogens, viz. E. coli, S. aureus and S. typhi (Table 3). Similar studies were conducted by Oluwajoba et al. (2013). Lactic acid bacteria isolated from millet and sorghum grains and fermented  products  were demonstrated to inhibit human pathogens such as S. aureus 25923, E. coli 25922, Pseudomonas aeruginosa 27853 and Enterococcus faecalis 29212 (Oluwajoba et al., 2013). Probiotic microbial strains such as L. fermentum, Bifidobacterium sp. and W. confusa are also reported to help in preventing and treating acute diarrhea (Lei et al., 2006). The major aim of using these probiotic strains should be to affect beneficially the gut microbial composition and activities. Hence, in the present study, when the selected nine bacterial isolates were also tested for their antibiotic resistance pattern, these were found resistance to tetracycline (except S8SF-4), streptomycin (except S6SF-44 and S8SF-4), kanamycin (except S8SF-4 and SKSB-14), chloramphenicol (except S8SF-4), ciprofloxacin (except S8SF-4), ampicillin (only PHFB-22, S8CF-32 and SKSF-8), penicillin (only PHFB-22, S8CF-32 and SKSF-8), erythromycin and vancomycin (except S6SF-44 and SKSB-14).  It is concluded that the selected nine isolates have good antagonistic potential as well as antibiotics resistance pattern.

The selection criteria of suitable probiotic microbial starter should be its ability to survive the acidic environment of the final fermented product and the adverse conditions of the gastrointestinal tract. Although, differences exist between species and specific strains, Lactobacilli (for instance, L. casei and L. plantarum showed longer shelf life than L. acidophilus) are generally considered to be intrinsically resistant to pH values higher than 3.0 (Hood and Zottola, 1988; Lee and Salminen, 1995). Hence, in the present investigation, the bacterial isolates were characterized for their probiotic properties including acid tolerance, bile tolerance, phenol tolerance and NaCl tolerance and all of them were found to tolerate acidic pH (2-3), bile (up to 0.5%), NaCl (up to 9%) and phenol (0.2%). Hence, it is concluded that the selected nine isolates have good probiotic properties.

In order to determine the identity of the nine potential probiotic bacteria, its 16S rDNA was sequenced and analyzed. A neighbor joining dendrogram was generated using the sequence from the nine potential probiotic bacteria (1400 bp) and representative sequences from the databases which revealed seven of the nine probiotic potential bacterial isolates belonged to Bacillus but different species whereas the other two isolates were found to be similar to Spingobacterium thalpophilum and Brevibacterium sp., respectively. Millet and sorghum flour have been used for isolation of lactic acid bacteria such as Lactobacillus plantarum, L. cellobiosus, L. pentosus, Leuconostoc mesenteroids and Pedicoccus pentosaceus (Okoronkwo, 2014). Mridula and Sharma (2015) reported a non-dairy probiotic drink from the sprouted cereals, legumes and soy milk using lactic acid bacteria. As found in the current investigation, Bacillus subtilis was reported to be isolated from the un-malted pearl millet grains in addition to Lactobacillus plantarum, Streptococcus lactis and Torulopsis glabrata  (Badau,  2006)  and  reported  to have probiotic potential (Kamgar et al., 2013). Microbes such as Lactobacillus sp., Bifidobacterium sp. and Weissella sp. are reported widely as probiotic microbes. Lactobacillus acidophilus was used to ferment a food mixture containing sorghum flour, whey powder and tomato pulp (Jood et al., 2012). However, in the present study, we report the isolation of Bacillus sp., Sphingobacterium sp. and Brevibacterium sp. as potential probiotic cultures. Hence, these cultures also can be exploited for making functional foods.

For any probiotic products/functional food, the ability of the probiotic microbial strain to attain high cell population is of primary importance. The probiotic microbial cell population of about 10⁷ cells ml^-1 at the time of consumption is considered functional (Gomes and Malcata, 1999). Probiotic foods obtained using a single microbial strain are not welcomed by the consumers due to rather sour and acidic taste and therefore probiotic strains are mixed (Saarela et al., 2000). Keeping this in mind, in the present investigation, all the probiotic potential bacteria were characterized for their compatibility so that these can be used in consortia. Compatibility studies between the nine probiotic potential bacteria revealed that there was no definite sign of suppression of four bacteria (S8SF-32, SKSF-55, SKSF-7 and SKSF-8) on any of the other bacteria while the remaining five bacteria showed inhibition over others, suggesting these were not compatible. Hence, it is concluded that the four of the bacterial isolates (S8SF-32, SKSF-55, SKSF-7 and SKSF-8) can be used as consortia in functional food product development. 

Cereals such as sorghum and millets are not only valuable sources of prebiotics and bioactive compounds useful for help in production of functional foods but also are sources of probiotic microbial cultures, which should be exploited for the production of new and innovative functional foods. Cereals and pseudocereals like Sorghum and pearl millet contain good quantities of potential prebiotic components, such as, water-soluble fibre (includes β-glucan and arabinoxylan), oligosaccharides (includes galacto- and fructo-oligosaccharides) and resistant starch important for the probiotic concept (Charalampopoulas et al., 2002). Hence, the multiple beneficial effects of cereals, especially the pseudo cereals such as sorghum and millets, can be used in association with good probiotic microbial cultures also isolated from these cereals in various ways to design novel cereal-based functional foods targeting different consumer segments having specific health requirements. Some of the important traits that need to be considered while selecting the cereals and the probiotic microbial culture include the composition of cereal  grains,  the  substrate  formulation, the growth capability of the starter probiotic culture, the stability of probiotic culture during storage, the organoleptic properties and the nutritional value of the final functional food product (Charalamoppulos et al., 2002). Cereals not only have the ability to grow probiotic microbes such as lactic acid bacteria in the human gut but also contain potential prebiotic compounds, whose functionality needs to be explored. However, the information available on the effects of cereal composition on the growth of probiotic microbes is limited. Therefore, the importance of substrate composition in conjunction with the nutritional requirements of the specific probiotic microbial strain is absolutely essential in order to make good quality functional food with validated health claims. In summary, the present work has demonstrated that the functional foods based on cereals, such as sorghum and millets, in association with good probiotic microbial stains can be exploited for designing novel cereal based functional foods for addressing food and nutritional security for millions of malnourished people living in the poorest of the poor countries of Africa and Asia.

The authors have not declared any conflict of interests.

The authors are grateful to the Ministry of Science and Technology, Government of India for their financial assistance (INT/SL/12/P-001). This work was undertaken as part of the CGIAR Research Program on dry land cereals. ICRISAT is a member of CGIAR Consortium. The authors acknowledge Dr Jana Kholova and Dr SK Gupta from ICRISAT for kindly providing the sorghum and pearl millet samples used in this study, respectively. The authors would also like to thank ICRISAT and all of the staff members of the biocontrol unit, including PVS Prasad, P Manohar, B Nagappa, D Barath and A Jabbar for their technical assistance.