#391_CARBALLO_bozza ! Ital. J. Food Sci., vol 28, 2016 - 598 PAPER POTENTIAL TECHNOLOGICAL INTEREST OF INDIGENOUS LACTIC ACID BACTERIA FROM ALGERIAN CAMEL MILK K. BELKHEIR1, J.A. CENTENO2, H. ZADI-KARAM1, N.-E. KARAM1 and J. CARBALLO*2 1Laboratoire de Biologie des Microorganismes et Biotechnologie (LBMB), Université Ahmed Ben Bella d’Oran 1, BP 1524, El M'Naouer, 31000 Oran, Algérie 2Food Technology Area, Faculty of Sciences, University of Vigo, 32004 Ourense, Spain *Corresponding author. Tel.: +34 988387052; fax: +34 988387001 E-mail address: carbatec@uvigo.es ABSTRACT Nine isolates of lactic acid bacteria (LAB) obtained from the predominant microbiota of different camel milk samples collected in South-West Algeria, were selected in accordance with their growth ability in (cow) milk. The isolates were phenotypically and genotypically assigned to the following species: 4 Leuconostoc mesenteroides subsp. dextranicum; 2 Lactobacillus brevis; 2 Lb. plantarum; and 1 Lactococcus lactis subsp. lactis. One isolate from each of the Leuconostoc and Lactobacillus species were selected on the basis of their highest proteolytic and aminopeptidase activities. The selected isolates were used in combination with a commercial mesophilic O-type culture to make fermented milks. Sulfury flavor was detected as predominant in the sensory analysis of the milks made with only the Ln. mesenteroides and Lb. brevis adjuncts, which were characterized by the highest abundances of sulfur volatile compounds. Butter flavor was perceived in the milks made with the Lb. plantarum(cit+) adjunct, and was related to the presence of acetoin. Finally, cheese flavor prevailed in the milks made with both the Lb. brevis and the Lb. plantarum adjuncts, characterized by their high contents of short-chain free fatty acids. The results suggest the potential interest of these microorganisms in the manufacture of dairy products, particularly the combination of the Lb. brevis and Lb. plantarum isolates for cheese making. Keywords: camel milk, lactic acid bacteria, Lactobacillus brevis, Lactobacillus plantarum, sensory analysis, volatile compounds ! Ital. J. Food Sci., vol 28, 2016 - 599 1. INTRODUCTION One of the main challenges of modern dairy industry is the development of a range of varied products to meet the needs and tastes of the various sectors of an increasingly numerous and demanding population. The request for novel cultured products development having different and improved sensorial attributes requires the use of microbial strains with interesting properties for application in dairy fermentations (WOUTERS et al., 2002). The strict hygiene measures and the high standardization of the production systems make the milk and dairy cultured products manufactured in the developed countries a poor source of novel and distinctive LAB strains. So, novel strains should be searched in little standardized raw materials and products, and in ecological niches with particular and unusual environmental conditions. Camel (Camelus dromedarius) milk, besides being produced in traditional low-tech systems, has compositional characteristics that make it very special milk with clearly different environmental conditions (FARAH, 1993; KONUSPAYEVA et al., 2009) that may influence the metabolic abilities of the microbiota that settles it. The majority of the studies carried out on camel milk in the world, and particularly in Algeria focuses on the problem of its low clotting ability, which has been widely investigated (FARAH and BACHMAN, 1987; BOUDJENAH-HAROUN et al., 2012). However, few works have been conducted regarding the microbiota of Camelus dromedarius milk or the biochemical and technological properties of LAB isolated from this source (ASHMAIG et al., 2009; DRICI et al., 2010; BENDIMERAD et al., 2012; AKHMETSADYKOVA et al., 2015). In this context, the aims of the present study were: (i) to identify LAB isolates with potential technological interest belonging to the predominant microbiota of Algerian camel milk, and (ii) to characterize and to assay the isolates with a view to their use as adjunct cultures in the manufacture of fermented milks and cheeses with differentiated sensory characteristics. 2. MATERIALS AND METHODS 2.1. Bacterial cultures, media, and growth conditions Nine LAB isolates obtained on pH 5.5 MRS medium (Oxoid Ltd., Basingstoke, UK) from six different camel milk samples collected in Bechar and Tindouf cities (southwest Algeria) were selected for this study. The selected isolates were able to clot cow milk after 12-24 h incubation at 30ºC. The bacterial cultures were maintained at –30ºC in MRS broth containing 20% glycerol (v/v). Working cultures were prepared by two consecutive transfers in MRS broth at 30 ºC. Lyophilized commercial mesophilic O-type culture FD- DVS R-704 (Chr. Hansen, Denmark), containing Lactococcus lactis subsp. lactis and Lc. lactis subsp. cremoris (phage-resistant) strains was used alone or in combination with selected indigenous LAB cultures for the preparation of fermented milks. This culture was stored at –20ºC and was directly inoculated in milk following the manufacturer’s instructions. 2.2. Phenotypic identification of LAB isolates Isolates were phenotypically assigned to the genus level as described by GARABAL et al. (2008). Utilization of citrate was also determined on citrate calcium agar (KCA) (NICKELS and LEESMENT, 1964), after 72 h incubation at 30ºC. Criteria followed for phenotypic identification were those compiled by WOOD and HOLZAPFEL (1995). Assignation to species level was made by means of the API 50 CHL carbohydrate fermentation strips ! Ital. J. Food Sci., vol 28, 2016 - 600 (bioMérieux, Marcy l’Etoile, France), following the manufacturer’s instructions. The results at 48 h were analyzed using the API CH Lab software package (bioMérieux). 2.3. Genotypic identification of LAB isolates For molecular identification of (presumptive) leuconostocs and lactobacilli isolates, chromosomal DNA from bacterial isolates was extracted from single-colony by using the SPEEDTOOLS DNA Extraction kit (Biotools B&M Labs, S.A., Madrid, Spain) following the manufacturer’s instructions. From the obtained DNA, 20 !L aliquots were made and stored at – 20ºC until needed. All DNA samples were tested using universal primers amplifying a 1000-bp region of the 16S rRNA gene 616V (forward): 5’-AGAGTTTGATYMTGGCTC AG-3’ and 699R (reverse): 5’-RGGGTTGCGCTCGTT-3’ (ARAHAL et al., 2008). The primers were synthesized by Invitrogen Co. (Invitrogen, Carlsbad, CA, USA), diluted at a final concentration of 1 !g/!L with sterilized deionized water upon reception and stored at –20ºC. PCRs were performed in a reaction volume of 50 !L, containing 20 pM each of forward and reverse primers and 1 mL DNA template and prepared by using the DreamTaq DNA polymerase kit (Thermo Scientific, Waltham, MA, USA). PCR amplifications were monitored in a Gene Amp thermal PCR System 2700 thermal cycler (Applied Biosystems, Foster City, CA, USA) under the following conditions: an initial cycle of 95ºC for 15 min; then 35 cycles of 95ºC for 30 s, 55ºC for 30 s, 72ºC for 1 min; and a final elongation step of 72ºC for 10 min. The PCR products were purified using the NucleoSpin® Gel and PCR Clean-up kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany). DNA sequencing was carried out with the v3.1 BigDye Terminator Cycle Sequencing kit (Applied Biosystems) on the ABI 3130xl capillary automate sequencer (Applied Biosystems) following the manufacturer’s instructions. The obtained sequences were aligned to 16S rRNA gene sequences in the Gen Bank NCBI data base (National Center for Biotechnology Information, http://www.ncbi.nlm.nih.gov) using the BLAST program (http://www.ncbi.nlm.nih.gov/BLAST) database. 2.4. Technological characterization of the selected LAB isolates Leuconostocs and lactobacilli isolates were characterized for the following abilities: (a) carbohydrate fermentation, using the miniaturized API 50CHL system (bioMérieux), after 48 h incubation at 30ºC; (b) acidifying activity in sterile (110ºC, 15 min) reconstituted (10%, w/v) skim milk (pH 6.70) (Oxoid) after 6 h incubation at 30ºC (IDF, 1995); (c) proteolytic activity in sterile reconstituted skim milk (Oxoid) after 24 h incubation at 30ºC, as evaluated by the OPA method (Church et al. 1983); (d) aminopeptidase (AP) activity in sterile reconstituted skim milk (Oxoid) using leucine- and lysine-p-nitroanilide (both reagents from Sigma-Aldrich Corp., St. Louis, MO, USA) as substrates (EL SODA and DESMAZEAUD, 1982); (e) diacetyl-acetoin production in sterile reconstituted skim milk (Oxoid) after 48 h incubation at 30ºC (IDF 1997); (f) extracellular proteolytic activity on calcium caseinate agar (Merck GmbH, Darmstadt, Germany) (RODRÍGUEZ-ALONSO et al., 2008); (g) extracellular lipolytic activity on 1% (w/v) tributyrin agar (Merck), containing 1% arabic gum (Panreac) (RODRÍGUEZ-ALONSO et al., 2008); and (h) antimicrobial activity of the nine isolates against each other and against the commercial culture R-704 (Chr. Hansen, Hørsholm, Denmark), determined by the agar well diffusion assay as described by CENTENO et al. (2002). All assays for testing technological abilities were performed in triplicate, and numerical results were expressed as the mean values obtained for each isolate. ! Ital. J. Food Sci., vol 28, 2016 - 601 2.5. Preparation of fermented milks Eight fermented or acidified milks were prepared per assay. One of the milks was acidified with the mesophilic commercial culture R-704 alone, and the other seven milks were fermented with the commercial culture plus one among three selected isolates (Leuconostoc mesenteroides subsp. mesenteroides C8M, Lactobacillus brevis C21B, Lactobacillus plantarum C22P), or their combinations (C8M+C21B, C8M+C22P, C21B+C22P, and C8M+C21B+C22P). Fermented milks were prepared in 1 L of retail pasteurized (80ºC, 20 s) homogenized whole (3.6 g fat 100 mL-1) milk (Leyma, A Coruña, Spain) contained in polyethylene bottles. The milks used for each assay corresponded to the same industrial batch and had been pasteurized the day before purchasing. The bottles were inoculated with 0.1 units of the DVS R-704 commercial culture. The DVS culture was previously rehydrated and strongly stirred in sterile (reconstituted) skim milk (Oxoid) at a ratio of 10 units per liter of milk, and then inoculated at 1% (v/v) in the pasteurized milk. The LAB isolates were cultured in sterile skim milk (Oxoid) at 30ºC for 16 h, and then inoculated at 1% (v/v) in the pasteurized milk. The inoculated milks were incubated at 30ºC for 24 h. Assays were made in triplicate. 2.6. Sensory analysis and volatile compounds produced in the fermented milks For sensory analysis, the (clotted) fermented milks were intensely agitated and then distributed in 50 mL amounts in sterile polyethylene containers. Flavors of the fermented milks were perceived by smelling and tasting, as evaluated by a panel of ten regular consumers of fermented milks who had been trained as previously described (RODRÍGUEZ-ALONSO et al., 2008). The judges were asked to agitate the milks before smelling and tasting. Odor and taste preference as well as overall acceptance were scored on a scale from 1 to 7. The acceptability indexes (AI) of the fermented milks were calculated by the formula: overall acceptance × 100/7 (DUTCOSKY, 1996). Volatile compounds were determined in 20 mL samples taken after agitation of the clotted milks, and stored at –80 ºC until analysis. Volatile compounds were extracted by using the solid phase microextraction (SPME) technique and detected by gas chromatography coupled to mass spectrometry (GC–MS). For samples equilibration, 10 mL headspace vials containing 1 g of fermented milk were sealed with a PTFE-faced silicone septum (Supelco, Bellefonte, PA, USA) and maintained at 35ºC in a thermo block (Memmert model 100-800, Schwabach, Germany) during 15 min. Then, a 75 µm film thickness carboxen/polydimethylsiloxane (CAR/PDMS) fibre (Supelco) was exposed to the headspace while maintaining the sample at 35°C during 30 min. The SPME adsorbed compounds were injected to the chromatograph with a splitless mode injection at 260ºC for 8 min. The separation of volatiles was performed on a Hewlett-Packard 6890N (Agilent Technologies, Santa Clara, CA, USA) gas chromatograph equipped with a DB-624 capillary column (30m × 0.25 mm id, 1.4 µm film thickness; J&W Scientific, Folsom, CA, USA), following the method described by LORENZO and FONSECA (2014). Compounds were identified by comparing their mass spectra with those contained in the NIST05 (National Institute of Standards and Technology, Gaithersburg, MD, USA) library, and by comparing their mass spectra and retention time with authentic standards (Supelco). Abundances of volatile compounds are provided as peak area units/106 values. 2.7 Statistical analysis Numerical data corresponding to technological characteristics, sensory analysis and volatile compounds were subjected to analysis of variance (ANOVA) and where statistical ! Ital. J. Food Sci., vol 28, 2016 - 602 differences were noted, differences among the distinct groups (isolates or fermented milks) were determined by the Duncan’s test at a significance level of P < 0.05. All statistical procedures were performed with the SPSS software, version 20.0 (SPSS Inc., Chicago, IL, USA). 3. RESULTS AND DISCUSSIONS 3.1. Isolation and identification of LAB isolates Low quantities (mean value of 2.50 log cfu mL-1) of LAB were found in the six samples of camel milk used for LAB isolation (data not shown). This fact may be attributed to the high content of lysozyme and ascorbic acid in camel milk (FARAH, 1993). Among the 45 LAB isolates initially obtained, 27 (60%) were phenotypically assigned to Leuconostoc/Weissella genus (data not shown). Similarly to the results of this study leuconostocs have often been found as the major LAB in camel milk (ASHMAIG et al., 2009; AKHMETSADYKOVA et al., 2015), this may be attributed to the higher resistance to lysozime of this microbial group as compared to other LAB (LIMONET et al., 2004). As regards the remaining 18 isolates, eight were identified as mesophilic lactobacilli, one was assigned to Lactococcus genus, and nine LAB isolates (unable to clot cow milk) could not be assigned to genus level (data not shown). The selected nine isolates able to clot (cow) milk after 12-24 h incubation at 30ºC were identified by means of the API 50 CHL system as follows: four Leuconostoc mesenteroides subsp. dextranicum (similarity level of 95.2-99.8%); two Lactobacillus brevis (99.7% similarity); two Lactobacillus plantarum (77.8-79.3% similarity); and one Lactococcus lactis subsp. lactis (72.3% similarity) (data not shown). Both the Lb. plantarum and the Lc. lactis subsp. lactis isolates were able to metabolize citrate on KCA medium (data not shown). The low proportion of leuconostoc isolates showing the ability to coagulate milk may be explained by the fact that these microorganisms are adapted to growth on vegetables and roots and therefore lack sufficient proteolytic ability to grow in milk (VEDAMUTHU, 1994). The lactococcal isolate was not further considered in this study because of the known susceptibility of lactococcal cells to bacteriophage infection, as lytic phage infection is at present a major cause of fermentation failure (GARNEAU and MOINEAU, 2011). Phenotypic identification of the selected (presumptive) leuconostocs and lactobacilli isolates was confirmed by sequencing of the fragments of the 16S rDNA gen amplified by PCR (data not shown). Sequences obtained from the four presumptive Ln. mesenteroides isolates produced significant (97-99%) alignments with the complete genome of Ln. mesenteroides subsp. dextranicum DSM 20484. Sequences from the two presumptive Lb. brevis isolates aligned (97%) with that of Lb. brevis ATCC 367. Lower percentages of similarities (91-94%) were found between the presumptive Lb. plantarum isolates and the Lb. plantarum JDM1 strain (genomes of all the type strains included in the NCBI data base). Similarly to the results of this study, the species Lb. plantarum and Lb. brevis have been isolated from Sudanese fermented camel milk (ASHMAIG et al., 2009), and Lc. lactis subsp. lactiscit+ has been found in Algerian fermented camel milk (DRICI et al., 2010; BENDIMERAD et al., 2012). 3.2. Technological properties of the selected leuconostocs and lactobacilli isolates The most relevant carbohydrate fermentation abilities exhibited by the selected LAB strains are shown in Table 1. Despite of their ability to hydrolyze lactose in the API 50 CHL test tube, all the selected isolates (Ln. mesenteroides subsp. dextranicum C2M, C5M, ! Ital. J. Food Sci., vol 28, 2016 - 603 C8M and C14M; Lb. brevis C21B and C27B; and Lb. plantarum C14P and C22P) showed low acidifying activities (mean pH values of 6.42-6.65, and mean titratable acidities of 0.06-0.21 g lactic acid 100 mL-1) after 6 h incubation at 30ºC (Table 1). The pH values of the milk cultures of the lactobacilli isolates were significantly (P < 0.05) lower than those of the leuconostocs isolates. The lowest mean pH values corresponded to the two isolates of Lb. brevis (Table 1). On the basis of their low acidifying activities, Lb. brevis, Lb. plantarum and Ln. mesenteroides have been generally classified as adjunct LAB to be used together with a lactococcal starter in order to enhance the acidification of milk (SETTANNI and MOSCHETTI, 2010). The proteolytic activities evaluated by means of the OPA test ranged between 0.124 lysine mM for one Lb. brevis isolate and 0.230 lysine mM for one Ln. mesenteroides isolate (Table 1). These values are much lower than those reported by HERREROS et al. (2003) for 17 isolates of leuconostoc and mesophilic lactobacilli obtained from raw goat’s milk cheeses (mean value equivalent to 0.296 lysine mM) and by GARABAL et al. (2008) for 42 isolates of mesophilic lactobacilli from raw cow’s milk cheeses (mean value equivalent to 0.298 lysine mM). Nevertheless, this low proteolytic activity could be a suitable trait if these bacteria are used as adjunct cultures, in order to prevent development of bitterness in the final product (NANDAN et al., 2010). All the leuconostocs and lactobacilli isolates exhibited both Leu- and Lys-aminopeptidase activities, with higher Lys-AP than Leu-AP values (Table 1). Differences among species were pronounced, with the two Lb. brevis isolates showing significant (P < 0.05) higher values for both activities (highest mean values of 200 U for Leu-AP and 530 U for Lys-AP) than the Lb. plantarum and Ln. mesenteroides isolates. Values obtained for the Lb. plantarum isolates were also significantly (P < 0.05) higher than those of the Ln. mesenteroides isolates (Table 1). High Leu- and Lys-aminopeptidase activities have been detected for cell-free extracts of some strains of Lb. brevis, Lb. plantarum and Ln. mesenteroides isolated from raw ewe’s and goat’s milk cheeses (MACEDO et al., 2000; HERREROS et al., 2003). Similarly to the results of the present study, NIETO-ARRIBAS et al. (2009) found low Leu- and Lys-AP activities for intact cells of 19 Ln. mesenteroides subsp. dextranicum isolates obtained from artisanal Manchego ewe’s cheese, with Lys-AP activity being higher than Leu-AP activity for all the isolates. The same authors (NIETO-ARRIBAS et al., 2010) reported higher values of both Leu-AP and Lys-AP activities for most of 10 Lb. plantarum isolates than those of Ln. mesenteroides isolates from the same cheese variety. Aminopeptidases play a key role in the degradation of bitter peptides and flavor formation during cheese ripening (URBACH et al., 1995; NANDAN et al., 2010). Only the two Lb. plantarum isolates able to metabolize citrate on KCA produced diacetyl- acetoin in skim milk (Table 1). The production of diacetyl-acetoin by Lb. plantarum C22P (120 mg diacetyl L-1) was significantly (P < 0.05) higher than those observed for Lb. plantarum C14P (45 mg L-1) and for the initially selected Lc. lactis subsp. lactis(cit+) isolate (35 mg L-1, data not shown). GARABAL et al. (2008) reported a mean production equivalent to 84 mg diacetyl L-1 for 33 isolates of mesophilic facultatively heterofermentative lactobacilli obtained from Galician (northwest Spain) raw cow’s milk cheeses. Production of diacetyl by LAB could be considered an interesting technological property since this compound is related to positive (butter) flavor and antimicrobial effect in dairy products (DRICI et al., 2010). None of the isolates exhibited extracellular proteolytic or lipolytic activities in spite of their growth on calcium caseinate agar and on tributyrin agar media (data not shown). ! Ital. J. Food Sci., vol 28, 2016 - 604 Table 1: Technological characterization of the selected LAB isolates obtained from camel milk (quantitative results are means of three replicates). Leuconostoc mesenteroides Lactobacillus brevis Lb. plantarum S.E.M. 1 P value Fermentation of: C2M C5M C8M C14M C21B C27B C22P C14P L- arabinose – – – – + + + + D-cellobiose – – – – – – + + D- fructose + + + + + + + + D-galactose + + – – + + + + D-glucose + + + + + + + + D-lactose + + + w2 + + + + D-maltose + + + + + + + + D-mannose + + + + – – + + D-melibiose – – – – + + + + D-raffinose – – – – – – – – L-rhamnose – – – – – – – – D-ribose w – w w + + + + D-saccharose + + + + – w + + L-sorbose – – – – – – – – D-trehalose + + + + – – + + D-xylose w w w w + + – – Acidifying activity (skim milk; 30 ºC, 6 h) pH 6.60b 6.65a 6.63a 6.65a 6.42d 6.45cd 6.48c 6.50c 0.003 0.000 Titratable acidity3 0.10 bc 0.06d 0.08cd 0.09c 0.21a 0.21a 0.14b 0.12b 0.0002 0.000 Proteolytic activity (OPA)4 0.210 ab 0.212ab 0.230a 0.189b 0.126d 0.124d 0.180bc 0.178c 0.009 0.012 Aminopeptidase activity5 Leu-AP 17.2d 14.6d 24.0d 19.4d 200a 197a 109b 51c 66 0.000 Lys-AP 73c 68c 73c 71c 530a 528a 129b 134b 161 0.000 Producton of diacetyl- acetoin6 0.0c 0.0c 0.0c 0.0c 0.0c 0.0c 120a 45b 6.9 0.000 a-dMean values within a row with different superscripts are significantly different (P < 0.05; Duncan’s test) 1S.E.M.: Standard error of the mean 2w: weak reaction 3Expressed as g lactic acid 100 mL-1 4Expressed as mM lysine 5Expressed as enzymatic units (1 enzymatic unit = amount of enzyme giving an absorbance of 0.001 units at 410 nm min-1) 6Expressed as mg of diacetyl per liter of milk ! Ital. J. Food Sci., vol 28, 2016 - 605 The absence of extracellular enzymatic activities in LAB other than enterococci has been generally reported (HERRERO et al., 1996; BUFFA et al., 2004; NIETO-ARRIBAS et al., 2009, 2010). According to HERRERO et al. (1996), LAB used as starter cultures should ideally present low lipolytic activity, as the degradation of milk fat must be slight in order to induce aroma production without giving rancid flavors. One isolate from each one of the identified species (Ln. mesenteroides subsp. dextranicum C8M; Lb. brevis C21B; and Lb. plantarum C22P) were selected in accordance with their expected better technological performance (highest proteolytic and AP activities, and production of diacetyl for Lb. plantarum isolates). It should be pointed out that the similarity in the results of biochemical and genotypic (data not shown), and technological assays for Lb. brevis C21B and C27B could indicate that these two isolates probably belong to the same strain, although they were obtained from different camel milk samples (data not shown). The selected isolates were used to make fermented milks in combination with the commercial culture R-704 (Chr. Hansen) after confirming the absence of antibacterial activity against each other and against the commercial culture (data not shown). 3.3. Sensory analysis and volatile compounds produced in the fermented milks The milks fermented with the commercial culture alone (control) and those cultured with the commercial starter in combination with both Ln. mesenteroides C8M and Lb. plantarum C22P, and in combination with the three selected isolates received the highest scores for sensory attributes (Table 2). The scores reached for the odor attribute were significantly (P < 0.05) higher for these last two fermented milks than for the other milks, and the scores for the taste attribute were significantly (P < 0.05) higher for the milk fermented with both the C8M and C22P adjuncts than for the other cultured milks. The highest acceptability index (86.1%) corresponded to the milk cultured with both the C8M and C22P isolates, followed by the milk fermented with the three selected isolates (76.7%) and by the control milk (74.5%) (Table 2). Table 2: Mean sensory scores and main flavor descriptors of the fermented milks made with the commercial FD-DVS R-704 O-type starter alone (control) and in combination with cultures of the selected isolates obtained from camel milk. Control +C8M +C21B +C22P +C8M +C21B +C8M +C22P +C21B +C22P +C8M +C21B +C22P S.E.M.1 P value Odor 5.0b 4.8bc 4.7bc 4.9bc 4.4c 5.8a 4.8bc 5.7a 0.060 0.039 Taste 5.3b 3.9d 3.8d 4.8bc 5.3b 6.2a 5.0bc 5.0bc 0.087 0.000 Main flavor descriptors acid yoghurt sulfury metallic sulfury garlic butter vanilla sulfury sour butter yoghurt cheese butter cheese butter AI (%)2 74.5bc 62.8e 61.5e 70.1d 70.2d 86.1a 71.7cd 76.7b 1.9 0.000 a-eMean values within a row with different superscripts are significantly different (P < 0.05; Duncan’s test) 1S.E.M.: Standard error of the mean 2AI (%): acceptability index of the fermented milk calculated by the formula: overall acceptance × 100/7 (in accordance with the used 7-point scale) The flavor descriptors used by the majority of the judges include: acid and yoghurt for the control milk; butter and yoghurt for the milk fermented with both the C8M and C22P adjuncts; and butter and cheese for the milk cultured with the three selected isolates (Table ! Ital. J. Food Sci., vol 28, 2016 - 606 2). The butter and vanilla notes described for the milks made with the Lb. plantarum C22P adjunct may be related to the production of diacetyl and acetoin by this culture. These compounds are responsible for butter and nuts flavors in cheese (KONDYLI et al., 2003; KAMINARIDES et al., 2007). Sulfury and garlic nuances were detected by most of the judges in the milk fermented with the sole Lb. brevis C21B adjunct, which showed the lowest AI (61.5%) (Table 2). The sulfur flavors are related to the production of sulfur compounds principally from L-methionine (Met). It has been reported that sulfur compounds provide particular flavor notes in cheese such as garlic taste (KAMINARIDES et al., 2007). The milks fermented with the C21B adjunct in combination with either the C8M or the C22P isolates received significant (P < 0.05) higher scores for the taste attribute than the milk made with the C21B isolate as the only adjunct culture (Table 2). This suggests that the “sulfury defect” was moderated due to the presence of other flavor compounds. Only ten volatile compounds were definitely identified in the fermented milks with the methodology used in the present study. These volatiles include: the alcohol 2-heptanol; the ketones 3-hydroxy 2-butanone (acetoin), 2-heptanone and 2-nonanone; the fatty acids acetic acid, butanoic acid, hexanoic acid and octanoic acid; and the sulfur compounds dimethyl disulfide (DMDS) and dimethyl trisulfide (DMTS) (Table 3). Table 3: Mean abundances (expressed as peak area units / 106) of each of the volatile compounds determined in the fermented milks made with the commercial FD-DVS R-704 O-type starter alone (control) and in combination with cultures of the selected isolates obtained from camel milk. Control +C8M +C21B +C22P +C8M +C21B +C8M +C22P +C21B +C22P +C8M +C21B +C22P S.E.M.1 P value ALCOHOLS 2-Heptanol ndd2 4.1c 5.2c 14.0a 8.6b 14.2a 10.8ab 9.8ab 0.67 0.003 KETONES 3-Hydroxy 2- butanone nd d ndd ndd 93a ndd 20.1b ndd 5.4c 21.8 0.031 2-Heptanone 1.1b 2.8b 1.5b 29.3a 0.8b 28.2a 4.8b 4.7b 3.06 0.001 2-Nonanone ndd 1.3b 0.4c 3.1a ndd 2.8a ndd ndd 0.39 0.043 FATTY ACIDS Acetic acid 128e 239c 149de 329a 160d 265bc 131e 281b 18.9 0.005 Butanoic acid 3.8e 5.8d 9.7bc 9.2bc 8.1c 7.4c 15.4a 10.7b 0.66 0.006 Hexanoic acid ndf 2.7e 7.3bc 6.5c 5.6d 6.7c 12.7a 7.8b 0.56 0.004 Octanoic acid ndc ndc 0.8ab 0.5b 0.5b 0.5b 1.2a 1.1a 0.073 0.019 SULFUR COMPOUNDS Dimethyl disulfide 0.7cd 2.0a 2.3a 1.3b 2.1a 1.0bc 0.5d 0.6d 0.21 0.011 Dimethyl trisulfide ndb ndb 0.3a ndb 0.2a ndb ndb ndb 0.01 0.001 a-fMean values within a row with different superscripts are significantly different (P < 0.05; Duncan’s test) 1S.E.M.: Standard error of the mean 2nd: Compounds not detected (considered as 0.0 values for statistical analysis) ! Ital. J. Food Sci., vol 28, 2016 - 607 The poor volatile profiles of the fermented milks, with the absence of aldehydes and the practical absence of alcohols, may be partially explained on the basis of the SPME fiber performance. It has been reported that the porous CAR/PDMS fiber coating shows a higher affinitiy for low-molecular-weight compounds, including a high proportion of ketones (LORENZO, 2014). In addition, absorption for the most compounds may reach equilibrium with a longer time of exposure than that assayed in this work (MARCO et al., 2004), although longer exposure times could result inappropriate due to oxidation phenomena that can produce competitive effects between compounds (LORENZO, 2014). The secondary alcohol 2-heptanol, and its probable precursor 2-heptanone were present in all the fermented milks made with the adjunct cultures, but not in the control milk. The highest abundances of (2-heptanone and) 2-heptanol were found in the milk fermented with the Lb. plantarum C22P adjunct. The abundances of 2-heptanol in the milks made with the sole C22P adjunct and with the combination of the C8M and C22P adjuncts were significantly (P < 0.05) higher than those of the milks made without the C22P adjunct culture. This compound, associated to herbaceous and fruity aromas has been considered as a key odorant in many cheese varieties (DELGADO et al., 2011). Regarding the ketones group, 3-hydroxy 2-butanone (acetoin) was only detected in the milks cultured with the Lb. plantarum C22P adjunct (Table 3), as it could be expected from the results of the technological assays. Acetoin is produced by the reduction of diacetyl originated from citrate metabolism, or by decarboxylation of α-acetolactate (McSWEENEY et al., 2000). Diacetyl, and to a lesser extent acetoin are generally appreciated for their buttery and nut-like notes (CURIONI et al., 2002), and may be responsible to some extent for the butter flavors described in the sensory analysis. The abundances of the methyl ketones 2-heptanone and 2-nonanone were significantly (P < 0.05) higher in the milks made with the sole Lb. plantarum C22P adjunct culture and with the combination of the C8M and C22P adjuncts than in the other fermented milks (Table 3). The methyl ketones 2- heptanone and 2-nonanone are associated with cheesy odors, particularly with blue cheese notes (CURIONI et al., 2002). Methyl ketones containing odd numbers of C atoms are produced from β-oxidation of fatty acids (McSWEENEY et al., 2000), which may have been released through the esterase and/or lipase activities of the lactobacilli. The four fatty acids identified were detected in all the fermented milks made with the adjunct cultures, with the exception of octanoic acid in the milk fermented with the sole C8M leuconostoc adjunct. The abundances of acetic acid were significantly (P < 0.05) higher in the milks made with the sole Lb. plantarum C22P adjunct culture than in the other fermented milks (Table 3). Citrate-fermenting microorganisms convert this compound to pyruvate, carbon dioxide, and acetic acid, together with various carbonyl compounds such as diacetyl and acetoin. Facultatively heterofermentative lactobacilli such as Lb. plantarum may also produce acetate from lactose or from amino acids (BUFFA et al., 2004). Acetate is primarily formed from citrate and this may be the main reason why this compound is more abundant in the milks fermented with the Lb. plantarum C22P adjunct. Acetate also comes from pyruvate originated in the glycolysis from carbohydrates, and this would explain its presence in milks fermented with adjuncts unable to metabolize citrate. Acetic acid is responsible for sour flavor in dairy products (KAMINARIDES et al., 2007). The abundances of butanoic and hexanoic acids were significantly (P < 0.05) higher in the milks made with the combination of Lb. brevis C21B and Lb. plantarum C22P adjunct cultures than in the other fermented milks (Table 3). MENÉNDEZ et al. (2000) found higher volatile free fatty acids contents (3.05 vs. 1.33 meq/100 g) in 1-day samples of cow’s cheeses made with a Lb. plantarum adjunct culture than in control samples. Butanoic acid may be derived from lipolysis of milk fat or produced through the fermentation of lactose and lactic acid (KAMINARIDES et al., 2007). However, and despite the fact that the strains tested in the present study did not exhibit (exocellular) lipolytic activity when growing on ! Ital. J. Food Sci., vol 28, 2016 - 608 tributyrin agar, hexanoic and octanoic acids are probably released from triglycerides through the action of non-specific bacterial esterases and lipases as reported by GONZÁLEZ DE LLANO et al. (1996) and BUFFA et al. (2004). Short-chain fatty acids are related to rancid and pungent flavors (McSWEENEY et al., 2000; DELGADO et al., 2011), and butanoic acid has been found to be a potent odorant in cheese (KAMINARIDES et al., 2007). The high contents of butanoic and hexanoic acids may relate to the cheese flavors detected in the milks made with both the C21B and C22P adjuncts. As regards sulfur compounds, the abundances of DMDS were significantly (P < 0.05) higher in the milks made with the Ln. mesenteroides C8M and the Lb. brevis C21B adjuncts, both alone or in combination, than in the other fermented milks (Table 3). DMTS was only detected in the milks made with the sole C21B adjunct and with the combination C8M+C21B. The thioesters DMDS and DMTS are expected to be formed mainly from the amino acid Met by the metabolism of LAB and secondary microbiota (URBACH, 1995; ENGELS et al., 1997). It has been reported that a number of lactobacilli strains, including Lb. brevis are able to degrade Met (ENGELS et al., 1997; SREEKUMAR et al., 2009) which agrees with our findings regarding the Lb. brevis C21B adjunct assayed in the present study. The sulfury flavors detected in the sensory analysis of milks made with the Ln. mesenteroides C8M and the Lb. brevis C21B adjuncts, and the garlic nuance described in the milk made with the sole C21B adjunct culture may be attributed to the high contents of sulfur compounds. 4. CONCLUSIONS The selected leuconostocs and lactobacilli isolates from camel milk used in the present study as adjunct cultures to manufacture fermented milks in combination with an acidifying starter, conferred different flavor nuances and were responsible for different volatile profiles of the products. Sulfury flavor was detected as a main flavor in the milks made with only the Ln. mesenteroides and Lb. brevis adjuncts, characterized by the highest abundances of volatile sulfur compounds. Butter flavor, related to the presence of the volatile 3-hydroxy 2-butanone (acetoin) was perceived in the milks made with the Lb. plantarum(cit+) adjunct, and cheese flavor, associated to high contents of volatile free fatty acids prevailed in the milks made with both the Lb. brevis and the Lb. plantarum adjuncts. The results suggest the potential interest of these LAB isolates in the manufacture of dairy products, in particular the combination of the Lb. brevis and Lb. plantarum isolates for cheese making. ACKNOWLEDGEMENTS Authors are grateful to Dr. José Manuel Lorenzo (Food Technology Center of Galicia, Ourense, Spain) for assistance with SPME–GC–MS methodology. This study was financially supported by the Government of Algeria, within The Program National Exceptional (PNE). K. Belkheir acknowledge receive a research fellowship from the Algerian Ministry of High Studies and Scientific Research, and the University of Relizane during these studies. This work was also partially supported by the Xunta de Galicia (The Spanish Regional Government) under the Consolidation and Restructuring Program of Competitive Research Units: Strategic Research Partnerships (2009/060). REFERENCES Akhmetsadykova S.H., Baubekova A., Konuspayeva G., Akhmetsadykova N., Faye B. and Loiseau G. 2015. Lactic acid bacteria biodiversity in raw and fermented camel milk. Afr. J. Food Sci. Technol. 6(3):84. http://dx.doi.org/10.14303/ajfst.2015.026. ! Ital. J. Food Sci., vol 28, 2016 - 609 Arahal D.R., Sánchez E., Macián M.C. and Garay E. 2008. Value of recN sequences for species identification and as a phylogenetic marker within the family “Leuconostocaceae”. Int. Microbiol. 11(1):33. http://dx.doi.org/10.2436/20.1501.01.42. Ashmaig A., Hasan A. and El Gaali E. 2009. Identification of lactic acid bacteria isolated from traditional Sudanese fermented camel’s milk (Gariss). Afr. J. Microbiol. Res. 3(8):451. Bendimerad N., Kihal M. and Berthier F. 2012. Isolation, identification, and technological characterization of wild leuconostocs and lactococci for traditional Raib type milk fermentation. Dairy Sci. Technol. 92(3):249. http://dx.doi.org/10.1007/s13594-012-0063-8. Boudjenah-Haroun S., Laleye C.L., Senoussi C., Moulti Mati F., Si Ahmed S. and Mati A. 2012. Coagulation of camel milk using dromedary gastric enzymes as a substitute of the commercial rennet. Am. J. Food Technol. 7(7):409. http://dx.doi.org/10.3923/ajft.2012.409.419. Buffa M., Guamis B., Saldo J. and Trujillo A.J. 2004. Changes in organic acids during ripening of cheeses made from raw, pasteurized or high-pressure-treated goats’ milk. Lebensm. Wiss. Technol. 37(2):247. http://dx.doi.org/10.1016/j.lwt.2003.08.006. Centeno J.A., Gaya P., Medina M. and Núñez M. 2002. Cross-inhibition among wild strains of Lactococcus lactis isolated from the same ecological niche. J. Food Protect. 65(1):205. Church F.C., Swaisgood H.E., Porter D.H. and Catignani G.L. 1983. Spectrophotometric assay using o-phthaldialdehyde for determination of proteolysis in milk and isolated milk proteins. J. Dairy Sci. 66(6):1219. http://dx.doi.org/10.3168/jds.S0022-0302(83)81926-2. Curioni P.M.G. and Bosset J.O. 2002. Key odorants in various cheese types as determined by gas chromatography- olfactometry. Int. Dairy J. 12(12):959. http://dx.doi.org/10.1016/S0958-6946(02)00124-3. Delgado F., González-Crespo J., Cava R. and Ramírez R. 2011. Formation of the aroma of a raw goat milk cheese during maturation analysed by SPME–GC–MS. Food Chem. 129(3):1156. http://dx.doi.org/10.1016/j.foodchem.2011.05.096. Drici H., Gilbert C., Kihal M. and Atlan D. 2010. Atypical citrate-fermenting Lactococcus lactis strains isolated from dromedary’s milk. J. Appl. Microbiol. 108(2):647. http://dx.doi.org/10.1111/j.1365-2672.2009.04459.x. Dutcosky S.D. 1996. “Sensory analysis of foods”. Ed. da Champagnat, Curitiba, Brasil. El Soda M. and Desmazeaud M.J. 1982. Les peptides-hydrolases de lactobacilles du groupe Thermobacterium. 1. Mise en evidence de ces activites chez Lactobacillus helveticus, L. acidophilus, L. lactis et L. bulgaricus. Can. J. Microbiol. 28(10):1181. http://dx.doi.org/10.1139/m82-174. El Zubeir I.E.M. and Jabreel S.O. 2008. Fresh cheese from camel milk coagulated with Camifloc. Int. J. Dairy Technol. 61(1):90. http://dx.doi.org/10.1111/j.1471-0307.2008.00360.x. Engels W.J.M., Dekker R., De Jong C., Neeter R. and Visser S. 1997. A comparative study of volatile compounds in the water-soluble fraction of various types of ripened cheese. Int. Dairy J. 7(4):255. http://dx.doi.org/10.1016/S0958- 6946(97)00003-4 Farah Z. 1993. Composition and characteristics of camel milk. J. Dairy Res. 60(4):603. Farah Z. and Bachman M.R. 1987. Rennet coagulation properties of camel milk. Milchwissenschaft 42(11): 689. Garabal J.I., Rodríguez-Alonso P. and Centeno J.A. 2008. Characterization of lactic acid bacteria isolated from raw cows’ milk cheeses currently produced in Galicia (NW Spain). LWT – Food Sci. Technol. 41(8):1452. http://dx.doi.org/10.1016/j.lwt.2007.09.004. Garneau J.E. and Moineau S. 2011. Bacteriophages of lactic acid bacteria and their impact on milk fermentations. Microb. Cell Factories 10 (Suppl. 1): S20. http://dx.doi.org/10.1186/1475-2859-10-S1-S20. González de Llano D., Rodríguez A. and Cuesta P. 1996. Effect of lactic starter cultures on the organic acid composition of milk and cheese during ripening–analysis by HPLC. J. Appl. Bacteriol. 80(5):570. http://dx.doi.org/10.1111/j.1365- 2672.1996.tb03259.x. Herrero M., Mayo B., González B. and Suárez J.E. 1996. Evaluation of technologically important traits in lactic acid bacteria isolated from spontaneous fermentations. J. Appl. Bacteriol. 81(5):565. http://dx.doi.org/10.1111/j.1365- 2672.1996.tb03548.x. ! Ital. J. Food Sci., vol 28, 2016 - 610 Herreros M.A., Fresno J.M., González Prieto M.J. and Tornadijo M.E. 2003. Technological characterization of lactic acid bacteria isolated from Armada cheese (a Spanish goats’ milk cheese). Int. Dairy J. 13(6):469. http://dx.doi.org/10.1016/S0958-6946(03)00054-2. IDF, 1995. IDF guideline-determination of acidifying activity of dairy cultures. In: IDF Bulletin 306, p. 34-36. International Dairy Federation, Brussels. IDF, 1997. Dairy starter cultures of lactic acid bacteria (LAB). Standard of identity. IDF Standard 149A. International Dairy Federation, Brussels. Kaminarides S., Stamou P. and Massouras T. 2007. Changes of organic acids, volatile aroma compounds and sensory characteristics of Halloumi cheese kept in brine. Food Chem. 100(1):219. http://dx.doi.org/10.1016/j.foodchem.2005.09.039. Kondyli E., Massouras T., Katsiari M.C. and Voutsinas L.P. 2003. Free fatty acids and volatile compounds in low-fat Kefalograviera-type cheese made with commercial adjunct cultures. Int. Dairy J. 13(1): 47. http://dx.doi.org/10.1016/S0958-6946(02)00144-9. Konuspayeva G., Faye B. and Loiseau G. 2009. The composition of camel milk: A meta-analysis of the literature data. J. Food Compos. Anal. 22(2):95. http://dx.doi.org/10.1016/j.jfca.2008.09.008. Limonet M., Cailliez-Grimal C., Linder M., Revol-Junelles A.M. and Millière J.B. 2004. Cell envelope analysis of insensitive, susceptible or resistant strains of Leuconostoc and Weissella genus to Leuconostoc mesenteroides FR 52 bacteriocins. FEMS Microbiol. Lett. 241(1):49. http://dx.doi.org/10.1016/j.femsle.2004.10.002. Lorenzo J.M. 2014. Influence of the type of fiber coating and extraction time on foal dry-cured loin volatile compounds extracted by solid-phase microextraction (SPME). Meat Sci. 96(1):179. http://dx.doi.org/10.1016/j.meatsci.2013.06.017. Lorenzo J.M. and Fonseca S. 2014. Volatile compounds of Celta dry-cured ‘lacón’ as affected by cross-breeding with Duroc and Landrace genotypes. J. Sci. Food Agr. 94(14):2978. http://dx.doi.org/10.1002/jsfa.6643. Macedo A.C., Vieira M., Poças R. and Malcata F.X. 2000. Peptide hydrolase system of lactic acid bacteria isolated from Serra da Estrela cheese. Int. Dairy J. 10(11):769. http://dx.doi.org/10.1016/S0958-6946(00)00111-4. Marco A., Navarro J.L. and Flores M. 2004. Volatile compounds of dry-fermented sausages as affected by solid-phase microextraction (SPME). Food Chem. 84(4):633. http://dx.doi.org/10.1016/S0308-8146(03)00288-7. McSweeney P.L.H. and Sousa M.J. 2000. Biochemical pathways for the production of flavor compounds in cheese during ripening: a review. Lait 80(3):293. http://dx.doi.org/10.1051/lait:2000127. Menéndez S., Centeno J.A., Godínez R. and Rodríguez-Otero J.L. 2000. Effect of Lactobacillus strains on the ripening and organoleptic characteristics of Arzúa-Ulloa cheese. Int. J. Food Microbiol. 59(1-2):37. http://dx.doi.org/10.1016/S0168- 1605(00)00286-5. Nandan A., Gaurav A., Pandey A. and Madhavan Nampoothiri K. 2010. Arginine specific aminopeptidase from Lactobacillus brevis. Braz. Arch. Biol. Technol. 53(6):1443. http://dx.doi.org/10.1590/S1516-89132010000600021. Nickels C. and Leesment H. 1964. Method for the differentiation and qualitative determination of starter bacteria. Milchwissenschaft 19: 374. Nieto-Arribas P., Seseña S., Poveda J.M., Palop L. and Cabezas L. 2009. Genotypic and technological characterization of Lactococcus lactis isolates involved in processing of artisanal Manchego cheese. J. Appl. Microbiol. 107(5):1505. http://dx.doi.org/10.1111/j.1365-2672.2009.04334.x. Nieto-Arribas P., Seseña S., Poveda J.M., Palop L. and Cabezas L. 2010. Genotypic and technological characterization of Leuconostoc isolates to be used as adjunct starters in Manchego cheese manufacture. Food Microbiol. 27(1):85. http://dx.doi.org/10.1016/j.fm.2009.08.006. Rodríguez-Alonso P., Garabal J.I. and Centeno J.A. 2008. Preliminary characterization of staphylococcal, micrococcal and yeast isolates obtained from raw cow milk cheeses currently produced in Galicia (NW Spain). Ital. J. Food Sci. 20(2):161. Settanni L. and Moschetti G. 2010. Non-starter lactic acid bacteria used to improve cheese quality and provide health benefits. Food Microbiol. 27(6):691. http://dx.doi.org/10.1016/j.fm.2010.05.023. Sreekumar R., Al-Attabi Z., Deeth H.C. and Turner M.S. 2009. Volatile sulfur compounds produced by probiotic bacteria in the presence of cysteine or methionine. Lett. Appl. Microbiol. 48(6):777. http://dx.doi.org/10.1111/j.1472- 765X.2009.02610.x. ! Ital. J. Food Sci., vol 28, 2016 - 611 Urbach G. 1995. Contribution of lactic acid bacteria to flavour compound formation in dairy products. Int. Dairy J. 5(8):877. http://dx.doi.org/10.1016/0958-6946(95)00037-2. Vedamuthu E.R. 1994. The dairy leuconostocs: use in dairy products. J. Dairy Sci. 77 (9):2725. http://dx.doi.org/10.3168/jds.S0022-0302(94)77215-5. Wood B.J.B. and Holzapfel W.H. 1995. The Genera of Lactic Acid Bacteria, p. 398-399. Blackie Academic and Professional, Glasgow. Wouters J.T.M., Ayad E.H.E., Hugenholtz J. and Smit G. 2002. Microbes from raw milk for fermented dairy products. Int. Dairy J. 12(2-3):91. http://dx.doi.org/10.1016/S0958-6946(01)00151-0. Paper Received December 30, 2015 Accepted May 19, 2016