Journal of Biotechnology 209 (2015) 31–40

Contents lists available at ScienceDirect

Journal of Biotechnology journal homepage: www.elsevier.com/locate/jbiotec

Simultaneous synthesis of mixtures of lactulose and galacto-oligosaccharides and their selective fermentation Cecilia Guerrero, Carlos Vera, Fernando Acevedo, Andrés Illanes ∗ School of Biochemical Engineering, Pontificia Universidad Católica de Valparaíso, Valparaíso, Chile

a r t i c l e

i n f o

Article history: Received 4 March 2015 Received in revised form 29 May 2015 Accepted 8 June 2015 Available online 12 June 2015 Keywords: Lactulose Galacto-oligosaccharides Probiotics Prebiotics ␤-galactosidase

a b s t r a c t Lactulose and galacto-oligosaccharides (GOS) are well recognized prebiotics derived from lactose. In the synthesis of lactulose with ␤-galactosidases GOS are also produced, but the ratio of lactulose and GOS in the product can be tuned at will, depending on the operation conditions, so to obtain an optimal product distribution in terms of prebiotic potential. The selectivity of fermentation of each carbohydrate alone as well as mixtures of both was determined using pH-controlled anaerobic batch cultures with faecal inoculum. Within the experimental range considered, lactulose/GOS molar ratio of 4 resulted in the highest selectivity for Bifidobacterium and Lactobacillus/Enterococcus, so this ratio was selected as the target for the synthesis of lactulose from fructose and lactose with Aspergillus oryzae ␤-galactosidase. Synthesis was optimized using response surface methodology, considering temperature, initial concentrations of acceptor sugars and fructose/lactose molar ratio as key variables, with the aim of maximizing lactulose yield at the optimal product distribution in terms of prebiotic potential (lactulose/GOS molar ratio of 4). Under optimal conditions (50 ◦ C, 50%w/w total initial concentrations of sugars and fructose/lactose molar ratio of 6.44), lactulose yield of 0.26 g of lactulose produced per g of initial lactose was obtained at the optimal product distribution. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Prebiotics have emerged in recent years as a new category of functional foods (Gibson, 2006; Lamsal, 2012). Prebiotics are mostly non-digestible oligosaccharides that are selectively fermented in the colon, increasing the number of Bifidobacterium and Lactobacillus in the intestinal microbiota. These microorganisms are generally considered to be health promoting, inhibiting the growth of pathogenic bacteria and stimulating immunity, possibly increasing resistance to infections (Gibson, 2006; Pan et al., 2009; Lamsal, 2012). Short chain fatty acids (SCFA) are products of the saccharolytic activity of the intestinal microbiota and exert various effects by supplying energy to the intestinal mucosa, lowering pH and stimulating calcium and water absorption (Blottiere et al., 1999; Pryde et al., 2002; Pan et al., 2009). Several non-digestible oligosaccharides (NDOs) have been considered as potential prebiotics, namely: fructo-oligosaccharides (FOS), galacto-oligosaccharides (GOS), xylo-oligosaccharides (XOS), isomalto-oligosaccharides (IMO), soybean-oligosaccharides (SOS), lactulose (Lu) and inulin (IN) (Rycroft et al., 2001; Sanz et al., 2005; Rodriguez-Colinas et al.,

∗ Corresponding author. Fax: +56 32 2273803. E-mail address: [email protected] (A. Illanes). http://dx.doi.org/10.1016/j.jbiotec.2015.06.394 0168-1656/© 2015 Elsevier B.V. All rights reserved.

2013). However, only a few satisfy all criteria to be considered as such, GOS and Lu being among them (Rycroft et al., 2001; Sanz et al., 2005; Gibson, 2006). The enzymatic synthesis of GOS and Lu with ␤-galactosidases has been well documented (Albayrak and Yang, 2002; Lee et al., 2004; Kim et al., 2006; Vera et al., 2012). ␤-Galactosidase is a commodity enzyme used in the food industry for the production of low-lactose milk and dairy products. In recent years the enzyme, although being a hydrolase, has been used as a catalyst for transgalactosylation reactions and the synthesis of GOS and Lu with ␤-galactosidases has been well documented (Albayrak and Yang, 2002; Lee et al., 2004; Kim et al., 2006; Vera et al., 2012). This is feasible as long as the hydrolytic potential of the enzyme is depressed, which can be done by proper engineering of reaction conditions. In transgalactosylation, one molecule of galactose is transferred from the non-reducing end of a ␤-galactoside to a hydroxyl-bearing receptor molecule other than water. In the case of the synthesis of GOS with ␤-galactosidases, lactose acts both as donor and acceptor of the transgalactosylated galactose (Albayrak and Yang, 2002; Vera et al., 2012). In general, ␤-galactosidases are rather non-specific with respect to galactose acceptor molecules, so other sugars may act as such (Lee et al., 2004; Kim et al., 2006; Guerrero et al., 2013). In the synthesis of Lu with ␤-galactosidases, lactose is the galactosyl donor and fructose is the acceptor; however, since lactose and

32

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

fructose are present in the reaction medium, both can be acceptors of transgalactosylated galactose, so inevitably a mixture of Lu and GOS is produced (Guerrero et al., 2011). If the task is to produce pure Lu, this is a drawback; however, both NDOs are high added value products whose prebiotic properties are well established (De Preter et al., 2006; Rastall, 2010) so their mixture is likely to preserve or even enhance the prebiotic effect. It has been demonstrated that by proper manipulation of reaction conditions (mostly by varying the ratio of substrates) it is possible to obtain a wide range of Lu/GOS ratios (Guerrero et al., 2011). The hypothesis is that one of such ratios is likely to be optimum in terms of prebiotic effect; this is based on previous studies with other oligosaccharide mixtures, such as GOS/FOS and FOS/IN, where increases in the concentration of Bifidobacterium and Lactobacillus/Enterococcus have been reported to be higher than with each NDO used alone, so the prebiotic effect was enhanced (Vos et al., 2006; Ghoddusi et al., 2007). Since GOS formation is inevitable during the enzymatic synthesis of lactulose, the effect of such mixtures on the most numerically predominant and functionally relevant populations of the intestinal microbiota was assessed to appreciate their prebiotic potential. After determining the best Lu/GOS ratio, their simultaneous synthesis from lactose and fructose was carried out with Aspergillus oryzae ␤-galactosidase, evaluating the effect of temperature, fructose/lactose ratio and total sugars concentration on the corresponding Lu and GOS yields and productivities, and on Lu/GOS selectivity. Then yield of both prebiotics was optimized at a determined Lu/GOS ratio using response surface methodology. 2. Materials and methods 2.1. Materials Lactulose (4-O-␤-D-galactopyranosyl-D-fructose) was provided by Sigma Chemical Co. (St Louis, MO, USA). The commercial GOS preparation (BiMuno) was further purified to 99% GOS at the Department of Food and Nutrition Sciences of the University of Reading, UK and commercial GOS preparation Cup-Oligo (70% GOS) was supplied by Kowa Europe GmbH (Düsseldorf, Germany). o-Nitrophenol (o-NP) and o-nitrophenyl-␤-D-galactopyranoside (o-NPG) and GOS standards (4␤-galactobiose and 3␣-4␤-3␣ galactotetraose) were supplied by Sigma (St Louis, MO, USA). D(+)-galactose, D-fructose, ␣-lactose monohydrate and all other reagents used for fermentation and enzymatic synthesis were analytical grade and provided either by Sigma, Merck (Darmstadt, Germany) or Oxoid Ltd (Basingstoke, UK). A commercial ␤-galactosidase preparation of Aspergillus oryzae, under the trade name Enzeco® Fungal Lactase Concentrate, was a kind gift from Enzyme Development Corporation (New York, USA). Specific activity of the enzyme was 196,000 IUH g−1 ; one international unit of activity of ␤-galactosidase (IUH ) was defined as the amount of enzyme hydrolyzing 1 ␮mole of o-NPG per minute at pH 4.5, 40 ◦ C and 30 mmol L−1 o-NPG. The enzyme was stored at 4 ◦ C and retained full activity throughout the work period.

sition, it was assumed that the area of each peak is proportional to the weight percentage of the respective sugar. This assumption was validated by a material balance. Standards of galactose, fructose, lactulose, 4␤-galactobiose and 3␣-4␤-3␣ galactotetraose were used to determine their retention times and check linear range of the measurements. 2.3. Determination of prebiotic capacity in pure cultures Bifidobacterium longum code NB667, Bifidobacterium lactis code Bb12 (ATCC 27536), Lactobacillus plantarum code CIDCA83114 and Lactobacillus rhamnosus code GC (ATCC53103). All strains were provided by the Research Centre in Food Cryobiology (CIDCA, La Plata, Argentina) being the selected strains for the determination of prebiotic capacity in pure cultures since these microorganisms are highly representative of the intestinal microbiota and considered as probiotics (Vernazza et al., 2006; Huebner et al., 2007; Vasiljevic and Shah, 2008). Fermentations were carried out at 37 ◦ C, 280 rpm and pH 7 under anaerobic conditions in a series of 10 sealed vials to preserve anaerobic conditions, each vial corresponding to a sample point during fermentation. Culture medium was the one reported by Rycroft et al. (2001) and Vernazza et al. (2006). The basal medium contained in g L−1 : peptone water: 2; yeast extract: 2; NaCl: 0.1; K2 HPO4 : 0.04; KH2 PO4 : 0.04; MgSO4 7H2 O: 0.01; CaCl2 6H2 O: 0.01; NaHCO3 : 2; haemin: 0.05; L-cysteine: 0.5; HCl: 0.5; bile salts: 0.5; 2 mL L−1 of Tween 80 and 10 ␮L L−1 of vitamin K1 . Due to the diversity of carbon sources, comparison among cultures was done on the basis of constant carbon atoms available for each substrate. Carbon sources evaluated were: lactulose, GOS (Cup Oligo) and mixtures thereof (at lactulose/GOS molar ratios of 0.5, 1 and 2). Glucose at 10 g L−1 was used as control carbon source, since it has no prebiotic effect and is rapidly metabolized by all the microorganisms tested. Cup Oligo contains 78.6 %w/w of GOS (Tri, tetra and pentasaccharides), 18.9% of disaccharides (GOS-2 and lactose) and less than 2% of monosaccharides. 2.4. pH-controlled batch cultures for determination of prebiotic capacity Water-jacketed 80 mL fermenters were filled with 22.5 mL of the basal medium described above and inoculated with 2.5 mL of faecal slurry. All tested substrates, Lu, purified GOS (99%w/w) and mixtures of both (Lu/GOS molar ratio of 0.5, 1 and 4), were added separately just before inoculation to give a final concentration of 1% (w/v). Faecal slurries (10% v/v) were prepared by homogenizing freshly voided human faeces in 0.1 M phosphate buffer solution (PBS) pH 7. Each vessel was magnetically stirred and maintained under anaerobic conditions with oxygen-free nitrogen. Culture pH was controlled automatically during the fermentations at 6.8 and temperature was maintained at 37 ◦ C (Rycroft et al., 2001; Sanz et al., 2005). Samples were removed from the fermenters at 0, 4, 8 and 24 h of incubation for enumeration of bacteria and determination of SCFA and lactic acid. The experiments were performed in triplicate, using one faecal sample from three different donors for each run of batch cultures with five different substrates each.

2.2. HPLC analysis of the reaction products 2.5. Enumeration of bacteria Substrates and products of synthesis were analyzed in a Jasco RI 2031HPLC system, provided with refractive index detector, isocratic pump (Jasco PU2080) and autosampler (Jasco AS 2055), using BP-100Ca++ columns (300 mm × 7.8 mm) for carbohydrate analysis (Benson Polymeric, Reno, USA). Samples were eluted with milli-Q water at a flow-rate of 0.5 mL min−1 . Temperatures in the column and in the detector were 80 ◦ C and 40 ◦ C, respectively. Chromatograms were integrated using the Jasco ChromPass software 1.7.403.1 (Maryland, USA). To determine sample compo-

Determination of biomass in pure cultures was done by turbidimetry at 650 nm and dry weight. In pH-controlled batch cultures fluorescent in situ hybridization technique (FISH) was used to quantify selected bacterial groups of the colonic microbiota. At 0, 4, 8 and 24 h of fermentation, 375 ␮L samples were removed from the fermenters and added to 1.125 mL of filtered 4% (w/v) paraformaldehyde solution pH 7.2, mixed and stored at 4 ◦ C for 6 h to fix the cells. Hybridization of the samples was carried out

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

33

Table 1 16S rRNA oligonucleotide probes used in this study. Probe name

Specificity

References

Ato291 Bac303 Bif164 Chis150 Erec482 Lab158 Rrec584 Prop853 Fpra655

Atopobium cluster Most Bacteroidaceae and Prevotellaceae, some Porphyromonadaceae Bifidobacterum spp. Most of the Clostridium histolyticum group (Clostridium cluster I and II) Most of the Clostridium coccoides-Eubacterium rectale group (Clostridium cluster XIVa and XIVb) Lactobacillus-Enterococcus group Roseburia genuos Clostridium cluster IX Faecalibacterium prausnitzii and relatives

Harmsen et al. (2000) Manz et al.(1996) Langendijk et al. (1995) Franks et al. (1998) Franks et al. (1998) Harmsen et al. (1999) Walker et al. (2005) Walker et al. (2005) Hold et al. (2003)

as described by Sarbini et al. (2011), using the appropriate genus specific 16 rRNA-targeted oligonucleotide probes labelled with the fluorescent dye Cy3 for the different target bacterial groups or the nucleic acid stain 4 ,6-diamidino-2-phenylindole (DAPI) for total cell counts (Table 1). The bacterial groups studied were chosen as representative of the most numerically predominant and functionally relevant in the human colon. A minimum of 15 random fields of view were counted per sample using a Nikon Eclipse E400 fluorescent microscope with filters for DAPI (excitation at 550 nm and emission at 461 nm) and Cy3 (excitation at 550 nm and emission at 564 nm). 2.6. Statistical analyses To determine if differences observed in the microbial populations obtained with the different carbon sources were statistically significant, ANOVA test was performed with the result that differences observed among populations cultured in the different carbon sources were in all cases significant at a confidence interval of 95%. 2.7. Enzymatic synthesis of lactulose and GOS with ˇ-galactosidase from Aspergillus oryzae Synthesis reactions were done with A. oryzae ␤-galactosidase in 150 mL Erlenmeyer flasks under magnetic stirring. Sugar substrates were dissolved in 100 mM citrate-phosphate buffer pH 4.5 previously heated at 95 ◦ C and then cooled down to the reaction temperature. Then, 10 g of enzyme solution in 100 mM McIlvaine citrate-phosphate buffer at pH 4.5 was added. Temperature, total initial concentration of sugars and fructose/lactose molar ratio were selected as key variables with respect to selectivity, expressed in terms of molar ratio of lactulose and GOS (RLu/GOS ), and yield (YLu ), as defined below (Guerrero et al., 2011). The effect of temperature (30–60 ◦ C), total initial concentration of sugars (30–70%w/w) and fructose/lactose molar ratio (1–12) were evaluated separately to select the appropriate ranges to be used in the optimization of synthesis. A Box–Behnken statistical design for three variables at three levels was selected. RLu/GOS and YLu were the parameters used as objective functions of the experimental design, both being determined at the maximum Lu concentration attained. Reactions were conducted at pH 4.5 (100 mM McIlvaine citrate-phosphate buffer), 200 IUH g−1 of lactose at varying temperatures (A), total initial concentration of sugars (B) and fructose/lactose molar ratio (C), according to the experimental design. Statistical analysis of data and enzymatic synthesis optimization were carried out using the software Design-Expert 8.0.6, trial version. Time course of the reaction was followed during 10 h. Product distribution was then determined by analyzing the amounts of Lu and GOS (GOS-3, GOS-4 and GOS-5) synthesized, numbers in GOS- representing the monosaccharide units of each GOS. The assays were carried out in duplicate, with standard deviations never exceeding 5%. Quantification of sugars was carried out as described in Section 2.2.

The following parameters were defined: -Lactulose yield (YLu ), which represents the fraction of the initial mass of lactose (ML ) that is converted into Lu (MLu ) at the maximum Lu concentration attained in the reaction: YLu =

MLu ML

(1)

Yield is expressed with respect to the limiting substrate lactose, but it can also be referred to the substrate in excess (fructose), in which case it represents the fraction of the initial mass of fructose that is converted into Lu at the maximum Lu concentration attained in the reaction. -GOS yield (YGOS ), which represents the fraction of the initial mass of lactose (ML ) that is converted into GOS (MGOS ) at the maximum Lu concentration attained in the reaction: YGOS =

MGOS ML

(2)

-Lu/GOS molar ratio (RLu/GOS ), which represents the molar ratio of Lu to total GOS at the maximum Lu concentration attained in the reaction: R

Lu GOS

=

[Lu] [GOS]

(3)

where brackets represent molar concentration. -Specific productivity of Lu (␲Lu ), which represents the amount of Lu produced (MLu ) per unit mass of enzyme preparation (ME ) and unit time (t) at the maximum Lu concentration attained in the reaction: Lu =

MLu ME × t

(4)

3. Results and discussion 3.1. Fermentation selectivity of lactulose and galacto-oligosaccharides mixtures in pure cultures The effect of lactulose, GOS and their mixtures was first evaluated in pure cultures of four probiotic bacteria representative of the intestinal microbiota. Fig. 1a shows the biomass concentration reached at stationary phase with the different substrates analyzed. Lu/GOS mixtures produced an increase in the biomass concentration obtained with B. lactis, B. longum and L. plantarum with respect to the control (glucose), while the opposite occurred with L. rhamnosus. In the case of Lu, increase with respect to the control and GOS was observed with all strains. These experiments confirm that the selected probiotic strains are able to metabolize all carbon sources analyzed separately. In fact, among pure sugars, higher growth rates in all strains were obtained in lactulose, being within the ranges reported for them (Sahota et al., 1982; Vernazza et al., 2006; Depeint et al., 2008; Rada et al., 2008; Cardelle-Cobas et al., 2011). It is now demonstrated that lactulose and GOS are also metabolized when present in mixtures.

34

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

3

3

(b)

2

2.5

Prebiotic activity score (PAS)

Biomass Concentration (g.L-1)

(a)

2

1.5

1

0.5

1

0

-1

-2

-3

-4

0

Fig. 1. (a) Biomass concentration obtained in anareobic cultures of Bifidobacterium lactis (䊏), Bifidobacterium longum ( plantarum (䊐) a 37 ◦ C. (b) Prebiotic activity score, as defined by Eq. 5, at 24 h of cultivation).

Fig. 1b shows the prebiotic activity score (PAS) calculated at 24 h of fermentations, according to the formula proposed by Huebner et al. (2007). PAS values were calculated according to Eq. (5) in order to determine the global effect of the different mixtures of lactulose and GOS on the bacterial populations: PAS =

Log(Pt ) − Log(Pt=0 ) Log(Ptg ) − Log(Ptg=0 )

(5)

where Pt = 0 and Pt are the concentrations of probiotic bacteria cultivated with prebiotic carbon source at 0 and 24 h, respectively, and Ptg = 0 and Ptg are their corresponding concentrations in glucose as carbon source (control). PAS is a sound parameter to compare microbial growth in probiotic and non-probiotic substrates. Therefore, substrates with a positive PAS are better metabolized than the control and are therefore selectively metabolized by the probiotic bacteria. In this way, PAS allows a first approximation to establish the prebiotic condition of the oligosaccharides and establish the existence or not of a synergy between them. As seen in Fig. 1b, Lu and Lu/GOS molar ratios higher than 1 produced positive values of PAS. However, since the microbial populations of the intestinal microbiota behave as a consortium, the determination of prebiotic effect in pure cultures is just a first approximation to determine it. 3.2. Fermentation selectivity of lactulose and galacto-oligosaccharides mixtures in pH-controlled batch cultures The variation in the predominant bacterial species in the intestinal microbiota as a consequence of the fermentation of the different carbon sources tested (Lu, GOS and Lu/GOS mixtures) is shown in Table 2. All tested substrates produced a significant increase in the concentration of Bifidobacterium, Lactobacillus/Enterococcus and Atopobium populations (enumerated by probes Bif 164, Lab 158 and Ato 291, respectively), which is in agreement with previously reported results for Lu (see Table 2) and GOS (Rycroft et al., 2001; Palframan et al., 2002; Sanz et al., 2005; Beards et al., 2010). Slight differences in Bifidobacterium spp. were observed, with a Lu/GOS molar ratio of 4 producing the highest increase. However, Lactobacillus/Enterococcus and Atopobium concentrations were similar with all carbon sources tested (Table 2). High Bifidobacterium and Lactobacillus/Enterococcus concentrations are desirable, as these

), Lactobacillus rhamnosus (

) and Lactobacillus

are generally considered to be the best health-promoting species within the intestinal microbiota with proven beneficial effects in digestion and nutrient absorption, pathogen prevention and stimulation of the immune system (Gopal et al., 2001; Vasiljevic and Shah, 2008). These bacteria are also associated with the production of SCFA and lactate, compounds to which a significant part of the prebiotic effect may be assigned (Rycroft et al., 2001; Sanz et al., 2005; Salazar et al., 2009). As seen in Table 2, a small increase in Bacteroides/Prevotella populations occurred in all carbon sources during fermentation. However, increase was higher for Lu/GOS mixtures than for both sugars alone, which is an agreement with previous reports about the use of Lu and GOS as carbon sources (Rycroft et al., 2001; Sanz et al., 2005). The Bacteroides/Prevotella is one of the main groups of bacteria involved in the digestion of complex oligo and polysaccharides in the colon (Salazar et al., 2009). Increase in the population of Bacteroides/Prevotella is related to elevated acetate and propionate production, even though it also depends on culture conditions (Salazar et al., 2009). However, propionate production is not only related to Bacteroides/Prevotella being also a final fermentation product of other microbial populations, such as Eubacterium and Clostridium cluster IX (enumerated by probe Prop 853) (Rycroft et al., 2001; Sanz et al., 2005; Beards et al., 2010); therefore, propionate production is also related to the levels of such bacterial populations as well (Walker et al., 2005). Eubacterium rectale/ Clostridium coccoides group, belonging to Clostridium cluster XIVa and XIVb (enumerated by probe Erec 482) increased during fermentation when using Lu, GOS and mixtures thereof, but no significant differences among the carbohydrate sources was observed. The bacterial groups enumerated by probe Chis 150, Erec 482, Rrec 584 and Fpra 655 (Table 1) are the main producers of butyrate in the human colon (Pryde et al., 2002; Walker et al., 2005; Salazar et al., 2009). Butyrate is of interest due to its potentially protective role in ulcerative colitis and colon cancer (Hague et al., 1995; Simpson et al., 2000; Pryde et al., 2002; Lupton, 2004). As seen in Table 2, Roseburia (enumerated by probe Rrec 584) and Faecalibacterium prausnitzii (enumerated by probe Fpra 655) populations decreased significantly during batch fermentations with all test substrates. The magnitude of decrease in these bacterial populations was more significant at higher Lu/GOS ratio, being

Carbon source Lu

GOS

Lu/GOS (0.5)

Lu/GOS (1)

Lu/GOS (4)

Time

Bif 164 (Log10 mL−1 )

Bac 303 (Log10 mL−1 )

Chis 150 (Log10 mL−1 )

Lac 158 (Log10 mL−1 )

Ato 295 (Log10 mL−1 ))

Erec 482 (Log10 mL−1 )

Rrec 584 (Log10 mL−1 )

Prop 853 (Log10 mL−1 )

Fpra 655 (Log10 mL−1 )

DAPI (Log10 mL−1 )

0h 4h 8h 24 h 4h 8h 24 h 4h 8h 24 h 4h 8h 24 h 4h 8h 24 h

7.59 (±0.03) 8.2 (±0.204)* 8.91 (±0.161)* 9.15 (±0.05)* 8.44 (±0.059)* 8.9 (±0.127)* 9.11 (±0.151)* 8.24 (±0.086)* 8.837 (±0.123)* 9.14 (±0.137)* 8.29 (±0.049)* 9.02 (±0.039)* 9.21 (±0.199)* 8.50 (±0.181)* 8.92 (±0.102)* 9.24 (±0.136)*

8.4 (±0.002) 8.72 (±0.092) 8.80 (±0.075)* 8.89 (±0.075)* 8.74 (±0.088)* 8.82 (±0.065)* 8.99 (±0.065)* 8.81 (±0.029)* 8.9 (±0.074)* 8.98 (±0.118)* 8.76 (±0.099)* 8.85 (±0.095)* 9.03 (±0.072)* 8.8 (±0.033)* 8.9 (±0.047)* 8.98 (±0.114)*

8.1 (±0.031) 8.73 (±0.074)* 8.68 (±0.005)* 8.58 (±0.122)* 8.68 (±0.042)* 8.64 (±0.101)* 8.54 (±0.049)* 8.69 (±0.087)* 8.57 (±0.012)* 8.5 (±0.083)* 8.67 (±0.068)* 8.6 (±0.05)* 8.4 (±0.079) 8.64 (±0.061)* 8.55 (±0.035)* 8.3 (±0.126)

6.66 (±0.067) 7.72 (±0.026)* 7.82 (±0.059)* 7.87 (±0.025)* 7.7 (±0.087)* 7.8 (±0.006)* 7.9 (±0.106)* 7.69 (±0.026)* 7.76 (±0.046)* 7.92 (±0.053)* 7.7 (±0.02)* 7.79 (±0.021)* 7.88 (±0.04)* 7.74 (±0.027)* 7.83 (±0.026)* 7.92 (±0.031)*

7.98 (±0.137) 8.57 (±0.029)* 8.82 (±0.127)* 9.29 (±0.011)* 8.59 (±0.07)* 8.79 (±0.134)* 9.25 (±0.073)* 8.53 (±0.04)* 8.85 (±0.152)* 9.2 (±0.095)* 8.62 (±0.04)* 9.01 (±0.114)* 9.22 (±0.121)* 8.56 (±0.045)* 9.05 (±0.113)* 9.28 (±0.1)*

8.44 (±0.038) 8.93 (±0.066)* 8.91 (±0.095)* 8.85 (±0.071)* 8.91 (±0.055)* 8.86 (±0.074)* 8.83 (±0.078)* 8.96 (±0.063)* 8.93 (±0.065)* 8.88 (±0.063)* 8.97 (±0.051)* 8.89 (±0.129)* 8.82 (±0.073)* 8.94 (±0.034)* 8.89 (±0.051)* 8.71 (±0.096)

7.85 (±0.049) 8.17 (±0.144) 7.93 (±0.176) 7.65 (±0.067) 8.06 (±0.119) 7.89 (±0.065) 7.68 (±0.052) 8.14 (±0.15) 7.84 (±0.085) 7.59 (±0.032) 8.2 (±0.13)* 7.86 (±0.105) 7.59 (±0.064) 8.26 (±0.131)* 7.8 (±0.012) 7.54 (±0.058)

8.16 (±0.05) 8.46 (±0.155) 8.48 (±0.172) 8.69 (±0.089)* 8.36 (±0.144) 8.44 (±0.144) 8.59 (±0.098)* 8.38 (±0.034) 8.47 (±0.074) 8.56 (±0.135)* 8.44 (±0.082) 8.49 (±0.097) 8.6 (±0.086)* 8.4 (±0.061) 8.48 (±0.025) 8.64 (±0.038)*

8.3 (±0.059) 8.59 (±0.115) 8.31 (±0.024) 8.1 (±0.059) 8.57 (±0.117) 8.31 (±0.071) 8.11 (±0.085) 8.59 (±0.07) 8.33 (±0.111) 8.11 (±0.044) 8.51 (±0.033) 8.33 (±0.169) 8.05 (±0.060) 8.36 (±0.04) 8.23 (±0.108) 7.98 (±0.022)

8.86 (±0.1) 9.61 (±0.15)* 9.63 (±0.138)* 9.89 (±0.115)* 9.51 (±0.11)* 9.67 (±0.164)* 9.9 (±0.074)* 9.51 (±0.147)* 9.65 (±0.158)* 9.92 (±0.083)* 9.57 (±0.136)* 9.71 (±0.156)* 9.9 (±0.081)* 9.54 (±0.057)* 9.71 (±0.171)* 9.92 (±0.095)*

* Significant differences from initial value (0) h were obtained at P ≤ 0.05. Standard deviations for three volunteers at 4, 8 and 24 h fermentation are shown in parentheses (n = 3). Lu, lactulose, GOS, galacto-oligosaccharides, Lu/GOS, 0.5 mole of lactulose per mole of galacto-oligosaccharides, Lu/GOS, 1 mole of lactulose per mole of galacto-oligosaccharides and Lu/GOS, 4 moles of lactulose per mole of galacto-oligosaccharides.

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

Table 2 Variation of bacterial populations (Log10 cell mL−1 ) in pH-controlled batch cultures, using lactulose (Lu), galacto-oligosaccharides (GOS) and mixtures of both as fermentation substrates at 0, 4, 8 and 24 h.

35

36

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

6

6

(a)

(b) 5

Selectivity index (SI)

Selectivity index (SI)

5 4 3 2

4

3

2

1

1

0

0

Fig. 2. Selectivity index score (SI) of different mixtures of lactulose (Lu) and galacto-oligosaccharides (GOS) from pH-controlled batch culture fermentations (a) at 8 h and (b) at 24 h. According to ANOVA test, significant differences were determined for the different substrates used at P ≤ 0.05.

0.4

0.25

(a)

(b) 0.2

Y GOS (g. g -1)

Y Lu (g. g -1)

0.3

0.2

0.1

0.15 0.1 0.05

0

1

4

6

8

0

12

Molar ratio of fructose/lactose (mole/mole)

4

6

8

12

0.16

12

(d)

(c) 10

π Lu (g.g-1.h-1)

R Lu/GOS (mol.mol -1)

1

Molar ratio of fructose/lactose (mole/mole)

8 6 4

0.12

0.08

0.04 2 0

1

4

6

8

12

Molar ratio of fructose/lactose (mole/mole)

0

1

4

6

8

12

Molar ratio of fructose/lactose (mole/mole)

Fig. 3. Effect of temperature on (a) YLu , (b) YGOS , (c) RLu/GOS and (e) ␲Lu in the enzymatic synthesis of lactulose with A. oryzae ␤-galactosidase at 50% w/w total initial sugars concentration, pH 4.5 and 200 IU/g at different fructose/lactose molar ratios. 䊏: 30 ◦ C, : 40 ◦ C, : 50 ◦ C, :55 ◦ C and 䊐: 60 ◦ C.

the molar ratio of 4 the one at which the concentrations of Roseburia and Faecalibacterium prausnitzii were the lowest. Clostridium histolyticum group (enumerated by probe Chis 150) is a very important component of the intestinal microbiota, being the main producer of butyrate, CO2 and H2 (Ghoddusi et al., 2007); however, they are also associated with gas formation and the associated abdominal discomfort. A slight increase in Clostridium histolyticum group was observed in Table 2 with all test substrates with the Lu/GOS molar ratio of 4 producing the lowest increase. In summary, results obtained for all bacterial probes (Table 1) with Lu, GOS and mixtures thereof do not differ to a great extent. Their

effect on most test bacteria shows that by using Lu/GOS mixtures an increase in fermentation selectivity for health-promoting bacteria can be achieved with respect to the oligosaccharides alone (Lu or GOS), which is in agreement with results reported for other prebiotic mixtures (Vulevic et al., 2004; Vos et al., 2006; Ghoddusi et al., 2007). In this way, Lu/GOS mixtures could be used effectively as prebiotics with no need for fractionation which is a considerable process advantage. Fig. 2 shows the selectivity index (SI) calculated at 8 h and 24 h of fermentations, according to the formula proposed by Ruiz-Matute et al. (2011). SI (Eq. (6)) values were calculated in order to deter-

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

0.4

0.25

(a)

(b) 0.2

YGOS (g.g -1)

0.3

YLu (g.g -1)

37

0.2

0.15 0.1

0.1 0.05 0 1

4

6

8

0

12

1

Molar ratio of fructose/lactose (mole/mole) 12

6

8

12

0.1

(d)

(c) 10

π Lu (g.g-1 .h-1)

R Lu/GOS (mol.mol -1)

4

Molar ratio of fructose/lactose (mole/mole)

8 6

0.08 0.06 0.04

4 0.02

2 0 1

4

6

8

12

Molar ratio of frctose/lactose (mole/mole)

0 1

4

6

8

12

Molar ratio of fructose/lactose (mole/mole)

Fig. 4. Effect of total initial sugars concentration on (a) YLu , (b) YGOS , (c) RLu/GOS and (d) ␲Lu in the enzymatic synthesis of lactulose with A. oryzae ␤-galactosidase at 50 ◦ C, pH 4.5 and 200 IU/g at different fructose/lactose molar ratios. 䊏: 30 %w/w,

: 40 %w/w,

: 50 %w/w, 55

: 60% w/w and 䊐: 70 %w/w.

Fig. 5. Response surfaces showing the effect of temperature (A), total initial concentration of sugars (B) and fructose/lactose molar ratio (C) on YLu and RLu/GOSin the reaction of synthesis with A. oryzae ␤-galactosidase at pH 4.5 and 200 IUH/g lactose. (a): effect of A and B at C = 8 on YLu and (d) on RLu/GOS; (b): effect of B and C at A = 55 ◦ C on YLu and (e) on RLu/GOS; (c): effect of A and C at B = 60 %w/w on YLu and (f) on RLu/GOS.

38

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

mine the global effect of the different mixtures of lactulose and GOS on the bacterial populations. According to ANOVA test, significant differences were determined for the different substrates used at P ≤ 0.05.



SI=

Bif164t Bif1640

    Erec482t   Bac303t   Chis150t  Lac158t +

Lac1580

+



Erec4820

totalcountt totalcount0

−



Bac3030

−

Chis1500

(6) where the subscripts t and 0 refer to bacterial counts at 8 h or 24 h, and 0 h, respectively. Bif stands for Bifidobacterium, Lac for Lactobacillus, Bac for Bacteroides and Chis for Chlostridium. According to Figs. 1 and 2, a Lu/GOS molar ratio of 4 was the one producing the highest PAS and SI, being therefore selected as the target for the synthesis of Lu with Aspergillus oryzae ␤-galactosidase. 3.3. Effect of temperature, sugars concentration and fructose/lactose molar ratio on the synthesis of a mixture of lactulose and galacto-oligosaccharides Fig. 3 shows the effect of temperature and fructose/lactose ratio on YLu , YGOS , RLu/GOS and ␲Lu during the synthesis of Lu. At constant fructose/lactose ratio, temperature effect is significant only in ␲Lu , with a consistent increase with temperature in the studied range, being the increase more pronounced in the 50–60 ◦ C range at all fructose/lactose ratios evaluated. Fig. 4 shows the effect of total initial sugars concentration on the parameters of Lu synthesis. An increase from 30 to 50%w/w produced a slight increase in YLu and a reduction in YGOS , while in the 60–70%w/w range both yields decreased. That effect was observed at fructose/lactose molar ratios of 4 or higher, where the synthesis of Lu prevailed over GOS; at a 1:1 molar ratio, however, YLu did not vary with concentration while YGOS increased significantly in the 30–50%w/w range and decreased at higher concentrations. RLu/GOS values were not affected in the concentration range studied at fructose-lactose molar ratios over 1 (Fig. 4c), while ␲Lu increased in the 30–50% w/w range and decreased in the 60–70%w/w range at fructose-lactose molar ratios over 1 (Fig. 4d). As shown in Figs. 3 and 4, at all temperatures and total initial concentration of sugars, fructose/lactose ratio has a strong effect on yield and productivity of Lu, and also in RLu/GOS ,. Substrate ratio is then a key operational variable determining product distribution (RLu/GOS ). RLu/GOS increased steadily with fructose/lactose molar ratio as the synthesis of lactulose prevails, being 70% and 86% at fructose/lactose molar ratios of 4 and 12, respectively. Based on these results the experimental design was done establishing the range of each variable as the one at which major differences in the parameters of lactulose synthesis were observed: 50–60 ◦ C, 50–70%w/w total initial sugars concentrations and fructose/lactose molar ratios 4–12. 3.4. Optimization of the synthesis of a mixture of lactulose and galacto-oligosaccharides with Aspergillus oryzae ˇ-galactosidase The optimization of the enzymatic synthesis of Lu/GOS mixture was performed according to the experimental design. Eq. (7) was obtained by regression from the data obtained for YLu , describing the effect of the coded variables on it: YLu = 4.8 × 10−4 × A − 0.011 × B + 0.045 × C + 2.4 × 10−3 × A × B + 4.1 × 10−3 × A × C − 5.1 × 10−3 × B × C +5.23 × 10−4 × A2 − 0.013 × B2 − 0.017 × C2 − 5.10−4 × A × B × C + 0.28

(7)

Fig. 6. (a) Response surfaces showing the values of temperature (A), total initial concentration of sugars (B) and fructose/lactose molar ratio (C) that allow obtaining a lactulose/GOS molar ratio (RLu/GOS) of 4. (b) Estimated values of lactulose yield (YLu) for RLu/GOS = 4 as function of A and B.

where A represents the temperature, B the total initial concentration of sugars and C the substrates molar ratio. The coefficient of determination of Eq. (7) was 0.99 with adjusted coefficient of determination of 0.989, which means that 98.9% of the variability of data is explained by the equation. The magnitude of the coefficients in Eq. (7) accounts for the impact of each variable in YLu and the interaction among variables. The highest coefficient corresponds to C, the variable with stronger influence on YLu . Fig. 5a–c depicts the response surfaces obtained, showing the effect of the operational variables on YLu and their interactions. The most significant variables in YLu were B and C. As seen in Fig. 5a, YLu decreased with the increase in B (from 0.28 glactulose glactose −1 at B = 50%w/w to 0.20 g g−1 at B = 70%w/w). Fig. 5b shows the effect of B and C on YLu ; YLu increased with C in the range from 4 to 12, reaching a value of 0.3 g g−1 at the latter ratio and B = 50 %w/w. Fig. 5c shows the interaction between A and C and their effect on YLu ; only C had a significant effect, YLu increasing from 0.22 glactulose glactose −1 at C = 4 to 0.31 g g−1 at C = 12.

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

From the response surfaces obtained (Fig. 5a–c), values of the variables that maximize YLu within the ranges considered were determined. Maximum YLu of 0.316 glactulose glactose −1 was predicted at 60 ◦ C, 56%w/w total initial concentration of sugars and fructose/lactose molar ratio of 12, with a confidence interval of 95%. Predicted value was confirmed experimentally by conducting the reaction of synthesis at the above conditions, YLu of 0.31 glactulose glactose −1 being obtained. At such conditions, RLu/GOS was 9.76 mol mol−1 and ␲Lu was 0075 g h−1 mg enz −1 . Yield of lactulose with respect to initial fructose was 0.048 glactulose gfructose −1 , which is low because of the excess fructose used; this will require to recover the excess fructose to recycle it to the reactor or else utilize a different reaction system.

ducting the synthesis reaction at the above conditions, RLu/GOS of 9.98 mol mol−1 being obtained. At such conditions, YLu was 0.28 glactulose glactose −1 and ␲Lu was 0.018 g h−1 mgenz −1 . Yield of lactulose with respect to the excess substrate fructose was 0.05 glactulose gfructose −1 . Even though maximum RLu/GOS can be attained within the range considered in the experimental design, it does not correspond to the maximum prebiotic effect. The above results show that by manipulating operational conditions, the Lu synthesis reaction can be driven to the desired RLu/GOS . Since the optimum RLu/GOS of 4 can be obtained at different combinations of A–C, it is necessary to define the region where this RLu/GOS is obtained at maximum YLu . From Eqs. (7) and (8), C is then expressed as a function of A and B for a RLu/GOS of 4, the following equation being obtained:

C = 13.6 + A × (1.23 + 0.91 × B) + 0.05 × B − 0.0045



0.5

107 + (1.7 × 106 − 4.2 × 104 × A) × A + 1.5 × 105 × B + A × (1.3 × 106 + 1.1 × 105 × A) × B + (1.6 × 105 + A × (4.4 × 103 + 4 × 104 × A)) × B2

YLu obtained at optimum conditions was significantly higher than previously reported for the synthesis of Lu with ␤galactosidases from various origins: 0.05 glactulose glactose −1 was reported by Lee et al. (2004) with ethanol permeabilized cells of Kluyveromyces lactis, 0125 glactulose glactose −1 , was reported by Kim et al. (2006) with ␤-galactosidase from S. solfataricus, and 0.28 glactulose glactose −1 was reported by Guerrero et al., (2011) with A. oryzae ␤-galactosidase. It was also significantly higher than values reported for the chemical synthesis of Lu: 0.21 glactulose glactose −1 reported by Paseephol et al. (2008) with calcium carbonate as catalyst and 0.20 glactulose glactose −1 reported by Villamiel et al. (2002) with sepiolite and alkaline ions as catalyst. However, it is similar to the value of 0.3 glactulose glactose −1 reported by Montgomery and Hudson (1930) with calcium hydroxide as catalyst, and significantly lower than the value of 0.87 glactulose glactose −1 (the highest to our knowledge) reported by Hicks and Parrish (1980) with boric acid and tri-ethanolamine as catalyst. Chemical synthesis has the disadvantage of using large amounts of environmentally offensive inorganic catalysts, producing significant product degradation and the formation of undesirable and hard to remove side-products making downstream processing cumbersome and costly (Aider and de Halleux, 2007). Consequently, biocatalysis is an attractive alternative approach. Following the same rationale as in the case of YLu , from the results for RLu/GOS , Eq. (8) was obtained describing the effect of the coded variables on it: R

Lu GOS

= 0.096 × A + 0.20 × B + 3.0 × C + 0.16 × A × B + 0.27 × A × C + 0.011 × B × C − 0.26 × A2 + 0.37 × B2 − 0.11 × C2 + 0.2 × A × B × C + 6.21

39

(8)

Coefficient of determination of Eq. (8) was 0.96 with adjusted coefficient of determination of 0.91. As in the case of YLu , the most significant variable was C. Fig. 5d–f shows the response surfaces obtained, showing the effect of the operational variables on RLu/GOS and their interactions. C had a strong influence on RLu/GOS (Fig. 5d–f), RLu/GOS increasing from 2.85 at C = 4 to 8.9 at C = 12. On the other hand, A and B had no significant effect on RLu/GOS (Fig. 5e), as seen from the coefficients in Eq. (8). From the surface of response obtained (Fig. 5) the values of variables that maximize RLu/GOS within the ranges considered were determined. Maximum RLu/GOS of 10.14 mol mol−1 was predicted at 60 ◦ C, 70%w/w total initial concentration of sugars and fructose/lactose molar ratio of 12, with a confidence interval of 95%. Predicted value was confirmed experimentally by con-

(9)

Eq. (9) represents the response surface accounting for the values of C required for each pair of values of A and B giving a RLu/GOS of 4, so maximizing the prebiotic potential of the reaction products (Section 3.2). Since a Lu/GOS molar ratio of 4 proved to be the best in terms of prebiotic effect (Figs. 1 and 2), experimental conditions were pursued for obtaining such a ratio (this analysis is valid for any desired Lu/GOS molar ratio). Fig. 6a shows the response surface for C described by Eq. (9). Replacing Eq. (9) in Eq. (7), a mathematical expression for YLu was obtained at different values of A and B, within the range of the experimental design, maintaining RLu/GOS in 4 (Fig. 6b). In this way, it is possible to determine the values of A and B that maximize YLu for RLu/GOS of 4. At RLu/GOS of 4, maximum YLu is 0.26 glactulose glactose −1 , which is 17.7% lower than the value obtained for maximum YLu at no matter what selectivity. Even so, the value of YLu obtained at optimum RLu/GOS is still higher than the values reported for the enzymatic synthesis of Lu up to now (Lee et al., 2004; Kim et al., 2006) and even higher than some of the values reported for the chemical synthesis (Passephol et al., 2008; Villamiel et al., 2002). 4. Conclusions Lu/GOS mixtures produced a slight increase in purportedly beneficial bacterial intestinal populations and higher reduction in the less health-positive populations than their separate use. Even though a molar ratio of 4 was considered the best in terms of prebiotic potential, no significant differences were observed for Lu/GOS molar ratios higher than 1. This is highly important since the production of GOS, which is inherent to the enzymatic synthesis of lactulose from lactose, can be reduced by operating at high fructose/lactose molar ratios and even almost completely arrested at values as high as 20; however, fructose concentrations in high stoichiometric excess are required, leaving a significant fraction of it unreacted. Therefore, there is space for determining an optimum Lu/GOS ratio from a process perspective, which might be somewhat below the value of 4 established here. The essential point is that this work proves that it is feasible to control the product composition in the enzymatic synthesis of lactulose by simply manipulating the fructose/lactose substrate ratio. Optimum composition will be established by considering both the quality of the product and process conditions of production. Acknowledgements Work financed by Fondecyt 1130059 and 3130339 from Conicyt, Chile. We acknowledge the generous donations of A. oryzae

40

C. Guerrero et al. / Journal of Biotechnology 209 (2015) 31–40

␤-galactosidase by Enzyme Development Corporation (New York, USA) and the donation of strains of Bifidobacterium lactis, Bifodobacterium longun, Lactobacillus rhamnosus and Lactobacillus platarum by Centro de Investigación y Desarrollo en Criotecnología de Alimentos (CIDCA), La Plata, Argentina. Valuable contribution of professor Robert Rastall and Dr. Sofia Kolida of the Department of Food and Nutritional Sciences, The University of Reading, UK, both in the development of the experimental work and the preparation of the manuscript, is acknowledged. References Aider, M., de Halleux, D., 2007. Isomerization of lactose and lactulose production: review. Trends Food Sci. Technol. 187, 356–364. Albayrak, N., Yang, S.T., 2002. Production of galacto-oligosaccharides from lactose by Aspergillus oryzae ␤-galactosidase immobilized on cotton cloth. Biotechnol. Bioeng. 77, 8–19. Beards, E., Tuohy, K., Gibson, G., 2010. Bacterial, SCFA and gas profiles of a range of food ingredients following in vitro fermentation by human colonic microbiota. Anaerobe 16, 420–425. Blottiere, H.M., Champ, M., Hoebler, C., Michel, C., Cherbut, C., 1999. Production and digestive effects of short chain fatty acids. Sci. Des Aliments 19, 269–290. Cardelle-Cobas, A., Corzo, N., Olano, A., Peláez, C., Requena, T., Ávila, M., 2011. Galactooligosaccharides derived from lactose and lactulose: influence of structure on Lactobacillus, Streptococcus and Bifidobacterium growth. Int. J. Food Microbiol. 149, 81–87. Depeint, F., Tzortzis, G., Vulevic, J., l´ıAnson, K., Gibson, G., 2008. Prebiotic evaluation of a novel galactooligosaccharide mixture produced by the enzymatic activity of Bifidobacterium bifidum NCIMB 41,171, in healthy humans: a randomized, double-blind, crossover, placebo-controlles intervention study. Am. J. Clin. Nutr. 87, 785–791. De Preter, V., Van Houtte, T., Huts, G., Swings, J., Rutgeerts, P., Ver Beke, K., 2006. Effect of lactulose and Saccharomyces boulardii administration on the colonic urea-nitrogen metabolism and the bifidobacteria concentration in healthy human subjects. Aliment. Pharmacol. Ther. 23, 963–974. Franks, A.H., Harmsen, H.J.M., Raangs, G.C., Jansen, G.J., Schut, F., Welling, W., 1998. Variations of bacterial populations in human feces measured by fluorescent in situ hybridization with group-specific 16S rRNA-targeted oligonucleotide probes. Appl. Environ. Microbiol. 64, 3336–3345. Ghoddusi, H.B., Grandison, M.A., Grandison, A.S., Tuohy, K.M., 2007. In vitro study on gas generation and prebiotic effects of some carbohydrates and their mixtures. Anaerobe 13, 193–199. Gibson, G.R., 2006. From probiotics to prebiotics and a healthy digestive system. J. Food Sci. 69, 141–143. Gopal, P.K., Sullivan, P.A., Smart, J.B., 2001. Utilisation of galacto-oligosaccharides as selective substrates for growth by lactic acid bacteria including Bifidobacterium lactis DR10 and Lactobacillus rhamnosus DR 20. Int. Dairy J. 11, 19–25. Guerrero, C., Vera, C., Plou, F., Illanes, A., 2011. Influence of reaction conditions on the selectivity of the synthesis of lactulose with microbial ␤-galactosidases. J. Mol. Catal. B Enzym. 72, 206–212. Guerrero, C., Vera, C., Illanes, A., 2013. Optimization of synthesis of oligosaccharides derived from lactulose (fructosyl-galacto-oligosaccharides) with ␤-galactosidases of different origin. Food Chem. 138, 2225–2232. Hague, A., Elder, D.J.R., Hicks, D.J., Paraskeva, C., 1995. Apoptosis in colorectal tumour cells: induction by the short chain fatty acids butyrate, propionate and acetate and by the bile salt deoxycholate. Int. J. Cancer 60, 400–406. Harmsen, H.J., Elferrich, P., Schut, F., Welling, G.W., 1999. A 16S rRNA.targeted probe for detection of lactobacilli and enterococci in faecal samples by fluorescent in situ hybridization. Microb. Ecol. Health Dis. 11, 3–12. Harmsen, H.J.M., Wildeboer-veloo, A.C.M., Grijpstra, J., Knol, J., Degener, J.N., Welling, G.W., 2000. Development of 16S rRNA-based probes for the Coriobacterium group and the Atopobium cluster and their application for enumeration of Coriobacteriaceae in human feces from volunteers of different age groups. Appl. Environ. Microbiol. 66, 4523–4527. Hicks, K.B., Parrish, F.W., 1980. A new method for the preparation of lactulose from lactose. Carbohydr. Res. 82, 393–397. Hold, G.L., Schwiertz, A., Aminov, R.I., Blaut, M., Flint, H.J., 2003. Oligonucleotide probes that detect quantitatively significant groups of butyrate-producing bacteria in human feces. Appl. Environ. Microbiol. 69, 4320–4324. Huebner, J., Wehling, R.L., Hutkins, R.W., 2007. Fuctional activity of commercial prebiotics. Int. Dairy J. 17, 770–775. Kim, Y.S., Park, C.S., Oh, D.K., 2006. Lactulose production from lactose and fructose by a thermostable ␤-galactosidase from Sulfolobus solfataricus. Enzyme Microb. Technol. 39, 903–908. Lamsal, B.P., 2012. Production, health aspects and potential food uses of dairy prebiotic galactooligosaccharides. J. Sci. Food Agric., http://dx.doi.org/10.1002/ jsfa.5712

Langendijk, P.S., Schut, F., Jansen, G.J., Raangs, G.C., Kamphuis, G.R., Wilkinson, M.H.F., Welling, G.W., 1995. Quantitative fluorescence in situ hybridization of Bifidobacterium spp. With Genus-specific 16S eRNA-Targeted probes and its application in fecal samples. Appl. Environ. Microbiol. 61, 3069–3075. Lee, Y.J., Kim, C.S., Oh, D.K., 2004. Lactulose production by ␤-galactosidase impermeabilized cells of Kluyveromyces lactis. Appl. Environ. Microbiol. 64, 787–793. Lupton, J.R., 2004. Microbial degradation products influence colon cancer risk: the butyrate controversy. J. Nutr. 134, 479–482. Manz, W., Amann, R., Ludwig, W., Vancanneyt, M., Schleifer, K.H., 1996. Application of a suite of suite of 16S rRNA-specific oligonucleotide probes designed to investigate bacteria of the phylum cytophaga-flavobacter-bacteroides in the natural environment. Microbiology 142, 1097–1106. Montgomery, E.M., Hudson, C.S., 1930. Relations between rotatory power and structure in the sugar group. XXVII. Synthesis of a new disaccharide ketose (lactulose) from lactose. J. Am. Chem. Soc. 52, 2101–2106. Pan, X.D., Chen, F.Q., Wu, T.X., Tang, H.G., Zhao, Z.Y., 2009. Prebiotic oligosaccharides change the concentrations of short-chain fatty acids and microbial population of mouse bowel. J. Zhenjiang Univ. Sci. B 10, 258–263. Palframan, R., Gibson, G.R., Rastall, R.A., 2002. Effect of pH and dose on the growth of gut bacteria on prebiotic carbohydrates in vitro. Anaerobe 8, 287–292. Paseephol, T., Darryl, M., Small, D.M., Sherkat, F., 2008. Lactulose production from milk concentration permeate using calcium carbonate-based catalysts. Food Chem. 111, 283–290. Pryde, S.E., Duncan, S.H., Hold, G.L., Stewart, C.S., 2002. The microbiology of butyrate formation in the human colon. FEMS Microbiol. Lett. 217, 133–139. Rada, V., Nevoral, J., Trojanová, I., Tománková, E., Smehilová, M., Killer, J., 2008. Growth of infant faecal bifidobacteria and clostridia on prebiotic oligosaccharides in in vitro conditions. Anaerobe 14, 205–208. Rastall, R.A., 2010. Functional oligosaccharides: application and manufacture. Annu. Rev. Food Sci. Technol., 305–339. Rodriguez-Colinas, B., Kolida, S., Barran, M., Ballesteros, A.O., Rastall, R.A., Plou, F.J., 2013. Analysis of fermentation selectivity of purified galacto-oligosaccharides by in vitro human faecal fermentation. Biotechnol. Prod. Process Eng. 97, 5743–5752. Ruiz-Matute, A.I., Brokl, M., Sanz, M.L., Soria, A.C., Cote, G.L., Collins, M.E., Rastall, R.A., 2011. Effect of dextransucrase cellobiose acceptor products on the growth of human gut bacteria. J. Agric. Food Chem. 59, 3693–3700. Rycroft, C.E., Jones, M.R., Gibson, G.R., Rastall, R.A., 2001. A comparative in vitro evaluation of the fermentation properties of prebiotic oligosaccharides. J. Appl. Microbiol. 91, 878–887. Sahota, S.S., Bramley, P.M., Menzies, I.S., 1982. The fermentation of lactulose by colonic bacteria. J. Gen. Microbiol. 128, 319–325. Salazar, N., Ruas-Madiedo, P., Kolida, S., Collins, M., Rastall, R., Gibson, G., de los Reyes-Gavilán, C.G., 2009. Exopolysaccharides produced by Bifidobacterium longun IPLA E44 and Bifidobacterium animalis subsp. Lactis IPLA R1 modify the composition and metabolic activity of human faecal microbiota in pH-controlled batch cultures. Int. J. Food Microbiol. 135, 260–267. Sanz, M.L., Gibson, G.R., Rastall, R.A., 2005. Influence of disaccharide structure on prebiotic selectivity in vitro. J. Agric. Food Chem. 53, 5192–5199. Sarbini, S.R., Kolida, S., Naeye, T., Einerhand, A., Brison, Y., Remaud-Simeon, M., Monsan, P., Gibson, G.R., Rastall, R.A., 2011. In vitro fermentation of linear and ␣-1,2-branched dextrans by the human fecal microbiota. Appl. Environ. Microbiol. 77, 5307–5315. Simpson, E.J., Chapman, M.A.S., Dawson, J., Berry, D., Macdonald, I.A., Cole, A., 2000. In vivo measurement of colonic butyrate metabolism in patients with quiescent ulcerative colitis. Gut 46, 73–77. Vasiljevic, T., Shah, N.P., 2008. Probiotics-from Metchnikoff to bioactives. Int. Dairy J. 18, 714–728. Vernazza, C.L., Gibson, G.R., Rastall, R.A., 2006. Carbohydrate preference, acid tolerance and bile tolerance in five strains of Bifidobacterium. J. Appl. Microbiol. 100, 846–853. Vera, C., Guerrero, C., Conejeros, R., Illanes, A., 2012. Synthesis of galacto-oligosaccharides by ␤-galactosidase from Aspergillus oryzae using partially dissolved and supersaturated solution of lactose. Enzyme Microb. Technol. 50, 188–194. Villamiel, M., Corzo, N., Foda, M.I., Montes, F., Olano, A., 2002. Lactulose formation catalysed by alkaline-subsituted sepiolites in milk permeate. Food Chem. 76, 7–11. Vos, A.P., Haarman, M., Buco, A., Govers, M., Knol, J., Garssen, J., Stahl, B., Boehm, G., M´ıRabet, L., 2006. A specific prebiotic oligosaccharide mixture stimulates delayed-type hypersensitivity in murine influenza vaccination model. Int. Immunopharmacol. 6, 1277–1286. Vulevic, J., Rastall, R., Gibson, G.R., 2004. Developing a quantitative approach for determining the in vitro prebiotic potential of dietary oligosaccharides. FEMS Microbiol. Lett. 236, 153–159. Walker, A.W., Duncan, S.H., McWilliam, C., Child, M.W., Flint, H.J., 2005. pH and peptide supply can radically alter bacterial populations and short-chain fatty acid ratios within microbial communities from the human colon. Appl. Environ. Microbiol. 71, 3692–3700.