The lactose operon from Lactobacillus casei is involved in the transport and metabolism of the human milk oligosaccharide core-2 N-acetyllactosamine

The lactose operon (lacTEGF) from Lactobacillus casei strain BL23 has been previously studied. The lacT gene codes for a transcriptional antiterminator, lacE and lacF for the lactose-specific phosphoenolpyruvate: phosphotransferase system (PTSLac) EIICB and EIIA domains, respectively, and lacG for the phospho-β-galactosidase. In this work, we have shown that L. casei is able to metabolize N-acetyllactosamine (LacNAc), a disaccharide present at human milk and intestinal mucosa. The mutant strains BL153 (lacE) and BL155 (lacF) were defective in LacNAc utilization, indicating that the EIICB and EIIA of the PTSLac are involved in the uptake of LacNAc in addition to lactose. Inactivation of lacG abolishes the growth of L. casei in both disaccharides and analysis of LacG activity showed a high selectivity toward phosphorylated compounds, suggesting that LacG is necessary for the hydrolysis of the intracellular phosphorylated lactose and LacNAc. L. casei (lacAB) strain deficient in galactose-6P isomerase showed a growth rate in lactose (0.0293 ± 0.0014 h−1) and in LacNAc (0.0307 ± 0.0009 h−1) significantly lower than the wild-type (0.1010 ± 0.0006 h−1 and 0.0522 ± 0.0005 h−1, respectively), indicating that their galactose moiety is catabolized through the tagatose-6P pathway. Transcriptional analysis showed induction levels of the lac genes ranged from 130 to 320–fold in LacNAc and from 100 to 200–fold in lactose, compared to cells growing in glucose.

SCIenTIFIC REPORTS | (2018) 8:7152 | DOI: 10.1038/s41598-018-25660-w is transported in L. casei BL23 by the PTS Gnb or by the PTS Lac , which are involved in LNB and lactose uptake, respectively 24,28 . The mutant strains BL385 (gnbC) 24 , that is disrupted in the gene encoding the EIIC domain of the PTS Gnb , BL153 (lacE) and BL155 (lacF) 28 , which are impaired in the EIICB and EIIA, respectively, of the PTS Lac , were tested for their capacity to ferment LacNAc. BL385 (gnbC) was able to grow in the presence of LacNAc as carbon source (Fig. 3a). Contrarily, strains BL153 (lacE) and BL155 (lacF) showed a poor growth with LacNAc which was similar to that of the negative controls (lactose supplemented and non-supplemented sugar-free MRS) (Fig. 3b,c). The growth pattern of strains BL153 (lacE) and BL155 (lacF) in the presence of glucose as a positive control is also shown (Fig. 3b,c). Analysis for sugar content in the supernatants demonstrated that LacNAc was consumed by strain BL385 (gnbC) but not by strains BL153 (lacE) and BL155 (lacF) (data not shown), indicating that the domains EIICB and EIIA encoded by lacE and lacF, respectively, are involved in the uptake of LacNAc (Fig. 1). It has been previously shown that the PTS from L. casei can accomplish sugar transport not coupled to phosphorylation 23 . Then, to further confirm the involvement of the PTS Lac in the transport of LacNAc and to test if transport though the EII permease was coupled to phosphorylation, the growth pattern of BL126 (ptsI), a mutant lacking the PTS-general component Enzyme I 33 , was tested in LacNAc as carbon source. BL126 (ptsI) did not grow in the presence of LacNAc (Fig. 3d), confirming that its utilization needs a functional complete PTS. These results suggest that LacNAc is internalized as a phosphorylated derivative. BL126 (ptsI) was grown with glucose and lactose as positive and negative controls, respectively. The growth pattern of strain BL126 in the presence of glucose differs from that of strains BL153 (lacE) and BL155 (lacF) (Fig. 3). This might be due to the fact that the latest strains contain a functional PTS for glucose uptake while strain BL126 (ptsI) can only transport this sugar by the proton-driven permease 34 . LacG is involved in the metabolism of LacNAc and lactose. To determine if the phospho-β-galactosidase LacG was involved in the utilization of LacNAc in L. casei BL23, a mutant in lacG was constructed (strain BL400). This mutant was cultured in sugar-free MRS supplemented with 4 mM LacNAc as carbon source (Fig. 4). The growth pattern showed that BL400 (lacG) strain did not ferment LacNAc and neither did lactose, the only substrate described until now for the lac operon [26][27][28] . These results indicated that LacG is necessary for the utilization of both disaccharides (Fig. 1b). Sugar content analysis of the culture supernatants detected LacNAc and lactose, respectively, in the supernatants from BL400 (lacG), while they were completely consumed by the wild-type BL23 strain (data not shown).
LacG (EC 3.2.1.85) belongs to the glycosyl hydrolase family 1 (GH 1; www.cazy.org), which includes β-glycosidases as well as phospho-β-glycosidases. In order to characterize that enzyme, the lacG gene was expressed in E. coli as a His-tagged protein and purified to homogeneity (data not shown). The purified protein displayed a molecular weight of 55 kDa, in agreement with the calculated mass of the 6x(His)-tagged protein (55,305 Da). 6x(His)LacG did not hydrolyze o-NP-β-D-galactopyranoside but it does when this substrate is phosphorylated (Table 1), suggesting a high selectivity toward phosphorylated compounds. The kinetic analysis showed a high Km and low Vmax for o-NP-β-D-galactopyranoside-6P, and it displayed an optimal pH of 7.0 and an optimal temperature of 41 °C (Table 1). 6x(His)LacG was unable to hydrolyze any of the natural oligosaccharides tested (Table 1), including lactose and LacNAc, possibly because these need to be phosphorylated before turned into substrates for this glycosidase.
The tagatose-6P pathway and the N-acetylglucosamine-6P deacetylase NagA are involved in the metabolism of LacNAc. The results described above suggest that, as occurs with lactose, LacNAc is transported and phosphorylated by the PTS Lac , and then hydrolyzed by the phospho-β-galactosidase LacG into GlcNAc and Gal-6P. It has been assumed that this phosphorylated sugar resulting from the lactose metabolism is catabolized through the Tag-6P pathway in L. casei 35,36 . To analyze this at the genetic level, the mutant strain L. casei BL393 (lacAB), deficient in the heteromeric Gal-6P isomerase of the Tag-6P route 24 , was cultured on lactose or LacNAc (Fig. 5). This strain showed a growth rate in lactose (0.0293 ± 0.0014 h −1 ) and LacNAc (0.0307 ± 0.0009 h −1 ) significantly lower than the wild-type on these disaccharides (0.1010 ± 0.0006 h −1 and 0.0522 ± 0.0005 h −1 , respectively) (wild-type versus lacAB mutant, P = 0.0004 (lactose); P = 0.0012 (LacNAc)). These results supported that lactose and LacNAc metabolism in L. casei utilizes the Tag-6P pathway for the catabolism of the galactose moiety (Fig. 1b). Additionally,  these results suggest that the residual growth showed by BL393 (lacAB) strain on those carbohydrates would be maintain by the catabolism of the glucose and GlcNAc moieties, resulting from the lactose and LacNAc hydrolysis, respectively.
Optimal temperature (°C) 41 Table 1. Activity and characterization of the phospho-β-galactosidase LacG. a Carbohydrates used as substrates. NP, nitrophenyl; Glc, glucose; Gal, Galactose; GlcNAc, N-acetylglucosamine; GalNAc, N-acetylgalactosamine; Man, mannose; Fru, fructose. b +, substrate is totally hydrolyzed after 16 h reaction in the conditions described in the "Materials and methods" section; −, no activity detected. c The enzyme activity was determined with onitrophenyl-β-D-galactopyranoside-6-P as the substrate. We have previously shown that in L. casei the nagA gene, encoding an N-acetylglucosamine-6P deacetylase, is involved in the metabolism of GlcNAc either free o derived from LNB 24 . Here, we tested the growth of the mutant strain L. casei BL388 disrupted in nagA on LacNAc as carbon source (Fig. 5). The results showed a growth rate for this mutant (0.0436 ± 0.0008 h −1 ) significantly lower (P = 0.004) than the wild-type (0.0522 ± 0.0005 h −1 ), suggesting that nagA is required for the catabolism of the GlcNAc moiety resulting from the hydrolysis of this disaccharide.
Transcriptional analyses of the lac operon. Northern blot analyses have previously shown that the lacTEGF operon from L. casei is induced by lactose and subjected to carbon catabolite repression by glucose 28 . In order to determine if the transcription of the lac genes are also regulated by LacNAc, RNA was isolated from L. casei wild-type strain BL23 grown in sugar-free MRS containing galactose, GlcNAc, glucose, LacNAc or LacNAc plus glucose and used for RT-qPCR analyses. Additionally, RNA obtained from cultures grown on lactose or lactose plus glucose were also included in these analyses to quantify the expression levels of the lac genes in these carbon sources (Fig. 6). Taking as a reference the transcript levels in cells growing in glucose, the lacT, lacE, lacG and lacF were induced by LacNAc and lactose. The induction levels ranged from 130 to 320-fold and from 100 to 200-fold in LacNAc and lactose, respectively. The expression levels were highly reduced when these carbohydrates were mixed with glucose (Fig. 6). These results indicate that the lac operon in addition to lactose is also induced by LacNAc and confirmed that it is repressed by glucose. The lac genes were barely expressed in the presence of galactose and GlcNAc, denoting that their induction relies on the presence of the disaccharide and not on the monosaccharides resulting from the hydrolysis.
The transcriptional antiterminator LacT is required for LacNAc metabolism. The L. casei LacT protein prevents transcription termination of the lac operon in response to the presence of lactose in the culture medium 26 . In order to determine the involvement of this transcriptional antiterminator in the metabolism of LacNAc, the mutant strain L. casei BL195 (lacT), deficient in LacT 26 , was cultured on LacNAc as carbon source (Fig. 7). BL195 (lacT) strain was also grown with glucose and lactose as positive and negative controls, respectively. The results show that this mutant strain exhibited a growth in the presence of LacNAc similar to that of the negative controls (lactose supplemented and non-supplemented sugar-free MRS), indicating that the transcriptional antitermiator LacT is also involved on LacNAc metabolism. Additionally, the results suggest that LacT antiterminates transcription of lac operon not only depending on the presence of lactose if not also on the presence of LacNAc in the growth medium.

Discussion
The disaccharide N-acetyllactosamine (LacNac) has an important role in many cell recognition processes such as parasite-host cell interaction 37 , autoimmune 38 and inflammatory 39 diseases, and also in cancer 40 . Additionally, LacNAc is a key structure present at human milk oligosaccharides and also at the glycan domains of glycoproteins and glycolipids present in the gastrointestinal tract [41][42][43] . We have demonstrated that L. casei is able to metabolize LacNAc. Curiously, the lac operon, which has been widely studied in this strain [26][27][28] , is also the responsible of the transport and catabolism of LacNAc. As previously described for the PTS Gnb , that is involved in the transport of N-acetyl-galactosamine, LNB and GNB in L. casei 24 , the PTS Lac represents a new example of a PTS able to transport two structurally related substrates, lactose and LacNAc. L. casei mutants deficient in either EIICB Lac or EIIA Lac 28 , were unable to grow in the presence of LacNAc. Due to the great biotechnological and economic  [44][45][46][47] . Lactose in these bacteria can be transported through proton symport permeases, lactose/galactose antiport systems 44,48 or via PTS 28,[49][50][51] . Analysis of the genome sequence of lactobacilli (http://www.ncbi.nlm.nih.gov/genomes) showed that genes encoding PTSs homologous to the PTS Lac from L. casei BL23, are present in the Lactobacillus casei/paracasei/rhamnosus/zeae group of phylogenetically related lactobacilli, and also in a few strains of Lactobacillus heilongjiangensis, Lactobacillus futsaii, Lactobacillus farciminis, Lactobacillus perolens, Lactobacillus fermentum, Lactobacillus sharpeae, Lactobacillus crustorum, Lactobacillus pobuzihii, Lactobacillus kimchiensis, Lactobacillus ruminis, Lactobacillus gasseri and Lactobacillus johnsonii, suggesting that lactose-specific PTSs are widely distributed among lactobacilli. For L. gasseri strain ATCC 33323, which is autochthonous of the gut, it has already been demonstrated that it contains two different PTSs involved in the transport of lactose and that the expression of both was induced by this carbohydrate 51 . Another strain indigenous of the gut is L. rhamnosous TCELL-1, for which it has also been demonstrated that the lac operon is induced by lactose 50 . Whether those PTS are functional for lactose and LacNAc remains to be investigated. Transcriptional analysis demonstrated here that the lac operon from L. casei is induced by lactose as well as by LacNAc and that induction levels are higher on LacNAc than on lactose. This might suggest that LacNAc is the substrate to which the lac operon had been adapted first. Indeed, many L. casei strains have been isolated from the human gastrointestinal tract 52,53 , which is very rich in LacNAc 5,41 . In E. coli, the lac operon is not regulated directly by lactose if not by allolactose, a transient product synthesized by the β-galactosidase, and this has generated controversy about the true physiological role of the lac operon in this bacterium 54 .
L. casei transport LacNAc or lactose through the PTS Lac resulting in the formation of LacNAc-P or lactose-P, which are further hydrolyzed inside the cell by the phospho-β-galactosidase LacG into Gal-6P and GlcNAc or glucose, respectively (Fig. 1b). We have biochemically characterized this enzyme and confirmed that can only hydrolyze phosphorylated substrates. This also occurs with all the proteins homologs to LacG that have been characterized until now, which have been isolated from Streptococcus mutants 55 , Saphylococcus aureus, Lactococcus lactis, L. casei strain 64H 56 and L. gasseri 57 . Contrarily, β-galactosidases isolated from Bifidobacterium bifidum are able to hydrolyze non-phosphorylated lactose and LacNAc 11,12 , showing two different mechanisms to metabolize these disaccharides in species that would compete for them in environmental niches such as the gastrointestinal tract. The Gal-6P generated after the hydrolysis of LacNAc-P or lactose-P by LacG is channeled through the Tag-6P pathway. We have showed that the mutant strain L. casei BL393 (lacAB) is impaired in the growth on lactose and LacNAc. The genes encoding the Tag-6P route in L. casei are present in the operon lacR1ABD2C, which includes a transcriptional regulator LacR1, the two subunits of the heteromeric Gal-6P isomerase (lacAB), a Tag-6P kinase (lacC) and a Tag-1,6-bisP aldolase (lacD2). Unlike L. casei, in Lactococcus lactis 58 and Streptococcus mutans 59 , the lac operon contains the genes lacRABCDFEG encoding the Tag-6P catabolic proteins in addition to the lactose-specific PTS and the phospho-β-galactosidase LacG.
L. casei species have been isolated from dairy products, plant material and reproductive and gastrointestinal tracts of humans and animals 52,53 , which reveals their great adaptability to different environments. Genome analyses have showed that gene loss and acquisition are the main events resulting in niche adaptation 60 . Additionally, lactobacilli also contain genes involved in sugar uptake, metabolism and regulation grouped in genomic islands 61 . Lactose metabolism is well known in the dairy industry, but few data is found about its metabolism by the gastrointestinal microbiota. Lactose and LacNAc are constituents of HMO molecules that reach the breastfeeding infant gut microbiota, and LacNAc and poly-LacNAc molecules are also present in high amounts in the newborn gut 62,63 . Here we showed that L. casei metabolizes both disaccharides by using the same transport system and catabolic enzymes, which could be another niche adaptation mechanism of this bacterium to optimize the metabolic machinery minimizing energy consumption in a very competitive environment such as the gastrointestinal tract. Furthermore, the present work evidences that a catabolic pathway designed for survival of lactobacilli in the children gut has become an important tool for the development of dairy fermented products.  Table 2. The L. casei strains were grown at 37 °C in MRS broth (Difco). E. coli was utilized as a cloning host and was grown in Luria-Bertani medium (Pronadisa) at 37 °C. E. coli DH10B transformants were selected with ampicillin (100 μg ml −1 ), E. coli BE50 with ampicillin (100 μg ml −1 ) and kanamycin (25 μg ml −1 ) and L. casei with erythromycin (5 μg ml −1 ). The vectors pRV300 64 and pQE80 (Qiagen) were used for disruption of genes in L. casei and overproduction of proteins, respectively. E. coli strains were transformed by electroporation with a Gene Pulser apparatus (Bio-Rad Laboratories) as indicated by the manufacturer, and L. casei strains were transformed as described previously 65 .
Culture of L. casei strains with lactose and LacNAc. The L. casei strains were cultured as previously described 24 on sugar-free MRS containing: bactopeptone (Difco), 10 g l −1 ; yeast extract (Pronadisa), 4 g l −1 ; sodium acetate, 5 g l −1 ; tri-ammonium citrate, 2 g l −1 ; magnesium sulphate 7-hydrate, 0.2 g l −1 ; manganese sulphate monohydrate, 0.05 g l −1 ; and Tween 80, 1 ml l −1 . LacNAc (Carbosynth, Compton, Berkshire, UK), lactose (Sigma-Aldrich, St. Louis, MO, USA), N-acetylglucosamine, galactose or glucose were added to the sugar-free MRS medium at a concentration of 4 mM. Bacterial growth was assayed in microtiter plates (100 μl culture broth per well) at 37 °C in a POLARstar Omega plate reader (BMG Labtech, Offenburg, Germany). The Gompertz model (GraphPad Software, San Diego, CA) was used for the analysis of the growth rates (μ). DNA manipulation and sequencing. DNA was obtained from L. casei BL23 as previously described 65 . Plasmid DNA was isolated from E. coli by using the kit illustra plasmidPrep Mini Spin (GE Healthcare, UK). Standard methods were used for recombinant DNA techniques 66 and PCR reactions were carried out with the Expand High Fidelity PCR System (Roche). The DNA sequencing reactions were performed by the Central Service of Research Support of the University of Valencia (Spain). Specific primers hybridizing within the proper DNA fragments and universal primers were used for sequencing. The analysis of DNA sequences was performed with the aid of the DNAMAN 4.03 software package (Lynnon BioSoft) and sequence similarities were analyzed with the BLAST program 67 .

Construction of a lacG mutant strain.
A DNA fragment containing part of lacG was obtained by PCR using L. casei BL23 chromosomal DNA and the oligonucleotides lacGfor (5′-CAAGGAAGACGGTAAAGG) and lacGrev (5′-CCAACGGATAGTCATTATG). The PCR product was cloned into pRV300 digested with EcoRV. The resulting plasmid pRVlacG was cleaved at the unique SphI restriction site, made blunt with the Klenow fragment, ligated and transformed to select a plasmid with a frameshift at the SphI site in lacG. L. casei was transformed with this plasmid and one integrant where single recombination occurred was selected and cultured at 37 °C without antibiotic selection for at least 200 generations. Cells were grown on MRS-agar plates and screened for erythromycin-sensitive phenotype by replica plating on MRS-agar plates with erythromycin. Antibiotic-sensitive clones were selected and, among them, one was chosen (BL400 strain) in which a double-crossover event conducted to the excision of the plasmid resulting in a mutated lacG copy, as was confirmed by sequencing of PCR-amplified fragment spanning the mutated region.  Disaccharide and monosaccharide analysis in culture supernatants. The cells from the L. casei cultures were eliminated by centrifugation and the supernatants were collected for sugar determination. The analyses were carried out by high-performance liquid chromatography (HPLC) with a Jasco PU2080Plus system. A Rezex RCM-Monosaccharide column (Phenomenex) (at 80 °C) was used and the samples elution was performed in isocratic mode using a flow rate of 0.6 ml min −1 . The mobile phase was water and a refractive index detector (Jasco RI-2031 Plus) was used for carbohydrate detection. Peaks in the chromatograms were identified by comparing the retention times with those of the standards (LacNAc, lactose, glucose, galactose and GlcNAc).
Expression and purification of His-tagged LacG. lacG coding region was amplified by PCR using genomic DNA of wild-type L. casei strain BL23 as template and the primers LacGBamHIFW (5′-TTTTGGATCCATGAGT AAACAGCTACCTCAAG) and LacGPstIRV (5′-TTTTCTGCAGTTAATCCGGAATGATGTGGG) with added BamHI and PstI restriction sites to the 5′-and 3′-ends (underlined). The obtained PCR product was digested with BamHI and PstI, cloned into pQE80 and the resulting plasmid pQElacG was used to transform E. coli BE50. DNA sequencing was carried out to confirm the correct construction. One clone, PE172 (pQElacG) was selected, grown in Luria-Bertani medium and induced with IPTG (1 mM) as described before 25 . The recombinant protein was purified and analyzed as described previously 24 .
His-tagged LacG enzyme activity. 100 μl reaction mixtures containing different o/p-nitrophenyl(N-P)-sugars (Table 1) at 5 mM were performed in 96-well microtiter plates. The LacG activity was measured at 37 °C in 100 mM Tris-HCl buffer, pH 7.0, and the reaction started by adding 1 μg of enzyme. The amount of released o/p-nitrophenol was tested by tracking the absorbance change with time at 404 nm using a microplate reader (POLARstar Omega, BMG Labtech, Offenburg, Germany). The optimal pH and temperature reaction, and kinetic analysis were determined with o-NP-β-D-galactopyranoside-6-phosphate as previously described 24 .

RNA isolation and Reverse Transcriptase quantitative PCR (RT-qPCR).
Total RNA was isolated from L. casei strain BL23 grown in sugar-free MRS supplemented with 4 mM of different sugars as previously described 24 . Cells were harvested at mid-exponential phase of growth (OD 550 of 0.3 for cultures on glucose, galactose or GlcNAc and 0.5 for cultures on LacNAc, lactose, a mix of LacNAc plus glucose or lactose plus glucose). The isolated RNA was digested with DNaseI and retro-transcribed using the Maxima First strand cDNA Synthesis Kit (Fermentas) 24 . The resulted cDNA was subjected to quantitative PCR for the genes lacT, lacE, lacG and lacF. RT-qPCR was performed using the Lightcycler 2.0 system (Roche), LC Fast Start DNA Master SYBR green I (Roche) and primers pairs that produce amplicons ranging from 70 to 200 bp in size. RT-qPCR was performed for each cDNA sample in triplicate and using the primers pairs: qPCRlacTfor (5′-TTGTAAGGGGACGTGGCATC)/qPCRlacTrev (5′-TTGTCGGGAAGTCTCGTTCG) (lacT), qPCRlacEfor (5′-TTGGCCATGAACACGATGGA)/qPCRlacErev (5′-CCGAAAGTGCATGGCACAAA) (lacE), qPCR-lacGfor (5′-AAGTCGAAGGAGCCACCAAG)/qPCRlacGrev (5′-GAACCGCCCCTGTTTATCCA) (lacG) and qPCRlacFfor (5′-GGTTTTGCACTTGTGGCGTA)/qPCRlacFrev (5′-CTGTTGGGCCTTCTCAACCA) (lacF). The reaction mixtures and cycling conditions were performed as previously described 24 . The pyrG, lepA and IleS genes were chosen as reference genes 68 . Cells growing in sugar-free MRS supplemented with glucose were used as reference condition. The relative expression based on the expression ratio between the target genes and reference genes was calculated using the software tool REST (relative expression software tool) 69 . The efficiency of all the primer pairs was between 1.9 and 2 (close to 100%). RT-qPCR reactions were performed in triplicate of two biological independent samples. Statistical analysis. Statistical analysis was performed using the Statgraphics Plus, ver. 2.1 (Statistical Graphics Corp., USA). One way analysis of variance (ANOVA) was used to assess the effects of the carbon source (galactose, GlcNAc, LacNAc, lactose, a mix of glucose and lactose, and a mix of glucose and LacNAc) on the expression levels of the lac genes. Student's t-test was used to detect statistically significant differences between growth rate from L. casei wild-type BL23 strain versus each mutant BL393 (lacAB) and BL388 (nagA) strains. Statistical significance was accepted at P < 0.05. Data availability statement. All data generated or analyzed during this study are included in this published article.