GlaR (YugA)—a novel RpiR‐family transcription activator of the Leloir pathway of galactose utilization in Lactococcus lactis IL1403

Abstract Bacteria can utilize diverse sugars as carbon and energy source, but the regulatory mechanisms directing the choice of the preferred substrate are often poorly understood. Here, we analyzed the role of the YugA protein (now designated GlaR—Galactose–lactose operon Regulatory protein) of the RpiR family as a transcriptional activator of galactose (gal genes) and lactose (lac genes) utilization genes in Lactococcus lactis IL1403. In this bacterium, gal genes forming the Leloir operon are combined with lac genes in a single so‐called gal–lac operon. The first gene of this operon is the lacS gene encoding galactose permease. The glaR gene encoding GlaR lies directly upstream of the gal–lac gene cluster and is transcribed in the same direction. This genetic layout and the presence of glaR homologues in the closest neighborhood to the Leloir or gal–lac operons are highly conserved only among Lactococcus species. Deletion of glaR disabled galactose utilization and abrogated or decreased expression of the gal–lac genes. The GlaR‐dependent regulation of the gal–lac operon depends on its specific binding to a DNA region upstream of the lacS gene activating lacS expression and increasing the expression of the operon genes localized downstream. Notably, expression of lacS‐downstream genes, namely galMKTE, thgA and lacZ, is partially independent of the GlaR‐driven activation likely due to the presence of additional promoters. The glaR transcription itself is not subject to catabolite control protein A (CcpA) carbon catabolite repression (CRR) and is induced by galactose. Up to date, no similar mechanism has been reported in other lactic acid bacteria species. These results reveal a novel regulatory protein and shed new light on the regulation of carbohydrate catabolism in L. lactis IL1403, and by similarity, probably also in other lactococci.


| INTRODUC TI ON
Lactose, a disaccharide comprised of galactose linked through a βglycosidic bond to the C 4 of glucose, is the dominant sugar found in milk. Lactic acid bacteria (LAB) are capable of growth in milk owing to an efficient use of lactose as a carbon source. Because of the high efficiency and economic relevance of lactose fermentation, numerous studies have focused on LAB. Lactose utilization genes have been characterized in many LAB species, and it has been shown that they can take up lactose by two principally different ways including the lactose-specific phosphotransferase system (lac-PTS) and secondary transporters such as lactose-galactose antiporters and lactose-H + symporters (reviewed by Aleksandrzak-Piekarczyk, 2013). The secondary transport systems transfer unphosphorylated lactose via specific permeases of the LacS subfamily (TC No. 2.A.2.2.3) belonging to the 2.A.2 glycoside-pentoside-hexuronide (GPH) family (Saier, 2000). After its import, lactose is hydrolyzed by β-galactosidase to glucose and galactose. Then, glucose is further metabolized via glycolysis, while the galactose moiety can be either released into the medium or converted into glucose-1-phosphate (Glc-1-P), which enters glycolysis following conversion to Glc-6-P. The conversion into Glc-1-P is performed by the action of four enzymes that constitute the Leloir pathway (De Vos, 1996;Poolman, 1993;Vaughan, van den Bogaard, Catzeddu, Kuipers, & de Vos, 2001). This pathway, discovered by L. F. Leloir andcoworkers in 1950s (reviewed in Frey 1996), consists of the crucial enzyme galactokinase (GalK) plus hexose-1-P uridylyltransferase (GalT) and UDPglucose 4-epimerase (GalE) that perform the conversion of galactose into glucose-1-P. Found more recently, an additional enzyme, the GalM mutarotase (aldose-1-epimerase), is involved in the interconversion of the galactose α-and β-anomers (Bouffard, Rudd, & Adhya, 1994).
Directly upstream of these genes encoding enzymes catalyzing lactose hydrolysis and/or galactose conversion, or within this operon, a gene encoding specific permease for lactose or galactose uptake may also be present Vaillancourt et al., 2002).
The uptake and metabolism of sugars is mastered by numerous regulatory proteins which form a regulatory network detecting environments and setting the catabolic abilities of the cell, thus helping to maintain energy efficiency. Based on their specificity, two groups of regulators are distinguished, general and secondary ones (Guédon, Jamet, & Renault, 2002;Mayo et al., 2010). In most low-GC gram-positive bacteria, the main general regulator is catabolite control protein A (CcpA) (Hueck & Hillen, 1995), which acts by binding to 14-nucleotide DNA target sites known as cre (catabolite responsive elements), conducting carbon catabolite activation (CCA) or repression (CCR) (Weickert & Chambliss, 1990).
The cre sites are found in promoter regions of the CCR-and CCAsensitive genes and the binding by CcpA to them is strongly stimulated by Ser46-phosphorylated HPr protein (Deutscher, 2008).
Sugar catabolism can also be mastered by specific secondary regulators, common in LAB and acting locally, falling to diverse protein families such as LacI, LysR, AraC, GntR, DeoR, RpiR, or BglG.
The mechanisms of transcriptional regulation of the Leloir pathway genes have been elucidated in some LAB species. Gal-lac operons are frequently regulated by specific transcription regulators, which belong to the LacI type. In Streptococcus (S.) thermophilus and S. mutans, GalR acts as a transcription activator and repressor of the lac and gal operons, respectively (Ajdić & Ferretti, 1998;Vaughan et al., 2001). In both species, the GalR-encoding galR gene is oriented divergently from the structural genes of the Leloir operon. In Lactobacillus casei, a potential transcription regulatory gene, galR, has been identified in the gal operon and is transcribed in the same direction (Bettenbrock & Alpert, 1998). In Lactobacillus helveticus, the inducible genes lacLM (encoding β-galactosidase) of the unusually organized gal and lac gene cluster are regulated at the transcriptional level by LacR repressor (Fortina, Ricci, Mora, Guglielmetti, & Manachini, 2003). No specific regulatory genes have been identified for the Leloir operon in L. lactis to date, albeit it has been demonstrated that expression of gal genes is under CcpAdependent catabolic repression Zomer, Buist, Larsen, Kok, & Kuipers, 2007).
We propose that YugA activates expression of lacS and the lacgal genes localized downstream by binding to the lacS upstream DNA region containing a putative promoter. Because of this newly identified regulatory function of YugA, we propose to re-name it GlaR (galactose-lactose operon Regulatory protein). To the best of our knowledge, this is the first report exploring a specific GlaRdependent regulatory mechanism of the Leloir pathway genes in L. lactis IL1403 at the molecular level. We examined the effects of glaR deletion and found that the lack of GlaR precludes the strain's growth in galactose-containing media and abolishes lacS gene expression. These results shed new light on the regulation of carbohydrate catabolism in this biotechnologically important bacterium and reveal a new regulatory protein. Notably, the described mechanism of control of galactose and lactose catabolism by enzymes of the Leloir utilization pathway is unique among LAB.

| Bacterial strains, media, and plasmids
Bacterial strains and plasmids used in this study are shown in Table 1.

| Construction of glaR deletion mutant and complementing plasmid
Lactococcus lactis IL1403 glaR deletion strain (L. lactis IL1403ΔglaR) was generated by double crossover between pGhost9 carrying DNA fragments flanking the glaR gene and the corresponding chromosomal region. The glaR upstream and downstream DNA fragments were amplified with, respectively, the glaRUPf/glaRUPr and glaRDOWNf/glaRDOWNr primer pairs ( Table 1). The obtained DNA fragments were cloned in the proper orientation in the integrative vector pGhost9, producing pGhost9ΔglaR. This deletion plasmid was transported into L. lactis IL1403 and homologous recombination was enforced by 10 −3 dilution of an overnight culture and incubation at nonpermissive temperature (38°C). Cells harboring pGhost9ΔglaR in the chromosome were cultivated at 38°C on G-M17Em. Removal from the chromosome and elimination of pGhost9 from L. lactis were performed by growing the integrants in G-M17 without antibiotic for at least 100 generations at the permissive temperature (28°C).
The genomic organization of the resulting glaR deletion strain (L. lactis IL1403ΔglaR) was confirmed by determining its sensitivity to Em and by sequencing of the mutated region.
To complement the glaR deletion, the glaR gene containing its putative promoter region was amplified using glaRUPf and glaRDOWNr primers (Table 1) and ExTaq polymerase giving the glaR(A) insert.

| Quantification of gene expression by reverse transcription-quantitative PCR (RT-qPCR)
RNA was isolated following manufacturer's instructions with the use of GeneMATRIX Universal RNA Purification Kit (EURx, Poland) from 10 ml of L. lactis IL1403 and L. lactis IL1403ΔglaR cultures grown in G-M17, C-M17, Gal-M17 or GalC-M17 and collected from midexponential phase (OD 600 = 0.6). RNA was isolated from at least three independent cultures.
First-strand cDNA was obtained from DNAse I (Sigma-Aldrich, USA)-treated RNA with random primers by the use of the RevertAid(TM) First-Strand cDNA Synthesis Kit (Thermo Fisher Scientific) according to manufacturer's instructions. qPCR assays on the cDNA were carried out in a 7500 Real-Time PCR System (Applied Biosystems, USA) and following the previously described methodology (Aleksandrzak-Piekarczyk et al., 2015). Specific primers for genes (Table 1) were created with Primer Express software (Applied Biosystems). The results were normalized to the L. lactis IL1403 reference genes tuf and purM coding for elongation factor TU and phosphoribosylaminoimidazole synthetase, respectively.

| Growth testing for carbon source utilization
Growth tests were performed using a Microbiology Reader Analyser, Bioscreen C (Oy Growth Curves Ab Ltd, Finland) in 200 μl of CDM with the required sugars (glucose, galactose or cellobiose). OD 600 of the bacterial cultures was recorded every 60 mins of growth up to 40 hr at 30°C. The assays were carried out in triplicate.

| Overproduction and purification of GlaR
The self-cleavable IMPACT TM affinity tag system (New England Biolabs, USA) was used to purify the GlaR protein. E. coli BL21 competent cells were transformed with the pTXB1 plasmid carrying the glaR gene. The obtained transformants were verified by colony PCR, with specific primers ptXB1for and glaRBamHrev. LB medium (600 ml) containing 100 μg/ml ampicillin was inoculated with a freshly grown colony and incubated at 37°C with shaking until an OD 660 of 0.5 was reached. After induction of the glaR gene expression using 0.3 mM IPTG, the culture was incubated overnight at 18°C. Then, the cells were pelleted by centrifugation (3000xg, 10 min, 4°C) and stored at −20°C until use. All subsequent purification steps were carried out at 4°C. The frozen cells were resuspended in 10 ml of column buffer A (25 mM Tris-HCl, pH 8.0; 500 mM NaCl; 10% glycerol) and were

| GlaR is crucial for L. lactis IL1403 growth on galactose
To assess the possible role of GlaR, a L. lactis IL1403ΔglaR mutant strain was constructed lacking the glaR gene and its growth was tested in CDM with different sugars and compared with its parental wild-type IL1403 strain. No significant differences were found between the growth of these two strains in G-CDM or C-CDM, but in a galactose-supplemented medium, the mutant lacking GlaR was unable to grow completely ( Figure 2).
Transformation of pGhost9glaR into IL1403ΔglaR that led to the creation of the L. lactis IL1403ΔglaRpGhost9glaR strain, fully reversed the effect of the glaR deletion, restoring the mutant's growth in medium supplemented with galactose.

| GlaR is a transcriptional activator of the gal-lac genes
To define the influence of GlaR on the expression of the gallac operon genes in response to various sugars, using RT-qPCR, for those genes varied between 2.6 (galE) and 8 (galM) (Figure 3b).
For the negative control yufC, the GlaR activation coefficient was close to 1, indicating-as expected-a lack of GlaR-dependent activation ( Figure 3b).
The lowest transcript levels of the genes studied were detected in L. lactis IL1403 wild-type and IL1403ΔglaR growing under repressive conditions (G-M17), most likely due to the downregulation of gal-lac genes by CcpA, as described previously . Expression of most of the Leloir genes increased in both strains in the medium supplemented with cellobiose ( Figure 3a) indicating a release from catabolic repression. Notably, lacS mRNA was not detected in either of these media in either of the strains.
In comparison with cellobiose, higher transcript levels of the gal-lac operon genes were detected when the wild-type strain was grown in media supplemented with galactose (Gal-M17 or GalC-M17) ( Figure 3a). The activation by galactose calculated as the ratio of expression in GalC-M17 and in C-M17 was the highest for the lacS gene, and for the other gal-lac genes, it varied from 1.6 (galE) to 6 (galM) (Figure 3b). In the glaR mutant downstream of lacS, these ratios were ca. 1 indicating that in the absence of GlaR, the galactosedependent activation of the gal-lac genes does not occur. Also for the negative control gene yufC, its expression levels with and without galactose were similar in both the wild-type strain and in the glaR mutant further confirming that it is not subject to galactose induction ( Figure 3B).

| GlaR activates expression of the Leloir operon by binding to the lacS promoter region
To identify the genomic region to which the GlaR protein binds specifically, an in vitro EMSA test was performed with selected upstream regions containing potential promoters of the Leloir operon genes (lacS, galM, galT, thgA, and galE) and of glaR and purified GlaR protein. An unrelated dsDNA containing the yufA upstream region was used as a control to test for nonspecific binding. No nonspecific interactions were detected at GlaR concentrations up to 4 μM; therefore, this concentration was used to investigate specific binding (Figure 4a). At this concentration, GlaR bound to the putative lacS promoter but it did not form specific complexes with any other putative promoters tested (Figure 4b).
Notably, GlaR bound to the lacS dsDNA also at lower concentrations (1-3.5 μM) (Figure 4a), indicating that the interaction is fairly strong.

| GlaR expression is inducible by galactose but insensitive to CcpA-mediated catabolite repression
CcpA is a master transcriptional regulator controlling carbohydrate utilization and metabolism genes in gram-positive bacteria including L. lactis (Hueck & Hillen, 1995;Zomer et al., 2007). As the promoter region of glaR contains a potential cre sequence (Figure 1) that could be recognized by CcpA, we sought to determine the role of
Members of the RpiR family harbor a DNA-binding HTH domain and a phospho-sugar-binding SIS motif, respectively, at their N-and C-terminal regions (Bateman, 1999;Teplyakov, Obmolova, Badet-Denisot, Badet, & Polikarpov, 1998 in question form a single operon with the promoter preceding the lacS gene. Indeed, using EMSA, we found that GlaR does bind specifically to a region upstream of lacS, but not to the putative promoters of the other genes downstream of lacS. Notably, the presence of GlaRindependent promoter-like regions upstream of these genes explains why they were expressed at a submaximal level even in the absence of galactose/or GlaR. We additionally confirmed that functional expression of lacS requires the action of GlaR by showing that the L. lactis IL1403ΔglaR was unable to grow in galactose medium. LacS permease is the main transporter used for galactose (but not for lactose; Aleksandrzak-Piekarczyk et al., 2005) uptake in IL1403 cells, and its inactivation leads to the galphenotype (our unpublished data).
Remarkably, also the transcription of glaR was induced substantially in galactose-containing medium in comparison with cellobiose, which could in part explain the effect of galactose on the GlaR-dependent expression of the gal-lac operon. It also suggested possible autoregulation of glaR expression by GlaR. Autoregulation is frequent in prokaryotic gene regulation strategies and has been reported for numerous transcription regulators (Gerlach, Valentin-Hansen, & Bremer, 1990;Meng, Kilstrup, & Nygaard, 1990;Morel, Lamarque, Bissardon, Atlan, & Galinier, 2001;Vaughan et al., 2001;Weickert & Adhya, 1993). However, we could not confirm a direct involvement of GlaR in glaR activation as no GlaR binding to the glaR promoter region was found by EMSA (Figure 4b). A plausible explanation includes an indirect control by GlaR (e.g., via an alternative regulator under the control of GlaR) or the action of another galactose-dependent but GlaR-independent mechanism.
Both lacS and glaR are preceded by cre boxes suggesting that their expression is under CcpA-driven carbon catabolite repression (CCR).
Indeed, in the presence of glucose, transcriptional arrest of all the genes under the control of the lacS promoter was detected, whereas cellobiose or galactose caused a relief from CCR. This phenomenon has already been studied in another L. lactis strain, MG1363 , in which the Leloir operon differs from the one of IL1403 but is also subject to CcpA-driven catabolic repression. In contrast, we found that that CcpA is not engaged in the regulation of glaR expression in L. lactis IL1403. One reason for this could be the two-nucleotide deviation of the cre sequence upstream of glaR (TaAAAACGaTTTCA) form the cre consensus WGWAARCGYTWWMA (Zomer et al., 2007). The two adenine mismatches may prevent or impair CcpA interaction with its operator and thus allow of the glaR transcription also in repressive conditions (glucose).
In summary, here, we have documented unusual mechanism of gal-lac operon activation in L. lactis IL1403 and, by similarity, probably also in other Lactococcus spp. No similar mechanism has been reported in other LAB species. This regulation relies on galactoseinducible and GlaR-dependent transcriptional activation of the lacS promoter inducing the lacS gene itself and the other lac and Leloir pathway genes located downstream.

ACK N OWLED G M ENTS
RT-qPCR assays amplification were performed at the Genetic Modifications Analysis Lab, IBB PAS. Proteomic analyses were performed at the Laboratory of Mass Spectrometry, IBB PAS. The equipment used for proteomic analysis was funded in part by the Centre for Preclinical Research and Technology (CePT), a project cosponsored by the European Regional Development Fund and Innovative Economy, The National Cohesion Strategy of Poland.

E TH I C A L S TATEM ENT
This article does not contain any studies with human or animals performed by any of the authors.

DATA ACCE SS I B I LIT Y
The authors declare that all data generated or analyzed during this study are included in this article.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.