Carbon catabolite repression and global control of the carbohydrate metabolism in Lactococcus lactis

In view of the economic importance of fermented dairy products considerable scientific attention has been given to various steps of fermentation processes, including the L-lactate formation of lactic acid bacteria (de Vos, 1996). In particular, the carbohydrate metabolism of L. lactis has been the subject of extensive research and several genes encoding proteins involved in the central carbohydrate metabolism have been described (Llanos et al., 1992; Llanos et al., 1993; Cancilla et al., 1995a; Cancilla et al., 1995b; Qian et al., 1997). Although several findings have established that the carbohydrate metabolism is subject to several forms of regulation, detailed information concerning this regulation and, in particular, the transcriptional control of the central carbohydrate metabolism is lacking (Collins and Thomas, 1974; Fordyce et al., 1982; Hardman et al., 1985; Garrigues et al., 1997).

A better understanding of the regulatory mechanisms involved would be an advantage for metabolic pathway engineering. Metabolic engineering is mainly aimed at the optimization of the metabolism and the diversion from L-lactate to other desired metabolites. The metabolite formation depends on the activity of enzymes of the central metabolic pathway and is therefore also subject to regulatory mechanisms in response to the carbon source provided. The research reported in this thesis has focussed on carbon catabolite repression (CCR), a global control system which regulates the transcription of genes involved in the carbohydrate metabolism depending on the carbon source availability (Hueck and Hillen, 1995). An overview of the present state of the art on CCR in Gram-positive bacteria is presented in Chapter 1.

The aim of the work presented in this thesis was to investigate the elements involved in CCR in L. lactis and their effects on the carbohydrate metabolism. Several cis- and trans-acting elements involved in specific and global control systems were identified and their role in the transcriptional and allosteric control of carbohydrate metabolism was characterized. The salient features of their sequences are summarized in the Appendix.

Chapter 2 describes the transcriptional and functional analysis of Tn5276-located genes involved in sucrose metabolism in L. lactis. The observation that the transcription of the previously cloned sacA gene was subject to glucose repression and the identification of a cre element in the promoter region of the sucrose genes lead to the choice of the sucrose genes as a model system to study the effects of CCR. (Rauch and de Vos, 1992). In addition to the sacA gene, encoding a sucrose-6-phosphate hydrolase, three new complete genes were identified. The sacB gene encodes a sucrose-specific EII protein of the phosphotransferase system (PTS) and its disruption resulted in the inability of the strain to utilize sucrose as carbon and energy source, thereby confirming the functionality of the sacB gene in the sucrose metabolism of L. lactis.. Downstream of the sacB gene the sacK gene was identified encoding a fructokinase. Partially overlapping sacA, the sacR gene was identified, the deduced protein sequence of which showed high homology to regulatory proteins of the LacI/GalR family. The L. lactis sucrose gene is the only gene cluster reported so far containing all three structural genes necessary for the complete catabolism of sucrose as well as a specific regulatory gene.

Transcriptional analysis of the sucrose gene cluster lead to the identification of three sucrose-inducible transcripts. One of 3.2 kb containing sacB and sacK which initiates from the sacB promoter and is likely to terminate at the inverted repeat located downstream of the sacK gene. Another transcript of 3.4 kb, was shown to contain the sacA and sacR genes and initiates from the sacA promoter. Furthermore, a third sucrose-inducible transcript of 1.8 kb was identified, which contains the only the sacR gene and initiates from a promoter which was mapped upstream of the sacR gene.

Disruption of the sacR gene resulted in the constitutive transcription of the sacBK and sacAR transcripts suggesting that SacR acts as a negative regulator of transcription. Under non-induced circumstances SacR most likely binds to the putative operator sites that were identified in the three promoters of the sucrose operon, resulting in repression of transcription. The presence of an inducer molecule (most likely sucrose-6-P) may result in the dissociation of SacR from the operator leading to transcription of the sucrose genes. The sacR gene is subject to a negative autoregulatory mechanism that results in higher levels of the repressor protein under induced circumstances compared to the non-induced situation. This control system allows for a very tight control of the expression of all the sucrose genes and to fast adaptation to environmental changes and resembles the system identified for the transcriptional control of the galactose genes in E. coli (Weickert and Adhya. 1993).

The transcription of the sacA and sacB promoters in the wild-type strain is subject to glucose repression. The disruption of the sacR gene resulted in the complete absence of the glucose repression observed in the wild-type. This suggested that the glucose repression is dependent on SacR and is most likely due to a reduced induction resulting from lower concentrations of inducer molecules rather than the activity of a general regulatory mechanism like CcpA-mediated CCR. The concentration of inducer molecules may be affected by inducer control mechanisms like inducer exclusion and inducer expulsion (see below). The tight regulation of the expression of the sac genes by the operon-specific regulator SacR and the apparent independency of the chromosomally encoded CcpA-mediated CCR, may be a consequence of their location on a conjugative transposon of non-lactococcal origin that may be transferred to a variety of hosts.

Chapter 3 deals with the detection of CcpA-like proteins in different Gram-positive bacteria including L. lactis. Polyclonal antibodies raised against purified Bacillus megaterium CcpA were used to screen protein extracts of several Gram-positive bacteria of high and low GC content. The results indicate that cross-reacting proteins were present in all Gram-positive bacteria tested, and suggest that a CcpA-mediated regulatory mechanism, like CCR, is a wide-spread phenomenon.

In Chapter 4 the cloning and analysis of the L. lactis ccpA gene is described. An L. lactis expression library was constructed and screened with the CcpA antiserum resulting in the isolation of the L. lactis ccpA gene. In contrast to the Staphylococcus xylosus and Lactobacillus casei ccpA genes, the expression level of the L. lactis ccpA gene does not vary significantly in response to the carbon source provided (Egeter and Brückner, 1996; Monedero et al., 1997). The observed negative autoregulation of ccpA in Staphylococcus xylosus and Lactobacillus casei probably provides the cell with a mechanism controlling CCR by varying the level of CcpA protein. The observed regulation of the expression level of the L. lactis ptsH gene (see below) might allow a similar regulation of CCR activity in L. lactis because it affects the concentration of HPr(Ser-P), which functions as a coregulator. Inactivation of the L. lactis ccpA gene resulted in a reduced growth rate on all sugars tested, suggesting an involvement of CcpA in the regulation of a key metabolic pathway.

Because the sucrose gene cluster, despite the presence of a cre element, appeared to be independent of CcpA-mediated CCR, a new model system to study CCR was required. The recently identified galactose gene cluster, containing the genes involved in the catabolism of galactose via the Leloir pathway, contains a cre element in the promoter region and was therefore a likely candidate for CcpA-mediated CCR (Grossiord et al., 1998). Disruption of the ccpA gene confirmed this involvement because the transcription of the gal operon in the resulting strain was partly relieved from CCR. However, the transcription of the gal operon was not completely relieved from CCR, suggesting that another regulatory mechanism was functional. Possible mechanisms for mediating the residual glucose repression are inducer exclusion and inducer expulsion.

The expression of the genes encoding the glycolytic enzymes pyruvate kinase and L-lactate dehydrogenase is subject to carbon source dependent regulation since higher activities of both enzymes were measured in cells grown on glucose compared to cells grown on galactose. The genes encoding pyruvate kinase and L-lactate dehydrogenase are located in an operon structure together with the gene encoding the glycolytic enzyme phosphofructokinase (Llanos et al., 1992; Llanos et al., 1993). This operon, designated las for lactic acid synthesis, contains a cre site in the promoter region and is therefore a likely candidate for CcpA-mediated regulation. The inactivation of the ccpA gene resulted in a four-fold reduction of the transcription of the las operon genes indicating that CcpA acts as a transcriptional activator. CcpA has been reported to act as a transcriptional activator of the Bacillus subtilis alsS and ackA genes encoding a-acetolactate synthase and acetate kinase, respectively (Grundy et al., 1993; Renna et al., 1993). However, the CcpA-mediated transcriptional activation of the L. lactis las operon is the first report of transcriptional control of genes encoding enzymes involved in the central carbohydrate metabolism in Gram-positive bacteria.

The lower transcription level of the las operon was reflected in reduced activities of pyruvate kinase and L-lactate dehydrogenase, resulting in a lower production of L-lactate. Furthermore, the fermentation pattern after growth on glucose had changed from almost homolactic, in case of the wild-type strain, to a more mixed-acid pattern in the ccpA knock out strain. The observation that the B. subtilis ccpA was capable of complementing the transcriptional activation, combined with the presence of cre sites in the promoter regions of glycolytic genes of different Gram-positive bacteria, strongly suggests that the observed transcriptional activation of glycolytic genes is not limited to L. lactis.

The analysis of the L. lactis ptsHI genes is described in Chapter 5. The ptsHI operon is transcribed as a 2.0-kb transcript from a single promoter mapped upstream of the ptsH gene. Furthermore, a 0.3-kb transcript was detected that contained only the ptsH gene. This transcript originates from the ptsH promoter and terminates at a stem-loop structure located downstream of the ptsH gene. This transcriptional organization most likely results in a higher expression of the ptsH gene, explaining the higher amount of HPr protein compared to enzyme I as observed in several bacteria, including Staphylococcus carnosus (Kohlbrecher et al., 1992).

The expression of the ptsHI genes appeared to be regulated since lower transcription levels were observed when the cells were grown on the non-PTS sugar galactose compared to the PTS sugar glucose. Induction of the ptsHI expression by glucose has also been observed in Bacillus subtilis and allows the cell to control the activity of the PTS in response to the carbon source availability (Stülke et al., 1997). The glucose induction of the Bacillus subtilis ptsHI genes is mediated via an antitermination mechanism and is dependent of a characteristic terminator structure located upstream of the ptsH gene. Because no obvious recognition sites for transcriptional regulators could be identified at relevant positions in the L. lactis ptsHI operon, the mechanism by which the transcriptional control of this operon operates remains to be clarified.

The disruption of both the ptsH and the ptsI genes resulted in the absence of growth on sucrose and fructose, indicating that these sugars are exclusively taken up by the PTS. The growth rate on glucose was severely reduced, suggesting that in addition to the PTS, another glucose uptake system is present. This finding is in agreement with the results of Thompson and coworkers who presented biochemical evidence that L. lactis uses the PTS and a non-PTS permease for the uptake of glucose (Thompson et al., 1985). Complementation of the ptsH and ptsI genes with the appropriate L. lactis genes under the control of an inducible promoter confirmed the functionality of both genes. Furthermore, the growth rate on galactose and maltose, two sugars that are most likely taken up via a non-PTS system was reduced two-fold. This observation suggests an involvement of the PTS with either protein activities involved in the galactose and maltose catabolism or the regulation of the expression of the encoding genes. In Gram-positive bacteria, the PTS has been reported to control catabolic pathways, like the Bacillus subtilis levanase or glycerol pathways (Stülke et al., 1995; Charrier et al., 1997) or the lactose uptake in Streptococcus thermophilus (Poolman et al., 1995), by HPr(His-P)-dependent phosphorylation of either enzymes or regulatory proteins resulting in enhanced or reduced activities.

In order to analyze the regulatory role of HPr(Ser-P) in the sugar metabolism of L. lactis, mutant HPr proteins were constructed that were affected in the phosphorylation of residue Ser-46. Overproduction in a wild-type strain of HPr(S46D) where residue Ser-46 has been changed into an aspartic acid, that mimics a phosphorylated serine, resulted in a reduction of the growth rate on galactose, whereas the growth rate on glucose was not affected. These results suggested that HPr(Ser-P) is involved in the CCR of the galactose metabolism. Whether this regulation occurs in combination with the inducer control mechanisms or at the transcriptional level in combination with CcpA remains to be determined.

In addition to its role in the CCR of the genes involved in the galactose metabolism, HPr(Ser-P) is also involved in the positive regulation of the enzymes encoded by the las operon. Expression of the gene encoding S46D HPr in wild-type cells grown on galactose resulted in increased activities of both pyruvate kinase and L-lactate dehydrogenase. Since the positive effect of the production of S46D HPr on the activities of pyruvate kinase and L-lactate dehydrogenase depends on the presence of the ccpA gene, it is feasible that the regulation occurs at the transcriptional level.

These findings established the function of HPr(Ser-P) as a signal molecule in several allosteric and transcriptional metabolic control systems in L. lactis . The occurrence of HPr(Ser-P) results in a reduced entry of new sugar phosphates into the glycolysis due to the inducer control mechanisms and CCR. In addition, the catabolite activation of the las operon results in an increased flux through the glycolysis. Consequently, the inducer control systems as well as the CcpA-mediated catabolite control can be seen as mechanisms to prevent the wasteful and possibly toxic accumulation of early glycolytic intermediates.

The studies described in this thesis have resulted in the characterization of different regulatory mechanisms involved in the control of the carbohydrate metabolism in L. lactis. The regulation of the expression of the Tn5276-located sucrose genes appeared to be dependent on the operon-specific regulator, SacR. This apparent independence of chromosomally encoded global regulation might be a result of the fact that these genes are located on transposons, which can be conjugally transferred to other species and therefore require a host-independent transcriptional control system. The analysis of the L. lactis ccpA gene lead to the identification of two CcpA-dependent regulatory systems i.e. the CcpA-mediated CCR of the galactose operon and the transcriptional activation of the glycolytic las operon. The CCR of the galactose operon mediated by CcpA confirmed previous reports on the role of CcpA in other Gram-positive bacteria. So far, CcpA-mediated transcriptional activation of gene expression was only identified in B. subtilis.

However, the observation that the L. lactis CcpA mediates the expression of genes encoding key enzymes of the glycolysis suggests that CcpA is involved in the global transcriptional control of the metabolic activity, in response to carbon source availability. The observation that the seryl-phosphorylated form of HPr is involved as coregulator in CCR as well as the CcpA-mediated transcriptional activation of gene expression established its important role as signal molecule reflecting the energy state of the cell.

The new information concerning the elements involved in the CcpA-mediated catabolite activation of the central carbohydrate metabolism can be used to accelerate the L-lactate formation in certains strains. The disruption of the ccpA gene most likely results in an increased intracellular concentration of early glycolytic intermediates like FDP, which might lead to an increased biosynthesis of e.g. extracellular polysaccharides since precursors thereof are derived from these metabolites. Data emerging from the L. lactis sequencing project (Bolotin et al., 1998) in combination with new technologies like the microarray technology (de Saizieu et al., 1998), allowing genome-wide monitoring of gene expression, and the knowledge of the global regulatory mechanisms presented in this thesis will facilitate the design of metabolic engineering strategies.