Lrp, a leucine-responsive protein, regulates branched-chain amino acid transport genes in Escherichia coli.

We investigated the relationship between two regulatory genes, livR and lrp, that map near min 20 on the Escherichia coli chromosome. livR was identified earlier as a regulatory gene affecting high-affinity transport of branched-chain amino acids through the LIV-I and LS transport systems, encoded by the livJ and livKHMGF operons. lrp was characterized more recently as a regulatory gene of a regulon that includes operons involved in isoleucine-valine biosynthesis, oligopeptide transport, and serine and threonine catabolism. The expression of each of these livR- and lrp-regulated operons is altered in cells when leucine is added to their growth medium. The following results demonstrate that livR and lrp are the same gene. The lrp gene from a livR1-containing strain was cloned and shown to contain two single-base-pair substitutions in comparison with the wild-type strain. Mutations in livR affected the regulation of ilvIH, an operon known to be controlled by lrp, and mutations in lrp affected the regulation of the LIV-I and LS transport systems. Lrp from a wild-type strain bound specifically to several sites upstream of the ilvIH operon, whereas binding by Lrp from a livR1-containing strain was barely detectable. In a strain containing a Tn10 insertion in lrp, high-affinity leucine transport occurred at a high, constitutive level, as did expression from the livJ and livK promoters as measured by lacZ reporter gene expression. Taken together, these results suggest that Lrp acts directly or indirectly to repress livJ and livK expression and that leucine is required for this repression. This pattern of regulation is unusual for operons that are controlled by Lrp.     

Characterization of Lrp, and Escherichia coli regulatory protein that mediates a global response to leucine

Exogenous leucine affects the expression of a number of different operons in Escherichia coli. For at least some of these operons, the leucine-related effect is mediated by a protein called Lrp (Leucine-responsive regulatory protein). The purification of Lrp to near homogeneity is described. Lrp is a moderately abundant, basic protein composed of two subunits of molecular mass 18.8 kDa each. In addition, the corresponding protein was purified from a strain having a mutation within the gene that encodes Lrp (lrp). This mutation (lrp-1) causes high constitutive expression of ilvIH, one of the operons controlled by Lrp (Platko, J. V., Willins, D.A., and Calvo, J.M. (1990) J. Bacteriol. 172, 4563-4570). The Lrp-1 and Lrp proteins have similar physical properties, but they show some differences in the characteristics with which they bind DNA upstream of the ilvIH promoter. The nucleotide sequences of the lrp and lrp-1 genes differ by only a single nucleotide, a C to G change that would substitute a Glu for an Asp at amino acid 114. Lrp has some amino acid sequence similarity to AsnC, a protein that regulates asnA expression (Kolling, R., and Lother, H. (1985) J. Bacteriol. 164, 310-315).     

The tdh and serA operons of Escherichia coli: mutational analysis of the regulatory elements of leucine-responsive genes.

The tdh promoter of Escherichia coli is induced seven- to eightfold when cells are grown in the presence of exogenous leucine. A scheme was devised to select mutants that exhibited high constitutive expression of the tdh promoter. The mutations in these strains were shown to lie within a previously identified gene (lrp) that encodes Lrp (leucine-responsive regulatory protein). By deletion analysis, the site of action of Lrp was localized to a 25-bp region between coordinates -69 and -44 of the tdh promoter. Disruption of a 12-bp presumptive target sequence found in this region of tdh resulted in constitutively derepressed expression from the tdh promoter. Similar DNA segments (consensus, TTTATTCtNaAT) were also identified in a number of other promoters, including each of the Lrp-regulated promoters whose nucleotide sequence is known. The sequence of the promoter region of serA, an Lrp-regulated gene, was determined. No Lrp consensus target sequence was present upstream of serA, suggesting that Lrp acts indirectly on the serA promoter. A previously described mutation in a leucine-responsive trans-acting factor, LivR (J. J. Anderson, S. C. Quay, and D. L. Oxender, J. Bacteriol. 126:80-90, 1976), resulted in constitutively repressed expression from the tdh promoter and constitutively induced expression from the serA promoter. The possibility that LivR and Lrp are allelic is discussed.     

Use of an in vivo titration method to study a global regulator: effect of varying Lrp levels on expression of gltBDF in Escherichia coli.

Most studies of global regulatory proteins are performed in vitro or involve phenotypic comparisons between wild-type and mutant strains. We report the use of strains in which the gene for the leucine-responsive regulatory protein (lrp) is transcribed from isopropyl-beta-D-thiogalactopyranoside (IPTG)-inducible promoters for the purpose of continuously varying the in vivo concentration of Lrp. To obtain a broad range of Lrp concentrations, strains were employed that contained the lrp fusion either in the chromosome (I. C. Blomfield, P. J. Calie, K. J. Eberhardt, M. S. McClain, and B. I. Eisenstein, J. Bacteriol. 175:27-36, 1993) or on a multicopy plasmid. Western blot (immunoblot) analysis with polyclonal antiserum to Lrp confirmed that Lrp levels could be varied more than 70-fold by growing the strains in glucose minimal 3-(N-morpholino)propanesulfonic acid (MOPS) medium containing different amounts of IPTG. Expression of an Lrp-regulated gltB::lacZ operon fusion was measured over this range of Lrp concentrations. beta-Galactosidase activity rose with increasing Lrp levels up to the level of Lrp found in wild-type strains, at which point expression is maximal. The presence of leucine in the medium increased the level of Lrp necessary to achieve half-maximal expression of the gltB::lacZ fusion, as predicted by earlier in vitro studies (B. R. Ernsting, J. W. Denninger, R. M. Blumenthal, and R. G. Matthews, J. Bacteriol. 175:7160-7169, 1993). Interestingly, levels of Lrp greater than those in wild-type cells interfered with activation of gltB::lacZ expression. The growth rate of cultures correlated with the intracellular Lrp concentration: levels of Lrp either lower or higher than wild-type levels resulted in significantly slower growth rates. Thus, the level of Lrp in the cell appears to be optimal for rapid growth in minimal medium, and the gltBDF control region is designed to give maximal expression at this Lrp level.     

Identification of endogenous inducers of the mal regulon in Escherichia coli.

The expression of the maltose regulon in Escherichia coli is induced when maltose or maltodextrins are present in the growth medium. Mutations in malK, which codes for a component of the transport system, result in the elevated expression of the remaining mal genes. Uninduced expression in the wild type, as well as elevated expression in malK mutants, is strongly repressed at high osmolarity. In the absence of malQ-encoded amylomaltase, expression remains high at high osmolarity. We found that uninduced expression in the wild type and elevated expression in malK mutants were paralleled by the appearance of two types of endogenous carbohydrates. One, produced primarily at high osmolarity, was identified as comprising maltodextrins that are derived from glycogen or glycogen-synthesizing enzymes. The other, produced primarily at low osmolarity, consisted of an oligosaccharide that was not derived from glycogen. We isolated a mutant that no longer synthesized this oligosaccharide. The gene carrying this mutation, termed malI, was mapped at min 36 on the E. coli linkage map. A Tn10 insertion in malI also resulted in the loss of constitutivity at low osmolarity and delayed the induction of the maltose regulon by exogenous inducers.     

Osmoregulation of the maltose regulon in Escherichia coli.

The maltose regulon consists of four operons that direct the synthesis of proteins required for the transport and metabolism of maltose and maltodextrins. Expression of the mal genes is induced by maltose and maltodextrins and is dependent on a specific positive regulator, the MalT protein, as well as on the cyclic AMP-catabolite gene activator protein complex. In the absence of an exogenous inducer, expression of the mal regulon was greatly reduced when the osmolarity of the growth medium was high; maltose-induced expression was not affected, and malTc-dependent expression was only weakly affected. Mutants lacking MalK, a cytoplasmic membrane protein required for maltose transport, expressed the remaining mal genes at a high level, presumably because an internal inducer of the mal system accumulated; this expression was also strongly repressed at high osmolarity. The repression of mal regulon expression at high osmolarity was not caused by reduced expression of the malT, envZ, or crp gene or by changes in cellular cyclic AMP levels. In strains carrying mutations in genes encoding amylomaltase (malQ), maltodextrin phosphorylase (malP), amylase (malS), or glycogen (glg), malK mutations still led to elevated expression at low osmolarity. The repression at high osmolarity no longer occurred in malQ mutants, however, provided that glycogen was present.     

The lrp gene product regulates expression of lysU in Escherichia coli K-12.

In Escherichia coli K-12, expression of the lysU gene is regulated by the lrp gene product, as indicated by an increase in the level of lysyl-tRNA synthetase activity and LysU protein in an lrp mutant. Comparison of the patterns of protein expression visualized by two-dimensional gel electrophoresis indicated that LysU is present at higher levels in an lrp strain than in its isogenic lrp+ parent. The purified lrp gene product was shown to bind to sites upstream of the lysU gene and to protect several sites against DNase I digestion. A region extending over 100 nucleotides, between 60 and 160 nucleotides upstream from the start of the lysU coding sequence, showed altered sensitivity to DNase I digestion in the presence of the Lrp protein. The extent of protected DNA suggests a complex interaction of Lrp protein and upstream lysU DNA.