Chapter 23

Sources of Nitrogen and Their Utilization

LAWRENCE J. REITZER

INTRODUCTION

This review focuses on the utilization of specific compounds as sole nitrogen sources by Escherichia coli, Salmonella typhimurium (formal designation, Salmonella enterica serovar Typhimurium), and occasionally Klebsiella aerogenes. Chapter 22 considers amino acids as carbon sources; its discussion of histidine, proline, and d-serine degradation is more complete than that presented here. Chapter 24 discusses the enzymes of ammonia and nitrogen assimilation as well as the response to nitrogen limitation. It should be noted that ethanolamine, purines, pyrimidines, N-acetylglucosamine, nitrate, and nitrite are not discussed here, although they can be used as nitrogen sources.

Nitrogen Sources and the Products of Their Catabolism

An exhaustive survey of nitrogen sources for S. typhimurium showed that about 25 different compounds can serve as nitrogen sources (36). These compounds are generally amino acids but also include purines, pyrimidines, and nitrate (at least during anaerobic growth). E. coli degrades a similar set of compounds (118). However, K. aerogenes can degrade many more compounds, including 19 amino acids and compounds such as urea, than either S. typhimurium or E. coli (12, 36, 118). For these organisms, ammonia is considered the preferred nitrogen source because it supports the fastest growth rate.

The role of ammonia in nitrogen source catabolism is unique—it is the only absolutely required product of such catabolism. Catabolism of a particular nitrogen source sometimes produces glutamate, either directly or through transamination. If not, then ammonia becomes necessary for the ATP-dependent synthesis of glutamate through the combined actions of glutamine synthetase and glutamate synthase (see chapter 24). Therefore, the production of glutamate spares the need for energy-consuming glutamate synthesis.

Pathways for the degradation of some nitrogen sources are reasonably well understood. For alanine and serine, ammonia is produced by known reactions, and this ammonia is used to synthesize glutamine and glutamate. For amino acids, such as asparagine and histidine, known enzymes produce both ammonia and glutamate. However, the degradation of other amino acids is only partially understood. For example, the reactions that produce ammonia from the catabolism of arginine, aspartate, glutamate, ornithine, proline, and putrescine have not been unambiguously determined.

The Ntr Response: Its Function and Regulation

Nitrogen-limited growth, i.e., growth in a glucose-containing medium in which another nitrogen source replaces ammonia, induces the synthesis of proteins that transport and degrade nitrogenous compounds and of glutamine synthetase, which assimilates ammonia. This coordinated response is called the Ntr (nitrogen regulated) response, and it has been thoroughly reviewed elsewhere (59, 118; see also chapter 24). Expression of many genes of nitrogen source catabolism does not require specific induction, which suggests that the primary function of the Ntr response is assimilation of any of a number of nitrogen sources that might be available when nitrogen becomes growth rate limiting. For example, growth with glutamine as a nitrogen source induces the periplasmic components of transport systems for aspartate, glutamate, lysine, ornithine, arginine, and histidine (54).

Many enzymes of nitrogen source catabolism are Ntr enzymes. The central regulators of this response are nitrogen regulator I (NR_I, also called NtrC), NR_II (also called NtrB), and σ ⁵⁴ complexed to core RNA polymerase (see chapters 24 and 86). The most important activator is NR_I, which when phosphorylated activates a number of genes, including some that code for other transcriptional activators (5).

Nitrogen availability controls the activities of the regulators of the Ntr response (see chapter 24). The signal of nitrogen status, the intracellular ratio of α-ketoglutarate to glutamine, is a complex signal that integrates the overall state of metabolism. A high ratio indicates relative nitrogen insufficiency and carbon (and energy) excess; a low ratio is an indicator of nitrogen excess and relative carbon and energy deficiency. It is important to note that relative nitrogen limitation implies carbon and energy sufficiency. Nitrogen limitation results in the phosphorylation of NR_I and expression of Ntr genes; nitrogen excess results in the net dephosphorylation of NR_I-phosphate, which does not activate transcription.

Pleiotropic Defects in Nitrogen Catabolism

Mutations in glnB (P_II), glnD (uridylytransferase/uridydyl-removing enzyme), glnG (NR_I or NtrC), glnL (NR_II or NtrB), or rpoN (ntrA, σ ⁵⁴) result in the inability to utilize a number of nitrogen sources (see chapter 24). Such lesions are pleiotropic because there is insufficient glutamine synthetase to assimilate ammonia during nitrogen-limited growth and because the reduced level of NR_I is insufficient for the expression of many Ntr genes. Mutations in the gltBDF operon, which codes for glutamate synthase, also result in pleiotropic defects because of their effects on the ratio of α-ketoglutarate to glutamine (see chapter 24). The phenotype of lrp mutants is similar to that of gltBDF mutants because Lrp is required for expression of gltBDF (see chapter 24). In summary, a hierarchy of regulators controls the genes of nitrogen source catabolism. Lrp directly controls expression of the gltBDF operon, whose products indirectly control the synthesis and the activity of NR_I. In turn, NR_I regulates the synthesis of specific catabolic enzymes and still more transcriptional regulators.

A number of metabolic lesions can pleiotropically affect nitrogen source utilization. Glutamine synthetase is required for ammonia assimilation during nitrogen-limited growth. Therefore, mutations that diminish glutamine synthetase activity impair the utilization of a variety of nitrogen sources. Two other enzymes are also required for nitrogen-limited growth. One of these enzymes is glutamate synthase. For nitrogen sources such as serine, that do not generate glutamate by transamination or as a product of catabolism, glutamate synthase is necessary for glutamate synthesis (118). The second required enzyme is the glutamine-dependent asparagine synthetase (see chapter 24). Enteric bacteria contain two asparagine synthetases; one requires ammonia as a nitrogen donor, and the second uses glutamine. The glutamine-dependent enzyme is absolutely required during nitrogen-limited growth apparently because the ammonia pool is too low for the ammonia-dependent enzyme.

Certain defects in an entire class of enzymes, the amidotransferases, can also have pleiotropic effects on nitrogen source utilization (see reference 127 for a review of these enzymes). The amidotransferases catalyze the transfer of the amide of glutamine in the synthesis of a variety of compounds. Glutamine is cleaved by one subunit or domain, and the amide is transferred to another. Defects in amide transfer can result in a growth requirement for a high level of ammonia, which can replace glutamine as a substrate. One example of such a phenotype is for some carbamoyl-phosphate synthetase mutants (64). A second example may be S. typhimurium nit mutants, which grow well only with ammonia in the medium (13). These mutants have only 2% of the wild-type level of NAD synthetase, an amidotransferase (B. L. Schneider and L. J. Reitzer, unpublished data).

In the sections that follow, the utilization of specific nitrogen sources is discussed. At the end, mechanisms that coordinate the response to nitrogen deprivation are considered.

AMMONIA

Ammonium ion enters metabolism after its transport through the inner membrane (reviewed in reference 52). Since un-ionized ammonia is small and uncharged, it is thought to be membrane permeable. The pK_a of ammonia is 9.25, which means that at physiological pH, most of the ammonia is protonated. In a medium with 0.2% ammonium sulfate (a common growth condition), ammonium ion and ammonia are both present at a significant concentration, and NH₃ transport may not need a specific transport system. For nitrogen-limited cells, the membrane permeability creates a potential problem because NH₃ can leak out. A comparison of the rate of ammonia assimilation to the calculated rate of diffusion through the membrane (a measure of NH₃ leakage) has suggested that six molecules of ammonium ion must be transported for each molecule assimilated by glutamine synthetase, which would imply a significant energy expenditure to assimilate ammonia during nitrogen-limited conditions (51, 52, 70).

Most bacterial cells appear to contain a specific ammonia carrier, which has an apparent K_m of about 10 μM, about 10-fold lower than the K_m of ammonia for glutamine synthetase (52). Its synthesis requires an intact Ntr response, and glutamine appears to allosterically regulate ammonia transport (41, 42, 52). Transport appears to involve antiport (exchange) with potassium ion, which is driven by an electrochemical potassium gradient (40). The structural genes for the transporter have not been identified, although a gene required for the activity of the ammonia transporter has been identified (28).

l- AND d-ALANINE

l- and d-alanine are excellent sources of nitrogen for E. coli, S. typhimurium, and K. aerogenes. An unusual aspect of alanine as a nitrogen source is that 0.2% alanine induces the Ntr response—even in ammonia-containing minimal medium (L. Reitzer, unpublished observation)—possibly because l- and d-alanine inhibit the activity of glutamine synthetase (see chapter 24). L-Alanine degradation begins with racemization to d-alanine by either of two different racemases; d-alanine is then broken down to pyruvate and ammonia by a membrane-bound d-amino acid dehydrogenase. The reader should consult chapter 24, which discusses the alanine racemases within the context of d-alanine synthesis, and chapter 22, which describes these enzymes and alanine transport systems in relation to the utilization of alanine as a carbon source. l-Alanine can be transported by either the cycA system or the high-affinity leucine-isoleucine-valine transport system, LIV-1. Since Lrp regulates both systems, alanine may be an effector of this control, since Lrp binds alanine (unpublished result in cited in reference 56).

GABA

E. coli K-12 cannot utilize γ-aminobutyrate (GABA) as a nitrogen or carbon source; however, mutants that can degrade GABA as a nitrogen source have been isolated, and secondary mutants derived from those mutants can use GABA as a carbon source (24). In contrast, E. coli B can use GABA as a nitrogen source but not as a carbon source (23). K. aerogenes can degrade GABA as the sole carbon and nitrogen source (34). Despite these differences, these organisms degrade GABA by a common pathway (Fig. 1) that involves transport by a specific permease, removal of the amino nitrogen by glutamate-succinic semialdehyde transaminase, and oxidation of the resulting succinic semialdehyde by succinic semialdehyde dehydrogenase to produce succinate.

Fig. 1The GABA pathway.

Source:

Proteins of the GABA Pathway and Their Functions

E. coli has a highly specific transport system for GABA; mutants with defective transport cannot use GABA as a nitrogen source (47, 76). The transaminase has been partially purified and characterized (4); mutants lacking the transaminase cannot utilize GABA as a nitrogen source (26). E. coli B and other bacteria have two succinic semialdehyde dehydrogenases; one is NADP dependent, and the other can use either NAD or NADP, although NAD is preferred (references 23 and 97 and references cited therein). The isolation of GABA nonutilizers has not resulted in mutants lacking dehydrogenase activity (26, 66).

GABA induces formation of the permease, the transaminase, and both dehydrogenases in E. coli and K. aerogenes (23, 24, 34, 47). GABA is thought to inactivate a specific repressor, the product of gabC; gabC mutants do not require specific induction (24). The true inducer for the NAD/NADP dehydrogenase appears to be succinic semialdehyde, a product of the catabolism of GABA, or p-hydroxyphenylacetate, a utilizable carbon source for E. coli and an intermediate in the degradation of aromatic compounds (23). Expression of the GABA genes also requires either carbon limitation or activation by the Ntr system (34, 126).

The Genes of GABA Catabolism

The gabDTP operon, specifying the dehydrogenase, transaminase, and permease, respectively, and gabC, encoding the putative repressor, have been cloned (4, 65, 76). The gene sequence was definitively established only after the whole gabDTP operon had been sequenced (4, 76). The gabC gene, encoding the putative repressor, has not been sequenced, although it is upstream of gabD (65). The promoters and transcripts of these genes have yet to be characterized.

ARGININE, AGMATINE, PUTRESCINE, AND ORNITHINE

Chapter 22 covers utilization of arginine, agmatine, putrescine, and ornithine as carbon sources and also discusses arginine transport systems. This section considers utilization of these compounds as nitrogen sources in E. coli and K. aerogenes. The ability to use agmatine as the sole nitrogen source is strain dependent for E. coli (see reference 98 and unpublished results cited on p. 336 of reference 21). Although S. typhimurium can utilize arginine as a nitrogen source, it cannot degrade agmatine, ornithine, or putrescine (36).

Pathways of Arginine Catabolism

The variety of potential arginine catabolic pathways is surprisingly large, and multiple pathways exist in a single organism (reference 21 provides an extensive review of these pathways). In the enteric bacteria, two pathways contribute to arginine catabolism. The first pathway is initiated by arginine decarboxylase (ADC) and agmatine ureohydrolyase, which degrades arginine to agmatine and then to putrescine (Fig. 2). Putrescine is converted to GABA after transamination and dehydrogenation reactions; GABA is then degraded as described elsewhere in this review. Succinylation of arginine initiates a second pathway of arginine catabolism, called the arginine succinyltransferase (AST) pathway (Fig. 3). Following succinylation, degradation of various succinylated intermediates produces glutamate and succinate.

Fig. 2The arginine decarboxylase-initiated pathway of arginine, agmatine, putrescine, and ornithine catabolism.

Source:

Fig. 3The arginine succinyltransferase pathway of arginine degradation. Succinyl-CoA, succinyl coenzyme A.

Source:

ADC-Initiated Catabolism: Mutants, Genes, and Proteins

Analysis of mutants suggests that ADC-initiated catabolism participates in catabolism of arginine, agmatine, ornithine, and putrescine in E. coli: speA (ADC-deficient) mutants grow at half the rate of a wild-type strain with arginine as a sole nitrogen source, and speB (agmatine ureohydrolase-deficient) mutants grow poorly with either arginine or agmatine as sole nitrogen source (104). A slight variation of this pathway is used to degrade ornithine. Ornithine is decarboxylated by ornithine decarboxylase to form putrescine, which is subsequently degraded as described above. E. coli mutants lacking this enzyme grow very slowly with ornithine as a nitrogen source (104). Some strains deficient in putrescine aminotransaminase and the subsequent enzyme grow less well with arginine, agmatine, putrescine, or ornithine as a nitrogen source; however, other mutants with low levels of these enzymes grow well with these compounds (104, 105). Presumably, the glutamate generated by transamination prior to GABA formation is sufficient for utilization of these compounds as nitrogen sources.

The E. coli genes for speA, speB, and speC, which specify ADC, agmatine ureohydrolase, and ornithine decarboxylase, respectively, have been cloned, although only the first two have been sequenced; all three are linked (11, 68, 113). The products of these genes have been purified (1, 99, 125). Arginine decarboxylase is at least partially located in the periplasm (15). The pat and prr genes of E. coli, which code for putrescine aminotransferase and pyrroline dehydrogenase, the fourth and fifth reactions of this pathway, have been mapped but not cloned (104). Neither enzyme has been purified, although putrescine transaminase in E. coli and K. aerogenes has been partially characterized (34, 50).

ADC-Initiated Catabolism: Regulation and Function

ADC is found only in those organisms in which ornithine, a precursor of putrescine, is not an intermediate in arginine catabolism (21, 108, 115). Therefore, it was suggested that the primary function of this pathway is to synthesize putrescine for cells grown in media with arginine, which inhibits and represses enzymes of putrescine synthesis from ornithine (115). Such a role is consistent with the factors that control ADC activity. ADC is feedback inhibited by putrescine and spermidine (125); however, to the extent that ADC is periplasmically localized, this inhibition may not be significant. ADC is also inhibited by three proteins, whose concentrations increase as polyamine levels increase (16, 85).

Some evidence suggests that ADC does not initiate arginine catabolism in E. coli or K. aerogenes. First, the synthesis of one of the proteins that inhibits ADC activity may be under Ntr control (16), which suggests that nitrogen limitation inhibits ADC activity. Second, nitrogen limitation does not affect ADC synthesis (105). Third, while speA (ADC-encoding) mutants grow less well with arginine, this phenotype could result from putrescine deficiency and therefore, does not imply that ADC has a catabolic function.

The regulation of ornithine decarboxylase is very similar to that of ADC: it is not an Ntr enzyme, and polyamines and spermidine inhibit its activity (1, 105).

In contrast, all of the enyzmes that degrade agmatine to succinate in E. coli are Ntr enzymes (105), which suggests that they have a catabolic function. In addition to this regulation, synthesis of agmatine ureohydrolase requires specific induction by agmatine (105, 114). The mechanism of this induction is somewhat unusual: speA and speB genes form a complex operon; speA is not under Ntr control, whereas speB is. Most transcription of speB initiates at the speA promoter, although a promoter for speB is present in the intercistronic region (113). Agmatine induces expression from the speB promoter (114); the mechanism of activation during nitrogen limitation has not been explored.

The regulatory pattern is somewhat different in K. aerogenes, which uses this pathway for degradation of agmatine, putrescine, and GABA but not for arginine (33, 34). With the exception of ADC, high-level expression requires sequential induction by intermediates in the pathway and, for at least two of these enzymes, induction by the Ntr system. This complex pattern of sequential induction ensures that appropriate portions of the pathway are expressed when needed.

In summary, the evidence described above suggests that the potential pathway of arginine degradation via GABA degrades agmatine, putrescine, and ornithine but not necessarily arginine.

Role of the AST Pathway in Arginine Catabolism

The AST pathway has been detected in K. aerogenes (108), E. coli, and S. typhimurium (Schneider and Reitzer, unpublished data). In these organisms, at least three enzymes of the AST pathway are nitrogen regulated (Schneider and Reitzer, unpublished data). Such a conclusion is verified if earlier results are reinterpreted to take into account that the arginine-inducible N ²-acetylornithine 5-aminotransferase of K. aerogenes (31) prefers succinylornithine as a substrate (108), which implies that it is the third enzyme of the AST pathway. It clearly is an Ntr enzyme (31). In both organisms, arginine or ornithine appears to be required for induction of the AST enzymes (31; Schneider and Reitzer, unpublished data). Mutants defective in enzymes of this pathway have yet to be isolated. Nonetheless, plasmid-bearing cells which have 10-times-higher concentrations of the AST enzymes grow twice as fast as isogenic plasmidless strains with arginine as a nitrogen source (Schneider and Reitzer, unpublished data). These results strongly suggest a catabolic function for the AST pathway.

Enzymes and Genes of the AST Pathway: the Arginine-Inducible Succinyl[acetyl]ornithine δ-Transaminase and the Enigmatic argM Locus of E. coli

With one exception, the enzymes of the AST pathway have not been characterized beyond their assay from crude extracts. The exception is the arginine-inducible succinyl[acetyl]ornithine δ-transaminase, which has been purified from E. coli and K. aerogenes (30, 31). In both organisms, this enzyme has been shown to prefer succinylornithine as a substrate (21, 108), but it can also use ornithine (8, 31). The arginine-inducible transaminase in E. coli was first discovered because mutants with high levels of this enzyme suppressed a defect in argD, which codes for the biosynthetic acetylornithine δ-transaminase (reference 95 and references cited therein).

A Clark-Carbon plasmid specifying succinyl[acetyl]ornithine δ-transaminase has apparently been identified (95); the insert was shown to map to min 38 of the E. coli chromosome (77). This map location creates an interesting problem because mutations that suppressed argD defects have been shown to map at min 87 at a locus called argM (cited in reference 95). The argM locus does not code for either the biosynthetic or the catabolic transaminases, since their genes are at min 74 and 38, respectively (77, 95). As described above, synthesis of the catabolic transaminase is nitrogen regulated in E. coli and K. aerogenes. Because mutations in glnL of the glnALG operon, which like argM maps to minute 87, can result in constitutive expression of Ntr genes, argM may be glnL.

AROMATIC AMINO ACIDS

E. coli can use tryptophan as a carbon source (chapter 22), but not as a nitrogen source in glucose-containing medium, because tryptophanase, the only known enzyme of tryptophan degradation, is very sensitive to catabolite repression. In contrast to E. coli, K. aerogenes can use tryptophan, phenylalanine, or tyrosine as the sole nitrogen source (86). In K. aerogenes, tryptophan degradation involves transport by a specific permease and transamination by a multispecific aromatic amino acid aminotransferase to produce glutamate and indolepyruvate, which is not metabolized (86). The aminotransferase, which has been purified, can use phenylalanine, tryptophan, and probably tyrosine as amino donors (87). Mutational loss of this enzyme results in the inability to utilize the aromatic amino acids (86). Synthesis of the tryptophan transport system is under Ntr control, whereas synthesis of the aminotransferase is constitutive (86).

ASPARAGINE

Asparagine as a Nitrogen Source

For E. coli and S. typhimurium, asparagine is one of the better nitrogen sources (13; however, see reference 122). Glutamate synthase-deficient (gltB) strains of E. coli, which are considered Ntr^–, can utilize aspartate, but not asparagine, as a sole nitrogen source (84), which implies that an essential component of asparagine regulation in E. coli is nitrogen regulated (see chapter 24). However, the Ntr component has yet to be identified.

The Proteins of Asparagine Catabolism and Their Functions

E. coli and K. aerogenes contain two asparaginases: an high-affinity periplasmic enzyme and a low-affinity cytoplasmic enzyme (17, 93); S. typhimurium has a periplasmic enzyme and presumably also contains a cytoplasmic enzyme (44). Because of their localization, there are two distinct pathways of asparagine utilization: asparagine transport, followed by hydrolysis in the cytoplasm; or periplasmic degradation, followed by transport of aspartate instead of asparagine.

Mutants with a low level of the cytoplasmic enzyme but a normal level of the periplasmic asparaginase cannot use asparagine as a sole nitrogen source (22, 122), which implies that the cytoplasmic asparaginase degrades asparagine as a sole nitrogen source. The synthesis of the cytoplasmic enzyme is considered constitutive (45). The factors that control synthesis of the periplasmic enzyme in E. coli and S. typhimurium imply that its primary function is generation of fumarate as an electron acceptor during anaerobic respiration (46, 107). In K. aerogenes, nitrogen limitation, not anaerobiosis, induces synthesis of the periplasmic asparaginase (93). A K. aerogenes mutant lacking this periplasmic enzyme grows somewhat more slowly with asparagine as a sole nitrogen source, especially with a low concentration of asparagine (93).

The periplasmic asparaginases from E. coli and K. aerogenes have been purified (38, 93). The genes for both E. coli enzymes have been cloned and sequenced (9, 45); homology between the enzymes is limited (45). The E. coli gene specifying the cytoplasmic asparaginase, ansA, is part of a bicistronic operon; the downstream gene codes for a protein of unknown function (45). The transcripts and promoters have not been analyzed. High- level expression of ansB, which codes for the periplasmic asparaginase, requires the cyclic AMP receptor protein and the product of the fnr gene, FNR (43), which accounts for catabolite repression and induction by anaerobiosis.

E. coli grown in an ammonia-containing (nitrogen-rich) medium has two kinetically distinct components for asparagine transport (123); asparagine transport has not been examined for cells grown in a nitrogen-limited medium.

ASPARTATE AND GLUTAMATE

Aspartate and glutamate can be used as sole nitrogen sources for growth of E. coli, K. aerogenes, and S. typhimurium (13, 48, 118); however, their catabolic pathways have not been defined. Although growth with aspartate as the sole nitrogen source induces the Ntr response, Ntr^– glnG mutants grow normally with aspartate (83), which suggests that essential components of aspartate catabolism are not necessarily Ntr regulated. An analysis of the products of aspartate catabolism from nitrogen-limited E. coli suggested the possibility that arginine is an intermediate in aspartate degradation (35). Consistent with this suggestion is the observation that cells with a plasmid that increases levels of arginine catabolic enzymes grow faster with aspartate as the sole nitrogen source (Schneider and Reitzer, unpublished data).

Transport

Schellenberg and Furlong characterized five distinct transport systems for glutamate and aspartate in E. coli W (100); a sixth system that operates during anaerobiosis and transports dicarboxylic acids, including aspartate, was recently discovered (27). The recent cloning of genes encoding components of two of these systems has provided the means to analyze the expression and functions of these systems (reference 116 and references therein); however, such studies have yet to be performed. At least two of these systems have been implicated in the use of aspartate or glutamate as a nitrogen source (48, 54).

Aspartate- and Glutamate-Catabolizing Enzymes

Aspartase catalyzes the degradation of aspartate to fumarate and ammonia and would be an obvious enzyme to degrade aspartate as a nitrogen source. However, aspartase is severely repressed when aspartate is the sole nitrogen source in E. coli or K. aerogenes (25; Reitzer, unpublished data), and aspartase-deficient strains grow as well as wild-type strains with aspartate or glutamate as a nitrogen source (Schneider and Reitzer, unpublished data). Aspartase has two functions: to provide fumarate, an electron acceptor, during anaerobic respiration (46, 107) and to degrade glutamate, and presumably aspartate, as a carbon source in E. coli (60).

There is no correlation between the ability to utilize glutamate as a carbon source and the level of glutamate decarboxylase, which suggests that this enzyme is not a glutamate catabolic enzyme (61). Glutamate dehydrogenase does not have a catabolic function because glutamate represses its synthesis and gdhA mutants grow normally with glutamate as a nitrogen source (see chapter 24).

GLUTAMINE

Glutamine is a signal of nitrogen sufficiency and, not surprisingly, supports rapid growth as a sole nitrogen source. However, as a sole nitrogen source, glutamine induces the Ntr response, probably a result of rapid intracellular glutamine catabolism, which in turn causes a low ratio of glutamine to α-ketoglutarate.

Transport

At least two independent systems, distinguishable by their affinity for glutamine, transport glutamine into E. coli (120, 121), S. typhimurium (7), and K. aerogenes (Reitzer, unpublished data). The E. coli systems have been the most intensively studied. An osmotic shock-sensitive, high-affinity system contains a periplasmic binding protein which is very specific for glutamine (120, 121). The glnHPQ operon, which codes for proteins of the high-affinity system, has been mapped and cloned, and its sequence has been determined (62, 79).

Nitrogen limitation induces expression of the high-affinity system in all three enteric organisms (7, 121; Reitzer, unpublished data). Transcription of the glnHPQ operon can initiate from two promoters, an Ntr promoter and a constitutive promoter (78). NR_I and integration host factor are required for expression of the Ntr promoter (18). Most studies of glnHPQ expression have involved mutants which can utilize glutamine as a carbon source. These mutants contain 2.5 to 6 times more of the glutamine-binding protein than a wild-type strain and have inexplicably acquired the ability to use glutamate as a carbon source (62, 120). These complications have not been taken into account in analyses of regulation.

The low-affinity glutamine transport system has not been well characterized because it is detectable only in the absence of the high-affinity system (120, 121; Reitzer, unpublished data). Glutamate inhibits its activity (120), which suggests that it might be one of the glutamate uptake systems. K. aerogenes mutants that have lost the high-affinity system grow as well as wild-type strains with 0.2% glutamine as a nitrogen source, which suggests that the low-affinity system can be the major system for glutamine uptake if the glutamine level is sufficiently high (Reitzer, unpublished data).

Enzymes of Glutamine Catabolism

Mutants deficient in glutamate synthase δ grow slowly with glutamine as the sole nitrogen source (see chapter 24), which suggests that glutamine synthase is the primary enzyme of glutamine catabolism. E. coli also contains two glutaminases, which could conceivably degrade glutamine; the glutaminases are more thoroughly described by McFall (in chapter 22). Since mutants deficient in either glutaminase have yet to be isolated and neither enzyme is under Ntr control, it is difficult to argue strongly for a function for either enzyme when glutamine is the nitrogen source.

HISTIDINE UTILIZATION AND REGULATION BY NAC

Histidine can serve as the sole source of carbon or nitrogen for K. aerogenes but not for either E. coli or S. typhimurium. Two activators of transcription, NR_I and NAC (nitrogen assimilation control), are required for histidine utilization in K. aerogenes during nitrogen-limited growth. E. coli and S. typhimurium cannot use histidine as a nitrogen source because the former lacks the genes of histidine utilization and the latter lacks NAC (5, 6). This section focuses on the role of NAC in histidine catabolism in K. aerogenes; chapter 22 and a review by Magasanik (58) consider the components of the hut system in more detail; chapter 24 considers NAC-dependent control of glutamate dehydrogenase and glutamate synthase; and the end of this chapter discusses the function of NAC during nitrogen-limited growth.

Four enzymes degrade histidine to glutamate, formamide, and ammonia. The genes for these enzymes are contained within two adjacent operons: hutIGC, which specifies two degradative enzymes and the HutC repressor, and hutUH, which encodes the other two catabolic enzymes. The two operons are transcribed in the same direction. Expression requires inactivation of the HutC repressor by urocanate, the product of the first reaction of the sequence, and growth in either a carbon-limited or nitrogen-limited medium. The binding site for the repressor in the single hutUH promoter has been determined (82).

Transcription of hutUH during nitrogen-limited growth requires NR_I and NAC. NR_I activates indirectly by stimulating expression of the nac gene from a σ ⁵⁴-dependent promoter (57, 102). NAC then initiates transcription of hutUH from the same start site as that used during carbon-limited growth, which implies that σ ⁷⁰-RNA polymerase, rather than σ54-RNA polymerase, initiates NAC-dependent transcription (5). This conclusion is consistent with the observation that NAC is homologous to LysR, which activates σ ⁷⁰-dependent promoters (102). The mechanism of NAC-dependent activation is under investigation (80, 81). Unlike NR_I activity, NAC activity is not regulated by nitrogen availability (103); the effects of other environmental factors on NAC activity have not been examined.

THREONINE, GLYCINE, AND l-SERINE

Under appropriate conditions, i.e., in leucine-containing media, threonine, glycine, and l-serine can be sole sources of nitrogen for wild-type E. coli and K. aerogenes. These amino acids are metabolically related because threonine is first degraded to glycine, then to l-serine, and finally to pyruvate and ammonia (Fig. 4). Tracer studies provided the first evidence for this pathway (72, 90). Even though growth with serine has been diagnostic of the Ntr response, components of this pathway are regulated not by the Ntr system but by the leucine-responsive protein Lrp, a global regulator of gene expression (chapter 94). This pathway of threonine catabolism can also be used as an energetically efficient pathway for the synthesis of C₁ units in the form of N⁵, N¹⁰-methylenetetrahydrofolate (91).

Fig. 4Degradation of threonine, glycine, and serine. CoA, coenzyme A.

Source:

Threonine

Genetic evidence suggests that threonine is degraded by the NAD-dependent threonine dehydrogenase, specified by tdh. A tdh null mutant cannot synthesize serine if the serine biosynthetic pathway from 3-phosphoglycerate pathway is mutationally inactivated (91). The second enzyme of the pathway, 2-amino-3-ketobutyrate coenzyme A ligase, cleaves 2-amino-3-ketobutyrate to glycine and acetyl coenzyme A. Both the dehydrogenase and the ligase have been purified (10, 69). Synthesis of threonine dehydrogenase is induced by leucine but is not affected by threonine (73, 90, 119). Lrp-dependent repression, which is alleviated by leucine, controls expression of tdh (94, 117). The genes for these enzymes comprise the tdh-kbl operon (2, 92); therefore, although regulation of ligase synthesis has not been examined, it is probably controlled in parallel with the dehydrogenase.

Glycine

The glycine cleavage system (GCV) is the third component of threonine degradation and the first component of glycine degradation. GCV cleaves glycine to NH₃, CO₂, and C₁ (as N ⁵,N ¹⁰-methylenetetrahydrofolate). GCV-deficient mutants cannot utilize threonine as a nitrogen source (90) or glycine as a nitrogen source (72) and cannot convert threonine or glycine to serine in mutants blocked in the other route of serine synthesis (89, 91). GCV is a four-protein complex in species from bacteria to mammals (49; reference 109 and references therein). In E. coli, three components are encoded within the gcvTHP operon (109), while the fourth, encoded by lpd, is also a component of the pyruvate and α-ketoglutarate dehydrogenase complexes (110). GCV activity is moderately induced by glycine and by entry into stationary phase; it is repressed by purines (major products of C₁ synthesis) (63, 124). Lrp is an essential activator of the gcvTHP operon, but leucine does not modulate this expression (55). GcvA, a LysR-like activator, mediates the slight induction by glycine and is partially responsible for repression by purines (124). Expression of gcvTHP does not require GcvA.

Serine hydroxymethyltransferase, the fourth enzyme of the pathway and the product of glyA, catalyzes the reversible condensation of a C₁ unit with glycine to form serine; the reaction is usually discussed in the direction of C₁ and glycine synthesis. The phenotype of glyA mutants is similar to that of GCV-deficient strains: neither can use threonine or glycine as a nitrogen source (72, 90) or threonine as a source of serine (91). This enzyme is moderately repressed by glycine, even when glycine is the sole source of nitrogen (72). Beyond this observation, nothing is known about the control of serine hydroxymethyltransferase activity during nitrogen-limited growth.

Serine

Multiple enzymes can catalyze the deamination of l-serine to pyruvate and ammonia. E. coli contains two serine deaminases, encoded by the sdaA and sdaB genes; only the former is expressed in minimal medium (111, 112). Although sdaA and sdaB mutants have been isolated, their ability to utilize various nitrogen sources has not been examined. Nitrogen limitation results in a modest increase (70%) of l-serine deaminase activity in E. coli but has no effect on activity in K. aerogenes (119). However, since E. coli grown in an ammonia-containing (i.e., nitrogen-rich) medium adapts immediately to growth with serine as a nitrogen source, the uninduced level of serine deaminase appears to be sufficient for growth (74). Leucine and glycine additively increase transcription of sdaA (39, 111). Leucine reverses Lrp-dependent repression (55, 117). Nothing is known about glycine-dependent induction. In addition to these effectors, threonine also induces, but only for K. aerogenes, not for E. coli, and only in carbon-limited media (39, 119). Expression of sdaA by DNA-damaging agents (71) and a complex posttranslational activation of serine deaminase activity (75) suggest functions other than nitrogen source catabolism.

The metC product, cystathionine β-lyase, probably has serine deaminase activity as a side reaction and may contribute to serine deamination in vivo (14). Growth with serine as the nitrogen source is significantly enhanced by methionine, which may overcome the serine-dependent inhibition (Schneider and Reitzer, unpublished data). Another serine-deaminating enzyme is a nonspecific isoleucine-sensitive threonine dehydrase in K. aerogenes (119).

Overview

In summary, Lrp, not components of the Ntr system, controls the formation of enzymes of this pathway. In addition to these catabolic proteins, Lrp also appears to control a serine transport system (37). There is no unifying mechanism for Lrp-dependent control: Lrp represses tdh-kbl and sdaA and leucine overcomes this repression, and Lrp activates gcvTHP. A mechanism of sequential induction is consistent with the possibility that these enzymes function as a pathway. The two enzymes of threonine degradation are induced by leucine and produce glycine. The enzymes of glycine and serine catabolism require specific induction by leucine, glycine, or both. Therefore, induction of the enzymes of threonine catabolism may result in elevated pools of glycine and subsequent induction of the enzymes of glycine and serine catabolism.

UREA

Urea can be utilized by wild-type species of Klebsiella and clinical isolates of E. coli and S. typhimurium. Urease, which degrades urea to ammonia and CO₂, has been implicated as a virulence factor in a number of diseases (19, 67). Urease from Klebsiella species is a complex heterotrimeric nickel-containing protein (19). There are two types of urease genes; plasmid-encoded genes are found in the clinical isolates, and chromosomal genes are present in K. aerogenes. The genes for urease, whether plasmid encoded or chromosomal, form a seven-gene cluster, which is probably expressed as an operon; four of the genes appear to be required for assembly (19, 20, 88).

There are two basic strategies of the regulation of urease formation. For K. aerogenes, induction requires nitrogen limitation and NAC (5, 20, 32). Such expression is not regulated by urea or nickel availability (20, 32). Expression of plasmid-encoded urease genes requires induction by urea and activation by the UreR protein, an AraC-like protein (19); these ureases are not under Ntr control.

COORDINATION OF THE RESPONSE TO NITROGEN-LIMITED GROWTH

There is no unifying theme in the regulation of specific nitrogen catabolic proteins. Synthesis of some of these proteins requires NR_I or another Ntr activator, while others require Lrp; some require specific induction, whereas others do not. Even if only Ntr genes are considered, at least four factors affect their expression. One important variable is the level of activator needed for expression. A low concentration of NR_I is sufficient to activate the glnALG and glnHPQ operons; higher concentrations are required for expression of nac, nifA, and genes of arginine utilization (3, 5, 53, 83). Additional fine tuning is brought about by the differential sensitivity of NAC-regulated Ntr genes to levels of NAC (103). This implies that during the transition to nitrogen-limited growth, when the concentration of NR_I is increasing, some Ntr operons are expressed before others.

A second variable that modulates the expression of Ntr genes is the degree of phosphorylation of NR_I (101), which can be phosphorylated not only by NR_II but also by high-energy compounds such as acetyl-phosphate that can directly phosphorylate NR_I (29).

A third factor that contributes to Ntr gene expression is negative autogenous control by Ntr regulators. High concentrations of NR_I reduce transcription of the gene coding for NR_I (106). It was estimated that this inhibition results in a twofold reduction of NR_I. Expression of nac is also negatively autogenously regulated: NR_I activates nac expression, and NAC downwardly modulates this expression (6).

The fourth factor that affects individual Ntr genes is regulation by secondary transcriptional regulators such as NifA and NAC. Their function appears to be to provide additional regulatory capabilities. NifA is inactivated by NifL, which appears to be an oxygen sensor (96). The regulatory consequences of NAC-dependent control are less obvious. It was suggested that NAC provides signal amplification (5). However, NAC-dependent genes require a high level of NR_I, and once expressed, these genes are not responsive to nitrogen availability (32, 103). Instead, NAC-dependent control diminishes responsiveness and may provide a form of short-term memory that may be competitively advantagous if a previous environment is reencountered.

In summary, the regulators of genes coding for proteins of nitrogen catabolism provide for a flexible and coordinated response to nitrogen deprivation.

ACKNOWLEDGMENTS

My work in this area has been funded by Public Health research grant GM38877 from the National Institute of General Medical Sciences. I am grateful to Jeri Trites for help in preparation of the manuscript.