Molecular Basis for Bacterial Growth on Citrate or Malonate

P. Dimroth

[SECTION EDITOR, AUGUST BÖCK]

Posted July 6, 2004

Institute of Microbiology, ETH Zurich, Schmelzbergstrasse 7, CH-8092 Zurich, Switzerland

Phone: +41-1-632 33 21, Fax: +41-1-632 13 78, Email: This e-mail address is being protected from spambots. You need JavaScript enabled to view it

The ability to use citrate or malonate as a carbon and energy source is widespread among bacteria, which is not surprising since these compounds are abundant in the environment, where they originate either from natural sources or from industrial production. Bacterial growth on either of these compounds is widely used as a diagnostic criterion to differentiate between genera and species of enterobacteria. In this chapter we will describe the catabolism of citrate and malonate by various enterobacteria under aerobic and anaerobic conditions. We will focus on the genetic equipment and on the proteins involved in the different degradation pathways. As we will see, the citrate- and malonate-degrading enzymes share remarkable biochemical properties which prompted us to discuss these in a common chapter.

BACTERIAL GROWTH ON CITRATE

Under aerobic conditions, citrate can be directly fed into the tricarboxylic acid cycle and only a citrate transporter is required in addition to the basic enzyme equipment of the cell. Hence, the well-known diagnostic test of aerobic growth on Simmon’s citrate medium (123) to differentiate between enterobacterial species and strains is directly related to the presence or absence of a citrate transporter. Salmonella and Klebsiella strains grow readily under these conditions because citrate transporters are synthesized. Most Escherichia coli strains, however, do not synthesize a citrate transporter aerobically, and no growth on citrate is observed (88).

For anaerobic growth on citrate, bacteria require a specific enzyme equipment for citrate degradation, in addition to a citrate carrier. Distinct pathways have evolved in clostridia, lactic acid bacteria, phototrophic bacteria, and enterobacteria, which are described in several reviews (2, 13, 49, 72). The most comprehensive picture for citrate fermentation is available for Klebsiella pneumoniae, in which all of the citrate fermentation enzymes have been well characterized biochemically. The corresponding genes have also been identified, and insight into the regulation of their expression is available. We will therefore focus here on the fermentation of citrate by this organism. Deviations from this prototype in other bacteria, particularly in E. coli and Salmonella enterica serovar Typhimurium, will be mentioned when appropriate.

An overview of citrate fermentation in K. pneumoniae is shown in Fig. 1 (18, 27, 28). After uptake into the cell by the Na⁺-dependent citrate carrier CitS (50, 104, 135), citrate is cleaved by citrate lyase into acetate and oxaloacetate (22, 126). The latter is subsequently converted to CO₂ and pyruvate by the membrane-bound oxaloacetate decarboxylase sodium ion pump (35, 36, 39, 128). These three reactions are specific for citrate fermentation, whereas the subsequent steps of pyruvate degradation are common to all fermentation pathways generating pyruvate as an intermediate (86). Pyruvate is converted to acetyl-coenzyme A (acetyl-CoA) and formate by pyruvate formate lyase, and part of the formate is cleaved to CO₂ and H₂ by formate hydrogen lyase (26). Acetyl-CoA is converted to acetyl phosphate by phosphotransacetylase, and acetate kinase converts acetyl phosphate and ADP to acetate and ATP.

Fig. 1Pathway of citrate fermentation in K. pneumoniae.

Also of interest in this pathway is the synthesis of NAD(P)H for biosynthetic reactions. While most bacteria growing fermentatively are faced with the problem of getting rid of the reducing equivalents (mainly NADH) formed by the oxidation of the growth substrates, citrate fermentation by K. pneumoniae includes no oxidative steps and NADH is not generated. However, as the mean oxidation status of citrate is above average for cellular components, the cells require reducing equivalents for citrate assimilation. It has been shown that hydrogen generated in the formate hydrogen lyase reaction (18) provides the reducing equivalents for NAD(P)H formation from NAD(P) by a soluble or a membrane-bound hydrogenase (129).

Overall, the citrate fermentation pathway produces 1 mol of ATP per mol of citrate by substrate-level phosphorylation in the acetate kinase reaction, and in addition, an electrochemical gradient of Na⁺ ions is generated by the oxaloacetate decarboxylase (35, 41). The chemical concentration gradient of Na⁺ is used to drive citrate uptake by the electroneutral H-citrate^2– (Hcit^2–)/Na⁺/H⁺ symporter (CitS) (104), and the electrical component of the ΔμNa⁺ is assumed to contribute driving force for ATP synthesis by the H⁺-translocating F₁F₀ ATP synthase, which is constitutively expressed in this organism. This type of ATP synthesis is termed decarboxylation phosphorylation (65). This is the only ATP-generating reaction in Propionigenium modestum, which grows from the fermentation of succinate to propionate and CO₂ (113). In the course of this fermentation, a methylmalonyl-CoA decarboxylase Na⁺ pump generates an electrochemical gradient of Na⁺ ions (63, 64) which serves as the driving force for ATP synthesis by an Na⁺-translocating F₁F₀ ATP synthase (89).

The same citrate fermentation pathway as in K. pneumoniae seems to operate in S. enterica serovar Typhimurium (100, 138). Both organisms are able to grow anaerobically on citrate as the sole carbon and energy source, because they dispose of an Na⁺-dependent citrate carrier (73, 85), citrate lyase, and an oxaloacetate decarboxylase Na⁺ pump under these conditions. The oxaloacetate decarboxylase has been purified and shown to have properties very similar to those of the enzyme from K. pneumoniae (138). The genes encoding oxaloacetate decarboxylase are part of the citrate fermentation gene cluster. In addition, there is a second copy of the oxaloacetate decarboxylase-encoding genes located downstream of the genes for an oxygen-labile l-tartrate dehydratase (140, 142). This enzyme converts l-tartrate into oxaloacetate, representing the first step of tartrate fermentation (107, 110). Oxaloacetate is then catabolized by the same pathway as in citrate fermentation (Fig. 1).

In contrast to K. pneumoniae and S. enterica serovar Typhimurium, E. coli is not able to grow anaerobically with citrate as the sole source of carbon and energy. This is due to its lack of oxaloacetate decarboxylase and the repression of 2-oxoglutarate dehydrogenase under anoxic conditions (1, 127). E. coli has been shown, however, to degrade citrate anaerobically in the presence of a cosubstrate such as glucose, lactose, pyruvate, or glycerol (94). Citrate is taken up under these conditions by the citrate carrier CitT, which is a member of the 2-oxoglutarate/malate translocator family and functions as a citrate/succinate antiporter (105). As in the other enterobacteria, citrate is degraded by citrate lyase to acetate and oxaloacetate (106). However, due to the lack of oxaloacetate decarboxylase, the oxaloacetate cannot be decarboxylated and is instead reduced to succinate. The reducing equivalents for these reactions are obtained from the oxidation of the cosubstrate (94). For this type of citrate fermentation, the uptake of citrate by the citrate/succinate antiporter appears to be particularly useful: succinate is produced from citrate in a 1:1 ratio, and the succinate gradient from the inside to the outside is used to drive the active transport of citrate into the cells.

Organization of the Genes

Figure 2 shows physical maps of the citrate fermentation genes in enterobacteria (11, 56). In K. pneumoniae the citrate fermentation genes are clustered as two operons (citS and citC operon) with divergent orientation (14). The citS operon includes the genes for the Na⁺-dependent citrate carrier (citS) (135), the three structural genes for oxaloacetate decarboxylase (oadGAB) (90, 121, 122, 141), and the two-component regulatory system (citAB) (14), where citA codes for the sensor kinase and citB codes for the response regulator (15). The genes of the citC operon encode citrate lyase ligase (citC), the three subunits of citrate lyase (citDEF), and an enzyme involved in the biosynthesis of the 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group of citrate lyase (citG) (14, 118). Interestingly, upon expression of the cit genes from K. pneumoniae in E. coli, catalytically active citrate lyase was not formed (14). Instead, an apo form of the enzyme was synthesized which lacked the phosphoribosyl dephospho-CoA prosthetic group (118). This indicates that an additional gene is required for the synthesis and/or attachment of the prosthetic group to the apoprotein. Evidence for the missing gene was obtained from comparison with the cit gene clusters from E. coli, S. enterica serovar Typhimurium, and Haemophilus influenzae. The E. coli cit operon contains an additional gene between citF and citG which was termed citX. The operon also includes the gene for a citrate transporter (citT) located downstream of citG. In divergent orientation to the citCDEFXGT gene cluster are the genes for the sensor kinase (citA) and the response regulator (citB). No citS or oadGAB genes are present in the E. coli genome, in accordance with the lack of a sodium-dependent citrate transporter and an oxaloacetate decarboxylase, respectively (11). The genome of serovarTyphimurium contains a cluster of citrate fermentation genes that is similar to that in K. pneumoniae but includes the citX gene between citF and citG (95). Another variation is found in H. influenzae. The citCDEF gene cluster continues with citX, which is fused with the citG gene. Downstream of citG is the gene for a citrate transporter (56).

Fig. 2Physical maps of the citrate fermentation genes in K. pneumoniae (K.p.), S. enterica serovar Typhimurium (S.t.), and E. coli (E.c.). The function of the encoded proteins is indicated. The genes having counterparts in K. pneumoniae are shown in dark gray; those without are indicated in light gray.

Conclusive evidence that the citX gene is required for the expression of a functional citrate lyase was obtained when citX from E. coli was coexpressed on a compatible plasmid with the citCDEFG gene cluster from K. pneumoniae (118). These cells synthesized holo-citrate lyase with its 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group, which was fully functional. Likewise, the homologous expression of the plasmid-encoded citCDEFXG genes of E. coli led to the formation of a functional citrate lyase (118). According to these results, a citX homologous gene had to be present on the K. pneumoniae genome at a location separate from the citCDEFG gene cluster. The citX gene was identified and shown to yield functional holo-citrate lyase when coexpressed with the other cit genes of K. pneumoniae on a compatible plasmid (120). Remarkably, citX clustered with three genes which presumably are also involved in citrate fermentation. One of these, citW, was shown to encode a citrate carrier (84), the third one to be identified on the K. pneumoniae genome in addition to citS (135) and citH (134). The two other genes clustered with citX were named citY and citZ (120). They encode a two-component regulatory system with high homology to the citA/citB system, which is essential for the expression for the citC and citS operons (15). A citZ insertion mutant could still grow on citrate anaerobically, and the citX content was comparable to that of the wild type (120). This indicates that the citY/citZ system is not essential for citrate fermentation, but a role of this two-component system in fine-tuning the expression of citrate fermentation genes in the natural environment is an attractive possibility to be addressed in further studies.

Proteins Involved in Citrate Fermentation

Citrate Carriers.

K. pneumoniae cells can produce three different citrate carriers, CitH, CitS, and CitW (50, 84, 121, 134, 135). CitH is responsible for citrate uptake under oxic growth conditions. The transported species is Hcit^2–. The transport is Na⁺ independent and driven mainly by ΔpH and to a minor extent by Δψ. Accordingly, a symport of Hcit^2– with three protons was proposed (133). As deduced from the gene sequence, CitH is a 48-kDa protein composed of 12 membrane-spanning segments and a central hydrophilic loop (134).

Under the conditions of citrate fermentation, the Na⁺-dependent citrate carrier CitS is synthesized (50, 79). CitS is a member of the 2-hydroxycarboxylate transporter (2-HTC) family, which are specific for substrates containing the 2-HTC motif like citrate, malate, or lactate and show strong sequence homology. The citS gene encodes a highly hydrophobic protein of 446 amino acids with a predicted molecular mass of 47.5 kDa (135). According to the hydropathy profile (70) and an extensive topological analysis of CitS using in vitro translation in the presence of dog pancreas microsomes (137), a topology of 11 transmembrane segments and three extensive loops is predicted. This is confirmed by studies with membrane-permeable and impermeable thiol reagents using full-length CitS with engineered cysteine residues (136). There have been several indications for a dimeric arrangement of CitS in detergent micelles and in lipid membranes. Early evidence was derived from affinity chromatography, in which a mixed population of the biotinylated CitS fusion protein and wild-type CitS eluted as a heterodimer (103, 104). Analyses by blue native polyacrylamide gel electrophoresis of His-tagged CitS indicated a molecular mass of the homodimer (~100 kDa). Furthermore, single-molecule studies with dual-labeled CitS and dual color detection provided strong evidence for the homodimeric association of CitS (83).

The transport kinetics of CitS was studied in bacterial membranes and with the purified carrier reconstituted into proteoliposomes (104). Citrate transport by CitS depends on Na⁺ (50) in a cooperative behavior, reflecting a Na⁺-to-citrate stoichiometry of at least 2:1 (135). The kinetics also indicated that Hcit^2- is the transported species of CitS (93). Driving force for citrate uptake is provided by ΔpNa⁺ and ΔpH but not by Δψ, indicating that the transport is an electroneutral event. Detailed kinetic studies were performed in the exchange mode and led to the model of the transport mechanism shown in Fig. 3 (103, 104). The dimeric transporter has on one monomer a high-affinity Hcit^2- binding site (K_s, ~140 μM) exposed to the outside and on the other monomer a low-affinity Hcit^2- binding site (K_s, ~14 mM) exposed to the inside. After the outside binding site has been occupied with Hcit^2- and two Na⁺ ions, and one hydroxyl and one Na⁺ ion have been bound to the inside binding site, a conformational change leads to the reorientation of the binding sites towards the opposite compartment. Simultaneously, the sites switch in their binding affinities so that citrate is released easily to the inside while the hydroxyl ion is released to the outside. The overall translocation mechanism catalyzed by CitS can thus be written as follows (104):

Fig. 3Model for citrate transport by the CitS homodimer. The overall translocation process is electroneutral and a transmembrane chemical gradient of the cotransported ions (ΔpNa⁺/ΔpH⁺) serves as driving force ().

Hcit^2- _out + 2 Na⁺ _out + Na⁺ _in + OH^- _in → Hcit^2- _in + 2 Na⁺ _in + Na⁺ _out + OH^- _out

The most highly conserved regions of CitS are the periplasmic loop V-VI and the cytoplasmic loop X-XI (4). Selected amino acid residues of both regions were subjected to site-specific mutagenesis. These studies revealed that asparagine 186 in the loop V-VI region appears to be involved in citrate binding (82). Single-molecule studies were performed with fluorophores introduced at engineered cysteines in the loop X-XI region. Upon citrate binding to CitS with the fluorophore at cysteine 398, complete fluorescence quenching of most of the single molecules was observed, indicating a citrate-induced conformational change in the loop X-XI domain of CitS (83). The presence of the high-affinity binding site for citrate in this region is compatible with mutational analyses with the related citrate transporter CitP from Leuconostoc mesenteroides (5, 6).

The third citrate transporter of K. pneumoniae is CitW (84). The citW gene is located directly upstream of citX, which is required for the biosynthesis of the prosthetic group of citrate lyase (120). The expression of these genes is probably regulated by the CitY/CitZ two-component regulatory system because the corresponding genes are located upstream of citX and in inverse orientation. As the highest levels of CitX are synthesized under anaerobic growth conditions on citrate minimal medium, CitW could play a specific role in citrate uptake into citrate-fermenting cells. From sequence comparison, CitW is a member of the 2-HTC transporter family which also includes CitS (4, 84,120). CitW consists of 454 amino acids and has a molecular mass of 48.2 kDa. The transporter catalyzes the Na⁺-independent uptake of citrate in exchange for acetate, with an affinity for Hcit^2- of 25 μM (84). As acetate is the major end product of citrate fermentation in K. pneumoniae, this transport mode seems to be adapted to the physiological needs for growth under these conditions. However, as CitS and CitW are both present in citrate-fermenting cells of K. pneumoniae, the participation of either carrier in citrate uptake under various environmental conditions remains to be explored.

The transporter-designed CitT acts as a citrate carrier in E. coli growing anaerobically on citrate in the presence of a suitable cosubstrate (see above) (94). The citT gene is located immediately downstream of the citCDEFXG cluster and is therefore part of the citC operon (105). The CitT protein consists of 487 amino acids and has a molecular mass of 53.1 kDa. Sequence comparisons revealed that CitT is related to the 2-oxoglutarate/malate translocator from spinach chloroplasts and bacterial gene products which have not yet been functionally characterized. CitT appears to be a member of a novel family of transporters involved in the transport of di- and tricarboxylic acids. Transport studies revealed that CitT catalyzes a homologous exchange of citrate or a heterologous exchange against succinate, fumarate, or tartrate (105). Since succinate is the end product of citrate fermentation in E. coli, it is likely that CitT functions in vivo as a citrate/succinate antiporter.

Citrate Lyase.

Citrate lyase catalyzes the Mg²⁺-dependent cleavage of citrate to acetate and oxaloacetate. The reaction represents the initial step in all known bacterial citrate fermentation pathways. Citrate lyase has been purified and characterized from a variety of different bacterial species (2, 40, 131), but most of the fundamental biochemical studies were performed with the enzyme from K. pneumoniae. The enzyme is composed of three different subunits, α (54.7 kDa), β (31.4 kDa), and γ (11.4 kDa), assembled into a hexameric complex of approximately 550 kDa in a 1:1:1 stoichiometry (43, 125). Whereas the α and β subunits possess catalytic activity, the γ subunit serves as an acyl carrier protein (ACP) providing the appropriate substrates for the catalytic subunits (42, 44). For this purpose the prosthetic group 2'-(5''-phosphoribosyl)-3'-dephospho-CoA is covalently bound via phosphodiester linkage to serine 14 of the apoprotein (10, 38, 47, 109, 124). In this form citrate lyase is still inactive, but gains catalytic power by converting the thiol group of the prosthetic group into the acetyl-thioester derivative (22). Citrate is then cleaved in two steps. In the first step, the α subunit catalyzes an acyl exchange reaction, using acetyl-S-ACP and citrate as substrates and forming citryl-S-ACP and acetate as products. Citryl-S-ACP then serves as substrate for the β subunit, where it is cleaved in an Mg²⁺-dependent reaction into acetyl-S-ACP and oxaloacetate (44, 48). The mechanism of the citrate lyase reaction is summarized in Fig. 4.

Fig. 4(A) Function of the different subunits of citrate lyase and activation of the enzyme by citrate lyase ligase. ACP, acyl carrier protein; HS-R, prosthetic group; α-subunit, acetyl-ACP:citrate ACP transferase; β-subunit, citryl-ACP lyase. (B) Chemical reactions allowing specific acetylation, deacetylation, and carboxymethylation of the prosthetic group of citrate lyase.

As the carbon-carbon bond of citrate itself cannot easily be cleaved by an aldol-type reaction, the enzyme elegantly overcomes this chemical problem by performing the critical reaction with the citryl thioester derivative (22). This is formed easily from the acetyl-S-ACP derivative in the first partial reaction, and subsequently the cleavage of the citryl-S-ACP derivative in the second partial reaction regenerates the acetyl-S-ACP derivative (44). Removal of the acetyl group from the enzyme by treatment with hydroxylamine or thiol compounds inactivates citrate lyase (22). The desacetyl citrate lyase can be reactivated by chemical acetylation with acetic anhydride (22) or by the natural enzymic acetylation with acetate, ATP, and citrate lyase ligase (CitC) (114). The mechanism of CitC follows the two-step reaction sequence shown below:

Acetate + ATP → acetyl-AMP + pyrophosphate

Acetyl-AMP + HS-ACP → AMP + acetyl-S-ACP

Desacetyl citrate lyase becomes irreversibly inactivated by treatment with iodoacetate (22). This compound reacts specifically with the SH group of the prosthetic group, converting it into the carboxymethyl derivative (43). The reaction proceeds even in the presence of an excess of dithioerythritol, and it is catalyzed by citrate lyase, which appears to recognize iodoacetate as well as acetic anhydride as a substrate analog (20, 43). Desacetyl citrate lyase or its carboxymethylated derivative was found to catalyze the cleavage of citrate in the presence of acetyl-CoA, with the intermediate formation of citryl-CoA (25). Thus, acetyl-CoA and citryl-CoA can substitute the natural ACP-bound acyl intermediates by virtue of structural similarities between CoA and the prosthetic group.

Two proteins are required for the biosynthesis of the 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group and its attachment to the apo-ACP (Fig. 5). First, CitG catalyzes the formation of the α-1,2-glycosidic bond between ATP and 3'-dephospho-CoA to yield the prosthetic group precursor 2'-(5''-triphosphoribosyl)-3'-dephospho-CoA (TPRDP-CoA) with the concomitant release of the adenine moiety of ATP (118). The precursor is not released into the environment but remains tightly bound to CitG. From this it is only released in the second step, where the precursor is transferred by CitX to apo-ACP, resulting in the formation of holo-ACP and pyrophosphate (119, 120). In holo-ACP the prosthetic group is linked as phosphodiester to serine 14 of the protein (10). Thus, CitG functions as a 2'-(5''-triphosphoribosyl)-3'-dephospho-CoA synthase and CitX functions as an apo-ACP phosphoribosyl-dephospho-CoA transferase.

Fig. 5Biosynthesis of the 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group of citrate lyase.

Oxaloacetate Decarboxylase.

Another key enzyme of citrate fermentation in K. pneumoniae is oxaloacetate decarboxylase. The decarboxylase has been characterized as a membrane-bound, biotin-containing sodium ion pump (34, 35, 36, 39). It is the prototype of a family of enzymes which convert the free energy of a decarboxylation reaction into an electrochemical gradient of sodium ions across the membrane (37, 41, 46). Other members of this family include methylmalonyl-CoA decarboxylase (63, 64), glutaconyl-CoA decarboxylase (21, 23, 24), and malonate decarboxylase from anaerobic bacteria (45). Each of these decarboxylases catalyzes an essential step in the respective fermentation pathway and thereby pumps sodium ions out of the cell.

The decarboxylase from K. pneumoniae is a multisubunit complex composed of subunits α (OadA, 63.5 kDa), β (OadB, 44.9 kDa), and γ (OadG, 8.9 kDa) in a 1:1:1 stoichiometry (54). The α subunit is water soluble and consists of two domains which are connected by a flexible proline/alanine linker peptide (122). The N-terminal domain harbors the carboxyltransferase catalytic site, where the carboxyl group from position 4 of oxaloacetate is transmitted to the biotin prosthetic group. The C-terminal domain contains the binding site for the biotin, and it is tightly connected with the C-terminal domain of the γ subunit and thus important for the stability of the complex (117, 122). The β subunit is a very hydrophobic integral membrane protein (90, 141, 142) catalyzing the decarboxylation of carboxybiotin coupled to the transport of Na⁺ ions across the membrane (51, 52). The topology of the β subunit consists of three membrane-spanning α-helices in the N-terminal part (helices I-III) and six membrane-spanning α-helices in the C-terminal part (helices IV-IX). These are connected by a hydrophobic linker (region IIIa) that is predicted to insert into the membrane from the periplasmic surface, but not to penetrate through it to the cytoplasmic surface (76). The γ subunit inserts into the membrane with one α-helix at the N terminus, while the remainder of the protein is hydrophilic (122, 142). The C-terminal domain carries a Zn²⁺ binding site contributed by Asp62, His77, His82, and an H₂O molecule as ligands (53, 117). The Zn²⁺ metal ion is part of the carboxyltransferase catalytic site, where its function is to polarize the carbonyl oxygen bond of oxaloacetate in order to facilitate the C–C bond cleavage. Without the γ subunit, the α subunit is a poor carboxyltransferase, operating at a rate of 0.13 s^–1, three orders of magnitude below the turnover of the holoenzyme (33).

A model of structure and function of the oxaloacetate decarboxylase Na⁺ pump is depicted in Fig. 6. The catalytic cycle starts with the transfer of the carboxylic group from position 4 of oxaloacetate to the biotin prosthetic group on the enzyme. The carboxybiotin thus formed switches from the carboxyltransferase catalytic site on OadA and OadG to the decarboxylase site on OadB. Here the decarboxylation takes place and the free biotin group is regenerated. During this Na⁺-dependent reaction, a periplasmically derived proton is consumed and two sodium ions are translocated from the cytoplasm into the periplasm (32, 52). Essential residues for this coupled vectorial reaction have been identified by site-specific mutagenesis of OadB (32, 77, 78). Based on these studies and other biochemical investigations, a model for the reaction mechanism was proposed (Fig. 7) (46, 78, 116). The model predicts that a number of highly conserved and functionally indispensable residues on helices IV and VIII and region IIIa of OadB are involved in the ion translocation mechanism. It is proposed that carboxybiotin formed at the carboxyltransferase site switches to the decarboxylase site on OadB, where it forms a stable complex, possibly with the side chain of R389 at the cytoplasmic surface of helix VIII. Site-directed sulfhydryl labeling studies with methanethiosulfonate reagents have identified helix VIII to align the channel for Na⁺ and H⁺ conductance across the membrane (139). Evidently, the proton which moves from the periplasmic reservoir through this channel must reach the carboxybiotin near the cytoplasmic surface to account for the consumption of a proton in the decarboxylation reaction (54). According to the model, the Na⁺ channel is initially open to the cytoplasm. In this conformation, the two different Na⁺ binding sites are of high affinity (K_m, ~1 mM). The first Na⁺ is thought to bind at a site near the periplasmic surface (center I), which includes D203 and probably also N373. Subsequently, the second Na⁺ binds to the Y229- and S382-including site (center II). According to the electroneutrality principle, it is envisaged that an Na⁺ ion would only be tolerated in the center of the membrane (center II) after charge balancing, requiring in this case the dissociation of a proton and removal from the site. The phenolic hydroxyl group of Y229 is sufficiently acidic for this purpose and is assumed to dissociate as the Na⁺ ion is approaching. The proton then moves to the carboxybiotin, where it is consumed during decarboxylation of this acid-labile compound. Concomitantly, the biotin prosthetic group leaves the site and OadB changes its conformation. This exposes the Na⁺ binding sites towards the periplasm and simultaneously decreases their Na⁺ binding affinities. The Na⁺ ions dissociate into the periplasmic reservoir while a proton enters the periplasmic channel and restores the hydroxyl group of Y229. Hence, each decarboxylation event is coupled to the transport of two Na⁺ ions from the cytoplasm to the periplasm and the consumption of a periplasmically derived proton.

Fig. 6Model of the catalytic mechanism of the oxaloacetate decarboxylase Na⁺ pump from K. pneumoniae. The model depicts the location and interaction between subunits as well as individual reactions carried out by these subunits during the catalytic cycle. B-H, biotin prosthetic group; B-CO₂^-, carboxybiotin.

Fig. 7Model for coupling Na⁺ and H⁺ movements across the membrane to the decarboxylation of carboxybiotin as catalyzed by oxaloacetate decarboxylase. The model shows the location of important residues of helix IV and helix VIII and of region IIIa of the β subunit. Also shown is the participation of these residues in the vectorial and chemical events of the Na⁺ pump. (A) The empty binding site region with enzyme-bound carboxybiotin (B-COO^?), exposing the Na⁺ binding sites towards the cytoplasm. (B) The first Na⁺ binding site at the D203-N373 pair has been occupied and the second Na⁺ enters the Y229-S382-including site with simultaneous release of the proton from the hydroxyl side chain of Y229. This displacement may lead to the rearrangement of hydrogen bonding in the network involving Y220 and carboxybiotin to deliver the proton to the latter compound. This catalyzes the immediate decarboxylation of the acid-labile carboxybiotin, including a conformational change (B → C) which exposes the Na⁺ binding sites towards the periplasm (C). The Na⁺ ions are subsequently released into this reservoir, while a proton enters the periplasmic channel and restores the hydroxyl group of Y229. (D) The Na⁺ binding sites are empty and exposed toward the periplasm and the biotin prosthetic group is not modified (B-H). Upon carboxylation of the biotin, the protein switches back into the conformation where the Na⁺ binding sites are exposed toward the cytoplasm (D → A) ().

The CitA/CitB Regulatory System.

Several classical investigations have established that the expression of the citrate fermentation genes is tightly regulated. The main stimuli for the synthesis of citrate lyase and oxaloacetate decarboxylase are the presence of citrate, anaerobiosis, and Na⁺ ions, while the additional presence of glucose prevents the expression of the corresponding genes (18, 28, 98, 99). This regulation is important physiologically since the synthesis of citrate lyase or oxaloacetate decarboxylase under inappropriate conditions would severely affect the function of the citric acid cycle, either by triggering a futile cycle of citrate synthesis and cleavage or by deprivation of oxaloacetate. More recently, the molecular basis for this regulation was identified. The citAB genes, which are located immediately downstream of oadB in the same orientation, were found to encode a two-component regulatory system which regulates the expression of the citC and citS operons (14, 15). Like other bacterial two-component regulatory systems (16, 101, 130), CitAB consists of a sensor kinase (CitA) and a response regulator (CitB). Binding of the appropriate signal molecules from the environment to CitA activates its intrinsic kinase activity and CitB becomes phosphorylated (97). This enhances the binding affinity to the appropriate DNA recognition site, allowing the phosphorylated CitB to bind and thereby to act as a transcriptional activator of the genes involved in citrate fermentation.

The sensor kinase CitA is a membrane-bound protein with a molecular mass of 62 kDa. It consists of an N-terminal periplasmic domain that is flanked by two transmembrane α-helices, a linker region, and the kinase domain, composed of the phosphorylation subdomain with the conserved H-box and the ATP binding subdomain with the N, G₁, F, and G₂ boxes (132, 143). A fusion protein (MalE-CitAC) composed of the maltose binding protein and the CitA kinase domain (amino acids 327-547) showed constitutive autokinase activity and transferred the γ-phosphate group of ATP to its cognate response regulator CitB (81). The autokinase activity of CitA was abolished by an H350L exchange, indicating that H350 represents the autophosphorylation site of CitA. The sensory properties of CitA were analyzed with the purified periplasmic domain containing a C-terminal His-tag (CitAPHis). This protein bound citrate specifically and with high affinity (K_d, ~5 μM at pH 7) in a 1:1 stoichiometry, indicating that the periplasmic domain of CitA acts as a highly specific receptor for citrate and hence that environmental citrate is the signal sensed by CitA. The crystal structure of the CitAP periplasmic sensor domain in complex with citrate has been solved (57, 108). Citrate is bound in a pocket with contact sites to two arginines, a lysine, and a histidine, which have been independently identified by calorimetric citrate binding studies of site-directed CitAP mutants (57).

The response regulator CitB is a water-soluble protein of 27 kDa composed of an N-terminal receiver domain with the phosphorylation site at aspartate 56 and a C-terminal DNA binding domain (15). The entire CitB and each of the individual domains have been overexpressed as His-tag fusion proteins in E. coli, purified, and characterized biochemically (97). CitB or the C-terminal domain binds specifically to the DNA segment connecting citC and citS, i.e., in the region between the divergently oriented citC and citS operons, where they are expected to act as transcriptional activators. Phosphorylation of CitB elicits structural changes resulting in an increase in its binding affinity by a factor of 50. This effect is in accordance with the anticipated role of signal transmission via phosphoryl group transfer reactions in the CitAB two-component regulatory system (97).

The signals regulating the expression of the citrate fermentation genes were also investigated with a chromosomally integrated citS'-'lacZ fusion construct (15). For maximal expression, citrate, anoxic conditions, and Na⁺ ions were required, in accordance with the requirement for the synthesis of the citrate fermentation enzymes in wild-type K. pneumoniae. In citrate minimal media containing added glucose or glycerol, expression of citS'-'lacZ was strongly reduced, supporting previous evidence for catabolite repression of the citrate fermentation genes (15). This evidence was confirmed and extended by binding studies with the cyclic AMP (cAMP) receptor protein (CRP) which is known to elicit catabolite repression in enterobacteria (12). The purified cAMP-CRP complex from K. pneumoniae bound to two sites in the citC-citS intergenic region, which were centered at position -41.5 upstream of the citC and citS transcriptional start sites (96). The binding was stimulated by the previous binding of the phosphorylated response regulator CitB. It was concluded that catabolite repression of the citrate fermentation genes is exerted by CRP and that in the absence of repressing carbon sources the cAMP-CRP complex serves to enhance the basal, CitB-dependent transcription level.

BACTERIAL GROWTH ON MALONATE

Environmental malonate stems in part from natural sources—in particular from plants but also from animals or as an end product of bacterial fermentations—and in part from industrial production. Aerobic growth on malonate is very common among gram-negative bacteria and has been described both for bacteria with a strictly respiratory type of metabolism, such as Azotobacter vinelandii (92), Pseudomonas aeruginosa (59), Rhizobium leguminosarum (58), and Acinetobacter calcoaceticus (87), and for facultatively anaerobic bacteria such as K. pneumoniae (91). In fact, 15 of the 30 genera of the Enterobacteriaceae are malonate positive, and this characteristic is an often used criterion in determinative bacteriology in this family (17). Anaerobic growth on malonate is more restricted. Fermentation of malonate as the sole carbon and energy source has been observed with Sporomusa termitida (19), Sporomusa malonica (31), and Malonomonas rubra (30). Citrobacter diversus (75) and K. pneumoniae (74) grow on malonate in the presence of yeast extract under anoxic conditions, and Rhodobacter capsulatus grows anaerobically on malonate in the light (29). Independent of the mechanism of energy conservation in these species, the decarboxylation of malonate is the key reaction in the decomposition of this compound. Malonate is chemically rather inert, especially at neutral pH, when both carboxylic acids are deprotonated. In fact, aqueous solutions of disodium malonate show no decarboxylation within 48 h at 125^oC (55). To overcome this difficulty, malonate is activated for C–C bond cleavage by transiently forming a thioester with the enzyme (45, 60, 115).

Malonate-Decarboxylating Enzymes

Aerobic malonate-degrading bacteria contain a water-soluble malonate decarboxylase as the key enzyme for the degradation of this substrate (45). Under these conditions bioenergetic demands are fulfilled easily, as the acetate derived is completely oxidized to CO₂ and H₂O in the highly exergonic reaction sequence of the citric acid cycle and respiratory chain. During fermentation of malonate, however, acetate and CO₂ are end products, and therefore the free energy of malonate decarboxylation (ΔG⁰, –17.4 kJ mol^-1) has to be conserved (30).

Malonate decarboxylase is induced upon aerobic growth of K. pneumoniae on malonate. The enzyme from this organism was the first malonate decarboxylase characterized biochemically and can be regarded as the prototype. Malonate decarboxylase is a water-soluble enzyme complex of 142 kDa that consists of four different subunits, α (65 kDa), β (34 kDa), γ (30 kDa), and δ (12 kDa), in an apparent stoichiometry of 1:1:1:1 (68, 115). The smallest subunit (γ, MdcC) is an acyl carrier protein (ACP) and provides the thiol moiety necessary for substrate activation in the form of a covalently attached 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group (115). The structure of this compound is the same as that of the prosthetic group of citrate lyase, and the catalytic mechanisms of malonate decarboxylase and citrate lyase are also very similar. The sequence of the catalytic events taking place on malonate decarboxylase is shown in Fig. 8 (45, 115). The first step is an exchange of the ACP moiety from acetyl-S-ACP to malonate, yielding malonyl-S-ACP and acetate. This ACP transferase reaction is catalyzed by the α subunit (MdcA). Subsequently, subunits β (MdcD) and γ (MdcE) catalyze the malonyl-S-ACP decarboxylation reaction. This reaction requires H⁺ as the second substrate and forms CO₂ and acetyl-S-ACP as products. It is obvious from this mechanism that the acetylated thiol cofactor of the ACP moiety is essential for malonate decarboxylase activity. Accordingly, the enzyme is inactivated by releasing the acetyl thioester residues by treatment with thiol reagents or hydroxylamine. The free thiol cofactor is acetylated again by treatment with acetic anhydride, and malonate decarboxylase activity is restored (115).

Fig. 8Reaction mechanism for malonate decarboxylation in K. pneumoniae and function of the individual proteins encoded by the mdc genes. MdcA, acetyl-S-ACP:malonate ACP transferase; MdcC, acyl carrier protein (ACP); MdcD,E, malonyl-S-ACP decarboxylase; MdcF, malonate transporter; MdcH, malonyl-CoA:ACP-SH transferase; MdcB, ATP:dephospho-CoA 5'-triphosphoribosyltransferase; MdcG, 2'-(5''-triphosphoribosyl)-3'-dephospho-CoA:apoACP 2'-(5''-phophoribosyl)-3'-dephospho-CoA transferase; MdcR, regulator protein.

The physiological acetylation of the thiol group of the ACP subunit involves two reactions. In the first reaction the malonyl residue of malonyl-CoA is transmitted to ACP-SH, yielding malonyl-S-ACP and CoA-SH. This reaction is catalyzed by a specific malonyl-CoA:ACP-SH transacylase (MdcH) (66). The enzyme is encoded by the mdcH gene, which is included in the mdc operon encoding malonate decarboxylase (68) (see below). In the second reaction the malonyl-S-ACP is decarboxylated to acetyl-S-ACP by the β and γ subunits of the malonate decarboxylase complex. The catalytic mechanism of MdcH involves the formation of a covalent malonyl-MdcH intermediate, from which the malonyl residue is subsequently transmitted to the ACP-SH moiety of malonate decarboxylase (66). MdcH is related to malonyl-CoA:ACP-SH transacylases of fatty acid synthetases (FabD) and polyketide synthetases. According to the crystal structure of FabD, the active site is related to α/β hydrolases, where Ser-His and Asp/Glu operate together as a charge relay system. In contrast to chymotrypsin, the Ser-His relay system of FabD is not stabilized by a side-chain carboxyl group but by the main-chain carbonyl group of Gln. Because the active site residues are conserved in all malonyl-CoA:ACP-SH transacylases, they can be identified in MdcH as Ser83, His187, and Asn328. The purpose of the charge relay system is to make Ser83 more acidic, producing a nucleophile that replaces the CoA moiety of malonyl-CoA to form malonyl-O-serine-MdcH. From this intermediate the malonyl residue is then transferred to the thiol moiety of the prosthetic group of malonate decarboxylase (66).

In its catalytically active acetylated form, malonate decarboxylase is specific for malonate as substrate with K_m and V _max values of 4.8 mM and 157 U/mg, respectively (115). The substrate is decarboxylated with maximum activity within a broad pH range between 5.0 and 8.5. As described above, the enzyme is inactivated by deacetylation and the desacetyl enzyme is reactivated by acetylation of the prosthetic group. Reactivation of the desacetyl enzyme is inhibited by incubation with silver nitrate, dicyclohexylcarbodiimide (DCCD), or thiol-directed reagents like p-hydroxymercuribenzoate or iodoacetic acid (115). Investigations on the DCCD-inhibited desacetyl enzyme revealed that the MdcA subunit accepts as substrates not only ACP-bound thioesters but also malonyl-CoA and acetyl-CoA (unpublished data). Hence, acetyl-CoA and malonate are converted to malonyl-CoA and acetate by virtue of this activity. The apparent activation of the desacetyl enzyme by acetyl-CoA in the presence of malonate is not due to the MdcH-catalyzed transfer of the acetyl group of acetyl-CoA to the thiol moiety of the prosthetic group, since MdcH is specific for malonyl-CoA (66). This activation rather involves the conversion of acetyl-CoA to malonyl-CoA, catalyzed by MdcA, and the subsequent transfer of the malonyl residue to the ACP moiety of the enzyme complex catalyzed by MdcH.

The biosynthesis of the prosthetic group has also been explored and was found to be very similar to the biosynthesis of the identical prosthetic group in citrate lyase. The synthesis starts with the MdcB-catalyzed condensation of dephospho-CoA with ATP to 2'-(5''-triphosphoribosyl)-3'-dephospho-CoA (69). In this reaction a new α(1''-2') glycosidic bond between the two ribosyl moieties is formed, and thereby the adenine moiety of ATP is displaced. MdcB therefore is an ATP:dephospho-CoA 5'-triphosphoribosyltransferase. The prosthetic group precursor binds by a strong but noncovalent interaction to MdcG, yielding MdcG_i. It is possible that the tight binding is necessary to protect the precursor in vivo from degradation (69). The second reaction is the transfer of the prosthetic group from MdcG_i to the apo-ACP, yielding holo-ACP and pyrophosphate and regenerating free MdcG. This enzyme can therefore be described as a 2'-(5''-triphosphoribosyl)-3'-dephospho-CoA:apo-ACP 2'-(5''-phosphoribosyl)-3'-dephospho-CoA transferase (69). Although the two reactions are catalyzed by individual enzymes in vitro, it cannot be excluded that in vivo MdcG forms a prosthetic group biosynthetic complex with MdcB and/or MdcC (ACP). MdcG indeed binds very tightly to holo-ACP. The physiological role for this tight binding may be to protect the prosthetic group from degradation reactions before the complex is assembled, or it may help to assemble the malonate decarboxylase complex. During assembly, MdcG must be released to make the prosthetic group accessible for malonylation and for interaction with the other subunits of the complex, and indeed isolated malonate decarboxylase does not contain MdcG (69). An unphysiological side reaction catalyzed by MdcG in vivo or in vitro in the absence of MdcB is the formation of adenylyl ACP from apo-ACP with ATP as the substrate (69). This reaction is considerably slower than the formation of holo-ACP in the presence of the physiological precursor 2'-(5''-triphosphoribosyl)-3'-dephospho-CoA and is therefore not observed in wild-type cells where all protein components of the prosthetic group biosynthesis are expressed.

The adenylyl transfer reaction catalyzed by MdcG is related to reactions catalyzed by nucleotidyl transferases, and sequence comparisons indeed indicate that MdcG is a member of the nucleotidyl transferase family of enzymes (67). These enzymes have a conserved motif with two aspartate residues which are essential for catalytic activity (3, 71). The motif is also present in MdcG, and mutation of either of these aspartates to alanine abolished the transfer of the prosthetic group to apo-ACP whereas the binding of the prosthetic group precursor was not affected (67). A model for the active site in MdcG adopted from the crystal structure of human DNA polymerase β (111) is shown in Fig. 9. The two catalytically important aspartic acid residues 134 and 136, together with a third conserved carboxylic amino acid, coordinate two magnesium ions which facilitate the nucleophilic attack of ACP-serine 25 on the α phosphate of the prosthetic group precursor.

Fig. 9Schematic of proposed phosphoribosyl-dephospho-CoA transfer by MdcG, showing the pentacovalent transition state according to the two metal ion catalysis mechanism (). The deprotonated hydroxyl group of serine 25 from the ACP makes an in-line nucleophilic attack on the α phosphate of the prosthetic group precursor under reversion of the conformation on the α phosphate. The transition state is stabilized by two magnesium ions, which are coordinated by the three invariant aspartates. A, adenine; R, 4'-diphosphopantetheine moiety of the prosthetic group.

Malonomonas rubra is a strictly anaerobic bacterium that grows on malonate as the sole carbon and energy source (30). The organism contains a malonate decarboxylase that is related to the enzyme from K. pneumoniae but includes additional features for energy conservation (60). The decarboxylase cannot be isolated as a complex but as individual water-soluble and membrane-bound components. Figure 10 summarizes the individual reactions involved in malonate decarboxylation in M. rubra. Like the K. pneumoniae malonate decarboxylase, the enzyme from M. rubra contains an ACP subunit (MadE) with a 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group with an attached acetyl thioester residue (8). The initial reaction of the catalytic cycle is also similar and involves the formation of malonyl-S-ACP and acetate from acetyl-S-ACP and malonate, catalyzed by an ACP transferase (MadA) (61). The following steps, however, are different. The free carboxyl group of malonyl-S-ACP is transferred by the carboxyltransferase (MadC,D) to a small biotin protein (MadF) (7, 9, 62), thereby regenerating the acetyl-S-ACP. The carboxybiotin protein diffuses to the membrane, where it is decarboxylated by the integral membrane protein MadB (9). This reaction is analogous to the decarboxylation of carboxybiotin by the OadB subunit of oxaloacetate decarboxylase and is coupled to Na⁺ ion translocation across the membrane. The electrochemical gradient of Na⁺ ions thus derived is used as driving force for ATP synthesis by an F₁F₀ ATP synthase. The initial acetylation of the thiol group of the ACP subunit is distinct from the K. pneumoniae malonate decarboxylase but similar to citrate lyase. Acetylation is achieved with acetate and ATP and a specific ligase (60).

Fig. 10Individual reactions catalyzing malonate decarboxylation coupled to Na⁺ translocation in M. rubra and genomic organization of the mad gene cluster.

Organization of Malonate Decarboxylase Genes

A number of operons encoding proteins required for malonate decarboxylation in aerobic or anaerobic bacteria have been sequenced. As representative examples we will describe the mdc operon from K. pneumoniae (68) and the mad operon from M. rubra (9). A physical map of these gene clusters and assignments of the functions of the proteins encoded is shown in Fig. 10. The mdc operon starts with a cluster of the four structural genes mdcACDE, encoding the subunits of malonate decarboxylase, interrupted by the mdcB gene, which encodes the biosynthesis of the prosthetic group precursor (69). Further downstream of mdcE is the gene mdcF, which encodes a malonate transporter (unpublished data). This is followed by mdcG and mdcH, which encode transfer of the prosthetic group to the apo-ACP and transfer of the malonyl residue from malonyl-CoA to the SH moiety of the prosthetic group, respectively (66, 69). In divergent orientation follows the mdcR gene (68). MdcR is a protein which is structurally related to regulator proteins of the LysR type. It has been shown by mutational analysis of mdcR that this protein indeed regulates mdc gene expression. E. coli transformed with a plasmid containing all nine mdc genes is able to grow aerobically on malonate as a sole carbon and energy source, in contrast to a strain with a deleted mdcR gene (68). Two binding sites for MdcR have been identified, one (P_mdc) situated at the 5' end of the mdcA gene, the other (P_mdcR) at the 5'-end of mdcR. MdcR functions as an activator for the transcription of mdcABCDEFGH with malonate as coinducer. Furthermore, MdcR negatively regulates its own synthesis, apparently independently of the presence of malonate (102).

The mad gene cluster from M. rubra (9) starts with madG, which encodes the enzyme for the prosthetic group precursor synthesis. This is followed by madB, which encodes the carboxybiotin decarboxylase, and madAECD, which encode the ACP transferase (MadA), the ACP (MadE), and the carboxyltransferase (MadCD), respectively. The madH gene codes for the ligase that acetylates the SH residue of the prosthetic group, and madK encodes the enzyme for the transfer of the prosthetic group to the ACP subunit. The madF gene codes for the biotin protein (7), and madLM encodes the malonate transporter (112). MadN is a transmembrane protein of unknown function.

Besides the structural genes for the different subunits of malonate decarboxylase and for the prosthetic group biosynthesis, each cluster also includes genes for specific malonate transporters. MdcF is the only hydrophobic protein encoded in the mdc operon of K. pneumoniae and therefore it was very tempting to assume it functions as a malonate transporter. Experimental evidence for this assumption was obtained by deleting mdcF from a plasmid containing the whole mdc operon of K. pneumoniae. The transformed E. coli strain was unable to degrade malonate but recovered this ability if mdcF was present on a separate plasmid. When grown on LB medium, the mdcF deletion mutant still synthesized malonate decarboxylase, indicating that the inability to grow on malonate was due to the lack of a malonate transport system and hence that mdcF encodes a malonate transporter (S. Hoenke, unpublished). In M. rubra, malonate uptake is catalyzed by a different transporter, composed of the two subunits MadL and MadM. It has been shown that malonate is imported into the cell as monoanionic H-malonate^– in a symport with one Na⁺ ion. Synthesis of MadLM in an E. coli strain expressing all mdc genes except mdcF complemented the malonate-negative phenotype of this strain in accord with the notion that MdcF, like MadLM, functions as malonate transporter (112).

Concluding Remarks

In this chapter we combined descriptions of the degradation of citrate or malonate in aerobic and anaerobic bacteria. Key enzymes specific for individual pathways are citrate lyase, oxaloacetate decarboxylase, and two different malonate decarboxylases in aerobic or anaerobic bacteria. These seemingly unrelated enzymes share remarkable common features that are depicted in Fig. 11. First, citrate lyase and the two different malonate decarboxylases contain the same 2'-(5''-phosphoribosyl)-3'-dephospho-CoA prosthetic group that is linked by a phosphodiester bond to an ACP. The prosthetic group biosynthesis is also the same, and the SH residue of the prosthetic group is acetylated in the catalytically active enzyme. This acetyl thioester is central for the activation of each specific substrate for the chemical C–C bond cleavage reaction. For this purpose, a specific ACP transferase converts the acetyl thioester into either the citryl or the malonyl thioester. These thioesters subsequently undergo C–C bond cleavage, releasing oxaloacetate or CO₂ and regenerating the acetyl thioester moiety. These enzymes therefore harbor the module for substrate activation and C–C bond cleavage. Malonate decarboxylase from the anaerobic bacterium M. rubra transfers the CO₂ released during C–C bond cleavage directly to the biotin moiety of a biotin carrier protein. The carboxybiotin is then decarboxylated by a membrane-bound decarboxylase, and simultaneously, an electrochemical gradient of Na⁺ ions is generated. An analogous reaction is catalyzed by oxaloacetate decarboxylase and was assigned the module for energy conservation. Malonate decarboxylase from M. rubra therefore combines both modules, whereas citrate lyase and the malonate decarboxylase from K. pneumoniae have the module for substrate activation and C–C bond cleavage and oxaloacetate decarboxylase has the module for energy conservation only.

Fig. 11Modular composition of the malonate decarboxylase from M. rubra. The enzyme is composed of a substrate activation module, which is related to citrate lyase and the non-energy-conserving malonate decarboxylase from K. pneumoniae. It also includes an energy conservation module with relationship to the Na⁺-translocating decarboxylases, e.g., oxaloacetate decarboxylase from K. pneumoniae.

Acknowledgments

Work in the author’s lab was supported by Swiss National Science Foundation and ETH Research Commission.