Results and Discussion

3.1. Biochemical and genetic studies on BDCs in rhizobia. Biotin-containing proteins are visualized with high specificity on protein blots developed with streptavidin-horseradish peroxidase conjugate. When protein extracts from various Rhizobium, Sinorhizobium and Mesorhizobium strains were analyzed by this method, three biotin-containing proteins with molecular masses of approximately 120, 78 and 14 kDa were detected (Araiza, Dunn 1998; Dunn et al. 1996, 2000). The 120 kDa protein is the subunit of the pyruvate carboxylase (PYC) in Rhizobium etli, Rhizobium tropici (Dunn et al. 1996), Rhizobium leguminosarum biovars viciae and trifolii (M. Dunn, unpublished) and S. meliloti (M. Dunn, G. Araiza, T. Finan, submitted). pyc::Tr\5 mutants of these rhizobia are unable to grow with pyruvate or sugars as sole carbon source but are symbiotically effective with their respective legume hosts (Arwas et al. 1986; Dunn et al. 1996; M. Dunn, G. Araiza, T. Finan, submitted; Ronson, Primrose 1979). PYC activity is present in M. loti, but not in Bradyrhizobium spp. and Azorhizobium caulinodans (M. Dunn, unpublished). The PYCs from R. etli CE3 and S. meliloti Rml021 have been studied in the most detail and their characteristics are as follows:

(i) Both PYCs function in the anaplerotic production of oxaloacetate from pyruvate, which is necessary for growth on pyruvate or sugars but not for symbiotic nitrogen fixation.

(ii) Both the R. etli and S. meliloti PYCs are encoded by chromosomal genes whose deduced products (1154 and 1152 amino acids, respectively) are 87% identical and contain typical BCCP, BC and CT domains. The holo-PYCs produced by both species are homotetramers with molecular masses of about 500 kDa.

(iii) PYC activity in both species is influenced primarily by the availability of biotin in the growth medium and to a lesser extent by the carbon source used for growth. Gene fusion studies have shown that pyc transcription is not significantly influenced under different growth conditions.

(iv) The R. etli PYC exhibits maximum activity in vitro with a much lower concentration of the allosteric activator acetyl CoA than that required by the S. meliloti enzyme. Like all previously characterized prokaryotic homotetrameric PYCs, the R. etli enzyme is inhibited by the allosteric effector L-aspartate. The PYC from S. meliloti provides a surprising exception to this rule since its activity is unaffected by L-aspartate in vitro.

We have shown that PYC plays a major role in determining the growth defects of an R. etli poly-P-hydroxybutyrate (PHB) synthase (phbC) mutant. Because the growth defects of the phbC mutant are identical to those of the R. etli pyc mutant (Cevallos et al. 1996; Dunn et al. 1996), we measured PYC activity and protein levels in the phbC mutant and found that they were significantly lower than in the wild-type strain. Interestingly, the inactivation of a second gene, called cfr (carbon flux regulator) in the phbC mutant restores PYC activity and growth (Dunn et al. 2000; Dunn, Araiza, Encarnación, Vargas, Mora, submitted; Encarnación, Vargas, Dunn, Mendoza, Mora, Mora, submitted). We are currently working to unravel the molecular mechanism by which PYC activity is regulated in the phbC and phbC cfr mutants.

Phosphoenolpyruvate carboxylase (PPC) performs the same anaplerotic role as PYC by catalyzing the non-biotin-dependent conversion of phosphoenolpyruvate to oxaloacetate. Some bacteria contain solely PYC or PPC as an anaplerotic enzyme, while others contain both (Scrutton 1978). It is interesting to note that Bradyrhizobium species and Azorhizobium caulinodans, which are thought to have evolved before the other three genera (Martínez-Romero, Caballero-Mellado 1996), lack PYC and contain only PPC, while the opposite is true of Rhizobium, Sinorhizobium and Mesorhizobium species (Araiza, Dunn 1998; M. Dunn, G. Araiza, T. Finan, submitted).

The activity of propionyl CoA carboxylase (PCC) has been detected in extracts prepared from free-living cells and bacteroids of many rhizobia (DeHertogh et al. 1964; Dunn et al. 2000). Biochemical analysis of R. etli CE3 has shown that the 78 kDa biotinylated protein detected on western blots corresponds to the a subunit of PCC. By analogy to PCC a subunits characterized in other organisms, this protein would contain the BCCP and BC domains of the enzyme. The R. etli

PCC also contains a 47 kDa, non-biotinylated PCC (3 subunit, which would contain the CT domain. The partially purified PCC from R. etli showed a distinct kinetic preference for propionyl CoA as a substrate, but will also carboxylate acetyl- and butyryl-CoAs with a much lower efficiency (Dunn et al. 2000). The selection of R. etli PCC mutants has so far been unsuccessful, but an S. meliloti PCC mutant (pccAr.TnS) (Charles, Aneja 1999) is devoid of PCC activity and unable to utilize propionate as a carbon source (M. Dunn and G. Araiza, unpublished). Charles and Aneja (1999) noted the presence of a gene encoding methylmalonyl CoA mutase (bhbA) adjacent to pccA and suggested that Pcc and BhbA participate in the final steps of the propionate degradation. This is supported by earlier biochemical evidence indicating an anaplerotic role for these enzymes in the production of succinyl CoA from propionate in rhizobia (DeHertogh et al. 1964).

The 14 kDa biotinylated protein produced by rhizobia is similar in size to the BCCP subunits of many bacterial acetyl CoA carboxylases (ACCs). We have not succeeded in detecting ACC activity in R. etli preparations enriched in this protein, perhaps because the ACC complex is highly unstable (Fall 1976).

3.2. Analysis of rhizobial genome sequences for BDC-encoding genes. Computer analysis of the complete genome sequences of S. meliloti and M. loti provide valuable data on BDC-encoding genes (Figure 1) and allow us to predict the entire complement of BDCs produced by these organisms. Consistent with the biochemical and genetic analysis of the rhizobial PYCs, the genome sequences of both organisms encode a single, chromosomally-localized pyc. In S. meliloti, this gene is the same as that which we recently cloned and functionally characterized (M. Dunn, G. Araiza, T. Finan, submitted). Each genome contains single copies of the genes encoding the four subunits of ACC. The accB encodes the BCCP subunit of the ACC, and the putative accB products in S. meliloti and M. loti are 16.5 and 15.9 kDa, respectively, fairly similar to the 14 kDa biotinylated protein visualized on Western blots. These data suggest that rhizobia contain an ACC dedicated to fatty acid synthesis and that PCC, which can also carboxylate acetyl CoA, does not function in this capacity.

The genes encoding the subunits of PCC are located on pSymb in S. meliloti and the pccA subunit is the same as that described by Charles and Aneja (1999). A 3-methylcrotonyl CoA carboxylase (MCC) also appears to be encoded on the S. meliloti pSymb, and this gene product would be predicted to function in leucine degradation. The mccA and mccB genes are flanked by genes encoding other enzymes (hmgL; 3-hydroxy-3-methylglutaryl-CoA lyase and ivdH\ isovaleryl-CoA dehydrogenase) involved in the degradation of leucine.

The prediction of genes encoding acyl CoA carboxylases in M. loti is complicated by the fact that the a and P subunit-encoding genes of PCCs and MCCs resemble one another at the primary sequence level. Two gene regions containing adjacent pccB and mccA genes are present on the M. loti chromosome, making it difficult to predict whether these encode the subunits of a MCC or a PCC. Non-adjacent, putative pccA and pccB genes are also present in M. loti, bringing the total number of BDCs in this organism to five.

Biochemistry, genetics and, more recently, genome analysis are providing insights into the characteristics and metabolic roles of BDCs in rhizobia. We should now be able to address whether any of the BDCs perform additional functions that are of unique importance to rhizobia, either as free-living populations in the soil or in symbiotic combination with a host plant. Because rhizobia produce multiple BDCs, they may provide a good model system to study how the biochemical and genetic regulation of this class of enzymes integrates with metabolism as a whole. Another interesting avenue of investigation involves determining how biotin protein ligase partitions biotin between the different BDCs.

Sinorhizobium meliloti Chromosome

[Äcetyl CoA carboxylase]

-i xerP H accA

Y00689 h

Mesorhizobium loti Chromosome

[Acetyl CoA carboxylase]

—I xerD hi OCC¿ y—| hyp, prot. |-tRNA transferase

—I trpA I acclr~ft fpic pyruvate carboxylase!

Y03896 TRmlb i Y03894II m rrrnti i-

ABC transporter

[Pyruvate carboxylase]


TRmla Yo3899

hyp. protein



|Proplonyl CoA carboxylase |

|3-methylcrotonyl CoA carboxylase"! Y21123

|Acyl CoA carboxylase 11

acyl CoA dehydrogenase


protein mccA h

[Acyl CoA carboxylase 2|

|Propionyl CoA carboxylase | —I tlhA "H pccA | \dnaQ— rUh |-| pcci? H hyP- prot").

Figure 1. Genome sequence encoding BDCs

4. References

Allen EK, Allen GN (1950) Bacterid. Rev. 14, 273-330

Araíza G, Dunn MF (1998) 16th North American Nitrogen Fixation Conference, abstract 1.20

Arwas R et al. (1986) J. Gen. Microbiol. 132, 2743-2747

Cevallos MA et al. (1996) J. Bacterid. 178, 1646-1654

Charles TC, Aneja P (1999) Gene 226,121-127

DeHertogh AA et al. (1964) J. Biol. Chem. 239, 2446-2453

Dunn MF (1998) FEMS Microbiol. Rev. 22, 105-123

Dunn MF et al. (1997) FEMS Microbiol. Lett. 157, 301-306

Dunn MF et al. (2000) In Pedrosa FO, Hungría M, Yates G, Newton WE (eds), Nitrogen Fixation:

From Molecules to Crop Productivity, pp. 379, Kluwer, Dordrecht, The Netherlands Dunn MF et al. (2000a) 17th North American Conference on Symbiotic Nitrogen Fixation, pp. 27 Encarnación S et al. (1995) J. Bacterid. 177, 3058-3066 Fall RR (1976) Biochim. Biophys. Acta 450, 475-480 Kaneko T et al. (2000) DNA Res. 7, 331-338 Lowe RH, Evans HJ (1962) Soil. Sci. 94, 351-356 Ronson CW, Primrose SB (1979) J. Gen. Microbiol. 112, 77-88 Scrutton MC (1978) FEBS Lett. 89, 1-9

Streit WR, Phillips DA (1996) Appl. Environ. Microbiol. 62, 3333-3338 West PM, Wilson PW (1940) Enzymol. 8, 152-162

Was this article helpful?

0 0

Post a comment