Existence of a True Phosphofructokinase in Bacillus sphaericus: Cloning and Sequencing of the pfk Gene

Abstract

Some strains of Bacillus sphaericus are entomopathogenic to mosquito larvae, which transmit diseases, such as filariasis and malaria, affecting millions of people worldwide. This species is unable to use hexoses and pentoses as unique carbon sources, which was proposed to be due to the lack of glycolytic enzymes, such as 6-phosphofructokinase (PFK). In this study, PFK activity was detected and the pfk gene was cloned and sequenced. Furthermore, this gene was shown to be present in strains belonging to all the homology groups of this heterogeneous species, in which PFK activity was also detected. A careful sequence analysis revealed the conservation of different catalytic and regulatory residues, as well as the enzyme's phylogenetic affiliation with the family of allosteric ATP-PFK enzymes.

Bacillus sphaericus strains are strictly aerobic and mesophilic bacilli which form spherical endospores (15). On the basis of DNA homology, this species has been divided in five homology groups (11, 17). In turn, group II was divided in two subgroups, IIA and IIB (11). Some strains in group IIA are biotechnologically very interesting, because they can produce insecticidal proteins which are active against mosquito larvae, in particular against various species of the genera Culex and Anopheles, which are known to be regular transmission vectors of filariasis and malaria (3, 4, 13, 14, 16).

These bacteria can metabolize a wide variety of organic compounds and amino acids (1), but they are unable to use hexoses and pentoses as unique carbon sources (18). In fact, these metabolic limitations hamper likely industrial developments with this species due to the high costs of the proteinaceous media used for toxin production compared to the costs of alternative media based on starch or molasses (5, 18, 26). Consequently, studies disclosing the true metabolic potential of B. sphaericus are essential for genetic-improvement programs.

The inability to metabolize carbohydrates in this species has been attributed to the absence of key enzyme activities in both the Embden-Meyerhof-Parnas pathway (phosphoglucoisomerase [EC 5.3.1.9], 6-phosphofructokinase [PFK] [EC 2.7.1.11]) and the Entner-Doudoroff pathway (phosphogluconate dehydratase [EC 4.2.1.12] and phospho-2-keto-3-deoxygluconate [EC 4.1.2.14]) (18). Furthermore, Russell et al. did not detect either glucose or sucrose transport (18). However, the observation that B. sphaericus 2362 could use N-acetylglucosamine as a carbon source (A. Alice, unpublished results) indicated that some metabolic enzymes had to be present in this microorganism.

PFK activity in B. sphaericus.

B. sphaericus and Bacillus subtilis strains were grown aerobically in Luria-Bertani (LB) medium at 30 and 37°C, respectively. At mid-log phase (optical density at 600 nm, 0.6 to 0.8), cells were harvested by centrifugation, washed twice with TDP buffer (20 mM Tris-HCl [pH 7.5], 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride), and resuspended in the same buffer at 4°C. Crude extracts were prepared by disrupting cells twice by 20-s sonication in the presence of glass beads and maintained refrigerated in an acetone-ice bath. Samples were centrifuged in a microcentrifuge at 17,400 × g for 15 min, and the supernatants were used for the enzymatic assays. Reactions were performed in a 500-μl final volume containing 50 mM Tris-HCl (pH 7.5), 5 mM Cl₂Mg, 2 U each of aldolase (Sigma Chemical Co.) and triose phosphate isomerase-glyceraldehyde-3-phosphate dehydrogenase (Sigma Chemical Co.), 1 mM fructose-6-phosphate (F-6-P) (Sigma Chemical Co.), 0.15 mM NADH, 0.5 mM ATP, and 10 to 50 μl of cell extract. NADH oxidation was determined by measuring the absorbance variation at 340 nm. Specific activity is expressed in nanomoles per minute per milligram of total protein. Protein concentrations were determined with a protein assay kit (Bio-Rad) with bovine serum albumin as the standard. PFK activity could be detected in all strains tested, and with the exception of some strains, they showed values similar to those obtained with B. subtilis 168 (Table 1). The detected activities were absolutely dependent on the presence of both ATP and F-6-P in the reaction mixture. Moreover, no ATP-PFK activity was detected when crude extracts were omitted from these mixtures, indicating that the commercial enzymes and reagents used in the assays were not contaminated with an ATP-PFK activity. The affinity constants (K_m), calculated from crude extracts, of ATP-PFK for ATP and F-6-P were estimated to be 0.11 and 0.29 mM, respectively, and the maximum rates (V_max) of metabolism were 28.5 nmol min⁻¹ for ATP and 25.9 nmol min⁻¹ for F-6-P.

TABLE 1.

Detection of PFK activity in B. sphaericus

Strain	DNA homology group	Sp act (nmol min−1 mg of protein−1)a
B. sphaericus strain
ATCC 14577	I	19.16 ± 1.14
2362	II A	23.08 ± 1.06
1593	II A	6.61 ± 3.9
2297	II A	12.55 ± 0.63
ATCC 7055	II B	20.05 ± 0.35
NRS592	III	6.82 ± 0.31
NRS400	IV	11.9 ± 1.54
NRS1198	V	25.88 ± 3.79
B. subtilis strain 168		27.7 ± 0.66

^aThe results are the means of results of three distinct experiments ± the standard deviations, except for strains 2362 (for which the results are the means of results of eight experiments) and 168 (for which the results are the means of results of five experiments).

PCR amplification and sequencing of the pfk gene of B. sphaericus 2362.

In order to ratify the previous finding, two degenerated primers, PFK1 (5′-TCCCMNGGNATGAAYGC-3′) and PFK2 (5′-CCNCGYTGDAYRTGNCC-3′), were designed from the alignment of different microbial ATP-PFK enzymes. Total DNA from B. sphaericus 2362 prepared as described previously (6) was subjected to PCR amplification. The reactions were performed in a 50-μl final volume containing the following reagents: 20 to 100 ng of DNA template, 1× buffer (Life Technologies), 1 U of Taq polymerase, 100 μM concentrations of each deoxynucleoside triphosphate, and a 1 μM concentration of each primer. Control reactions were performed with the DNA of B. subtilis 168. The temperature profile was as follows: initial denaturation at 94°C for 5 min followed by 30 cycles of 94°C for 30 s, 51°C for 60 s, and 72°C for 90 s and then an additional extension step of 72°C for 5 min.

A fragment of the expected size (720 bp) was amplified and sequenced with an ABI PRISM dRhodamine Terminator Cycle Sequencing Ready Reaction kit with AmpliTaq DNA polymerase and an automatic ABI 310 DNA sequencer (Applied Biosystems). Analysis of the translated sequence showed great similarity to ATP-dependent PFK (ATP-PFK) enzymes of B. subtilis, Bacillus stearothermophilus, and Paenibacillus macquariensis. A 658-bp fragment upstream of the 5′ end of the amplicon and a 1,542-bp fragment downstream of the 3′ end could also be amplified by reverse PCR. At least two independent clones were sequenced. Sequence analysis of all the fragments was performed with the BLAST server of the National Center for Biotechnology Information at the National Institutes of Health, Bethesda, Md. (http://www.ncbi.nlm.nih.gov). This analysis showed three ORFs (Fig. 1). The central ORF (962 bp) showed high identity to ATP-PFK amino acid sequences of B. subtilis (68% identical residues), B. stearothermophilus (67% identical residues), and P. macquariensis (52% identical residues), and as consequence, it was named pfk. The deduced amino acid sequence corresponded to a protein of 319 amino acids with a predicted molecular mass of 33,948 Da. A putative ribosome binding site (AGGAGG) was detected 7 bp upstream of the start codon, a distance that has been reported to be optimal for efficient translation in B. subtilis and Escherichia coli (24). Also, a putative promoter region was identified upstream of this gene with potential −10 and −35 regions similar to those described for σ^A in B. subtilis (9) and a putative −16 region was identified (25).

FIG. 1.

Organization of the B. sphaericus 2362 genomic region containing the pfk gene. orf1 is similar to the B. subtilis accA gene, and orf3 resembles the B. subtilis pykA gene (see the text). Black arrows indicate the positions and directions of primers used both in the RT-PCR assays (FFK5 and FFK6) and in the synthesis of the probe used in the Southern blot analysis (PFK1 and PFK2). Relevant restriction sites and their positions in the sequence are also indicated.

Upstream from pfk, we found an incomplete ORF (orf1, 312 bp) with an intergenic region of 311 bp which encodes a protein with similarity to the acetyl coenzyme A carboxylase carboxyltransferase alpha subunit (AccA) protein of B. subtilis. Moreover, downstream of the pfk gene there was another ORF (orf3) possibly encoding a protein with high identity to pyruvate kinase of Bacillus psychrophilus and B. subtilis. This gene organization (accA pfk pykA) is similar to that observed in B. subtilis and Bacillus halodurans (12, 22).

In order to determine if this cloned gene was transcribed, a reverse transcription-PCR (RT-PCR) assay was performed on B. sphaericus 2362 cells grown in LB medium (optical density at 600 nm, 0.6 to 0.8). Total RNA was isolated by using Trizol (Life Technologies) heated at 70°C and treated with RNase-free DNase (Promega) to eliminate contaminating DNA. For RT reactions, 1 to 5 μg of total RNA was used for the synthesis of the first-strand cDNA with Moloney murine leukemia virus reverse transcriptase enzyme (Ambion) and the primer FFK6 (5′-TAAGTCGACATTAGTAACTCCACTGC-3′). The RT reactions (20-μl mixtures) were carried out according to the manufacturer's recommendations. Then, 5 μl of each RT reaction product mixture was used as the template for PCR amplifications. These amplifications were carried out as described above with an annealing temperature of 60°C and with the primers FFK5 (5′-CACAAGCTTATTCAGCGTGGTGG-3′) and FFK6. A fragment with the expected size (617 bp) was detected with primers FFK5 and FFK6, while PCR products were not found in control reactions when the RT reaction was carried out without reverse transcriptase enzyme (Fig. 2).

FIG. 2.

RT-PCR of the pfk gene. RNA was extracted from B. sphaericus 2362, and the RT-PCRs were performed as described in the text. Lane 1, 100-bp DNA ladder; lane 2, RT-PCR with RNA from 2362; lane 3, RT-PCR without reverse transcription enzyme in the RT reaction.

Analysis of the deduced protein sequence.

The sequence of the PFK from B. sphaericus was aligned and compared with sequences of other bacterial species. Multiple alignments were performed with CLUSTAL X software (23) (Fig. 3). Ten out of the 11 amino acids that constitute the substrate-binding domain in the E. coli enzyme (21) are conserved in B. sphaericus (Fig. 3). Curiously, the R162 residue, which binds F-6-P and is important for the inhibition of the B. stearothermophilus ATP-PFK by phosphoenolpyruvate (PEP) (20), is changed to N162 in B. sphaericus PFK. The role of this particular site in allosteric inhibition was recently studied, with a finding that an R162A mutation decreased by 30-fold ATP-PFK's affinity for F-6-P but that it diminished the effectiveness of PEP inhibition by only one-third (10). Further biochemical work is required to determine the incidence of the R162N change in B. sphaericus PFK and its effects on both ATP-PFK's affinity for F-6-P and PEP inhibition.

FIG. 3.

Alignment of the PFK amino acid sequences from different microorganisms. The consensus sequence is given. The numbers on the right indicate the residue positions in the sequences. A, F, and E refer to residues implicated in the binding of ATP, F-6-P, and PEP, respectively, for E. coli ATP-PFK and are indicated in bold. B. sph., B. sphaericus (accession number AY034597); B. sub., B. subtilis (O34529); B. ste., B. stearothermophilus (P00512); P. mac., P. macquariensis (Q59214); B. hal., B. halodurans (BAB06883); L. del., Lactobacillus delbrueckii subsp. bulgaricus (P80019); E. coli, E. coli (AAC76898).

Another 10 residues are implicated in the ATP binding of E. coli ATP-PFK (21). In the deduced sequence of B. sphaericus ATP-PFK, this domain is also highly conserved, with eight identical amino acids, a conservative substitution (R77 to K77), and a nonconservative substitution (F73 to C73); however, these two differences were also present in the very well-characterized ATP-PFK from B. stearothermophilus. Also, residue D187, important for PEP inhibition and ADP or GDP activation of the enzyme (2), was conserved in B. sphaericus.

Phylogenetic analysis of the evolutionary relationships between these proteins was realized with the programs supplied in the PHYLIP 3.5c package (8). This method showed that this enzyme clustered convincingly with other enzymes of the ATP-PFK family (data not shown).

The presence of these conserved residues and the similarity to other ATP-PFK enzymes suggests that PFK of B. sphaericus 2362 is also an allosteric enzyme.

Complementation of the E. coli pfk mutant.

In order to confirm that the cloned gene encodes a functional ATP-PFK enzyme, a DNA fragment amplified with the primers PFKN (5′-TCGGGATCCAAAAAAATTGCTG-3′ containing a BamHI site [underlined]) and PFKC (5′-CATAAGCTTTTATATTGATAGTTC-3′, which contains a HindIII site [underlined]) and carrying the complete pfk gene from B. sphaericus 2362 was cloned into the expression vector pQE30 (QIAGEN), which had been digested with BamHI-HindIII. The recombinant plasmid, called pALE-232, was used to transform E. coli DF1020 [pro-82 glnV44(AS) λ-ΔpfkB201 recA56 endA1 Δ(rhaD-pfkA)200 thi-1 hsdR17; E. coli Genetic Stock Center], which has no ATP-PFK activity (7). This E. coli strain does not grow on mannitol as the carbon source. This deficiency was restored only when the pALE-232 plasmid was introduced and not with the pQE30 control vector. When ATP-PFK activities from these same strains cultivated in the presence of IPTG (isopropyl-β-d-thiogalactopyranoside) were compared, E. coli DF1020 cells carrying the B. sphaericus 2362 pkf gene showed an ATP-PFK activity of 27 nmol min⁻¹ mg of protein⁻¹ while E. coli cells harboring pQE30 showed less than 1 nmol min⁻¹ mg of protein⁻¹. These results proved that the cloned gene indeed encodes an ATP-PFK enzyme from B. sphaericus.

pfk genes in other strains of B. sphaericus.

In order to test whether the presence of the pfk gene in strain 2362 was a particular feature of this strain or strains in the IIA homology group, we performed a Southern blot experiment using an internal fragment of the pfk gene as the probe against DNA from different strains representing each of the homology groups described in the B. sphaericus taxon. DNA of B. sphaericus strains was digested with EcoRI and analyzed by electrophoresis on 0.8% horizontal agarose slab gels. Southern blot analysis was performed by standard procedures (19), and after being blotted, membranes were hybridized with a fragment of the pfk gene, amplified by PCR with the primers PFK1 and PFK2. This probe was labeled with a Bioprime DNA labeling system (Life Technologies). The detection was carried out with a chemiluminescence Phototope-Star CDP-STAR (New England BioLabs) system as described by the manufacturer.

Figure 4 shows that a clear signal was detected in all strains, indicating that the pfk gene (or a very similar one) is present in all DNA homology groups.

FIG. 4.

Southern blot of chromosomal DNAs from different strains of B. sphaericus with a pfk gene probe. Chromosomal DNAs from different strains were digested with EcoRI. Lane 1, ATCC 14577 (I); lane 2, ATCC 7055 (IIB); lane 3, NRS 592 (III); lane 4, NRS 400 (IV); lane 5, NRS 1198 (V); lane 6, 2362 (IIA); lane 7, 1593 (IIA); lane 8, ASB13052 (IIA); lane 9, 2297 (IIA); lane 10, DNA ladder. The positions of certain marker bands are indicated as references.

Conclusions.

Although B. sphaericus had a well-known inability to use sugars as unique carbon sources, results obtained in this study provide the first evidence that one of the enzymes of the upper half of the glycolytic pathway is present in B. sphaericus. Conservation of the catalytic and regulatory domains of the protein strongly supports its functionality in glycolysis or gluconeogenesis. Complementary studies indicated the presence of a PEP phosphotransferase system in this bacterium (A. F. Alice and C. Sanchez-Rivas, Abstracts of the 9th International Conference on Bacilli, abstr. 107, 1997), and there is also evidence of the existence of an N-acetylglucosamine-specific PEP phosphotransferase system in B. sphaericus (A. Alice, unpublished results), suggesting that B. sphaericus may in fact be able to translocate and catabolize certain sugars like N-acetylglucosamine, most probably through glycolysis.

This study may establish the foundations for other studies of the metabolism of this important mosquito pathogen, and this knowledge might eventually be used to find better carbon sources for the industrial production of toxins.

Nucleotide sequence accession number.

The nucleotide sequence presented in this work was entered in the GenBank nucleotide database under the accession number AY034597.