Aromatic Amino Acid Auxotrophs Constructed by Recombinant Marker Exchange in Methylophilus methylotrophus AS1 Cells Expressing the aroP-Encoded Transporter of Escherichia coli

Abstract

The isolation of auxotrophic mutants, which is a prerequisite for a substantial genetic analysis and metabolic engineering of obligate methylotrophs, remains a rather complicated task. We describe a novel method of constructing mutants of the bacterium Methylophilus methylotrophus AS1 that are auxotrophic for aromatic amino acids. The procedure begins with the Mu-driven integration of the Escherichia coli gene aroP, which encodes the common aromatic amino acid transporter, into the genome of M. methylotrophus. The resulting recombinant strain, with improved permeability to certain amino acids and their analogues, was used for mutagenesis. Mutagenesis was carried out by recombinant substitution of the target genes in the chromosome by linear DNA using the FLP-excisable marker flanked with cloned homologous arms longer than 1,000 bp. M. methylotrophus AS1 genes trpE, tyrA, pheA, and aroG were cloned in E. coli, sequenced, disrupted in vitro using a Km^r marker, and electroporated into an aroP carrier recipient strain. This approach led to the construction of a set of marker-less M. methylotrophus AS1 mutants auxotrophic for aromatic amino acids. Thus, introduction of foreign amino acid transporter genes appeared promising for the following isolation of desired auxotrophs on the basis of different methylotrophic bacteria.

The nonpathogenic Gram-negative bacterium Methylophilus methylotrophus is able to grow efficiently using C₁ substrates (methanol, methylamine, or trimethylamine) as the sole source of carbon and energy, and it uses the ribulose monophosphate pathway for fixation of formaldehyde produced by the oxidation of methanol (36). Methanol has received considerable attention by the fermentation industry as an alternative substrate to the more generally used sugars from agricultural crops. It can be synthesized either from petrochemicals or renewable resources, such as biogas (48), and therefore the production of methanol does not compete directly with human food supplies. Methylotrophs can therefore be considered potentially useful strains for industrial biotechnology. M. methylotrophus AS1 is an obligate methylotroph originally isolated from activated sludge, and it has been deposited in the National Collections of Industrial, Marine and Food Bacteria (NCIMB; no. 10515). This organism was extensively studied in the 1970s and has been industrialized on a large scale for the manufacturing of single-cell proteins (SCP) from methanol (56, 63). During that period, a significant amount of research was conducted on the direct production of amino acids by fermentation from methanol (3, 58). Although initially promising, these efforts ultimately proved relatively unsatisfactory and impractical, due primarily to the rather poor set of genetic tools that had been developed for methylotrophs.

Over the last 5 years, several genomes of methylotrophs have been sequenced (8, 20, 29, 37, 65, 67), and significant progress in elucidating their metabolism has been achieved (14). The number of tools available for the genetic and metabolic engineering of methylotrophic bacteria has been expanded greatly (1, 15, 21, 43), and strategies to produce fine and bulk chemicals by methylotrophs have been described (5, 42, 57, 61). All of these factors led to renewed interest in the construction of methylotrophic strain producers, and the larger knowledge base has enabled more targeted engineering of these bacteria (55).

Although M. methylotrophus AS1 has been extensively studied with regard to the industrial scale production of SCP (56, 63) and the oxidation of methanol at the initial stages of carbon and energy metabolism (13, 28), there has been little metabolic analysis of amino acid biosynthesis in this organism. Moreover, selection of auxotrophic mutants of obligate methylotrophs for broadening convenient genetic tools remains a particularly complicated task (19). Although the isolation of several auxotrophs for M. methylotrophus AS1 has been described (6, 23, 40), their numbers are limited. Development of different methods for the isolation of the mutants did not lead to construction of a collection of auxotrophic mutants that could assist in the investigation of amino acid biosynthetic pathways in M. methylotrophus AS1.

As for the l-lysine (Lys) synthesis, systematic research was carried out by specialists at Ajinomoto Co., Inc., Japan, beginning with the investigation of the Lys biosynthetic pathway in M. methylotrophus AS1 (23, 61) and continuing with the construction and improvement of a Lys producer (22, 24, 33, 34). This was followed by optimization of fed-batch fermentation for overproduction of Lys from methanol (35).

The aim of our investigation was to generate strains based on M. methylotrophus AS1 with the potential for industrial production of aromatic amino acids (AroAAs). It is known that mutants with relaxed feedback inhibition of key biosynthetic enzymes should be isolated at the initial steps of the construction of the amino acid producers and that the relevant degradation pathways should be blocked due to selection of the corresponding auxotrophic strains (7, 31, 49).

In this study, a novel method for the construction of AroAA auxotrophic mutants of M. methylotrophus AS1 is described. This method is based on the introduction of a foreign gene encoding a specific amino acid transporter into the genome of M. methylotrophus AS1. The resulting recombinant methylotrophic strain, which possesses increased permeability to the AroAAs and their analogues, was mutated by recombination-mediated substitution of the target chromosomal genes of aromatic pathways by a flippase recombinase (FLP)-excisable marker from artificial linear DNA. This approach led to the construction of a set of M. methylotrophus AS1 marker-less mutants with destroyed genes of AroAA biosynthesis. Thus, introduction of foreign amino acid transporter genes appeared promising for the following isolation of desired auxotrophs on the basis of different methylotrophic bacteria.

MATERIALS AND METHODS

Bacterial strains, plasmids, and cultivation conditions.

Strains and plasmids used in the study are shown in Table 1. M. methylotrophus cells were grown at 37°C on minimal medium SEIIa (23) of the following composition: 1.9 g/liter K₂HPO₄, 1.56 g/liter NaH₂PO4 × 2H₂O, 5 g/liter (NH₄)₂SO₄, 200 mg/liter MgSO₄× 7H₂O, 72 mg/liter CaCl₂ × 2H₂O, 5 μg/liter CuSO₄ × 5H₂O, 25 μg/liter MnSO₄ × 5H₂O, 23 μg/liter ZnSO₄ × 7H₂O, and 9.7 mg/liter FeCl₃ × 6H₂O. Methanol was added to the liquid medium at a final concentration of 2%, or 1% methanol and 1.2% Bacto agar (Difco) were added to create the solid medium. Escherichia coli strains were cultured at 37°C in Luria-Bertani (LB) medium (52), on LB agar containing 1.2% Bacto agar, or on M9 minimal medium supplemented with 0.2% glucose. The following antibiotic concentrations were used for M. methylotrophus: 100 μg/ml ampicillin (Ap), 2 μg/ml tetracycline (Tc), 50 μg/ml streptomycin (Sm), 10 μg/ml kanamycin (Km), and 20 μg/ml chloramphenicol (Cm). For E. coli strains, the following antibiotic concentrations were used: 200 μg/ml Ap, 10 μg/ml Tc, and 50 μg/ml Km.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)	Source or reference
Strains
M. methylotrophus
AS1	Wild type	NCIMB 10515
AS1::aroPEco	Modified strain AS1 containing the E. coli gene aroP integrated in the chromosome	This study
AS1::aroPEco-trpEMme::FRT	TrpE mutant	This study
AS1::aroPEco-tyrAMme::FRT	TyrA mutant	This study
AS1::aroPEco-pheAMme::FRT	PheA mutant	This study
AS1::aroPEco-aroGMme::FRT	AroG mutant	This study
AS1::aroPEco-trpEMme::FRT- tyrAMme::FRT	TrpE TyrA double mutant	This study
E. coli
TG1	F− Δ(lac-pro) supE thi hsdΔ5 [F′ traD36 proAB+ lacIq lacZΔM15]	VKM IMG-341
S17-1	Tpr Smr; F−recA pro thi hsdR−hsdM+ RP4-2-Tc::Mu-Km::Tn7	ATCC 47055
MG1655	F− λ−ilvG rfb-50 rph-1	CGSC 6300
JW1256	BW25113ΔtrpE::kan	4
JW2581	BW25113ΔtyrA::kan	4
JW2580	BW25113ΔpheA::kan	4
JW0737	BW25113ΔaroG::kan	4
JW2582	BW25113ΔaroF::kan	4
JW1694	BW25113ΔaroH::kan	4
BW25113-ΔaroGFH	BW25113ΔaroG ΔaroF ΔaroH with FLP-mediated Kmr excision	This study
JW1256-ΔtrpE	JW1256 with FLP-mediated Kmr excision	This study
JW2581-ΔtyrA	JW2581 with FLP-mediated Kmr excision	This study
JW2580-ΔpheA	JW2580 with FLP-mediated Kmr excision	This study
Plasmids
pAYCTER3	Apr; IncQ broad-host-range cloning vector derived from pAYC32	1
pAYCTER-aroPEco	pAYCTER3 with the 1.9-kb BamHI-EcoRI PCR fragment containing the E. coli gene aroP	This study
pAYCTER-brnQEco	pAYCTER3 with the 1.7-kb SmaI PCR fragment containing the E. coli gene brnQ	This study
pMIV5-[FRT-Kmr-FRT]-Mob	Apr Kmr; pMIV5-Mob with the 1.5-kb HindIII-NdeI fragment from pKD4 containing the Kmr gene-flanked FRT sites	1
pMIV5-[FRT-Kmr-FRT]-Mob-aroPEco	pMIV5-[FRT-Kmr-FRT]- Mob with 1.9-kb PCR fragment containing the E. coli gene aroP	This study
pFLP31	Apr Smr; pAYCTER3 with the 3.3-kb SmaI-BamHI fragment from pCP20 containing the genes λcI857ts λcro-FLP	1
pPW121	Cloning vector	51
pTP310	Tcr; pRK310 with the 5.7-kb BamHI fragment from pUC-MuAB containing the genes MuAB ner cts	1
pKD4	Kmr Apr; pANTSγ[FRT-Kmr-FRT]	16
pPW-(trpEGDC)Mme	pPW121 with 5.2-kb Sau3A fragment containing M. methylotrophus genes trpEGDC	This study
pPW-(tyrAMme)	pPW121 with 2.4-kb Sau3A fragment containing the M. methylotrophus gene tyrA	This study
pPW-(pheA-hisH-aroG)Mme	pPW121 with 4.6-kb Sau3A fragment containing M. methylotrophus genes pheA hisH aroG	This study

Bacterial mating.

Plasmids were transferred into strains of M. methylotrophus AS1 by biparental mating using E. coli S17-1 bearing the respective plasmid as described in reference 1.

DNA manipulation.

Plasmid DNA isolation, E. coli transformation, restriction enzyme digestion, ligation, end-blunting, and Southern hybridization were carried out as described by Sambrook and Russell (52). Restriction enzymes, T4 DNA ligase, E. coli DNA polymerase Klenow fragment, calf intestine alkaline phosphatase, and Pfu polymerase were obtained from Fermentas (Lithuania), and P1vir phage-mediated generalized transduction experiments were carried out as described by Miller (44).

PCR.

All PCR analyses were performed using Pfu polymerase (Fermentas) with 1 μM of each primer, 200 μM of deoxynucleoside triphosphates, and typically 200 ng of genomic DNA or 25 ng of plasmid DNA as a template.

Electroporation of M. methylotrophus AS1.

The culture of M. methylotrophus AS1 was grown to an optical density at 600 nm of 1.0 in the SEII medium. Cells were harvested from a 5-ml culture, placed on ice, and washed three times in ice-cold distilled water. Finally, cells were resuspended in 100 μl ice-cold water, DNA was added (200 ng for plasmids; 500 ng for linear fragments), and electroporation was carried out in a 0.2-cm cuvette at 2.5 kV, 25 μF, and 200 Ω. A total of 1 ml of SEII medium was added, and the cells were grown for 3 h at 37°C with vigorous shaking. The electroporation efficiency was approximately 5.2 × 10⁴ transformants per 1 μg of plasmid DNA.

Construction of plasmids with aroP and brnQ.

The plasmids and primers used in this work are listed in Table 1 and in Table S1 in the supplemental material, respectively. The E. coli gene aroP (aroP_Eco) was amplified by PCR using chromosomal DNA from E. coli MG1655 as a template and the primers P24 and P25. The 1.9-kb PCR fragment containing aroP was digested with BamHI/EcoRI and ligated into the broad-host-range cloning vector pAYCTER3 (1) digested with BamHI/EcoRI to obtain pAYCTER-aroP_Eco. The plasmid for integration of aroP into the chromosome of M. methylotrophus AS1, pMIV5-[FRT-Km^r-FRT]-Mob-aroP_Eco, was constructed by blunt-end PCR cloning of the Pfu polymerase-amplified fragment containing aroP_Eco into the integrative plasmid pMIV5-[FRT-Km^r-FRT]-Mob (1), which was partially digested with PvuII (since two sites for PvuII are present in pMIV5-[FRT-Km^r-FRT]-Mob) and dephosphorylated with calf intestine alkaline phosphatase.

The E. coli gene brnQ was amplified using chromosomal DNA from E. coli MG1655 as a template and the primers P26 and P27. The 1.7-kb PCR fragment containing brnQ_Eco was digested with SmaI and ligated into the same restriction site as pAYCTER3 to create pAYCTER3-brnQ_Eco.

Cloning of the genes from AroAA biosynthesis pathways of M. methylotrophus AS1.

Chromosomal DNA of M. methylotrophus AS1 was partially digested with Bsp143I (Sau3AI), and fragments greater than 2 kb were purified from agarose gels and ligated into the BglII site of pPW121 (51). The BglII restriction site is located in the vector plasmid inside a specially modified structural part of the λcI repressor gene. Cloning into this site leads to Tc^r gene expression driven by the derepressed λP_R-promoter, providing direct selection for the inserts. The resulting ligation mixture was used to transform the appropriate E. coli mutant strain and select Tc^r clones with complemented mutation. The plasmid DNA was isolated from such clones, and after confirmation of the complementation event by retransformation to the appropriate E. coli strain, the mutant strain was further sequenced. The plasmids pPW-(trpEGDC_Mme), pPW-(tyrA_Mme), and pPW-(pheA-hisH-aroG_Mme) containing 5,227-bp, 2,445-bp, and 4,689-bp Bsp143I (Sau3AI) fragments, respectively, with genes for biosynthesis of AroAAs (Fig. 1) were selected using this approach.

FIG. 1.

Genetic organization of the Bsp143I (Sau3AI) fragments containing the genes from aromatic amino acid biosynthesis pathways of Methylophilus methylotrophus AS1 cloned in this work. The 5,227-bp Bsp143I fragment contains trpE, trpG, trpD, cds1, and trpC. The 4,689-bp Bsp143I fragment contains the 3′ part of serC, pheA, hisH, aroG, and the 5′ part of pqqB. The 2,445-bp Bsp143I fragment contains tyrA and the 5′ part of aroA. The genes are shown as arrows, and the partial ORFs located on the fragment are indicated with apostrophes. The sites for restrictases used for gene disruption and subcloning are shown.

Construction of plasmids and linear DNA fragments for mutagenesis.

To inactivate trpE_Mme, an FLP-excisable Km^r marker was amplified using Pfu polymerase, primers P1 and P2 (see Table S1 in the supplemental material), and pKD4 as a template. The resultant 1.5-kb fragment carrying FRT-Km^r-FRT was ligated into the NcoI site of pPW-(trpEGDC_Mme) that had been blunted by the Klenow fragment of E. coli DNA polymerase, and the plasmid pPW-(trpEGDC_Mme)-Km was constructed. A linear DNA fragment obtained after digestion of pPW-(trpEGDC_Mme)-Km by BamHI, and Eam1104I was used for trpE_Mme mutant construction.

To inactivate tyrA_Mme, the 1.5-kb FRT-Km^r-FRT carrier fragment was amplified from pKD4 by PCR analysis using primers P5 and P2 and ligated into the SwaI site of pPW-(tyrA_Mme). The resultant plasmid, pPW-(tyrA_Mme)-Km, was digested with SmaI and BamHI to obtain a linear fragment for the tyrA_Mme mutant construction. The primer pairs P28/P29 and P30/P31 were used to amplify the linear fragments containing FRT-Km^r-FRT flanked with “arms” of different lengths from pPW-(tyrA_Mme)-Km as a template in order to optimize the mutant construction procedure.

To inactivate pheA_Mme, the linear fragment containing FRT-Km^r-FRT flanked with arms of 1,000 bp in length was constructed by overlap extension PCR (18, 66). The primers P8 and P9 were used for left arm amplification, P10/P11 were used to amplify the right arm, and P12/P13 were employed for FRT-Km^r-FRT amplification. The primers P8 and P10 contain 21-nt regions at their 5′ extremities that are homologous to the 3′ and 5′ terminals of the FRT-Km^r-FRT marker. The reaction was carried out for 30 cycles as follows: denaturation at 95°C for 30 s, primer annealing at 58°C for 45 s, and primer extension at 72°C for 2 min. Then, 50-ng samples of the left and right arm were used as primers, and 50 ng of the fragment containing FRT-Km^r-FRT was used as a template to amplify the entire fragment. The reaction was carried out as follows: denaturation at 95°C for 1 min, primer annealing at 55°C for 1 min, and primer extension at 72°C for 3 min.

For aroG_Mme, the left arm was amplified using the primers P14/P15, the right arm was amplified using the primers P16/P17, and FRT-Km^r-FRT was amplified using the primers P18/P19. Primers P14 and P16 contain 18- and 19-nt regions, respectively, at their 5′ extremities that are homologous to the 5′ and 3′ terminals of the FRT-Km^r-FRT marker. The reaction was carried out as follows: denaturation at 95°C for 30 s, primer annealing at 48°C for 45 s, and primer extension at 72°C for 2 min. Finally, 50 ng of each resulting fragment was mixed together, and P15 and P17 were used for PCR amplification of the entire fragment. The reaction was carried out as follows: denaturation at 95°C for 1 min, primer annealing at 48°C for 1 min, and primer extension at 72°C for 6 min.

Sequencing.

The nucleotide sequencing of both strands was performed using an ABI Prism 3100 genetic analyzer with a BigDye Terminator v3.1 cycle sequencing kit (both from Applied Biosystems) according to the manufacturer's instructions. DNA sequencing data were analyzed using the Methylobacillus flagellatus KT genome database provided by GenBank (http://www.ncbi.nlm.nih.gov/nuccore/CP00284), and nucleotide and protein data were analyzed using BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi).

Elimination of FRT-Km^r-FRT marker from the chromosome.

pFLP31 was transferred into strain M. methylotrophus with the FRT-Km^r-FRT marker in the chromosome by mobilization (using plasmid-encoded Ap^r as the selective marker), and the Km^r Ap^r colonies were isolated. The isolated colonies were suspended at a final concentration of up to 10⁷ cells/ml in 5 ml of SEIIa with Ap liquid medium, and the suspension was heated at 42°C for 20 min, followed by incubation in a shaker at 37°C for 16 to 18 h to induce FLP recombinase synthesis. The culture was then plated on nonselective SEIIa medium to obtain individual colonies. These were analyzed for the occurrence of the Km and Ap markers on the solid SEIIa medium in the presence of the appropriate antibiotic. In a typical experiment, approximately 80% of the resulting colonies were Km^s due to the FLP-mediated excision of the FRT-Km^r-FRT marker.

Nucleotide sequence accession numbers.

All determined nucleotide sequences were deposited in the GenBank/EMBL/DDBJ databases. The accession numbers are as follows: GQ184460 for (trpEGDC)_Mme, GQ184461 for tyrA_Mme, and GQ184459 for (pheA-hisH-aroG)_Mme.

RESULTS

Unsuccessful attempts to isolate auxotrophic mutants of M. methylotrophus AS1 after MNNG treatment.

Auxotrophic mutants have been difficult to isolate from obligate methylotrophs and particularly M. methylotrophus AS1 using standard methods of mutagenesis (6, 19, 46). We attempted various forms of mutagenesis using N-methyl-N′-nitro-N-nitrosoguanidine (MNNG), not only the standard methods that are typically used for E. coli cells (44) but also methods with some specific modifications described previously for M. methylotrophus AS1 (23, 40) and for Methylobacillus flagellatus KT (62). No auxotrophs deficient in AroAA biosynthesis were found among 20,000 clones of the M. methylotrophus AS1 strain that were treated with MNNG in the same manner. One of the explanations of that failure was that corresponding AroAAs added to the medium might not permeate the bacterial cytoplasmic membrane to allow sufficient mutant growth (19, 25, 59).

To test this hypothesis, we carried out a set of experiments to determine the natural resistance of M. methylotrophus AS1 to such toxic compounds as analogs of AroAAs. It is well known that the addition of a high concentration of some amino acids (e.g., l-valine [Val]) can be detrimental to the growth of the strain due to the feedback inhibition of key enzymes participating in the biosynthesis of related amino acids (17). Thus, we tested the ability of M. methylotrophus AS1 to grow with excess Val in the medium.

During our experiments, we found that M. methylotrophus AS1 had a basal level of resistance to the 5-methyltryptophan that was significantly higher (more than a hundredfold) than typical for E. coli strains. Moreover, M. methylotrophus AS1 could grow in the presence of high concentrations of Val (up to 5 g/liter) that significantly exceeded values inhibitory for other Gram-negative bacteria. These results indirectly confirmed suggestions regarding low permeability of the cytoplasmic membrane of M. methylotrophus AS1 to amino acids and their analogs.

Introduction of E. coli aroP increased the sensitivity of M. methylotrophus AS1 to analogs of AroAAs.

We next introduced a recombinant plasmid with the E. coli gene aroP (aroP_Eco) in an attempt to increase the permeability of the AS1 membrane to AroAAs. As a member of the APC superfamily of transporters (69), AroP_Eco is the permease that transports Phe, Tyr, and Trp (11) across the inner membranes of E. coli bacteria using the proton motive force mechanism (10).

As a control, another E. coli gene, brnQ, was used. BrnQ_Eco has a highly similar ortholog in Salmonella enterica serovar Typhimurium that belongs to the LIVCS family and is thought to function as a sodium/branched-chain amino acid (BCAA) symporter (47). BrnQ_Eco likely corresponds to the LIV-II BCAA transport system, which has been shown to transport Leu, Val, and Ile in E. coli (2).

The pAYCTER-aroP_Eco and pAYCTER3-brnQ_Eco plasmids were constructed and mobilized into M. methylotrophus AS1 (see Materials and Methods). M. methylotrophus AS1 and two of its plasmid carrier derivatives possessing different E. coli transporter genes were plated on methanol-containing solid medium supplemented with increased concentrations of 5-methyltryptophan and Val as bacterial growth inhibitors (Table 2). The presence of aroP_Eco dramatically increased the sensitivity of M. methylotrophus cells to 5-methyltryptophan but had no influence on the high level of Val resistance. In contrast, brnQ_Eco carrier methylotrophic cells clearly manifested Val sensitivity in comparison to the control strain while retaining high resistance to the aromatics. Introduction of E. coli amino acid transporters therefore likely increased the permeability of the M. methylotrophus inner membrane for the corresponding amino acids or their analogs. This approach appeared promising for construction of auxotrophic mutants.

TABLE 2.

Resistance of M. methylotrophus AS1 strains to 5-methyl-tryptophan and Val^a

M. methylotrophus strain	Growth in the presence ofb:
	5-Methyltryptophan (mg/liter)				Val (g/liter)
	0	5	10	50	0	1	2.5	5
AS1 pAYCTER3	+	+	+	+	+	+	+	+
AS1 pAYCTER-aroPEco	+	−	−	−	+	+	+	+
AS1::aroPEco	+	−	−	−	+	+	+	+
AS1 pAYCTER-brnQEco	+	+	+	+	+	−	−	−

^aM. methylotrophus AS1 strains were grown on solid SEIIa medium.

^b+, growth; −, no growth.

We constructed a special recipient strain of M. methylotrophus AS1 with aroP_Eco inserted into the bacterial chromosome. The plasmid pMIV5-[FRT-Km^r-FRT]-Mob-aroP_Eco carrying the aroP_Eco gene and FLP-excisable Km^r marker in the mini-Mu unit was used to integrate aroP into the M. methylotrophus AS1 chromosome using the method of Abalakina et al. (1) based on the phage Mu-driven integration system. The presence of one copy of the integrated mini-Mu unit in the chromosome was confirmed by PCR and Southern hybridization (results not shown). The Km^r marker was excised by FLP recombinase as previously described (1), and one of the marker-less Mu integrants that retained sensitivity to 5-methyltryptophan (Table 2) was thereafter referred to as the AS1::aroP_Eco strain.

Cloning of trpEGD_Mme genes via functional complementation of the E. coli trpE mutant and inactivation of trpE in AS1::aroP_Eco.

For the cloning of different M. methylotrophus AS1 aromatic pathway genes, corresponding E. coli strains from the Keio collection were used, including in-frame, single-gene knockout mutants for all nonessential genes of E. coli K-12 (4).

Strain JW1256 from this collection carried the FLP-excisable Km^r marker inserted into and replacing the central part of the first gene from the trpEDCBA_Eco operon. The Km^r marker was excised, and the resulting E. coli JW1256-ΔtrpE strain was used as the recipient for isolation of trpE_Mme by complementation of auxotrophicity. The selected Tc^r, Trp⁺ transformants contained the plasmid pPW-(trpEGDC)_Mme with the inserted 5.2-kb DNA fragment (Fig. 1). The nucleotide sequence analysis of this fragment revealed three directly aligned open reading frames (ORFs) encoding proteins that had a high level of homology (77.3% identity for TrpE, 83.3% for TrpG, and 85.5% for TrpD) with the known TrpEGD subunits of anthranilate synthase, glutamine amidotransferase, and/or anthranilate phosphoribosyl transferase (EC 4.1.3.27/2.4.2.18) of Methylobacillus flagellatus KT.

trpE_Mme was disrupted by inserting a 1.5-kb DNA fragment containing FRT-Km^r-FRT amplified from pKD4 (16). This insertion was designed such that after integration into the M. methylotrophus AS1::aroP_Eco chromosome and FLP-mediated excision of Km^r, the “scar” sequence with the residual FRT site would inactivate the trpE_Mme gene by in-frame insertion, thereby likely preventing polar effects (Fig. 2A).

FIG. 2.

(A) Scheme of the trpE in-frame insertion mutant construction, structure of scar sequence, and in-frame fusion protein after FLP-mediated excision of the Km^r gene. Positions of the primers (P1 and P2) used for PCR amplification of Km^r from pKD4 and of the primers (P3 and P4) used for verification of the mutation are indicated with arrows. (B) PCR verification of the trpE mutation is as follows: lane 1, DNA ladder; lane 2, PCR of genomic DNA of M. methylotrophus AS1::aroP_Eco with primers P3/P4; lane 3, PCR of genomic DNA of M. methylotrophus AS1::aroP_Eco-(trpE_Mme::[FRT-Km^r-FRT]) with primers P3/P4; lane 4, PCR of genomic DNA of M. methylotrophus AS1::aroP_Eco-trpE_Mme::FRT with primers P1/P4.

M. methylotrophus AS1::aroP_Eco cells were electroporated with the linear DNA fragment containing trpE_Mme::[FRT-Km^r-FRT], with more than 2 kb of the methylotrophic DNA flanking the marker at both ends. Km^r clones were then selected on methanol-containing medium supplemented with 100 mg/liter Trp. As a control, the standard AS1 strain was used as the recipient. Approximately 100 Km^r clones auxotrophic for Trp were selected in a typical experiment with the AS1::aroP_Eco strain. We were unable to obtain any recombinant Km^r clones under the same conditions when the control strain was used. PCR analysis confirmed the inactivation of trpE_Mme due to recombination-mediated marker exchange in the chromosomes of 10 tested Km^r clones (Fig. 2B). The Km^r marker was removed from two clones of M. methylotrophus AS1::aroP_Eco-(trpE_Mme::[FRT-Km^r-FRT]) using an FLP recombinase expression plasmid. Eighty percent of the resulting clones were Km^s. PCR analysis confirmed the elimination of the Km^r marker and the presence of the 75-bp scar sequence in trpE_Mme in the chromosomes of 10 tested Km^s clones (Fig. 2). All obtained Km^s clones remained auxotrophic for Trp.

In summary, we were able to obtain the stable marker-less auxotrophic strain of M. methylotrophus AS1::aroP_Eco-trpE_Mme::FRT using (i) a specially constructed strain with the aroP_Eco transporter gene in the chromosome, (ii) linear DNA carrying trpE_Mme that was inactivated by excisable Km^r, (iii) recombination-mediated marker exchange between electroporated DNA synthesized by PCR and the bacterial chromosome, and (iv) FLP-mediated marker curing.

Cloning and inactivation of a set of genes involved in aromatic pathways and optimization of experimental procedures.

The same approach was used for the cloning and inactivation of other genes from aromatic pathways, and construction of auxotrophic mutants was based on the AS1::aroP_Eco strain. DNA fragments carrying tyrA_Mme, pheA_Mme, and aroG_Mme encoding chorismate mutase/prephenate dehydrogenase (EC 1.3.1.12/5.4.99.5), chorismate mutase/prephenate dehydratase (EC 4.2.1.51/5.4.99.5), and 3-deoxy-d-arabino-heptulosonate 7-phosphate synthase (DAHPS) (EC 2.5.1.54) were initially selected from the library of M. methylotrophus AS1 genes by complementation of the corresponding mutations from E. coli K-12 strains.

The Km^s variant of JW2581-ΔtyrA_Eco from the Keio collection (4) was used for tyrA_Mme identification. The plasmid pPW-(tyrA_Mme) complemented the ΔtyrA_Eco mutation and contained a 2.4-kb fragment (Fig. 1). An ORF that is highly homologous to TyrA from M. flagellatus KT (60.4% identity) was found in this fragment and disrupted by the insertion of the excisable Km^r marker. This was followed by electroporation of the linear DNA fragment into M. methylotrophus AS1::aroP_Eco. Km^r clones that could grow on methanol-containing medium supplemented with 100 mg/liter Tyr were selected, and this led to the isolation of the tyrA_Mme::[FRT-Km^r-FRT] mutant. The disrupted structure of the gene was confirmed by PCR with P5/P2 and P6/P7 primer pairs (data not shown).

The mutant construction procedure was optimized using the tyrA_Mme::[FRT-Km^r-FRT] linear DNA fragments, with the marker flanked by homologous DNA arms of different lengths. More than 10 of the desired clones were obtained per trial when the areas homologous to both target arms around the Km^r marker were ≥1,000 bp.

It was possible to design the linear DNA fragments for mutant construction using overlap extension PCR (26) with target arms of ≥1,000 bp. We later used this strategy for the construction of the pheA_Mme and aroG_Mme mutants (see below).

A double-auxotrophic M. methylotrophus AS1::aroP_Eco-trpE_Mme::FRT-tyrA_Mme::[FRT-Km^r-FRT] strain was also constructed. The disrupted tyrA_Mme-carrying DNA fragment was electroporated into the marker-less trpE_Mme mutant strain described above. Km^r clones grown on the methanol-containing medium supplemented with 100 mg/liter each of Trp and Tyr were then isolated.

pheA_Mme was isolated using E. coli JW2580-ΔpheA_Eco as the recipient strain for complementation. One of the recombinant plasmids from the Tc^r, Phe⁺-selected clones carried a rather large (4.7-kb) DNA fragment that contained three directly aligned ORFs according to sequence analysis (Fig. 1). These were homologous to PheA, HisH, and AroG from M. flagellatus KT (63.4%, 71.9%, and 73.5% identity, respectively), where the corresponding genes were located in different parts of the bacterial genome.

In addition to pheA_Mme, the putative aroG_Mme gene that likely encodes DAHPS attracted our attention because its ortholog, AroG_Eco, regulates the carbon flux into the general aromatic pathway in E. coli (50). Moreover, only one gene for DAHPS, aroG, was annotated in the M. flagellatus KT genome. To test whether the putative gene indeed had aroG_Mme function, a special E. coli strain was constructed. It is known that E. coli K-12 has three genes corresponding to the DAHPS isoenzymes aroG_Eco, aroF_Eco, and aroH_Eco; these genes code for proteins that are sensitive to feedback inhibition by Phe, Tyr, and Trp, respectively (50). Three strains from the Keio collection, JW0737 (ΔaroG_Eco::[FRT-Km^r-FRT]), JW2582 (ΔaroF_Eco::[FRT-Km^r-FRT]), and JW1694 (ΔaroH_Eco::[FRT-Km^r-FRT]), were used to construct a new marker-less, DAHPS-negative strain. We removed the Km^r marker from the chromosome of the initial recipient, JW0737, with consequent P1 duction of two marked deletions from other strains. At each step, FLP-mediated marker excision finalized the stepwise process of chromosomal rearrangements. The resulting strain, BW25113-ΔaroGFH_Eco, could grow on minimal medium supplemented with glucose, all AroAAs, and p-aminobenzoic acid (PABA). The plasmid pPW-(pheA-hisH-aroG)_Mme carrying the DNA fragment from the putative operon restored the growth of E. coli BW25113-ΔaroGFH_Eco on the minimal medium containing glucose; this was likely due to compensation of absent E. coli DAHPS function by the expression of aroG_Mme.

PCR-mediated overlap extension was used to construct two linear DNA fragments that were electroporated into M. methylotrophus AS1::aroP_Eco for inactivation of the pheA_Mme or aroG_Mme gene in the chromosome. For the electroporated DNA fragment containing ΔpheA_Mme::[FRT-Km^r-FRT], Km^r clones were obtained only on selective medium containing methanol supplemented with all AroAAs and PABA. We could not obtain any clones when Phe was the sole aromatic supplement. We propose that the polar effect of pheA_Mme disruption could have led to untimely transcription termination of the corresponding operon with aroG_Mme as the distal cistron. When the Km^r marker was excised, the residual scar restored the ORF in ΔpheA_Mme::FRT, and the resulting methylotrophic strain AS1::aroP_Eco-ΔpheA_Mme::FRT could grow on methanol-containing medium supplemented with 200 mg/liter Phe alone.

The DNA fragment with ΔaroG_Mme::[FRT-Km^r-FRT] was used for the construction of an auxotrophic mutant that could grow only when all essential aromatics were in the medium. Km^r removal did not change the nutrient demands of the marker-less strain. These results demonstrate unambiguously that only one gene encoding DAHPS, aroG_Mme, was present in the M. methylotrophus AS1 genome, as had also been noted earlier for M. flagellatus KT.

Thus, by using M. methylotrophus AS1::aroP_Eco as the recipient, a set of auxotrophic mutants that possess disrupted AroAA pathways was obtained using the optimized experimental procedure.

DISCUSSION

The isolation of auxotrophic mutants remains a rather complicated task for molecular genetic analysis of obligate methylotrophs (19). Various contradictory explanations have been offered for the failure of the mutagenesis of certain methylotrophic bacteria. Different groups have proposed either extremely high toxicity of the mutagens used or the failure of mutagens to permeate the cells (27). At the same time, a large group of experiments has directly or indirectly indicated that the inability of nutrient substances to permeate the methylotrophic membrane was the main factor that hampered the isolation of the desired auxotrophs using selective medium (19, 68). Indeed, the absence of intracellular transport of radioactively labeled Glu was convincingly demonstrated for M. methylotrophus AS1 (68).

It is therefore not surprising that successful auxotrophic mutant isolation was obtained when changes in bacterial membranes occurred due to unknown mutations. It is likely that these changes took place when the strains became resistant to antibiotics (40) or when the selection of auxotrophic mutants yielded mutants that were able to take up hydrophobic BCAA at high concentrations in the selective medium (62).

It is typically assumed that AroAAs can efficiently penetrate the bacterial cell membrane, as they are hydrophobic (41). Although their passive permeability is high, bacteria usually have additional uptake systems. Gram-negative E. coli possess the AroP_Eco system. This common transporter accepts all three AroAAs, as well as several substrate-specific transport systems (Mtr, TnaB, TyrP, PheP) (31, 50). A common AroP system for all AroAA uptake has been described for Gram-positive bacteria, in particular Corynebacterium glutamicum (30). At the same time, there are no indications concerning the presence of specific amino acid transporter genes in the known genomes of obligate methylotrophs.

Indeed, both significantly increased resistance of methylotrophic bacteria to 5-methyltryptophan and resistance to high concentrations of Val (which usually inhibits the growth of Gram-negative bacteria) were detected in our experiments. This resistance dramatically decreased after the introduction of the E. coli transporter genes aroP and brnQ in methylotrophic cells. Here, we detected specificity of the heterologous proteins for transported substrates: AroP_Eco had decreased resistance only to analogs of AroAAs, while BrnQ_Eco had decreased resistance only to Val.

Many E. coli genes are able to be expressed in M. methylotrophus AS1 under the control of their native regulatory regions (1). We therefore anticipated the possible transcription of aroP_Eco and brnQ_Eco in these methylotrophic cells and did not carry out preliminary modifications of gene transporter promoters, as this likely had to be done in another bacterial host. Nevertheless, the fact that the protein products of the cloned genes acted as specific transporters integrated in the heterologous bacterial membrane could not have been predicted in advance and was of experimental use.

We therefore recommend the introduction of foreign amino acid transporter genes for the isolation of other desired methylotrophic auxotrophs. It seems possible to detect the functional activity of the foreign transporter by observing decreased resistance in the resulting recombinant strains to the structural analogs of the corresponding amino acids. The approach developed in the present study may therefore be used for the isolation of mutants from different organisms.

The construction of mutants by recombination-mediated marker exchange was not original per se and has been used for different methylotrophs in addition to M. methylotrophus AS1 (6). This method included the introduction of suicide plasmid DNA that could not replicate in the corresponding bacterial host and the substitution of the plasmid marker for the corresponding target in the bacterial chromosome. If the selection pressure was absent, however, it was difficult to discriminate the rare clones that carried the desired substitutions from the majority of variants that obtained the marker via single-cross-mediated integration of the whole suicide plasmid into the target gene. Use of linear DNA for marker exchange increased the selectivity of the isolation of the desired recombinants because the marker could be inserted into the chromosome only through double-cross-mediated recombination.

Under standard conditions, linear DNA introduced into E. coli is degraded by the powerful RecBCD nuclease, and therefore it is necessary to either use specially constructed recipient strains, i.e., recBC and sbcB mutants (for an example, see reference 45), or inhibit the RecBCD and SbcCD nuclease using the λ Gam protein, thereby preserving linear DNA and allowing it to be used as a substrate for recombination (reviewed in references 12 and 53). The introduction of DNA into bacterial cells by electroporation, however, renders the preliminary inactivation of the RecBCD and SbcCD nuclease unnecessary for successful participation of the introduced linear DNA in recombination-mediated rearrangements of the bacterial chromosome, not only in E. coli but also in different Gram-negative bacteria (9, 38, 39).

In the present study, electroporation was used to allow recombination-mediated insertion of the linear DNA fragment into the chromosome of M. methylotrophus AS1. It was shown that a sufficient number of recombinant clones for genetic analysis could be isolated using standard experimental conditions if the length of the flanked homology around the selective marker and its target gene in the bacterial chromosome was ≥1,000 bp. Thus, construction of a linear DNA fragment for further insertion could be achieved in vitro using a PCR-mediated overlap extension procedure.

The synergy of the two experimentally achieved results, (i) obtaining the specialized recipient carrying the foreign transporter gene in its genome and (ii) optimization of the linear DNA fragment constructed for insertion mutagenesis, finally led to isolation of M. methylotrophus AS1-based auxotrophic mutants with disrupted AroAA biosynthetic pathways. These mutants and the methods that we developed for their construction were used later for the design of strains that overproduce Phe from methanol (32, 60; Yomantas et al., unpublished data).