Molecular Characterization of Two Lactate Dehydrogenase Genes with a Novel Structural Organization on the Genome of Lactobacillus sp. Strain MONT4

Abstract

Two lactate dehydrogenase (ldh) genes from Lactobacillus sp. strain MONT4 were cloned by complementation in Escherichia coli DC1368 (ldh pfl) and were sequenced. The sequence analysis revealed a novel genomic organization of the ldh genes. Subcloning of the individual ldh genes and their Northern blot analyses indicated that the genes are monocistronic.

Lactobacillus sp. strain MONT4 was isolated from high-temperature-fermenting grape musts (2). It is a unique organism due to its capability of fermenting l-arabinose and ribose exclusively to d(−)- and l(+)-lactate via a homolactic fermentation pathway (2). Unlike heterolactic fermentations of pentoses that yield equimolar amounts of lactic and acetic acids, homolactic fermentations of pentoses yield only lactic acid and are unknown among lactobacilli. Our long-range goal is to metabolically engineer Lactobacillus sp. strain MONT4 for ethanol production, as it perfectly meets the criteria essential in a commercial biomass-to-ethanol process (25). Metabolic engineering of Lactobacillus sp. strain MONT4 for ethanol production requires the introduction of ethanol production genes into this organism. Moreover, inactivation of the lactate dehydrogenase (ldh) genes is necessary to eliminate coproduct formation.

Escherichia coli (pfl ldh) lacks the ability to grow anaerobically on sugars and thus can be used to clone genes from fermentative pathways (1, 3, 6, 7, 8, 9, 10, 19, 21, 24). In order to clone the ldh genes, genomic DNA from Lactobacillus sp. strain MONT4 was partially digested with Sau3AI and was ligated to BamHI-digested pJDC9 (5). E. coli DC1368 transformants were selected on Luria-Bertani medium containing antibiotics (erythromycin, 1 mg per ml; kanamycin, 60 μg per ml; chloramphenicol, 60 μg per ml) after 48 h of incubation under anaerobic conditions. Four antibiotic-resistant transformants were obtained. One of them, designated E. coli DC1368(pUI100), contained a plasmid with a 5-kb insert (Table 1). When the cell extracts from the transformants were subjected to native polyacrylamide gel electrophoresis and lactate dehydrogenase (LDH) activity staining (11), E. coli DC1368(pUI100) expressed both l(+)-lactate dehydrogenase (LLDH) and d(−)-lactate dehydrogenase (DLDH) enzymatic activities (Fig. 1).

TABLE 1.

Bacterial strains and plasmids

Strain or plasmid	Relevant characteristics	Source or referencea
E. coli
E. coli DH5α		Invitrogen Life Technologies
E. coli DC1368	thr-1 leu-6 thi-6 lacY tonA22 rspL ldhA::Kan pfl::Cam	David P. Clark
E. coli DC1368(pUI100)	Emr, LLDH positive, DLDH positive	This study
E. coli DH5α(pJDC9::ldhD)	Emr, DLDH positive	This study
E. coli DH5α(pJDC9::ldhL)	Emr, LLDH positive	This study
Lactobacillus
Lactobacillus sp. strain MONT4 (DSMZ 20605)		DSMZ
L. acidophilus NRRL B-4495		USDA-ARS Culture Collection
L. amylovorus NRRL B-4538		USDA-ARS Culture Collection
L. brevis NRRL B-4527		USDA-ARS Culture Collection
L. buchneri NRRL B-1837		USDA-ARS Culture Collection
L. casei subsp. casei NRRL B-1922		USDA-ARS Culture Collection
L. coryniformis subsp. coryniformis NRRL B-4391		USDA-ARS Culture Collection
L. coryniformis subsp. torquens NRRL B-4390		USDA-ARS Culture Collection
L. curvatus subsp. curvatus NRRL B-4562		USDA-ARS Culture Collection
L. curvatus WSU		Stephanie Clark
L. delbrueckii subsp. delbrueckii NRRL B-763		USDA-ARS Culture Collection
L. delbrueckii subsp. lactis NRRL B-4525		USDA-ARS Culture Collection
L. farciminis NRRL B-4566		USDA-ARS Culture Collection
L. fermentum NRRL B-4524		USDA-ARS Culture Collection
L. fructosus NRRL B-1841		USDA-ARS Culture Collection
L. gasseri NRRL B-14168		USDA-ARS Culture Collection
L. helveticus NRRL B-4526		USDA-ARS Culture Collection
L. hilgardii NRRL B-1843		USDA-ARS Culture Collection
L. jensenii NRRL B-4550		USDA-ARS Culture Collection
L. malefermentans NRRL B-1953		USDA-ARS Culture Collection
L. mali NRRL B-4563		USDA-ARS Culture Collection
L. paracasei subsp. paracasei NRRL B-4560		USDA-ARS Culture Collection
L. pentosus NRRL B-227		USDA-ARS Culture Collection
L. plantarum NRRL B-4496		USDA-ARS Culture Collection
L. reuteri NRRL B-14171		USDA-ARS Culture Collection
L. rhamnosus NRRL B-442		USDA-ARS Culture Collection
L. ruminis NRRL B-14853		USDA-ARS Culture Collection
L. salivarius subsp. salicinicus NRRL B-1949		USDA-ARS Culture Collection
Plasmids
pJDC9	6.95 kb; EmrlacZ; strong transcriptional terminators	5
pUI100	11.9 kb; pJDC9::ldhD-ldhL; Emr	This study
pJDC9::ldhD	8.6 kb; Emr	This study
pJDC9::ldhL	8.2 kb; Emr	This study

^aDSMZ, German Collection of Microorganisms and Cell Cultures. USDA-ARS, U. S. Department of Agriculture—Agricultural Research Services.

FIG. 1.

Enzymatic activity staining of native polyacrylamide gels after electrophoresis of cell extracts from E. coli DC1368 and E. coli DC1368(pUI100) containing the cloned lactate dehydrogenase genes on a 4.687-bp genomic insert from Lactobacillus sp. strain MONT4. Three enzymatic staining methods have been employed: LLDH enzymatic activity staining using l(+)-lactic acid as substrate (A); DLDH enzymatic activity staining using d(−)-lactic acid as substrate (B); and LDH enzymatic activity staining using both l(+)-lactic acid and d(−)-lactic acid as substrates (C). Lanes 1, SeeBlue prestained standard (nine polypeptides from 4 to 250 kDa; Invitrogen Life Technologies); lanes 2, purified LDH enzyme (A and C, LLDH; B, DLDH; both from Boehringer Mannheim); lane 2a: purified DLDH; lanes 3, cell extract from E. coli DC1368; lanes 4, cell extract from E. coli DC1368(pUI100). White and black block arrows indicate the positions of purified LLDH and DLDH enzymes, respectively. Dotted and wavy-lined block arrows indicate the positions of Lactobacillus sp. strain MONT4 LLDH and DLDH enzymes, respectively, present in E. coli DC1368(pUI100) cell extracts. Thin-line arrows indicate the positions of the top four prestained standards (myosin [250 kDa], bovine serum albumin [98 kDa], glutamic dehydrogenase [64 kDa], and alcohol dehydrogenase [50 kDa]).

The DNA sequence of the genomic insert from pUI100 was determined (GenBank accession no. AY301012). Eleven open reading frames (ORF) were identified. A 1,002-bp ORF that potentially encodes a 334-amino-acid protein was identified as DLDH. BLAST searches revealed that the nucleotide and deduced amino acid sequences for this ORF have identities of 78 to 90% and 43 to 62%, respectively, to the sequences of published DLDH enzymes from other lactic acid bacteria (LAB) (3, 4, 10, 12, 14, 15, 16, 17) (Fig. 2). The nucleic acid and deduced amino acid sequence of the ORF was compared (http://www.jgi.doe.gov) to that of unpublished LAB genome sequences (Bifidobacterium longum DJO10A, Brevibacterium linens BL2 Bio, Lactobacillus brevis ATCC 367, Lactobacillus casei ATCC 334, Lactobacillus delbrueckii subsp. bulgaricus ATCCBAA-365, Lactobacillus gasseri ATCC 33323, Lactococcus lactis subsp. cremoris SK11, Leuconostoc mesenteroides subsp. mesenteroides LA81 [ATCC 8293], Oenococcus oeni PSU1 [ATCC BAA-331], Pediococcus pentosaceus ATCC 25745, and Streptococcus thermophilus LMD-9). At the amino acid level, a 25 to 60% identity and a 45 to 75% similarity within stretches of 244 to 332 amino acids were detected between the deduced amino sequence of the ldhD gene and the translated and annotated genome sequences of 11 LAB strains.

FIG. 2.

Alignment of the deduced amino acid sequence of ldhD from Lactobacillus sp. strain MONT4 and the published amino acid sequences of the ldhD genes from other LAB and LAB-like organisms. L., Lactobacillus; Leu., Leuconostoc; O., Oenococcus; P., Pediococcus. I, II, and III represent multiple amino acid sequences from the same species. Letter colors: black, dissimilar sequence; black with gray background, similar sequence; blue, conservative sequence; red with gray background, identical sequence; green, weakly similar sequence.

Downstream of ldhD, a 963-bp ORF that potentially encodes a 321-amino-acid protein was identified as LLDH. BLAST searches indicated that the nucleotide and deduced amino acid sequences for this ORF have identities of 80 to 85% and 59 to 76%, respectively, to the published LLDH enzymes from other LAB strains (8, 9, 13, 14, 18, 19, 20, 22, 23) (Fig. 3). Asp₁₆₈ and Ile₂₅₀ (amino acids involved in catalysis) and Arg₁₇₁, Gln₁₀₂, Thr₂₄₆, and Arg₁₀₉ (amino acids involved in substrate binding and specificity [8, 13]) were identified. BLAST searches were used to compare the nucleic acid and deduced amino acid sequences of the ORF to the unpublished genome sequences from 11 LAB strains. At the amino acid level, 33 to 75% identity and 52 to 89% similarity were detected, within stretches of 305 to 320 amino acids, between the deduced amino sequence of the ldhL gene and the translated and annotated genome sequences of nine LAB strains.

FIG. 3.

Alignment of the deduced amino acid sequence of ldhL from Lactobacillus sp. strain MONT4 and the published amino acid sequences of ldhL genes from other LAB and LAB-like organisms. B., Bifidobacterium; Lc., Lactococcus; L., Lactobacillus; P., Pediococcus; S., Streptococcus. I, II, and III represent multiple amino acid sequences from the same species. Letter colors: black, dissimilar sequence; black with gray background, similar sequence; blue, conservative sequence; red with gray background, identical sequence; green, weakly similar sequence.

Subcloning of the ORFs that potentially code for LDH enzymes was carried out. Two DNA fragments of 1,674 and 1,297 bp that include the ldhD and ldhL genes, respectively, were amplified, ligated to pJDC9, introduced into E. coli DH5α, and screened by blue-white screening. Two erythromycin-resistant white-colored transformants containing the subcloned ldhD and ldhL genes were designated E. coli DH5α(pJDC9::ldhD) and E. coli DH5α(pJDC9::ldhL), respectively (Table 1). E. coli DH5α(pJDC9::ldhD) expressed only DLDH activity, while E. coli DH5α(pJDC9::ldhL) expressed only LLDH activity.

As determined by DNA sequencing, the ldhD gene from Lactobacillus sp. strain MONT4 is located upstream of the ldhL gene. Such structural organization of the ldh genes on genomes of LAB strains has not been reported until now. When the unpublished genomes of 11 LAB strains were investigated, no ldhD or ldhL gene was found to be located upstream or downstream from each other in any of the 8 organisms that appeared to contain both ldhD and ldhL genes. Our goal is to understand the biological significance of this novel organization by generating and characterizing isogenic mutants.

Northern blots were performed to study the expression of the ldh genes as a function of growth using three probes (663-bp probe for ldhD, 819-bp probe for ldhL, and 1,755-bp probe for ldhD-ldhL). Low and high levels of transcription were observed at late lag and late exponential stages of growth, respectively. Expression of the ldhL genes in Lactobacillus spp. seems to be species dependent, as the ldhL from Lactobacillus helveticus revealed a high level of expression at the exponential phase of growth (22). The ldh genes from Lactobacillus sp. strain MONT4 were found to be monocistronic, with a transcript size of 1.1 kb each. The only polycistronic ldh gene from LAB strains is the one from Lactococcus lactis. It is part of an operon along with the genes encoding phosphofructokinase and pyruvate kinase enzymes (18).

Based on a neighbor-joining tree generated with 16S rRNA gene alignments, 27 strains representing 24 Lactobacillus species (Table 1) were selected for Southern blots. Low- and high-stringency Southern blots of the PstI-digested total genomic DNA were conducted using the ldhD, ldhL, and ldhD-ldhL probes. Unlike low-stringency conditions, no hybridization under high-stringency conditions was detected between the ldh probes and the genomic DNA from 24 Lactobacillus species, suggesting that differences exist among LDHs at the nucleotide level.

Our present research deals with the inactivation of the ldh genes and metabolic engineering of the ldh-deficient strains for ethanol production.