Efficient Homofermentative l-(+)-Lactic Acid Production from Xylose by a Novel Lactic Acid Bacterium, Enterococcus mundtii QU 25

Abstract

Enterococcus mundtii QU 25, a newly isolated lactic acid bacterium, efficiently metabolized xylose into l-lactate. In batch fermentations, the strain produced 964 mM l-(+)-lactate from 691 mM xylose, with a yield of 1.41 mol/mol xylose consumed and an extremely high optical purity of ≥99.9% without acetate production.

Lactic acid bacteria (LAB) ferment sugars through different pathways, resulting in homo-, hetero-, or mixed acid fermentation. Few LAB strains produce lactate from xylose, the major constituent of hemicelluloses in lignocellulosic biomass and the second most abundant sugar, next to glucose, in nature (14). The phosphoketolase (PK) pathway in LAB converts 2 of the 5 carbons in xylose to acetic acid, increasing the purification cost of lactic acid (6, 16, 20). Hence, the maximum theoretical yield of lactate is limited to 1 mol/mol of xylose (0.6 g lactate per gram of xylose). Therefore, LAB strains using the PK pathway for xylose metabolism are not effective for the industrial-scale production of lactic acid. The lack of industrially suitable LAB for efficient conversion of xylose into lactate has been cited as a major technical obstacle for development of the poly(lactic acid) industry.

Lactococcus lactis IO-1 (JCM 7638), which was isolated and characterized in our laboratory, was reported to produce l-lactate at a high yield of 0.68 g/g of consumed xylose (exceeding 1.0 mol/mol). This result was achieved by fermenting xylose to lactate via 2 different pathways, the PK pathway and the pentose phosphate (PP)/glycolytic pathway (20). The PP/glycolytic pathway quantitatively converts xylose into a 3-carbon intermediate, pyruvate, providing the potential to produce 1.67 mol of lactate per mole of pentose (15, 16) (homolactate fermentation). In rare cases, homolactate fermentation of pentose is carried out by some streptococci. Fukui et al. (5) reported that a Streptococcus sp. (Enterococcus sp.) and Lactobacillus thermophilus T₁ metabolized d-xylose, l-arabinose, and d-ribose to only lactate under anaerobic culture conditions. In 1978, Barre (2) isolated 1 Lactobacillus species strain from wine, named Lactobacillus strain MONT4. This strain exclusively fermented arabinose and ribose, yielding dl-lactate. To the best of our knowledge, few researchers have performed studies on the development of engineered LAB strains to increase the lactate yield from xylose. Recently, Okano et al. (12, 13) used Lactobacillus plantarum NCIMB 8826. This strain is deficient in the l-lactate dehydrogenase (l-LDH) gene, and its PK genes have been replaced with a heterologous transketolase gene. Owing to this development, the L. plantarum ΔldhL1-xpk::tkt strain had been used to produce 41.2 g/liter (458 mM) of d-lactate from 46.4 g/liter (309 mM) xylose after 60 h of fermentation. However, to our knowledge, there has been no report on homo-l-(+)-lactate fermentation from xylose by LAB, in particular wild-type strains. Our laboratory has recently isolated a variety of new LAB from many natural sources. Here, we report the efficient fermentation of xylose to l-(+)-lactate with minimal by-products by one of these isolates, discovered in an ovine fecal sample collected from Fukuoka Zoo, Japan, and identified as Enterococcus mundtii QU 25 NITE BP-965 (1).

For growth of E. mundtii QU 25, 1 ml of glycerol stock was transferred into 9 ml of mMRS medium (MRS broth with xylose instead of glucose) supplemented with 1% xylose in a 15-ml screw-cap tube and incubated at 30°C for 24 h. A 4-ml aliquot of this culture was then transferred to a 50-ml vial containing 36 ml mMRS medium. The inoculum was incubated at 30°C for 8 h before inoculation at 10% (vol/vol) in a jar fermentor. Batch fermentations were carried out at 200 rpm in a 1-liter jar fermentor (Biott, Tokyo, Japan), with a working volume of 0.4 liters mMRS medium supplemented with 25 g/liter xylose (final concentration, 166 mM). Analysis of cell growth and dry cell weight (DCW) was evaluated as described previously (19). Xylose and fermentation products were determined using high-performance liquid chromatography as described previously (11, 19). The optical purity of the lactate product was measured using a BF-5 biosensor (Oji, Hyogo, Japan) according to the manufacturer's protocol. The purity of l-lactate was evaluated as follows: percent optical purity = (l-lactate concentration − d-lactate concentration)/(l-lactate concentration + d-lactate concentration) × 100. To prepare the crude cell extract for enzyme assays, cells cultured at a low (ca. 166 mM) or high (ca. 666 mM) xylose concentration for 12 h were harvested, suspended with the respective buffers, disrupted using a French pressure cell, and centrifuged at 7,190 × g for 15 min at 4°C. The obtained supernatants were used as crude cell extract. Activity of phosphoketolase in PK pathway was measured spectrophotometrically as ferric acetyl hydroxamate produced from the enzymatically generated acetyl phosphate at 43°C as previously described (10, 18). One unit of phosphoketolase activity is defined as the amount of enzyme catalyzing the formation of 1 μmol of acetyl phosphate per min from xyluose-5-phosphate. Activities of transaldolase and transketolase in PP/glycolytic pathway were measured at 43°C by coupling glyceraldehyde-3 phosphate production from either fructose-6-phosphate and erythrose-4-phosphate or ribulose-5-phosphate and xyluose-5-phosphate to NADH oxidation via triose-phosphate isomerase and glycerol phosphate dehydrogenase, as previously described (20). The enzyme activities were determined by subtracting NADH oxidase activity from the resulting rate of absorbance decrease at 340 nm. The activity of NADH oxidase was measured at 43°C in the same reaction buffer with transaldolase and transketolase containing NADH as the sole substrate. One unit of enzyme activity is defined as the amount of the enzyme catalyzing the formation of 1 μmol glyceraldehyde-3 phosphate for transaldolase and transketolase and 1 μmol NAD⁺ for NADH oxidase per min.

Under non-pH-controlled batch fermentations at 30°C, the rates of biomass accumulation and xylose consumption were very low, with a maximum lactate concentration of 35 mM after 24 h of cultivation (Table 1). The residual xylose concentration was 131 mM. The inefficient conversion of xylose to lactate may be attributed to the low pH (5.0 after 24 h of fermentation), which has a negative effect on cellular metabolism.

TABLE 1.

l-Lactate fermentation with xylose by Enterococcus mundtii QU 25 under different pHs and temperatures^a

Controlled pH	Temp (°C)	Mean ± SD maximum lactate concn (mM)	Time point (h)	Maximum lactate productivity (mM/h)
NA	30	35.0 ± 0.1	24	2.61
6.0	30	66.0 ± 7.6	24	4.35
6.5	30	126 ± 15.0	24	8.72
7.0	30	131 ± 1.6	24	12.1
7.5	30	87.0 ± 1.2	24	9.73
7.0	37	141 ± 20.0	24	14.5
7.0	41	155 ± 2.9	24	22.9
7.0	43	200 ± 3.1	16	21.9
7.0	45	168 ± 0.7	16	21.1
7.0	47	15.0 ± 0.1	24	1.95

^aThe xylose concentration at the beginning of fermentation is 166 mM. Averages with standard deviations are based on three independent fermentations. NA, not applicable.

pH-controlled fermentations were carried out at 30°C to reduce the inhibitory effects of free lactate on the producer cells according to the method described previously (21). pH 7.0 provided the highest lactate concentration produced (131 mM) and the maximum lactate productivity (12.1 mM/h), which increased by 274% and 365%, respectively, compared to the levels for non-pH-controlled batches (Table 1). Furthermore, by optimizing the cultivation temperature, the flux to lactate increased at elevated temperatures. When the temperature was increased from 30°C to 43°C at pH 7.0, the amount of lactate produced and the maximal lactate productivity rate were increased by 52% and 81% (200 mM and 21.9 mM/h), respectively (Table 1).

In an attempt to obtain higher lactate concentrations, fermentations were carried out at pH 7.0 and 43°C with 3 levels of initial xylose concentrations: 334 mM (50.1 g/liter) (Fig. 1 A), 480 mM (72.0 g/liter) (Fig. 1B), and 691 mM (103 g/liter) (Fig. 1C). Table 2 summarizes the results of batch cultures at various xylose concentrations. The optical purity of l-lactate was ≥99.9% at all xylose levels. With 334 mM xylose, fermentation was almost completed within 24 h, with a production level of 490 mM lactate at a high yield of 1.51 mol/mol of xylose consumed (0.9 g of lactate per gram of xylose consumed), which is substantially higher than the highest reported yield of lactate (1.37 mol/mol) from xylose by wild-type LAB (5). Very small amounts of by-products were detected, i.e., ≤17 mM acetic acid, ≤13 mM formic acid, and ≤34 mM ethanol. With 480 mM xylose, fermentation was almost completed at 48 h, with a production of 668 mM lactate at a yield of 1.47 mol/mol of xylose consumed (0.88 g of lactate per gram of xylose consumed). There was a remarkable decrease in by-product formation, with maximum concentrations of 12, 9, and 15 mM for acetic acid, formic acid, and ethanol, respectively. With the highest level of xylose tested (691 mM), complete fermentation of xylose to 964 mM (86.7 g/liter) lactate was achieved after 96 h of incubation. Lactate yield was calculated to be 1.41 mol/mol of xylose consumed (equivalent to 0.83 g of lactate per gram of xylose consumed), and no acetate was detected at the end of fermentation. The maximum specific productivities were almost the same at all xylose levels (Table 2). Furthermore, to investigate lactate production by strain QU 25 in nitrogen/vitamin-poor medium, we used a medium containing small amounts of nitrogen and vitamin sources (5 g/liter yeast extract) with the other components (K₂HPO₄, triammonium citrate, MgSO₄·7H₂O, MnSO₄·5H₂O, and Tween 80) instead of mMRS medium (10 g/liter peptone, 8 g/liter beef extract, 4 g/liter yeast extract). As the result, a slight decrease in lactate (771 ± 5.02 mM) was observed at high yield of 1.41 ± 0.009 mol/mol from 658 mM xylose under the optimum conditions, which indicated the economic feasibility of lactate production using strain QU 25.

FIG. 1.

Profiles of lactate fermentation with xylose by Enterococcus mundtii QU 25. Fermentations were conducted in a 1-liter jar fermentor with a 0.4-liter working volume at 43°C, pH 7.0, and 200 rpm. The initial xylose concentrations were 334 mM (A), 480 mM (B), and 691 mM (C). Symbols: open circles, xylose concentration; closed circles, lactate concentration; open triangles, acetic acid concentration; closed triangles, formic acid concentration; open squares, ethanol concentration; closed squares, dry cell weight. Data points represent the means and standard deviations of results from 3 independent experiments. The standard deviation is less than that corresponding to the size of the symbol if no error bars are seen.

TABLE 2.

Lactate fermentation with high concentrations of xylose by Enterococcus mundtii QU 25^a

Initial xylose concn (mM)	Maximum cell mass (g/liter)	Maximum lactate concn (mM)	Yield of products (mol product/mol xylose consumed)				Maximum specific productivityc (mmol/g [DCW] cells/h)	Maximum lactate productivity (mM/h)	Lactate yield/acetate yield (mol/mol)
			Lactic acidb	Acetic acid	Formic acid	Ethanol
334	5.24	490	1.51 ± 0.18	0.08 ± 0.01	0.03 ± 0.01	0.1 ± 0.03	16.2	42.7	18.3
480	4.47	668	1.47 ± 0.06	0.011 ± 0.004	0.04 ± 0.03	0.03 ± 0.02	14.0	32.1	134
691	3.56	964	1.41 ± 0.09	0.00	0.02 ± 0.01	0.009 ± 0.001	16.4	38.0	Unlimited

^aFermentations were done at 43°C and pH 7. Averages with standard deviations are based on three independent fermentations.

^bThe maximum theoretical yield for lactate by the pentose phosphate/glycolytic pathway (1.67 mol of lactate per mol of xylose).

^cThe maximum specific productivity for lactate.

Ilmen et al. (8) constructed a Pichia stipitis strain that produced 644 mM (58 g/liter) lactate, with a yield of up to 0.58 g/g of xylose. The recombinant Corynebacterium glutamicum strain produced lactate from xylose under anaerobic conditions, with a yield of 0.54 g/g (9). During our manuscript preparation, Wang et al. (23) reported a high concentration, 752 mM (67.7 g/liter), of lactate in batch fermentation by a newly isolated Bacillus sp. strain; however, this product did not show optical purity. In addition, they used very high concentrations of calcium carbonate (60% [wt/wt] of xylose) as a neutralizing agent, which would probably produce a considerable amount of calcium sulfate (gypsum) during the conversion of calcium lactate to free l-lactate in addition to the high consumption of sulfuric acid during this process. Moreover, the resulting gypsum poses considerable economic and ecological problems (3, 4, 17, 22). Furthermore, their strain showed substrate inhibition at a higher xylose concentration, with a sharp decrease in lactate production, and produced less than 333 mM (<30 g/liter) of lactate from 681 mM (102 g/liter) xylose. However, our strain produced a higher lactate concentration of 964 mM from 691 mM xylose, with very high optical purity (≥99.9%).

An analysis of products at the end of fermentation suggested a metabolic pathway of strain QU 25 from xylose. Strain QU 25 could be presumed to utilize the PP/glycolytic pathway, not the PK pathway, as the main pathway for xylose metabolism. This hypothesis was proposed because lactate represented 95 to 97% of the total fermentation products, and the yield of lactate to xylose reached up to 1.51 mol/mol (Table 2). This value is very close to the maximum theoretical yield obtained through the PP/glycolytic pathway (1.67 mol/mol) and is comparable to the 1.48-mol/mol yield of lactate produced via only the PP/glycolytic pathway by PK gene-disrupted L. plantarum mutant in homolactate fermentation with xylose (12). In addition, very small amounts of formic acid, acetic acid, and ethanol were produced during xylose fermentation by the QU 25 strain, which are probably formed by pyruvate formate lyase or pyruvate dehydrogenase, phosphotransacetylase/acetate kinase, and aldehyde and alcohol dehydrogenases, respectively, as previously reported (14, 20). Note that these low levels of by-products would probably result in lower purification cost (7). To prove our hypothesis, we investigated the key enzymes relating to PK and PP/glycolytic pathways in the respective cells grown at low and high xylose concentrations (166 mM and 666 mM). Phosphoketolase activity was detected with 17.7 ± 0.390 U/mg protein in cells grown at a low xylose concentration but not at a high xylose concentration. Higher activities of transaldolase and transketolase (0.377 ± 0.016 and 0.366 ± 0.057 U/mg protein, respectively) were detected in cells grown at a high xylose concentration than in those grown at a low xylose concentration (0.314 ± 0.011 and 0.286 ± 0.062 U/mg protein, respectively). These results indicate that only the PP/glycolytic pathway is used for xylose metabolism of strain QU 25 at a high xylose concentration; however, the PK and PP/glycolytic pathways are used for xylose catabolism at a low xylose concentration. To the best of our knowledge, no wild LAB strain was previously reported to utilize the PP pathway only for the fermentation of xylose, which facilitates the near-complete conversion of xylose into optically pure l-lactic acid.

In conclusion, the production of l-lactate from xylose by E. mundtii QU 25 is feasible and more efficient than that of any other strains reported thus far. This strain may provide an ideal wild-type microorganism for economical l-lactate production from renewable biomass substrates. It should be especially emphasized that this is the first report on homo-l-(+)-lactate fermentation with xylose by LAB.