Introduction

Lactic acid is widely used in food, feed, cosmetic, and pharmaceutical industries [2, 19] as well as in producing poly lactic acid (PLA), one of the most promising biodegradable polymers [15]. PLA can occur as poly (L-lactic acid) (PLLA), poly (D-lactic acid) (PDLA) or a stereocomplex of PLLA and PDLA which has significantly improved mechanical performances, thermal resistance, and hydrolysis resistance compared to PLLA or PDLA [12, 15]. The demand for high optical purity of D-lactic acid is increasing due to its potential applications in making thermostable PLA. Therefore, it is of importance to produce optically pure D-lactic acid in a cost-effective way.

Optically pure lactic acids required for making PLA can be readily produced by microbial fermentation using different carbon sources [8, 14]. High optical purity of L-lactic acid has been commercially produced by microbial fermentation [8]. During microbial production of D-lactic acid, usually small amount of L-lactic acid is present. It has been reported that the optical purities of D-lactic acid were lower than 98.3 % using bacteria such as E. coli and Sporolactobacillus sp. [17, 18, 21]. It is thus important to obtain D-lactic acid with very high purity because the presence of a small amount of L-lactic acid would severely affect its commercial applications. For example, the optical purity of lactic acid is a crucial parameter for controlling the physical and thermochemical properties as well as the biodegradation rate of PLA [15].

Separation of a small amount of L-lactic acid from D-lactic acid in a cost-effective way is very difficult using the conventional methods such as chromatography and electrophoresis [6, 10]. It has been reported that mutagenesis of the strains can improve the optical purity of D-lactic acid. High yield of D-lactic acid with high optical purity can be obtained from a genetically modified strain under oxygen deprivation conditions [9]. If D-lactic acid or its polymer is to be used in food or personal care industries, microbes used for producing D-lactic acid might need to be carefully selected. Usually, non-GMO strains are preferred for the D-lactic acid production. Therefore, it is useful to explore an alternative way to provide D-lactic acid with high optical purity. Studies have shown that lactic acids can be produced by bacteria using diverse carbon resources [20]. It would be useful if bacterial stains can use economic carbon sources such as lignocelluloses and starch. If the carbon sources are from lignocellulose, acetic acid is unavoidably generated during the pre-treatment as a result of hydrolysis of the highly acetylated hemicellulose [5, 7]. In this case, the removal of acetic acid from lactic acid products in a cost-effective way is also very important.

Here, we report a method for producing optically pure D-lactic acid by the combined use of two native bacterial strains that we isolated from the natural environment in Singapore. An additional advantage is that acetic acid, which is usually produced in case of hetero-fermentation or unavoidably existed in lignocellulose hydrolysate, can also be efficiently removed by this method.

Materials and Methods

Microorganisms and Culture Media

Bacillus sp. ADS3 and Weissella sp. S26 were isolated from the soil samples collected in Singapore. Soil samples were mixed with autoclaved water. Water was then plated on LB plates. Colonies were inoculated into 2 mL of LB medium and cultured at 37 °C for 12 h. Any colony that can reduce the pH of the medium was further purified. Among these screened bacteria, two of them showed the ability to produce lactic acids, and they were identified as Weissella sp. and Bacillus sp., respectively, using the conventional bacterial identification methods. Weissella sp. S26 is capable of producing lactic acid rich in D-isomer (i.e., 95 %). Bacillus sp. ADS3 is able to preferentially digest acetic acid and L-lactic acid in the mixture of acetic acid, L- and D-lactic acids.

The agar plates for cultivating Bacillus sp. ADS3 contained (per liter) 20 g sodium acetate, 5.0 g (NH4)2SO4, 5 g KNO3, 2.0 g NaH2PO4, 0.1 g MgSO4·7H2O, 0.1 g MnSO4·4H2O, 0.1 g FeSO4·7H2O, 1.0 g yeast extract (BD), and 20.0 g agar (BD), with pH 7.0. The same medium containing no agar was used for cultivating Bacillus sp. ADS3 in submerged cultures. Bacillus sp. ADS3 was cultivated in 1-L flasks at 30 °C and 150 rpm.

The modified MRS medium was used for preparing seed culture of Weissella sp. S26. It contained (per liter) 10.0 g peptone, 8.0 g meat extract, 4.0 g yeast extract (BD), 20.0 g glucose (or xylose), 2.1 g sodium chloride, 1.0 g Tween 80, 2.0 g K2HPO4, 2.0 g triammonium citrate, 0.2 g MgSO4·7H2O, and 0.05 g MnSO4·4H2O, with pH 6.2.

Three media were used for producing D-lactic acid by Weissella sp. S26. The first one was glucose or xylose medium containing (1 L) 20.0 g or 50.0 g of glucose or xylose, 10.0 g peptone, 4.0 g yeast extract, and 2.0 g triammonium citrate, with pH 6.6. The second medium was starch-MRC medium containing (1 L) 10.0 g peptone, 8.0 g meat extract, 4.0 g yeast extract, 1.0 g Tween 80, 2.0 g K2HPO4, 2.0 g triammonium citrate, 0.2 g MgSO4·7H2O, 0.05 g MnSO4·4H2O, 2.1 g NaCl, and 50 g starch, with pH 6.0. The third medium was the same with the second one except that the tapioca starch (Ng Nam Bee Marketing Pte. Ltd, Thailand) was processed before adding to the mixture by hydrolyzing 50 g of starch in 500 mL of sterilized water using 0.5 mL of α-amylase (liquozyme, Novozymes) and 5 mL of glucoamylase (Spirizyme, Novozymes) at 85 °C for 2 h.

Production of D-Lactic Acid-Rich Product by Weissella sp. S26

Weissella sp. S26 seed was prepared by inoculating Weissella sp. S26 into 100 mL MRS medium in 500 mL flasks followed by shaking at 37 °C for 28 h. The mixture was centrifuged at 8000×g for 15 min, and the pellets were collected, re-suspended, and inoculated into a 2-L fermenter containing 1 L medium. When the second medium was used, 0.5 mL of liquozyme and 5 mL of Spirizyme were added into the fermenter together with the seed inoculum. The fermentation was carried out at 37 °C, 200 rpm with a N2 flow rate of 0.2 vvm. The pH was maintained at 6.0 by automatic addition of 5.0 M NaOH or 10.0 M HCl during the fermentation. Samples were regularly collected for analysis during fermentation. After completion of the fermentation, the broth was heated at 100 °C for 20 min to kill the bacterial cells, and the pH was then adjusted to 8.0 with 5.0 M NaOH. The resulting product was used for further treatment by Bacillus sp. ADS3 to remove acetic acid and L-lactic acid.

Preparation of Bacillus sp. ADS3 Cells

Five to ten colonies of Bacillus sp. ADS3 were picked up from the agar plate and inoculated into 20 mL of liquid medium which was then incubated at 37 °C, 200 rpm overnight. The culture was transferred into a 200-mL medium and incubated at 37 °C, 200 rpm overnight. The culture was collected by centrifugation at 8000×g for 15 min. The cell pellets were ready for use in the subsequent steps for removing acetic acid and L-lactic acid.

Removal of L-Lactic Acid from the Fermentation Broth Rich in D-Lactic Acid by Bacillus sp. S3

Bacillus sp. ADS3 pellets from the 200-mL culture were added into 1 L of sterilized Weissella sp. S26 fermentation broth. The mixture was incubated at 30 °C, 150 rpm for several days. Samples were collected regularly for analysis by HPLC.

Degradation of Pure Acetic Acid or L-Lactic Acid by Bacillus sp. ADS3

To investigate the mechanism of acetic acid and L-lactic acid removal by Bacillus sp. ADS3, 100 mL of acetic acid (5 g/L) or L-lactic acid (5 g/L) whose pH was adjusted to 8.0 by 5 M NaOH was added into 200 g of wet Bacillus sp. ADS3 cells which were produced by cultivating in 200 mL of liquid medium for 16 h followed by centrifugation at 8000×g for 15 min. The mixture was stirred at 30 °C, 200 rpm continuously, and liquid samples were taken regularly for HPLC analysis. After completion of the conversion, the cells were harvested by centrifugation, and production of polyhydroxyalkanoates (PHAs) was analyzed. The supernatant was analyzed by LC/MS to detect the nominal masses of metabolites produced during the acetic acid or lactic acid digestion.

Analytical Methods

The optical purity of lactic acid was detected by HPLC (LC-10AT, refractive index detector SPD-10A, Shimadzu, Kyoto, Japan) with a Mitsubishi MCI® Gel CRS15W column (50 × 4.6 mm, Mitsubishi Chemicals, Tokyo, Japan) at 30 °C using 2 mM of CuSO4 as the mobile phase at 0.4 mL/min. The optical purity of D-lactic acid was calculated based on the following equation [20]:

$$ \mathrm{Optical}\kern0.5em \mathrm{purity}=\frac{\mathrm{D}-\mathrm{lactic}\kern0.5em \mathrm{acid}-\mathrm{L}-\mathrm{lactic}\kern0.5em \mathrm{acid}}{\mathrm{D}-\mathrm{lactic}\kern0.5em \mathrm{acid}+\mathrm{L}-\mathrm{lactic}\kern0.5em \mathrm{acid}}\times 100\% $$

The concentrations of xylose, glucose, and acetic acid were analyzed by a Shimadzu HPLC system equipped with a Bio-Rad Aminex HPX-87H column (300 × 7.8 mm, Bio-Rad, Herculse, CA, USA). A standard curve was first plotted using pure xylose, glucose, or acetic acid. Samples during fermentation were first cleared by centrifugation at 8000×g, 4 °C for 10 min before being injected to the column for analysis.

Results

Production of D-Lactic Acid from Glucose by Combined use of Weissella sp. S26 and Bacillus sp. ADS3

Weissella sp. S26 can produce lactic acid rich in D-isomer from various sugars. As the types and concentrations of sugars may affect the modes of lactic acid fermentation, we tested the lactic acid fermentation using five different media containing 2 % glucose, 5 % glucose, raw starch, starch hydrolysate, and 2 % of xylose, respectively.

When 2 % of glucose was used for the lactic acid production by Weissella sp. S26, the final lactic acid titer reached 15 g/L at 20 h. Acetic acid was also detected at 1.3 g/L (Fig. 1a). After adding Bacillus sp. ADS3 cells into the sterilized fermentation broth, both acetic acid and L-lactic acid started to be degraded (Fig. 1b). Acetic acid was completely degraded within 45 h when the D-lactic acid optical purity reached 96.6 % (Fig. 1b). When glucose (5 %) was used for the fermentation (Fig. 2), the lactic acid titer reached 31.1 g/L at 26 h. Acetic acid was completely removed at 60 h when the D-lactic acid optical purity reached 98.2 % (Fig. 2b).

Fig. 1
figure 1

Time courses of lactic acid production by Weissella sp. S26 (a) and removal of L-lactic acid and acetic acid from the fermentation broth by Bacillus sp. ADS3 (b). After fermentation, the broth was sterilized at 100 °C for 20 min and cooled down to 30 °C before adding Bacillus sp. ADS3 cells

Fig. 2
figure 2

Time courses of D-lactic acid production by Weissella sp. S26 using 5 % glucose (a) and removal of L-lactic acid from the fermentation broth by Bacillus sp. ADS3 (b). After fermentation, the broth was sterilized at 100 °C for 20 min and cooled down to 30 °C before adding Bacillus sp. ADS3 cells

Production of D-Lactic Acid from Starch by Combined use of Weissella sp. S26 and Bacillus sp. ADS3

Starch was used as a cheaper carbon source to produce lactic acid in two cases, simultaneous saccharification and fermentation and saccharification followed by fermentation.

For the simultaneous saccharification and fermentation, the starch was added into the medium together with liquozyme and Spirizyme to start the process. The lactic acid titer reached 24.2 g/L at 37 h (Fig. 3a). Longer fermentation time did not improve the lactic acid production, which might be attributed to the fact that the cell density reached the maximum at which point bacterial cells do not produce lactic acid. The acetic acid concentration reached 11.8 g/L at 67 h (Fig. 3a). After sterilization, the addition of Bacillus sp. ADS3 cells completely removed the acetic acid within 115 h (Fig. 3b) when the D-lactic acid optical purity reached up to 99.5 % (Fig. 3b).

Fig. 3
figure 3

Time courses of D-lactic acid production by Weissella sp. S26 in simultaneous saccharification and fermentation of starch (a) and removal of L-lactic acid and acetic acid from the fermentation broth by Bacillus sp. ADS3 (b). After fermentation, the broth was sterilized at 100 °C for 20 min and cooled down to 30 °C before adding Bacillus sp. ADS3 cells

In the case of starch saccharification followed by fermentation, the lactic acid titer reached 24 g/L at 23 h, while the acetic acid concentration continuously increased and reached 5.8 g/L at 45 h (Fig. 4a). After sterilization, the addition of Bacillus sp. ADS3 cells completely removed the acetic acid at 90 h (Fig. 4b) when the optical purity of D-lactic acid reached up to 99.5 %.

Fig. 4
figure 4

Time courses of D-lactic acid production by Weissella sp. S26 in saccharification followed by fermentation of starch (a) and removal of L-lactic acid and acetic acid from the fermentation broth by Bacillus sp. ADS3 (b). After fermentation, the broth was sterilized at 100 °C for 20 min and cooled down to 30 °C before adding Bacillus sp. ADS3 cells

Production of D-Lactic Acid from Xylose by Combined use of Weissella sp. S26 and Bacillus sp. ADS3

Xylose is a potential carbon source for industrial production of lactic acid as it is rich in lignocellulose. When 2 % of xylose was used, the lactic acid concentration reached 13.7 g/L at 48 h with 13.2 g/L of D-lactic acid and 8.5 g/L of acetic acid (Fig. 5a). The broth was sterilized and treated with Bacillus sp. ADS3 cells. The acetic acid was completely removed at 120 h when the optical purity of D-lactic acid reached 98.5 % (Fig. 5b).

Fig. 5
figure 5

Time courses of D-lactic acid production from xylose by Weissella sp. S26 (a) and removal of L-lactic acid and acetic acid from the fermentation broth by Bacillus sp. ADS3 (b). After fermentation, the broth was sterilized at 100 °C for 20 min and cooled down to 30 °C before adding Bacillus sp. ADS3 cells

Discussion

D-lactic acid has not yet been commercially produced, although its market demand is increasing due to the production of thermostable PLA. A lot of researches have been reported for D-lactic acid production, but almost all of them used genetically modified strains. In many cases, a small amount of L-lactic acid is still contained in the D-lactic acid product after the fermentation. The removal of this small amount of L-lactic acid from D-lactic acid in a cost-effective way is very difficult [6, 10]. Genetically modified Corynebacterium glutamicum was shown to be able to produce D-lactic acid in a minimal medium in 30 h, which showed potential application of such strains [9]. In this study, we show a method that can produce D-lactic acid using economic carbon sources and non-genetically modified bacteria. This method will be useful if the D-lactic acid or its product is used in food industry or using economic carbon resources as a carbon source for fermentation. Our method can not only remove L-lactic acid from D-lactic acid, but also simultaneously remove acetic acid from the lactic acid product. The contamination of lactic acid product by acetic acid is usually unavoidable especially when the sugars are from lignocellulose hydrolysate in which acetic acid is inevitably existed as a result of the hydrolysis of the acetylated hemicellulose. It has been noted that the processing time is long in the current study especially since economic carbon sources are used (Fig. 2). Recycling of bacterial cells may be a way to reduce the operation cost. This method is advantageous in terms of its simple operation and simultaneous removal of acetic acid and L-lactic acid. Although it needs some time in the step of L-lactic acid removal, no special equipment is required. Bacillus sp ADS3 can also be used to process the fermentation product of other D-lactic acid-producing bacteria. It is expected to be a cost-effective and commercially competitive method for producing D-lactic acid from various sugars especially when the sugars are from lignocellulose where the removal of acetic acid is essential. The D-lactic acid produced by the combined use of two native strains is also very suitable for applications where non-GMO products are needed, such as in food and personal care industries.

It is worth mentioning that although Bacillus sp. ADS3 is capable of digesting both acetic acid and lactic acid, it preferentially digests acetic acid if both are available. For the mixture of L- and D-lactic acids, it preferentially digests L-lactic acid than D-lactic acid, possibly due to the less toxicity of L-lactic acid than of D-lactic acid [11]. Therefore, stopping the digestion reactions at a right time is crucial for getting high purity of lactic acid free of acetic acid or high optical purity of D-lactic acid free of L-lactic acid.

Regarding the mechanism of acetic acid and L-lactic acid removal by Bacillus sp. ADS3, we preliminarily presumed that polyhydroxyalkanoates (PHAs) might be enriched in the cell biomass, which are usually accumulated by some bacteria as an intracellular energy reserve under nutrient-starvation conditions or even under nutrient non-limiting conditions by some bacteria such as Ralstonia eutropha and Alcaligenes latus [13, 17]. It has also been reported that some organic acids such as acetic acid, propionic acid, and lactic acid can be used by PHA-producing strains for PHA synthesis [16]. Some Bacillus species such as Bacillus thuringiensis R1 [3], Bacillus megaterium [4], Bacillus subtilis, Bacillus firmus, Bacillus sphaericus, and Bacillus pumilus [1] have been reported to accumulate PHAs in various amounts. However, after analyzing the compositions of the cell biomass, we could not detect any production of PHAs in both cases of pure acetic acid and pure L-lactic acid degradation by Bacillus sp. ADS3. However, we did detect the production of larger amounts of HCO3 in the liquid phases by LC/MS analysis (data not shown). Therefore, it is highly possible that both acetic acid and lactic acid are converted to CO2 by Bacillus sp. ADS3. Further investigation of the mechanism is still needed to have a better understanding of the biological process.

In summary, we show that Weissella sp. S26 can produce lactic acid using various carbon sources. D-lactic acid with high optical purity can be obtained by combined use of Weissella sp. S26 and Bacillus sp. ADS3. Our study provides an economic way to produce high quality D-lactic acid.