www.elsevier.com/locate/jbiosc

Journal of Bioscience and BioengineeringVOL. xx No. xx, 1e6, 2014

Fed-batch fermentation for enhanced lactic acid production from glucose/xylosemixture without carbon catabolite repression

Mohamed Ali Abdel-Rahman,1,2 Yaotian Xiao,1 Yukihiro Tashiro,3,4 Ying Wang,1 Takeshi Zendo,1

Kenji Sakai,3 and Kenji Sonomoto1,5,*

Laboratory of Microbial Technology, Division of Systems Bioengineering, Department of Bioscience and Biotechnology, Faculty of Agriculture, Graduate School, Kyushu University,6-10-1 Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan,1 Botany and Microbiology Department, Faculty of Science (Boys), Al-Azhar University, Nasr City PN:11884, Cairo,

Egypt,2 Laboratory of Soil Microbiology, Division of Systems Bioengineering, Department of Bioscience and Biotechnology, Faculty of Agriculture, Graduate School,Kyushu University, 6-10-1 Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan,3 Institute of Advanced Study, Kyushu University, 6-10-1 Hakozaki, Higashi-ku,

Fukuoka 812-8581, Japan,4 and Laboratory of Functional Food Design, Department of Functional Metabolic Design, Bio-Architecture Center,Kyushu University, 6-10-1 Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan5

Received 5 June 2014; accepted 18 July 2014Available online xxx

* Correspondtems BioenginAgriculture, GFukuoka 812-8

E-mail adhotmail.comgmail.com (Y.u.ac.jp (K. Sak

1389-1723/$http://dx.doi

Please citemixture wi

There has been tremendous growth in the production of optically pure L-lactic acid from lignocellulose-derivedsugars. In this study, Enterococcus mundtii QU 25 was used to ferment a glucose/xylose mixture to L-lactic acid. Main-tenance of the xylose concentration at greater than 10 g/L achieved homo-lactic acid fermentation and reduced theformation of byproducts. Furthermore, carbon catabolite repression (CCR) was avoided by maintaining the glucoseconcentration below 25 g/L; therefore, initial concentrations of 25 g/L glucose and 50 g/L xylose were selected. Sup-plementation with 5 g/L yeast extract enhanced the maximum xylose consumption rate and consequently increasedlactic acid production and productivity. Finally, a 129 g/L lactic acid without byproducts was obtained with a maximumlactic acid productivity of 5.60 g/(L$h) in fed-batch fermentation with feeding a glucose/xylose mixture using ammo-nium hydroxide as the neutralizing agent. These results indicate a potential for lactic acid production from glucose andxylose as the main components of lignocellulosic biomasses.

� 2014, The Society for Biotechnology, Japan. All rights reserved.

[Key words: Glucose/xylose mixture; L-Lactic acid production; Fed-batch fermentation; Carbon catabolite repression; Enterococcus mundtii QU 25]

Lactic acid has a wide application in food, cosmetic, pharma-ceutical, textile, leather, and polylactic acid industries (1). Lacticacid has two isomers, L(þ) and D(�) lactic acids. Optically pure L- orD-lactic acid is more valuable than racemic DL-lactic acid becausethe physical properties and degradability of polylactic acid aredependent on the stereochemistry of the individual lactic acidisomers (2). In addition, L-lactic acid is the preferred isomer in foodpreservation and pharmaceutical industries as humans have only L-lactate dehydrogenase whereas D-lactic acid is often harmful tohuman metabolism (3).

Industrial production of lactic acid is achieved via chemicalsynthesis ormicrobial fermentation. Optically pure isomers of lacticacid can be produced only by fermentation process. Currently, lacticacid production by fermentation uses food feedstock such asstarches and sugars (1), which could cause food shortages and priceincreases. Therefore, use of a non-food feedstock such as lignocel-lulosic biomass offers a non-competitive, cost-attractive, and envi-ronmentally friendly substrate for the production of lactic acid (4).

ing author at: Laboratory of Microbial Technology, Division of Sys-eering, Department of Bioscience and Biotechnology, Faculty ofraduate School, Kyushu University, 6-10-1 Hakozaki, Higashi-ku,581, Japan. Tel./fax: þ81 92 642 3019.dresses: mohamedali@kyudai.jp (M.A. Abdel-Rahman), vacrite@(Y. Xiao), tashiro@agr.kyushu-u.ac.jp (Y. Tashiro), wangyingcqu@Wang), zendo@agr.kyushu-u.ac.jp (T. Zendo), kensak@agr.kyushu-ai), sonomoto@agr.kyushu-u.ac.jp (K. Sonomoto).

e see front matter � 2014, The Society for Biotechnology, Japan..org/10.1016/j.jbiosc.2014.07.007

this article in press as: Abdel-Rahman, M. A., et al., Fed-batch fthout carbon catabolite repression, J. Biosci. Bioeng., (2014), h

Ninety percent of the global plant biomass is lignocellulosic,amounting to approximately 200 � 109 tons per year. From this,8e20 � 109 tons remain potentially accessible (5). Lignocellulosicbiomass consists mainly of cellulose, hemicellulose, and lignin.Cellulose is a homo polymer composed of glucose subunits, whilehemicellulose is heterogeneous, mainly composed of xylose sub-units, and considered to be the second most abundant sugar innature after glucose (6,7). The percentage of cellulose and hemi-cellulose in lignocellulosic biomass are dependent on the type oflignocellulosic biomass, i.e., cottonseed hairs contain 80e95% cel-lulose and 5e20% hemicellulose; leaves contain 15e20% celluloseand 80e85% hemicellulose; and grasses contain 25e40% celluloseand 25e50% hemicellulose (4,6). Complete hydrolysates derivedfrom various lignocellulosic biomasses show different ratios ofglucose and xylose. However, effective utilization of lignocellulosicbiomass for lactic acid production is hindered by a number of dif-ficulties, mainly because most lactic acid bacteria (LAB) cannotutilize xylose. In addition, carbon catabolite repression (CCR) isobserved in most LAB, where preferred sugars such as glucoserepress the utilization of other non-favourable sugars such asxylose and arabinose (8,9). Furthermore, the formation of byprod-ucts limits the yield of lactic acid and increases the purification cost(4,10).

Interestingly, Enterococcus mundtii QU 25, a LAB strain isolatedin our laboratory, can efficiently ferment either glucose (11) orxylose (12) as the sole carbon source to produce optically pure

All rights reserved.

ermentation for enhanced lactic acid production from glucose/xylosettp://dx.doi.org/10.1016/j.jbiosc.2014.07.007

2 ABDEL-RAHMAN ET AL. J. BIOSCI. BIOENG.,

L-lactic acid under optimal conditions of pH 7.0 and 43�C. It hasbeen reported to utilize the pentose phosphate pathway at highxylose concentrations, and produce lactic acid homofermentativelywith minimal byproduct formation (12). Therefore, this studyaimed to produce homofermentative optically pure L-lactic acidfrom a glucose/xylose mixture using E. mundtii QU 25 and optimizeconditions for the efficient fermentation of lactic acid.

MATERIALS AND METHODS

Strain and medium E. mundtii QU 25 NITE BP-965 was used in this study(13) and stored at �80�C in 2-mL vials containing 15% glycerol until use. ModifiedMan, Rogosa, and Sharpe (mMRS) medium was used for cell growth, inoculumpreparation, and fermentation, as previously reported (13). Glucose/xylosemixture was added to mMRS medium as indicated in each experiment. The pH ofthe medium was adjusted to 7.0 using 1 M NaOH.

Effect of xylose concentration on byproduct formation OnemL of glycerolstock was inoculated to themedium at 10% (v/v) and incubated for 24 h at 43�C. FourmL of this fermentation broth was transferred to 50-mL plastic tubes containing36 mLmedium and cultivated for 12 h. The pre-culture broth was then inoculated at10% (v/v) in a 1-L jar fermentor (Biott, Tokyo, Japan) containing a 0.4-L workingvolume of mMRSmediumwith either 10 g/L glucose/50 g/L xylose (G10X50) or 10 g/L glucose/10 g/L xylose (G10X10). Batch fermentations were carried out at 43�C and200 rpm. The pH was controlled at 7.0 by automatic addition of 10 M NaOH duringthe fermentation. Five mL samples were removed at different time intervals and cellgrowth and sugar and fermentation product were analysed.

Effect of different ratios of glucose and xylose in the sugar mixtures onCCR In this experiment, different concentrations of the glucose/xylose mixturein mMRSmediumwere used: 25 g/L glucose/50 g/L xylose (G25X50), 50 g/L glucose/50 g/L xylose (G50X50), and 50 g/L glucose/25 g/L xylose (G50X25). Batch fer-mentations were carried out using a 1-L jar fermentor with a 0.4-L working volumeunder the same conditions as described above.

Effect of nitrogen source supplementation on lactic acid fermentation Inthis experiment, a 5 g/L yeast extract or 5 g/L ammonium sulphate was added tomMRS medium containing G25X50. Batch fermentations were performed in a 1-Ljar fermentor under the same conditions described above.

Fed-batch fermentation with different neutralizing agents In thisexperiment, fed-batch fermentation was performed in mMRS medium (G25X50)supplemented with 5 g/L yeast extract with the pH controlled at 7.0 by 10 M sodiumhydroxide or 10 M ammonium hydroxide. Mixtures of 25 g/L glucose/50 g/L xyloseand 5 g/L yeast extract were fed into the jar fermentor at 27 h and 54 h to maintainthe concentrations of xylose and glucose at greater than 10 g/L and lower than 25 g/L,respectively.

Enzyme assays To prepare crude extracts, batch fermentations were per-formed in a 1-L jar fermentor using mMRS medium containing either G25X50 orG50X50 with the pH controlled at 7.0 by 10 M sodium hydroxide. Cells were har-vested at 6 h (glucose remaining) and 26 h (glucose exhausted), and disrupted by 3passages through a French Press Cell (Thermo Fisher Scientific, Waltam, MA, USA) at20,000 psi and centrifuged at 7190 �g for 15 min at 4�C. The supernatant was usedas crude cell extract for the assays of xylose isomerase (XI) and xylulose kinase (XK)as described previously (14).

Analytical methods Cell growth was measured by spectrophotometry (UV-1600, Shimadzu) at a wavelength of 562 nm. Dry cell weight (DCW) was calculated

0

0.5

1

1.5

2

2.5

3

0

10

20

30

40

50

60

0 12 24 36 48 60 72

Glu

cose

, xyl

ose,

lact

ic a

cid

(g/L

)

Time (h)

A B

FIG. 1. Lactic acid fermentation with varying concentrations of glucose/xylose in mMRS media50 g/L. Open circle, glucose; closed circle, xylose; open triangle, acetic acid; closed triangle, l2 mol of lactic acid via EmbdeneMeyerhof pathway. One mole of xylose yields 1 mol of acetpentose phosphate pathway.

Please cite this article in press as: Abdel-Rahman, M. A., et al., Fed-batch fmixture without carbon catabolite repression, J. Biosci. Bioeng., (2014), h

as described previously, with an OD of 1.0 corresponding to 0.218 g/L E. mundtii QU25 DCW (13). Glucose, xylose, and fermentation product levels were analysed usinga high-performance liquid chromatography (HPLC) system (US HPLC-1210, Jasco,Tokyo, Japan) equipped with a SUGAR SH-1011 column (Shodex, Tokyo, Japan) asdescribed previously (13). The optical purity of lactic acid was measured using aBF-5 biosensor (Oji Scientific Instruments, Amagasaki, Japan) according to themanufacturer’s protocol. Protein concentration in the crude cell extract wasdetermined using a BCATM Protein Assay Kit (Thermo, Rockford, IL, USA).

RESULTS

Effect of xylose concentration on byproductproduction Using xylose as a sole carbon source, we reportedon metabolic shift from homo-fermentation to hetero-fermenta-tion at high and low initial xylose concentration, respectively (12).To study the relationship between xylose concentration and aceticacid production, a mixtures of glucose and xylose in batchfermentations with G10X10 (Fig. 1A) and G10X50 (Fig. 1B) wereinvestigated. Lactic acid levels of 10.7 g/L and 50.7 g/L wereattained with acetic acid production of 1.98 g/L and 0.260 g/Lfrom G10X10 and G10X50 g/L, respectively. With G10X10, aceticacid was produced throughout the fermentation, while no aceticacid production was observed with G10X50 until the xyloseconcentration fell below 10 g/L. In addition, G10X50 fermentationresulted in a higher yield of lactic acid (0.876 g/g) and lower yieldof acetic acid (0.005 g/g) than G10X10 fermentation (0.557 g/gand 0.103 g/g, respectively). These results indicated that a higherinitial xylose concentration resulted in a higher concentrationand yield of lactic acid in the fermentation using a mixture ofglucose and xylose, and that we can restrict byproductproduction by maintaining the concentration of xylose in thefermentation broth at higher than 10 g/L.

Effect of different ratios of glucose and xylose in the sugarmixtures on CCR Various lignocellulosic biomasses can behydrolysed to yield diverse ratios of sugar mixtures (4), thereby,influencing sugar utilization and CCR. To investigate the effects ofthe sugar mixture on lactic acid fermentation, E. mundtii QU 25was grown in batch fermentations with 3 ratios of sugar mixtures:G50X25, G50X50, and G25X50. In all fermentations, glucose wascompletely utilized within 18 h with similar consumption rates,while xylose consumption was quite different (Table 1, Fig. 2).With G50X25 (Fig. 2A) and G50X50 (Fig. 2B), some xylose wasconsumed in the presence of glucose while E. mundtii QU 25 couldconsume only 42.9% and 61.4% of xylose, respectively, even after18 h when the glucose was exhausted. On the other hand, G25X50utilized glucose and xylose simultaneously, with 98.0% of xyloseconsumed linearly (with a constant consumption rate) (Table 1,

Ace

tic

acid

(g/L

)

Time (h)

0

0.5

1

1.5

2

2.5

3

0

10

20

30

40

50

60

0 12 24 36 48 60 72

by E. mundtii QU 25. (A) Glucose 10 g/L and xylose 10 g/L; (B) glucose 10 g/L and xyloseactic acid. Ethanol was not detected in these fermentations. One mole of glucose yieldsic acid and 1 mol of lactic acid via phoshoketolase pathway or 1.67 mol of lactic acid via

ermentation for enhanced lactic acid production from glucose/xylosettp://dx.doi.org/10.1016/j.jbiosc.2014.07.007

TABLE 1. Kinetic parameters of lactic acid fermentation in mixtures with different ratios of glucose and xylose by E. mundtii QU 25 at controlled pH 7.0 and 43�C.

Carbon source Timea (h) Max. lactic acid (g/L) DCWmaxb (g/L) Yieldc (g/g) PLA

d [g/(L$h)] Max. PLAe [g/(L$h)] Max. glucoseconsumptionratef [g/(L$h)]

Max. xyloseconsumption rateg

[g/(L$h)]

Xyloseutilizationratioh (%)

G50X25 72 55.9 4.27 0.823 0.776 5.16 [3e6 h] 6.87 [3e6 h] 0.930 [6e9 h] 61.4G50X50 60 62.0 3.78 0.800 1.03 4.96 [3e6 h] 5.72 [3e6 h] 1.17 [6e9 h] 42.9G25X50 72 62.3 3.61 0.812 0.865 5.21 [3e6 h] 5.22 [3e6 h] 1.42 [6e9 h] 98.0

a Fermenation time at the maximum lactic acid concentration.b Maximum dry cell weight.c Yield of lactic acid at max lactic acid concentration.d Productivity of lactic acid during the indicated fermentation period.e Maximum productivity of lactic acid in the indicated period.f Maximum glucose consumption rate in the indicated period.g Maximum xylose consumption rate during the indicated period.h Percentage of xylose consumption to the initial xylose concentration.

VOL. xx, 2014 LACTIC ACID PRODUCTION FROM MIXED SUGARS WITHOUT CCR 3

Fig. 2C). As shown inTable 1, themaximumxylose consumption ratewas 1.42 g/(L$h) with G25X50, which was higher than that forG50X50 and G50X25 at 1.17 g/(L$h) and 0.930 g/(L$h),respectively. In addition, G25X50 exhibited a higher lactic acidconcentration (62.3 g/L) and maximum lactic acid productivity[5.21 g/(L$h)] than G50X50 [62.0 g/L and 4.96 g/(L$h)] andG50X25 [55.9 g/L and 5.16 g/(L$h)]. It is likely that the higherutilization of xylose by E. mundtii QU 25 led to higher lactic acidproduction. These results demonstrate that the initialconcentration of glucose is a key factor determining xyloseutilization for lactic acid fermentation from mixed sugars usingE. mundtii QU 25, and that CCR can be avoided by maintaining theglucose concentration at less than 25 g/L.

Upstream xylose metabolic enzyme activities by E. mundtiiQU 25 in CCR One of the mechanisms for CCR of xylose con-sumption has been reported to be reduced enzyme activities relatedto upstream xylose metabolism in some LAB such as Lactobacillusbrevis (10) and Lactobacillus plantarum (15). Therefore,wemeasuredXI and XK activities in E. mundtii QU 25 batch fermentations withG50X50 (with CCR) and G25X50 (without CCR) at 6 h (in thepresence of glucose) and 26 h (after glucose exhaustion) (Fig. 2). Inthe presence of glucose, activities of XI (0.784 � 0.042 U/mgprotein) and XK (0.925 � 0.069 U/mg protein) with G25X50 were2.63-fold and 2.56-fold higher than those with G50X50(0.298 � 0.064 U/mg protein, 0.362 � 0.052 U/mg protein). Evenafter the glucose was completely consumed, the activities of XIand XK were maintained at high levels (0.495 � 0.022 U/mgprotein and 0.511 � 0.056 U/mg protein, respectively) withG25X50 while the activities with G50X50 decreased to0.112 � 0.047 U/mg protein and 0.117 � 0 U/mg protein,respectively. These results suggest that the low initialconcentration of glucose enhanced the activities of XI and XK,

0

15

30

45

60

75

0 12 24 36 48 60 720

15

30

45

60

75

0 12 24Glu

cose

, xyl

ose,

lact

ic a

cid

(g/L

)

Time (h) Tim

A B

FIG. 2. Lactic acid production by E. mundtii QU 25 from varying ratios of glucose and xylose iclosed circles, xylose; closed triangles, lactic acid.

Please cite this article in press as: Abdel-Rahman, M. A., et al., Fed-batch fmixture without carbon catabolite repression, J. Biosci. Bioeng., (2014), h

which resulted in a release of CCR following xylose consumption.Surprisingly, XI and XK activities in E. munditii QU 25 were notelevated after glucose exhaustion, which we attribute to the lowxylose consumption (Fig. 2).

Effect of nitrogen source supplementation on lactic acidproduction In order to demonstrate the effects of nitrogensources on sugar utilization and lactic acid productionwith G25X50by E. mundtii QU 25, supplementation of yeast extract and ammo-nium sulphate were investigated separately (Fig. S1). As shown inTable 2, supplementation with 5 g/L yeast extract stimulated cellgrowth and yielded a higher maximum DCW (4.63 g/L) thannormal medium without extra yeast extract (DCW, 3.61 g/L),indicating that the nitrogen source of the mMRS medium wouldbe insufficient for the cell growth of E. mundtii QU 25. However,the maximum DCW did not increase with ammonium sulphatesupplementation (DCW, 3.63 g/L). Although all fermentationsachieved simultaneous consumption of glucose and xylose(Fig. 2C and Fig. S1), xylose consumed with ammonium sulphatesupplementation was much lower than that with yeast extractsupplementation. Furthermore, the highest maximum xyloseconsumption rate among the 3 fermentations [3.35 g/(L$h)] wasobtained with yeast extract supplementation. The fermentationsupplemented with yeast extract exhibited higher lactic acidconcentration (67.2 g/L), and higher lactic acid productivity[0.933 g/(L$h)] than those with no supplementation [62.3 g/L, and0.865 g/(L$h)] and with supplementation of ammonium sulphate[52.1 g/L, and 0.724 g/(L$h)]. Surprisingly, supplementation ofyeast extract significantly increased the yield of lactic acid, goingfrom 0.812 g/g with no addition to 0.909 g/g. These resultssupport that supplementation of 5 g/L yeast extract to mMRSmedium stimulates not only cell growth and xylose utilizationbut also lactic acid production (concentration, productivity, and

0

15

30

45

60

75

0 12 24 36 48 60 72

C

36 48 60 72

e (h) Time (h)

n mMRS media containing (A) G50X25; (B) G50X50; (C) G25X50. Open circles, glucose;

ermentation for enhanced lactic acid production from glucose/xylosettp://dx.doi.org/10.1016/j.jbiosc.2014.07.007

TABLE 2. Kinetic parameters of lactic acid fermentation by E. mundtii QU 25 in mMRSmedia containing G25X50 supplementedwith different nitrogen sources at controlled pH7.0 and 43�C.

Supplementation(5 g/l)

Timea (h) Max. lacticacid (g/L)

DCWmaxb (g/L) Yieldc (g/g) PLA

d [g/(L$h)] Max. PLAe [g/(L$h)] Max. glucose consumptionratef [g/(L$h)]

Max. xylose consumptionrateg [g/(L$h)]

No addition 72 62.3 3.61 0.812 0.865 5.21 [3e6 h] 5.22 [3e6 h] 1.42 [6e9 h]Yeast extract 72 67.2 4.63 0.909 0.933 5.29 [3e6 h] 4.57 [3e6 h] 3.35 [6e9 h]Ammonium sulphate 72 52.1 3.63 0.869 0.724 5.16 [3e6 h] 5.72 [3e6 h] 0.767 [6e9 h]

a Fermentation time at the maximum lactic acid concentration.b Maximum dry cell weight.c Yield of lactic acid at the maximum lactic concentration.d Productivity of lactic acid during the indicated fermentation time.e Maximum productivity of lactic acid in the indicated period.f Maximum glucose consumption rate in the indicated period.g Maximum xylose consumption rate in the indicated period.

4 ABDEL-RAHMAN ET AL. J. BIOSCI. BIOENG.,

yield) by E. mundtii QU 25 in fermentations using a mixture ofglucose and xylose. The mMRS medium supplemented with 5 g/Lyeast extract was used for further studies.

Fed-batch fermentation with different neutralizingagents The above results indicate that glucose and xylose con-centrations should be maintained at lower than 25 g/L and athigher than 10 g/L, respectively, to achieve the simultaneous con-sumption of sugars without CCR and to restrain byproduct pro-duction by E. mundtii QU 25. In order to enhance the concentrationof lactic acid produced without CCR, we investigated the effects offed-batch fermentations using different neutralizing agents onlactic acid production. Fed-batch fermentations with 2 rounds ofsubstrate feeding were performed in mMRS media (G25X50) sup-plemented with 5 g/L yeast extract at 43�C and pH 7.0 using eithersodium hydroxide or ammonium hydroxide as the neutralizingagent. Mixtures of 25 g/L glucose and 50 g/L xylose were fed at 27 hand 54 h tomaintain the xylose concentration at greater than 10 g/L(Fig. S2).

Using sodium hydroxide, the maximum DCW of 4.63 g/L in fed-batch fermentation (Table 3) was the same as with batch fermen-tation (Table 2), while simultaneous consumption of glucose andxylose was observed even after substrate feeding in the fed-batchfermentation (Fig. S2-A). This resulted in an increase in the lacticacid concentration (to 113 g/L) in the fed-batch fermentation,compared to 67.2 g/L in the batch fermentation (Tables 2 and 3)with no byproduct formation, which indicates homo-lactic acidfermentation. However, immediately after the second round ofsubstrate feeding, the utilization rate for both sugars was quite lowand 15.0 g/L glucose and 59.4 g/L xylose remained in the broth after168 h (Fig. S2-A).

Using ammonium hydroxide as a neutralizing agent as shown inTable 3, the maximum DCW increased from 4.63 g/L using sodiumhydroxide to 5.40 g/L using ammonium hydroxide, and thus themaximum productivity of lactic acid was enhanced from 5.31 g/(L$h) to 5.60 g/(L$h). Additionally, overall productivity alsoincreased from 0.673 g/(L$h) to 0.768 g/(L$h). Although E. mundtii

TABLE 3. Kinetic parameters of fed-batch fermentation using sodium

Neutralizing agent Timeb (h) Max. lacticacid (g/L)

DCWmaxc (g/L) Yieldd (g/g) PLA

e [g/

Sodium hydroxide 168 113 4.63 0.809 0.67Ammonium hydroxide 168 129 5.40 0.785 0.76

a Fed-batch fermentations were carried out with controlled pH 7.0 at 43�C in mMRS mG25X50 plus 5 g/L yeast extract were carried out at 27 and 54 h.

b Fermentation time at maximum lactic acid concentration.c Maximum dry cell weight.d Yield of lactic acid at the maximum lactic concentration.e Productivity of lactic acid during the indicated fermentation time.f Maximum productivity of lactic acid in the indicated period.g Maximum glucose consumption rate in the indicated period.h Maximum xylose consumption rate in the indicated period.

Please cite this article in press as: Abdel-Rahman, M. A., et al., Fed-batch fmixture without carbon catabolite repression, J. Biosci. Bioeng., (2014), h

QU 25 did not consume all the xylose after the second round ofsubstrate feeding, glucose was completely consumed and theresidual xylose concentration decreased to 41.4 g/L after 168 h offermentation (Fig. S2-B), whichwasmuch lower than 59.4 g/L usingsodium hydroxide (Fig. S2-A). 129 g/L L-lactic acid with 99.5% op-tical purity was produced after 168 h of cultivation using ammo-nium hydroxide, which was much higher than the 113 g/L obtainedusing sodium hydroxide. These results indicate that ammoniumhydroxide was a satisfactory neutralizing agent that could stimu-late cell growth and lactic acid production by E. mundtii QU 25 byavoiding the sodium ion-induced inhibition.

DISCUSSION

There are 3 serious drawbacks in lactic acid fermentation by LABfrom mixed sugars (glucose, xylose, and arabinose) derived fromlignocellulosic biomass. First, few LAB can metabolize xylose tolactic acid. The second drawback is byproduct formation (mostlyacetic acid) due to heterofermentation of xylose via the phospho-ketolase pathways, and the third drawback is CCR of a non-favourable sugar (e.g., xylose) consumption caused by a preferredsugar (e.g., glucose) (4). In previous studies, we reported homo-L-lactic acid fermentation using glucose (11) or xylose of highconcentration (12) as the sole carbon source by our isolate,E. mundtii QU 25. In addition, this strain produced not only lacticacid but also acetic acid at low xylose concentration with low lacticacid yield, which resulted from the activation of metabolic enzyme(phosphoketolase) responsible for hetero-lactic acid fermentation(12). This study aimed at investigating the characteristics of lacticacid fermentation by E. mundtiiQU 25 from glucose/xylose mixture,consisting of lignocellulosic biomass.

With 10 g/L glucose, the batch fermentation at high xyloseconcentration (G10X50) exhibited homo-lactic acid productionwith little acetic acid production (0.260 g/L) and high lactic acidyield (0.876 g/g) while hetero-lactic acid production was observedwith high acetic acid (1.98 g/L) and low lactic acid yield (0.557 g/g)

hydroxide and ammonium hydroxide as neutralizing agents.a

(L$h)] Max. PLAf [g/(L$h)] Max. glucose consumptionrateg [g/(L$h)]

Max. xylose consumptionrateh [g/(L$h)]

3 5.31 [3e6 h] 3.52 [3e6 h] 2.26 [6e9 h]8 5.60 [3e6 h] 5.93 [3e6 h] 2.43 [6e9 h]

edia (initial sugars, G25X50) supplemented with 5 g/L yeast extract. Feeding with

ermentation for enhanced lactic acid production from glucose/xylosettp://dx.doi.org/10.1016/j.jbiosc.2014.07.007

VOL. xx, 2014 LACTIC ACID PRODUCTION FROM MIXED SUGARS WITHOUT CCR 5

from low xylose concentration (G10X10) (Fig. 1). Therefore, it wassuggested that the similar fermentation patters by E. mundtii QU 25should be shown from glucose/xylose mixture with our previousstudy using xylose as the sole carbon source (12). Furthermore, wecould indicate the threshold level of xylose concentration (�10 g/L)for achieving homo-lactic acid fermentation from glucose/xylosemixture.

CCR has been known as general phenomenon in most LAB,which results in an inhibition of simultaneous utilizations of non-favourable and preferred sugars (8,9). In this study, simultaneousutilization was clearly observed with low glucose concentration(G25X50) (Fig. 2C), however, E. mundtii QU 25 showed apparentCCR of xylose utilization with high glucose concentrations(G50X25 and G50X50) (Fig. 2A and B). These results suggested thatmaintaining the glucose concentration at less than 25 g/L is asignificant factor for better xylose consumption and restriction ofCCR by E. mundtii QU 25 using glucose/xylose mixture. In a previ-ous report using Enterococcus faecalis RKY1, CCR of maltose con-sumption was observed in sugar mixtures of glucose (75 g/L) andmaltose (75 g/L), but CCR was eliminated with glucose levels of lessthan 25 g/L (16). Therefore, to achieve lactic acid fermentation byE. mundtii QU 25 without CCR from mixed sugars derived fromlignocellulosic biomass, it is desirable for the lignocellulosicbiomass to contain higher proportions of hemicellulose than cel-lulose, such as leaves (80e85% and 15e20%, respectively) andcoastal bermuda grass (35.7% and 25%, respectively) (4).

The mechanisms for CCR of xylose utilization by some lactoba-cilli have been considered to be related to the repression of enzymeactivities of upstream xylose metabolism (10). However, little isknown of the mechanism in E. mundtii species. Our results ofenzyme assays suggested that CCR of xylose consumption fromsugar mixtures by E. munditii QU 25 is likely to result from lowactivities of XI and XK, similar to that in L. brevis (10). Moreover,genetic repression of metabolic enzymes has been reported as amechanism of CCR in Firmicutes bacteria (9,17). Generally, 2 pro-teins are thought to regulate genetic expression, phosphocarrierprotein (Hpr) and catabolite control protein A (CcpA). Serine-phosphorylated Hpr couples with CcpA under high intracellularmetabolite concentrations (glucose 6-phosphate and fructose 1,6-biphosphate) derived from glucose (favourable sugar) metabolism,and the complex then represses transcription of genes or operonsencoding metabolic enzymes for non-favourable sugars. Ourenzyme activity results suggest that transcription of the genesencoding XI and XK would decrease with high glucose concentra-tions in mixed sugar fermentations by E. mundtiiQU 25 via a similarmechanism, as bacteria of the same genus (E. faecalis) have beenreported to possess the ccpA gene (18). Nevertheless, to the best ofour knowledge, the CCRmechanism of xylose utilization by glucose

TABLE 4. Comparison of glucose/xylose fermentation by Enterococcus mundtii

Strain(s) Mixed sugarconcentrationa

(g/L)

Lactic acid(g/L)

Yieldb

(g/g)PLA

c [g/(L$h)] Iso

p

E. mundtii QU 25 G25X50 67.2 0.909 1.12 L

129 0.785 0.768L. brevis ATCC 14869 G10X10 12.5 0.579 0.568 e

L. plantarum DldhL1::PxylAB-xpk1::tkt-Dxpk2::PxylAB

G25X25 39.7 0.79 1.53 D

G50X25 58.6 0.79 2.25 D

G75X25 74.2 0.780 2.85 D

L. rhamnosus/L. brevis(mixed culture)

G15X5 14.8 0.73 0.40 e

E. casseliflavus/L. casei(mixed culture)

G100X50 95 ca. 0.63 0.495 L

a G, glucose (g/L); X, xylose (g/L).b Yield of lactic acid at the maximum lactic acid concentration.c Productivity of lactic acid.

Please cite this article in press as: Abdel-Rahman, M. A., et al., Fed-batch fmixture without carbon catabolite repression, J. Biosci. Bioeng., (2014), h

has been poorly characterized in LAB. Further research is planned toclarify these points in order to improve the efficiency of xyloseutilization in the presence of high concentrations of glucose.

In this study, we also found two significant factors to affect lacticacid fermentation by E. mundtii QU 25 from glucose/xylose mixturewith avoiding CCR: nitrogen source and neutralizing agents. Actu-ally, supplementation of 5 g/L yeast extract to mMRS medium(Fig. S1-A) enhanced cell growth, xylose consumption rate, con-centration and yield of lactic acid (Table 2), compared to fermen-tation with no supplementation (Fig. 2C). Stimulation of lactic acidproduction by nitrogen source supplementation has also been re-ported with several LAB (19,20). For example, an increase in yeastextract supplementation from 2 g/L to 5 g/L during Lactobacilluscoryniformis subsp. torquens fermentation of Hydrodictyon retic-ulatum hydrolysates containing mainly glucose and mannose (20)enhanced the carbon source utilization ratio from 68.5% to 93.7%,the lactic acid concentration from 26.8 g/L to 36.5 g/L, and the lacticacid yield from 0.334 g/g to 0.457 g/g. On the other hand, fed-batchfermentation showed higher lactic acid concentration (129 g/L) andDCW (5.40 g/L) using ammonium hydroxide as neutralizing agentthan those (113 g/L and 4.63 g/L, respectively) using sodium hy-droxide (Table 3; Fig. S2). It has been reported that L. plantarum JCM1149 showed weak cell growth with a sodium concentration ofgreater than 50 g/L and lack of growth when the sodium concen-tration was greater than 60 g/L (21). Totally approximately 50 g/Lsodium ions were fed to the fermentor using sodium hydroxide asthe neutralizing agent, which could result in the decreasedfermentation efficiency by E. mundtii QU 25. In addition, there is areport on enhanced cell growth and lactic acid productivityfollowing Lactobacillus delbrueckii subsp. lactis fermentation usingammoniumhydroxide comparedwith sodium hydroxide due to theutilization of ammonium ion as the nitrogen (2). After all, the lacticacid concentration improved from 62.3 g/L (Table 1 and Fig. 2C) to129 g/L (Table 3 and Fig. S2-B) from glucose/xylose mixture in fed-batch fermentation with 2 rounds of substrate feeding by thesupplementation of yeast extract using ammonium hydroxide,achieving simultaneous consumption of glucose and xylose underthe maintenance of concentrations at lower than 25 g/L and athigher than 10 g/L, respectively.

Table 4 summarises several studies on avoiding CCR. We firstestablished a homo-L-lactic acid production in fed-batch fermen-tation using mixed sugars without CCR, which resulted in 129 g/LL-lactic acid at a relatively high lactic acid yield of 0.785 g/g.Therefore, E. mundtii QU 25 could be used for lactic acid productionfrom low-cost feedstock of lignocellulosic biomass as leaves orswine wastes.

In conclusion, we investigated lactic acid fermentation frommixed glucose and xylose using xylose-utilizing E. mundtii QU 25.

QU 25 with those by other lactic acid bacteria reported in the literature.

mer andopticalurity (%)

Acetic acid(g/L)

Fermentationmode

Temperature; pH(neutralizing agent)

Reference

-(99.5) 0.0 Batch 43�C; 7.0 (10 M NaOH) This study0.0 Fed-batch 43�C; 7.0 (10 M NH3 solution) This study4.5 Batch 43�C, No control 10

-(99.1) 0.7 Batch 36�C; 6.0 (10 M NH3 solution) 15

-(99.6) 0.10 Batch-(99.5) 0.0 Batch

1.22 Batch 37�C; 5.0 (10 g/l CaCO3) 22

-(96) 5.7 Batch 30�C; 6.8 (4 M NaOH) 23

ermentation for enhanced lactic acid production from glucose/xylosettp://dx.doi.org/10.1016/j.jbiosc.2014.07.007

6 ABDEL-RAHMAN ET AL. J. BIOSCI. BIOENG.,

Homo-lactic acid production with low byproduct formation andsimultaneous mixed sugar utilization in batch fermentation wasaccomplished by maintaining the concentrations of xylose andglucose over 10 g/L and below 25 g/L, respectively. Yeast extractsupplementation and usage of ammonium hydroxide as aneutralizing agent improved both the lactic acid concentration andproductivity. Finally, we succeeded in attaining a high L-lactic acidconcentration (129 g/L) with 99.5% optical purity without CCR infed-batch fermentation, the highest reported in published litera-ture to date.

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jbiosc.2014.07.007.

ACKNOWLEDGMENTS

This study has been supported by a Grant-in-Aid (P12088) forJSPS Fellows toMohamed Ali Abdel-Rahman from the Japan Societyfor the Promotion of Science (JSPS), Japan. This research wasfinancially supported in part by the Sumitomo Corporation (Japan).

References

1. Abdel-Rahman, M. A., Tashiro, Y., and Sonomoto, K.: Recent advances inlactic acid production by microbial fermentation processes, Biotechnol. Adv.,31, 877e902 (2013).

2. Tashiro, Y., Kaneko, W., Sun, Y., Shibata, K., Inokuma, K., Zendo, T., andSonomoto, K.: Continuous D-lactic acid production by a novel thermotolerantLactobacillus delbrueckii subsp lactis QU 41, Appl. Microbiol. Biotechnol., 89,1741e1750 (2011).

3. Adnan, A. F. M. and Tan, I. K. P.: Isolation of lactic acid bacteria from Malaysianfoods and assessment of the isolates for industrial potential, Bioresour. Tech-nol., 98, 1380e1385 (2007).

4. Abdel-Rahman, M. A., Tashiro, Y., and Sonomoto, K.: Lactic acid productionfrom lignocellulose-derived sugars using lactic acid bacteria: overview andlimits, J. Biotechnol., 156, 286e301 (2011).

5. Lin, Y. and Tanaka, S.: Ethanol fermentation from biomass resources: currentstate and prospects, Appl. Microbiol. Biotechnol., 69, 627e642 (2006).

6. Hendriks, A. and Zeeman, G.: Pretreatments to enhance the digestibility oflignocellulosic biomass, Bioresour. Technol., 100, 10e18 (2009).

7. Oshiro, M., Shinto, H., Tashiro, Y., Miwa, N., Sekiguchi, T., Okamoto, M.,Ishizaki,A., andSonomoto,K.:Kineticmodelingandsensitivity analysisofxylosemetabolism in Lactococcus lactis IO-1, J. Biosci. Bioeng., 108, 376e384 (2009).

8. Brückner, R. and Titgemeyer, F.: Carbon catabolite repression in bacteria:choice of the carbon source and autoregulatory limitation of sugar utilization,FEMS Microbiol. Lett., 209, 141e148 (2002).

Please cite this article in press as: Abdel-Rahman, M. A., et al., Fed-batch fmixture without carbon catabolite repression, J. Biosci. Bioeng., (2014), h

9. Görke, B. and Stülke, J.: Carbon catabolite repression in bacteria: many waysto make the most out of nutrients, Nat. Rev. Microbiol., 6, 613e624 (2008).

10. Kim, J. H., Shoemaker, S. P., and Mills, D. A.: Relaxed control of sugar utili-zation in Lactobacillus brevis, Microbiology, 155, 1351e1359 (2009).

11. Abdel-Rahman, M. A., Tashiro, Y., Zendo, T., and Sonomoto, K.: Improvedlactic acid productivity by an open repeated batch fermentation system usingEnterococcus mundtii QU 25, RSC Adv., 3, 8437e8445 (2013).

12. Abdel-Rahman, M. A., Tashiro, Y., Zendo, T., Hanada, K., Shibata, K., andSonomoto, K.: Efficient homofermentative L-(þ)-lactic acid production fromxylose by a novel lactic acid bacterium, Enterococcus mundtii QU 25, Appl.Environ. Microbiol., 77, 1892e1895 (2011).

13. Abdel-Rahman, M. A., Tashiro, Y., Zendo, T., Shibata, K., and Sonomoto, K.:Isolation and characterisation of lactic acid bacterium for effective fermenta-tion of cellobiose into optically pure homo-L-(þ)-lactic acid, Appl. Microbiol.Biotechnol., 89, 1039e1049 (2011).

14. Wang, Y., Abdel-Rahman, M. A., Tashiro, Y., Xiao, Y. T., Zendo, T., Sakai, K.,and Sonomoto, K.: L-(þ)-Lactic acid production by co-fermentation of cello-biose and xylose without carbon catabolite repression using Enterococcusmundtii QU 25, RSC Adv., 4, 22013e22021 (2014).

15. Yoshida, S., Okano, K., Tanaka, T., Ogino, C., and Kondo, A.: Homo-D-lacticacid production from mixed sugars using xylose-assimilating operon-inte-grated Lactobacillus plantarum, Appl. Microbiol. Biotechnol., 92, 67e76 (2011).

16. Yun, J. and Ryu, H.: Lactic acid production and carbon catabolite repressionfrom single and mixed sugars using Enterococcus faecalis RKY1, Process Bio-chem., 37, 235e240 (2001).

17. Deutscher, J.: The mechanisms of carbon catabolite repression in bacteria,Curr. Opin. Microbiol., 11, 87e93 (2008).

18. Leboeuf, C., Auffray, Y., and Hartke, A.: Cloning, sequencing and character-ization of the ccpA gene from Enterococcus faecalis, Int. J. Food Microbiol., 55,109e113 (2000).

19. Nancib, A., Nancib, N., Meziane-Cherifa, D., Boubendira, A., Fickb, M., andBoudrantb, J.: Joint effect of nitrogen sources and B vitamin supplementationof date juice on lactic acid production by Lactobacillus casei subsp. rhamnosus,Bioresour. Technol., 96, 63e67 (2005).

20. Nguyen, C. M., Kim, J., Song, J. K., Choi, J. G., Choi, Y. H., Jang, K. S., andKim, J.: D-Lactic acid production from dry biomass of Hydrodictyon reticulatumby simultaneous saccharification and co-fermentation using Lactobacillus cor-yniformis subsp. torquens, Biotechnol. Lett., 34, 2235e2240 (2012).

21. Cai, Y., Ohmomo, S., Ogawa, M., and Kumai, S.: Effect of NaCl-tolerant lacticacid bacteria and NaCl on the fermentation characteristics and aerobic stabilityof silage, J. Appl. Microbiol., 83, 307e313 (1997).

22. Cui, F., Li, Y., and Wan, C.: Lactic acid production from corn stover using mixedcultures of Lactobacillus rhamnosus and Lactobacillus brevis, Bioresour. Technol.,102, 1831e1836 (2011).

23. Taniguchi, M., Tokunaga, T., Horiuchi, K., Hoshino, K., Sakai, K., andTanaka, T.: Production of L-lactic acid from a mixture of xylose and glucose byco-cultivation of lactic acid bacteria, Appl. Microbiol. Biotechnol., 66, 160e165(2004).

ermentation for enhanced lactic acid production from glucose/xylosettp://dx.doi.org/10.1016/j.jbiosc.2014.07.007