Enhanced D-lactic acid production from renewable resources using engineered APPLIED GENETICS AND MOLECULAR BIOTECHNOLOGYEnhanced D-lactic acid production from renewableresources using engineered Lactobacillus plantarumYixing Zhang 1 & Praveen V. Vadlani 1,2 & Amit Kumar 3 & Philip R. Hardwidge 3 &Revathi Govind 4 & Tsutomu Tanaka 5 & Akihiko Kondo 5Received: 11 June 2021 /Revised: 24 August 2021 /Accepted: 16 September 2021# Springer-Verlag Berlin Heidelberg 2021Abstract D-lactic acid is used as a monomer in the produc-tionofpoly-D-lacticacid(PDLA),whichisusedtoformheat-resistant stereocomplex poly-lactic acid. To produce cost-effective D-lactic acid by using all sugars derived from bio-mass efficiently, xylose-assimilating genes encoding xyloseisomerase and xylulokinase were cloned into an L-lactate-deficient strain, Lactobacillus plantarum. The resulting re-combinant strain, namely L. plantarum NCIMB 8826ΔldhL1-pLEM-xylAB, was able to produce D-lactic acid (atoptical purity >99 %) from xylose at a yield of 0.53 g produce D-lactic acid, such as cyanobacterium, in which anengineered D-lactate dehydrogenase and a lactic acid trans-porter were introduced in Synechococcus elongatus (Li et al.2021) and in yeast, in which the coding region of pyruvatedecarboxylase I was deleted and D-lactate dehydrogenasegene from Leticonostoc mesenteroides was introduced intothe genome of Saccharomyces cerevisiae (Ishida et al.2006).Lacticacidbacteria(LAB)havebeentraditionallyusedfor lactic acid production. LAB are Gram-positive microor-ganisms that exist within plants, meat, and dairy products thatproduce lactic acid with high yield and productivity. LAB areeither homofermentative or heterofermentative based on theirend products. Homofermentative LAB such as L. delbrueckii(Zhang and Vadlani 2021) produce lactic acid as major end-product through the Embden-Meyerhof pathway (EMP) andare preferable for commercial-scale lactic acid production(Abdel-Rahman et al. 2021). Most homofermentative LABcannot use pentose sugars, the dominant sugars of hemicellu-lose, which leads to low productivity. On the other hand,heterofermentative LAB, such as Lactobacillus brevis (Guoet al. 2021), and Lactobacillus pentosus (Bustos et al. 2005),usethephosphoketolase(PK)pathway,whichcleavespentosesugarsto glyceraldehyde 3-phosphate (GAP) and acetylphos-phate followed by converting GAP into lactic acid. Xylosefermentation in these heterofermenters involves isomerizationof xylose to xylulose and phosphorylation of xylulose toxylulose-5-phosphate (Lockman et al. 1997); however, otherby-products, such as acetic acid, ethanol, and/or formic acidthat result in low lactic acid yield and additional cost in thepurification step, were also produced. Lactobacillusplantarum is a facultative heterofermentative strain that fer-ments hexosesugarsthrough EMP,butitalsohasaninduciblePK pathway; arabinose was converted to xylulose-5-phosphate and further converted to lactic acid and acetic acidthrough the PK pathway in L. plantarum (Okano et al 2021a;Helanto et al. 2007). The goal of utilizing all major biomasssugars can be achieved during the production of lactic acid ifxylose-assimilating genes are introduced into L. plantarumandconvertxyloseintoxylulose-5-phosphate,anintermediatein the PK pathway. Introduction of DNA into Lactobacilli ischallenging, mainly because of the unavailability of suitablecloning vectors and efficient transformation systems (Posnoet al. 1991a). The ability of plasmid to replicate itself andexpress foreign genes is usually unpredictable, which alsoadds difficulty to applying recombinant DNA technology toLactobacillus strains (Serror et al. 2002). pLEM415 plasmidderived from pLEM3, which was isolated from Lactobacillusfermentum, has been used to express heterologous genes indifferent Lactobacillus strains (Fons et al. 1997; Rochat et al.2006). Awell-defined constitutive promoter is preferred to aninducible promoter because inducible systems are not alwayseasy to manage under industrial conditions (Ahmed 2006). Aconstitutive clpC promoter from L. fermentum BR11displayed high activity and encouraged heterologous gene ex-pression in Lactobacilli strains (McCracken et al, 2000;Okano et al. 2021b; Okano et al 2021c).Lactic acid bacteria are fastidious microorganisms; theyrequire a wide range of growth factors including amino acids,vitamins, and fatty acids (Yadav et al. 2021). Complex nitro-gen sources are usually used to grow lactic acid bacteria, andyeast extract (YE) is the most effective for both microbialgrowth and lactic acid production (Kwon et al. 2000); how-ever, YE is not cost-effective for the production of lactic acidfor commodity chemicals. The cost of YE is estimated tocontribute as much as 30 % of the total production cost oflactic acid (Li et al. 2006). Soybean meal is a major residueof soybean oil extraction and contains 44 % crude protein andall essential amino acids, including high levels of glutamicacid, a strong lactic acid bacteria growth promotant (Batalet al. 2000; Pollack and Lindner 1942). Corn stover is one ofthe most abundant agricultural residues in the USA and insome regions in the world, and constitutes corn leaves, stalks,and husks (Li et al. 2004). Raw corn stover contains 49.6 %glucan, 25.1 % xylan, and 23.7 % lignin and is used as foragefor cattle, but recently, the feedstock has been extensivelyevaluated for ethanol production (Guragain et al. 2021;Zambare and Christopher 2021).In this study, we constructed a recombinant plasmid forxylose assimilation and introduced it into L. plantarumNCIMB 8826 ΔldhL1. D-lactic acid production of this strainwas investigated using corn stover and soybean meal extract(SBME) as substrates. The composition of the fermentationmediumwasoptimizedbyresponsesurfacemethodology,andoptimalconditionswereusedtoproduceD-lacticacidinafed-batch fermentation.Materials and methodsBacterial strains and plasmidsL. brevis ATCC 367 was purchased from the American TypeCulture Collection (Manassas, VA, USA). L. plantarumNCIMB 8826 ΔldhL1 and pCU-PxylAB plasmid containingthe clpC promoter were developed in our previous study(Okano et al. 2021b; 2021c), and pLEM415 plasmid was do-nated by Serror et al. (2002). L. plantarum NCIMB 8826ΔldhL1 and L. brevis ATCC 367 were grown in de Man,Rogosaand Sharpe (MRS)broth at37and 30°C, respectively(OXOID Ltd. Basingstoke, Hampshire, England).Escherichia coli DH5α was used to manipulate pLEM415-based DNA, which was growing in Luria-Bertani (LB) medi-um at 37 °C. Antibiotics were added when necessary:100 μg mL ?1 ampicillin for E. coli and 25 μg mL ?1 erythro-mycin for L. plantarum. Table 1 shows the microorganisms,plasmids, and primers used in this study.Appl Microbiol BiotechnolFeedstock preparationCorn stover was obtained from the Kansas State UniversityAgronomy Farm in Manhattan and Tribune, Kansas, whichwas pretreated with 1 % (v/v) sodium hydroxide according tothemethoddescribedbyGuragainetal.(2021).Soybeanmealwas obtained from the O.H. Kruse Feed TechnologyInnovation Center in Manhattan, Kansas. Soybean meal ex-tract was prepared by using a method modified from Zhanget al. (2021). Soybean meal (60 g) was mixed with 600 mLwaterandshakenatroomtemperaturewithanagitationrateof150 rpm (Innova 2350, New Brunswick Scientific, CT, USA)for 1 h. The soybean meal slurry was then centrifuged at 10,000×g for 10 min (Sorval RC 5C Plus, GMI Inc., MN, USA).The supernatant was collected and used as SBME for lacticacid production. Amino acid content of SBME was analyzedby Agricultural Experiment Station Chemical Laboratory,University of Missouri (Columbia, MO, USA).Construction of recombinant pLEM415-xylAB plasmidThe genomic DNA of L. brevis ATCC 367 was extractedusing an IBI genomic DNA mini kit (MidSci, St. Louis,MO, USA) according to the manufacturer’s instructions. ThexylAB operon from L. brevis ATCC 367’s genome was ampli-fied by PCR using xylAB-F and xylAB-R primers, which weredesigned based on the sequence of the xylAB operon derivedfrom L. brevis ATCC 367’s genome (GenBank accessionnumber NC_008497.1). The amplified 2.98-kb DNA frag-ment was then digested with XhoI and PvuII and ligated intoXhoI- and EcoRV-digested pLEM415 vector. The promoterclpC fragment from the pCU-PxylAB vector was amplifiedby PCR using clpC-F and clpC-R primers, and the amplifiedfragmentwas thenligated intothe pLEM415vectorharboringthe xylAB operon using recombinant sites KpnI and XhoI. Theresulting plasmid designed to express xylose isomerase andxylulokinase under control of the clpC promoter was desig-nated pLEM415-xylAB (Fig. S1) and was sent to MolecularCloning Laboratories (South San Francisco, CA, USA) forsequencing. The sequence thus obtained was verified for in-frame cloning using MEGA 6 (Tamura et al. 2021).pLEM415-xylAB was then transformed into L. plantarumNCIMB 8826 ΔldhL1 using the method modified from Naritaet al. (2006). L. plantarum NCIMB 8826 ΔldhL1 was culti-vated overnight in a test tube containing 5 mL of MRS broth,and the overnight culture was then diluted 100 times withfresh MRS broth and cultivatedat37 °C until the OD 600 valuereached 0.5 to 0.8. Cells were washed five times with washbuffer (272 mM sucrose, 7 mM HEPES, 1 mM MgCl 2 ,pH 7.4) and suspended in 1 mL electroporation buffer (washbuffer with 20 % (w/v) PEG6000). Fifty microliters of com-petent cells were mixed with 0.1～0.3 μg of plasmid DNA andincubated on ice for 30 min. Before electroporation, 1 μL ofTypeOne Restriction Inhibitor (Epicenter Technologies Corp.Madison, WI, USA) was added. Samples were then subjectedto a 2.5-kV, 25-μFD, and 200-Ω electric pulse in a 0.2-cmcuvette by using a Gene pulser Xcell electroporator (Bio-Rad, Hercules, CA, USA). Fresh MRS broth (500 μL) wasimmediately added, and cells were incubated for 2 h at 37 Cbefore plating on MRS agar supplemented with 25 μg mL ?1Table 1 Bacterial strains and plasmidsStrains and plasmids Relevant characteristics AntibioticresistanceReference or sourceStrainsEscherichia coli DH5α lacZ ΔM15, recA1, endA1 InvitrogenL. plantarum NCIMB 8826 ΔldhL1 L. plantarum NCIMB 8826 L-lactate dehydrogenasegene1 deletionOkano et al. 2021cL. brevis ATCC 367 Source of xylAB gene ATCCL. plantarum NCIMB 8826 ΔldhL1-pLEM415-xylABL. plantarum NCIMB 8826 ΔldhL1 strain carryingxylose assimilation plasmidThis studyPlasmidspLEM415 Escherichia coli-Lactobacillus shuttle vector ErythromycinAmpicillinSerror et al. 2002pCU-PxylAB Expression vector containing the xylAB operon underthe control of clpC UTLS promoterErythromycinOkano et al. 2021bPrimersxylAB-F TAACTCGAGGGAGGGCTTTTATAATTATGACxylAB-R TAACAGCTGCTAAAGCTCCGCTCGCCGATAGTCTAAclpC-F GCGGTACCCTTAAAATATAGTCATAGAATTAGGGCGclpC-R GCCTCGAGTAATCTTGACCATTATTTTACCACACTTRestriction enzyme cleavage sites are underlinedAppl Microbiol Biotechnolerythromycin. Plates were incubated at 37 C for 2 to 3 days.The resulting transformant was designated L. plantarumNCIMB 8826 ΔldhL1- pLEM415-xylAB and used in the fer-mentation experiments.D-lactic acid production from pure sugarsAfed-batch xylosefermentationexperimentwasconductedina 7-L fermenter with 5 L working volume (Bioflo 110, NewBrunswick Scientific Inc., Enfield, CT, USA). L. plantarumNCIMB 8826 ΔldhL1-pLEM415-xylAB was grown in MRSbroth with 25 μg mL ?1 erythromycin until OD value reachedabout 5 and was used toinoculate at5 % (v/v) tothe fermentercontaining 5 L modified MRS medium with 40 g L ?1 of xy-lose supplemented with 10 g L ?1 of peptone, 5 g L ?1 of YE,2g L ?1 ofammoniumcitrate,2 g L ?1 ofK 2 HPO 4 , 0.1 g L ?1 ofMgSO 4 .7H 2 O, and 0.05 g L ?1 of MnSO 4 .4H 2 O.Erythromycin was added at a final concentration of25 μg mL ?1 .Temperature was controlled at 37 °C with agita-tion of 150 rpm. pH was maintained at 6.5 by adding 10 Nsodium hydroxide.Amixedsugarfermentationexperimentwasconductedina2-L fermenter with1.5 L working volume(BiostatB, SatoriusAG, Goettingen, Germany). Fermentation medium contained37.5 g L ?1 glucose and 19.7 g L ?1 xylose and was supple-mented with 10 g L ?1 of peptone, 5 g L ?1 of YE, 2 g L ?1 ofammonium citrate, 2 g L ?1 of K 2 HPO 4 , 0.1 g L ?1 ofMgSO 4 .7H 2 O, and 0.05 g L ?1 of MnSO 4 .4H 2 O and25 μg mL ?1 of erythromycin. Fermentation conditions wereidentical to the xylose fermentation experiment.D-lactic acid production from corn stoverSequential saccharification and fermentation (SHF) and si-multaneous saccharification and fermentation (SSF) experi-ments with corn stover were carried out in 150-mL conicalflasks. In SHF experiments, 2 g of dried alkali-treated cornstover was hydrolyzed by Cellic CTec2 obtained fromNovozyme Inc. (Franklinton, NC, USA). The dosage ofCTec2 was added at 8 FPU per gram of corn stover.Saccharification was carried out at 50 °C for 48 h and centri-fuged at 10.000×g for 10 min (Sorvall RC 5C Plus, GMI Inc.,MN, USA). Supernatant was collected and pH adjusted to 6.5by sodium hydroxide. Corn stover hydrolysate was supple-mented with all the components (except sugars) of the modi-fied MRS medium to make a final volume of 50 mL, and 3 %(w/v) of CaCO 3 was added to buffer the pH. Erythromycinwas added at a final concentration of 25 μg mL ?1 .Fermentation was performed at 37 °C with 150 rpm agitation.In SSF experiments, 2 g of alkali-treated corn stover wassupplemented with all the components (except sugars) of themodifiedMRSmediumandvolumewasadjustedto50mLby50 mM sodium citrate buffer (pH 5). L. plantarum NCIMB8826ΔldhL1-pLEM415-xylABinoculumwasaddedat5%(v/v) and Cellic CTec2 was added at 8 FPU per gram of cornstover. SBMEwasevaluated tosubstituteYEfor D-lacticacidproduction from corn stover via SSF process. SBME wasadded at 10 % (v/v) with 5 g L ?1 peptone, 2 g L ?1 of ammo-nium citrate, 2 g L ?1 of K 2 HPO 4 , 0.1 g L ?1 of MgSO 4 .7H 2 O,and0.05gL ?1 ofMnSO 4 .4H 2 O.Erythromycinwasaddedatafinal concentration of 25 μg mL ?1 . Fermentation conditionswere the same as described in SHF experiments.The fed-batch SSF experiment was carried out in 500-mLconical flasks with working volumes of 100 mL. Dried alkali-treated corn stover (4 g), Cellic CTec2 (5.6 FPU g ?1 of cornstover), SBME (15 % v/v), peptone (3 g L ?1 ), salts (2 g L ?1 ofammonium citrate, 2 g L ?1 of K 2 HPO 4 , 0.1 g L ?1 ofMgSO 4 .7H 2 O, and 0.05 g L ?1 of MnSO 4 .4H 2 O), and inocu-lum (5 % v/v) were added at the beginning of fermentation.CaCO 3 (3 g) was also added in the beginning to maintain pHin the flasks. Erythromycin was added at a final concentrationof25μgmL ?1 .Feedwasappliedevery36handcontained2gof corn stover, 1.5 g of CaCO 3 , and 15 mL of SBME alongwith Cellic CTec2 at a dosage of 5.6 FPU g ?1 of corn stover.Statistical experimental designResponse surface methodology was used to optimize key fac-torsaffectinglacticacidproduction,whichwereenzymeload-ing, SBME concentration, and peptone concentration in abatch shake flask. Design Expert V. 8.0.7.1 (Stat-Ease Inc.,Minneapolis, MN, USA) was used to generate experimentaldesign, assess the response of dependent variables, and alsogenerate response surface plots.Three independent factors (enzyme loading, SBME con-centration, and peptone concentration) and their respectivelevels are given in Table S1. Box-Behnken design (Box andBehnken 1960) was adopted to optimize the levels of thesethree factors. A total of 17 runs comprising 5 replicates in thecentral point were carried out in random order. Lactic acidconcentration was the response. A second-order quadraticmodel was fitted for the experimental results. Validation ofoptimized conditions was carried out with four replications.Analytical proceduresCell growth was measured by a spectrophotometer at a wave-length of 600 nm (UV-1650PC, Shimadzu, Torrance, CA,USA). Concentrations of lactic acid, acetic acid, glucose,and xylose were measured using a high-performance liquidchromatography (HPLC) system equipped with a refractiveindex detector (RID-10A) and a Rezex ROA organic acidcolumn (300 × 7.8 mm, Phenomenex Inc., Torrance, CA,USA). Samples were centrifuged at a speed of 15,000×g for10 min (Eppendorf, Hauppauge, NY, USA), and the superna-tant was acidified with 1 N H 2 SO 4 and centrifuged at 15,Appl Microbiol Biotechnol000×g for 15 min to remove CaSO 4 precipitant. Supernatantwas diluted 10 times with deionized water before analysis.0.005 N H 2 SO 4 was used as mobile phase at an elution speedof 1 mL min ?1 , column temperature was maintained at 80 °C,and RID detector temperature was maintained at 40 °C. Theoptical purity of lactic acid was measured by the method de-scribed by Zhang and Vadlani (2021).ResultsD-lactic acid production from pure sugars usingL. plantarum NCIMB 8826 ΔldhL1-pLEM415-xylABXylose-assimilating gene (xylAB)-transformed strains wereselected on an MRS plate containing 25 μg mL ?1 of erythro-mycin. Cultivation with xylose as the sole carbon source wascarried out using L. plantarum NCIMB 8826 ΔldhL1 as acontrol to confirm that transformed stains were able to usexylose. After 48 h of cultivation, the OD 600 of L. plantarumNCIMB 8826 ΔldhL1 was 0.7, but L. plantarum NCIMB8826 ΔldhL1-pLEM-xylAB showed remarkably increasedgrowth (OD 600 was 3.6); these results indicate that introduc-tion of xylAB genes into L. plantarum NCIMB 8826 ΔldhL1resulted in successful assimilation of xylose by L. plantarumNCIMB 8826 ΔldhL1-pLEM-xylAB.Fermentation with 40 g L ?1 of xylose was performed toevaluate the lactic acid fermentation ability of L. plantarumNCIMB 8826 ΔldhL1-pLEM-xylAB from pure xylose. Asshown in Fig. 1a, 19.7 g L ?1 of D-lactic acid was producedalong with 12.8g L ?1 ofaceticacidatthe end ofthe first stageof fermentation. The yield of D-lactic acid from xylose in thefirst stage was 0.53 g g ?1 , which was comparable to that of L.brevis (0.50 g g ?1 ). After 56 h of fermentation, 700 mL offermentation broth was pumped out, 700 mL offresh mediumcontaining 200 g xylose was added, and fermentation contin-ued for the next 112 h. In the end of fermentation, 30.1 g L ?1of D-lactic acid was produced with 20.5 g L ?1 of acetic acid.A mixed sugar experiment using glucose and xylose at a2:1 ratio was conducted to mimic sugar composition in enzy-matic hydrolysate of alkali-treated corn stover. As shown inFig. 1b, almost all glucose was consumed at 36 h, whereasxylosewasconsumedmoreslowlythanglucoseandalmostallxylose was consumed at 48 h. At the end of fermentation,47.2 g L ?1 of D-lactic acid and 8.9 g L ?1 of acetic acid wereobtained. Yield of lactic acid from both glucose and xylosewas 0.84 g g ?1 , and productivity was 0.98 g L ?1 h ?1 .Simultaneous utilization of glucose and xylose byL g g ?1 and productivity of0.51 g L ?1 h ?1 (Zhang and Vadlani 2021).D-lactic acid production from corn StoverFigure 2a shows the fermentation profile of using corn stoverhydrolysate. Glucose (16.3 g L ?1 ) was consumed within 12 h,and xylose (8.6 g L ?1 ) was consumed within 24 h. D-lacticacid 19.4g L ?1 alongwithaceticacid4.6gL ?1 was produced.The yield of D-lactic acid from total sugar was 0.78g g ?1 , andproductivity was 0.27g L ?1 h ?1 .The SSF process was appliedto convert corn stover to D-lactic acid to improve lactic acidproductivity (Fig. 2b). Glucose and xylose concentration inthe medium were maintained at less than 1 g L ?1 during theentire process. As shown in Table 2, D-lactic acid concentra-tion reached up to 26.8 g L ?1 , with overall yield of 0.67 g g ?1and productivity of 0.74 g L ?1 h ?1 . D-lactic acid yield andproductivity from corn stover by this recombinant strain wereFig. 1 D-lactic acid production from pure sugars in fermenter. a Xylosefermentation and b mixed sugars (xylose and glucose) fermentation.Symbols represent xylose (filled circle), glucose (open circle), lacticacid (filled triangle), acetic acid (open triangle), and cell density (filleddiamond)Appl Microbiol Biotechnolgreatly improved compared with D-lactic acid yield(0.50gg ?1 )andproductivity(0.32gL ?1 h ?1 )fromcornstoverby homofermentative strai。