Eos

HC

a

ARRAA

KPSAlF

1

felamiamcbCom

(

h0

Journal of Biotechnology 188 (2014) 112–121

Contents lists available at ScienceDirect

Journal of Biotechnology

j ourna l ho me pa ge: www.elsev ier .com/ locate / jb io tec

nhancing precursors availability in Pichia pastoris for theverproduction of S-adenosyl-l-methionine employing moleculartrategies with process tuning

arit Ravi Kant, M. Balamurali, Sankaranarayanan Meenakshisundaram ∗

entre for Biotechnology, Anna University, Chennai 600 025, India

r t i c l e i n f o

rticle history:eceived 5 April 2014eceived in revised form 1 August 2014ccepted 18 August 2014vailable online 23 August 2014

eywords:ichia pastoris-Adenosyl-l-methionineTP-Methionine

a b s t r a c t

S-Adenosyl-l-methionine (SAM) is an important metabolite having prominent role in treating variousdiseases. Due to increasing demand of SAM, improvement in its production is essential. For this purpose, S-adenosyl-l-methionine synthetase gene (sam2) was overexpressed in the present study, and we studiedthe effect of coexpression of methionine permease (mup1) and adenylate kinase (adk1) genes. Fromthe recombinant strains expressing individual genes, we observed that SAM2 synthetase is the primarylimiting factor and its overexpression is essential to increase the SAM productivity. Coexpression ofmup1 with sam2 did not enhance SAM production, while coexpression of adk1 with sam2 improved SAMproduction, clearly indicating that ATP is the primary limiting precursor in SAM production. However,coexpression of all three genes synergistically improved SAM productivity with better L-methionine (L-met) conversion efficiency in every stage, and it was 77% more compared to overexpressing sam2 alone.

ed batch cultivation Sparging pure oxygen reduced cultivation time. Feeding nitrogen source and additional L-met duringinduction phase enhanced SAM yield by 38.4% and 55.1%, respectively. Moreover, building up biomassbefore induction resulted in 145% increase in specific yield and 83% higher L-met conversion efficiency.This is the first report on increasing both the precursors L-met and ATP availability through molecularstrategies using microorganisms for the production of SAM.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

S-Adenosyl-l-methionine (SAM) is an essential metaboliteound in all the living cells and is one of the most widely usednzyme substrates after ATP. SAM plays a major role in cellu-ar biochemistry as a precursor to methylation, aminopropylation,nd transsulphuration pathways (Fontecave et al., 2004). Thisolecule has gained importance due to its therapeutic applications

n treating diseases such as depression, liver disease, osteoarthritis,nd cancer (Bottiglieri, 2002; Lu and Mato, 2012). S-Adenosyl-l-ethionine synthetase (EC 2.5.1.6, SAM2, SAM synthetase or MAT)

atalyses the biosynthesis of SAM from L-met and ATP. SAM coulde produced by chemical, enzymatic, and fermentation routes.

hemical methods employ rigorous conditions for chiral separationf the molecule, while in enzymatic method, high cost of ATP hasade the process economically not viable. The progress in the SAM

∗ Corresponding author. Tel.: +91 44 2235 8365; fax: +91 44 2235 0299.E-mail addresses: meenakshi@annauniv.edu, smsuin@gmail.com

S. Meenakshisundaram).

ttp://dx.doi.org/10.1016/j.jbiotec.2014.08.017168-1656/© 2014 Elsevier B.V. All rights reserved.

production has been extensively reviewed by Chu et al. (2013). SAMis accumulated in microbial cells in different quantities, and Saccha-romyces cerevisiae (S. cerevisiae) is able to accumulate this in largequantities (Gawel et al., 1962). Two SAM synthetase genes, sam1and sam2, have been found in yeast (Thomas et al., 1988; Thomasand Surdin-Kerjan, 1991); among the two genes, sam2 is insensitiveto L-met concentration. Chan and Appling (2003) analysed the reg-ulation of SAM levels in yeast and reported that the excess SAM pro-duced in L-met supplemented medium is sequestered to vacuoles,resulting in hyper accumulation of this molecule in yeast. Enhance-ment of SAM production is carried out in three ways: screeningdifferent yeast strains and their mutants for hyper producers,genetic engineering the strains for overproducing SAM, and processoptimisation studies for increasing the volumetric productivity.

Upon screening various microorganisms, Shiozaki et al. (1984)found that yeasts accumulate large quantities of SAM. The authorsdemonstrated that intracellular SAM2 synthetase enzyme activity

and SAM accumulation have positive correlation in the hyper-producing strains. In order to identify SAM producers, screeningvarious mutant strains of S. cerevisiae has also been attempted usingethionine/nystatin resistant screens. In ethionine resistant strains,

H. Ravi Kant et al. / Journal of Biotechnology 188 (2014) 112–121 113

Table 1Overview of vectors, strains and recombinant P. pastoris strains used in this study.

Vectors and strains Relevant characteristics Source

VectorspPICZB PAOX1/(ZeocinR)/His− InvitrogenpPIC6A PAOX1/(BlasticidinR)/His− InvitrogenpPIC3.5K PAOX1/(GeneticinR)/His+ InvitrogenE. coli host strainDH5� Shuttle host for cloning InvitrogenPichia pastoris host strainGS115 Mut+/His− InvitrogenSingle gene over-expressing strainsGS115/sam2 PAOX1-pPICZB-sam2/(ZeocinR)/Mut+ This studyGS115/adk1 PAOX1-pPIC6A-adk1/(BlasticidinR)/Mut+ This studyGS115/mup1 PAOX1-pPIC3.5K-mup1/(GeneticinR)/Mut+ This studyDouble gene over-expressing strainsGS115/sam2-adk1 PAOX1-pPICZB-sam2/PAOX1-pPIC6A-adk1/Mut+ This studyGS115/sam2-mup1 PAOX1-pPICZB-sam2/PAOX1-pPIC3.5K-mup1/Mut+ This studyTriple gene over-expressing strain

IC6A-

asip

(2igmbatstc

mfioddanS

tifpoecif

ipMbBmism

GS115/sam2-adk1-mup1 PAOX1-pPICZB-sam2/PAOX1-pP

ccumulation of SAM occurs by increase in L-met resistance SAM2ynthetase enzyme (Shiomi et al., 1995; Cao et al., 2012), whilen nystatin-resistant strains, the utilisation of SAM by ergosterolathway is reduced (Shobayashi et al., 2006).

Overexpression of sam2 gene in S. cerevisiae and Pichia pastorisP. pastoris) increases SAM production (He et al., 2006; Luo et al.,008). Improvements in SAM production has also been attempted

n P. pastoris using DNA shuffled methionine adenosyltranferaseene for improved enzyme activity (Hu et al., 2009a) and chimericethylenetetrahydrofolate reductase for avoiding feedback inhi-

ition of SAM in L-met metabolism (Chan and Appling, 2003). Inddition to over expressing SAM2 synthetase enzyme in P. pas-oris, He et al. (2006) also attempted knockout of cystathione �ynthase (CBS) enzyme and succeeded in overproducing SAM. Inhis knockout, CBS blocked the conversion of SAM and L-met toysteine.

In order to improve SAM levels in recombinant P. pastoris, opti-isations of L-met feeding (Hu et al., 2007, 2009b), ammonium

eeding (Zhang et al., 2008a) and alternate glycerol-methanol feed-ng (Hu et al., 2008) have also been attempted. In these studies, byptimum feeding, either L-met or ATP, the precursors for SAM pro-uction, was enhanced. To improve oxygen uptake rate in high cellensity cultivation of recombinant P. pastoris, oxygen vectors, suchs 1% heptanes, was used to improve oxygen solubility in combi-ation with other carbon feeding; as a result, there was increase inAM production level (Zhang et al., 2008b).

In the above mentioned studies, the authors either attemptedo enhance the enzyme activity by mutation or genetic engineer-ng studies or tried to increase the precursor availability by processeeding conditions. Genetic engineering strategies to enhance therecursor availability have not been employed except coexpressionf Vitreoscilla haemoglobin gene in P. pastoris for improving ATP lev-ls (Chen et al., 2007). This made us interested in undertaking theurrent project to address the improvement in precursor availabil-ty using genetic engineering strategies with process optimisationor overproduction of SAM.

While studying L-met uptake by S. cerevisiae, Isnard et al. (1996)dentified high affinity (MUP1) and low affinity (MUP3) methionineermeases, which are different from other amino acid transporters.enant et al. (2006) reported that transport of L-met is achieved

y seven membrane permeases Mup1, Mup3, Agp1, Agp3, Bap2,ap3, and Gnp1. They found that in the presence of L-met, the

ajor methionine transporters, Mup1 and Mup3, are repressed

n the yeast. In the current study, we have proposed overexpres-ion of mup1 to improve transport of L-met supplemented in theedium.

adk1/PAOX1-pPIC3.5K-mup1/Mut+ This study

In a study with another methylotrophic yeast, Candida boidinii(C. boidinii), Tani et al. (1994) found that when late exponentialphase cells are used for ATP production, the productivity becomeslow due to the decline in the ADK1 enzyme activity. As ATP is one ofthe precursors in our production system and also P. pastoris beingmethylotrophic yeast, we proposed to overexpress adk1 gene.

The aim of our study was to enhance the precursor availability inSAM production by combining above strategies of overexpressingmup1 and adk1 genes along with sam2 and improvement in L-metconversion efficiency by process strategies.

A schematic metabolic map for SAM biosynthesis with asso-ciated pathways is depicted in supplementary file of the presentstudy.

2. Materials and methods

2.1. Strains and culture conditions

Table 1 lists the strains and vectors used in this work. E. colistrains used for creation of recombinant vectors were cultured inlow salt Luria-Bertani (LB) medium (pH 7.0) containing 10 g tryp-tone, 5 g yeast extract, and 5 g NaCl per litre. P. pastoris strainsfor creating recombinant strain were cultured in the yeast extractpeptone dextrose (YPD) medium (pH 7.0) containing 10 g yeastextract, 20 g peptone, and 20 g dextrose per litre. Suitable antibioticmarkers as in the vector were used for screening the recombinantsin respective media, supplemented with 1.5% agar for both thestrains.

The following three mediums were used for expression studies:(i) the buffered complex medium with glycerol (BMGY) containing10 g yeast extract, 20 g peptone, 100 mmol potassium phosphatebuffer (pH 6.0), 13.4 g yeast nitrogen base without amino acid,20 g glycerol, 20 g (NH4)2SO4, 4 × 10−4 g biotin, and 8 g L-met perlitre; (ii) the basal salts medium (BSM) for P. pastoris cultivationcontaining 26.7 ml of 85% phosphoric acid (H3PO4), 0.93 g calciumsulphate (CaSO4·2H2O), 18.2 g potassium sulphate (K2SO4), 14.9 gmagnesium sulphate (MgSO4·7H2O), 4.13 g potassium hydroxide(KOH), 40.0 g glycerol, 5 ml PTM1 (Stratton et al., 1998) solution perlitre, and pH was adjusted to 4.8 with 28% (w/v) aqueous ammonia(NH4OH); (iii) Pichia trace minerals 1 (PTM1) salt solution contain-ing 6.0 g cupric sulphate (CuSO4·5H2O), 0.08 g sodium iodide (NaI),3.0 g manganese sulphate (MnSO4·H2O), 0.2 g sodium molybdate

(Na2MoO4·2H2O), 0.02 g boric acid (H3BO3), 0.5 g cobalt chloride,20 g zinc chloride, 65 g ferrous sulphate (FeSO4·7H2O), 0.2 g biotin,5.0 ml 98% (w/w) sulphuric acid (H2SO4) per litre; this stock solu-tion was filter sterilised and stored at 4 ◦C.

114 H. Ravi Kant et al. / Journal of Biotec

Table 2Primers used in this study.

Primer name Restrictionsites

Primer sequences (5′–3′)a

sam2 forward XhoI CCGCTCGAGATGTCCAAGAGCAAAACsam2 reverse XbaI GCTCTAGATTAAAATTCCAATTTCTTTGGadk1 forward ApaI CTGGGGCCCATGTCTAGCTCAGAATCCATTAGadk1 reverse XhoI CCGCTCGAGTTAATCCTTACCTAGCTTGmup1 forward SnabI TTCTCCTACGTAATGTCGGAAGGAAAGmup1 reverse NotI CCGGCGGCCGCTCATTACAGCGATTTTTGAOX1 forwardb GACTGGTTCCAATTGACAAGC

a

c

2p

epaDafirvtp

2

BwPtpeeEaaatttiwPs(pseg

TP

The restriction enzyme sites used for cloning are underlined.b AOX1 forward along with gene reverse primers used to confirm the presence of

loned gene in correct frame by genomic DNA PCR.

.2. Construction of the recombinant plasmids pPICZB-sam2,PIC6A-adk1 and pPIC3.5K-mup1

Table 2 lists the primers used in this study and restrictionnzyme sites in the primers. Recombinant plasmids pPICZB-sam2,PIC6A-adk1 and pPIC3.5K-mup1 were constructed using the PCRmplified sam2, adk1 and mup1 genes of S. cerevisiae S288c genomicNA. PCR conditions were optimised in a gradient PCR (eppendorf),nd Table 3 lists the conditions for all amplified genes. PCR ampli-ed sam2, adk1 and mup1 genes were digested with respectiveestriction enzymes in the primer and ligated to similarly digestedectors of pPICZB, pPIC6A, and pPIC3.5K, respectively. The resul-ant plasmids were denoted as pPICZB-sam2, pPIC6A-adk1, andPIC3.5K-mup1.

.3. Construction of recombinant P. pastoris strains

The recombinant plasmid pPICZB-sam2 was linearised withglII, while the other plasmids (pPIC6A-adk1 and pPIC3.5K-mup1)ere linearised with SacI for transformation. Competent cells of

. pastoris host (GS115) strain were made as given in Pichia pro-ocols (Cregg and Russell, 1998), and linearised constructs wereurified with a QIAGEN mini prep kit. About 8–10 �g of lin-arised plasmid was mixed with 80 �l competent cells in 0.2 cmlectroporation cuvette and pulsed (1500 V, 250 �, 50 �F) usingCM® 630 electroporator (BTX, San Diego, CA, USA). Immedi-tely cells were resuspended in 1 ml of 1.0 M ice-cold sorbitolnd kept for incubation. These cells were plated on the YPDgar plates containing ZeocinTM, BlasticidinTM and GeneticinTM forhe selection of recombinant P. pastoris harbouring the respec-ive expression cassette. The colonies appeared on plates afterhree to four days at 28 ◦C (Thompson et al., 2001). The pos-tive transformants containing the desired expression cassette

ere screened and confirmed by PCR (Linder et al., 1996) usingAOX1 forward and insert reverse primers. Recombinant P. pastoristrains, GS115/pPICZB-sam2 (GS115/sam2), GS115/pPIC6A-adk1GS115/adk1) and GS115/pPIC3.5K-mup1 (GS115/mup1) were pre-

ared and named as single gene overexpressing strain. Similarly,trains were made in combinations by exploiting the differ-nt selection pressure in the vectors and named as doubleene (GS115/sam2-adk1; GS115/sam2-mup1) and triple genes

able 3CR conditions for the amplification of sam2, adk1 and mup1 gene.

S. no. Gene name PCR condition

Initial denaturation Cycle

Denaturation

1 sam2 95 ◦C/5 min 95 ◦C/90 s

2 adk1 95 ◦C/5 min 95 ◦C/90 s

3 mup1 95 ◦C/5 min 95 ◦C/90 s

hnology 188 (2014) 112–121

(GS115/sam2-adk1-mup1) overexpressing strains as listed inTable 1.

2.4. Analysis

Biomass was measured in dry cell weight (DCW) as well as inoptical density at 600 nm (OD600). Protein estimation was carriedout by Bradford method. Protein expression profiles were analysedby 10% sodium dodecyl sulphate poly acrylamide gel electrophore-sis (SDS-PAGE) method. The extraction of SAM from the P. pastoriscultures was performed using perchloric acid method with 0.75 Nperchloric acid instead of 0.4 N, as originally described by Wanget al. (2001).

SAM was quantified in high performance liquid chromatography(HPLC, SHIMADZU) using a reversed-phase YMCR ODS C-18 analyt-ical column (4.6 mm × 250 mm, 5 �m particle size) equipped witha UV detector (254 nm). The mobile phase consisted of two buffers:buffer-1 composed of a mixture of 20 mM citric acid with 10 mMsodium dihydrogen orthophosphate per litre millipore water, whilebuffer-2 containing 100% HPLC grade acetonitrile. HPLC bufferwas prepared by mixing 56 volumes of buffer-1 and 44 volumesof buffer-2 in 0.4% (w/v) of sodium lauryl sulphate. Buffer wasdegassed by ultrasonication and filtered through 0.2 �m membranefilters. After equilibrating the HPLC system with buffer, the samplewas injected, and the separation was obtained using isocratic flow.The flow-rate was maintained at 1.5 ml min−1, and the detectionwas monitored at 254 nm. Retention time for SAM was at 12 min.SAM standard plot was constructed using the standard from SIGMA.

2.5. Flask cultivation

A single colony was transferred from YPD agar plate into 3 mlYPD medium and grown at 28 ◦C in orbital shaker until OD600 valuereached 15. The 1 ml of this pre-inoculum was sub-cultured to a50 ml sterile growth medium (BGY) in Erlenmeyer flasks. The cellswere harvested aseptically, once OD600 reached 25, the cells weretransferred to 15 ml expression medium (BMY) with 0.8% (w/v) L-met. Starting the induction with 0.5% (v/v) of methanol initially, themethanol concentration was increased to 1% (v/v) after 24 h andcontinued till 96 h at an interval of 24 h. Samples were harvestedevery 24 h to determine OD600, DCW and SAM yield.

2.6. Bioreactor cultivation

Inoculum for the bioreactor was prepared by transferringrecombinant P. pastoris strains from YPD agar plate to 3 ml YPGmedium and grown at 28 ◦C, 220 rpm until OD600 value reachedto 15. The 3 ml pre-inoculum was sub-cultured aseptically to a150 ml sterile BSM medium in 500 ml Erlenmeyer flasks. OnceOD600 reached 25, the inoculum was transferred to the bioreactor.

Fermentation studies were carried out using a BSM supple-

mented with glycerol as the major carbon source. We used a3.7-l bioreactor (KLF2000, Bioengineering AG, Switzerland) with 2-linitial working volume with the provision of pH and dissolved oxy-gen (DO) controlling system online. Temperature was maintained

No. of cycles Final extension

Annealing Extension

58 ◦C/90 s 72 ◦C/60 s 30 72 ◦C/10 min58 ◦C/60 s 72 ◦C/60 s 30 72 ◦C/10 min49 ◦C/90 s 72 ◦C/90 s 30 72 ◦C/10 min

H. Ravi Kant et al. / Journal of Biotec

Table 4Feeding strategies and bioreactor conditions used for different cultivations of P.pastoris GS115/sam2-adk1-mup1.

Batch Feeding strategies & bioreactor conditions

c1 Normal fed-batch condition (no change in BSM mediumcomposition)

c2 c1 + [20%, DO level maintained by mix of air & 100% pure O2]c3 c2 + [10 g l−1 (NH4)2SO4 pulsed during induction phase (every 6 h,

1 g l−1 pulsed)]c4 c2 + [3 g l−1 glutamine + 2 g l−1 L-met after 10 h of L-met feeding]c5 c4 + [8 g l−1 additional L-met feeding during induction for 20 h]c6 c5 + 40 g l−1 of glycerol containing 12 g l−1 of PTM1 trace salts was

fed at 2.67 g l−1 h−1, i.e. extended glycerol feeding to build high cell

aa1wt

cittooswwf7

2G

odaDiIttbanc2cpbg

3

opocpa

density before induction

t 28 ◦C, and pH was controlled at 4.8 using aqueous ammoniand phosphoric acid. Agitation was increased between 600 and000 rpm to maintain DO above 20%. In the later stage of batcheshen DCW increased above 40 g l−1, DO levels came down to lower

han 20%. Polyoxypropylene glycol (PPG) was used as the antifoam.Cells were grown in BSM medium until initial glycerol depleted

ompletely, and the depletion was indicated by a dramatic increasen the DO to 100%. Glycerol feeding was then initiated to increasehe cell biomass under limited conditions: 40 g l−1 of glycerol con-aining 6 g l−1 of PTM1 trace salts was fed at 2.67 g l−1 h−1. At the endf the glycerol fed-batch, methanol feeding was initiated at a ratef 5 ml l−1 pulses. Methanol pulses were given whenever the DOpikes appeared, indicating the depletion of carbon source. Alongith methanol pulses during the induction phase, L-met feedingas started after eighth hour of post induction at a rate of 8 g l−1

or 10 h. Samples were taken out after every 12 h of induction until2 h for determining the SAM yield and biomass.

.7. Optimisation of cultivation for the P. pastorisS115/sam2-adk1-mup1

Cultivation of control batch in the above conditions was carriedut with the GS115/sam2-adk1-mup1 strain. All the bioreactor con-itions were maintained as mentioned above except for increasedgitation. Pure oxygen was sparged into the system to maintainO level more than 20% throughout the cultivation. Modifications

n the feeding strategies were carried out as shown in Table 4.n c1 cultivation, normal fed batch conditions were maintainedhroughout the process. In c2 cultivation, pure oxygen was spargedo overcome DO limitation by maintaining 20% DO throughout fed-atch culture. In c3 cultivation, similar to c2 conditions with 10 g l−1

mmonium sulphate was fed during induction phase to provideitrogen source, which is necessary for amino acid synthesis. In4 cultivation, similar to c2 conditions with 3 g l−1 glutamine and

g l−1 L-met fed during induction phase. In c5 cultivation, similar to4 conditions with additional 8 g l−1 L-met was fed during inductionhase, whereas, in c6 cultivation, to build high cell density biomassefore induction, similar to c5 conditions with additional 40 g l−1

lycerol during growth phase was fed.

. Results

Our objective was to construct recombinant P. pastoris for the

ver production of pharmaceutically important SAM. For this pur-ose, we cloned the sam2, adk1 and mup1 genes under the controlf methanol induced alcohol oxidase1 (AOX1) promoter for intra-ellular expression by transforming the genes constructs into P.astoris GS115 host. We created different clones individually andlso by combining the above three genes.

hnology 188 (2014) 112–121 115

3.1. Construction and confirmation of strains

We amplified sam2, adk1 and mup1 genes from S. cerevisiaegenome and sub-cloned in P. pastoris expression vectors pPICZB,pPIC6A, and pPIC3.5K, respectively. The recombinant plasmidconstructs, pPICZB-sam2, pPIC6A-adk1 and pPIC3.5K-mup1 weretransformed into P. pastoris host GS115 to form GS115/pPICZB-sam2, GS115/pPIC6A-adk1, and GS115/pPIC3.5K-mup1 strains.Colonies were screened in respective antibiotic marker plates.We confirmed the presence of genes using PCR with AOX1 for-ward primer and gene-specific reverse primer. From each set ofrecombinant strain, five colonies were selected, and SAM produc-tion was tested using shake flask cultivation using BMY mediumwith L-met supplementation. Table 5 shows the results. WhileGS115/sam2 clone produced 50 fold higher SAM than the hoststrain, GS115/adk1 clone produced only three fold higher SAM;GS115/mup1 did not improve SAM production. We selected thehighest expressing GS115/sam2 clone for the creation of doublegene overexpressing strains. Recombinant plasmids pPIC6A-adk1and pPIC3.5K-mup1 were transformed into P. pastoris GS115/sam2for generating GS115/sam2-adk1 and GS115/sam2-mup1 strainsusing respective antibiotic markers in the vector. Here we alsoconfirmed the presence of genes as mentioned above. Shake flaskstudies were carried out similar to single gene expressing clones inthese strains also, and we found that coexpression of adk1 enhancedSAM production by 22% than the single gene strain GS115/sam2.Here also, the mup1 gene failed to enhance the SAM production. Weselected the highest expressing GS115/sam2-adk1 strain for creat-ing triple gene expressing strain and used the recombinant plasmidpPIC3.5K-mup1. After confirming the integration of genes, shakeflask studies were performed. We found increase in SAM produc-tion in the triple gene strains further by 81% than single gene strainGS115/sam2 and 48% than double gene strain GS115/sam2-adk1.Expression of the three enzymes in all the clones was confirmedby SDS-PAGE analysis (Fig. 1). We observed the basal expressionof these proteins because of the presence of inherent sam2, adk1,and mup1 genes. Basal expression of sam2 was more comparable toadk1 and mup1. However, in the induced samples of recombinantstrains, we found higher expression levels of these enzymes.

3.2. Analysis of intracellular ATP concentration and SAMaccumulation in shake flask cultivations

In order to verify the improvement in ATP availability in theadk1 harbouring strains, ATP analysis was carried out after induc-tion in all the strains harbouring adk1 gene along with the GS115host and GS115/sam2 strains. Table 6 presents the results. ATP lev-els after induction in the control strain GS115 was higher at 48 hof induction (2.50 ± 0.20 �mol g−1 DCW) compared to GS115/sam2(0.48 ± 0.03 �mol g−1 DCW). Therefore, it is evident that ATP isone of the limiting factor in the SAM production. When adk1 genewas introduced to the host strain after induction at 48 h, the ATPlevels increased significantly (4.70 ± 0.25 �mol g−1 DCW) and thisincrease can be attributed to the enhancement of SAM productionin other recombinant strains harbouring adk1 gene GS115/sam2-adk1 and GS115/adk1-mup1-sam2 (Table 6).

3.3. Biomass and SAM accumulation in fed batch cultivations

As SAM production is intracellular and its volumetric pro-ductivity increases with increase in cell concentration, bioreactorcultivation for all the single, double and triple clones was carried

out in BSM medium. We used the three stage cultivation strat-egy of P. pastoris with L-met supplementation for these batches.Figs. 2 and 3.c1 (for triple clone) present the profile of biomassconcentration and SAM production in all the batches. Table 7 shows

116 H. Ravi Kant et al. / Journal of Biotechnology 188 (2014) 112–121

Table 5Screening of recombinant P. pastoris strains for SAM production.

S. no. Strains SAM yield (mg l−1)

Colony 1 Colony 2 Colony 3 Colony 4 Colony 5 Average Standarddeviation

1 GS115/host 5.01 4.30 5.00 4.82 4.70 4.77 0.29Single gene over-expressing strains2 GS115/sam2 221.10 219.70 180.40 220.10 217.09 211.68 17.553 GS115/adk1 14.55 16.38 12.30 12.61 15.91 14.35 1.864 GS115/mup1 4.05 5.60 5.62 6.25 4.33 5.17 0.94Double gene over-expressing strains5 GS115/sam2-adk1 276.10 262.58 251.31 262.50 242.02 258.90 12.896 GS115/sam2-mup1 192.30 225.07 248.11 241.21 220.92 225.52 21.69Triple gene over-expressing strain

42

H

tsip

FGaE(

7 GS115/sam2-adk1-mup1 323.10 372.30

ighest SAM production among the colonies is given in bold.

he SAM yield with respect to biomass concentration, L-met conver-ion efficiency, and SAM productivity. In all the batches, cell densityncreased between 65 and 70 g l−1 of DCW in 72 h. Compared to P.astoris host, all the strains produced more SAM. In the single gene

ig. 1. Protein expression profile of recombinant P. pastoris strains. About 50 �l of cell lyS115/sam2, GS115/adk1 and GS115/mup1. (a and c)Lane 1: uninduced sample (control); lt 24 h and 48 h. (b) Lane 1: molecular weight protein marker; lane 2: uninduced samplxpression of double gene overexpressing strain GS115/sam2-adk1 and GS115/sam2-mb,d-f) lane 1: molecular weight protein marker; lane 2: uninduced sample (control); lan

3.31 401.50 398.70 383.78 38.45

strains, the SAM production was the maximum in the GS115/sam2clone, indicating clearly that the overexpression of the sam2 geneis essential. The genes mup1 and adk1 enhanced SAM productivityonly when they were coexpressed with sam2 gene. Among the two

sate was loaded in the gels. (a–c) Expression of single gene overexpressing strainane 2: molecular weight protein marker; lanes 3 and 4: induction samples collectede (control); lanes 3 and 4: induction samples collected at 24 h and 48 h. (d and e)up1. (f) Expression of triple gene overexpressing strain GS115/sam2-adk1-mup1;es 3 and 4: induction samples collected at 24 h and 48 h.

H. Ravi Kant et al. / Journal of Biotechnology 188 (2014) 112–121 117

Table 6Analysis of ATP concentration, energy charge, and SAM yield in post induction samples during shake flask cultivations.

S. no. Strains ATP concentration (�mol g−1 DCW) Energy chargea (EC) SAM yield (mg l−1)

24 h 48 h 24 h 48 h 24 h 48 h

1 GS115/host 1.86 ± 0.12 2.50 ± 0.20 0.46 ± 0.013 0.44 ± 0.012 4.63 ± 0.31 5.35 ± 0.442 GS115/sam2 2.09 ± 0.15 0.48 ± 0.03 0.50 ± 0.015 0.30 ± 0.010 91.15 ± 5.21 223.99 ± 13.273 GS115/adk1 3.77 ± 0.21 4.70 ± 0.25 0.64 ± 0.016 0.68 ± 0.016 8.94 ± 0.73 19.38 ± 2.134 GS115/sam2-adk1 3.48 ± 0.20 2.70 ± 0.12 0.60 ± 0.015 0.59 ± 0.014 123.74 ± 10.3 278.03 ± 14.54

0

pnaW

FPl

5 GS115/sam2-adk1-mup1 2.77 ± 0.13 2.33 ± 0.10

a Energy charge was calculated as per Atkinson (1968).

recursor enhancing genes, coexpression of mup1 with sam2 didot have any significant effect on productivity, but its expressionlong with other precursor genes enhanced the production of SAM.hen all the three genes were expressed simultaneously, the yield

ig. 2. Profile of biomass (DCW) and SAM yield during fed batch cultivation of recombina. pastoris GS115/adk1; (d) P. pastoris GS115/mup1; (e) P. pastoris GS115/sam2-adk1; and

-methionine feeding. Data are presented as means ± SE of three replicates.

.59 ± 0.014 0.57 ± 0.013 347.63 ± 16.1 434.41 ± 21.13

was enhanced by 75% compared to GS115/sam2 and GS115/sam2-mup1 and was 32% higher than GS115/sam2-adk1. One of thecritical parameters in SAM production is L-met conversion effi-ciency as it is an expensive substrate and a nutritional source

nt P. pastoris strains. (a) P. pastoris host strain GS115; (b) P. pastoris GS115/sam2; (c)(f) P. pastoris GS115/sam2-mup1. GF, glycerol feeding; MF, methanol feeding; LMF,

118 H. Ravi Kant et al. / Journal of Biotechnology 188 (2014) 112–121

Fig. 3. Profile of biomass (DCW), SAM yield and DO during fed batch cultivation of recombinant P. pastoris triple gene overexpressing strain GS115/sam2-adk1-mup1 withdifferent culture conditions (c1–c6). c1: normal fed-batch conditions for GS115/sam2-adk1-mup1; c2: similar to c1 with constant DO maintained at 20%; c3: similar to c2with addition of 10 g l−1 (NH4)2SO4 during induction phase; c4: similar to c2 with addition of 3 g l−1 glutamine and 2 g l−1 L-met after 10 h L-met feeding in induction phase;c5: similar to c4 with 8 g l−1 additional L-met feeding during induction for 20 h; c6: similar to c5 conditions with addition of 40 g l−1 of glycerol containing 12 ml l−1 of PTM1

trace salts was fed at 2.67 g l−1 h−1 before induction. GF, glycerol feeding; MF, methanol feeding; LMF, l-methionine feeding. GF1, extended glycerol feeding in c6; LMF2,additional 8 g l−1 l-methionine feeding in c5 and c6. Data are presented as means ± SE of three replicates.

Table 7Summary of biomass, volumetric SAM yield, L-met conversion efficiency, specific SAM yield and SAM productivity during the fed-batch cultivation of recombinant P. pastorisstrains.

Strains Biomass (g l−1) L-met feed (g) SAM yield (g l−1) L-met conversionefficiency (%)

Specific SAM yield(g SAM)(g DCW)−1

SAM productivity(g SAM)(g DCW)−1(h)−1

GS115/host 64.49 20 0.03 0.17 0.0005 0.00001GS115/sam2 68.67 20 1.33 7.47 0.0194 0.00040GS115/adk1 66.93 20 0.10 0.55 0.0015 0.00004GS115/mup1 66.97 20 0.18 1.22 0.0027 0.00005GS115/sam2-adk1 69.32 20 1.78 9.99 0.0257 0.00053GS115/sam2-mup1 66.70 20 1.34 7.51 0.0201 0.00042GS115/sam2-adk1-mup1 69.13 20 2.36 13.25 0.0341 0.00071

H. Ravi Kant et al. / Journal of Biotechnology 188 (2014) 112–121 119

Table 8Summary of biomass, volumetric SAM yield, L-met conversion efficiency, specific SAM yield and SAM productivity during the fed-batch cultivation optimisation forrecombinant P. pastoris strain GS115/sam2-adk1-mup1.

Batch conditions Biomass (g l−1) L-met feed (g) SAM yield (g l−1) L-met conversionefficiency (%)

Specific SAM yield(g SAM)(g DCW)−1

SAM specific productivity(g SAM)(g DCW)−1(h)−1

c1 69.30 20 2.36 13.25 0.0341 0.00071c2 73.09 20 2.66 14.92 0.0364 0.00076c3 69.93 20 3.30 18.47 0.0472 0.00079c4 74.36 25 3.93 17.63 0.0529 0.00110

13.424.2

fceo1ci

3o

cfFprdiTtcctcmHtitaIatwt1sctylgcip2cw

4

t

c5 70.14 45 5.39c6 116.48 45 9.73

or cellular metabolism. On comparing the bioconversion effi-iency of SAM2 expressing clones, GS115/sam2 clone showed 7.47%fficiency that was further increased upon coexpression with thether two precursor enhancing genes, reaching a maximum of3.25% for GS115/sam2-adk1-mup1. Similarly, the SAM yield coeffi-ient with respect to biomass and SAM productivity also increasedn these clones (Table 7).

.4. Effects of fermentation process conditions on SAMverproduction

To further improve SAM productivity and L-met conversion effi-iency, the process conditions were optimised as given in Table 4or recombinant P. pastoris strain GS115/sam2-adk1-mup1. Whileig. 3 depicts the profile of biomass concentration, DO, and SAMroduction for all the optimised fed batches, Table 8 presents theespective SAM yield, L-met conversion efficiency, and SAM pro-uctivity. In the cultivation c1, when the biomass concentration

ncreased above 10 g l−1, the DO saturation dropped less than 5%.he DO level was unresponsive to changes in aeration and agita-ion, resulting in oxygen-limited condition in the cultivation. In the2 cultivation, this limitation was overcome by switching the DOontrol from agitation to pure oxygen sparging. Though in c2 cul-ivation, SAM yield was similar to the oxygen limited cultivation1, the growth rate of culture in the biomass build-up phase wasuch higher, resulting in reducing the cultivation time by 24 h.ence, this condition was maintained in the subsequent cultiva-

ions. To enhance the availability of nitrogen source during thenduction phase, which is an essential substrate for adenine syn-hesis, we adopted two cultivation methods, one with addition ofmmonium sulphate (c3) and another with additional L-met (c4).n both the cultivations, c3 and c4, SAM yield increased by 22%nd 48%, respectively, and L-met conversion efficiency increasedo 18.47% and 17.63%, respectively. Considering the above results,e added additional 8 g l−1 L-met during the induction phase in

he subsequent c5 cultivation. This indeed improved SAM yield by00% compared to the c2 cultivation; however, the L-met conver-ion efficiency lowered to 13.43%, a 25% decrease compared to the3 and c4 cultivations. But in all the batches, we observed a sequen-ial increase from 0.034 g-g−1 DCW to 0.0769 g-g−1 DCW in SAMield per unit cell mass. Though L-met conversion efficiency wasower in c5 batch, the volumetric and specific yield of SAM wasood (5.39 g-l−1 and 0.0769 g-g−1 DCW). Hence, in the subsequentultivation, before the induction phase, biomass concentration wasncreased with additional glycerol in the limited carbon fed batchhase. In the c6 cultivation, the conversion efficiency improved to4.23% with a 265% and 80% increase in SAM yield compared to the2 and c5 cultivations, respectively. The specific SAM to cell yieldas maintained as in c5.

. Discussion

Several studies have been reported on maximising SAM produc-ion by overexpressing SAM synthetase in different yeast systems

3 0.0769 0.001283 0.0835 0.00139

such as S. cerevisiae, Saccharomyces sake, P. pastoris, Klyveromyceslactis (Chu et al., 2013). Among all these systems, P. pastoris wasreported to be hyper producers, because the cell density can bebuilt to very high levels, and SAM being an intracellular metabo-lite, the volumetric SAM yield and SAM specific productivity werehigher. In SAM production, apart from overexpressing the enzymecatalysing the reaction, increase in the availability of the reactants(L-met and ATP) is essential for higher SAM production.

In an attempt to improve the precursor availability, we coex-pressed two genes mup1 and adk1. First, to check whether theimprovement in the precursors alone enhances SAM produc-tion, the single expressing strains (GS115/mup1 and GS115/adk1)were created and compared with strain expressing catalyst(GS115/sam2). From the results (Tables 5 and 7), it is evident thatoverexpressing sam2 gene is essential for SAM production, and itincreased the production by 10 fold compared to the mup1 andadk1 strains. However, compared to the host strain, there wasslight increase in SAM production in these strains. In the nextstep, these genes were coexpressed with sam2 gene. Among thetwo strains created, GS115/sam2-adk1 strain was able to produce1.78 g l−1 SAM (Fig. 2e) compared to 1.33 g l−1 in the GS115/sam2alone (Fig. 2b). This improvement in yield by 30% is due to theincreased availability of ATP. Sakai et al. (1995) attempted highlevel ATP production using methylotrophic yeast C. boidinii. In theirearlier study (Tani et al., 1994), they carried out the enzymatic anal-ysis of ATP production in C. boidinii and showed that conversionof AMP to ATP catalysed by ADK1 could be the rate-limiting stepfor ATP production. In an attempt to improve the ATP productivity,the authors overexpressed adk1 gene from S. cerevisiae in C. boidiniiand it enhanced productivity. As P. pastoris is also a methylotrophicyeast, we made an attempt for the first time overexpressing adk1gene in this strain and found that it indeed improved ATP produc-tion which in turn resulted in higher SAM production (Table 7). TheATP levels and adenylate energy charge (AEC) (Atkinson, 1968) inthe host cells, GS115/sam2, and the recombinants harbouring adk1gene were compared (Table 6). In that the ATP levels was lower inthe host strain and GS115/sam2 strain while all the strains harbou-ring adk1 gene were found to increase substantially. In P. pastorisfed batch for production of recombinant galactosidase using GS115mut+ strain, Plantz et al. (2006) estimated AEC at the end of glyc-erol growth phase and initiation of methanol induction phase. Theyobserved that the AEC dropped to 0.6 when methanol inductionphase was started and decreased further during production phase.In this study we also found the AEC values were 0.46 ± 0.013 24 hafter induction in the host strain and maintained at the same levelduring the induction phase. But in GS115/sam2 strain this AECdropped further to 0.3 ± 0.010 after 48 h of induction since ATP isutilised for SAM production. In the case of strains harbouring adk1gene we found the AEC levels enhanced by 40% compared to hoststrain (0.60 ± 0.015) and also maintained during the SAM produc-

tion phase. This proved that integration of adk1 gene in methanolutilising P. pastoris enhanced the ATP production.

On the other hand, coexpression of another precursor genemup1 in GS115/sam2-mup1 strain did not improve the SAM

1 Biotec

puatsmstpmpoospiASrSiaibmcaaiaFttto

tttthoepvthc

ibagt(

digsintta1

20 H. Ravi Kant et al. / Journal of

roduction (Table 7). The amino acid transport in yeast and itsptake by cells is mediated by several amino acid permeases, whichre regulated transcriptionally and post translationally in responseo the variation in intracellular amino acid concentrations. Whiletudying methionine uptake by S. cerevisiae by different per-eases, Isnard et al. (1996) found that MUP1 is a high affinity

pecific permease for methionine. Menant et al. (2006) reportedhat methionine transport is achieved through seven membraneermeases. They found, with supplementation of extracellular L-et, the specific methionine permeases Mup1, Mup3 and general

ermease Agp3 were repressed transcriptionally by 20 fold, whilether general permeases were improved by three to ten-fold. Basedn these observations, the L-met flux by general permeases may beufficient for SAM production in the GS115/sam2-mup1. The SAMroduction improvement in GS115/sam2 was also not observed

n GS115/sam2-mup1 strain due to limitation of other precursorTP. As ATP is also used for cellular energy, its availability forAM production is a possible limiting factor. Shiozaki et al. (1986)eported that competition between cellular energy demand andAM production resulted in limitation of ATP for SAM productionn S. sake. To improve the ATP level, Hu et al. (2007) attemptedlternate feeding strategy of methanol and glycerol during thenduction phase and succeeded in accelerated SAM synthesis ratey enhanced ATP generation during glycerol feeding. In terms ofolecular approach, Chen et al. (2007) found coexpression of intra-

ellular Vitreoscilla haemoglobin in P. pastoris increased metabolicnd respiratory activity that would improve energy metabolismnd ATP synthesis rate; coexpression with SAM2 synthetase alsomproved SAM production. All these studies clearly indicated ATPs one of the limiting factors in enhancing the SAM production.rom our findings on double gene expressing strains, it is evidenthat among the two precursors ATP is the major limiting factorhan L-met. This could be further substantiated by improvement inhe L-met conversion efficiency by 30% in GS115/sam2-adk1 thanther two strains GS115/sam2 and GS115/sam2-mup1 (Table 7).

In the triple gene expressing strain GS115/sam2-adk1-mup1,he SAM production improved to 2.36 g l−1 (Fig. 3c1) comparedo 1.78 g l−1 obtained from GS115/sam2-adk1 (Fig. 2e). This showshat for improved ATP production in the adk1 coexpressing strain,he L-met flux by general permeases is not sufficient; on the otherand, enhancement of L-met production by high affinity methi-nine permease has resulted in higher production of SAM. Hut al. (2009b) reported improving intracellular L-met availability byrocess strategies improved the SAM productivity and L-met con-ersion rate. In the triple gene expressing strain, improvement inhe intracellular methionine availability by overexpressing mup1, aigh affinity methionine specific permease and enhanced the L-metonversion efficiency further by 32.6% (Table 8).

Oxygen limitation is one of the common problems encounteredn high cell density cultivation. In our attempt to increase the solu-ility of oxygen, pure oxygen was used to avoid oxygen limitationnd this resulted in improved growth rate of the organism duringrowth phase which in turn resulted in the reduction of cultivationime by 24 h along with marginal improvement in SAM productionFig. 3c2).

Zhang et al. (2008a) studied the effect of ammonia on SAM pro-uction and found that the availability of nitrogen source during

nduction phase enhanced SAM production. In the improved tripleene expressing strain, this strategy with addition of ammoniumulphate in the induction phase in the c3 batch along with increas-ng the concentration of L-met (as it can be used as both carbon anditrogen source in the c4 batch) was tested. In both the batches,

he SAM production increased (Fig. 3c3 and c4), but increase inhe specific yield and productivity were higher with increasedddition of L-met (Table 8). Increasing L-met concentration to8 g l−1 in the induction phase in c5 batch resulted in enhancing

hnology 188 (2014) 112–121

the SAM production (Fig. 3c5) as well as SAM yields per gramof biomass. This could be due to the saturation in accumulationof maximum amount of SAM per cell. To confirm this, in thesubsequent batches, biomass concentration was improved beforeinduction phase (Table 8); conditions in the induction phasewere maintained as in c5 batch. Effect of cell densities on SAMproduction was investigated by Zhang et al. (2008c), and theyfound that building cell density before induction enhanced theSAM’s yield. In c6 batch, both SAM production and L-met conver-sion efficiency improved by 80% as compared to the c5 batch. Thespecific SAM yield per gram of cell was 0.0835 g-g−1 DCW, whichwas marginally higher than the c5 batch (Table 8).

Regarding SAM overproduction He et al. (2006) reported highestproduction of 13.5 g l−1 by knocking-in and knocking-out strate-gies of SAM utilising pathway. But in their studies they have tosupplement expensive glutathione and histidine to achieve the pro-duction target. Shiozaki et al. (1986) reported SAM production of10.8 g l−1 and they also supplemented more methionine and biotinfor achieving this productivity. Zhang et al. (2008a) reported SAMproduction of 11.63 g l−1 with the induction time of 96 h. In allthese studies, the reported SAM yield with respect to biomass waslower than the yield obtained in our study and in addition eitherthe induction time was more than the present study or expensivesupplements, such as glutathione and histidine, were made in thecultivations.

In conclusion, we created the recombinant P. pastoris overex-pressing strains for the overproduction of SAM and found that theoverexpression of SAM2 synthetase enhanced SAM production. AsATP was the limiting factor in SAM production, cloning adk1 geneimproved the ATP and SAM production. Furthermore, coexpressionof sam2, adk1 and mup1 genes along with building high cell densityculture before induction, synergistically improved the SAM pro-duction and specific SAM yield to 9.73 g l−1 and 145%, respectively.Henceforth, to the best of our knowledge, this is the first reportof improvement in precursor availability by molecular strategiesin sam2 engineered P. pastoris with mup1 and adk1, resulting inenhanced SAM production. Our findings revealed that among thetwo precursors, ATP is the major limiting factor. However, with theimproved ATP supply, L-met flux by general permeases is limited.Improvement in L-met flux with high affinity permease synergisti-cally improved SAM production.

Acknowledgements

The authors are thankful to Council of Scientific and IndustrialResearch (CSIR), Government of India for providing Senior ResearchFellowship (SRF) to Ravi Kant Harit and All India Council for Techni-cal Education (AICTE), Government of India for providing NationalDoctoral Fellowship (NDF) to M. Balamurali. The authors are thank-ful to Department of Biotechnology (DBT), Government of India forfunding this project.

Appendix A. Supplementary data

Supplementary material related to this article can be found,in the online version, at http://dx.doi.org/10.1016/j.jbiotec.2014.08.017.

References

Atkinson, D.E., 1968. The energy charge of the adenylate pool as a regulatory param-eter. Interaction with feedback modifiers. Biochemistry 7 (11), 4030–4034.

Bottiglieri, T., 2002. S-Adenosyl-l-methionine (SAMe): from the bench to thebedside-molecular basis of a pleiotrophic molecule. Am. J. Clin. Nutr. 76 (Suppl.),1151S–1157S.

Cao, X., Yang, M., Xia, Y., Dou, J., Chen, K., Wang, H., Xi, T., Zhou, C., 2012.Strain improvement for enhanced production of S-adenosyl-l-methionine in

Biotec

C

C

C

C

F

G

H

H

H

H

H

I

L

L

L

M

H. Ravi Kant et al. / Journal of

Saccharomyces cerevisiae based on ethionine-resistance and SAM synthetaseactivity. Ann. Microbiol. 62, 1395–1402.

han, S.Y., Appling, D.R., 2003. Regulation of S-adenosylmethionine levels in Saccha-romyces cerevisiae. J. Biol. Chem. 278, 43051–43059.

hen, H.X., Chu, J., Zhang, S.L., Zhuang, Y.P., Qian, J., Wang, Y.H., Hu, X.Q., 2007. Intra-cellular expression of Vitreoscilla haemoglobin improves S-adenosylmethionineproduction in a recombinant Pichia pastoris. Appl. Microbiol. Biotechnol. 74 (6),1205–1212.

hu, J., Qian, J., Zhuang, Y., Zhang, S., Li, Y., 2013. Progress in the research ofS-adenosyl-l-methionine production. Appl. Microbiol. Biotechnol. 97, 41–49.

regg, J.M., Russell, K.A., 1998. Transformation methods. Methods in Molecular Biol-ogy: Pichia Protocols, vol. 103. Humana Press, Totowa, NJ, pp. 27–39.

ontecave, M., Atta, M., Mulliez, E., 2004. S-adenosylmethionine: nothing goes towaste. Trends Biochem. Sci. 29, 243–249.

awel, L.J., Turner, J.R., Parks, L.W., 1962. Accumulation of S-adenosylmethionine bymicroorganisms. J. Bacteriol. 83, 497–499.

e, J., Deng, J., Zheng, Y., Gu, J., 2006. A synergistic effect on the pro-duction of S-adenosyl-l-methionine in Pichia pastoris by knocking in ofS-adenosyl-l-methionine synthase and knocking out of cystathionine-� syn-thase. J. Biotechnol. 126 (4), 519–527.

u, H., Qian, J., Chu, J., Wang, Y., Zhuang, Y.P., Zhang, S.L., 2009a. DNAshuffling of methionine adenosyltransferase gene leads to improvedS-adenosyl-l-methionine production in Pichia pastoris. J. Biotechnol. 141,97–103.

u, H., Qian, J., Chu, J., Wang, Y., Zhuang, Y.P., Zhang, S.L., 2009b. Optimizationof l-methionine feeding strategy for improving S-adenosyl-l-methionine pro-duction by methionine adenosyltransferase overexpressed Pichia pastoris. Appl.Microbiol. Biotechnol. 83, 1105–1114.

u, X.Q., Chu, J., Zhang, S.L., Zhuang, Y.P., Wang, Y.H., Zhu, S., Zhu, Z.G., Yuan, Z.Y.,2007. A novel feeding strategy during the production phase for enhancing theenzymatic synthesis of S-adenosyl-l-methionine by methylotrophic Pichia pas-toris. Enzyme Microb. Technol. 40, 669–674.

u, X.Q., Chu, J., Zhang, Z., Zhang, S.L., Zhuang, Y.P., Wang, Y.H., Guo, M.J.,Chen, H.X., Yuan, Z.Y., 2008. Effects of different glycerol feeding strategies onS-adenosyl-l-methionine biosynthesis by PGAP-driven Pichia pastoris overex-pressing methionine adenosyltransferase. J. Biotechnol. 137, 44–49.

snard, A.D., Thomas, D., Surdin-Kerjan, Y., 1996. The study of methionine uptakein Saccharomyces cerevisiae reveals a new family of amino acid permeases. Mol.Biol. 262, 473–484.

inder, S., Schliwa, M., Kube-Granderath, E., 1996. Direct PCR screening of Pichiapastoris clones. Biotechniques 20, 980–982.

u, S.C., Mato, J.M., 2012. S-adenosylmethionine in liver health, injury, and cancer.Physiol. Rev. 92, 1515–1542.

uo, Y., Yuan, Z., Luo, G., Zhao, F., 2008. Expression of secreted His-tagged S-adenosylmethionine synthetase in the methylotrophic yeast Pichia pastoris and

its characterization, one-step purification, and immobilization. Biotechnol. Prog.24, 214–220.

enant, A., Barbey, R., Thomas, D., 2006. Substrate-mediated remodeling of methi-onine transport by multiple ubiquitin-dependent mechanisms in yeast cells.EMBO J. 25, 4436–4447.

hnology 188 (2014) 112–121 121

Plantz, B.A., Sinha, J., Villarete, L., Nickerson, K.W., Schlegel, V.L., 2006. Pichia pastorisfermentation optimization: energy state and testing a growth-associated model.Appl. Microbiol. Biotechnol. 72, 297–305.

Sakai, Y., Rogi, T., Takeuchi, R., Kato, N., Tani, Y., 1995. Expression of Saccharomycesadenylate kinase gene in Candida boidinii under the regulation of its alcoholoxidase promoter. Appl. Microbiol. Biotechnol. 42, 860–864.

Shiomi, N., Fukuda, H., Kimura, A., 1995. Improvement of S-adenosylmethionineproduction by integration of the ethionine resistance gene into chromo-somes of the yeast Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol. 42,730–733.

Shiozaki, S., Shimizu, S., Yamada, H., 1986. Production of S-adenosyl-l-methionineby Saccharomyces sake. J. Biotechnol. 4, 345–354.

Shiozaki, S., Shimizu, S., Yamada, H., 1984. Unusual intracellular accumulationof S-adenosyl-l-methionine by microorganisms. Agric. Biol. Chem. 48 (9),2293–2300.

Shobayashi, M., Mukai, N., Iwashita, K., Hiraga, Y., Iefuji, H., 2006. A new method forisolation of S-adenosylmethionine (SAM)-accumulating yeast. Appl. Microbiol.Biotechnol. 69, 704–710.

Stratton, J., Chiruvolu, V., Meagher, M., 1998. High cell-density fermentation. Meth-ods in Molecular Biology: Pichia Protocols, vol. 103. Humana Press, Totowa, NJ,pp. 107–120.

Tani, Y., Yonehara, T., Sakai, Y., 1994. Improvement of ATP production with cells of amethylotrophic yeast, Candida boidinii, by genetic engineering. Proc. Japan Acad.70, 53–57.

Thomas, D., Rothstein, R., Rosenberg, N., Surdin-Kerjan, Y., 1988. SAM2encodes the second methionine S-adenosyl transferase in Saccharomycescerevisiae: physiology and regulation of both enzymes. Mol. Cell. Biol. 8,5132–5139.

Thomas, D., Surdin-Kerjan, Y., 1991. The synthesis of the two S-adenosyl-methioninesynthetases is differently regulated in Saccharomyces cerevisiae. Mol. Gen. Genet.226, 224–232.

Thompson, J.R., Reqister, E., Curotto, J., Kurtz, M., Kelly, R.W., Kramer, P.M., Yang,S., Pereira, M.A., Tao, L., 2001. Reversed phase high-performance liquid chro-matography procedure for the simultaneous determination of SAM and SAH inmouse liver and the effect of methionine on their concentrations. J. Chromatogr.B: Biomed. Sci. Appl. 762, 59–65.

Wang, W., Kramer, P.M., Yang, S., Pereira, M.A., Tao, L., 2001. Reversed phasehigh-performance liquid chromatography procedure for the simultaneousdetermination of SAM and SAH in mouse liver and the effect of methionine ontheir concentrations. J Chromato B: Biomedical Sciences and Applications. 762,59–65.

Zhang, J., Wang, X., Zheng, Y., Fang, G., Wei, D., 2008a. Enhancing yield ofS-adenosylmethionine in Pichia pastoris by controlling NH4

+ concentration. Bio-process Biosyst. Eng. 31, 63–67.

Zhang, J.G., Wang, X.D., Zhang, J.N., Wei, D.Z., 2008b. Oxygen vectors used for S-

adenosyl-l-methionine production in recombinant Pichia pastoris with sorbitolas supplemental carbon source. J. Biosci. Bioeng. 105, 335–340.

Zhang, J., Wang, X., Su, E., Fang, G., Ren, Y., Wei, D., 2008c. A new fermentationstrategy for S-adenosylmethionine production in recombinant Pichia pastoris.Biochem. Eng. J. 41, 74–78.