Journal of Biotechnology 141 (2009) 97103

Contents lists available at ScienceDirect

Journal of Biotechnology

journa l homepage: www.e lsev ier .com

DNA shufing of methionine adenosyltransferaseS-adenosyl-l-methionine production in Pichia pa

Hui Hu, J uanState Key Labor y (ShaEast China Uni

a r t i c l

Article history:Received 22 OReceived in reAccepted 17 M

Keywords:S-adenosyl-l-mMethionine adDNA shufingPichia pastoris

n imase (scherhe shed fohe Mrtingeducin thy relaof SAM

recombinant strain GS115/DS56, which showed a favorable foreground in industrial scale production ofSAM.

2009 Elsevier B.V. All rights reserved.

1. Introduc

S-Adenomethyl donto acting astranssulfurainterest intive disordeTherefore, sduce SAM,2005) and K

Methionknown enzand l-methPichia pasto(Mato et awere constr

Abbreviatiocell weight; l-tion per minutliquid per min

Correspon Correspon

E-mail addjuchu@ecust.e

0168-1656/$ doi:10.1016/j.jtion

syl-l-methionine (SAM) is the major intracellularor in all living organisms (Park et al., 1996). In additiona methyl donor, SAM also plays an important role intion and polyamine synthesis reactions. ConsiderableSAM production has arisen because it can treat affec-rs, liver disease, and neurological effectively (Lu, 2000).everal strains have been screened successively to pro-such as yeast (Shiozaki et al., 1986; Shobayashi et al.,luyveromyces lactis (Mincheva and Balutsov, 2002).ine adenosyltransferase (MAT, EC2.5.1.6) is the onlyyme to synthesize SAM in organisms. It catalyzes ATPionine (l-Met) to synthesize SAM. The recombinantris (Li et al., 2002; Yu et al., 2003) and Escherichia colil., 1995; Yang et al., 2002) that over-expressed MATucted for SAM production. A positive correlation was

ns: ATP, adenosine triphosphate; DO, dissolved oxygen; DCW, dryMet, l-methionine; MAT, methionine adenosyltransferase; rpm, rota-e; SAM, S-adenosyl-l-methionine; VVM, volume of gas per volume ofute.ding author. Tel.: +86 021 64253021; fax: +86 021 64253702.ding author. Tel.: +86 021 64250736; fax: +86 021 64253702.resses: jiangchaoqian@ecust.edu.cn (J. Qian),du.cn (J. Chu).

revealed between SAM accumulation and MAT activity. Therefore,it is of great importance to enhance MAT activity for SAM accumu-lation.

IncreasingMAT expression level is a routinemethod to enhancethe enzyme activity. However, SAM, but not MAT, is the desiredproduct of the bioprocess. Over-expression of MAT gene also mightinhibit cell growth and SAM production due to the competitionof resource and energy. Therefore, molecular evolution of MAT toincrease its specic enzyme activity is a better way to enhanceMAT activity for SAM accumulation. DNA is widely used for in vitromolecular evolution (Crameri et al., 1998). It involves the recombi-nation of multiple homologous DNA sequences to create a hybridgene library for selection. Thismethodhas been shown to be a pow-erful engineering approach to tailor theproperties of enzymes, suchas thermostability (Suen et al., 2004; Emond et al., 2008), activity(Tatsuya and Yoshinori, 2007; Ryu et al., 2008), and pH stability (Liuet al., 2005) of the enzymes.

The methylotrophic yeast P. pastoris has been developed intoa commercially important host for the production of heterologousproteins because it has theunique ability to growonminimalmediaat very high cell densities with the strong, tightly regulated alco-hol oxidase promoter (Cereghino and Cregg, 2000). Furthermore,excess SAM in yeast is sequestered in the vacuole, so yeast itself canaccumulate high level of SAM (Chan and Appling, 2003; Megumiet al., 2007). Combined with these advantages, the recombinant P.pastoris has the potential to accumulate higher level of SAM.

see front matter 2009 Elsevier B.V. All rights reserved.biotec.2009.03.006iangchao Qian , Ju Chu , Yong Wang, Yingping Zhatory of Bioreactor Engineering, National Engineering Research Center for Biotechnologversity of Science & Technology, 130 Meilong Road, Shanghai 200237, PR China

e i n f o

ctober 2008vised form 16 February 2009arch 2009

ethionineenosyltransferase

a b s t r a c t

S-Adenosyl-l-methionine (SAM) is aenhance methionine adenosyltransfernant Pichia pastoris, MAT genes from Ewere recombined by DNA shufing. Tstruct recombinant strains and screenIn the two best recombinant strains, trecombinant strains containing the sta65% respectively. The analysis on the dthe K18R mutation probably resultedinsights into structure-enzyme activitevolution of MAT. Finally, a 6.14g l1/ locate / jb io tec

gene leads to improvedstoris

g, Siliang Zhangnghai),

portant molecule for normal cell function and survival. ToMAT, EC2.5.1.6) activity and thus SAM production in recombi-ichia coli, Saccharmyces cerevisiae, and Streptomyces spectabilisufed genes were transformed into P. pastoris GS115 to con-r high MAT activity and enhanced SAM production mutants.AT activities were respectively 201% and 65% higher than theMATgenes, and the SAMconcentration increasedby103% anded sequences of ve representativeMAT variants showed thate increased activity of the best MAT, which may provide newtionship of MAT and might shed light on the further rationalproduction was reached in a 500 l bioreactor with the best

98 H. Hu et al. / Journal of Biotechnology 141 (2009) 97103

In this study, three homologous MAT genes from Saccharmycescerevisiae, E.coliDH5 and Streptomyces spectabilis (GenBank acces-sion nos. M23368, AE000377, and AF117274), which share 5262%nucleic acid identity, were shufed to generate a hybrid MAT genelibrary. The shufed MAT genes were then transformed into P.pastoris GS1ity and SAMachieved intion, whichthe recombsequence aconstructioaltered actimumSAMcbioreactor,production

2. Materia

2.1. Genera

E. coli Ddard growtPlasmid pPvector inP. ptoris GS115of Pichia exlands). TheDNA ligatioof E.coli DH(Takara, ToTaKara Corp

2.2. Culture

The MDtoris strainsthe copy nurecombinanExpression

The cultiasks or 15described pin the inductions of meto a nal cl-Met (Bafecentrationsynthesis.

For 5 l bbatch phasdescribed psterile 50%solution (Inabout 16hwas inducetrace salts sl-Met powdthree timesof induction

The cultsimilar to th(with 1.2% ((v/v) PTM1tively.

Table 1Primers used for amplication of homologous MAT genes.

Source Primers (53)a

E.coli DH5, CCAGAATTCATGGCAAAACACCTTTTTACGTC(forward primer)

siae

bilis

ences) are u

asmi

MATm thilis wPrimMATstrucPlasmon ainearlectecin penetn YP) weST, rI and(deno

NA sh

ee 5K/E

AOXTGGnepted tos, 199s chacyclet 72usin

a, Tobrarnsformed into E.coli DH5 to maintain and propagate thed libraries. Recombinant E.coli DH5 colonies (104) werened, and recombinant plasmids were extracted and puried.ried recombinant plasmids were linearized by BglII diges-d transformed into P. pastoris GS115 for screening.

reening procedures

ombinant strains harboring shufed MAT genes were cul-in 15ml tube and the intracellular SAM concentration wasd after cultivation. Strains with improved SAM productiononrmed by cultivation in 250ml shake asks. Recombinantharboring wild-type MAT genes were used as the control.15 to enhance SAM production. The highest MAT activ-concentration of 463Ug1 DCW and 1.64g l1 werethe best strain (GS115/DS56) in shake ask cultiva-had 201% and 102% improvement respectively over

inant strain harboring MAT gene from S. cerevisiae. Thelignment, conservation analysis, and tertiary structuren were performed to speculate the mechanism of thevities of ve selected shufed MATs. Finally, the maxi-oncentration inGS115/DS56 reached 6.14g l1 in a 500 lwhich showed a favorable foreground in commercialof SAM.

ls and methods

l methods

H5 was used for routine subcloning employing stan-h media and conditions (Sambrook and Russell, 2001).IC3.5K was used for constructing MAT gene expressionastorisGS115, and the transformationofDNA intoP. pas-was performed as described by the instruction manualpression kit (Invitrogen BV, Groningen, The Nether-kits for plasmid isolation, restriction enzyme digestion,n, DNA purication, and isolation of the genomic DNA5 and S. cerevisiae were purchased from TaKara Co.kyo, Japan). Shufed MAT genes were sequenced atoration (Takara, Tokyo, Japan).

condition

plate was used to screen His+ recombinant P. pas-, and the YPD-Geneticin plate was used to estimatember of MAT gene integrated into the chromosome oft P. pastoris strain as described by the manual of PichiaKit (Invitrogen BV, Groningen, The Netherlands).vation of recombinant P. pastoris strains in 250ml shakeml tubes for SAM production was similar to thosereviously (Chen et al., 2007), with some modicationstion phase. During the induction phase, the concentra-thanol were measured and 100% methanol was addedoncentration of 1.2% (v/v) every 12h, and the sterileng, China) powder was added every 24h to a nal con-of 0.1% (w/v) 12h after induction to initiate the SAM

ioreactor (FUS, Guoqiang, China) cultivation, the initiale procedure was carried out according to the methodreviously (Hu et al., 2007). After the batch phase, the(v/v) glycerol solution with 1.2% (v/v) PTM1 trace saltsvitrogen BV, Groningen, The Netherlands) was fed forwith a constant rate of 20.2 g l1 h1. Then, the straind by 100% methanol, fed together with 1.2% (v/v) PTM1olution with a constant rate of 6.8 g l1 h1. The sterileer (total 15g l1 culture broth) was supplemented for, at a concentration of 5 g l1 every 20h from 12 to 52h(Hu et al., 2007).

ivation in a 500 l bioreactor (B.Braun, Genmany) wasat in a 5 l bioreactor, but the feeding rate of 50% glycerolv/v) PTM1 trace salts solution) andmethanol (with 1.2%trace salts solution) was 23.8 and 7.5 g l1 h1, respec-

S. cerevi

S. specta

a Sequand NotI

2.3. Pl

Theed frospectab2007).

Allto contively.digestiwere land seGenetiYPD-Ggrow ohigherGS115/by BglIstrain

2.4. D

ThrpPIC3.using 5GCAAAMATgesubjecFrancePCRwaby 50cycle aed by(Takargene liand trashufecombiThe pution an

2.5. Sc

Recturedassayewere cstrainsTATGCGGCCGCTTACTTCAGACCGGCAGCATC(Reverse Primer)CCAGAATTCATGTCCAAGAGCAAAACTTTC(forward primer)TAAGCGGCCGCTTAAAATTCCAATTTCTTTG(Reverse Primer)CCAGAATTCGTGTCCCGCCGTCTCTTCACCTCG(forward primer)TATGCGGCCGCTCAGTGAACAAGCGGCAGGAG(Reverse Primer)

added to introduce restriction endonuclease recognition sites (EcoRInderlined.

d and strain construction

genes from S. cerevisiae and E.coli DH5 were ampli-eir genomic DNA, respectively. The MAT gene from S.as amplied from plasmid pWHM3-SAM (Chen et al.,ers used are listed in Table 1.geneswere inserted into the EcoRI-NotI sites of pPIC3.5Kt pPIC3.5K/EC, pPIC3.5K/SA, and pPIC3.5K/ST, respec-id constructionswere conrmedby restriction enzyme

nd DNA sequence analysis. The recombinant plasmidsized by BglII and transformed into P. pastoris GS115d on MD plates. Positive clones were cultured on YPD-late. The resultant strains which could only grow onicin plate containing 0.25mgml1 Geneticin (could notD-Geneticin plate containing 0.5mgml1 Geneticin orre selected and denoted as GS115/SA, GS115/EC, andespectively. The plasmid pPIC3.5K was also linearizedtransformed into P. pastoris GS115 to generate a controlted as GS115/3.5K).

ufing

1.4 kb DNA fragments containing the MAT genes fromC, pPIC3.5K/SA, and pPIC3.5K/ST were amplied by1 (5-GACTGGTTCCAATTGACAAGC-3) and 3AOX1 (5-CATTCTGACATCC-3) primers. Equal amounts of threereparationsweremixed, anddigestedwithDNase I, thenshufing procedure as described previously (Zhao and7)withminormodications. Theprogramofprimerlessnged as following: one cycle at 94 C for 3min, followeds at 94 C for 60 s, 30 C for 60 s, 72 C for 90 s and oneC for 7min. The primerless PCR product was ampli-g 5AOX1 and 3AOX1 primers. LA Taq DNA polymerasekyo, Japan) was used for all PCRs. The resulting MATies were inserted into the EcoRI-NotI sites of pPIC3.5K

H. Hu et al. / Journal of Biotechnology 141 (2009) 97103 99

Fig. 1. Comparison of MAT activity (a), SAM concentration (b), and DCW (c) in GS115/3.5K (1) GS115/SA (2), GS115/EC (3), and GS115/ST (4). All strains were induced withmethanol for 96h in shake ask cultivations. Data are means SE of three replicates.

2.6. Analytical methods

To determine the dry cell weight (DCW), cells were harvestedby centrifugation at 12,000 rpm for 5min, washed twice withdeionized wcentration wMAT activitShiozaki etrequired toper hour at

2.7. ConserMAT

The consDomainswrpsb.cgi).MAT wereswissmodeof MAT froor humandoi:10.2210The selectibetween thsites of thereported stJohn and Ge

3. Results

3.1. Express

Three reharboring t

elative SAM concentration in the recombinant strains harboring shufede. All strains were induced with methanol for 60h in tube cultivations.

inant strains harboring the wild-type MAT genes were used as the controlconcentration in GS115/SA was dened as 1.

ilis were cultured in the shake asks and strain GS115/3.5Ked as the control. After inducedwithmethanol for 96h, MATand SAM concentration were much higher in GS115/SA,

EC, and GS115/ST than that in GS115/3.5K. MAT activityM concentration in the best recombinant strain (GS115/SA)53.6819.87Ug1 DCW and 0.810.11g l1, respectively,were about 21 times and 134 times higher than those in3.5K, respectively. It couldbeconcluded thathighMATactiv-improved SAM production were obtained in recombinantexpressing exogenous MAT genes (Fig. 1a and b). Thereobvious difference in the cell mass among the four strains,indicated that MAT expression and SAM production did notdly affect cell growth (Fig. 1c).

Fig. 3. Compa(6). All strainsater, and dried to a constant weight at 80 C. SAM con-as assayed as described by Wanger et al. (1984) and

y was assayed according to the method reported byal. (1984). One unit (U) of enzyme was dened as thatcatalyze the transformation of 1mol l-Met into SAM37 C.

vation analysis and tertiary structure construction of

ervative sites ofMATwere analyzed by NCBI ConservedSearch (http://www.ncbi.nlm.nih.gov/Structure/cdd/The tertiary structures of the shufed and wild-typeconstructed using the Swiss-model server (http://

l.expasy.org/SWISS-MODEL.html). Reported structuresm E. coli (pdb code: 1xra) (Takusagawa et al., 1996)(pdb code: 2p02, Structural Genomics Consortium,/pdb2p02/pdb) were used as construction templates.on of template depended on the homology degreee target and the modeled MAT. The putative activeshufed and wild-type MAT were deduced from theructure of MAT from E.coli (Takusagawa et al., 1996;orge, 1999) through DNA alignment.

ion of exogenous MAT genes in P. pastoris

combinant strains (GS115/SA, GS115/EC, and GS115/ST)he wild-type MAT genes from S. cerevisiae, E.coli, and S.

Fig. 2. RMAT genRecomband SAM

spectabwas usactivityGS115/and SAwere 1whichGS115/ity andstrainswas nowhichmarkerison of MAT activity (a), SAM concentration (b), and DCW (c) in GS115/SA (1) GS115/DS16were induced with methanol for 96h in shake ask cultivations. Data are means SE of t(2), GS115/DS56 (3), GS115/DS34 (4), GS115/DS69 (5), and GS115/DS4hree replicates.

100 H. Hu et al. / Journal of Biotechnology 141 (2009) 97103

Fig. 4. The ter2403) (b), and6384) (d), pDindicated the dreferences to c

3.2. Screeni

To obtaicerevisiae, Ewere recomgeneswereformed intofor screenintion of SAMtiary structures of the shufed and wild-type MAT constructed by Swiss-model server.pDS4 (residues 2392) (c)were constructed by Swiss-model server using the structure of

S16 (residues 7384) (e), and pDS69 (residues 6384) (f) were constructed by Swiss-modeifferent structure between pDS16 and pSA. The putative active sites of MAT are shown inolor in this gure legend, the reader is referred to the web version of the article).

ng of MAT gene variants obtained by DNA shufing

n MAT with higher specic activity, MAT genes from S..coli, and S. spectabilis (5262% nucleic acid identity)bined by DNA shufing. The Libraries of shufed MATinserted into pPIC3.5K at the EcoRI-NotI sites and trans-P. pastoris GS115 to construct the recombinant strainsg. Approximately 400 cloneswere screened for produc-. The SAM concentration in GS115/SA was dened as 1,

relative SAMand 0.58, re(3.2%) prodclones (45.60.58 and 1,concentrati

AmongGS115/DS56and GS115/The tertiary structures of pST (residues 2392) (a), pDS56 (residuesMAT from E.coli as a template. The tertiary structures of pSA (residuesl server using the structure ofMAT fromhuman as a template. Arrowsgreen and the mutations are shown in red (for interpretation of the

concentration in GS115/EC and GS115/ST were 0.71spectively. Among the 412 clones screened, 13 clonesuced a relatively higher SAM concentration above 1, 188%) produced a comparable SAM concentration betweenand 211 clones (51.2%) produced a relative lower SAMon less than 0.58 (Fig. 2).the 412 clones tested, the two best (GS115/DS16 and) and three negative clones (GS115/DS4, GS115/DS34,DS69) were picked for further study. All ve strains

H. Hu et al. / Journal of Biotechnology 141 (2009) 97103 101

were cultured in shake asks. At the end of cultivation, MAT activ-ity and SAM concentration in GS115/DS56 and GS115/DS16 weresignicantly higher than in GS115/SA (Fig. 3a and b), which sug-gested that these two strains harbored the improved MAT genesthat enhanced the SAM production. MAT activity and SAM con-centration in the best strain GS115/DS56were 463.3424.68Ug1DCW and 1.640.14g l1, respectively, which were 2.01 and 1.02times higher than that of strain GS115/SA, respectively. The threenegative strains were poor in MAT expression and SAM production(Fig. 3a and b), which suggested that these three strains harboredthe negative MAT genes generated by DNA Shufing. The cell massof the two best and three negative clones were nearly the same(Fig. 3c), which agreed with the result shown in Fig. 1c.

3.3. Sequence analysis of the MAT gene variants

TheshufedMATgenes integrated in thechromosomeof theveselected strains were sequenced. The amino acid sequences of theshufed MAT (denoted as pDS16, pDS56, pDS4, pDS34, and pDS69,respectively

The amin98% and 97%respectivelyI337V) appD155G, andin the DNAplete translstop codonand thus almpDS34 wou

The amiidentity witComparedD341G) apppDS4 (K187tion (395thlead to a tru

To furtheMAT, tertiaconstructedmutations,pDS16 (Fig.f). Meanwh(Table 2).

The K18sites inpDSthat only thenzyme (Fiprobably restructure ofwas away fr

Table 2The conservative analysis of the mutations appeared in pDS4, pDS16, pDS56, andpDS69.

Proteins Mutations/conservationa

pDS16 5th/N 337th/NpDS69 23rd/N 48th/N 155th/N 278th/YpDS56 18th/Y 31st/N 65th/N 341st/N 395th404th/NpDS4 187th/N 281st/Y 398th/N

a Y represents conservative and N represents not conservative.

servative analysis ofmutations in pDS16 also indicated that the twomutationswerenonconservative (Table 2). So themechanismof theimproved activity of pDS16 was still enigmatic.

The conservative analysis of mutations (Table 2) showed thatthere was only one conservative mutation in pDS4 (S281F) andpDS69 (K278G), respectively.TheS281Fmutation inpDS4was closeto the putative active sites of the enzyme (Fig. 4c) and the K278Gmutation in pDS69 was exactly at one of the putative active sites

). Thsitesd thaut

ll awore, tmutaation

oductors

in G, the15/DShat oed mum2h oGS1

(Fig.wastionhe m5/DStion

cussi

suc/DS5ake

Fig. 5. Compa tangle), GS115/SA (triangle), GS115/EC (pentagram), and GS115/ST (circle)cultured in a 5) were deduced from their DNA sequences.o acid sequences of pDS16, pDS69 and pDS34 had 99%,identity with MAT from S. cerevisiae (denoted as pSA),

. Compared with pSA, two site mutations (K5E andeared in pDS16 and four site mutations (C23A, A48D,K278G) appeared in pDS69. A stop codon appearedsequence of pDS34, which would lead to the incom-ation of pDS34 with only 117 amino acid residues. Thisresulted in the nearly complete inactivation of pDS34ost no accumulation of SAM inGS115/DS34. Therefore,

ld not be discussed further.no acid sequences of pDS4 and pDS56 had 99% and 98%hMAT from S. spectabilis (denoted as pST), respectively.with pST, four site mutations (K18R, L31P, I65V, andeared in pDS56 and three site mutations appeared inE, S281F, andR398D). Therewas also a frame shiftmuta-404th) in the DNA sequence of pDS56, which wouldncated MAT with 404 amino acid residues.r analyze the structure-enzyme activity relationship ofry structures of the shufed and wild-type MAT wereusing Swiss-model server. The results showed that

which appeared in all the selected MAT variants except4e), did not affect the tertiary structure (Fig. 4b, c, andile, the conservative sites of MAT were also analyzed

R substitution is the only mutation in the conservative56 (Table 2). The tertiary structureof pDS56also showedeK18Rmutationwasnear theputative active sites of theg. 4b). These results suggested that the K18R mutationsulted in the increased activity of pDS56. The tertiarypDS16 was changed, but the I337V mutation in pDS16om the putative active sites of MAT (Fig. 4e). The con-

(Fig. 4factivetive. AnMAT (Mwere aTherefS281Finactiv

3.4. Prbioreac

StraFig. 5ain GS1than tincreasmaximafter 6that ofasksstrainsproduc500 l. Tin GS11applica

4. Dis

We(GS115both sh

rison of MAT activity (a), SAM concentration (b), and DCW (c) in GS115/DS56 (recl bioreactor. Arrows indicated the time for l-Met addition.e C23A mutation in pDS69 was also near the putativeof the enzyme (Fig. 4f), but this site was not conserva-e same C23A mutation also happened in the wild-typeino et al., 1996). Other mutations in pDS4 and pDS69ay from the putative active sites of MAT (Fig. 4c and f).he decreased activity of pDS4 was probably due to thetion and the K278G mutation was likely to result in theof pDS4.

tion of SAM by strain GS115/DS56 in 5 l and 500 l

S115/DS56 was cultured in a 5 l bioreactor. As shown inmaximal MAT activity of 587.2Ug1 DCWwas achieved56 after 42h of induction, which was 2.8 times higherf GS115/SA. SAM concentration in GS115/DS56 alsoarkedly due to the improvement of MAT activity. TheSAM concentration in GS115/DS56 reached 5.59g l1

f methanol induction, which was about 28% higher than15/SA (Fig. 5b). Similar with the cultivation in shake1c and Fig. 3c), the cell mass of the four recombinantnot changed with the varied MAT activity and SAM(Fig. 5c). Finally, the cultivation scale was scaled up toaximum SAM concentration of 6.14g l1 was obtained56, which showed a favorable foreground for industrial.

ons

ceeded in constructing a new recombinant strain6) with increased MAT activity and SAM production inasks and 500 l bioreactor by expressing an improved

102 H. Hu et al. / Journal of Biotechnology 141 (2009) 97103

MAT gene. The increased MAT activity played a key role in the highSAM production.

There are two ways to increase the heterogenous enzyme activ-ity in the recombinant strains: one is to increase the expressionlevel and thHowever, inwill consumgrowth. It wcoli JM101 h(Bruno et amonooxyge1992). In th(MAT), butsubstrate inthe better wto improveactivityofMactivity. Firsbinant straiall the recoplate and scontainingGeneticin pselected. Thgene should(Scorer et aseveral amisequence oMAT transcvaried MATto the differMAT activitspecic actobtained by

No crossbut we foungenes. It mcompletelybetween thuncompletethe three hobases long.tity, a veryPCR shouldperature instudies.

In ourincreased Mduction. Thin GS115/Dconcentratiing the SAMonly increaalso happeGS115/SA, timprovemeSAM concenTherefore, MproductionSAM produ

Itwasdeand glyceroand SAM pring the MATdecreased (hemoglobin

strain harboring the MAT gene from S. spectabilis to improve ATPsynthesis and thus enhanced SAM production (Chen et al., 2007).Thus, it could be concluded that the ATP level was another limitingfactor for SAM production. The l-Met feeding strategy also affected

M pron strst str

biny sho

wled

arem Nch a7AAai M

searc

nces

., Parkatalytron. Mo, J.Lphic y., App

yces ceX., Chular exuctio1212, A., Ranes fr291.S., Andonsanmostalle, Oerichiagrowt

, Chu, Jvel feic synme M., Geohetaseu, J., T

ia pastZhu,densiechno., Sun,nohyBao 22000.., Paj

hionin712A,M.R.enosyovere, S.,osylm

1595a, K.Pchedrobiol., Tai, Jethion.Kang,xidasechnok, J., RSprin.A., Cld seleigh-lee other is to improve the specic activity of the enzyme.creasing the expression level of heterogenous enzymee more resource and energy, and even inhibit the cellas reported that full induction of styAB expression in E.ad led to a strong inhibition of cell growth andwashoutl., 2008). The similar effect was also observed in alkanenase-overproducing, E. coli (Favre-Bulle and Witholt,is study, the desired product is not the enzyme itselfthe MAT-catalyzed product SAM. And ATP is anotherthis MAT-catalyzed SAM biosynthesis reaction. Thus,ay to increase MAT activity and SAM accumulation is

the specic activity of MAT. In this study, the enhancedATvariant shouldbeattributed to the improvedspecictly, theMAT gene copy should be the same in all recom-ns harboring the shufed orwild-typeMAT genes. Sincembinant strains were cultured on the YPD-Geneticintrains which could only grow on YPD-Geneticin plate0.25mgml1 Geneticin but could not grow on YPD-late containing 0.5mgml1 Geneticin or higher wereis selection ensures that only one copy of the MATbe integrated into the chromosome of the P. pastoris

l., 1994). Secondly, the similar expression cassette (onlyno acid residue substitutions happened in the codingf all the MAT gene variants) would lead to the similarription level in each recombinant strain. Therefore, theactivities in the recombinant strains were mainly dueent coding sequences of the MAT variants, and the highy in GS115/DS56 should be attributed to the improvedivity of pDS56 encoded by the recombinant MAT geneDNA shufing.overs happened in the ve selected shufedMAT genes,d a crossover appeared in another two sequenced MATeant that the three homologous MAT genes were notshufed.We supposed that the lownucleic acid identitye three homologousMAT genes (5262%) resulted in thed shufing because the conserved nucleic acids amongmologous MAT gene sequences were no more than 10To force crossover based on such short sequence iden-low effective annealing temperature in the primerlessbe used (Stemmer, 1994). Thus, the annealing tem-the primerless PCR could be optimized in the further

best recombinant strain GS115/DS56, the degree ofAT activity was much higher than that of SAM pro-e MAT activity was 83% higher in GS115/DS56 thanS16 in the shake ask cultivations, while the SAMon and specic SAM concentration (calculated by divid-

concentration (g l1) by the dry cell weight (g l1))sed about 22% and 24%, respectively. This phenomenonned in the 5 l bioreactor cultivations. Compared withhemaximumMAT activity in GS115/DS56 had 2.8 timesnt, but the maximum SAM concentration and specictration only increased about 28% and 13%, respectively.AT activity was not the only limiting factor for SAM

and too high MAT activity would not further increasection signicantly.monstrated that alternateormixed feedingofmethanoll during the induction phase improved ATP synthesisoduction in the recombinant P. pastoris strain harbor-gene from S. cerevisiae although the MAT activity was

Li et al., 2002; Hu et al., 2007). Furthermore, Virtreoscillagene was transformed into the recombinant P. pastoris

the SAadditiothe bethe comstrateg

Ackno

WePrograResearNo.200Shanghthis re

Refere

Bruno, BbiocEnvi

Cereghinlotro

Chan, S.Yrom

Chen,H.celluprod1205

Crameriof ge288

Emond,H.,Mther

Favre-BuEschcell937.

Hu,X.Q.A nomatEnzy

John, C.Tsynt

Li, D.Y., YPich

Liu, P.Y.,cell-Biot

Liu, Z.QlactoXue

Lu, S.C.,Mato, J.M

met0647

MautinoS-advivo

Megumiaden(6),

MinchevenriMik

Park, J.Hl-m2185

Ryu, K.,peroBiot

SambrooCold

Scorer, CRapifor hduction. Liu et al. (2006) compared ve different l-Metategies and found that feeding l-Met continuously wasategy for SAM production. Therefore, further studies oned effects of MAT activity, ATP level, and l-Met feedinguld be carried out to optimize the SAM production.

gements

grateful to National Basic Research Program (973o. 2007CB714306), the National High Technologynd Development Program of China (863 Program100601) and Science and Technology Commission ofunicipality (07pj14027) for their nancial supports toh.

, J.B., Lars,M.B., Schmid, A., 2008. NADH availability limits asymmetricic epoxidation in a growing recombinant Escherichia coli strain. Appl.icrobiol. 74 (5), 14361446.., Cregg, J.M., 2000. Heterologous protein expression in the methy-east Pichia pastoris. FEMS Microbiol. Rev. 24, 4566.ling, D.R., 2003. Regulation of S-adenosylmethionine levels in Saccha-revisiae. J. Biol. Chem. 278, 4305143059., J., Zhang, S.L., Zhuang, Y.P., Qian, J.C.,Wang, Y.H., Hu, X.Q., 2007. Intra-pression of Vitreoscilla hemoglobin improves S-adenosylmethioninen in a recombinant Pichia pastoris. Appl. Microbiol. Biotechnol. 74 (6),.illard, S.A., Bermudez, E., Stemmer,W.P., 1998.DNAshuingof a familyom diverse species accelerates directed evolution. Nature 391 (6664),

re, I., Jaziri, K., Potocki-Veronese, G., Mondon, P., Bouayadi, K., Kharrat,, P., Remaud-Simeon,M., 2008. Combinatorial engineering to enhancebility of amylosucrase. Protein Sci. 17, 967976.., Witholt, B., 1992. Biooxidation of n-octane by a recombinantcoli in a two-liquid-phase system: effect of medium components on

h and alkane oxidation activity. Enzyme Microb. Technol. 14, 931

., Zhang, S.L., Zhuang, Y.P.,Wang, Y.H., Zhu, S., Zhu, Z.G., Yuan, Z.Y., 2007.eding strategy during the production phase for enhancing the enzy-thesis of S-adenosyl-l-methionine by methylotrophic Pichia pastoris.icrob. Technol. 40, 669674.rge, D.M., 1999. The bifunctional active site of S-adenosylmethionine. J. Biol. Chem. 274 (46), 3290932914.ian, L.W., Ji, X.S., Yuan, Z.Y., 2002. Production of SAM by recombinantoris. Chin. J. Biotechnol. 3, 295299.D.H., Tan, T.W., 2006. Effect of feeding pre-l-methionine on high-ty fermentation for S-adenosyl-l-methionine production. Chin. J.l. 22 (2), 268272.Z.H., Zheng, P., Leng, Y.J., Qian, S., 2005. Directed evolution of d-

drolase by error prone PCR and DNA shufing. Sheng Wu Gong Cheng1 (5), 773781.S-adenosylmethionine. Int. J. Biochem. Cell Biol. 32, 391395.ares, M.A., Mingorance, J., Avarez, L., 1995. Production of S-adenosyl-e (SAM) by fermentation of transformed bacteria. European patent1., Barra, J.L., Rosa, A.L., 1996. eth-1, theneurospora crassa locus encodinglmethionine synthetase: molecular cloning, sequence analysis and inxpression. Genetics 142 (3), 789800.Tsutomu, F., Haruyuki, I., 2007. Effects of accumulated S-ethionine on growth of yeast cells. Biosci. Biotechnol. Biochem. 711597.., Balutsov, V.M., 2002. Isolation of S-adenosyl-l-methionine from anKluyveromyces lactis mutant grown in whey media. Prikl. Biokhim.. 38 (4), 389392..Z., Roessner, C.A., Sott, A.T., 1996. Enzymatic synthesis of S-adenosyl-ine on the preparative scale. Bioorg. Med. Chem. 4 (12), 2179

J.H., Wang, L., LeeJ, E.K., 2008. Expression in yeast of secreted lignine with improved 2 4-dichlorophenol degradability by DNA shufing.logy 135 (3), 241246.ussell, D.W., 2001. Molecular Cloning:A Laboratory Manual, third ed.g Harbor Laboratory Press, New York, NY.are, J.J., McCombie, W.R., Michael, A., Romanos, Sreekrishna, K., 1994.ction using G418 of high copy number transformants of Pichia pastorisvel foreign gene expression. Biotechnology 12 (2), 181184.

H. Hu et al. / Journal of Biotechnology 141 (2009) 97103 103

Shiozaki, S., Shimizu, S., Yamaha, H., 1984. Unusual intracellular accumulation ofS-adenosyl-l-methionine bymicroorganisms. Arric. Biol. Chem. 48, 22932300.

Shiozaki, S., Shimizu, S., Yamada, H., 1986. Production of S-adenosyl-l-methionineby Saccharomces sake. J. Biotechnol. 4 (6), 345354.

Shobayashi, M., Mukai, N., Iwashita, K., Hiraga, Y., lefuji, H., 2005. A new method forisolation of S-adenosylmethionine (SAM)-accumulating yeast. Appl. Microbiol.69, 704710.

Stemmer, W.P.C., 1994. DNA shufing by random fragmentation and reassem-bly: in vitro recombination for molecular evolution. Proc. Natl. Acad. Sci. 91,1074710751.

Suen, W.C., Zhang, N., Xiao, L., Madison, V., Zaks, A., 2004. Improved activity andthermo-stability of Candida antarctica lipase B by DNA family shufing. Protein.Eng. Des. Sel. 7 (2), 133140.

Takusagawa, F., Kamitori, S., Misaki, S., George, D.M., 1996. Crystal structure of S-adenosylmethionine synthetase. J. Biol. Chem. 271 (1), 136147.

Tatsuya, T., Yoshinori, H., 2007. Development of innovative Pediocin PA-1 by DNAShufing among Class IIa Bacteriocins. Appl. Environ. Microbiol. 73, 52925299.

Wanger, J., Danzin, C., Huot-Olivier, S., Claveric, N., Palfreyman, M.G., 1984. Highperformance liquid chromatographic analysis of S-adenisylmethionine and itsmetabolites in rat tissues: interrelationship with changes in biogenic catechollevels following treatment with l-dopa. J. Chromatogr. 290, 247262.

Yang, J., Wang, M., Sun, J., Wei, P.H., Tang, W.G., 2002. Cloning of S-adenosyl-l-methionine synthetase gene from Saccharomyces ceverisiae and its expressionin E. coli. J. China Pharm. Univ. 3, 253255.

Yu, Z.L., Wu, X.J., Li, D.Y., Yang, S., Zhou, Z., Cai, J.Z.Y., 2003. Enhancement of theproduction of SAMby overexpression of SAMsynthetase in Pichia pastoris. ShengWu Wu Li Xue Bao, Shanghai 35, 127132.

Zhao,H.M., Frances,H.A., 1997.OptimizationofDNAshufing forhighdelity recom-bination. Nucleic Acids Res. 25 (6), 13071308.

DNA shuffling of methionine adenosyltransferase gene leads to improved S-adenosyl-l-methionine production in Pichia pastorisIntroductionMaterials and methodsGeneral methodsCulture conditionPlasmid and strain constructionDNA shufflingScreening proceduresAnalytical methodsConservation analysis and tertiary structure construction of MAT

ResultsExpression of exogenous MAT genes in P. pastorisScreening of MAT gene variants obtained by DNA shufflingSequence analysis of the MAT gene variantsProduction of SAM by strain GS115/DS56 in 5l and 500l bioreactors

DiscussionsAcknowledgementsReferences