Ácido hialuronico, huimin yu et al 2008

8/4/2019 Ácido Hialuronico, Huimin Yu et al 2008

http://slidepdf.com/reader/full/acido-hialuronico-huimin-yu-et-al-2008 1/9

A RTICLE

A High-Throughput Screen for Hyaluronic AcidAccumulation in Recombinant Escherichia coli

Transformed by Libraries of EngineeredSigma FactorsHuimin Yu, 1,2 y Keith Tyo,1 Hal Alper,1 Daniel Klein-Marcuschamer, 1

Gregory Stephanopoulos 1

1Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge,Massachusetts 02139; telephone: 617-253-4583; fax: 617-253-3122; e-mail: [email protected] 2Department of Chemical Engineering, Tsinghua University, Beijing, ChinaReceived 16 January 2008; revision received 1 April 2008; accepted 14 April 2008

Published online 24 April 2008 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/bit.21947

ABSTRACT:Hyaluronic acid (HA) is an important bio-material with functional medical and cosmetic applications.As its synthesis has been recently reported in recombinantbacteria, it is of interest to develop a high throughputscreening method for the rapid isolation of HA accumulat-ing strains transformed by combinatorial libraries. Here wereport a novel two-step screening strategy to select for betterHA-producing recombinant Escherichia coli strains trans-formed by mutation libraries of rpoD and rpoS, coding forthe s

D and s S factors of the RNA polymerase, respectively.

The rst screen, based on translucent colony morphology identication, was used to qualitatively distinguish HA-producing strains on agar plates from non-HA producingstrains that exhibit dense colony morphology. The secondscreen was based on the photometric measurement of analcian blue staining solution that precipitates with HA,creating an inverse relationship between HA concentrationand alcian blue absorbance. The color attenuation tteda second-order polynomial between HA concentrationand OD 540 absorbance. Using the alcian blue absorbancequantication, 74 translucent colonies from the HA-rpoDlibrary and 78 translucent colonies from the HA-rpoS library were isolated and cultured foroptimal strain selection. Threerepresentative superior recombinants with high, mediumand low increase of HA accumulation, respectively, wereidentied by the screen from the HA-rpoD and HA-rpoSmutant library. Further ask culture conrmed that resultsof the library screen were reliable and the superior recom-binant D72 highly accumulated HA of 561.4 mg/L with aproductivity of $ 265 mg HA/g dry cell. Sequencing resultsshowed that the mutant rpoD gene in D72 is in a truncatedprotein that lacks the conserved regions 3 and 4 of the s

D .Generally, this two-step high throughput screen presents a

promising strategy for selecting superior HA-producingstrains from large scale mutation libraries.Biotechnol. Bioeng. 2008;101: 788–796.ß 2008 Wiley Periodicals, Inc.KEYWORDS:Hyaluronic acid; high throughput screening;translucent colony identication; alcian blue staining;global transcriptional machinery engineering; sigma factorengineering

IntroductionGlobal transcriptional machinery engineering (gTME)allows modication of the transcriptome by changing thebinding specicity of RNA polymerase to different pro-moters through amino acid mutations in binding factors,thereby having the potential to impact complex cellularphenotypes. It has been successfully applied for theimprovement of ethanol tolerance and productivity inSaccharomyces cerevisiae (Alper et al., 2006) and morerecently in Escherichia coli (Alper and Stephanopoulos,2007). As such, it is a promising approach for improvingthe industrial production of different target products by

engineered microbes. As is the case with all large-scalecombinatorial library approaches, gTME is best applied witha reliable high throughput screening platform for selectingmutants with improved cellular phenotypes.

Hyaluronic acid (Hyaluronan, HA) is a valuable functionalbiopolymer. Its importance stems from its structural, rheolo-gical, physiological, and biological properties, leading to a widerange of applications in the health, cosmetic and clinical elds(Goa and Beneld, 1994; Lauren, 1998). Microbial productionof HA using the group C or group A Streptococcus has beenpursued as alternative to chemical extraction from chicken

yThe author was a visiting scholar when she completed this research.Correspondence to: G. StephanopoulosContract grant sponsor: Singapore-MIT AllianceContract grant number: SMA-2Contract grant sponsor: National Key Basic Research Project 973 of ChinaContract grant number: 2007CB714304

788 Biotechnology and Bioengineering, Vol. 101, No. 4, November 1, 2008 ß 2008 Wiley Periodicals, Inc.



comb (DeAngelis et al., 1993; Kim et al., 1996; Kumari andWeigel, 1997; Kakizaki et al., 2002; Ogrodowski et al., 2005).Recently, new methods using novel recombinant productionstrains that utilize inexpensive media and avoid pathogenicity have been developed as substitute of Streptococcusfermentation.The construction of a recombinant gram-positive Bacillussubtilis has been reported (Widner et al., 2005), and morerecently, our group has developed a recombinant E. coli, Top10/pMBAD-sseABC , to produce HA (Yu and Stephanopoulos,

2008).The above methods applied rational metabolic engineering

in constructing the HA producing strains (Stephanopoulosand Kelleher, 2001; Stephanopoulos and Sinskey, 1993),addressing issues of precursor supply and pathway kineticsand regulation (Stephanopoulos and Simpson, 1997; Val-lino and Stephanopoulos, 1994). While such methods have yielded nontrivial improvements in many instances, espe-cially when combined with bioreactor operation andoptimization strategies (Follstad et al., 1999; Kiss and Steph-anopoulos, 1991; San and Stephanopoulos, 1984; Stephano-poulos et al., 1979), they cannot exploit potentially bene cialeffects of distal genes or complex cellular regulatory networks.Combinatorial approaches have been developed for thispurpose giving rise to the concept of inverse metabolicengineering (Jin et al., 2005; Jin and Stephanopoulos, 2007).The goal of such methods is to elicit global changes in the cellthat favor conditions of increased production, and/or makechanges to cell physiology that improve environmentaltolerance to overcome limited productivity.

A key element to the implementation of such strategies isthe availability of high throughput screens for isolating cellscapable of high product accumulation. In this work, wepresent such screens for the case of HA production in E. coli.

Alcian blue is a water soluble copper-phthalocyaninedye, C56 H68 Cl4 CuN 16 S4 , which can be used for the staining of sulfated and carboxylated acid mucopolysaccharides (Penney et al., 2002). It is believed to form salt linkages with the acidgroups of mucopolysaccharides due to the presence of copperin the molecule, which decreases the blue color. Since HA is amucopolysaccharide, it is feasible to construct an alcian bluestaining method for quantifying the HA concentration.

The above screen was tested with a culture of E. coli cellstransformed with mutation libraries of the subunits of thetranscription machinery. A two-step high throughput screen-ing platform that consisted of translucent colony identi cationin combination with the alcian blue staining was used for the

selection of high HA production variants from the libraries.This allowed the identi cation of mutant transcription factorsthat enhanced the speci c HA productivity by as much as 30%over the originally engineered parental strain.

Materials and Methods

DNA Manipulations, Plasmids, and Bacterial Strains

All DNA manipulations, such as genomic DNA isolation,restriction enzyme digestion and ligation, were performed

by standard procedures (Sambrook et al., 1988) or followingthe specic manufacturer ’s instructions. Restrictionenzymes were purchased from New England Biolabs, TaqDNA polymerase and primers were ordered from Invitrogen(Grand Island, NY). Plasmid pMBAD (Yu and Stephano-poulos, 2008) was constructed by the introduction of a 62 bpmulti-cloning sites (MCS) sequence containing XbaI-BamHI-StuI-KpnI-SacI-EcoRI-HindIII restriction sites intothe plasmid of pBAD (Invitrogen) with an ampicillin

resistance marker. E. coliTop10 (Invitrogen) was used as theexpression host of the plasmid pMBAD- sseABC , which wasconstructed by the insertion of the fragment sseABC into thebackbone of pMBAD (Yu and Stephanopoulos, 2008). ThesseABC operon consists of the genes sehasA, hasB, and hasC .The sehasA was synthesized by assembly PCR (Hoover andLubkowski, 2002) according to the protein sequence of the HA synthase from Streptococcus equisimilis (NCBI-AAB87874.1, GI:2655100). The functionality of hasB andhasC were provided by the genes ugd and galF of E. coliK12MG1655, coding for the UDP-glucose 6-dehygrogenaseand the glucose-1-P uridyltransferase, respectively. E. coliTop10/pMBAD- sseABC is an L-arabinose inducible recom-binant strain for HA production (Yu and Stephanopoulos,2008), while E. coli Top10/pMBAD was used as the nullcontrol. E. coli DH5 a (Invitrogen) was used for routinetransformations as described in the protocol.

Library Construction

A low copy host plasmid (pHACM) was constructed aspreviously described (Alper and Stephanopoulos, 2007).The genes encoding the s

D subunit and the s S subunit of

RNA polymerase, denoted as rpoD, and rpoS, respectively,were amplied from E. coli genomic DNA, using thefollowing primers: rpoD-F-SacI: AACCTAGGAGCTCTGA-TTTAACGGCTTAAGTGCCGAAGAGC and rpoD-R-Hin-dIII: TGGAAGCTTTAACGCCTGATCCGGCCTACCGAT-TAAT, and rpoS-F-SacI: AACCTAGGAGCTCAGACTGG-CCTTTCTGACAGATGCTTACT and rpoS-R-HindIII: AA-CCTAGGAGCTCAGACTGGCCTTTCTGACAGATGCTT-ACT. Fragment mutagenesis was performed using theGenemorph II Random Mutagenesis kit (Stratagene, LaJolla, CA) with various concentrations of initial template toobtain low, medium, and high mutation rates as described inthe product protocol as well as previously described (Alper

and Stephanopoulos, 2007). Following the error-pronePCR, the mutated fragments of rpoD and rpoS were puri edusing a Qiagen PCR cleanup kit, digested by the respectiverestriction enzymes overnight (HindIII/SacI for rpoD,HindIII/SacI for rpoS), ligated overnight into a digestedpHACM backbone, and nally transformed into E. coliDH5 a competent cells. Cells were plated on LB-agar platesand scraped off to create a liquid library. The total library size was approximately 106 . The plasmid library wasextracted using the Qiagen Miniprep kit (Qiagen, Valencia,CA) and stored at À 808 C. An approximately equal

Yu et al.: Sigma Factor Library Screen for HA Production 789Biotechnology and Bioengineering



concentration of the plasmid library of pHACM- rpoD M andpHACM- rpoS M was transformed into E. coli Top10/pMBAD-sseABC by electroporation and plated on selectiveplates after dilution. The HA-producing libraries of Top10/(pMBAD- sseABC , pHACM- rpoD M ) and Top10/(pMBAD-sseABC , pHACM- rpoS M ) were abbreviated as HA-rpoD andHA-rpoS libraries, respectively. Strain Top10/pMBAD-sseABC with single plasmid, and strains with doubleplasmids: Top10/ (pMBAD- sseABC , pHACM-blank),

Top10/(pMBAD- sseABC , pHACM- rpoD), and Top10/(pMBAD- sseABC , pHACM- rpoS) which containing theempty pHACM plasmid or the wild type rpoD and rpoSgene, respectively, were used as controls or references formutant evaluation.

Translucent Colony Screening Optimization

Different growth media were tested for optimizing thetranslucent colony phenotype of HA-producing strains. Allmedia were prepared using the following concentrations of supplements, as speci cally mentioned in each medium,

MgSO4 Á7H 2 O, 0.25 g/L; MgCl2 , 0.95 g/L; sorbitol, 15 g/L;leucine, 0.2 g/L; L-arabinose (inducer), 0.1 g/L; ampicillin,100 mg/L; chloramphenicol 34 mg/L. Six different modi edmedia were used for optimizing the translucent colony screening, including M9 M (M9 supplemented with 10 g/Lglucose, MgSO4 Á7H 2 O, leucine, ampicillin, and L-ara-binose), R M (R medium (Wang and Lee, 1998) supple-mented with leucine, ampicillin and L-arabinose), MOPS M

(MOPS medium (Teknova, Hollister, CA, Neidhardtet al., 1974) supplemented with leucine, ampicillin, andL-arabinose), MM1 M (MM1 medium (Bellemann et al.,1994) supplemented with MgSO 4 Á7H 2 O, leucine, ampicil-

lin, and L-arabinose), LBMA (LB medium supplementedwith 15 g/L glucose, MgCl2 , ampicillin, and L-arabinose)and LBSMA (Bellemann et al., 1994) medium (LB Mediumsupplemented with sorbitol, MgCl 2 , ampicillin andL-arabinose) were used for the medium optimization of the translucent colony screen.

High Throughput Quantication of HAby Alcian Blue Staining

The alcian blue solution was prepared by the followingprocedure: 1.0 g alcian blue 8GX (Sigma–Aldrich, St. Louis,

MO) was dissolved in 100 mL 3% glacial acetic acid andthe pH was adjusted to 2.5 using acetic acid (WebPath:Internet Pathology Laboratory http://library.med.utah.edu/WebPath/webpath.html). The solution was ltered througha 0.45 m m syringe lter (VWR, Edison, NJ), and a crystal of thymol was added. It was stored at room temperature andfound to be stable for 6 months. The optimized procedurefor high throughput HA quanti cation is as follows: 400 m Lof fermentation broth containing HA was aliquoted into a1.5 mL centrifuge tube pre- lled with 550 m L 3% acetic acid.Fifty microliters Alcian blue solution was added followed by

vortexing, and the mixture was microwaved for 30 s; aftercentrifugation, the tube was cooled at room temperature for2.5 h. Then, the solution was centrifuged at 10,000 rpm for1 min, and 200 m L of supernatant were loaded into a 96-wellplate, and the OD 540 was measured using the plate reader.A standard curve was generated using 400 m L of 50, 100,200, 300, and 500 mg/L commercial HA standards (VWR).All experiments were repeated three times except wherespecically noted.

Phenotype Selection, Media, and Culture Conditions

LBSMAC (LBSMA supplemented with chloramphenicol)solid medium was used for the translucent colony screeningof HA-producing libraries. Selected translucent colonieswere transferred to 2 mL LBAC medium and culturedovernight in 30 Â 115 mm closed top centrifuge tubesshaking at 37 8 C. Inoculums (2%, V/V) of the stationary phase culture were used to culture the selected clone inanother tube with 1 or 2 mL LBM AC medium (LB mediumsupplemented with MgCl 2 , ampicillin and chlorampheni-col). These cultures were incubated at 37 8 C for 2.5 h(OD 600 $ 0.8), and induced with L-arabinose. After 5 h, thecultures were supplemented with 10 g/L glucose to allow accumulation of HA. Cultures were stopped at 24 h, andHA concentration was quantitatively measured by thealcian blue method in a 96-well plate measuring OD 540 by a Packard Fusion 96-well plate reader. For one batch of screening, usually 38 transparent library colonies weresimultaneously quanti ed with 2 dense colonies of theparental Top10/pMBAD- sseABC as controls.

The optimal selected HA-rpoD and HA-rpoS library strains were plated, inoculated, and cultured in 40 mLLBMAC /300 mL asks at 378 C with 200 rpm orbital shaking

for further HA productivity testing. Cell density wasmonitored spectrophotometrically at 600 nm by an Amer-sham Biosciences Ultraspec 2100 Pro. The inducer, 0.1 g/LL-arabinose, was added at around 2.5 h when OD 600 reached0.8. Glucose (10 g/L) was later supplemented at 5 h.Additional glucose (6 g/L) was supplemented at 24 h, andmeanwhile the pH of the broth was adjusted to 7.0 –7.5 usingNaOH (4 mol/L). Broth was harvested at 48 h to assay theHA titer by a modi ed turbidimetric method usingcetyltrimethylammonium bromide (CTAB) (Chong andNielsen, 2003; Ferrante, 1956) with both Top10/(pMBAD-sseABC , pHACM-blank) and Top10/pMBAD- sseABC ascontrols. Meanwhile, strain D0, Top10/(pMBAD- sseABC ,pHACM- rpoD) containing an extra wild-type rpoD gene,and strain S0, Top10/(pMBAD- sseABC , pHACM- rpoS)containing an extra wild-type rpoS gene, were alsoconstructed and simultaneously cultured for parallelcomparison with the mutants.

HA Titer Measurement by a Modied CTAB Method

HA titers were measured by a modi ed turbidimetricmethod (Chong and Nielsen, 2003; Ferrante, 1956). Acetate

790 Biotechnology and Bioengineering, Vol. 101, No. 4, November 1, 2008



buffer was prepared as 0.2 M sodium acetate and 0.15 MNaCl in water and adjusted pH 6.0 using glacial acetic acid,and CTAB reagent was prepared as 2.5 g/L CTAB in 0.5 M(20 g/L) NaOH solution. Fermentation broth samples wereincubated rst with an equal volume of 0.1% w/v sodium-dodecyl-sulfate (SDS) at room temperature for 10 minto free the capsular HA (Chong and Nielsen, 2003).Subsequently, the HA product was precipitated out from themedium samples with 2 volumes of ethanol (Ogrodowski

et al., 2005) incubating at 48

C for 1 h. The precipitate wascollected by centrifugation (2,000 g for 20 min at roomtemperature) and resuspended in 1 volume of deioned waterfor 10 min. Then the re-dissolved samples were applied tothe modi ed CTAB assay as follows: add 0.15 mL sampleinto 0.35 mL acetate buffer, mix then add 1.0 mL CTABreagent; mix and incubate at room temperature for 5 min;measure OD 400 and calculate the HA titer by 20 Â the valuecalculated from the standard curve, which was prepared by assaying the correlation of the turbidity OD 400 to differentconcentration of HA diluted by a concentrated 0.5 g/Lstandard with MW of 6.8 Â 105 (Lifecore Biomedical, Inc.,Chaska, MN).

Results and Discussion

Translucent Colony Formation and Identi cation ofHA-Producing Recombinant E. coli Cells

In light of the conventional method typically used toidentify high-HA-producing strains of Streptococcus spp.and B. subtilis by viscous colony morphology on solidmedium (Kim et al., 1996; Widner et al., 2005), screeningbased on colony-morphology was also employed foridentifying HA-producing cells in recombinant E. coli. Aslisted in Materials and Methods Section, six modi ed mediadenoted as M9 M , R M , MOPSM , MM1M , LBMA, and LBSMA,were tested for mucoid colony morphology due to thesecretion of HA. Results showed that both Top10/pMBAD-sseABC , the HA-producing strain, and Top10/pMBAD, thenon-HA producing strain, could not grow well on M9 M , R M ,or MOPS M solid media. Colonies of Top10/pMBAD- sseABC appeared on MM1 M plates after 3 days incubation at 37 8 C,but did not show any distinct morphological traits com-pared to Top10/pMBAD. Similar results were observed forLBMA solid medium. However, for the sorbitol-containing

medium of LBSMA, the translucent colonies of Top10/pMBAD-sseABC were apparently different from the densecolony morphology of Top10/pMBAD. As shown inFigure 1, overnight cultures of Top10/pMBAD- sseABC and Top10/pMBAD plated simultaneously on LBSMAformed notably different colonies with translucent ordense morphology, respectively. The observed differencebetween the two types of strains can be used for qualitativeidenti cation of HA-producers in recombinant E. coli.Further studies showed that translucent morphology couldbe observed with strains producing as little as $ 50 mg/L HA.

High Throughput Quantication of HA-ProducingStrains Using Alcian Blue Staining

While colony morphology is adequate to discern largedifferences in HA productivity, a more quantitative ap-proach was necessary to screen for incremental improve-ments in HA accumulation. Therefore, a novel screeningmethod that is scalable in throughput and signi cantly morequantitative in predicting HA titer was developed anddesignated alcian blue staining.

Absorbance scan of a pure alcian blue solution from200 nm to 800 nm yields two positive absorbance peaks at334 and 605 nm (Fig. 2a). However, after adding 10, 50, or100 m L alcian blue solution into 1.0 mL of 200 mg/LHA and using the corresponding alcian blue solutionwithout HA as blank control, the absorbance patternwas signicantly changed showing decreased absorbance atdifferent wavelengths (correspondingly Fig. 2b –d), suchas 380, 560 and 700 nm for the 50 m L alcian blue mix.Furthermore, a visible precipitation of alcian blue was

observed. By comparing the absorbance intensities of each peak, it was found that the 50 m L alcian blue stainingsystem showed the strongest negative absorbance relative tothe solution without HA (Fig. 3), and it was thus selectedas the optimal concentration of dye for subsequentexperiments.

Subsequently, a calibration curve was constructed by measuring absorbance at 540 nm of a mixture of 50 m L/mLalcian blue staining solution and six different HA con-centrations: 25, 50, 100, 200, 300 and 500 mg/L HA. As canbe seen in Figure 4, a second-order polynomial formula ts

Figure 1. Translucent colony morphology of HA-producing recombinant E. coli and dense colony morphology of non-HA producing recombinant E. coli . Translucentcolonies are marked by dashed arrows, and dense colonies are marked by solidarrows. A reversing treatment in Photoshop 7.0 was performed to obtain a bettercontrast between the dense and translucent morphologies.




the HA-OD 540 response curve in the range of 50 –500 mg/LHA concentration, and a linear t can be observed from100–500 mg/L HA in alcian blue if incubation is increasedbeyond 1 h. By setting the HA-alcian blue binding time at

2.5 h and tting the OD 540 -HA using a second-orderpolynomial formula (Fig. 5), a correlation of R2 $ 0.99945was obtained indicating that the alcian blue staining methodcan quantitatively predict HA concentrations.

Figure 2. Absorbance spectra of pure alcian blue solution and alcian blue mixed with HA. a: scanned spectrum of 10 m L alcian blue solution in 990 m L 3% acetic acid buffer,using the buffer as blank control; ( b, c, and d), negative absorbance of 10, 50, and 100 m L, respectively, alcian blue and HA solution in a total volume of 1 mL using the correspondingalcian blue solution without HAas blank control. Thescanned samples were prepared as follows: 10, 50, and 100 m L alcian blue solution were mixed with 500 m L of 400 mg/L HA and3%acetic acid bufferwas addedto 1 mL,the mixturewas microwaved 30s, cooledfor1 h atroomtemperatureandcentrifuged1 min at10,000 rpm. Thesupernatant was loadedinto the UV-cuvette, and the spectrum scanned from 200–800 nm. Optimal absorbance peaks: (a), 334 and 605 nm, positive; (b), 340 and 620 nm, negative; (c), 380, 560, and 700 nm,negative; (d), 400, 540, and 730 nm, negative. Experiments repeated three times.

Figure 3. Intensity comparison of the HA-stained alcian blue solution.

Figure 4. HA quantication by alcian blue staining. Different standing time forHA and alcian blue binding at room temperature was evaluated (30 min, 1, 2, 5, and5.0h) withinthe HAconcentrationrangeof 0–500 mg/L. [Colorgure canbe seen intheonline version of this article, available at www.interscience.wiley.com.]




High Throughput Screening of HA-rpoD andHA-rpoS Libraries

The above screen was applied to the identi cation of sigmafactor mutants eliciting increased HA production in thepreviously engineered E. coli. Libraries of the s

D factor(Alper and Stephanopoulos, 2007) which controls theexpression of around 1,000 genes responsible for normalexponential growth (Gregory et al., 2005; Helmann andChamberlin, 1988), and the s

S factor that orchestratesthe stationary phase phenotype in response to cessationof growth caused by various stresses (Venturi, 2003) werescreened. By random mutagenesis of the genes rpoD(coding for s

D ) and rpoS (coding for s S), two libraries

were constructed, transformed into the parental HA-producing strain and screened for HA production asdescribed in the Materials and MethodsSection.Each library has three levels of mutation frequency, denoted as high (H),moderate (M) and low (S).

Using the rst identi cation step (translucency), 74 rpoDmutants and 78 rpoS mutants were selected from $ 1,000 of colonies on solid plates, and subsequently tested for HA

accumulation by the alcian blue method. The parentalstrain carrying only the plasmid for HA synthesis (Top10/pMBAD-sseABC ) was simultaneously cultured and used as acontrol. The selection results are plotted in Figure 6, inwhich the D72 strain (Top10/(pMBAD- sseABC , pHACM-rpoD M72 )) harvested from the high mutation frequency

Figure 5. Standard curve of alcian blue staining for HA quanti cation.A second-order polynomial t was used. HA (mg/L)¼ 926.33818–2077.15527Â

OD540 þ 1228.74084Â OD2540, R 2 ¼ 0.99945. (Insert) the ve-well of alcian blue solution

added with different concentrations of HA. The binding time of the alcian blue andHA was 2.5 h. [Colorgure can be seen in the online version of this article, available at

www.interscience.wiley.com.]

Figure 6. Library screening of optimal E. coli for HA accumulation using alcian blue quanti cation. Control strain [Dashed line —100%], Top10 /pMBAD-sseABC . D2 [D2arrow], Top10/(pMBAD-sseABC , pHACM-rpoD M2 ); D33 [D33 arrow], Top10/(pMBAD-sseABC , pHACM-rpoD M33 ); D34 [D34 arrow], Top10/(pMBAD-sseABC , pHACM-rpoD M34 );D72 [D72 arrow], Top10/(pMBAD-sseABC , pHACM-rpoD M72 ); S47 [S47 arrow], Top10/(pMBAD-sseABC , pHACM-rpoS M47 ). The details for the library screening were stated inMaterials and Methods Section. All samples were measured in duplicate and reproduced two times. [Color gure can be seen in the online version of this article, available atwww.interscience.wiley.com.]




in the HA-rpoD library showed the highest increase of HA concentration relative to the control (100% line).The D33 (Top10/(pMBAD- sseABC , pHACM- rpoD M33 ))and D34 (Top10/(pMBAD- sseABC , pHACM- rpoD M34 ))also enhanced the HA titer at the same level. Meanwhile,the D2 strain (Top10/(pMBAD- sseABC , pHACM- rpoD M2 ))exhibited a signicant but lower increase to the control.In the HA-rpoS library, almost all of the mutants causeda noticeable decrease in HA accumulation. Only one

strain, S47 (Top10/(pMBAD- sseABC , pHACM- rpoS M47

)),produced HA at levels similar to the control.

Further Characterization of Improved RecombinantE. coli Mutants for High-HA Production

For classifying the mutants obtained from the screeningabove, three representative strains of D72, D2 and S47 wereranked as high, medium and low increase strain for HAaccumulation, respectively. These strains were cultivated inshake asks simultaneously to con rm the reliability of the high throughput alcian-blue staining screen, usingthe strain of Top10/(pMBAD- sseABC , pHACM-blank) ascontrol C0. Strain Top10/pMBAD- sseABC , the control usedin the library screen, was also used as control in the shakeask experiments (C1). Strain D0 (Top10/(pMBAD-sseABC , pHACM- rpoD)) and S0 (Top10/(pMBAD- sseABC ,pHACM- rpoS)) containing the wild-type rpoD and rpoSgene on the stringent plasmid pHACM, respectively, werecultured as references to observe the function of the rpoD orrpoS overexpression or mutation. The culture volume wasscaled up to 40 mL medium in a 300 mL ask, and all of thestrains were cultured for 48 h to measure the HA titer thencalculate the HA productivity per cell weight, as listed inTable I.

It can be seen that three mutants of D72, D2, and S47showed high, medium and low HA accumulation in shakeasks, similar to the trend observed in the high-throughputlibrary screening step. This con rms that the alcian-bluestaining developed in this work can be quantitatively used

for high throughput screen of superior HA-producingstrains.

Control C0 exhibited lower HA titer than control C1, dueto the presence of an additional plasmid. Mutant D72 yielded $ 561mg/L HA with the highest speci c productivity of 264.8 mg HA/g cell, $ 35% productivity increasescompared to control C0 and $ 15% to control C1. Analysisof D0 and S0, which overexpress the wild-type rpoD andrpoS on a plasmid, gave an interesting result. Overexpression

of the wild-type rpoD in D0 improved HA accumulationonly slightly compared to control C0, but overexpression of the wild type rpoS in S0 caused an increase in cell density, anear 36% higher HA titer with respect to C1, and $ 72%increase with respect to C0. This means that overexpressionof the wild-type rpoS can also be benecial to bothcell growth and HA accumulation, although the mutantS47 did not exhibit a signi cant improvement. Since rpoSorchestrates the expression of stress genes, and it has beensuggested that free RNA polymerase is limiting for sigmabinding (Farewell et al., 1998), rpoS-overexpression couldhelp in coping with the HA-producing stress.

In comparison to D0, the mutant D72 exhibited re-markable HA titer and productivity increase. Therefore, wesequenced and aligned the plasmid-containing rpoD M72 inD72 and rpoD in D0 to verify the function of the mutant s

D

factor. Results showed that the rpoD gene overexpressed inplasmid pHACM- rpoD in strain D0 is identical with thesequence of Genbank J01687.1. For rpoD M72 , however, thereoccurred not only seven site-mutations of 131 T/C, 494 T/C,508 G/A, 608 A/T, 629 A/T, 830 T/C and 994 G/T, but alsoa CAGGCGATCA (10 bp) deletion in the region of 1,309 –1,318 bp of rpoD, which resulted into a coding-shiftthereafter and a TGA stop codon present at 1,357 bp of rpoD72M . Correspondingly, there generated seven aminoacid replacements including I44T, F165S, A170T, D203V,N210I, M277T, D332Y distributed in the region 1 and non-conserved region 2 of the 613 aa (amino acid) s

D factor(Lonetto et al., 1992), as shown in Figure 7. A short mutantterminal of PALSRIRRAPSVFRCI appeared at 437 aa and

Table I. Comparison on cell growth and HA accumulation of the selected mutants from the libraries a .

Strains DCW (g/L) HA titer (mg/L) Productivity (mg HA/g cell)

Increase to C0 (%) Increase to C1 (%)

Titer Productivity Titer Productivity

C0 2.06 404.8 Æ 7.0 196.5 Æ 3.4 — — — —C1 2.21 509.8 Æ 12.5 230.7 Æ 5.6 — — — —D72 2.12 561.4 Æ 5.4 264.8 Æ 2.5 38.7 34.8 10.1 14.8D2 2.33 548.2 Æ 7.6 235.3 Æ 3.7 35.4 19.7 7.5 2.0S47 2.11 479.0Æ 20.6 227.0 Æ 9.8 18.3 15.5 À 6.0 À 1.6D0 2.11 425.0 Æ 5.9 201.4 Æ 2.8 5.0 2.5 À 16.6 À 12.7S0 2.91 695.6Æ 9.7 239.0 Æ 3.3 71.8 21.6 36.4 3.6

b Experiments reproduced three times and the culture conditions were described in Materials and Methods Section. The HA titer was measured by the modi ed CTAB method. The control strain C0 is the Top10/(pMBAD- sseABC , pHACM-blank) with two plasmids and the chromosomal copies of wild-type rpoD and rpoS. Control C1 is the Top10/(pMBAD- sseABC ) which used as the control strain in library screening, harboring a single plasmid and thechromosomal copies of wild type rpoD and rpoS. D0 is Top10/(pMBAD- sseABC , pHACM- rpoD) with an extra copy of unmutated rpoD on pHACM; S0 isTop10/(pMBAD- sseABC , pHACM- rpoS) with an extra copy of unmutated rpoS on pHACM; D2, Top10/(pMBAD- sseABC , pHACM- rpoD M2 ); D72, Top10/(pMBAD- sseABC , pHACM- rpoD M72 ); S47, Top10/(pMBAD- sseABC , pHACM- rpoS M47 ) all have an extra copy of mutant rpoD or rpoS in the pHACMplasmid.




terminated at 452 aa, forming a truncated s D in D72 without

the region 3 and region 4 that conserved in the s D family.

The mechanism by which D72 alters the transcriptionpro le, giving rise to the observed phenotype, is not obviousfrom the sequence data. One possibility is that the effects of this mutant are indirect, through heightening of a genera-lized stress response. Another is that they are due to itsability to retain some, but not all of the functions of thesigma D protein. Although such analysis is out of thescope of the present paper, the sequence-speci c observa-tions do not affect the main result here presented: thesuccessful development of a high-throughput method forisolating improved HA-producing clones from combina-torial libraries.

ConclusionsWe reported a two-step high-throughput HA assay andscreening strategy for recombinant E. coli that overproducehyaluronic acid. The rst step is a qualitative identi cationof translucent colony on sorbitol-containing solid medium,and this approach can probably work well in other Gram-negative hosts for visualization of different exopolysacchar-ide capsule, for example, amylovoran synthesis in Erwinia

Amylovora (Bellemann et al., 1994). The second is the highthroughput quantitative assay of HA concentration by alcianblue staining. Although alcian blue had been used as a stainin surgical pathology for slide staining of acid mucopoly-saccharides (WebPath: Internet Pathology Laboratory), it isthe rst report that we can use it in the solution form andto ef ciently accomplish a high throughput screening of HA-producing E. coli. Since alcian blue can specically bind and aggregate with HA, the alcian blue staining canbe universally used for HA quanti cation in all of HA-secreted producers, such as Streptococcus and Bacillus

subtilis. Certainly we can also use similar procedures tocomplete the screening of other acid mucopolysaccharides.

Based on this two-step high throughput screening method,we performed the screens using two random mutationlibraries of HA-rpoD and HA-rpoS inducing diversity of transcriptional regulation of gene expression during growthand stationary phase, respectively. We speci cally selectedthree representative mutants of D72, D2, and S47 with high,medium and low increase of HA accumulation, respectively,from the mutant pool, to carry out the scaled-up culture andfurther test the reliability of the high-throughput screening.HA titer and productivity results indicated that D72, D2,and S47 maintained the productivity trend when comparedto either control C0 (Top10/(pMBAD- sseABC , pHACM-blank)) or control C1 (Top10/(pMBAD- sseABC )), con rm-ing that the two-step high-throughput screening in this work is a promising approach that can effectively identify the highHA-producing mutant from gTME and any other functionallibraries.

Sequencing results of rpoD M72 in D72 in comparison tothe wild type rpoD in D0 showed that the s

D expressed inthe D72 mutant is a truncation subunit remaining themost conserved region 2 but losing the conserved region3 and region 4, which have been proposed to contain ahelix-turn-helix DNA binding motif, bind the RNA

polymerase core enzyme, bind certain transcription acti-vators and play a role in promoter-identi cation of the RNApolymerase holoenzyme, respectively (Wosten, 1998). Inour previous work, another truncation s

D factor made upof region 4 was also isolated in the second round gTMEscreening for improving ethanol tolerance of E. coli cells(Alper and Stephanopoulos, 2007), suggesting that, becausesigma factors have several functions involved in globalregulation of the transcriptome, truncations losing con-served regions may generate signi cant phenotypicimprovements. For future research, we will carry out new

Figure 7. Location of mutations in plasmid-expressed s D in mutant D72 from the high mutation frequency library of HA-rpoD in comparison with the wild-type s

D in D0, thesame as the sequence of Genbank: J01687.1. The mutagenesis in rpoD M72 resulted in the generation of a truncated s

D factor losing both the conserved region 3 and region 4.[Color gure can be seen in the online version of this article, available at www.interscience.wiley.com.]




experiments for optimization of the HA-producing host andvector, and then speci cally dedicate our efforts to enhancethe HA productivity by using combinatorial libraries. Asa reference, the industrial platform for HA production,a strain of Streptococcus, reaches titers in fermentor (Kimet al., 1996) and ask cultures that ranged from 6 to 10 g/Land 1 to 2 g/L, respectively, a target we wish to attain withfurther efforts.

We acknowledge supports from the Singapore-MIT Alliance (SMA-2)and the National Key Basic Research Project 973 of China(2007CB714304).

REFERENCES

Alper H, Stephanopoulos G. 2007. Global transcription machinery engi-neering: A new approach for improving cellular phenotype. MetabolEng 9:258–267.

Alper H, Moxley J, Nevoigt E, Fink GR, Stephanopoulos G. 2006. Engineer-ing yeast transcription machinery for improved ethanol tolerance andproduction. Science 314:1565 –1568.

Bellemann P, Bereswill S, Berger S, Geider K. 1994. Visualization of capsuleformation by Erwinia amylovora and assays to determine amylovoransynthesis. Int J Biol Macromol 16(6):290 –296.

Chong BF, Nielsen LK. 2003. Amplifying the cellular reduction potential of Streptococcus zooepidemicus. J Biotechnol 100:33–41.

DeAngelis PL, Papaconstantinouan J, Weigel PH. 1993. Isolation of aStreptococcus pyogenesgene locus that directs hyaluronan biosynthesisin acapsular mutants and in heterologous bacteria. J Biol Chem268(20):14568–14571.

Farewell A, Kvint K, Nystrom T. 1998. Negative regulation by RpoS: A caseof sigma factor competition. Mol Microbiol 29(4):1039 –1051.

Ferrante ND. 1956. Turbidity measurement of acid mucopolysaccharidesand hyaluronidase activity. J Biol Chem 220:303 –306.

Follstad B, Balcarcel R, Wang DIC, Stephanopoulos G. 1999. Metabolic ux analysis of hybridoma continuous culture steady state multiplicity.Biotechnol Bioeng 63:675 –683.

Goa KL, Beneld P. 1994. Hyaluronic acid: A review of its pharmacology and use as a surgical aid in ophthalmology and its therapeutic potentialin joint disease and wound healing. Drugs 47:536 –566.

Gregory BD, Nickels BE, Darst SA. 2005. An altered-specicity DNA-binding mutant of Escherichia coli s

70 facilitates the analysis of s70

function in vivo. Mol Microbiol 56(5):1208 –1219.Helmann JD, Chamberlin MJ. 1988. Structure and function of bacterial

sigma factors. Ann Rev Biochem 57:839 –872.Hoover DM, Lubkowski J. 2002. DNAWorks: An automated method for

designing oligonucleotidesfor PCR-based gene synthesis.Nucleic AcidsRes 30(10):e43.

Jin YS, Stephanopoulos G. 2007. Multi-dimensional gene target searchfor improving lycopene biosynthesis in Escherichia coli. Metab Eng9(4):337–347.

Jin YS, Alper H, Yang YT, Stephanopoulos G. 2005. Improvement of xyloseuptake and ethanol production in recombinant Saccharomyces cerevi-siae through inverse metabolic engineering approach. Appl Env Microb71(12):8249–8256.

Kakizaki I, Takagaki K, Endo Y, Kudo D, Ikeya H, Miyoshi T, BaggenstossBA, Tlapak-Simmons VL, Kumari K, Nakane A, Weigel PH, Endo M.2002.Inhibitionof hyaluronansynthesisin Streptococcus equiFM100 by 4-methylumbelliferone. Eur J Biochem 269:5066 –5075.

Kim JH, Yoo SJ, Oh DK, Kweon YG, Park DW, Lee CH, Gil GH. 1996.Selection of a Streptococcus equimutant and optimization of cultureconditions for the production of high molecular weight hyaluronicacid. Enz Microb Technol 19:440 –445.

Kiss RD, Stephanopoulos G. 1991. Metabolic activity control of L-lysinefermentation by restrained growth fed-batch strategies. BiotechnolProg 7:501–509.

Kumari K, Weigel PH. 1997. Molecular cloning, expression, and character-ization of the authentic hyaluronan synthase from group C Strepto-coccus equisimilis. J Biol Chem 272(51):32539–32546.

Lauren TC. 1998. The chemistry, biology and medical applications of hyaluronan and its derivatives. London: Portland Press Ltd.

Lonetto M, Gribskov M, Gross CA. 1992. The s70 family: Sequence

conservation and evolutionary relationships. J Bacteriol 174(12):3843–3849.

Neidhardt FC, Bloch PL, Smith DF. 1974. Culture medium for Entero-bacteria. J Bacteriol 119(3):736 –747.

Ogrodowski CS, Hokka CO, Santana MHA. 2005. Production of hyalur-onic acid by Streptococcus. Appl Biochem Biotechnol 121 –124:753–761.

PenneyDP, PowersJM, Frank M, Willis C, ChurukianC. 2002. Analysis andtesting of biological stains: The biological stain commission proce-dures. Biotechnic Histochem 77(5&6):237 –275.

Sambrook J, Fritsch EF, Maniatis T. 1988. Molecular cloning: A laboratory manual, 2nd edition. Cold Spring Harbor, NY: Cold Spring HarborLaboratory Press.

San KY, Stephanopoulos G. 1984. Studies on on-line bioreactor identi ca-tion. IV. Utilization of pH measurements for product estimation.Biotech Bioeng 26:1209–1218.

Stephanopoulos G, Kelleher J. 2001. How to make a superior cell. Science292:2024–2026.

Stephanopoulos G, Simpson TW. 1997. Flux ampli cation in complex metabolic networks. Chem Eng Sci 52:2607 –2627.

Stephanopoulos G, Sinskey AJ. 1993. Metabolic engineering: Issues andmethodologies. Trends Biotechnol 11:392 –396.

Stephanopoulos G, Fredrickson AG, Aris R. 1979. The growth of competingmicrobial populations in a CSTR with periodically varying inputs.AIChE J 25:863–872.

Vallino JJ, Stephanopoulos G. 1994. Carbon ux distributions at theglucose-6 phosphate branch point in Corynebacterium glutamicumduring lysine overproduction. Biotechnol Prog 10:320 –326.

Venturi V. 2003. Control of rpoS transcription in Escherichia coli andPseudomonas: Why so different? Mol Microb 49(1):1 –9.

Wang FL, Lee SY. 1998. High cell density culture of metabolically engi-neered Escherichia colifor the production of poly (3-hydroxybutyrate)in a dened medium. Biotechnol Bioeng 58(2&3):325 –328.

Widner BR, Behr S, Dollen V, Tang M, Heu T, Sloma A, StembergD, DeAngelis PL, Weigel PH, Brown S. 2005. Hyaluronic acid pro-duction in Bacillus subtilis. Appl Environ Microbiol 71(7):3747 –3752.

Wosten MMSM. 1998. Eubacterial sigma-factors. FEMS Microbiol Rev 22:127–150.

Yu HM, Stephanopoulos G. 2008. Metabolic engineering of Escherichia colifor biosynthesis of hyaluronic acid. Metab Eng 10:24 –32.


Ácido hialuronico, huimin yu et al 2008

Documents