View
3
Download
0
Category
Preview:
Citation preview
ww
w.bi
opro
scal
e-co
nfer
ence
.org
4th BioProScale Symposium
6 to 8 April 2016Berlin – Germany
Institute for Biotechnology and Fermentation in Berlin (IfGB) Seestrasse 13, 13353 Berlin, Germany
In cooperation with
Institute for Biotechnology and Fermentation in Berlin
Organisers Technische Universität Berlin – Chair of Bioprocess Engineering & Institute for Biotechnology and Fermentation in Berlin (IfGB)
Location Technische Universität Berlin, Institute for Architecture, Lecture Hall A151 Strasse des 17. Juni 152, 10623 Berlin (Charlottenburg), Germany
Three-day symposium about industrial scale bioprocess intensification from process development to large-scale understanding
Bioprocess intensification through Process Analytical Technology (PAT) and Quality by Design (QbD)
Industrial-scale bioprocess operation Manufacturing at large scale from process to product, robustness vs. repeatibility,
scaleability of operations, single-use techniques, process analytical technologies and process control
Scale down approaches and process analytical technologies for advanced process design
Comprehensive process monitoring, non-invasive sensors for all process scales, mul-tiparameter and soft-sensors, addressing inhomogeneities by sensor applications and multiposition sampling, modelling and control approaches, influence of early process steps on later down-stream operations, understanding heterogeneity and population development
Quality by design in bioprocess development High throughput applications, model based DoE, in-line process analytics, data
handling and analysis, multivariate data processing, scale down methods, evolutio-nary strategies for strains and processes, computation based process development
Chair of Bioprocess Engineering
• Manifold expression and production strategies
• Record-breaking productivities
• MeOH-induced and MeOH-free processes
• Tailor-made glycoproteins – Pichia Glyco Switch®
www.vtu-technology.com
Pichia pastoris Protein Expression Excellence
AutomAting your imagination
#imagineWithHamilton
Where will your imagination take your science?
4th BioProScale Symposium 2016
3
About the organisers
Technische Universität Berlin: Department of Biotechnology – Chair of Bioprocess Engineering
TheresearchattheChairofBioprocessEngineeringattheTUBerlinisdirectedtothedevelopmentandapplicationofnewmethodsforfasterbioprocessdevelopment,includinggenetic,cultivation,andanalyticaltoolswithaspecialfocusontheindustrialscale.Itaimsspecificallyinunder-standingtheimpactofreactorinhomogeneitiesonthemicrobialmetabolismandadaptation,bothaffectingprocessrobustness.Thisknowledgeisappliedtodesignmolecularbiologicalandprocessengineeringsolutionsandthuscontributestotheunderstandingandimprovementofmicrobialprocessesofbothfundamentalandindustrialinterests.Bycombiningstate-of-the-artcultivation,sensoranddataanalysis,automation,andmechanisticmodellingtechnologieswithmolecularbiologicalandphysiologicaltechniques,theactivitiesattheChairofBioprocessEngineeringcontributetoimprovetheefficiencyofbioprocessesandthustothesocietaladvancementofIndustrialBiotechnology.
www.bioprocess.tu-berlin.de
IfGB – Institut für Gärungsgewerbe und Biotechnologie zu Berlin
Foundedin1874,undertheumbrellaoftheInstituteofFermentationandBiotechnologyinBerlin(IfGB)fermentationorientedresearchandeducationhasbeenconductedformorethen140years–alwaysinclosecooperationwiththeTechnischeUniversitätBerlin(resp.itspredesessorinstitutions).Since2003theVersuchs-undLehranstaltfürBrauereiinBerlin(VLB)e.V.isthesouleholderofIfGB.
Since2003,underthebrandnameIfGBservicesandtrainingforthespiritsindustryanddistillershavebeenoffered.Startingin2009ourserviceandtrainingprogrammeswillbeexpandedintothefieldofbiotechnology–againinclosecooperationwiththeInstituteofBiotechnologyofTUBerlin.
www.ifgb.de
IfGB is a brand of VLB Berlin e.V.www.vlb-berlin.org
Scientific advisory boardFrankDelvigne(UniversityofLiège,Belgium)
KristV.Gernaey(DTU,Denmark)
StefanJunne(TUBerlin,Germany)
ClausLattemann(Sanofi,France)
Welcome addressDearColleagues,LadiesandGentlemenDearGuestsandStudents
Myco-workersandIwouldliketowarmlywelcomeyoutoBerlintothe4thBioProScaleSymposium.Thissymposiumisdedicatedto100yearsofthefedbatchtechnique,whichwasinitiallydiscoveredanddevelopedbyFriedrichHayduckin1915inBerlinatthe“VereinderSpiritusfabrikanteninDeutschlandinBerlin”,apredecessororganizationoftheIfGB.Wealsocelebrate40yearsofBioprocessEngineeringinBerlinwiththeappointmentofMatthiasReussasprofessorforBioverfahrenstechnikattheTUBerlinin1976,and30yearsofthedegreeofBiotechnologyatouruniversity.
Withmorethan200participants,theBioproscaleSymposiumhasbecomeabroadlyacceptedimportantplatformforthediscus-sionofindustrial-scalebioprocessissueswithafocusoncellphysiology.Thespecificcharacterofthissymposiumistheviewonthemethodologyofcharacterizationandcontrollinglarge-scalebioprocessesindependentfromthespecificprocess,andthuswewillhavetalksrangingfrombioenergytobiopharmaceuticalproduction.Importantaspectsforabettercontrolofindustrialscaleproductionarenovelprocessanalyticaltools,modelandsoftsensorapproaches.Keysforthesuccessfulimplementationofbioprocessestowardsabio-basedeconomyarethecomprehensiveunderstandingofscaleupphenomenaandtheavailabilityofeffectivetoolstosimulatelargescaleconditions,andfinallytotesttherobustnessofanewprocessalreadyinalaboratoryenvironment.Currently,weseeanincreasingacceptancefortheideatoconsiderscaleissuesalreadyattheearlyproductdevelopmentstage.Thisprovokesmanynewdevelopmentsinthebiore-actorfield,butalsoneedsparallelprogressinsensoranddatahandlingtechnologies.Integrationofthisknowledgeprovidesaparadigmchangeinthewayofhowtotranslatebiomolecularresearchintonewbioproducts.
Weareverydelightedtohaveawealthofpromisinglecturesinourprogramandwouldliketoextendourgratitudeespeciallytothespeakerswhofollowedourinvitationandwillshareanddiscusstheirexpertisewithus,aswellastoourexhibitorsandsponsors,whoprovidedasubstantialbasisforapleasantatmosphere.IwishyouallaveryinterestingsymposiumandagreatstayinBerlin!
Professor Dr. Peter NeubauerTechnische Universität Berlin – Chair of Bioprocess Engineering
www.bioproscale-conference.org
PeterNeubauer(TUBerlin,Germany)
HenkNoorman(DSM,TheNetherlands)
MatthiasReuss(UniversityofStuttgart,Germany)
RalfTakors(UniversityofStuttgart,Germany)
RolandWeis(VTUTechnology,Austria)
4
13:00 Welcome address and introductionPeterNeubauer(DepartmentofBioprocessEngineering,TUBerlin,Germany)
13:30 Plenary lecture: Reflections of a 40-years ripening process for modeling of bioreactors – from well mixed towards more reality (L01)MatthiasReuss(StuttgartResearchCenterSystemsBiologySRCSB,UniversityofStuttgart,Stuttgart,Germany)
Industrial scale operation
Chair AlvinNienow(UniversityofBirmingham,Birmingham,UK),PeterNeubauer(TUBerlin,Berlin,Germany)
14:15 Keynote lecture: Continuous improvements in the understanding of large scale aerobic fermentation processes (L02) StuartMichaelStocks(Novozymes,Bagsvaerd,Denmark)
14:45 Keynote lecture: Large-scale production of two molecules of pharmaceutical interest in Saccharomyces cerevisiae (L03)ClausLattemann(SANOFI,Paris,France)
15:15 Keynote lecture: Microbial GMP manufacturing at large-scale – challenges and considerations (L04)TorstenSchmidt(Lonza,Visp,Switzerland)
15:45 Sponsor Talk: Bioprocess scale-up from small to large pilot scale using Eppendorf fermentation systems (L05)ChristofKnocke(EppendorfBioprocessCenter,Jülich,Germany)
15:55 Coffeebreakwithpostersessionandexhibition
10:05 Fermentation process gradient effect on Corynebacterium glutamicum 1945 ΔACT3 PTUF-LDCCOPT producing cadaverine (L18)WilliamsOlughu(LoughboroughUniversity,Loughborough,UK)
10:20 Performance loss of Corynebacterium glutamicum cultivations under scale-down conditions (L19)AnjaLemoine(TUBerlin,Germany)
10:35 Studies on the clavam biosynthetic pathway: A controling step for clavulanic acid production (L20)HowardRamirez-Malule(UniversidaddeAntioquiaUdeA,Medellín,Colombia)
10:50 Sponsor Talk: Single-use online capacitance biomass measurement supporting process scale-up from 10 L to 1000 L in cell culture application (L21)StuartTindal(SartoriusStedimBiotech,Göttingen,Germany)
11:00 Coffeebreakwithpostersessionandexhibition
Metabolic adaptation and population dynamics
Chair MarcoOldiges(ForschungszentrumJülich,Germany),ChristianReitz(TUBerlin,Germany)
9:00 Keynote Lecture: Impacts of large-scale gradients mirrored on metabolic and transcriptional regulation (L15)RalfTakors(UniversityofStuttgart,Germany)
9:30 Rapid feast-famine cycles for studying metabolic response and adaptation representative for large-scale (L16)LiangWu(DSMBiotechnologyCenter,Delft,TheNetherlands)
9:50 Metabolic regulation of carbon metabolism: Key element for robustness to bioreactor inhomogeneities (L17)MichaelLimberg(ForschungszentrumJülich,Germany)
Programme at a glance, Wednesday, 6 April 2016
Programme
Bioprocess and bioreactor modelling
Chair ChristophHerwig(TUWien,Vienna,Austria),NicolasCruz-Bournazou(TUBerlin,Berlin,Germany)
16:40 Keynote lecture: Hybrid modelling and multi-parametric control of bioprocesses (L06)ChristophHerwig(TUWien,Vienna,Austria)
17:00 Industrial case study: Data science investigations to improve bioprocess scale-up (L07)PatrickSagmeister(ExputecGmbH,Vienna,Austria)
17:15 Multivariate data- and knowledge-driven tools for robust and efficient cell culture process development (L08)MichaelSokolov(ETHZürich,Zürich,Switzerland)
17:30 Implementation of computational fluid dynamics (CFD) as a Quality by Design (QbD) tool for industrial process development (L09)ThomasWucherpfennig(BoehringerIngelheimPharma,Biberach,Germany)
17:50 Scale down of industrial scale fermentors using Euler-Lagrange CFD: A penicillin case study (L10)CeesHaringa(DelftUniversityofTechnology,Delft,TheNetherlands)
18:05 A modelling framework for bioreactor simulations connecting physical and biological heterogenities (L11)JérômeMorchain(Laboratoired‘IngénieriedesSystèmesBiologiquesetdesProcédés,Toulouse,France)
18:20 Investigation of scale-up effects on mammalian cell culture process performance and critical quality attributes (L12)MatthiasBrunner(TUWien,Vienna,Austria)
18:30 Keynote lecture: Modelling populations of microorganisms – The Anaerobic Digestion Model No. 1 as an example (L13)KristV.Gernaey(TechnicalUniversityofDenmarkDTU,Lyngby,Denmark)
19:00 Keynote lecture: Need for speed: Tools for enhanced microbial bioprocess development (L14)MarcoOldiges(ForschungszentrumJülich,Germany)
19:30 Poster session, Exhibition & Welcome Reception 21:30 End
Programme at a glance, Thursday, 7 April 2016
4th BioProScale Symposium 2016
5
11:40 Keynote Lecture: Exploiting the heterogeneity of biotic and abiotic phase in order to increase bioprocess robustness: Toward new bioreactor design (L22)FrankDelvigne(UniversityofLiège,Belgium)
12:10 Multiparameter monitoring of gradients in the liquid phase of fermentation processes (L23)AnikaBockisch(TUBerlin,Germany)
12:25 Quantification of spatial heterogeneity in large bioreactors (L24)AndersNørregaard(TechnicalUniversityofDenmarkDTU,Lyngby,Denmark)
12:40 Sponsor Talk: Connected in situ measurement system (L25)YoannGasteuilandCélineVinson(smartINST,Lyon,France)
12:50 Sponsor Talk: Optical sensors for bioprocess development and control (L26)GernotJohn(PreSensPrecisionSensing,Regensburg,Germany)
12:55 Lunchbreak,postersessionandexhibition
Scale down approaches and process analytical technologies for advanced process design
Chair:KristGernaey(TechnicalUniversityofDenmarkDTU,Lyng-by,Denmark),StefanJunne(TUBerlin,Berlin,Germany)
14:00 Keynote Lecture: Bioprocesses under magnification: Learning from single cells about large-scale processes (L27)AlexanderGrünberger(ForschungszentrumJülich,Germany)
14:30 Development of a perfusion system in a microbioreactor using sedimentation as a scale down tool for ATF perfusion bioreactors (L28)SteffenKreye(Glycotope,Berlin,Germany)
14:45 Keynote Lecture: Issues that are relevant when scaling-up hMSC microcarrier production processes in stirred bioreactor (L29)RegineEibl(ZürichUniversityofAppliedSciencesZHAW,Wädenswil,Switzerland)
15:15 Sponsor Talk: Evaluation of a stirred small scale single-use bioreactor for microbial applications (L30)ThomasDreher(SartoriusStedimBiotech,Göttingen,Germany)
15:25 Scalability of cultivations in the 2-dimensional ro-cking single-use bioreactor CELL-tainer (L31)Anna-MariaMarbà-Ardébol(TUBerlin,Germany)
15:40 Towards efficient microaerobic processes using engineered Escherichia coli strains (L32)AlvaroR.LaraandKarimJaen(UniversidadAutónomaMetropolitana-Cuajimalpa(UAM),México)
16:10 Coffeebreakwithpostersessionandexhibition
16:50 Keynote Lecture: Novel fluorescence-based on-line capable monitoring of cell cycle distribution and growth rate for mammalian cell cultures (L33)GrischaFuge,An-PingZeng(TUHamburg,Harburg,Germany)
17:20 Multiposition sampling, multiparameter monitoring and electrooptical tracking of microbial activity: New tools for better understanding biogas production (L34)ErichKielhorn(TUBerlin,Berlin,Germany),JörnBeheim-Schwarzbach(IASP,Berlin,Germany)
17:45 In-line photon density wave spectroscopy for bioprocess monitoring (L35)RolandHass(UniversityofPotsdam,Potsdam,Germany)
18:00 Keynote lecture: Stirred, not shaken – Unsolved issues and new problems in scale-down for scale-up (L36)AlvinW.Nienow(UniversityofBirmingham,Birmingham,UK)
19:30 Conference Dinner RestaurantZillemarkt,Bleibtreustr.48a,10623Berlin(SBahnhofSavignyplatz)
Programme at a glance, Friday, 8 April 2016
Quality by design in bioprocess development
Chair:RalfTakors(UniversityofStuttgart,Stuttgart,Germany),ErichKielhorn(TUBerlin,Berlin,Germany)
9:00 Keynote lecture: Next generation mAb production: Continuous manufacturing, PAT, and real time release (L37)DouglasRichardson(Merck,NewJersey,USA)
9:30 Keynote lecture: High-level protein production with Pichia: Case studies of development path from strain generation over bioreactor (L38)RolandWeis(VTUTechnology,Grambach,Austria)
10:00 Keynote lecture: ESETEC 2.0, a new, secretory Escherichia coli strain for the production of complex recombinant proteins at industrial scale (L39)GuidoSeidel(WackerBiotech,Jena,Germany)
10:30 Different process strategies for the Glutathione overproduction in Saccharomyces cerevisiae (L40)MartinSenz(ResearchandTeachingInstituteforBrewinginBerlinVLB,Berlin,Germany)
10:50 Coffeebreakwithpostersessionandexhibition
11:40 Keynote lecture: Parallel small-scale bioreactors with on-line monitoring to prove the consistency of medium ingredients and pre-cultures and to efficiently assess process characteristics (L41)JochenBüchs(RWTHAachen,Germany)
12:10 Towards intensifying design of experiments: An industrial Escherichia coli feasibility study (L42)MoritzvonStosch(NewcastleUniversity,Newcastle,UK)
12:30 Robot based bioprocess development by automated on-line optimal re-design in parallel bioreactors (L43)M.NicolasCruz-Bournazou(TUBerlin,Berlin,Germany)
12:45 Plenary lecture: From advanced PAT to QbD-compliant quasi-continuous integrated pharmaceutical production (L44)ReinerLuttmann(HamburgUniversityofAppliedSciences,Hamburg,Germany)
13:30 Concluding remarks PeterNeubauer(TUBerlin,Germany)
13:45 End of symposium
6
Opening Session
13:00 Welcome address and introductionPeter NeubauerTechnische Universität Berlin, Institute for Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, peter.neubauer@tu-berlin.de
13:30 Plenary lecture: Reflections of a 40-years ripening process for modeling of bioreactors – from well mixed towards more reality (L01)Matthias ReussStuttgart Research Center Systems Biology (SRCSB), University of Stuttgart, reuss@ibvt.uni-stuttgart.de
Thelectureaimsata“Tourdeforce”througha40yearsperiodofdevelopmentofbioreactormodeling.Outofmypersonalperspectiveitisobvioustoselecttheyear1976asastartingpointforthisretrospective-theyearinwhichIstartedwiththeinstallationofthedepartmentofBioprocessEngineeringwithintheInstituteofBiotechnologyattheTUBerlin.Themodelsusedatthattimewerebasedonthewell-knowntextbookassumptionofwellmixedsystemappliedtosinglephaseandgas-liquidapplications.Thefirstmilestonesforconsiderationofmorerealisticmixingcon-ditionsatlargerscaleweretackledwiththeaidofsocalledmicro-macromixermodelsforsimulatingglucosegradientsduringfed-batchoperationsofbaker´syeastproductioninstirredtankreactors.Thesemodelswerelaterextendedtocouplingmasstransfergas-liquidandmixinginhighlyviscousnon-Newtonianfermentationbroths.Themodelingac-tivitieswerelinkedwithextensiveexperimentalinvestigationsofrecirculationtimedistributionsforsingleandmultipleimpellersystems.Inthemidoftheeightieswestartedwiththeapplicationofcompartmentmodels,whichwerelaterextendedtomultiphaseapplicationsattheUniversityofStuttgart.WesoonrealizedthelimitationofthesemodelsafarastheproblemsofquantitativepredictabilityofthemanyparametersatdifferentscaleofoperationisconcernedandconsequentlysteppedintotheapplicationofCFD.Inthebeginningoftheninetiestheseapplicationswereperformedwiththeaidofhybridmodelingapproaches-anintegrationofcompartmentmodelsandCFD.Afteraperiodofstudyingalternativestothek-Emodelsthetoolboxofsimulatingthefluiddynamicconditionsinsingleandmultipleflowwascoupledtoagent-basedsimulationsofthestructuredbiophaseviaEuler-Lagrangemodelingstrategies.Forthefirsttimeitwaspossibletosimulatethelifelinesofsinglecellsandpopulationsinlargescaleoperations.Thepresentationiswrappedupwithadiscussionoffutureperspectivesofmodel-basedscaledownstrategiesandapplicationsofparallelprocessingforsimulationsofverylargebioreactorsusedinbio-industry.
Industrial scale operationChair AlvinNienow(UniversityofBirmingham,Birmingham,UK),PeterNeubauer(TUBerlin,Berlin,Germany)
14:15 Keynote lecture: Continuous improvements in the understanding of large scale aerobic fermentation processes (L02)Stuart Michael StocksNovozymes A/S, Bagsvaerd, Denmark, stus@novozymes.com
NovozymesisalargeandgrowingbiotechnologyconcernheadquarteredinDenmark,producingtensofthousandsoftonsoffinishedenzymeproductseachyearwithfood,feedandtechnicalapplications,aswellasasmallinterestinpharmaceuticalproduction.ScaleupandscaledownofproductionprocessesisacoreactivitysupportedbyanumberofengineersandscientistsbasedinR&D,pilotplantsandproductionfacilitieslocatedaroundtheglobe.Weaimtoimproveourunderstandingofscaleupandscaledown,sowecanbesmarterabouthowwedoit.Muchofwhatwedoisconfidential,butovertheyears,incollaborationwiththeDanishTechnicalUniversity(DTU),andotherinstitutions,wehavebeenactiveinpursuitofnewknowledgeandmodelswhichcandescribefermentationprocessperformanceacrossanumberofscalesandhavemadesomeprogresstowardsthisgoal.Whilethisjourneyisbynomeanscomplete,resultsofworkdoneonagitatorgeometry,oxygentransferandmodellingoftheprocesseshastakenusagooddis-tance;muchofthisdatacannowbepresented.Inaddition,therearenowsomefurtheropportunitiestobecomewiser,namely,thedevelopmentoflowcostofftheshelfprogrammableelectronicspackagesandthematurationoffluorescencespectroscopygivesusanunprecedentedopportunitytopeerintootherwiseopaquestainlesssteelreactors,andvalidateourprogresstowardsbetterunderstandingwhatwearedoing.
14:45 Keynote lecture: Large-scale production of two molecules of pharmaceutical interest in Saccharomyces cerevisiae (L03)Claus LattemannSANOFI - C&BD Biochemistry, Paris, France, claus.lattemann@sanofi.com
Reportsontheproductionofheterologousmetabolitesinmicrobialsystemshavebeenpublishedrecentlydemon-stratingthatvariouscompoundclassescomprisingnaturalandnon-naturalmoleculescanbeproducedinmicroorgan-isms.Productionofsuchcomplexheterologousmoleculesinmicroorganismsinlargevolumefermentationhowever,requiresthecoordinateddevelopmentofproductionstrainsandthecorrespondingfermentationprocess.Twoexam-plesoflargevolumeproductionofactivepharmaceuticalingredientsinbaker’syeastwillbediscussed.
Wednesday, 6 April 2016
Programme
4th BioProScale Symposium 2016
7
15:15 Keynote lecture: Microbial GMP manufacturing at large-scale – challenges and considerations (L04)Torsten SchmidtHead of Microbial Manufacturing, Lonza AG, Visp, Switzerland, torsten.schmidt@lonza.com
Microbialprocessesforproductionofbiopharmaceuticalsarecharacterizedbyahugevarietyofdifferentprocesssteps.Basedontheselectedhoststrainandexpressionsystemtherearemultipledifferentoptionsforprocessingthetargetprotein.Fermentationtiter,stepyieldsandproductpurityarethetypicalselectioncriteriainthisprocessdevelopmentphase.Lacksinscalabilityandprocessrobustnessareconsideredwithalowerpriority.Forthefirstproductionofearlyclinicalmaterialthosepotentialissuescanbeeasilysolvedbytheflexibilityofsmall-scalefacilitiesandoversizing.Whenitcomestolarge-scaleothersolutionsareneeded.Basedonthedesignofthefacilitylarge-scaleplantsforlate-phaseclinicalorcommercialmanufacturingarelessflexible.AlsoCostofGoods(COGS)haveamoreimportantroleatlarge-scale.
15:45 Sponsor Talk: Bioprocess scale-up from small to large pilot scale using Eppendorf fermentation systems (L05)Ulrike Becken1, Bin Li2, Christof Knocke3, Jens Schiffler1, Bruno Sommer Ferreira4, Ma Sha2
1Eppendorf AG Bioprocess Center, Jülich, Germany, 2Eppendorf Inc., USA, 3Vaudaux-Eppendorf AG, Schönenbuch, Switzerland, knocke.c@eppendorf.ch, 4Biotrend, Portugal
Theultimategoalinbioprocessdevelopmentistherealizationofcommercialproduction.Currently,thescale-upoffermentationprocesses,whichiscriticaltothesuccessofindustrialfermentationforbioproduction,isreceivingmuchattention.Eppendorffermentationsystemscoverawiderangeofworkingvolumesfromlessthan1Ltoaslargeas2,400L.Weinvestigatedtheirscale-upcapabilitiesfromsmalltopilotscale.Withinthispresentation,wediscussengi-neeringparameterscriticalforscale-up,suchasvesselandimpellergeometry,tipspeed,mixingtime,oxygentransferrate(OTR),kLa,andpowernumber.Furthermore,wepresentrecentcustomerdataontheproductionofthebiopoly-merpolyhydroxybutyrate(PHB)usingBacillussacchari.Successfulscale-upoftheprocessfrom2Lto200Lexemplifiesthescale-upcapabilitiesofEppendorffermentationsystems.
15:55 Coffeebreakwithpostersessionandexhibition
Bioprocess and bioreactor modellingChair ChristophHerwig(TUWien,Vienna,Austria),NicolasCruz-Bournazou(TUBerlin,Berlin,Germany)
16:40 Keynote lecture: Hybrid modelling and multi-parametric control of bioprocesses (L06)Christoph HerwigChristian Doppler Laboratory for Mechanistic and Physiological Methods for Improved Bioprocesses, Research Division Biochemical Engineering, Institute of Chemical Engineering, Vienna University of Technology, Vienna, Austria, christoph.herwig@tuwien.ac.at
Thegoalofbioprocessingistooptimizeprocessvariables,suchasproductquantityandqualityinareproducible,scal-ableandtransferablemanner.However,bioprocessesarehighlycomplex.Alargenumberofprocessparametersandrawmaterialattributesexist,whicharehighlyinteractiveandmayvaryfrombatchtobatch.Thoseinteractionsneedtobeunderstood,andsourceofvariancemustbeidentifiedandcontrolled.Whilepurelydatadrivencorrelations,suchaschemometricmodelsofspectroscopicdata,maybeemployedfortheunderstandinghowprocessparametersarerelatedtoprocessvariables,theycanhardlybedeployedoutsideofthecalibrationspace.Currently,mechanisticmodels,modelsbasedonmechanisticlinksandfirstprinciples,areinthefocusofdevelopment.Theyareperceivedtoallowtransferabilityandscalability,becausemechanisticscanbeextrapolated.Moreover,themodelsdeliveralargerangeofhardlymeasureablestatesandphysiologicalparameters.Forimplementationofmechanisticmodels,however,modelsneedtobesimplifiedandlinkedtoprocessparametersforrealtimeexecution.Forthis,hybridmodels,hencelinksbetweenDataDrivenandMechanisticModelsmaybeahelpfulsolution.Moreover,modelsneedtobedeployedinthecontrolcontext:Bioprocessesneedtobecontrolledontheonehandondifferentparameterssimultaneously(e.g.constantprecursorconcentrationandspecificgrowthrate)andontheotherhandmayhavedifferentobjectivefunc-tions(maximumproductivityandcorrectproductquality).Hencenovelsolutionsandcasestudiesformultipleinputandoutputcontrolsneedtobedeveloped,astheyalreadyexistinothermarketsegments.Thecurrentcontributionwantstodisplaycurrentsolutionsandcasestudiesofdevelopmentanddeploymentofhybridmodelsandmulti-parametriccontrolofbioprocesses.Thefollowingelementscanbecovered:
y Hybridmodelsolutions,combinationsofdatadrivenandmechanisticmodels.y Workflowshowmechanisticmodelscanbedevelopedfromdatadrivenapproachesandviceversa.y Discussionbetweenexplorative(DoE-based)andmodelbasedexperimentaldesigny Implementationofhybridmodelsinthereal-timecontexty ModelsasPATtool:Demonstrationsofcasesinwhichmodelswereasolutiontomeasurelessy Observersolutionsforreal-timeparameteroptimizationy Multiparametriccontrolandeventpredictiony LifeCyclemanagementofmodelsy Knowledgemanagementusinghybridmodels
8
17:00 Industrial case study: Data science investigations to improve bioprocess scale-up (L07)Patrick Sagmeister1, Sandra Abad2, Christoph Herwig3
1Exputec GmbH, Vienna, Austria, patrick.sagmeister@exputec.com, 2Boehringer Ingelheim RCV, Vienna, 3Vienna University of Technology
Scale-upofbioprocessesreferstothetransferofprocessesfromlaboratorytopilotandproductionscale.Thisisanessentialandcriticalstepforthecommercializationofanybiotechnologicalproduct.Thisstepcanbeaccompaniedbyunexpecteddeviationsimpactingonprocessperformanceandproductquality.Scale-dependentdeviationsaretypi-callyrecognizedverylateinthecommercializationofbiotechnologicalproducts,therebyexhibitingahigheconomicimpact,suchasfailedbatchesandreprocessingunderGMPregulations.Therefore,itisoftheutmostimportancetoachieveathoroughandscience-basedunderstandinginordertomakethescale-upofbioprocessesevenmorerobustandpredictable.Inthepresentedindustrialcasestudyisdemonstratedhowthescale-upofanindustrialE.colimanufacturingprocesswasimprovedbasedonadatascienceapproach,leveragingintegrateddatafromprocessdevelopment,pilotingandmanufacturing.Firstly,wepresenthowadatascienceframeworkwasputtopracticetoidentify-andunderstandscale-dependentchanges.Secondly,wediscusshowscale-upimprovementstrategies,includingthedefinitionofscale-downmodels,werederivedfromtheobtainedmechanisticinsights.Thirdly,wediscusshowasystematicanalysisofavailablebigdatafrommanufacturingdatabasescanbeleveragedtomakebioprocessscale-upmorepredictableandderiveclear-andinterpretabledesignguidelinesforthedesignandscale-upofbioprocesses.
17:15 Multivariate data- and knowledge-driven tools for robust and efficient cell culture process development (L08)Michael Sokolov, Alessandro Butte, Massimo MorbidelliETH Zurich, Zurich, Switzerland, alessandro.butte@chem.ethz.ch
Inrecentyears,theEuropeanMedicinesAgency(EMA)andtheUnitedStatesFoodandDrugAdministration(FDA)havestressedtheneedofgreaterprocessunderstandingandqualitycontrolintheareaofpharmaceuticaldevelopment,manufacturingandqualityassurance,allwhichissummarizedintheProcessAnalyticalTechnology(PAT)initiative.Inbioprocesses,andparticularlyincellcultures,extensivedataacquisitionandanalysisarelikelytoprovideversatileandsignificantreal-timeinformation.Similarly,theQualitybyDesign(QbD)initiativewaslaunchedinordertopromotetheneedforbuildingproductqualityintotheprocess.Inprocessdevelopment,thiscanbeachievedbyadeepprocessunderstandingcombinedwithabroadinvestigationoftheproductqualityinordertoconnectthosetwodomainsandensurerobustproductqualityinanefficientmanner.Allofthesegoalsgoalongwithverylargedatasets,whichusuallyfeatureahighdegreeofcorrelatedvariables.Thismotivatedtodevelopmultivariateapproaches,whichreducethosehigh-dimensionalproblemstothemajoruniquepropertiesoftheprocessaswellasthecorrespondingproduct.Thisworkpresentsseveralcasesstudies,wheredifferentmultivariatedata-andknowledge-driventoolsweresuccessfullyappliedatvari-ousstagesofcellcultureprocessscreeninganddevelopment.Inparticular,highthroughputscreeningforearlyprocessdevelopment,productqualitypredictionatmultiplescalesduringprocesscharacterization,processmonitoringbyRamanspectroscopyaswellashybridapproachesincorporatingadeterministicprocessmodelbackbonewillbediscussed.Allthepresentedresultsgowellbeyondthepos-sibilitiesofferedbycommercialtoolsandprovideanimportantbasisforriskreductionandefficiencyduringprocessdevelopment.
17:30 Implementation of Computational Fluid Dynamics (CFD) as a Quality by Design (QbD) tool for industrial process development (L09)Thomas Wucherpfennig1, Erik Hasenfus1, Johannes Wutz2, Ralf Takors2 und M. Berger1 1Biopharma Process Development, Biberach an der Riss, Germany, Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach an der Riss, Germany, thomas.wucherpfennig@boehringer-ingelheim.com, 2Institute of Biochemical Engineering, University of Stuttgart, Stuttgart, Germany
Thedemandforcomplextherapeuticproteinsandantibodiesfortreatmentofvariousdiseaseshasincreasedcon-tinuouslyinthelastdecades,promptingthecontinuousdevelopmentofnewprocesses.Forthedevelopmentofthesenovelprocesses,theQualitybydesign(QbD)approachhasgainedpopularitysincetheFDA´sPAT-publication.However,thispracticerequiresathoroughunderstandingofaproductandespeciallyofthemanufacturingprocessandequipment.Toolstofurthersuchanequipmentunderstandinge.g.forthesimulationofthecomplexturbulentflowsinstirredtankbioreactorshavebeenaroundforsometime.Theirapplicationinthebiomanufacturingindustry,however,hasremainedscarceduetothedemandofcomputationalpower,complexunderlyingmathematicsandthehighamountoftrainingneeded.Inthisstudyitisdemonstrated,howthesimulationofscale-upcriterialikeP/V,kLa,ormixingtimescanfacilitatearobustscale-uporbeusedinafacilityfitanalysistodeterminewhetheradevelopedprocessissuitedtotheproduc-tionequipmentatvarioussites.Furthermore,aspecificapplicationexampleisprovidedbyusingCFDmethodsfortheset-upofaprocessdesignspaceforanacceptablerangeofthevolumetricoxygenmasstransfercoefficient(kLa).Agitationandaerationaresomeofthefirstparameterswhicharetestedduringprocessdevelopment;basedontheseparametersandthefillingvolumeasinputparametersanDOEstudy(MODDE10,Umetrics,Sweden)wasconductedinanaerated2Lstirredbenchtopbioreactor.ForeachexperimentalsetupatwophaseEuler-LagrangeCFDmodelwasestablishedandakLavaluedeterminedinsilico(Fluent16.6,ANSYSInc.,Canonsburg,USA).AprocessdesignspacewasestablishedformaintainingarequiredkLavalueofa6.5±1h-1neededtosupporthighdensitycellculture.Atafillingvolumeof1.5Lbetween0.09and0.1LL-1min-1andagitationratesbetween197and300rpmaresufficienttomaintainakLavalueof6.5h-1withaprobabilityof95%inthestudiedbioreactorsystem.Allsimulatedvalueswerevalidatedexperimentallyandshowedagoodfit.Theconductedinsilicoap-proachcouldbeusedinfutureprocesscharacterizationstudiesandsavevaluableexperimentaltime.
Programme
4th BioProScale Symposium 2016
9
17:50 Scale down of industrial scale fermentors using Euler-Lagrange CFD: A penicillin case study (L10)Cees Haringa1, Wenjun Tang2, Jianye Xia2, Amit T. Deshmukh3, Matthias Reuss4, Joseph J. Heijnen5, Robert F. Mudde1, Henk J. Noorman3,5 1Chemical Engineering department, Delft University of Technology, Delft, The Netherlands, c.haringa@tudelft.nl, 2East China University of Technol-ogy, Shanghai,China, 3DSM Biotechnology Center, Delft, The Netherlands, 4Stuttgart Research Center Systems Bioloogy (SRCSB), University of Stuttgart, 5Biotechnology department, Delft University of Technology, Delft, The Netherlands
Micro-organismslivinginsideindustrialscalefermentersmaybeexposedtorapidvariationsintheextracellularenvi-ronment,forexample,variationsintheobservedsubstrateconcentration.Boththemagnitudeandfrequencyofthesevariationsmayimpactthemetabolismoftheorganismandtherebytheindustrialviabilityoftheprocess.Toestimatetheirupfrontviascale-downsimulation,itisessentialtogaininsightinthemagnitudeandfrequencyatwhichorgan-ismsareexposedtothesevariations.BasedonpioneeringworkofLapinetal.andDelvigneetal.,weproposetheuseofanEuler-LagrangeCFDframeworktostudytheseaspectsonanindustrialscale.TheLagrangianpointofviewallowstoquantifytheobservedfermenta-tionenvironmentfromthepointofviewoftheorganism,andanalysetheobserved‘lifelines’toextractthefrequenciesandmagnitudeofenvironmentalvariations.Fromthisinformationitispossibletodeterminethedesignparametersforascaledownsimulator,asspecifiedbyNoorman,andset-upanindustriallyrelevantscale-downsimulation.Wepresentacasestudy,basedonthefermentationofPenicilliumchrysogenum,inwhichweshowcasetheproceduretoextracttherelevantinformationfromCFDsimulations,andtheconsecutivedesignofscale-downsimulatorsbasedonthisinformation.Withthis,weshowacomprehensivemethodologyforthedesignofquantitativelyrepresentativescale-downsimulatorsbasedonCFDdata,andhowthismethodologycanbeappliedinpracticalsituations.Thisisamulti-partyresearchproject,betweenDSM-SinochemPharmaceuticals,TUDelft,EastChinaUniversityofScienceandTechnologyandGuojia,subsidizedbyNWOandMoST.
[1]ALapinetal.ModelingthedynamicsofE.colipopulationsinthethree-dimensionalturbulentfieldofastirred-tankbioreactor,astructuredsegregatedapproach.ChemEngSci61(14):4784-4797,2006.[2]FDelvigneetal.Amethodologyforthedesignofscale-downbioreactorsbytheuseofmixingandcirculationstochasticmodels,BiochemEngJ28(3):256-268,2006.[3]HJNoorman,Anindustrialperspectiveonbioreactorscale-down:whatwecanlearnfromcombinedlarge-scalebioprocessandmodelfluid-studies.BiotechnolJ,6(8):934-943,2011
18:05 A modelling framework for bioreactor simulations connecting physical and biological heterogeneities (L11)Maxime Pigou1,2,3 and Jérôme Morchain1,2,3
1Laboratoire d’Ingénierie des Systèmes Biologiques et des Procédés, Toulouse, France, morchain@insa-toulouse.fr, 2INRA, Ingénierie des Systèmes Biologiques et des Procédés, Toulouse, France, 3CNRS, Toulouse, France
Largescalegradientsofsubstrateandoxygenareoftenobservedatthereactorscalebecauseofinsufficientmacromix-ingandheterogeneousgas-liquidtransferrates.ItappearsthatthemostpromisingapproachresidesinthecouplingofpopulationbalancemodelsalongwithCFDandmetabolicapproaches.Thisrequiresahighlevelofexpertiseinseparatescientificdomains.Moreovertheset-upofsuchhighresolutionmodelsisverytimeconsuming.Inordertoeasethetaskwehavedevelopedasimulationtool,namedADENON,thatcombineseachtypeofmodels(hydrodynamics,populationbalanceandmetabolic)inageneralframework.RatherthanusingafullyresolvedCFDmodel,hydrodynamicsdescriptionisbasedonacompartmentapproach.Apopulationbalancemodeldescribingadaptationdynamicsisformulatedintermsofuptakeratedistribution.Thepopulationbalanceapproachsolvestheproblematicofaccessingindividualhistory,atlowercostthanLagrangiantrackingmethods,whilehandlingtheissuethatriseswhentwopopulationswithdifferentpropertiesaremixed.Finallyametabolicmodelisinvokedtocalculatethefateofthesubstrateinsidethecell,bytakingintoaccountthecellstateandlocalconcentrations.Thespecificityhereliesinthefactthatthereisnooptimizationproceduretocomputethemetabolicfluxes.Altogether,thesemodellingchoicesallowforasimultaneouscalculationofallaspectsoftheproblematareasonablecomputationalcost.Severalexamplesofdynamicsimulationofmicrobialcultivationinunsteadycontinuous,multi-stageandlargescalebioreactorswillbepresentedandtheresultscomparedtoavailableexperimentaldata.
18:20 Investigation of scale-up effects on mammalian cell culture process performance and critical quality attributes (L12)Matthias Brunner1,2, Jens Fricke1,2, Christoph Herwig1,2
1Research Division Biochemical Engineering, Institute of Chemical Engineering, Vienna University of Technology, Vienna, Austria, matthias.brunner@tuwien.ac.at, 2Christian Doppler Laboratory for Mechanistic and Physiological Methods for Improved Bioprocesses, Institute of Chemical Engineering, Vienna University of Technology, Vienna, Austria
Inlarge-scalemammalianprocessesheterogeneitiesofcriticalprocessparametersaremostlikelytooccur,duetodif-ferentgassingstrategiesandelevatedmixingtimescomparedtosmallscalefermentations.Understandingtheeffectoftheseinhomogeneitiesonmammaliancellcultivationsisanessentialrequirementforprocessscale-up.Furthermore,comprehensiveknowledgeoftherelationshipbetweentheprocessandtheproductscriticalqualityattributes(CQAs)isnecessarytosatisfyQualitybyDesign(QbD)guidelines.InthiscontextacomprehensiveDesignofExperiments(DoE)approachincludingvariousproductqualityanalyticswasperformedtoinvestigatetheinfluencesandtheinteractionsofscaleuprelevantparametersaspH,dissolvedoxygentension(pO2)anddissolvedcarbondioxidetension(pCO2)onchinesehamsterovary(CHO)cellbatchperformanceandmonoclonalantibody(mAb)quality.ThroughindependentcontrolofprocesspH,pO2andpCO2wecouldderivesingleprocessparametereffectsandinteractionsonCHOphysiologyaswellasonmAbchargeandglycosylationvari-ants.Finally,processpHwasrecognizedasthekey-parameter,exertingstrongeffectsonCHOmetabolismandmAbCQAs.Subsequently,wefurthermoreinvestigatedtheeffectofpermanentpH-shiftstoelevatedlevels,similartothosepresentinlarge-scaleduetobaseaddition(pH7.8and9.0).Therefore,onecompartmentbatchfermentationsincludingpH-shiftswereconductedandshort-termaswellaslong-termeffectsofexternalpH-shiftsonintracellularpHandoverallCHOmetabolismcouldbederived.
10
Programme
Withinthiscontributionacomprehensivestudyregardinglarge-scaleheterogeneitiesonmammaliancellcultureprocessperformanceandcriticalqualityattributeswillbepresented.ThisincludesDoEresultsregardingtheeffectofcriticalscale-upparametersonprocessperformanceattributesandindustry-relevantCQAs.Furthermore,theeffectoflarge-scalepHinhomogeneitiesonintracellularpHandCHOphysiologywillbeevaluatedoutofonecompartmentexperiments.
18:30 Keynote lecture: Modelling populations of microorganisms – the Anaerobic Digestion Model No. 1 as an example (L13)Krist V. Gernaey1, Ulf Jeppsson2, Damien Batstone3 and Xavier Flores-Alsina1
1Department of Chemical and Biochemical Engineering, Technical University of Denmark (DTU), Lyngby, Denmark, kvg@kt.dtu.dk, 2Division of Industrial Electrical Engineering and Automation, Department of Biomedical Engineering, Lund University, Lund, Sweden, 3Advanced Water Man-agement Centre, The University of Queensland, St Lucia, Brisbane, Queensland, Australia.
Manyindustrialscalefermentationprocessesrelyonpurecultures.However,thereareexceptions,forexampleayogurtproduction.Withthecurrentfocusontheproductionofbiofuelsandchemicalsfromrenewables,newproc-essesaredeveloped,andthereisalsoincreasedfocusonthepotentialbenefitsthatcanbeobtainedbyusingmixedculturesinsteadofpurecultures.Forprocessdesignandoperation,itisimportanttodevelopmathematicalmodelsthatcandescribeaprocessrelyingonsuchamixedculture.Fromamodellingpointofview,andtakingapurecultureasareference,thecomplexityofmodellingamixedcultureishigher,sincetheco-existenceofseveralpopulationsofmicroorganismsandtheirinteractionshavetobedescribed.Inthewastewatertreatmentfield,almostallprocessesrelyonmixedcultures,andmathematicalmodellinghasduringthepastdecadesplayedanimportantroleinbuildingupprocessknowledge,andindevelopmentofnoveloperationandcontrolstrategies.TheIWAAnaerobicDigestionModelNo.1(ADM1)istakenasanexamplehere,andstartingfromtheoriginalmodel,aseriesofrecentlypublishedexten-sionsoftheADM1arehighlightedtofunctionallyupgradetheADM1toallowforplant-widephosphorus(P)removalsimulation.ThecloseconnectionbetweentheP,sulfur(S)andiron(Fe)cyclesindeedrequiresasubstantial(andunavoidable)increaseinmodelcomplexityduetotheinvolvedthree-phasephysico-chemicalandbiologicaltransformations.
19:00 Keynote lecture: Need for speed: Tools for enhanced microbial bioprocess development (L14)Johannes Hemmerich, Joachim Koepff, Holger Morschett, Marco OldigesResearch Center Jülich, Institute of Bio- and Geosciences: IBG-1 Biotechnology, Jülich, Germany, m.oldiges@fz-juelich.de
Thebiotechnologicalproductionoffineandbulkchemicalsaswellasheterologousproteinproductionusingmicrobialsystemsisoneimportantfieldinthegrowingbioeconomy.OneoftheimportantkeystosuccessistheincreasingspeedofgeneticmanipulationspossibleforplatformorganismslikeEscherichiacoliorCorynebacteriumglutamicumandothers.Thisallowstoengineermicrobialstrainsinafastwayandeasilyprovidesstrainlibrariesharbouringlargebiologicalvariance.However,thecapabilityofdetailedphenotypingofsuchgenotypelibraryatwell-definedbioprocesslevelisordersofmagnitudeslower.Thus,itrepresentsasubstantialbottleneckinstrainengineeringaswellasbioprocessdevelopment.Especially,thisholdstrueforthechallengeofidentifyingthebestperformingstrainunderbioreactorprocesscondi-tionsinsteadofartificialscreeningconditions.Allthisdemandsforincreasedexperimentalthroughputinmicrobialphenotypingatwell-definedbioprocessconditions.Tofullyreleasethepotentialofrapidstrainandbioprocesscharac-terisationnotonlythecultivation,butalsoupstreamprocessing,analyticsanddatatreatmentneedtobeaccelerated,inordertoprovideavaluableintegratedtoolformicrobialstrainengineering,syntheticbiology,gene-functionrelationshipaswellasbioreactionengineeringandbioprocessdevelopment.ThisisillustratedtakingmicrobialapplicationexampleswithCorynebacteriumglutamicum,StreptomyceslividansandChlorellavulgarisascasestudies,showinghowdevelopmentsinminiaturizedcultivationtechnologycombinedwithsmartlabautomationanddataprocess-ingfacilitatesmicrobialphenotypingandbioprocessdevelopment.Suchelevatedthroughputinmicrobialcultivationtechnologyandbioprocessengineeringprovideacomprehensivedatabasisandpavesthewaytogenerateimprovedknowledgeaboutmicrobialsystemsintermsofmetabolism,regulationandapplication.
19:30 Poster session & Exhibition & Welcome Reception Foyer of the Institute for Architecture
21:30 End
ChairofBioprocessEngineering
4th BioProScale Symposium 2016
11
Metabolic adaptation and population dynamics Chair MarcoOldiges(ForschungszentrumJülich,Jülich,Germany),ChristianReitz(TUBerlin,Berlin,Germany)
9:00 Keynote Lecture: Impacts of large-scale gradients mirrored on metabolic and transcrip-tional regulation (L15)Michael Löffler, Joana Simen, Ralf Takors Institute of Biochemical Engineering (IBVT), University of Stuttgart, Stuttgart, Germany, takors@ibvt.uni-stuttgart.de
IndustrialworkhorsessuchasEscherichiacoliareappliedinlargescalebioreactorsandaretherebyexposedtogradientsmirroringimpactsofhydrostaticpressure,insufficientmixingconditionsandspatiallysegregatedcellularactivities.Cellscirculatinginproductionvesselsareexposedtofrequentlychangingmicro-environmentalconditionsthattriggernotonlymetabolicbutalsotranscriptionalcellularresponses.Yet,cellularperformanceinlargescaleisrequestedtomeetexpectationsoflabscalestudiesinordertofulfileconomicdemands(Takors,2012).AtIBVTascale-downdeviceconsistingofastirredtankbioreactorandaplugflowreactorwasinstalledforsimulatinglargescalegradients.Incontrasttothecommonapproachscale-downexperimentswereperformedundersteady-stateconditions.Consequentlyanaccuratelydefinedreferencescenariowassetthatallowedthein-depthelucidationofshort-andlong-termmetabolicandtranscriptionalresponsesofE.coliontheexposedheterogeneities.Morethan600geneswerefoundtobeup-ordown-regulated(threshold1.5fold)duringthefirst110sinPFR.Distinctpatternsofshort-andlong-termadaptationstrategieswerededucedwithrespecttocarbon(C-)andnitrogen(N-)gradients.Theoccur-renceofstablesubpopulationswasobserved.Biologicalmotivatedcriteriaforscale-upwereformulated.TakorsR.2012.Scale-upofmicrobialprocesses:Impacts,toolsandopenquestions.J.Biotechnol.160(1-2):3-9.
9:30 Rapid feast-famine cycles for studying metabolic response and adaptation representa-tive for large-scale (L16)Liang Wu1, Lodewijk de Jonge2, Camilo Suárez Méndez2, Amit Deshmukh1, Sef Heijnen2, Walter van Gulik2, Aljoscha Wahl2, Henk Noorman1,2
1DSM Biotechnology Center, Delft, the Netherlands, henk.noorman@dsm.com, 2Department of Biotechnology, Delft University of Technology, the Netherlands
Inlarge-scalebioreactors,microorganismsfacecontinuouschangesintheirenvironment.Mostprominentareglu-cose,oxygenandshearratedynamics,butalsopH,temperatureandotherfactorscanplayarole.Inlab-scaleresearch,suchdynamicsconditionsareoftennotapplied,oratbestbasedonqualitativescale-downreasoning.Asaresult,theperformanceofcellsinlarge-scalecanbedisappointingcomparedtoexpectationsraisedfromlabtests.Metabolicmechanismsbehindsuchscale-uplossescanbemultiple,butrelativelylittleisquantitativelyknowntoday.Wewillreportonsomerecentresultsfromstudieswithrepetitiveinputcycles,appliedatlab-scale,whereweinvesti-gatedthemetabolicresponseandadaptationoffungiandyeast.Akeyobservationfromthesefeast-faminestudiesisahighturnoverofintracellularstoragecompounds,associatedwithATPfutilecycles.Alsootherregulatorymechanismsarebecomingmoreclearly,gainingvaluableinsightinscale-upperformanceandscale-downdesignrequirements.
9:50 Metabolic regulation of carbon metabolism: Key element for robustness to bioreactor inhomogeneities (L17)Michael Limberg, Tita Aryani, Mathias Joachim, Wolfgang Wiechert, Marco Oldiges Forschungszentrum Jülich / Institute of Bio- and Geosciences- IBG-1: Biotechnology, Jülich, Germany. m.limberg@fz-juelich.de
Performancelossesduringthescale-upfromlaboratoryintoproductionscaleandtheassociatedincreaseofproduc-tioncosts,imperilsthecompetitivenessofsustainablebiotechnologicalproduction.Onereasonforthesedeficienciesisbasedonthedecreasingmixingqualityandtheresultingglucose,oxygenandpHgradientformationattheindus-trialscale.Butnotallorganismsseemtobenegativelyaffectedbysuchinhomogeneouscultivationenvironmentsinasimilarmanner.EspeciallytheindustrialplatformorganismCorynebacteriumglutamicumprovidesanastonishingdegreeofrobustnesstooscillationsthroughsuchgradients.Thisraisesthequestionwhichkeypathwaysareneces-saryforsuchanabilitytoadaptandhowtheyareregulated.Thispotentiallycanleadtonoveltargetsforthedesignofevenmorerobustproductionstrains.ThisstudypicksupthisquestionbygeneratingabroadrangeofoscillatingpH,oxygenandsubstratesupplycondi-tionsandinvestigatingtheirinfluenceona1,5-diaminopentaneproducingCorynebacteriumglutamicumstrain.Thereforeatwocompartmentscale-downdevicecomposedoftwoconnectedstirredtankreactors(STR-STR)wassetup.Thisdeviceprovidesthepossibilityofmonitoringandcontrollingtheinhomogeneouscultivationparametersineachreactorforinvestigatingthesoleinfluenceofoneparameterorthecombinationofseveralinhomogeneities.DuringthisstudyonecompartmentprovideswellaeratedandpHneutralcultivationconditions.Theadditionalcompartmentwasusedforthegenerationofabroadspectrumofadversegrowthenvironmentswhichmimicdifferentsectionsinlargescalebioreactors.Thesesectionsarecharacterizedbyvaryingoxygenavailability,changingpHandoscillatingsubstratesupplyconditions(excessandlimitation).Dependingontheresidencetime(upto3min)andthecombinationofdisturbingparametersinthiscompartmenttheintermediaryformationofsideproductslikeorganicacidsaswellasgrowthrateweresignificantlyinfluenced.ByinvestigatingtheintracellulartranscriptandproteomelevelsthecompartmentspecificactivationofdifferentstressresponsemechanismsaswellasregulatoryactivationandoverexpressionofNAD+regeneratingpathwaysunderO2-limitationwereobserved.Thismatcheswithoscillatingchangesofmetaboliteconcentrationlevelsintheglycolysisandpentosephosphatepathway.Thestudyisthusthefirststeptowardsaglobalunderstandingofbiologicalrobustnessandpredictingtheirlimits.
Thursday, 7 April 2016
12
Programme
10:05 Fermentation process gradient effect on Corynembacterium glutamicum 1945 ΔACT3 PTUF-LDCCOPT producing cadaverine (L18)Williams Olughu1, Chris Rielly1, Chris Hewitt2 and Alvin Nienow3
1Loughborough University, UK, w.c.olughu@lboro.ac.uk, 2Aston University, UK, 3University of Birmingham, UK
Typicallyonscale-upofabioprocess,itisuneconomicaltomaintainhighspecificpowerinputsandhencethemixingtimetendstoincreaseandthelocalmasstransfercoefficientsarereduced,withincreasingsizeofthefermenter.Anincreaseinthetimescalerequiredforthefermentersystemtohomogeniseincreasesthelikelihoodofthepresenceofphysicalandchemicalgradients,i.e.thesubstrate,dissolvedoxygenconcentrations,temperatureandpHbecomemorespatiallyinhomogeneousonalargerscale.Ascellscirculatewithinsuchalargereactor,theyrespondtothevariedenvironmentwithinthesystem,whichmayresultinadeclineintheirproductivityorachangeintheformationofmetabolites.Hereatwo-compartmentmodelconsistingofaplugflowreactor(PFR)andastirredtankreactor(STR)wasusedtoinvestigatehowC.glutamicum1945Δact3puf-ldccopt(modifiedtoproducecadaverine)respondstogradients(pH,dissolvedO2,nutrient)andvaryingmixingtimes,inanattempttomimicwhathappensinalarge-scaleindustrialfed-batchprocess.GlucoseandpHcontrollingagentswereintroducedtoeitherthePFRorSTRreactorstocreatechemicalgradientsofsubstrateand/orpHconcentrationsindifferentregionsoftheflow;thePFRsectionhadvolumesof10%and20%oftheSTRsection.Theresultsshowednodeclineinbiomassproductivityinallsimulationsstudied,butinthesimulationwhichwasclosesttomimickingalarge-scalefed-batchenvironment,a26%lossinfinalcadaverinetitrewasrecordedwhenthemeanresidenceinthePFR(PFRƮmean)wassetat60s.Afurtherincreaseincadaverinelossof36%wasobservedwhenthePFRƮmeanwasincreasedto120s.Thissuggeststhatasthecellsfacedanincreasinglystressfulenvironment,moreenergywasredirectedtoproducestabilisingaminoacidstomaintainhomoeostasis,resultinginalossofcadaverineproductivity.Forthefirsttime,adualstainingprotocolinflowcytometrywasdevelopedforC.glutamicumtoprobeviabilityattheindividualcelllevel.Theflowcytometryanalysesshowedthatcellviabilityimprovedinallcasesstudied,above95%attheendoftheprocess,whichsuggeststhatthecellsbecamebetteradaptedtotheirenvironmentasfermentationprogressed.SamplestakenfromthetopsectionofthePFRhadthehighestconcentrationoflactateandatdrycelldensitiesabove25g/LwastwicethetitreintheSTR.Thisindicatesthatmixedacidfermentationoccurredpredominantlyintheoxygen-depletedPFRsectionofthistwo-compartmentmodel.Comparingthisstudywithothersimilarworks(withE.coli,B.subtilisandS.cerevisiae)confirmsthatonscaleupofbiologicalsystems,thedegreeofresponseorlossofproductivityisorganismspecific.
10:20 Performance loss of Corynebacterium glutamicum cultivations under scale-down conditions (L19)Anja Lemoine1 , Michael Limberg2, Marco Oldiges2, Peter Neubauer1, Stefan Junne1
1Technische Universität Berlin, Department of Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, anja.lemoine@tu-berlin.de, 2Research Centre Jülich, Institute of Bio- and Geosciences, IBG-1: Biotechnology, Jülich, Germany
Scale-downreactorsarefrequentlyusedtostudytheimpactofoscillationsinthelaboratory-scale,astheyoccurinlarge-scalenutrientlimitedfed-batchprocesses.Corynebacteriumglutamicumisawidelyappliedhostorganismofindustrialrelevanceforlarge-scalelysineproduction.Recentexperimentsinmulti-compartmentreactorsindicatedrobustnessofC.glutamicumagainstenvironmentaloscillationsinoxygenandsubstrateavailability[1,2].However,incontrasttoindustrialapplication,mineralsaltmediaismostlyusedintheseexperiences.Inthisstudy,wecultivatedC.glutamicumwithcomplexmediabasedonmolasses,sucroseandcornsteepliquorinathreecompartmentreactor.Theset-upofthescale-downexperimentswerebasedoncomputationalfluidflowstudiesperformedforanindustrial-scalereactorwithfeedadditioninthebottomoftheliquidphase.Threedistinctzonesweremimicked:(i)substratestarvationandoxygenlimitation,(ii)substratelimitationandoxygenexcess,and(iii)substrateexcessandoxygenlimitation[3].Incontrasttopreviousstudies,weobservedceasedgrowthandareducedaccumulationofthemainproductcadaverineaftertenhoursoffeeding,whilethesubstrateuptakeofsucrosewasreducedby20%.Atthesametime,severalorganicacids,amongthemlactateandacetatewereaccumulated,bothwitharateof2mmolL-1h-1.Proteomeanalysisshowedhigherconcentrationlevelsofseveralenzymes,whichindicatesageneralstressresponsetotheinhomogeneousconditions.Insitudeterminationofthecellsizedistributionwithlaser-lightbackreflectionshowedagglomerationunderoscillatorycultivationcondi-tions.Thesefindingsindicatethatthemorphologicalandphysiologicalstatusofcellsarebothdrasticallyinfluencedbytheunfavorablecultivationconditionsastheyappearinthelateproductionphaseatanelevatedcelldensity.
[1]KaessFetal.ProcessinhomogeneityleadstorapidsideproductturnoverincultivationofCorynebacteriumglutamicum.MicrobCellFact,2014.13.[2]OldigesMetal.Scale-downstudyofoscillationsinoxygenandsubstratesupplyforCorynebacteriumglutamicum.NewBiotechnol,2014.31:S50-S50.[3]LemoineAetal.ResponseofCorynebacteriumglutamicumexposedtooscillationscultivationconditionsinatwo-andanovelthreecompartmentscaledownbioreactor.BiotechnolBioengin,2015.116(6):1220-1231.
10:35 Studies on the clavam biosynthetic pathway: a controling step for clavulanic acid production (L20)Howard Ramirez-Malule1, Albeiro Restrepo2, Wilson Cardona3, Stefan Junne4, Peter Neubauer4, Rigoberto Rios-Estepa1
1Grupo de Bioprocesos, Departamento de Ingeniería Química, Universidad de Antioquia UdeA, Medellín, Colombia, howardrm13@gmail.com, 2Grupo de Química Física Teórica, Instituto de Química, Universidad de Antioquia UdeA, Medellín, Colombia, 3Grupo de Química de Plantas Colombianas, Instituto de Química, Universidad de Antioquia UdeA, Medellín, Colombia, 4Chair of Bioprocess Engineering, Department of Bio-technology, Technische Universität Berlin, Berlin, Germany.
Clavulanicacid(CA),producedbyStreptomycesclavuligerus(Sc),isasecondarymetabolitewithclinicalinterestduetoitspotentialtoinhibitserineβ-lactamases;nevertheless,lowyieldsarecommonlyobtainedduringcultivationproc-esses.ThesynthesisofCAismediatedbythesocalledclavambiosyntheticpathway,forwhichsomestepsstillremainunclear;thus,improvedknowledgeaboutitsintricacieswouldhelptoeventuallyidentifycontrollingsteps,specificallythestepsinvolvedintheconversionof(3S,5S)-clavaminicacidinto(3R,5R)-clavulanicacid.Inthiswork,achemostatcultivationofScwasconducted;acetateaccumulationoccurredwithinthetimeframeinwhichCAwasaccumulated.Althoughtheacetylgroupispresentinβ-lactamintermediatecompoundssuchasN-acetyl-glycyl-clavaminicacid(NAG-clavam),itisstillunknownhowitisincorporated.Here,weproposeareactionmechanism–basedonacomputa-tionalsimulationapproach–foracetateincorporationduringNAG-clavamformation,inCAbiosynthesis.
4th BioProScale Symposium 2016
13
10:50 Sponsor Talk: Single-use online capacitance biomass measurement supporting process scale-up from 10 L to 1000 L in cell culture application (L21)Jochen Scholz, Stuart Tindal, Sebastian RuhlSartorius Stedim Biotech GmbH, Göttingen, Germany, stuart.tindal@sartorius.com
Onlinecapacitancemeasurementhasalreadybecomeawell-establishedPATtoolinbiopharmaceuticalapplicationsutilizingtraditionalmulti-usebioreactorandfermentationequipment.AccordingtoAspenReport2013andReportandSurveyofBiopharmaceuticalManufacturingCapacityandProduction(2015),over40%ofthebiomanufacturingrespondersusethistechnologyforbiomassprocessmonitoringorcontrol.AsthisPATtoolemergesinthemulti-useprocessworld,simultaneouslythereisarapiduptaketosingle-use(SU)cultivationequipmentinprocessdevelopmenttoproductionscale(2000L).Consequently,SUcapacitancesensorsforbiomassmonitoringhavebeendevelopedandpresentedtothecommunitythroughoutthepastyears.Recently,BioPAT®ViaMassSUcapacitancemeasurementbecamecommerciallyavailablewithsignificantuptakeinmarket.Capacityadaptationandbioprocesstransferaretwokeyrequirementsthatpresentsignificantchallenges.Duringthescale-upprocess,differenttypesofbioreactorsareused.Today,SUcultivationequipmentisavailablewithgoodscalabilityfromsmallprocessdevelopmenttoproductionscale.Here,alignedgeometricalaspectratiosandwellcharacterizedprocessparametersarethekeyforgoodcomparability.Inthisscale-upcontext,onlinebiomasscapacitancemeasurementcanplayakeyroletomonitorconsistentcellgrowthrateandcultivationperformance.ThistalkdemonstratestheprocessscalabilityofonlinecapacitancemeasurementinaCHOcellculturefrom10Lto1000L.Additionally,across-systemcomparabilityofconventionalstirredtankandrockingmotioncultivationbioreactorsisshown.Theinvestigationsconcludebyshowingacleargoldenbatchtrajectoryofcellgrowthandproductivityinallscalesandbothagitationsystems.Asverification,theonlinecapacitancemonitoringwascorroboratedbymultipleofflinereferencetechniquesduringthefrequentsamplingoftheprocess.Inconclusion,theSUcapacitancetechnologydisplaysthesamecapabilitiesasthemulti-usetechnologyandenablestheidenticalprocesscontrolpotentialatallscales/systems.
11:00 Coffeebreakwithpostersessionandexhibition
11:40 Keynote Lecture: Exploiting the heterogeneity of biotic and abiotic phase in order to increase bioprocess robustness: Toward new bioreactor design (L22)Frank DelvigneUniversity of Liege, Belgium, F.Delvigne@ulg.ac.be
Cell-to-cellvariationrelatedtophenotypicdiversificationstrategiesisknowntoinducevariabilityandlackofpredict-abilityinbioprocesses.Actually,thesinglecelltoolboxcomprisesmodellingproceduresandadvancedexperimentaldevicesthatcouldbeusedtodecipherthisimportantphenomenoninprocess-relatedconditions.Modellingproceduresareactuallyfocusedonthestochasticityassociatedwiththeintrinsiccomponentofbiologicalnoise,theothercomponents,i.e.extrinsicandintracellulardiffusionareoftenomitted.However,inbioprocesscondi-tions,extrinsiccomponentofnoiseisofcriticalimportancesinceitisknownthatenvironmentalfluctuationsarelikelytooccurinlarge-scalebioreactors[4].Alternativemodellingproceduresarethusneededinordertoovercomeactuallimitations.Theactualsinglecellexperimentaltoolboxcompriseshighresolutionandhighfrequencyopticaldevices[5],butneedstobeadaptedforthemonitoringofrelevantcultivationdevices,suchasstirredbioreactor[1-3].Thesecondlimitationoftheactualtoolboxreliesonitsdependenceoffluorescentmarkers.Thisconstraintswillbeillustratedthroughtwobiosensingstrategies,andalternativeapproachescomprisingsub-populationculturingwillbeproposed.Alternatively,abetterunderstandingofphenotypicplasticityduringbioprocessescanleadtonewbioreactordesign.
[1]BaertJ.tal.2015.Phenotypicvariabilityinbioprocessingconditionscanbetrackedonthebasisofon-lineflowcytometryandfitstoascalinglaw.BiotechnolJ10,1316–1325.[2]BrognauxA.etal.2013.Alow-cost,multiplexable,automatedflowcytometryprocedureforthecharacterizationofmicrobialstressdynamicsinbioreactors.MicrobCellFact12.[3]DelvigneF.etal.2015.Dynamicsingle-cellanalysisofSaccharomycescerevisiaeunderprocessperturbation:comparisonofdifferentmethodsformonitoringtheintensityofpopulationhet-erogeneity.JChemTechnolBiotechnol90,314–323.[4]DelvigneF,GoffinP.2014.Microbialheterogeneityaffectsbioprocessrobustness:Dynamicsinglecellanalysiscontributetounderstandingmicrobialpopulations.BiotechnolJ9,61–72.[5]DelvigneF.etal.2014.Metabolicvariabilityinbioprocessing:implicationsofmicrobialphenotypicheterogeneity.TrendsBiotechnol32,608–616.
12:10 Multiparameter monitoring of gradients in the liquid phase of fermentation processes (L23)A. Bockisch1 , J. Biering2, S. Sachse3, W. Vonau3, S. Junne1, P. Neubauer1
1Technische Universität Berlin, Department of Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, anika.bockisch@tu-berlin.de, 2Versuchs- und Lehranstalt für Brauerei in Berlin e.V., Berlin, Germany, 3Kurt-Schwabe-Institut Meinsberg e.V., Meinsberg, Germany
Inlargescalefermentationlikeinthebrewingprocess,processconditionsasthelocalpowerinputandfluidflowmightbeuneven.Thiswouldleadtogradients.Sincetheavailablesensortechnologyisnotdesignedfortheconsiderationofsuchheterogeneities,theknowledgeaboutthemagnitudeofgradientsinbioprocessesisratherlow.Offlinesamplesoronlinedatafromdevicesinstalledatonesinglepositionofthetankarenotrepresentativeforthewholeliquidphase.Computationalfluiddynamicsisoftennotabletodescribegradientformation,asthisdependsonthecells’metabolicactivity,whichevenaltersduringaprocess.Inordertoimproveprocessmonitoringandtoidentifycriticalreactorzonesinafermentationprocess,mobilemulti-parametersensortoolshavebeendevelopedfortheinsituandonlinemonitoringofvariousprocessparameters(pH-value,dO2,dCO2,redoxpotential,conductivity,temperature,pressure)inindustrialbioreactors.Thetime-dependentandspatialdistributioncanbecorrelatedtodatafromofflinestudieslikemetaboliteconcentrationsandlipidcon-tentmeasuredwithHPLCandGCmethodsaswellasthecells’physiologicalstateinvestigatedbyflowcytometry.Whenthisdataisapplied,theimpactofgradientsonthecellularperformancecanbedescribed.
14
Programme
12:25 Quantification of spatial heterogeneity in large bioreactors (L24)Anders Nørregaard1, Stuart M. Stocks2, Brian Madsen3 and Krist V. Gernaey1
1Department of Chemical and Biochemical Engineering, Technical University of Denmark, Lyngby, Denmark, andno@kt.dtu.dk, 2Novozymes A/S, Krogshoejvej 36, Bagsværd, Denmark, 3Novo Nordisk A/S, Hallas Allé, Kalundborg, Denmark
Substrategradientsareknowntoexistinproductionscalestirredtankbioreactorsduetoimperfectmixinganddiffer-encesinthemasstransferrateofoxygen[1,2].Accesstoexperimentalmixingandmasstransferdatafromlargescaleislimited,andtheresultsaredependentonthegeometryofthevesselsandimpellersinuse.Amethodformeasuringoxygengradientsinsituinaproductionscalebioreactorwithminimalalterationofthevesselisproposed.Themethodutilizesdifferenttypesoffiberopticaloxygensensorsthatareintroducedtothebioreactor.Theflexibilityofthesensorsmakesthemethodapplicableintheindustryandfacilitatesthecollectionofexperimentaldataatlargescalewithoutdisturbingtheworkflowattheproductionplant.Thesetuphasbeentestedina1m3pilotscalemixingvesselwithamodeloxygentransferexperiment.Furthermoreithasbeentestedinapilotscalebioreactorduringyeastfermentationinordertoinvestigatethedurabilityofthesensorsduringheatsterilizationcyclesaswellasfoulingfrombiologicalgrowth.Thesetupisabletoquantifydifferencesintheoxygendrivingforceinthemodelsystem,indicatingtheexist-enceofmasstransfergradients.
[1]LarssonGetal.1996.Substrategradientsinbioreactors:Originandconsequences.BioprocessEng14(6),281–289.[2]OosterhuisNM,KossenNW1984.Dissolvedoxygenconcentrationprofilesinaproduction-scalebioreactor.BiotechnolBioeng26,546–550.
12:40 Sponsor Talk: Connected in situ measurement system (L25)Yoann Gasteuil, Céline VinsonsmartINST 213 Rue de Gerland, Lyon, France, celine.vinson@smartinst.fr
Theflowinamixingtanklikeabioreactordependsonseveralparametersasthereactorgeometry,thefluidproperties,theliquidvolume,theimpellergeometryandtherotationspeed.Numericalsimulationshelptooptimizetheproc-esses,buttheyusuallyprovideonlyacrudeapproximationoftherealdynamics.ThesmartINSTsystemallowstruemeasurementofthewayeachcontrolparameterimpactstheflowandthereforethemixingandinternalreactorprocess.ThesystemconsistsofthesmartCENTER,whichisthedatalogger,adataanalyzer,andcommunicationcenterwithsmartPARTs.ThesmartPARTsareembeddedsensors,whichfreelymovearoundthecoreofthefluidinordertomonitorsimultaneously,andinreal-time,theevolutionofproductsandprocesses.AsmartPARTinthereactorcharacterizestheflowandprovidesnewpossibilitiesforprocessoptimization,scale-upandscale-downstudiesandreactoroptimization.ThesmartPARTcharacterizestheflowthankstonewdatalikeagitationrateandhydrodynamicsignature.Theagitationraterepresentstheenergysuppliedinthefluid.ThisrateisobtainedinrealtimethankstoanaccelerometerembeddedinthesmartPART.Thehydrodynamicsignaturecharacterizesflow,providingdetailedunderstandingofit,arisingfromtheagitationratestatisticsconnectedtocontrolparametersettings.Itformsameasurablequantitativedescriptionoftheflow,allowing:
y Flowcomparisony Detectionofzonesofverystrongagitation,whichcandamagecellsy Detectionofdeadareas
Moreover,thewirelessmeasurementdevices,smartPARTs,measurealsoatthemostrelevantplace:atthecoreofthefluidwherethereactionistakingplace.ItcanmeasurepH,turbidity,conductivityandtemperature,withintheprocess.Thisuniquefeaturemakesthesystemextremelyreactiveandreliable.Havingtheabilityofin-linemeasurementsoftheseparametersmakestheinstrumentverysuitedforadvancedprocessdesignusingPATandQbDapproach.Such,itcanalsocontributetobettercontrolledprocesses,beingthebasisofreproducibleproductquality.Duringthepresentation,wewilldescribethetoolstocharacterizeaflowanddetailthebenefitsforprocessdescription,optimizationandcontrol.
12:50 Sponsor Talk: Optical Sensors for bioprocess development and control (L26)Gernot John PreSens Precision Sensing GmbH, Regensburg, Germany, g.john@presens.de
Startingendofthe90schemical-opticalsensorsarenowwidelyusedinbioprocessdevelopmentandproduction.Thesamesensorisavailablefordifferentsizes.ExamplesandlatestdevelopmentsareshownforO2,pHandCO2sensors.
12:55 Lunchbreakwithpostersessionandexhibition
4th BioProScale Symposium 2016
15
Scale down approaches and process analytical technologies for advanced process design
Chair: KristGernaey(TechnicalUniversityofDenmarkDTU,Lyngby,Denmark),StefanJunne(TUBerlin,Berlin,Germany)
14:00 Keynote Lecture: Bioprocesses under magnification: Learning from single cells about large-scale processes (L27)Alexander Grünberger, Wolfgang Wiechert and Dietrich KohlheyerJülich research Center, Germany, a.gruenberger@fz-juelich.de
Ourunderstandingoflarge-scalebioprocessesisstilldominatedbyanaveragecellapproach.However,asconfirmedinrecentyears,isogenicbacterialpopulationscanbephysiologicallyheterogeneous.Thispresentationwillgiveashortoverviewonmicrofluidicsingle-cellcultivation(MSCC)anditsapplicationinthefieldofbioprocessdevelop-ment.Microfluidicsingle-cellbioreactorsofferpreciselycontrolledexternalconditionsandallowsingle-cellanalysisatfullspatio-temporalresolution.OntheexampleofCorynebacteriumglutamicumIwilldemonstratehowpicoliterbioreactorscanbeusedtogetadeeperunderstandingofmicrobialbioprocesses.Finally,existingbottlenecksandthechallengeoftheintegrationofMSCCintotheexistingbioprocessdevelopmentworkflowwillbediscussed.
14:30 Development of a perfusion system in a microbioreactor using sedimentation as a scale down tool for ATF perfusion bioreactors (L28)Steffen Kreye, Rainer Stahn, Antje Danielczyk, Steffen GoletzGlycotope GmbH, Berlin, Germany, steffen.kreye@glycotope.com
Withthebiotechindustrymovingtocontinuousbioprocessingandfocusingincreasinglyonproductquality,perfusionbioreactorsarebecomingmoreandmoreimportant.Whereasbatchorfed-batchbioreactorscanbesuccessfullyrep-resentedinasmallscalewithsufficientthroughputforearlystageprocessdevelopment,the1Lbenchtopbioreactorremainsthesmallestscaledownunitforperfusionculture.Here,wedemonstratethedevelopmentofasedimentationbasedperfusionsysteminthesingle-useambrsystemwithgoodcomparabilityto1LATFperfusionbioreactorswhencomparingcellgrowth,viabilityandproductquality,especiallyglycosylation,forGlycoExpressTM(GEX)andCHOcells.Thesystemisbettersuitedforclonescreening,mediaoptimizationandthepredictionofproductqualitycomparedtobatchorchemostatculture.Furthermore,datawillpresentedofGlycotope’sproprietaryGlycoExpressTMtechnologyshowingextremelyconsistentglycosylationofdifferentglycoproteinsbetweenbatches,batchsizes,reactorsizes,proc-esscontrolstrategies,DSPscalesandproductionsite.
14:45 Keynote Lecture: Issues that are relevant when scaling-up hMSC microcarrier production processes in stirred bioreactor (L29)Regine Eibl, Valentin Jossen, Dieter EiblZürich University of Applied Sciences, School of Life Sciences and Facility Management, Institute of Biotechnology, Switzerland, regine.eibl@zhaw.ch
Thepotentialofhumanmesenchymalstemcells(hMSCs)asagentsforallogeneiccelltherapiesisundoubted.InordertoprovidetherequiredamountsofclinicalgradehMSCs,theavailabilityofefficientscale-upproceduresisstringentlyrequired.Promisingresultshavebeenachievedinstirredbioreactorsthatoperatewithmicrocarriersatbenchtop-andpilotscale.UsageofComputationalFluidDynamics(CFD)togetherwithbioengineeringdata(fluidflow,shearstress,microcarrierdistribution)hascontributedtoanincreaseinhMSCexpansionefficiency.However,formationoflargemicrocarrier-cell-aggregatesmayresultinmasstransferlimitationsandinhomogeneitiesintheculturebroth.Inourpresentationfocusingonmicrocarrier-basedstirredbioreactors,scale-upapproachesforhMScsfromadiposetis-sue(hADSCS)andbonemarrow(hBM-MSCS)arepresented.Inaddition,wediscussanownapproachforthe50Lscale.Cellnumbersbetween1*1010and3.6*1010cellswereachievedwithin7days,whilemaintainingstemcellpropertiesandqualities.Finally,thesuperiorityoftheNS1ucriterionandtheimportanceoftime-dependentmicrocarrier-cell-aggregatesizeforrapidandsuccessfulscale-upofhMSCexpansionsaredescribed.
15:15 Sponsor Talk: Evaluation of a stirred small scale single-use bioreactor for microbial applications (L30)Thomas Dreher1,Marco Leupold1, Ute Husemann1, Mwai Ngibuini2, Gerhard Greller1
1Sartorius Stedim Biotech GmbH, D-Göttingen, Germany, thomas.dreher@sartorius.com, 2Sartorius Stedim Biotech Ltd., Royston, UK
Forasuccessfulprocesstransfer,aswellasforbioprocessdevelopmentandoptimizationasuitablesmallscaleap-proachisessentialtoreducetimeandcosts.Tofurtherincreasetheefficiencyofthesetransferssingle-usetechnol-ogyisveryattractive.Unfortunately,smallscalesingle-usesystemsoftencannotmimictheconditionsofestablishedbioreactors,intermsofoxygentransferandheatremoval,especially,formicrobialapplications.Consequently,havingsuchamodelinsmallscaleisofspecialinterest.Basedmentionedlimitationsasmallscalesingle-usebioreactor,theambr®250modular(250mLmaximalworkingvol-ume),wasdevelopedandevaluated.Thedesigncriteriaandgeometricalratiosaresimilartoestablishedbioreactors.Duetointegratedpumps,processeswithhighautomationeffortscanberealized.Afurtherfeatureistheintegratedexhaustgasanalyzer,fulfillingallrequirementsforPAT.Tobenchmarktheambr®250modularanindustrialrelevantEscherichiacolihighcelldensityprocesswasperformed.Theachievedresultswerecomparedwithaconventionalglassbioreactor(Uni-Vessel®glass5L),wherethesameprocesswascarriedout.ItwaspossibletoreachapeakcelldensityofOD600=335(drycellweight=120g/L).Noprocessdeviationsoccurredduringthefermen-tationintermsofoxygentransferorheatremoval.Theresultsarecompletelycomparabletotheconventionalglassbioreactor.Overall,theambr®250modularisausefultoolforthetransfer,developmentandoptimizationofindustrialbioprocesses.Furthermore,processdevel-opmentcyclescanbeshortenedandthecostsarereducedincludingahighflexibilityduetotheadvantagesofthesingle‐usetechnology.
16
15:25 Scalability of cultivations in the 2-dimensional rocking single-use bioreactor CELL-tainer (L31)Anna Maria Marbà-Ardébol, Peter Neubauer, Stefan Junne Technische Universität Berlin, Institute for Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, a.marbaardebol@campus.tu-berlin.de
Whilescale-upofstirredSUBscanfollowtraditionalproceduresofconventionalSTRs,thescalingofrockingSUBsismoredemanding.However,theselastoneshavetheadvantagethattheyaremoreflexibleintermsoffillingvolumeduetotheabsenceofstirrers.Moreover,inthecaseofthe2-DrockingCELL-tainer,thepossibilitytoincreasethework-ingvolumeduringthecultivationdoesnotonlyreducecontaminationrisks,butalsoreducescellstressduetotransfersteps.Animportantparameterforthescalabilityofmicrobialaerobiccultivationisalsothegasmasstransfercoefficient(kLa),whichcanbemaintainedfromthemLtothe120Lscale.
15:40 Towards efficient microaerobic processes using engineered Escherichia coli strains (L32)Alvaro R. Lara1, Karim E. Jaén2
1Universidad Autónoma Metropolitana-Cuajimalpa (UAM), Mexico, alara@correo.cua.uam.mx, 2Posgrado en Ciencias Naturales e Ingeniería, Universidad Autónoma Metropolitana-Cuajimalpa Ciudad de México, Mexico
Operationalandeconomicconstraintsinlarge-scalebioreactorsoftenresultinlocalorglobalmicroaerobicconditions,whichleadtolessefficientbioprocesses.Escherichiacoli(E.coli)adaptstomicroaerobicitybyactivatingfermenta-tionpathwaysthataccumulateacidicby-products,indetrimentofgrowthrateandbiomassyield(YX/S).Thiswillbeillustratedwithastudyshowingtheeffectsofconstantandoscillatingdissolvedoxygen(DOT)conditionsontheproductionofplasmidDNA(pDNA).Constantmicroaerobicity(3%airsat.)affectedthemetabolism,butnotthepDNAproduction.Incontrast,heterogeneousDOTseverelyimpactedthequalityoftheproducedpDNA.Therefore,themetabolismofE.coliwasmodifiedtobettercopewithconstantmicroaerobicity.Forthatpurpose,genescodingforglobalregulatorslikeCreBandArcA,orforfermentativepathwayswereinactivated.Theperformanceofawild-type(W3110)andengineeredE.colistrainswasevaluatedinbatchculturesatconstantlowDOT.BycombiningthepartialeliminationoffermentationpathwaysandtheexpressionoftheVitreoscillahemoglobin(VHb),a32%decreaseoncar-bonwasteasby-products,24%increaseonYX/Sand13%increaseofgrowthratewereobtained.FluxbalanceanalysisofthebeststrainestimatedmajordifferencesinthefluxesthroughthepentosephosphatepathwayandtricarboxylicacidcycleasconsequenceofVHbpresence.Overall,ourresultsshowthatE.colicanbegeneticallymodifiedtoovercomesomeofthedisadvantagesofmicroaerobicgrowth,whichispotentiallyusefulforbetterbioreactorscale-upandoperation.
16:10 Coffeebreakwithpostersessionandexhibition
16:50 Keynote Lecture: Novel fluorescence-based on-line capable monitoring of cell cycle distribution and growth rate for mammalian cell cultures (L33)Grischa Fuge, Yaeseong Hong, Uwe Jandt and An-Ping ZengInstitute of Bioprocess and Biosystems Engineering, TU Hamburg, Germany, grischa.fuge@tuhh.de
Areliableon-linemeasurementofthecellgrowthrateishighlydesirablebutstilloutofreachforprocesscontrolinmostcases.Formammaliancellcultureon-linemonitoringofthecellcycledistributionisofparticularinterest,sinceitdeterminesthecellgrowthrateandproductivity.Wehavedevelopedanon-linecapablemonitoringtechniqueforthecellcycledistributionofmammaliancellculturesinbioreactors,whichisbasedontwostablyexpressedFluorescent Ubiquitination based Cell Cycle Indicator(FUCCI)proteinsinthecells.AsdemonstratedwithtwoCHOcelllinederivatives,thedetectionofthespecificfluorescencepropertieshasproventobeareliable,qualitativeread-outforcellgrowth.
17:20 Multiposition sampling, multiparameter monitoring and electro-optical tracking of mi-crobial activity: New tools for better understanding biogas production (L34)Jörn Beheim-Schwarzbach*1,Erich Kielhorn2, Alexander Angersbach3, Emma Ritzi1, Manja Laqua2, Anna Vaupel2, Boris Habermann1, Peter Neubauer2, Stefan Junne2
1Institute of Agricultural and Urban Ecological Projects, Philippstraße 13,10115 Berlin, joern.beheim-schwarzbach@iasp.hu-berlin.de; 2Department of Bioprocess Engineering, Technical University of Berlin, Berlin, Germany, erich.kielhorn@tu-berlin.de. 3EloSystems GbR, Berlin, Germany
Gradientsmightappearintheusuallylaminarmixedbiogasfermentation.Therefore,innova-tivesamplingandmonitoringtoolshavebeendevelopedtoachieveabetterunderstandingofgradientsintheliquidphaseinordertoidentifysuitableareasforsampling.Microsensorsforthedeterminationoftemperature,pH-value,redoxpotentialandCO2concentrationhavebeenoptimizedforlong-termuseintheculturebroth.Theyarecombinedtoamulti-parameterprobeandintegratedintoportablelancesystemsthatallowthemovementofsensorsinsidethereactorforthedetectionofgradientsinsitu.Sincetheestimationofthecells’physiologicalstateisdifficult,electro-opticalmonitoringofthepolarizabilityisappliedtogaininformationaboutcellactivity.Itisshownthatthemicrobialactivitycorrelateswellwithprocessparameterssuchasthemethaneformationrate.Thepresentedtoolsallowanimprovedmonitoringoftheliquidphaseandmicrobialactivityinindustrialplantsforonsiteprocessoptimi-zation.Theycontributetoabetterunderstandingandconcomitantlytoareductionofoperationalrisks,especiallyataflexiblefeedstockload,orwheneveraprocessrunsunstable.
Programme
4th BioProScale Symposium 2016
17
17:45 In-line photon density wave spectroscopy for bioprocess monitoring (L35)Roland Hass, Oliver ReichUniversity of Potsdam, Germany, rhass@uni-potsdam.de
Highlyconcentratedbioprocessescauseanalyticalchallengesforprocessspectroscopyduetothesignificantamountoflightscatteringgeneratedbythecultivatedbiomaterial.PhotonDensityWave(PDW)Spectroscopycanindepend-entlydetermineabsorptionandscatteringpropertiesofhighlyturbidliquidmaterials.Withthehelpoffiber-opticalprobesitcanbeusedasnovelprocessanalyticaltechnologyforbioprocessmonitoring.BesidesanintroductionintothePDWtechnology,exampleslikethemashingprocessduringbeerproduction,thecultivationofScenedesmusrubescensinnon-conventionalphotobioreactorsandenzymaticmilkcoagulationwillbediscussed.Itisfoundthatparameterslikebiomassgrowthandbiomassdecompositionorthecaseinconversionproc-essinfreshmilk(hydrolysis,aggregation,gelling)arewellmonitored.
[1]VargasRuizSetal.OpticalmonitoringofmilkfatphasetransitionwithinhomogenizedfreshmilkbyPhotonDensityWavespectroscopy,IntDairyJ26,120–126,2012.[2]HassRetal.IndustrialapplicationsofPhotonDensityWavespectroscopyforin-lineparticlesizing,ApplOptics52,1423–1431,2013.Alsoin:VirtualJournalforBiomedicalOptics8.p.1423–1431,2013.[3]HassRetal.OpticalmonitoringofchemicalprocessesinturbidbiogenicliquiddispersionsbyPhotonDensityWavespectroscopy,AnalBioanalChem407,2791–2802,2015.
18:00 Keynote lecture: Stirred, not shaken – Unsolved issues and new problems in scale-down for scale-up (L36)Alvin W. NienowEmeritus Professor of Biochemical Engineering, University of Birmingham, UK, a.w.nienow@bham.ac.uk
Mostbioprocessesaredevelopedatthebenchscale(mLtoL)andthenscaled-uptostirredbioreactorsofmany10soreven100sofcubicmetresiftheproductoftheprocessofferscommercialpossibilities.Suchscale-uphasoftenbeenbasedonratherarbitraryrulessuchasimpellertipspeed.However,abetterapproachistoconsiderwhatwouldbethedesiredoperatingconditionsatlargescale,especiallytakingintoaccountreasonableoperatingcostsandthendevisesmallscaleexperimentsthatsimulatethelargescaleoperatingconditions-scale-downforscale-up.OneofthemajordifferencesassociatedwithlargescaleoperationistheincreasedmixingtimewhichismanifestedastemporalandspatialinhomogeneitieswithrespecttopH,nutrients,dO2,etc.;andmuchofthisconferenceisdevotedtodevisingsmallscaletestswhichcangiveaninsightintohowthesefluctuationsimpactoncelldensity,productyieldandquality.Otherimportantphysicalfeatureschangewithscale,suchasReynoldsnumberandsuperficialgasvelocity.Theformeroftheseleadstochangesinflowregimesothat,forexample,atverysmallscale(ambr15,whichiscurrentlyverypopularforcloneselectionfortherapeuticproteinproduction),theflowbecomestransitionalratherthanturbulent.Alternatively,withrheologically-complex,veryhighviscositybroths(asfoundwithmycelialfermentations),theflowmaybetransitionalorevenlaminaratthebenchscalebutturbulentornearlysoatthecommercial.Inadditiontotheaboveconsiderationofphysicalbioreactorscale,thereisalsoanimpactofthescaleofthebiologicalentity.Suchascaleisimportantwhenconsideringtheimpactoffluiddynamicstressonproc-essperformance.Acurrentexampleofthisscaleissueisthedifferencebetweenthecultureofanimalcells(~15-18μm)infreesuspensionandofstemcellsonmicrocarriers(~200μm).Thislecturewillgiveabroadbrushairingtoalltheseissues.
19:30 Conference Dinner
Restaurant Zillemarkt, Bleibtreustr. 48a, 10623 Berlin (Charlottenburg)(S Bahnhof Savignyplatz)
ChairofBioprocessEngineering
18
Quality by design in bioprocess developmentChair: RalfTakors(UniversityofStuttgart,Stuttgart,Germany),
ErichKielhorn(TUBerlin,Berlin,Germany)
9:00 Keynote lecture: Next generation mAb production – Continuous manufacturing, PAT, and real time release (L37)Douglas Richardson, David Pollard, John P. HigginsMerck, New Jersey, USA, douglas.richardson14@merck.com
Recentsuccessinmonoclonalantibodybasedimmuno-oncologytherapeuticssuchasKeytrudahaveinitiatedamanufacturingrevolutionwithintheBiopharmaceuticalindustry.Preparationforawaveofimmuno-modulatoryrecep-tor-focusedclinicalcandidateshighlightstheneedforamoreflexiblemanufacturingandtestingnetwork.Tosupportthisrevolutiondevelopmenteffortsfocusingonsingle-usetechnology,continuousmanufacturing,processanalyticaltechnology,IT/MVDA,andrealtimereleasearebeingpursued.Thispresentationwilldescribetheinvestments,needs,andchallengesinnext-generationmAbproductionandtesting.
9:30 Keynote lecture: High-level protein production with Pichia: Case studies of development path from strain generation over bioreactor (L38)Roland WeisVTU Technology GmbH, Grambach, Austria, roland.weis@vtu.com
ProteinproductionprocessesusingPichiapastorisareincreasinglyrecognizedashighlyefficientbymanyindustries.Especiallyforbiopharmaceuticals,key-to-successistherapidandreliablegenerationofhigh-levelproductionstrainsmatchingproteinquantityandqualityattributes.Aseamlesstransferabilityoflab-scaleresultstomorecontrolledbio-reactorconditionsisofutmostimportance.Theapplicationofaversatilepromoterlibraryforfine-tunedexpressionofboth,thetargetgene(s)aswellasspecificorgeneralauxiliarygene(s)createsadiversityofexpressionstrainsthatcanonlybeanalyzedindepthwithasolidhighthroughputcultivationandscreeningsystem.VTU´sreliable96-wellscalecultivationandscreeningplatformdelivershigh-levelproductioncandidatestrainswhichperformequallywellunderbioreactorconditionssothattime-consumingandlaboriousintermediatestepslikeshakeflaskcultivationinordertoselectspecificstrainsareobsolete.ThecreationofnovelAOX1-promotervariantswithuniqueexpressioncharacteristicsinglycerol-orglucose-fedprocesses,fullydevoidofmethanolnowenablemethanol-freehigh-levelproteinproduction,aswell.Alargenumberofmodelproteinsweresubjectedtothishigh-throughputscreeningplatformfollowedbydirecttransferintobioreactors,oftenyieldingmultig/Llevelsofsecretedrecombinantprotein.Straingenerationandbioreactorproductionproceduresofsomeselectedexamplesincludinggrowthfactors,HSA-fusionproteinsaswellasenzymeswillbedisclosedindetail.
10:00 Keynote lecture: ESETEC 2.0, a new, secretory Escherichia coli strain for the production of complex recombinant proteins at industrial scale (L39)Guido SeidelWacker Biotech, Jena, Germany, guido.seidel@wacker.com
SeveralyearsagoWACKERBIOTECHdevelopedauniqueE.colistrainwhichiscapabletosecretecomplexproteinsdi-rectlytothefermentationmedia.Wefocusedourresearchonthenextwaveofbiologicssocalledantibodyfragments(FAbs).OverthelasttwoyearsweredevelopedandengineeredourworkhorseandwehaveintroducedESETEC2.0asanoptimisedplatformfortheproductionofFAbs(mono-,bi-ormultispecific)andothercomplexproteinsassolubleproteinssecretedtothefermentationmedia.Thereforecelldisruptionand/orrefoldingstepsarenotnecessarytoproc-esstheproductofinterest.ESETEC2.0®combinedwithourfermentationprocessisabletoreachfermentationtitersinthe[g/l]-range.Thetalkwilldescribeournewengineeredandoptimisedexpressionsystem,theindustrialfermentationprocessandwewillgivesomeinformationinourintegratedantibodyfragmentpurificationscheme.Wehavedevelopedseveralintegratedprocessesbyusingcelllineselectionwithmediumthroughputscreening,alignedwithfermentationandpurificationbasedonstate-of-the-artanalytics.Ourprocessesarescalabletoindustriallevelstosupplyclinicalprogramswithlargeamountsofhighqualitydrugsubstances.
10:30 Different process strategies for the glutathione overproduction in Saccharomyces cerevisiae (L40)Martin Senz1, Eric Lorenz2, Maximilian Schmacht1
1Research and Teaching Institute for Brewing in Berlin (VLB), Department Bioprocess Engineering and Applied Microbiology, Berlin, Germany, m.senz@vlb-berlin.org Berlin, Germany, 2 Technische Universität Berlin, Chair of Bioprocess Engineering, Berlin, Germany
Glutathione(GSH)playsakeyroleinprotectivemechanismsagainstreactiveoxygenspeciesineukaryoticandprokaryoticorganisms.Duetoitsanti-oxidativecharacteritiswidelyusedinpharmaceutical,cosmeticandfoodindustry.Forinstance,inbakeriesGSHenrichedyeastcellscanbeusedasdoughmodifierhavingtheadvantageofnoneedfordeclarationinthefinalproduct.TherearevariousstrategiesforthefermentationprocessofGSHproduction,themostcommonbeingfed-batchprocedurewiththeyeastSaccharomycescerevisiae.Foranefficientproductionprocess,factorslikemediumcostsandprocessrobustnessatvariousscales,aswellasachievinghighcelldensitywithhighintracellularGSHcontent,areessential.ThistalkprovidesacompilationofdifferentprocessoptimizationstrategiesfortheproductionofGSH-enrichedyeasts.
Programme
Friday, 8 April 2016
4th BioProScale Symposium 2016
19
Incontextofqualitybydesign,studiesaddressingthefermentationformat,feedingstrategy,mediumcomposition,precursorforGSHtransformationanditstypeofsupplementation,arepresented.Promisingprocessformatswerevariedconcerningtheappliedyeaststrainandprocessscale.Therefore,thehereshownresultsofferinsightsandprospectsforcurrentandfurtherprocessdevelopmentsinGSHproduction.
10:50 Coffeebreakwithpostersessionandexhibition
11:40 Keynote lecture: Parallel small-scale bioreactors with on-line monitoring to prove the consistency of medium ingredients and pre-cultures and to efficiently assess process characteristics (L41)Jochen BüchsLehrstuhl für Bioverfahrenstechnik, RWTH Aachen, Germany, jochen.buechs@avt.rwth-aachen.de
ManufacturingofpharmaceuticalproductsstrictlyhastocomplywithGMP,PATandQbDrequirements.Therefore,robustandeasytousetoolsandmethodologiesareurgentlyneededtoassessthequalityofcomplexmediumingre-dientsandtoprovetheconsistencyofpre-cultures.Inaddition,moreprocessunderstandingisrequiredtopredicttheimpactofprocessdeviationsonproductquality.Theapplicationofparallelshakenbioreactorssystems(200µLto20mLscale)isintroducedwhichallowafastandefficientassessmentofbioprocesscharacteristics.Manydifferentstartingconditionscanbechosenandcriticalcomponentscanbepulsedintorunningculturestoinvestigatetheinfluenceofthesevariationsonproductquality.
12:10 Towards intensifying design of experiments: An industrial Escherichia coli feasibility study (L42)Moritz von Stosch1, Jan-Martijn Hamelink2, Rui Oliveira3
1Chemical Engineering and Advanced Materials, Newcastle University, UK - moritz.von-stosch@ncl.ac.uk, 2GSK Vaccines, Laval, Canada, 3REQUIMTE/DQ, Faculty of Science and Technology, University Nova de Lisboa, Campus de Caparica, Caparica, Portugal
ThecombinationofDesignofExperiment(DoE)withmultivariatedataanalysisiscommonlyusedinupstreambioproc-essdevelopmenttoevaluatetheimpactofprocessparametersonthefinalproducttiterand,wherepossible,productquality.However,foranincreasingnumberofprocessparameters,i.e.:designfactors,thenumberofexperimentsincreasesexponentially,reachingnumbers,whichareinfeasibletoexecute.Thiscontributionintroducesanovelconcept,referredtoasintensifiedDesignofExperiments(iDoE),i.e.:intra-exper-imentvariationofthedesignfactors.TheiDoEwasexperimentallyimplementedforanE.coliprocessinindustryandtheresultingdatausedtodevelopadynamichybridmodel.ThemodelwasvalidatedandevaluatedagainstdatafromatraditionalstatisticalDesignofExperiments(DoE).Theresultsshowthatthemodel(whichwasdevelopedbasedonthedynamicexperiments)canaccuratelydescribe12experimentsofclassicalDoE.TheresultssuggestthatiDoEincombinationwithdynamicmodelscanbeusedtosignificantlyreducethenumberofexperimentsandtimelines.Inaddition,themethod-ologydeliversdetailedprocessunderstanding,thussupportingtheimplementationofQbDandPATinprocessdevelopment.
12:30 Robot based bioprocess development by automated on-line optimal re-design in parallel bioreactors (L43)Mariano Nicolas Cruz Bournazou1, Tilman Barz2, Florian Glauche1, Peter Neubauer1
1Chair of Bioprocess Engineering, Institute of Biotechnology, Technische Universität Berlin, Berlin, Germany, nicolas.cruz@mailbox.tu-berlin.de, 2Energy Department, AIT Austrian Institute of Technology GmbH, Vienna, Austria
Modernbio-industrydemandsafasterandmorerobustprocessdevelopment.Toachievethis,advancesinsensortechnology,miniaturization,parallelization,andautomationneedtobesupportedbypropercomputeraidedtoolswhichallowtheoptimaluseofallexistingresourcesincombination,canpredictthebehaviorofthesystemunderdif-ferentconditions,andcreateapaththatsupportsaconsistentflowofinformationbetweenallstagesandscalesduringdevelopment.Duetotheinherentcomplexityofbiologicalsystems,simpleadaptivemodelsandefficientautomaticexperimentalfacilitiesarerequiredtoallowafastandcheapfittingofthemodeltosmallvariations(e.g.betweensimilarstrains,dif-ferentoperatingconditions,changesinsubstratecomposition,etc.).Wepresentaframeworkfortheonlineoptimalexperimentalre-designofparallelnonlineardynamicprocessesandapplyittotheE.coliW3010fed-batchineightmini-bioreactorsinanautomaticliquidhandlingfacility.Bythis,an“intelligent”automaticfacilitythatcanfitsimplemodelstorealdatainminimaltimeisavailable.Thisworkshowsthatthere-computationoftheoptimalsetuponthefly,whiletheautomatizedexperimentisrunning,ispossible.Asaresultafiftyfoldloweraveragevariationcoefficientontheparameterestimates(4.83%insteadof235.86%)comparedtoasequentialOptimalExperimentalDesignprovesthestrengthsoftheapproach.
12:45 Plenary lecture: From advanced PAT to QbD-compliant quasi-continuous integrated pharmaceutical production (L44)Reiner LuttmannResearch Center of Bioprocess Engineering and Analytical Techniques, Hamburg University of Applied Sciences, Germany, reiner.luttmann@haw-hamburg.de
NowadaysdreamsofQbD-compliantpharmaceuticalproductionarebasedonsmallfullcontinuousoperatedinte-gratedprocesseswithaglobalmonitoringofbiologicalvariablesandofcoursewithdisposabletechnology.Thesegoalsarefarawayfromtodayandmoreovercontradictory.Thiscontributiondescribesawaytorealizesomeofthesetaskswithstateoftheartequipmentandprocessstrategies,explainedataprocessdevelopmentofpotentialMalariavaccineproductionwithPichiapastoris.Afirstintensificationofprocessingisbasedonarecursiveexpressionofpharmaceuticalproteinsandcanbeachievedbyatwo-stageupstreamstrategy,includingarepeatedcellbreedingwithasubsequentproteinproductioninaparallelreactor.
20
TheseprocessesweredevelopedinafullyautomatedmanneraccompaniedbyintensivePATapplications.Foradvancedon-linemonitor-ingofbiotechnologicalvariablesspectroscopicin-linemethodsusingNIR-,Raman-and2D-fluorescence-spectrometerandtheassociatedMVDAareappliedfordetectionofcelldensityandglycerolbutalsofortotalsecretedproteinsandcellinternalAOX.Moreoveranat-lineapplicationofIMACwithHPLCenablesthediscriminationofsecretedproteinsandthereforeatargetproteinevaluation.AQbD-compliantproductionofpharmaceuticalproteinsneedsariskanalysisandtheidentificationofCriticalProcessParameterswithappropriateDesign-andControl-Spaces.Thereforeafullyautomatedbioreactorplantwasdevelopedforoptimizationofexpressioncondi-tions.ThisplantimagesthesequentialproductionstrategyandfulfillstheprocessdevelopmentboundaryconditionsofICH-Q8.Thedevelopmentofaquasi-continuousoperatedproductionplantstartswiththeproceduralintegrationofallunitoperationsofanoverallprocess.Thisinturnresultsinatwo-stepupstreamandafive-stepdownstreamprocess,consistingofa10lcellbreedinganda30lproteinproductionreactor,acellclarificationviaaseparator,aretentionofcellulardebrisusingamicro-filtration,andaconcentrationofthesecretedproductbyultra-filtrationwithsubsequentbufferexchangethroughdia-filtration,followedbyafinalpurificationusingcolumnchromatography.Thethreemainoperationscellbreeding,proteinproductionandalldownstreamoperationshererunseriallyover3days,butalsoinparal-lelwithaone-day-offset.Withthisstrategy,aquasi-continuousprocessflowispossible.However,aQbD-compliantproductionprocessrequiresaknowledge-basedsupportandmonitoring.Thisincludes,forexample,anadditionaloptimizationofallindividualdownstreamstepsusingDoE,fromwhichtherespectivedesignspacesandcontrolspaceshavealsotobeidentified.Amonitoringofthereproducibilityoftheentirerepeatedprocessflowrequirestheuseofanon-lineevaluationalongwithaGoldenBatchapproach.ForthispurposetheplantisequippedwithcorrespondingsoftwaremodulessuchasSIPAT,SIMCAQ,andSIMCA-online,andalsowithMATLABforenhanceddataprocessingandin-silicoprocessdevelopment.
13:30 Concluding remarksPeter Neubauer Technische Universität Berlin, Institute for Biotechnology, Chair of Bioprocess Engineering – peter.neubauer@tu-berlin.de
13:45 Endofsymposium
Programme
ChairofBioprocessEngineering
Pleasenote:
5th BioProScale SymposiumApril2018,Berlin
4th BioProScale Symposium 2016
21
Metabolic adaptation and population dynamics
Combination of scale-down approach with 13C-labeling experiments (P01)Pooth Viola1, Nöh K., Wiechert W., Oldiges M.2
1 Institute of Bio- and Geoscience, IBG-1: Biotechnology, v.pooth@fz-juelich.de, 2 Forschungszentrum Jülich GmbH, Jülich, Germany
Imperfectmixinginducesgradientformationinindustrialscalebioreactors.Thereisevidencethatmicroorganismsadapttheirmetabolismtosuchinhomogeneousconditions.Nevertheless,bioprocessesmaysufferfromreducedprofitabilityasmicrobialmetabolismanditsperformancearehighlydependentonenvi-ronmentalconditions.Thus,scale-downbioreactorsystemsareusedasapowerfultooltomimicgradientsduringearlylabscaleoptimization.Inordertofollowmetabolicchanges,13C-labelingexperimentsserveasbasisformetabolicfluxanalysis.Underdynamicprocessconditionsonlyveryshortperiodsofmetabolicsteady-state,be-ingapreconditionforacarbonlabelingexperiment,arepresent.So,forshorttransientbiologicalchangesanisotopicallynon-sta-tionarycarbonlabelingexperiment(INST-CLE),whereatransientlabelenrichmentintheintracellularmetabolitepoolscanbeob-served,isthepreferredexperimentalapproach(Nöhetal.2007).Currently,thereareno13Cmetabolomedatafromscale-downbioreactorexperimentsavailableformetabolicfluxanalysis.Espe-ciallyfastsamplingandfurthersampleprocessing,aswellastheanalysisofmetabolomedatafromdynamicscale-downcultiva-tionsremainchallenginganddemandforsinglebioreactorsetupwithperiodicoscillations.Atthispoint,westartedwiththebatchandfed-batchcultivationofEscherichiacoliK12inatwocompart-mentscale-downbioreactor(STR²,stirredtankreactor)underoscillatingoxygensupply.Duetotheoscillationbetweenaerobicandnon-aeratedbioreactorcompartmentsdifferentgrowthphenotypeswereobserved.Theadditionalsubstrateexcessorsubstratedepletionatfed-batchcultivationsrevealedimpactsonby-productmetabolism.TheresultsofthedynamicconditionsoftheSTR²willbecom-paredtoasingleSTRwithperiodicoxygenlimitationduringcultivation.TheconditionsoftheSTR²scale-downbioreactorwillbetransferredtothesingleSTRwithperiodicoscillations,becauseonlythattypeofbioreactorissuitableforINST-CLEsduetotheintegratedrapidsamplingdeviceandsubstratepulseunit(Nöhetal.2007).Afterthecompleteestablishmentoftheworkflow,thesystemistobeappliedundervariouslimitingconditionstarget-ingindustrialproductionstrains.Thereby,acloserlookintothemetabolicstateofthecellunderdynamiccultivationconditionsisfacilitated.
[1]Nöhetal.(2007).JBiotechnol129(2):249-267
Bacterial phase diagrams: Analyzing and predicting cell-to-cell heterogeneity of microbial gene expression (P02)Iska Steffens1, Alexander Grünberger1, Dennis Binder2, Thomas Drepper2, Wolf-gang Wiechert1, Dietrich Kohlheyer1
1 Institute of Bio- and Geosciences, IBG-1: Biotechnology, Systems Biotechno-logy, Forschungszentrum Jülich GmbH (Juelich Research Center), Germany, steffensiska@yahoo.de, 2 Institute of Molecular Enzyme Technology, Heinrich-Heine-University Düssel-dorf, Forschungszentrum Jülich GmbH (Juelich Research Center), Germany
Phenotypicheterogeneityrepresentsvaryingcellbehaviorofclonalpopulationsandisamajorissueinthecontextofindustrialbioprocessing.Populationheterogeneityisconsideredtoenhance
thefitnessofmicrobialpopulationsinnaturalenvironments,butappearstoberesponsibleforfailure,reducedcontrolandloweryieldinbacterialproductionprocesses.Accordingly,novelanalyticaltoolsandstrategiesneedtobedevelopedtopredictandcontrolbacterialpopulationheterogeneity[1].Inthiscontribution,weintroducetheconceptof“Bacterialphasediagrams”(BPDs).Similartochemicalphasediagrams,BPDscanbeusedtovisualizeandpredictthestate/behaviorofmicrobialgeneexpression(e.g.homogeneousorheteroge-neousgeneexpression)atcertainenvironmentalconditions.Inordertodeterminesingle-cellresponses,novelpicoliterbioreactorcultivationplatformswereused[2,3].AsaproofofprincipleT7-RNA-PolymerasebasedEYFPproductioninE.coliBL21(DE3)[4]wasinvestigated.Thismodelsystemgener-atesanon-gradualandpartiallyinhomogeneousinductionbehavioroverabacterialpopulation.Herewepresentthemanipulationofexpressionbehavior,bymeansofvaryinginductor(hereIPTG)andrepressorconcentrations(hereglucose).Especiallyforlowamountsofinducermolecules,aninhomogeneousall-or-nothingexpressionresponsecanbeobserved.Latestresults,potentialapplicationandongoingresearchdirectionswillbeshown.Ourresultsshowthattheconceptof“bacterialphasedia-grams”hasthepotentialtopredictbacterialbehaviouringene-expression.Thispavesthewayforanimprovedunder-standingofcell-to-cellheterogeneityandcouldleadtonovelstrategiesofcontrollingmicrobialbioprocesses.
[1]GrünbergerAetal.,2014,CurrOpinBiotechnol29:15-23.[2]GrünbergerAetal.,2012,LabonaChip,12(11):2060-2068.[3]GrünbergerAetal.,2015,CytometryA,DOI:10.1002/cyto.a.22779.[4]BinderDetal.,2014,IntegrativeBiol6:755-765.
Chemical evolution of a bacterial proteome (P03)Christian Schipp1,2, Robert Kaml2, Stefan Oehm2, Michael Georg Hösl2, Stefan Junne1, Peter Neubauer1, Nediljko Budisa2
1 Chair of bioprocess engineering, TU Berlin, Berlin, Germany, christian.schipp@tu-berlin.de, 2 Biocatalysis, Department of chemistry, TU Berlin, Berlin, Germany
Theintroductionofalternativebiochemicalbuildingblockstothestandardizedchemistryofalivingcellremainsanimpor-tantmilestoneinsyntheticbiology.Changingthefundamen-talcomponentsoflivesuchasaminoacidsornucleicacidshasmostlydetrimentaleffectsonbiochemicalprocesses.Therefore,thesystemmustadapttowardsthenewsubstrateinanevolvingmannerrewiringitsgeneticbasistoovercomethestructurallyunfavorablereplacement.Duringalong-termevolutionexperiment(LTEE),atryptophanauxotrophE.colistrainwasforcedbyaserial-transferregimeinatryptophan(Trp)limitingenvironmenttoadapttowardsthenon-canonicalanalogL-beta-(thieno[3,2-b]pyrrolyl)-alanine([3,2]Tpa).Genomicandproteomicanalysisrevealeddeepchangesongeneticandtranslationallevel.SomemutationsobviouslyledtotheshutdownofE.coli’sstringentresponse,whichisawell-knownartifactofLTEE.Nevertheless,thereplacementofTrpto[3,2]Tpainmorethan20.000UGGcodonsyieldedinrobustgrowingdescendantsinminimalglucosemediumonlycontainingthesurrogate[3,2]Tpaasatryptophansource.Thisstudyillustratesanapproachfortheconstructionofsyntheticcellswithalternativebiochemicalbuildingblocksexploitingtheflexibilityofthegeneticcode.Furthermore,theexperimentwilldeepentheunderstandingtheevolvabilityoflivingcellsinrespecttonon-canonicalsubstrates.
Poster abstracts
22
Let there be light – Quantitative analysis of phototrophic cyanobacteria in controlled microfluidic environments (P04)Christian Dusny, Kristin Lindstaedt, Andreas Schmid
Department Solar Materials, Helmholtz Centre for Environmental Re-search GmbH – UFZ, Leipzig, Germany, christian.dusny@ufz.de
Recentbiotechnologyincreasinglyfocussesonharness-ingthephotosyntheticmachineryofcyanobacteriafortheproductionofenergycarriersandchemicalbuildingblocks[1].However,inordertofacilitateindustrialprocessesbasedonbioartificialphotosynthesis,thecyanobacterialcellasaliving,light-drivencatalystwithitscomplexmolecularmechanismsandcellularfunctionshastobeunderstoodindetailbymeansofsinglecellanalysis[2].Mostsurprisinglyanddespiteofthehugepotential,singlecellstudiesdealingwiththeanalysisofphototrophicmicroorganismsarescarceandstilllimitedtofeasibilitystudieswithoutproperphysi-ologicalcharacterization.Inourcontributionwedemonstratehowmicrofluidictechnologiescanbeappliedtostudysinglephototrophicorganismsandanswerquestionsconcerningtheirbiotechnologicalapplicationatthescaleofasinglecell.
[1]AngermayrSAetal.2009.Energybiotechnologywithcyanobacte-ria.CurrOpinBiotechnol20:257-263.
[2]OliverJW,AtsumiS.2014.Metabolicdesignforcyanobacterialchemicalsynthesis.PhotosynthRes120:249-261.
LEANPROT: Systems biology platform for the creation of lean-proteome Escherichia coli strains (P05)Sebastian Hans1, Nicolas Cruz Bournazou1, Egils Stalidzans2, Trygve Brautaset3, Klaus Bensch4, Raivo Vilu5, Peter Neubauer1
1 Technische Universität Berlin, Institute of Biotechnology, Chair of Bio-process Engineering, Berlin, Germany, 2 Latvia University of Agriculture, Faculty of Information Technology, Biosystems group, Jelgava, Latvia, 3 Norwegian University of Science and Technology, Department of Biotechnology, Trondheim, Norway, 4 biotechrabbit GmbH, Germany, 5 Competence Center of Food and Fermentation Technologies, Tallinn, Estonia
Therapiddevelopmentofgeneticengineeringand-omicstechnologiestogetherwithnewmetabolicmodelingtoolsallowsthecreationofrecombinantchassiswithsuperiorchar-acteristics.However,thecurrentstrainsstillcontainanumberofsynthesisandregulatorypathwayswhichhaveevolvedinnature,butwhicharenotneededinabioprocessandconsumeenergyandmetabolicresourcesandconsequentlylowerthefinalproductionyields.ThereforeintheLEANPROTERA-SysAPPprojectwecon-sidertodevelopleanproteomestrainsbasedonasystemsbiology-basedstrainengineeringapproach.Theproteomeisoptimizedbyiterativecyclesofmodelingandbiologicalexperimentswhichareperformedbyparallelconsidera-tionofdifferentparametersasDesignofExperiments(DoE)underwell-definedprocessrelatedenvironmentalconditions.Ourapproachincludestheuseofasinglecellmodel,whichincludesthesequestrationofmetabolitesandenergyintothedifferentmetabolicroutesunderconsiderationofthecellcycleandcellvolumedynamics(basedonamodelbyCooperHelmstetterDonachie).Webelievethatthisnewwayofap-proachingtheproblemofproteinproductiontogetherwithstateoftheartchemicalengineeringoptimizationprinciplescouldopenanewavenueforthecreationofsuperiorcellfactories.
Bioprocess modelling
Computational analysis of hydrodynamics and light distribution in photo-bioreactors for algae biomass production (P06)Varun Loomba, Gregor Huber, Eric von Lieres
Forschungszentrum Juelich, Jülich, Germany, v.loomba@fz-juelich.de
Microalgaecanbedirectlyusedinhealthfoodorasbio-filtersforwastewatertreatment.Theyalsohavenumerouscommercialapplicationsincosmetics,aquacultureandchemicalindustryasasourceofhighlyvaluablemolecules,e.g.,polyunsaturatedfattyacids[1].Moreover,theyareincreasinglyrecognizedasapromis-ingsourceforbiodieselproduction[2].Torealizethefullpotentialofmicroalgae,optimaloperatingconditionsfortheircultivationinphoto-bioreactors(PBR)needtobeidentifiedinordertomaxi-mizeproductivity,lipidcontent,andefficiencyofphotosynthesis.ThemostimportantparametersaffectingPBRperformancearereactorshape,lightintensitydistribution,algaegrowthandothermetabolicproperties.ThepresentedstudyaimsatoptimizingtheseparametersusingComputationalFluidDynamics(CFD)simulationswiththeCOM-SOLMultiphysicssoftware.Specifically,flatpanelphoto-bioreac-torswithturbulentmixingduetoairspargingandone-sidedlight-ingarestudied.First,flowprofilesofbothliquidandgasphasesarecomputedusingtheEuler-Eulerapproachforanalyzingtheairsparginganddetectingpotentialdeadzones.Then,lightinten-sitydistributionsarecalculatedinsidedifferentPBRtypes,basedonabsorptionandlightscatteringbyalgaeandgasbubbles.Subsequently,thepathsofindividualalgaearetraced,andtheenvironmentalconditionstheyareexposedto,arerecordedovertime,inparticularaerationandlightintensity.Resultsoftheabovedescribedsimulationstageswillbepresentedanddiscussed.
[1]Spolaoreetal.Commercialapplicationsofmicroalgae,JBiosciBioeng101(2006):87-96.
[2]Bitogetal.Applicationofcomputationalfluiddynamicsformodelinganddesigningphotobioreactorsformicroalgaeproduction:Areview,ComputElectronAgr76(2011):131-147.
Nutrient distribution in perfusion based picoliter bioreactors (P07)Christoph Westerwalbesloh, Alexander Grünberger, Birgit Stute, Sophie Weber, Wolfgang Wiechert, Dietrich Kohlheyer and Eric von Lieres
Forschungszentrum Jülich GmbH, Jülich, Germany, c.westerwalbesloh@fz-juelich.de
Cell-to-cellheterogeneityhasbeenfoundtohavethepotentialtosignificantlyinfluencebiotechnologicalprocesses.Microfluidicde-vicesoffernewopportunitiesforresearchandoptimizationofmi-croorganismsonsingle-celllevel,whichcangivefurtherinsightsinunderlyingcausesandresultingeffectsofheterogeneity[1].Arecentlydevelopedmicrofluidicplatformallowscultivationofsev-eralhundredmicrocolonies,eachconsistingofuptoseveralhun-dredcells,inseparatecultivationchambersononedevice.Data,e.g.growthrates,canbegeneratedbyautomatedtime-lapsemicroscopywithspatiotemporalresolutiononsingle-celllevel[2].Especiallyperfusionbasedcultivationchambertypesareexpectedtooffergoodenvironmentalcontrolwithregardtothecultivationmedium,i.e.constantsubstrateandlowproductandbyproductconcentration.Duetolength-scalesinthemicrometerrangeandcorrespondingsmallvolumesofseveralpicolitersperreactionchamberitiscurrentlyimpossibletodirectlymeasurethosecon-centrationsandtheirgradientswithinthedevices.Thereforethemicrofluidicdeviceemployedin[2]wasmodeledandsimulatedusingCOMSOLMultiphysics.Theliquidvelocityfieldandthemasstransferwithinthesupplychannelsandcultivationchamberswerecalculatedtogaininsightinthespatialdistributionofsuppliednu-
Poster abstracts
4th BioProScale Symposium 2016
23
trientsandmetabolicproductssecretedbythecultivatedbacteria.Thegoalwastoidentifypotentialsubstratelimitationsorproductaccumulationswithinthecultivationdeviceandtheresultinginhomogeneityexperiencedbysinglecells.Themetabolicuptakeandproductionrates,colonysize,andgrowthmediumcomposi-tionwerevariedcoveringawiderangeofoperatingconditions[3].Thislaysthefoundationforfurtherstudiesandtheoptimizationofexistingpicoliterbioreactorsystems.
[1]AGrünbergeretal.Single-cellmicrofluidics:opportunityforbioprocessdevelopment.CurrOpinBiotechnol29,15–23,2014.
[2]AGrünbergeretal.Spatiotemporalmicrobialsingle-cellanalysisusingahigh-throughputmicrofluidicscultivationplatform.CytomPartA87(12),1101–1115,2015.
[3]CWesterwalbeslohetal.ModelingandCFDsimulationofnutrientdistributioninpicoliterbioreactorsforbacterialgrowthstudiesonsingle-celllevel.LabChip,15,4177–4186,2015.
Combination of modeling and environmental life cycle assessment approach for demand driven biogas production (P08)Funda Cansu Ertem, Juan A. Arzate, Peter Neubauer, Stefan Junne, M. Nicolas Cruz-Bournazou
Technische Universität Berlin, Institute for Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, fcertem@mailbox.tu-berlin.de
Asuitablestrategytoeventhefluctuatingenergyprovisionbysomerenewableenergysourcesisaflexibleoperationofbiogasplants.Nevertheless,thisprocessneedstoremainstable,profit-ableandenvironmentallyfriendly.Inordertoidentifyreasonableandsuitableoperationpoints,processmodellingiscombinedwithanLCA.Inthisproof-of-conceptstudy,abiogasprocesswassimulated,whichflexiblygeneratesenergyfromtheco-digestionofmaize,grassandcattlemanure,whileemittingthelowestamountofgreenhousegasemissions.TheprogramwasstructuredinMatlabtosimulateanybiogasprocessbasedontheAMOCOmodelandcombinedwithequationsthatprovideclimatechange,acidificationandeutrophicationpotentialsofthewholeproduc-tionsystembasedonReCiPemidpointv.1.06methodology.Thesimulationwasoptimizedbasedonrealdataofbiogasplantsandexistingliteratureresearch.
Effect of light modeling on the prediction of microalgae growth (P09)Christopher McHardy, Cornelia Rauh
Technische Universität Berlin, Chair of Food Biotechnology and Food Process Engineering, Berlin, Germany, christopher.mchardy@tu-berlin.de
Thepredictionofgrowthofmicroalgaerequiresaccuratemod-elingoflightpropagationinphotobioreactors.AlthoughLambertsLawprovidesasimplewaytoapproximatelightattenuation,itdoesnotcaptureanyscatteringeffectsand,inconsequence,growthpredictionsdeviatefromexperimentalresults.Toreachhigherlevelofaccuracy,numericalsimulationoflightdistributioninphotobioreactorsmayserveasanalternative.Inthepresentcontribution,theeffectsoflightmodelingonpredictionsofmicro-algaegrowthareinvestigatedbynumericalsimulations.
Modelling dissolved oxygen and glucose gradients in pulse-based fed batch culture of Escherichia coli (P10)Emmanuel Anane, Christian Reitz, Franziska Vera Ebert, Mariano Nicolas Cruz, Stefan Junne, Peter Neubauer
Chair of Bioprocess Engineering, Institute of Biotechnology, Technische Universität Berlin, Berlin, Germany, anane@tu-berlin.de
Localadditionofsubstrateandair,coupledwithinsufficientmixinginindustrialscalefed-batchculturesresultintheoc-currenceofzonesthroughoutthebioreactor.E.colicellsthatareintermittentlyexposedtothesezonesreactmetabolicallybyproducingacetate[1],andphysiologicallybyaccumulat-ingnon-canonicalaminoacids(NCAA)inbothcellularandrecombinantproteins[2].Inthecurrentwork,wedevelopadynamicalmechanisticmodeltopredicttheresponseofanE.colistrainforrecombinantproteinproductiontoperturba-tionsinglucosefeedandoxygentransferinfed-batchculture.Themodelisvalidatedwithexperimentaldatafrompulse-basedfed-batchculturesofE.coli.Thesimulationsshowagoodagreementwiththeexperimentaldataandprovideaframeworkformodel-basedprocessmonitoringandoptimi-sationoffermentationstrategies.
[1]ESWernerssonetal.Substrategradientsinbioreactors:originandconsequences.14,1996.
[2]MBiermannetal.Traceelementassociatedreductionofnorleu-cineandnorvalineaccumulationduringoxygenlimitationinarecombinantEscherichiacolifermentation.Microb.CellFact.,vol.12,p.116,2013.
Bioprocess scale-up and scale-down
How to successfully scale up a microcarrier ba-sed expansion process of human mesenchymal stem cells from adipose tissue (P11)Valentin Jossen1, Cedric Schirmer1,2, Regine Eibl1, Matthias Kraume2, Ralf Pörtner3, Dieter Eib1
1 Zurich University of Applied Sciences, Institute of Chemistry and Biotechnology, Switzerland, valentin.jossen@zhaw.ch, 2 Berlin University of Technology (TU Berlin), Institute of Process Engineering, Germany, 3 Hamburg University of Technology (TUHH), Institute of Bioprocess and Biosystems Engineering, Germany
Humanmesenchymalstemcellsisolatedfromadiposetissue(hASCs)areofincreasinginterestforclinicalapplications.Duetotheirgoodclinicalefficacyandtolerability,theycanbeusedforallogeneictherapies(e.g.acutemyocardialinfarc-tion).However,therequirednumberoftherapeuticallyactivehASCsforallogenicapplicationsisintheregionofatrillioncellsperyear.Therefore,alternativesto2-dimensionalplanarcultivationsystems,whicharetypicallyusedtopropagatethem,andsuitablescaleupstrategiesfortheexpansionprocedureareurgentlyrequired.Promisingalternativestotraditionalsystemsarestirredandwave-mixedsingle-usebioreactors.Inthesedynamicallymixedsystems,therequiredgrowthsurfacefortheadherentcellsisprovidedbymicrocar-riers(MCs)thataresuspendedinthebioreactor.However,thisraisesthequestionofhowaMC-basedexpansionprocessforhASCscanbescaledup.AsuccessfulscaleupapproachforaMC-basedproductionprocessispresentedforadiposederivedstromal/stemcellsinastirred(35L;pilotscale)andawave-mixed(2L;benchtopscale)single-usebioreactor.Todeterminetheoptimumme-dium-microcarrier-combination,extensivescreeningstudiesbasedontheexperimentaldesignwereperformedinspinnerflasks.Inaddition,experimentalandnumericalinvestigationswithaspinnerflask,aBIOSTATSTR50LandaBIOSTATCulti-BagRM2Lwerealsocarriedout.Thisallowedtheoptimum
24
cultivationconditionstobepredicted,whilekeepingtheshearstresslevelslow.Thescaleupapproachresultedinpeakcelldensitiesof2.9x105hASCsmL-1(1.015x1010hASCs)intheBIOSTATSTR50Landin1.9x105hASCmL-1(2.85x108hASCs)intheBIOSTATCultiBagRM2L(proof-of-conceptculti-vation).Duringthecultivationthecellsmaintainedtheirstemcellpropertiesandsurfacemarkers.
Scalability of multiple strain co-fermentation (P12)Lene Dybdal Christiansen, David Kold
Chr. Hansen, Hørsholm, Denmark, dkldp@chr-hansen.com
Duringscale-upoffermentationprocesses,scaledependentchangescanhaveasignificantimpactontheprocess.Varia-tioninprocessparametersuchaspHandtemperature,whichcanbeeasilyinvestigatedinlaboratoryscale,mayshowtobeinsignificant,comparedtochangesinmediapreparationorprocesstime.Scale-upoffermentationprocessesweremultiplestrainsareco-fermentedcanbeespeciallychallenging,aseachstraincanhavedifferentsensibilitytoprocesschanges.Thismayleadtoachangedstrainbalanceandtheinteractionsbetweenthedifferentstrainsmaybeaffected.Identifyingthesignificantparameterstocontrolandinvestigatemaythereforprovecomplex.Inthisstudy,theeffectofchangesinthemediapreparationbetweenlaboratoryandproductionscalefermentationsofamultiplestrainlacticacidbacteriacultureisinvestigated.Furthermore,itisdemonstrated,howthescaleofthereactoritselfmayimpacttheculturecomposition,andwhichchal-lengescanarisewhentransferringmodelsfromlaboratoryscaletoanindustrialsetting.
Impact of dissolved oxygen tension on plasmid DNA production by Escherichia coli (P13)Karim Jaen1, Juan C. Sigala2, Roberto Olivares2, Alvaro R. Lara2
1 Universidad Autónoma Metropolitana-Cuajimalpa, Mexico City, Mexico, jaenkarim@gmail.com, Posgrado en Ciencias Naturales e Ingeniería; 2 Departamento de Procesos y Tecnología
DespitethegrowingpotentialofplasmidDNA(pDNA)asatherapeuticagent,littleisknownabouttheimpactoflarge-scaleconditionsonpDNAproduction.Thereareafewreportsontheeffectofdissolvedoxygentension(DOT)heterogenei-tiesonpDNAproduction,however,theresultsarenotclear.Inthepresentcontribution,weinvestigatedtheeffectofconstantaerobic(DOT=30%airsat.),microaerobic(DOT=3%)andoscillatoryoxygenavailabilityontheproductionofapDNAvaccinemodel(pVAX1)byE.coliW3110recA-in1Lbatchbioreactorcultures.DOToscillationswerecreatedbystirrerspeedshiftsfrom100to1200RPMevery10minthatcausedDOToscillationswithdecreasingamplitudefrom0-90%airsat.Oscillatoryconditionsdidnotaffectthespecificgrowthrate,butresultedinadecreaseofbiomassyieldonglucoseofca.24%,ahigheraccumulationoffermentationby-productsanda30%increaseoftherespiratoryquotient(RQ),comparedtoconstantmicroaerobicconditions.ThiscanbeattributedtothelowerO2uptakeundersuchconditions.DOToscillationscausedperiodicchangesoftherespira-toryactivity,asevidencedbyoff-gasanalyses.Interestingly,temperatureoscillationswerealsoobservedinresponsetoDOTchanges.pDNAyield,titerandproductivitywerenotsignificantlyinfluencedbythedifferentDOTconditions.Un-derconstantaerobicandmicroaerobicconditions,thepDNAsupercoiledfraction(SCF)remainedcloseto90%throughoutthecultures,andtheDNAfidelity(analyzedbyIlluminase-
quencingtechnology)wascloseto100%.Incontrast,underDOToscillations,SCFdroppedto82%andshowedmoreheterogeneityasrevealedbyagarosegelandchipmicrofluidicselectrophoresis.Overall,ourresultsshowthatdespitecellularbehaviorisstronglyinfluencedbytheDOT,pDNAproductionisaffectedmainlyonitsqualityattribute(SCF),ratherthaninproductivity,byDOThetero-geneities.
Increasing efficiency of baculovirus-based r-protein productions in orbitally shaken single-use bioreactors (P14)Ina Dittler1, Britta Badertscher1, Katharina Blaschczok1, Tim Bürgin2, Tibor Anderlei2, Peter Neubauer3, Dieter Eibl1 and Regine Eibl1
1 Zurich University of Applied Sciences, Wädenswil, Switzerland, ina.dittler@zhaw.ch, 2 Kühner AG, Birsfelden, Switzerland, 3 TU Berlin, Department of Biotechnology, Berlin, Germany
Insectcellsusedinconjunctionwiththebaculovirus-expressionvectorsystemprovideasuitableproductionsystemforthede-velopmentandmanufacturingofrecombinantproteinproductssuchasvirus-likeparticlevaccines.Inordertospeedupvac-cinedevelopmentandproduction,andtoreactmorequicklytopandemicscenarios,newstrategiesforefficientbaculovirus-basedproductiontechnologiesneedtobeinvestigated.Amongtheseistheimplementationofsingle-use(SU)bioreactors,whichhavebeenprovenascost-effective(48%lesscapitalcost),andready-to-usesystemswithoutthetime-intensivecleaningandsterilizationprocessesrequiredbystainlesssteelbioreactors.Furthertime-andcost-savingscanbeachievedbyusingthetiterlessinfected-cellpreservationandscale-up(TIPS)methodasanalternativetotheclassical2-phaseproductionprocess.TheTIPSmethodenablesalarge-scaleproteinexpressionin100Lbioreactorswithcryopre-served,baculovirus-infectedinsectcells(SpodopterafrugiperdaSf9)inonescale-upstep.Itsmajoradvantagesincludetheelimi-nationoftiteringvirusstocksupernatants,andthestablestorageofhighlyconcentratedrecombinantbaculovirusesenclosedintheinsectcells.Inthisstudy,acombinationoftheTIPSmethodwithSUbioreac-torswasinvestigatedwiththeaimofincreasingtheefficiencyofbaculovirus-basedproductionprocessesinorbitallyshakenbioreactorsfrommL-toL-scale.AftersuccessfulestablishmentoftheTIPSmethodbyinfectingSf9cellswiththerecombinantbacu-lovirusAutographacalifornicamulticapsidenucleopolyhedrovirus(AcMNPV),amodelprotein,thesecretedalkalinephosphatase(rSEAP),wasexpressedin125mLshakeflasks.Subsequently,theproductionprocesswastransferredto3LshakeflasksandtotheorbitallyshakenSUbioreactorSB10-X.TherSEAPactivitiesobtainedwerecomparableinallthreeorbitallyshakensystems(152–173UmL-1).Infact,theywereupto40%higherthanthosedescribedforstirredandwave-mixedSUbioreactorsusingtheclassical2-phaseproductionprocess.
Performance loss of Corynebacterium glutamicum cultivations under scale-down conditions (P15)Anja Lemoine1 , Michael Limberg2, Marco Oldiges2, Peter Neubauer1, Stefan Junne1
1 Technische Universität Berlin, Department of Biotechnology, Chair of Biopro-cess Engineering, D-13355 Berlin, Germany, anja.lemoine@tu-berlin.de, 2 Research Centre Jülich, Institute of Bio- and Geosciences, IBG-1: Biotechnology, 52425 Jülich
SeeL19
Poster abstracts
4th BioProScale Symposium 2016
25
Simulating large scale conditions in a scale-down bi-oreactor: Impacts on cell physiology of recombinant Escherichia coli (P16)Christian Reitz, Ping Lu, Stefan Junne, Peter Neubauer
Technische Universität Berlin, Department of Biotechnology, Chair of Biopro-cess Engineering, Berlin, Germany, Christian.reitz@tu-berlin.de
Escherichiacoliisanimportantproductionstrainforindustrialandpharmaceuticalrecombinantproteinproductioninlarge-scalefed-batchbioprocesses.Limitedmixingcapacitiesleadtothedevelopmentofgradientsconcerningthenutrientandoxygenavailabilitywhenacriticallevelinscaleandcelldensityisreached.E.colirespondstohighsubstrateconcentrationsincombinationwithoxygenlimitationwithanincreasedproductionofmetabo-litesbasedonpyruvateviaoverflowmetabolismandmixed-acidfermentation.Furthermore,anincreasedfluxintothebranched-chainaminoacidpathwaycanbeseenleadingtoanincreasedproductionalsoincludingnon-canonicalaminoacidslikenorva-line[].Non-canonicalaminoacidscanbeincorporatedintopro-teins,e.g.norvalineasasubstituteforleucine.Also,methionineisknowntobeexchangedbynorleucine[].Theseexchangesmayhaveeffectsontheproductionandqualityofarecombinantpro-tein.Tobetterunderstandtheimpactofheterogeneousoscillatingconditionsonthephysiologicalstateofthecellsandqualityoftheproducedrecombinantprotein,experimentsinascale-downmulti-compartmentreactorsystembasedon[]areperformedandinadditioncomparedtoglucose-pulsed-basedstirredtankreactorcultivations.Metabolomeanalysisfromsamplesofcultivationsofarecom-binantE.coliK-12W3110strainoverexpressingaleucine-richrecombinantproteinshowsanaccumulationofmetabolitesderiv-ingfrompyruvate(acetateandlactate)aswellasanincreasedproductionofnon-canonicalbranched-chainaminoacidsunderscale-downconditions.Furthermore,weseeincreasedincorpo-rationofthenon-canonicalaminoacidsnorvaline,norleucineandβ-methyl-norleucinewhenoscillatingcultivationconditionsareapplied,influencingthequalityoftherecombinantproduct.Theseresultshaveimplicationsforthedesignofmolecularandengineeringapproachesfordevelopingrobustprocessesaimingtowardsahighqualityofthedesiredproteinproduct.
[1]JSoinietal.AccumulationofaminoacidsderivingfrompyruvateinEscherichiacoliW3110duringfed-batchcultivationinatwo-compart-mentscale-downbioreactor“,AdvBiosciBiotechnol2(5)2011:336–39.
[2]ZIRandhawaetal.IncorporationofNorleucineatMethioninePositionsinRecombinantHumanMacrophageColonyStimulatingFactor(M-CSF,4-153)ExpressedinEscherichiacoli:StructuralAnalysis“,Biochem33(14):4352–62.
[3]SJunneetal.ATwo-CompartmentBioreactorSystemMadeofCom-mercialPartsforBioprocessScale-downStudies:ImpactofOscillationsonBacillusSubtilisFed-BatchCultivations.BiotechnolJ6(8):1009–17.
Bioprocess development
Tools enabling Real-Time Raman identification of biofilm fouling (P17)Martin Kögler1, Bifeng Zhang2, Li Cui2, Yunjie Shi3, Marjo Yliperttula1, Timo Laaksonen1, Tapani Viitala1, Kaisong Zhang2
1 Division of Biopharmaceutics and Pharmacokinetics, University of Helsin-ki, Helsinki, Finland, martin.kogler@helsinki.fi, 2 Key Laboratory of Urban Pollutant Conversion, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, P. R. China, 3 DSXDSYS-R&D-Design and Manufacturing Company, Xiamen, P. R. China
Thisposterdescribesaproof-of-conceptforacompactreal-timesurface-enhancedRamanspectroscopy(SERS)-onlinesensingapproachfordetectionofbacteriaindrinkingwatermembranefiltration.Inthisstudywecreatedacustom-designedflow-cellthatmimicsacross-flowmembranefiltrationsystem.Thisenablesonetomeasurechangesinsurface-foulants,suchasBrevundi-
monasdimiuta(BD)bacteriaandadenine,underconditionsthataresimilartoconventionalmembranefiltrationsystems.FormeasurementsweusedacommonportableRaman-spectrometerwithalaboratoryRaman-probemountedintoaspecificallydesignedflow-cellwithanopticalwindowandthelaser-beamfocusedontospeciallydevelopedgoldnanoparticle(AuNP)SERS-sensingareaonfilter-membranes.Thisallowedreal-timedetectionoflowconcentrationsofsurface-foulantsimmediatelyafterinoculationintoanultra-purewaterreservoirunderpressure-drivenfiltrationcondi-tionswithalimitofdetection(LOD)ataround10-7M.Wecomparedtheseonlineresultswithstaticmeasurementsfromanoffline,sample-takingapproach,usingaconfocalRaman-laboratory-microscope.Thedevelopedgoldnanoparticle(AuNP)SERS-sensing-areaonthemembranesprovedtobestableoveralongperiodofsurfacefoulinginvestigationsandtosuppressthestronginterferingRaman-signaloriginatingfromthecompositionlayerofmostfiltrationmembranes.
Bioprocess optimization using culture transfer in multi-reactor system (P18)Sten Erm, Kaarel Adamberg, Raivo Vilu
Competence Center of Food and Fermentation Technologies, Tallinn, Estonia, sten@tftak.eu
Thephysiologyofproducingorganismisknowntoaffecttheoutcomeofproductionprocesses,thusphysiologymustbewellcontrolledinoptimizationstudies.Continuouscultivationfixesthephysiologicalstateoftheorganism,butistimeandmaterialconsuming.Wehavedevelopedasystemcombiningthehighthroughputofbatchcultivationswiththecontrolledenvironmentofcontinuousones.Inthissystemmicroorgan-ismsarecultivatedinonebioreactorfollowedbyculturedistributioninanetworkofreceiverbioreactorsandcontinu-ationofindependentsteadystateexperimentstherein-ef-fectivelythephysiologicalstateofthecultureismultiplied.Inadditiontostartingexperimentsfrompreciselyestablishedphysiologythenovelsystemisverysuitablefortheproduc-tionofrecombinantproteinsorotherre-searchandpro-ductionapplicationswhichentailphysiologydisruption,asgrowthandproductionunitsaredecoupled.Incurrentstudytheefficiencyanalysisofthesystemispresented,proofofconceptexperimentaldataofrecombinantproteinproduc-tionoptimizationisincluded.
Bio-Butanol production with Clostridium aceto-butylicum in a multi-stage continuous process (P19)Peter Götz, Katja Karstens, Sergej Trippel
Biotechnologie, Beuth University of Applied Sciences, Berlin, Germany, goetz@beuth-hochschule.de
Finitenessoffossilhydrocarbonresourcesandenvironmentalpollutionresultingfromtheuseoftheseresourcesmakepro-ductionofrenewableenergysourcesnecessary.Wellknownproductsinthisareaareethanolandbutanol.Comparedtoethanol,butanolhashigherenergycontentandcanbeblendedwithgasoline.Inaddition,butanolactsasabulkrawmaterialinthechemicalindustry.TheOPTISOLVprojectconsortiumconsistsofindustrialandacademicpartnersfromItalyandGermany.Withintheframeworkoftheproject,newcontinuousprocessesforsolventproductionwithClostridiumacetobutylicumarebeingdeveloped.Experimentalinvestigationsupportedbymathematicalmodellingisusedtoevaluatefeasibleandproductivecontinuousbioreactorconfigurations.Whileinthewidelyusedbatchprocesses,thetwometabolicphasesofC.
26
acetobutylicum–acetogenesisandsolventogenesis–followeachotherintime,inamultistagecontinuousbioreactorsys-temtheyarespatiallyseparated.Achangeofmetabolicstatesthroughoutthemulti-stagebioreactorsystemallowstheircharacterization.Multistageprocessescanbeoperatedwithbothsuspendedcells(cellsaretravelingthroughreactorsandareexposedtoenvironmentalchanges)andimmobilizedcells(constantenvironmentofthecells).Inourgroup,wefocusonmodelingthedynamicsofmicro-bialmetabolicnetworkanditsregulationaswellasonthegenerationofdatafromcontinuouscultivationexperiments.Wealsoattemptoptimizingthebio-butanolproductionproc-essinacascadeofcontinuousstirredtankreactors(CCSTR)usingmodelpredictions.Severaloptimizationpossibilitiesareeligiblelikeanadding/removalofreactorstages,changeofresidencetimesinthestagesbyreactorvolumeadaptationandrecirculationorfeedingbetweenthestages.Currentlyoursystemachievesafinalbutanolconcentrationof11gBUTL-1withavolumetricproductionrateof1gBUT(Lh)-1.Mathematicalmodelingofthesuspendedcellssystemisstillachallenge.Theagent-basedmodel,developedinourgroup,includessimulationofchemicalconcentrationsineachbioreactorstageandbiologicalconversionratesofbiomassineachmetabolicstate.Agentsrepresentsubpopulationswithproperphysiologicalproperties.Movingtheseagentsaccordingtotherulesofresidencetimedistributions,thesesubpopulationsarechanginginsizeandaremovingtonewenvironments.Theperformanceofeachreactorstagewillthenbecalculatedfromthemeanmetabolicreactionsofthesubpopulations.
Improving mammalian cell culture process development by combining model-based simulations with DoE tools (P20)Johannes Möller1, Fabian Freiberger1, Regine Eibl2, Dieter Eibl2, Ralf Pörtner1
1 Institute of Bioprocess and Biosystems Engineering, Hamburg Univer-sity of Technology, Hamburg, Germany, johannes.moeller@tuhh.de, 2 Institute of Biotechnology, Biochemical Engineering and Cell Cultivation Technology, Zurich University of Applied Sciences, Wädenswil, Switzer-land
Bioprocessesinvolvingmammaliancellsaremostfrequentlybasedonoptimizedfeedingprofileswithinafed-batch.StatisticalDesignofExperiment(DoE)toolsarewidelyusedforthiskindofprocessoptimizationbecauseoftheirsimplestructureandeasyhandling.Limitationsoccurwithrespecttothehighnumberanddurationofexperimentsduringprocessdevelopment.Evenifhigh-throughputsystemscanperformthesenumbersofexperimentsinparallel,theheu-risticrestrictionofboundariesresultsinstepwiseiterationswithmultipleruns.Thisistime-consuming,cost-intensiveandfurthercomplicatesthepathfromprocessdevelopmenttoprocessestablishment.TheuseofDoEtoolsincombina-tionwithanappropriategrowthmodelmightbeavaluablemethodforevaluatingfed-batchstrategiesanddecreasingtheexperimentalspacerequiredforstatisticaloptimizationmethodsinsilicobeforeexperimentsarecarriedout.Toprovethisconcept,amodelwasusedtodescribethedynamicsofcellmetabolismofCHO-XM-111cellsinatwo-stepgrowthprocesswithmediumexchangefollowedbyafed-batch.Modelparameterswerefittedtoaverageddatafromthreeparallelshakeflaskcultivations,andmodelpredictionswereusedtodecreasetheexperimentalspaceandthenumberofcultivationparametersinsilico.Therefore,theconcentrationofL-glutamine(feed),constantfeedrate,pointintimeforthemediumexchange,andthestartoffeedingwereidentifiedascriticalprocessparameters.Atfirst,simulationswereappliedtominimizeexperimentalspace.Therefore,thepointintime
formediumexchangeandthestartingpointforfeedingweredeterminedfromabatchsimulationataminimalL-glutamineconcentrationof0.1mmolL-1.Secondly,toincreasethetotalcellnumberinafed-batch,theL-glutamineconcentrationandconstantfeedratewereoptimizedbysimulatedDoE.Intotal,twosimulatedanditerativeresponsesurfaceplotswereevaluated.Insteadofperformingmultipleexperimentstogeneratedataforresponsesurfaceplots,datagainedfromsimulationwasused.Finally,oneappropriatecombinationofcultivationparameterswastestedandcomparedwiththeinitialmodel.Themethodshownissuitableforthegenerationofdeeperproc-essunderstanding,e.g.thelinkageofdifferentprocessparameterstoqualityattributes.Furthermore,cultivationstrategiesformam-maliancelllinescanbecomparedandevaluated.Thisresultedinasignificantreductioninthenumberofexperimentsrequiredduringprocessdevelopmentandprocessestablishment.
Advanced pre-cultivation method for recombinant Escherichia coli cells (P21)Basant Ali1, Antje Neubauer2, Antti Vasala2 and Peter Neubauer1
1 Chair of Bioprocess Engineering, Technische Universität Berlin, Berlin, Germa-ny, basantali@mailbox.tu-berlin.de, 2 BioSilta Oy, Oulu, Finland
E.coliculturesarewidelyusedforrecombinantproteinproduc-tion.Cultivationconditionsplayanimportantroleinfinalproteinyieldandcorrectproteinfolding.Thequalityofthestartercultureaffectstherecombinantproteinproductionprocess,howevertheredonotexistanystandardsforcontrolledpre-cultures.TheEnBasetechnology,basedonfed-batchprinciple,iswidelyappliedforenhancedrecombinantproteinproduction.Hereweinvesti-gate;whetherEnBasebasedEnPressomediacouldalsoprovidebenefitsforstartercultures.Thisstudy,attheexampleofaprocessforproductionofathermophilicnucleosidephosphorylaseinE.colishowsthatEnPressoBmediumbasedcultures,whichcanbedirectlyinoculatedfromafrozenstockculture,providearepro-ducibleandrobuststartercultureforproteinproductionproc-esses.
Enhanced plasmid DNA production by enzyme con-trolled glucose release and an engineered Escheri-chia coli (P22)Alvaro R. Lara1, Karim E. Jaén2
1 Universidad Autónoma Metropolitana-Cuajimalpa (UAM), México, alara@correo.cua.uam.mx, 2 Posgrado en Ciencias Naturales e Ingeniería, Universi-dad Autónoma Metropolitana-Cuajimalpa Ciudad de México, Mexiko
Toevaluatethecombinationofaculturemediumemployingglu-coamylase-mediatedglucosereleasefromagluco-polysaccharideandanE.colistrainengineeredinitsglucosetransportsystemforimprovingplasmidDNA(pDNA)production.TheproductionofpDNAwastestedusingE.coliDH5growninshake-flasksandtherecentlydevelopedVH33D(recAdeoR)-en-gineeredstrain,whichutilizesglucosemoreefficientlythanwildtypestrains.Threeglucoamylaseconcentrationsforreleasingglu-cosefromthepolysaccharidecarbonsourcewereused:1,2and3U/L.Bothstrainsreachedsimilarcelldensitiesrangingfrom5to8.8g/Lunderthedifferentconditions.ThehighestpDNAyieldsonbiomass(YpDNA/X)forbothstrainswereobtainedwhen3Uenzyme/Lwereused.Undertheseconditions,35±3mgofpDNAl-1wereproducedbyDH5aafter24hofculture.Underthesameconditions,theengineeredstrainproduced66±1mgpDNA/Lafter20h.pDNAsupercoiledfractionswerecloseto80%forbothstrains.ThepDNAconcentrationachievedbytheengineeredE.coliwas89%higherthanthatofDH5a.Thecombinationoftheengi-neeredstrainandenzyme-controlledglucosereleaseisanattrac-tivealternativeforpDNAproductioninshake-flasks.
Poster abstracts
4th BioProScale Symposium 2016
27
Enzymatic production of nucleotide analogues for click chemistry-based applications (P23)Maryke Fehlau, Birthe Rausch, Anke Wagner, Peter Neubauer
BioNukleo UG and Department of Bioprocess Engineering, Technische Univer-sität Berlin, Berlin, Germany, maryke.fehlau@bionukleo.de
Theuseofenzymesasefficientandsustainablebiocatalystsisespeciallyinterestinginthosefieldsofindustrialbiotechnologythatarecurrentlydominatedbychemicalproductionprocesses.Oneofthosefieldsistheareaofmolecularbiologywheremodi-fiednucleotidesareincreasinglyusedforlabelingofPCRproductsandtheproductionoffluorescentprobesforfluorescenceinsituhybridization(FISH)applications.Recentlytheapplicationofcop-percatalyzedclickchemistryisgainingmoreandmoreinterestforthoseapplicationsusing5-ethynyldUTP,C8-alkyne-dUTPorC8-alkyne-dCTPassubstrates.Inthepresentedstudyweareshowingtheenzymaticproductionofphosphorylatedbuildingblocksusing5-ethynyl-deoxyuridineasamodelcompound.Theheterologousexpressionofallinvolvedenzymeswasoptimized.Finally,thebiocatalystsweresuccessfullyproducedashis-orGST-taggedproteinsinLBorEn-Presso®1mediumandpurifiedbyaffinitychromatography.Theproductionof5-EthynyldUMPwasachievedbyusingdeoxy-nucleotidekinase2withconversionratesbetween50and100%.Theactivitiesofguanylatekinase(GUK),CMP-UMPkinase(CMPK)andapyruvatekinase(PK)werealreadyshownwithnaturalsub-strates.NDPsandNTPswereformedwithyieldsof30to70%.Theactivitiestowards5-EthynyldUMPremaintobeinvestigatedforthoseenzymes.
[1]UkkonenKetal.2011.High-yieldproductionofbiologicallyactiverecombinantproteininshakeflaskculturebycombinationofenzyme-basedglucosedeliveryandincreasedoxygentransfer.MicrobCellFact10,107.
[2]SerraIetal.2014.ImmobilizedDrosophilamelanogasterDeoxyribo-nucleosideKinase(DmdNK)asaHighPerformingBiocatalystfortheSynthesisofPurineArabinonucleotides.Adv.Synth.Catal.356,563-570.
Biological activity of halogenated cladribine and fludarabine derivatives (P24)Heba AbdElRahman1,2, Maryke Fehlau1,3, Anke Wagner1,3, Peter Neubauer1
1 Technische Universität Berlin, Berlin, Germany, 2 National Research Centre NRC, Cairo, Egypt, hebayehya@hotmail.com, 3 BioNukleo UG, Berlin, Germany
Yearsaftertheapprovalofcytarabineandedoxudineasanti-cancerandantiviralagents,respectively,in1969,nucleosidesanaloguesstillholdtheirplaceasanimportantclassofmedicinalcompounds.Thedevelopmentofnewmoleculesisfocusedontheimprovementofdrugpropertieswithregardto:longtermtoxicity,resistanceandbioavailability.Inthepresentproject,itistheobjectivetofindthebestconditionsfortheproductionofatleast25differentpurinenucleosideanaloguesusingthermophilicenzymesinthescaleof100mgwithpurity≥99%.Thestructuresareconfirmedusingdifferentspectroscopicanalyses(MS/NMR).Finallydifferentbiologicaltestswillbecarriedout.Duringtheinvitrobiologicaltests,it’sintendedtoanalyzethevirostaticactivityandtoperformcytotoxicassaysaccountingforbothsolidtumorsandhematologicmalignancies.
Reconstructing the biosynthetic machinery of Ruminococcin A, a lanthipeptide identified from a strictly anaerobic Gram positive bacteria, in Escherichia coli (P25)Ongey Elvis Legala1, Michel Fons2, Peter Neubauer1
1 Chair of Bioprocess Engineering, Department of Biotechnology, Tech-nische Universität Berlin, Berlin, Germany, elvis.ongey2@gmail.com, 2 ISM2/BiosCiences UMR CNRS 6263, Interactions Moléculaires Microbiote – Muqueuse Intestinale, Service 342, Faculté des Sciences et Techniques de St Jérôme, Université Paul Cézanne Aix-Marseille III, Marseille, France
RuminococcinA(RumA)isatrypsin-dependentlanthipetidethatisproducedbythestrictanaerobicGram-positivebacteriumRuminococcusgnavusE1,isolatedfromhumanintestinalmicrobiome.ThebiosynthesispathwayofRumAin-volves,amongstothers,themodifyingenzyme(RumM)whichcatalyticallyinstallsthreethioethercrosslinksandanα,β-unsaturatedamino,didehydrobutyrineintoitscorestructure.Theenhancedactivityofthepeptideagainstpathogenicclostridiafurthermakesitaplausibletargetforthetreat-mentofhumaninfectionsandapplicationsinlivestock.DuetothenatureofR.gnavusE1,standardcultivationstrategiesareobviouslynotoptimalforlarge-scaleproductionandsubsequentpharmaceuticalcommercializationofRumA.Inthiswork,weaimedatengineeringthebiosyntheticrouteofRumAinE.colitoenhancethedevelopmentofabioprocessschemeforaviableindustrialproduction.Weisolatedthegenesforthestructuralpeptide(rumA)andthemodifyingenzyme(rumM)fromR.gnavusE1andco-expressedtheminE.coliunderseparateregulatorymechanismsinordertobalanceexpressioninawaytoguaranteeaminimalloadforthecell.Wealsoinvestigatedthebehavioroftheengineeredstrainwithrespecttothetoxicityofthetargetproducttothehost.OurresultsshowthatactiveRumAisproducedinE.coliwhichpresentsaverygoodbasisforscale-upoptimization.Toourknowledge,thisisthefirsttimethatalanthipeptidefromastrictlyanaerobicGram-positivemicrobeisheterologouslyexpressedinaGram-negativehost.
Platform technologies for automated bioprocess development (P26)Florian Glauche, Peter Neubauer
Technische Universität Berlin, Chair of Bioprocess Engineering, Berlin, Germany, florian.glauche@tu-berlin.de
Bioprocessdevelopmentistimeandcostintensiveduetothelargedesignspacetobeinvestigated.Laboratoryautomationcombinedwithexperimentaldesignstrategiesoffernewwaystoincreaseexperimentalthroughput.However,oftenkeypa-rameterssuchareselectedfromscreeningexperimentswhichdonotresembletheconditionsatproductionscale.AtthechairofBioprocessEngineeringoftheTUBerlin,alaboratoryofthefutureconcepthasbeenestablishedwhichfeaturesautomatedhigh-throughputcultivationsystemscombinedwithfed-batchgrowthmedia,chemoopticalsensorsandamodulardatabasefordataevaluationandmodelling.Theplatformtechnologieswereappliedforofrecombinantpro-teinproductionprocessdevelopmentfrommicrobialhosts.
28
Novel bioprocess types based on CHO lysate cell-free protein synthesis systems (P27)Lena Thoring1, 2, Doreen A. Wuestenhagen1 and Stefan Kubick1
1 Fraunhofer Institute for Cell Therapy and Immunology (IZI), Branch Bioanalytics and Bioprocesses Potsdam-Golm (IZI-BB), Department of Cell-free and Cell-based Bioproduction, Potsdam, Germany, 2 Technical University of Berlin, Department of Biotechnology, Berlin, Germany
Nowadays,differenttypesofbioprocessesareavailablefortheproductionofpharmaceuticallyrelevanttargetpro-teins.Ingeneral,hostorganismsusedduringfermentationprocessescanbedividedintofourgroupsthatcompriseE.coli,yeast,insectcellsandmammaliancells.Theselectionofsuitableorganismsdependsontheproteinproductionstructureofthedesiredproteinproductprocessitself.Humanproteinsinparticularplayanessentialroleintherapeuticalapplications.Duetotheircomplexstructure,posttranslationalmodificationsandelaboratefoldingprocedures,thesetypesofproteinsneedspecialproductionrequirementsandsyn-thesismilieus.Thereforethemostfavorablesystemsfortheproductionoffunctionallyactivehumanproteinarebasedonculturedmammaliancells.Currently,ChineseHamsterOvary(CHO)cellsareoneofthemostimportantandwell-knownmammaliancelllinesusedforindustrialproduction.CHOcellfermentationprocesseswerealreadyimplementedinindustrialproductionduringthe1980s.Overthelastdecades,thesystemwascontinuouslydevelopedandprovidesvariouspossibilitiesthatcouldbeadaptedforspecificapplications.Besidestheconventionalinvivoproteinproduction,alterna-tivebioprocessmodes,socalledcell-freeproteinsynthesisplatformsweredeveloped.Cell-freeproteinsynthesissystemsenableproteinproductioninanopenmannerusingcelllysates.Thisprocedureprovidestheadvantagethatmilieuconditionscanbeeasilyadaptedtoeachtypeofprotein,byforexampleadjustingredoxconditionsoradditionofchaper-ons.Additionally,theproteinproductionprocessinthecell-freesystemisfastercomparetoinvivoprocesses.Cell-freesystemsharbortheentirefunctionallyactivetranslationalma-chinery,includingribosomes,translationfactorsandtRNAs.Prokaryoticcell-freesystemsenablehighproteinproductionrates,buttheyarelimitedinproteinfoldingandmodifica-tions.Forhumantargetproteinsandsocalled“difficult-to-ex-press”proteins,cell-freesystemsareavailablethatarebasedoneukaryoticlysates.Eukaryoticcell-freesystemsderivedfromSf21cellsandmammaliancelllinesharborendogenousmicrosomalstructuresduetoamildcelldisruptionandlysatepreparationprocedure.Endogenousmicrosomes,obtainedfromtheendoplasmaticreticulum,enabletoimplementpost-translationalmodificationsanddirectlyintegratemembraneproteinsintotheirappropriatelipidenvironment.Toclosethegapbetweenconventionallabscalecell-freepro-teinsynthesisandinvivoproteinproductionprocesses,wearedevelopingcell-freeproductionprocessesbasedonCHOcelllysates.CHOcell-freeproteinsynthesisisimplementedintworeactionsmodeswhichdistinguishthemselvesbyreac-tionconditions,proteinyieldandprocesscosts.The“simply-to-perform”batchmode,aonepotreaction,providesproteinyieldsupto50µg/ml.Synthesisofvariousproteinscouldbedemonstratedinbatchmodeincludingintegralmembraneproteins,e.g.GPCRs,ionchannelsandglycoproteinsaswellascytosolicproteins.Additionally,thepossibilitytoperformcell-freesynthesisinthepresenceofnon-canonicalaminoacidwasshownbyincorporationintotargetprotein.Inthismanner,newopportunitieswillbeprovidedthatenableforexamplespecificfluorescencelabelingofproteins.Besidesthementionedbatchmodeanovel,dialysisprocessmodeisdevelopedforcell-freeproteinsynthesis.Acompartmentali-zationoftheprocessreactionintoareactionchamberandanadditionalfeedingchamberleadstoaprolongedreaction
time,therebyincreasingthetotalproteinyieldbyremovinginhibi-torybyproductsanddeliveringenergycomponentsalongwiththediffusiongradient.Differentprocessoptimizationstrategiesforthedialysismodeenhanceproteinyieldsupto10timesyieldscomparedtobatchbasedsynthesis.Withregardtothepresentedresultsourtechnologyoffersanenormouspotentialintermsofanovelproductionprocessfor“difficulttoexpress”proteinsandtheirsubsequentfunctionalanalysis.
Enhancement of lipopeptide yield by integrated bioprocess development (P28)Julia Glazyrina1, P. Neubauer1, P. Götz2
1 Technical University of Berlin, Institute of Biotechnology, Department of Bioprocess Engineering, Berlin, Germany, julia.glazyrina@tu-berlin.de. 2 Beuth University of Applied Sciences Berlin, Bioprocess Engineering, Berlin, Germany
Lipopeptidearesurfaceactivecompoundsandcanbeusedasbio-surfactantsinmanyindustrialareasincludingallapplicationfieldsofchemicalsurfactantssuchasagriculture,foodproductionandcosmeticsandalsoinspecialfieldslikebioremediation.Moreover,theyhaveantibioticactivitythatallowstousethelipopeptidesasplantprotectionagents.Currentlyrelativelylowyieldandhigh-costdownstreamprocessingmaketheproductionprocessoflipopeptideeconomicallyinefficient.Inthisprojectabioprocessforproductionofthelipopeptidfengycin,startingfromstrainscreening,viaoptimizationofmediaandcultivationconditionstodownstreamprocessing,usinganin-situproductremoval(ISPR)strategybyfoamfractionation,wasdeveloped.Itwasshownthatusingacost-effectivecultivationmediumbasedonpotatofruitjuice(side–productfromstarchproduction),opti-mizedcultivationconditionssuchaslowtemperatureandoxygen-limitationandalsobyapplyingtheintegratedfoamfractionationsystem,afive-foldincreaseoffengycinyieldwasachieved.Attheendofthebioprocessdevelopment,concentrationoffengycininthecultivationmediaofmorethan1gL-1couldbeobtained.
Docosahexaenoic acid (DHA) production of new strain of Indonesian Thraustochytrid Aurantiochytrium sp. LR52 in single use TubeSpin 600 system (P29)Andri Hutari1,2, Wahyu Hidayati3, Apon Zaenal Mustopa4, Peter Neubauer1
1 Technische Universität Berlin, Department of Biotechnology, Chair of Biopro-cess Engineering, Berlin Germany, andrihutari@gmail.com, 2 University of Muhammadiyah Prof. Dr. Hamka, Biology, Jakarta, Indonesia, 3 University of Muhammadiyah Prof. Dr. Hamka, Pharmacy, Jakarta, Indonesia, 4 Indonesian Institute of Science (LIPI), Research Center for Biotechnology, Bogor, Indonesia
ThraustochytridsAurantiochytriumsp.areuniquemarinemicro-organismintermofdocosahexaenoicacid(DHA)productionbe-causetheygrowfast,producehighcelldensity,highproportionsoflipidandDHAcontent.DHAhasalreadyprovenasimportantfattyacidforinfantdevelopment,maintenanceofnormalbrainfunction,visualacuity,andpreventionofneurologicaldisordersandinflammatorydiseasesinadult.TheapplicationofDHA-pro-ducingAurantiochytriumisrapidlyexpandingforfoodadditive.Inaddition,driedAurantiochytriumhasGRASstatusforuseasfeedtobroilerchickens,layinghens,andaquaculture.ThegenusAuran-tiochytriumsp.(before2007isSchizochytrium)belongingtotheLabyrinthulomycetesofthekingdomChromophyta(Straminipila).Aurantiochytriumsp.producedoverallDHAproductivitymorethantwicehigherthananotherestablishedspeciesfromdino-flagellateCrypthecodiniumcohniiwhereasthecultivationtimeofAurantiochytriumwasshorterthanthatofC.cohnii.Inaddition,AurantiochytriumalsoshowedalesserincidenceofunwantedmicrobialcontaminationthanoccurredinvariousCrypthecodin-iumcohniifermentations.ThishasattractedmanyresearchersandindustrytoscreennewstrainofAurantiochytriumandoptimize
Poster abstracts
4th BioProScale Symposium 2016
29
thefermentationperformance.Recently,wehaveisolatedastrainofAurantiochytriumfromIndonesianmangrovehabitat,namelyLR52.CultivationofAurantiochytriumsp.LR52inTubeSpin600withlowartificialsaltwater(ASW)concentrationof3.6gL-1yieldeddriedbiomassandDHAcontentmorethan63and18gL-1,respectivelyin6days.TheDHAproductionrateofLR52isoneofthehighestvaluereportedforanyDHAproducingmicroorgan-ismsinshakeflaskscale.
Process analytical technologies
Spectroscopic bioprocess monitoring and control of CHO cell cultivations (P30)Jens Claßen, Shaobin Lu, Marko Sandor, Dörte Solle, Thomas Scheper
Leibniz Universität Hannover/Institut für Technische Chemie, Hannover, Ger-many, classen@iftc.uni-hannover.de
Processanalyticaltechnology(PAT),aninitiativeoftheAmericanFoodandDrugAdministration(FDA),providestoolsforholistic,knowledge-basedandriskavoidingcontrolofmanufacturingprocessesinchemicalandpharmaceuticalindustry.Onetoolisonlinemonitoringofcriticalprocessparameters(CPP)byspectro-scopicmethodscombinedwithmultivariatedataanalysis[1,2].Thisisparticularlyimportantforcost-intensivemanufacturingofbiopharmaceuticalstoenableanearlyprocessfaultdetectionandrealtimeproductrelease.Mammaliancellcultivation(CHO-K1cellline)in15lstirredtankreactorischosenasabenchmarkprocessformonitoringandcontrol.MonitoringofapoptosisiscarriedoutbyUV/Visspectros-copy.Bynon-invasiveMIRspectroscopysubstrateandmetaboliteconcentrations(glucoseandlactate)couldbemonitoredduringthecultivation.Theresultsofspectroscopicinlinemeasurementandconsequentialdevelopedpartialleastsquaresregressionmodels(PLS)areshownexemplarily.Basedontheonlineglucosepredictionanautomatedfeedstrategyisbuilt-ontopreventanearlycelldeathbysubstratelimitation.Theprocessstrategyischargedfrombatchtofed-batchprocessbythisapplicationandhighercellcountswererealized.[1]SandorMetal.Comparativestudyofnon-invasivemonitoringvia
infraredspectroscopyformammaliancellcultivations,JBiotechnol.,2013,168,636–645.
[2]BiechelePetal.Sensorsystemforbioprocessmonitoring,EngLifeSci,2015,00,1-20
Mathematical model of adherent vero cell growth and poliovirus production in animal component free medium (P31)Gerco van Eikenhorst, R. V. Ursache, Y. E. Thomassen, W. A. M. Bakker
Institute for Translational Vaccinology, Process Development, Bilthoven, the Netherlands, gerco.van.eikenhorst@intravacc.nl
Sabin–IPV(orsIPV,inactivatedpoliovaccinebasedonattenuatedSabinstrains)isanticipatedtoreplacetheoralpoliovaccine(OPV)fortheendgameinpolioeradication[1,2].OptimizationofsIPVproductionwillleadtoabettereconomicallyfeasiblevaccine[3].TheaimofourstudywastodevelopadescriptivemathematicalmodelabletocapturedynamicsofadherentVerocellgrowthandpoliovirusinfectionkineticsinanimalcomponentfreemediumtofurthersupportsIPVprocessoptimizations.TobeabletomonitorVerocellgrowthclosely,theexperimentalbioreactorsetupincludedaFogalebiomassprobeandacellfreesamplingdeviceforonlinemetaboliteanalysesusingaNovaBioProfileFLEX.Onlinebiomassdatacorrespondedwellwiththeofflinemeasuredcellconcentrations,alsotheon-andofflinemainmetabolites(glucose,lactate,glutamine,glutamate,ammonia)werecomparable(Fig1).Thedevelopedmodelpredictsthecelldensity,metabolitesprofilesandviralyieldsintime.Theviralgrowthmodelimpliedthatthemultiplicityofinfection
(MOI)andthetimeofinfection(TOI)donotaffectmaximalpoliovirusyields.However,theydoaffecttheprocesstime,whichmaybereducedbyselectingalowTOIandahighMOI.Additionally,weobservedacorrelationbetweenviraltitersandD-antigen,ameasureforimmunogenicity,ofSabinpoliovirustype1.Thedevelopedmodelisadequateforfurtherstudiesofthecellmetabolismandinfectionkineticsandmaybeusedtoidentifycontrolstrategiestoincreaseviralproductivity.Increasedviralyieldscanpotentiallyreducethecostsofpoliovaccineswithcouldhavelargeimplicationsonpublichealth.Inaddition,thebiomassprobeandtheonlinemetaboliteana-lyzermaybeusedincombinationwiththemodeltomonitorthecellandvirusgrowthinaprocessanalyticalapplication(PAT).
Multiparameter monitoring of gradients in the liquid phase of fermentation processes (P32)A. Bockisch1, J. Biering2, S. Sachse3, W. Vonau3, S. Junne1, P. Neubauer1
1Technische Universität Berlin, Department of Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, anika.bockisch@tu-berlin.de, 2Versuchs- und Lehranstalt für Brauerei in Berlin e.V., Berlin, Germany, 3Kurt-Schwabe-Institut Meinsberg e.V., Meinsberg, Germany
SeeL23
Multiposition Sampling, Multiparameter Monitoring and Electrooptical Tracking of Microbial Activity: New Tools for better understanding biogas production (P33 )Jörn Beheim-Schwarzbach*1,Erich Kielhorn2, Alexander Angersbach3, Emma Ritzi1, Manja Laqua2, Anna Vaupel2, Boris Habermann1, Peter Neubauer2, Stefan Junne2
1Institute of Agricultural and Urban Ecological Projects, Berlin, Germany, joern.beheim-schwarzbach@iasp.hu-berlin.de, 2 Department of Biopro-cess Engineering, Technical University of Berlin, Berlin, Germany, erich.kielhorn@tu-berlin.de, 3 EloSystems GbR, Berlin, Germany
SeeL34
Online-monitoring tools in batch and (semi-)perfusion cultivations for vaccine production (P34)Daniel Vazquez-Ramirez1, Alexander Nikolay1, Yvonne Genzel1, Udo Reichl1,2
1 Max Planck Institute for Dynamics of Complex Technical Systems, Sandtorstraße 1, 30106 Magdeburg, Germany, vazquez_ramirez@mpi-magdeburg.mpg.de, 2 Otto-von-Guericke University Magdeburg, Chair of Bioprocess Engineering, Universitätsplatz 2, 30106 Magdeburg, Germany
Theincreasingavailabilityofonline-monitoringtoolsenablesabetterunderstandingandcontrolofanimalcellculture-basedprocesses.However,real-timeprocessmonitoringofcellconcentrations,nutrientsandotherparametersisstillbeinglimitedinlarge-scalevaccinemanufacturing.Neverthe-less,forfurtherintensificationofvaccineproduction,i.e.theestablishmentofhigh-cell-density(HCD)perfusionprocesses,online-monitoringisofcrucialimportance.Inthepresentwork,thesuitabilityofdevicesforonline-monitoringofpHanddissolvedoxygenconcentrationinshakeflasks,aswellasofglucose,lactateandcellconcentrationsin1Lbioreactorsforvirusproductionprocesseswastested.Highcellconcentrations(~50×106cells/mL)wereachievedeitherthroughmediumrenewalbyperiodiccentrifugation(semi-perfusion)in125and250mLshakeflasksorbyusingahollowfiber-based(ATForTFF)perfusionsystemfor1Lbiore-actors.Forshakeflaskcultivations,oxygenconcentrationandpHvalueweremonitoredduringcellproliferationandMVAviruspropagationinsuspensionAGE1.CR.pIXcellsusingthe
30
SFR®system(Presens).Inbioreactors,theconcentrationsofglucoseandlactateweremeasuredonlineduringperfusioncultivationofAGE1.CR.pIXandbatchcultivationofBHK-21cellsusingaBioPAT®Traceprobe(Sartorius).Additionally,viableandtotalcelldensitiesweremonitoredonlineusingtheIncyteandDencyteeprobes(Hamilton),respectively,duringthepropagationofMVAandinfluenzavirusA/PR/8/34(H1N1)inAGE1.CR.pIXcells,andyellowfevervirus(YFV17D)inadherentVeroandBHK-21cells.TheSFR®systemshowedthatpHvaluesof7.2to7.4anddis-solvedoxygenconcentrationsabove86%canbemaintainedwhilecultivatingAGE1.CR.pIXcellsupto45×106cells/mLinshakeflasksunderasemi-perfusionregime.Interestingly,areducedoxygenuptakerateofthecellscouldbeobservedwhendelayingthemediumrenewal.Thisseemedtocorrelatewiththebeginningoflactateconsumptionusuallyobservedinbatchprocesseswhenglucoseconcentrationsdropbelow~2.5g/Landlactateconcentrationsincreaseabove~2.0g/L.Duringthebioreactorcultivations,theBioPAT®TracereadoutscorrelatedwellwithofflineglucoseandlactatemeasurementsfromaBioProfile®100Plusandallowedperiodicadjustmentsofperfusionratestokeepoptimalglucoseconcentrations.However,particularlyduringBHK-21cultivations,otherkeymetabolitesratherthanglucosealsoseemedtoplayanim-portantroleforoptimalcellgrowth.Concerningthetotalandviablebiomass,online-measure-mentsobtainedwiththeDencyteeandIncyteprobesduringthecellgrowthphaseofAGE1.CR.pIX,BHK-21andVerocellscorrelatedfairlywellwithofflineanalysescarriedoutwithaVi-CELL®system(BeckmanCoulter).Asexpected,theonlinemonitoringofcellconcentrationwasproblematicforcultiva-tionsofBHK-21cellsathighercellconcentrationsandafterinfectionwithYFVduetochangesincellphysiologyaswellascellmorphology.Towardslate-stageinfectiontimepoints,measurementswerenotreliable.However,duringMVAviruspropagationinAGE1.CR.pIXcells,acorrelationwithofflineanalysiswasobservedevenatlate-stageinfectiontimes.Here,celllysisduringviruspropagationcouldbefollowedbythefrequencyandcapacitancereadoutsoftheIncyteprobe.Inaddition,itwasobservedthatstepwiseadditionoflargeamountsofbasetoadjustthepHvalue,andtheincreaseingasspargingcouldinterferewiththeestimationoftheviableandthetotalcellconcentrations,respectively.TheSFR®offersreliableandvaluableinformationofpHvaluesanddissolvedoxygenconcentrationsforcharacterizationofcelllinesinshakeflasks.AlthoughtheBioPAT®measurementscouldbeusedformanualperfusioncontrolbasedonglucoseconcentrations,additionalspecificnutrientrequirementsofeachcelllineshouldbetakenintoaccountandmonitoredifnecessary.OnlinedataobtainedfromtheIncyteandDen-cyteesensorscanbeusefultomonitorcultivations,andformanualandautomaticcontrolofperfusioncultures.TheIncytealsoenablesfrequencyandcapacitancemeasurementsthatmaysupportprocesscontrolduringvirusproduction.
A PAT tool for real-time determination of spore quality in filamentous fungi bioprocesses (P35)Daniela Ehgartner, Jens Fricke, Christoph Herwig
Institut für Verfahrenstechnik, Umwelttechnik und Techn. Biowissen-schaften, Vienna University of Technology, Vienna, Austria, daniela.ehgartner@tuwien.ac.at
Sporeinoculumqualityinfilamentousbioprocessesisacriticalparameterassociatedwithviablesporeconcentrationandsporegermination[1].Itinfluencespelletmorphologyand,consequently,processperformance[2].Anessentialstepbeforeinoculationisthedeterminationoftheviablesporeconcentration,inordertoapplyqualitycontrolanddecreasebatch-to-batchvariability.Thestate-of-the-artmethodto
investigatethisvariableistedious,associatedwithsignificantinherentbias,andnotapplicableinrealtime.Therefore,itisnotusableasaprocessanalyticaltechnology(PAT).PATaimstodesign,analyseandcontrolpharmaceuticalproductioninordertoensureproductquality[3].Monitoringofsporegerminationisnotonlynecessarytoinvesti-gatefurthersporequalityattributesbutcanalsobeusedasvalida-tionofthebeforehandmentionedmethod.Qualityattributesconnectedtosporegerminationarenotonlytheamountofger-minatingsporesbutalsohowlongsporesneedforgerminationandwhetherallsporesgerminateatthesametimeornot.Thoseelementsweresofarmonitoredusingimageanalysis[4],whichishamperedbycomplexmediumbackgroundoftenobservedinfilamentousbioprocesses[5].Themethodpresentedhereisbasedonthecombinationofvi-abilitystainingandlarge-particleflowcytometrywhichenablesmeasurementsinreal-timeandhenceaimstobeapplicableasPATtool.Itiscompatiblewiththecomplexmediumbackground,andallowsthequantificationofmetabolicallyactivespores.Further-more,sporeswellingasapreviousstepofsporegerminationandgerminationitselfmaybemonitored.Adistinctionofgerminatedsporesfromnotgerminatedsporeswasbasedonalogisticregres-sionusingmultiparametericdatafromforwardscatterandgreenfluorescencechannel.Inanindustrialbioprocesswithfilamentousfungi,agoodcorrela-tiontoCFUcountswasfound.Themorphologicalparametersassporeswellingandsporegerminationaswellasmetabolicactivitywerefollowedovertheinitialprocessphasewithclosetemporalresolution.Thevalidationofthemethodtomeasuresporegermi-nationshowedanerrorofsporeclassificationoflessthan5%.Differencesinsporegerminationforvarioussporeinoculaagesandsporeinoculumconcentrationsweremonitored.Notallsporesshowingmetabolicactivitygerminated.Theamountofsporesgerminatedwasfoundtobedependentonthequalityofthesporeinoculum.Thiscombinationofviabilitystainingandflowcytometryisapromisingtooltonotonlydeterminesporeinoculumqualitybe-forebatchinoculuationbutalsotoat-linemonitorsporeswellingandsporegerminationinordertoreactinreal-timeondeviationsandhencedecreasebatch-to-batchvariability.ThereforethemethodcouldbeapplicablefortheimplementationasPATtoolinfilamentousbioprocesses[6].
[1]J.Nielsen&P.Krabben,Biotechnol.Bioeng.1995,46(6),588-598.[2]K.G.Tucker&C.R.Thomas,Biotechnol.Techn.1994,8(3),153-156.[3]FoodandDrugAdministration,DHHS.2004,Rockville.[4]P.W.Coxetal.,Microbiol.1998,144(4),817-827.[5]A.E.Poschetal.,Microb.cellfact.2012,11(1),1-14.[6]D.Ehgartneretal.,Fung.Gen.Biol.2016,Manuscriptinpreparation.
Scalability of cultivations in the 2-dimensional rocking single-use bioreactor CELL-tainer (P36)Anna Maria Marbà-Ardébol, Peter Neubauer, Stefan Junne
Technische Universität Berlin, Institute for Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, a.marbaardebol@campus.tu-berlin.de
SeeL31
Poster abstracts
4th BioProScale Symposium 2016
31
Analysis of the single-cell size distribution as a pro-cess analytical tool in bioprocesses (P37)Anna Maria Marbà-Ardébol1, J. Emmerich2, S. Junne1, P. Neubauer1
1Technische Universität Berlin, Institute for Biotechnology, Chair of Bioprocess Engineering, Berlin, Germany, a.marbaardebol@campus.tu-berlin.de, 2 SOPAT GmbH, Berlin, Germany
Atlineandonlinemicroscopyisapromisingtechniquetogainvaluabledataaboutpopulationheterogeneityandevensubstrateconversionandproductsynthesis,asthesestepshaveanimpactonthecellsizeinvariousbioprocesses,e.g.cellulargrowth,intra-cellularlipidaccumulation,etc.Inthisstudy,real-timemonitoringofthesinglecellsizedistributionisinvestigatedforthequantifica-tionofintracellularaccumulationofthepolyunsaturatedfattyaciddocosahexaenoicacidintheheterotrophicmicroalgaeCrypte-codiniumcohnii.Thenovelphoto-opticalmicroscopyprobeofSOPATwasappliedinsituandtheperformancewascomparedtoatline3-dimensionalholographicinterferometricmicroscopy(Ovizio),whichprovidedalsodataaboutthesurfacestructure,andtraditionalflowcytometry.Itwasproventhatboth,theatlineandtheinsitumicroscopy,arewell-suitedforthepredictionofthefattyacidaccumulationincells,andthuscanbeappliedasaprocessanalyticaltool.FurtherstudiesinvestigatedtheimpactofoscillatingcultivationconditionsonthepopulationheterogeneityinSaccharomycescerevisiaecultures.Bothtoolsaresuitabletoreducetimesinprocessdevelopmentandoptimization.
Electrooptical monitoring of polarizability in Lactobacillus plantarum batch and fed-batch fermentations (P38)Klaus Pellicer Alborch1, Alexander Angersbach2, Peter Neubauer1, Stefan Junne1
1 Chair of Bioprocess Engineering, Department of Biotechnology, Technische Universität Berlin, Berlin, Germany, klaus.pelliceralborch@tu-berlin.de, 2 EloSystems GbR, Berlin, Germany.
Thedegreeofpolarizabilityofcellsdependsparticularlyonthecells’physiologicalstate[1]andisrelatedtotheirviability:cellswithhighmetabolicactivityandintactenergyhouseholdpossessalargerpolarizability[2].Theanisotropyofpolarizability(AP)canbemonitoredonlineapplyingelectroopticalmethods[2]withthehelpofautomatedsamplepreparation.Inpreviousreports,weshowedthatthemonitoringoftheAPofrod-shapedbacteriainbatchcultivationsallowedthepredictionofmetabolicsynthesisrates[3,4].TheobjectiveofourresearchistheinvestigationofwhetherdifferentgrowthconditionsandvariablefeedratesinanaerobicL.plantarumATCC8014batchandfed-batchcultivationsinflu-encetheAP,aimingtoidentifysuitablecultivationconditions.Theeffectsofsubstratepulses,temperaturevariation,pHshiftsandpresenceofoxygenhavebeeninvestigatedconcerninginfluencesontheAP.Inallcases,thegrowthphasecouldbedividedinsev-eralstages,basedonAPspectraoffourfrequencies(200,400and900kHzand2.1MHz).Themonitoringofthisparameterenableadeeperinsightintotheactualphysiologicalstateofcellsinrealtimeandwithoutanystainingorfluorescentmarkers,asusedinflowcytometry.Finally,theautomatedmeasurementapproachallowsitsapplicationasaprocessanalyticaltool.
Simulation of biomass transport and deposition in expanded bed adsorption chromatography (P39) Vikas Yelemane, Martin Kangwa and Marcelo Fernández-Lahore
Downstream Bioprocessing Laboratory, Life science and Chemistry, Jacobs University, Bremen, Germany, v.yelemane@jacobs-university.de
Integrativeapproachesindownstreambioprocessplayanimportantroleindevelopingcosteffectivepurificationproc-essforbio-moleculeswithincreasedrecovery[1,2].Expandedbedadsorption(EBA)isoneofsuchintegratedunitopera-tionsinproteinpurificationwithestablishedhandsonappli-cationinresearchaswellasindustrialsectors.However,EBAhasbeenevidencedtosufferfromthebiomassinteractioninanunclarifiedfeedstock.Hence,understandingtherationalebehindthebiomassadsorbentinteractionformsthebasisofresearchwhileusingEBArelatedapproaches.Additionally,investigatingthepotentialstrategiestoreducethisinterac-tionformachallengingscientificquestionthathastobead-dressedwhichhasgainedcurrentresearchinterestinthisarea[3].PresentstudyconsistsofresultsobtainedfrombiomassdepositionexperimentsonexpandedbedusingYeastandCHOcellsasamodelorganismandStreamlineDEAEandFast-lineMabDirectProteinAasmodeladsorbents.Processwasmodelledinordertoextractquantitativeinformationsuchas,attachmentefficiency(α),single-collectorcontactefficiency(η),particledepositionratecoefficient(kdsec-1),attachmentrateka1[1/s]anddetachmentratekd1[1/s].Theseparam-etersarecomparedwithprocessmodificationbychangingmobilephasepropertiesoradsorbentsurfaceproperties.
[1]M.FernándezLahoreOetal.ExpandedBedChromatography,SurfaceEnergeticsofBiomassDeposition,in:EncyclopediaofIndustrialBiotechnology,JohnWiley&Sons,Inc.,2009.
[2]JThommesetal.Theinfluenceofcelladsorbentinteractionsonproteinadsorptioninfluidisedbeds,AbstrPapAmChemSoc216(1998)U246-U246.
[3]RVennapusaetal.Assessingadsorbentbiomassinteractionsduringexpandedbedadsorptionontoionexchangersutilizingsurfaceenergetics,JChromatogr.A,1181(2008)9-20.
On-line state estimation applied to 550L filamentous fungal fermentations (P40)Lisa Mears1, Stuart M. Stocks2, Mads O. Albaek2, Gürkan Sin1, Krist V. Gernaey1
1 CAPEC-PROCESS Research Centre, Department of Chemical and Biochemical Engineering, Technical University of Denmark , Denmark, 2 Fermentation Pilot Plant, Novozymes A/S, Bagsværd, Denmark
Stateestimationallowsonlinepredictionofstatevariableswhicharenotabletobemeasuredon-linebyuseofavailableonlinedata.Forbioprocesses,keystatesofinterestincludebiomassconcentrationandproductconcentrationwhichareseldomavailableasdirectmeasurements[2].Thislackofknowledgeofthecurrentstatesinthesystemleadstochal-lengesforprocessmonitoringandcontrol.Thereisthereforeaninterestinstateestimationmethods,inordertomodelthesekeyprocessstatesbasedonavailableon-linemeasure-ments[1].Inthisworkafirstprinciplemodelisappliedto550LscalefilamentousfungalfermentationprocessesoperatedatNovozymesA/S.Themodelcomprisesofanonlineparameterestimationblock,coupledtoaphysicalmodelofthesystem.Theparameterestimationblockutilizeson-lineoffgasmeas-urementsandnitrogensourceadditioninordertomodelchangingreactionratesinthesystem.Basedonprocessstoichiometry,thecurrentratesofproductandbiomassfor-mationarethenidentified[3].Thisparameterestimateisthenusedasaninputtoadynamicmechanisticprocessmodel,
32
whichdescribesthemasstransfercapabilitiesofthesystembasedontheoperatingconditions,includingstirrerspeed,aerationrateandheadspacepressure[4],[5].Thefullmodelthereforeprovidesinsightintothecurrentmetabolicstateofthesystem,andtheoxygentransfercapacityinreal-time.Thisprocessmodelissuccessfullyappliedon-linetoeightnewpilotscalebatchesasastateestimatorinordertopredictthebiomassandproductconcentration,fromrobust,avail-ableon-linemeasurements.Suchstateestimatorsarevalu-abletoolsforonlineprocessmonitoring,andcontrolstrategydevelopmentforon-lineprocesscontrolandoptimization.
[1]P.Sagmeister,P.Wechselberger,M.Jazini,A.Meitz,T.Langemann,andC.Herwig,“Softsensorassisteddynamicbioprocesscontrol:Efficienttoolsforbioprocessdevelopment,”Chem.Eng.Sci.,vol.96,pp.190–198,Jun.2013.
[2]B.Sonnleitner,“AutomatedMeasurementandMonitoringofBioprocesses :KeyElementsoftheM3CStrategy,”no.November2012,pp.1–33,2013.
[3] R.T.J.M.VanDerHeijden,J.J.Heijnen,C.Hellinga,B.Romein,andK.C.aM.Luyben,“Linearconstraintrelationsinbiochemicalreac-tionsystems:I.Classificationofthecalculabilityandthebalance-abilityofconversionrates,”Biotechnol.Bioeng.,vol.43,no.1,pp.3–10,1994.
[4] M.O.Albaek,K.VGernaey,M.S.Hansen,andS.M.Stocks,“Evalu-ationoftheenergyefficiencyofenzymefermentationbymecha-nisticmodeling.,”Biotechnol.Bioeng.,vol.109,no.4,pp.950–61,Apr.2012.
[5]M.O.Albaek,K.VGernaey,M.S.Hansen,andS.M.Stocks,“Mod-elingenzymeproductionwithAspergillusoryzaeinpilotscalevesselswithdifferentagitation,aeration,andagitatortypes.,”Biotechnol.Bioeng.,vol.108,no.8,pp.1828–40,Aug.2011.
Influence of agitation intensity on the morpho-logy, rheology and productivity of Trichoderma reesei in fed-batch fermentations (P41)Daniela Quintanilla1, Kim Hansen2, Anna Eliasson Lantz1, Krist V. Ger-naey1
1 Department of Chemical and Biochemical Engineering, Technical Uni-versity of Denmark, Building 229, 2800 Lyngby, Denmark 2 Novozymes A/S, Krogshoejvej 36, 2880 Bagsværd, Denmark
Filamentousmicroorganismsarewidelyusedashostsintheproductionofindustrialenzymes.Nevertheless,theyhavemajordisadvantages,duetotheunavoidableoxygentransferlimitationstheypresent.Thisisaconsequenceofthehighviscositiesofthemediatheydevelop,whichisrelatedtohighbiomassconcentrations,growthrateandmorphology.Anobviousstrategytoovercomethisproblemistoincreaseagitationpower.However,thequestionariseswhetherthefragmentationormorphologicalchanges,whichoccurduetoshear,haveaninfluenceonproductivityinthatcase.
Inthiswork,wedeterminedtheinfluenceofagitationintensityonthemorphology,rheologyandproteinproductioncapabilityofTrichodermareeseiRUT-C30inbenchscalefermenters.Eightfed-batchfermentationswereconductedin2Lbioreactorsatdifferentagitationspeedsandwithtwodifferentglucoseconcentrationsinthemediumduringthebatchphase.Thefermentationswerefedatveryslowfeedratesensuringthatnooxygenlimitationscondi-tionswereencountered;lactosewasusedascarbonsourceinthefed-batchphase.Allthefermentationsreceivedthesameratiooffed-carbon/initialglucoseconcentration.Thefermentationswereconductedatfourdifferentagitationsspeeds(600,1000,1400and1800rpm).Forthefermentationsranatveryhighagitationspeeds,higherCPR(carbondioxideproductionrates)wereobservedforbothmediaconcentrations;theincreasedagitationspeedalsoresultedinlowerbiomassconcentrations.Thissuggeststhatthefungusisrespiringathigherrateswhentheagitationspeedincreasestosustaintheeffortcausedbythehighshearrate,whichisinagree-mentwith[1].However,theyieldsforextracellularproteinovertotalfedcarbonwerenotsignificantlyaffected.Themorphologyandrheologywereconsiderablyaffected,i.e.higherfragmenta-tionwasobservedatthehigheragitationspeeds–asexpected–whichledtolowerviscosities.Thisstudyhasallowedustoconcludethatagitationintensitycanbemanipulatedinordertoreduceoxygentransferlimitationsandimprovebulkmixing.Itcanalsobeusedtoreduceviscositybyaffectingthemorphologywithoutaffectingenzymeproduction.Thisisinagreementwiththeresultsfrom[2].
[1] R.LejeuneandG.V.Baron1995“EffectofagitationongrowthandenzymeproductionofTrichodermareeseiinbatchfermentation,”Appl.Microbiol.Biotechnol.,43(2)249–258,1995.
[2] A.Amanullah,L.H.Christensen,K.Hansen,A.W.Nienow,andC.R.Tho-mas,“Dependenceofmorphologyonagitationintensityinfed-batchculturesofAspergillusoryzaeanditsimplicationsforrecombinantproteinproduction,”Biotechnol.Bioeng.,77(7)815–826.
www.bioproscale-conference.org
Poster abstracts
Exhibitors
Recommended