CA2073974A1 - Process for the high density fermentation of escherichia coli in an agitator vessel fermentor - Google Patents

Abstract

Abstract The invention relates to a process for the high cell-density fermentation of Escherichia coli in an agitated tank fermenter.

Description

`~ Title ef the inve~tio~ zSr~
Proce~ for the high cell-den~ity far~entation of E~cherlchia coli in an agita~ed tank ferment2r Field of thQ inventio~
S The invantion relate~ to a process for the high cell--den~ity fermentation of E. coli ~train~, specific ally tho~e Yuitable a~ ho ts for vectors with recombinant expression systems. The field of the invention lie~ in those branches of indu~try in a national economy which hsve a biotechnological ~ector and employ in the la~tar B. coli strainY as producer organisms.
Prior art E~ coli i~ often used a~ cellular host for the production of recombinant DNA product~. Beside-~ a high intracellular concentration of the de3ired product, a crucial preregui~ite for a high overall yield is ~he achievement of high concentrations of cells in the fer~enter.
The problems which ari~e in the cul~ivation at high cell den~ities in a conventional agitated ~ank fermenter are those of growth inhibition owing to the initial ~ub~trate concentxations being too high, the con3umption of a~sential con~tituent~ of the nutrient medium during the cour~e of cultivation, the formation of metabolic by-product~ which have inhibitory effect~, and the limited capacity of oxygen introduction.
Variou~ ~trategies c.~n b~ used to attempt to meet the increasing oxygen dema~d o~ a growiny B. coli ~ulture, nam~lys increa~ing the ~tirrer ~peed and the ga~
introduction rate, introducing air enriched in o~ygen (Jung et al. 1988; Pa~3 et al. 1989; RrUger 1989;
~ppstein et al. 1989), introducing pure oxygen ~a3 (Bauer and Sh~loach 1974; Shiloach and Bauer 1975; Bauer and White 1976; ~auer and Ziv 1976; ~ori et al. 1979; Gleiser and ~au~r 1981; ~itzutani et al. 1986; Bailey et al.
1987 ; ~an et al . 1987; ~r~ger 1989), cultivation at low temperature~ (Shiloa~h and Bauer 1975; ~auer and Whi~e 1976; Bauer and Ziv 1976) and under elevated pre~ure (T~ai et al. 1987).

3 7~
E. coli ha~ been fermented in a glucose/mineral ~alt medium with exce~s oxygen up to 16 . 5 g of dry biomass of X/l (Reiling et al. 1985). R~Uger (1989) and Eppstein et al . ( 1989 ) achieved 19 g of X/l . Growth at higher ~ell densitie~ required a fed-batch technique in order, on the one hand, to eliminat~ inhibikion and limitation of ~ub~tancss and, on the other hand, to prevent the formation of inhibitory metabolic by-product~. It was necessary in all casas to metex in other nutrient~ as well besides the C source and ammonia as N
qour~e and pH regulator. U~ing a glucose/mineral salt medium, Eppstein et al . ( 1989 ) achieved 39 g of a~/l by additional subsequent feeding with salts. Fas~ et al.
(lgB9) achiered 45 g of XJl by metering~ glut:ose ~olution, which al~o contained magnesium 3ulphate, into a glucoYe/mineral salt ~edium which had been initially introduced. Additional ~ubsequent feeding with salts by Pan et al. (1987) resulted in ~5 g of X/l.
Other fed-batch fermentations were carried out in gluco~s/mineral ~alt media whic~, beforQ inoculation thereof, additionally contained yea~t extract or auto-lysad yeast powder. 38 g of ~Jl were obtained without additional ~ub~equent feeding during cell growth apar~
from glucose and ammonia (Mori et al. lg79~. Additional matering in of ~alt-~ produced higher biomas~ concentra-tions with 47 g of ~/l (Bausr and White 1976), 55 g of ~/1 (Shiloach and Bauer 1975) and 68 g o~ X~l (B~uer and ZiY 1976). Additional metering in of ~alts, trace ele-ments and vitamins re ulted in 70 g of X/l (Fie~chko and 3n Rit~ch 1986). It wa~ possible to obtain very high call densitie~ of 110 g o~ X/l ~Cutayar and Psillon 1989) and 125 g o~ X/l (~ori et al. 1979) by co~plex ub~equant feeding with ~alt~, yea~t extract and trace el~ment~ in addition to the customary me~ering in of gluco~e and ammonia. Further ~uppl~mentation of a glucose/mineral ~alt medium which already contained yea~t extrac~ before inocula~ion with Bacto tryp~one or pep~one, and the metering in of thase complax substrat2~ ~uring the fermontation re~ulted in no further increase in th _ 3 ~
bioma3s concentration (~ailey et al. 1987; Tsai et al.
1987). Very high biomass concentrations o~ 80 to 105 g of X/l (~ppstein et al. 198~) were achievad by combina~ion of th~ ~ed-batch technique de~crib~d above (me~ering in of glucose and salts) by introduction of air enriched in oxygen. Kr~ger (1989) achieved 112 g of ~/1 by introd~c-ing pure oxygen gas.
All the proc~s~es described above are associated with the di~ad~an~age tha~ thay represent very co~plex subsequent feeding ~y~tem~. ~he addition of all the additional n~trients apart fro~ glucose, or gluco~e mixcd with magne~ium sulphate, a~d æmmonia requires specific metering in of nutrient sub~trates, which co~pri~e either o~ly intermittent addition~ or time-controllQd addition phas3~. This also applie~ to the inpst of the ga~eous oxygen ~ub~trata. The aim is a process for achieving high cell den~ities u~ing a gluco~e/minsral salt medium without the nece~sity for fur~her nu~rient~, apart from tha usual gluco~e ~olution mixed with mag-ne~ium sulphate, and ammonia, with the u~u~l introductiono~ air, and where po~sible without the nec2s~ity for enrichment with pure oxygan.
E. coli cells which act as ho~t~ for vectors and thu~ are used for the production of recombinant DNA
product3 must b~ fermented appropriately for the ex-pression y~tem to be employed for ~ha recombinant product. ThiY mean~ that E. coli ho~t-~ with expre~ion systæms which c~n b~ ~witched on are inltially fermentQd at high eell d~nsitis~ and then the ~witching on of the recombinant D~A e~pre~ion i~ carried outc Al~o known for . coli ar~ a n~mber o~ homologou~ and h2terologous expre~ion ~y3tem~ which are con~titutive or ~or which expr~ssion i5 higher at low~r SpQCi~iC growth rate~ than at the ~aximu~ ~pecific growth rate pos~ible, ~, in the 3S particular nut~ient medium. ~he ~pecific growth rate i~
defined as follow~:

~ . d~/dt.

~ 4 -In .~ome ca~e a ~rowth of E. coli with ~
has bèen achieved. Thus, Zabriskie and Arcuri (1986) achieved growth of the culture with ~ by feeding the carbon source into the fermentation solution at a constant rate which was lower than that which would have resulted in ~. However, the di ad~antage of this procedure i8 that ~ change continuously. If tha metering rate were kept con~tant throughout the fed-batch proces~, ~ would f~ll continuously.
Allen and Luli ~1985) developed a ~trategy for the intermittent read~u~tment of the ~pecific growth rate ~ with increasing cell den~ity by readjusting the ~eeding rate~ of the carbon source from time to tLme. The implicit disadvantage of thi~ ~trat~gy i~ that ~ changes lS co~tinuou~ly in a tolerance range which d~pend~ on the frequency of the off-line measurement~ of gluco~e and cell d3nsi~y, although growth is at ~ < ~. Changes in the 3pecific growth rate entail fluctuation3 in mata-boli~m and may ha~e a de~tabili~ing effect, especially when there i~ production of m~tabolic by-product~ with inhibitor~ effect~. Lee and Mohler (1989) prs~anted a controlled growth-rate fermentation in order to achieve grow~h at constan~ ~. The core of the procedure is continuou~ mea~urem~nt of ths carbon dioxide released by the cell~ during growth, the instan~a~eous growth ra~e calcula~ed from this by c~mputer, and the imm~diate ad~ust~nt, which i3 nece~ary to main~ain the required gxowth rate, o~ th~ metering rate of th~ carbon souxco with the coupled feeding o~ the C ~ource. ~ee and Mohler ~0 (1989) were able in ~hi~ way to cultiva~a ~. coli a~
approximately con~tant ~ < ~ the range of cell d~nsities ~om 0.3 t~ 60 g o~ ~/1 (EP 031594~ ~2). T~
3trategy of maintaining ~ constant via msa3urement of ~he variable C02 is, however, as~ociated in principle with the di~ad~antage that CO2 is ~ot in every case a growth-corre-lated ~ignal. ~his particularly applies when change~ inmetaboli~m take place or a~aplerotic bio ynthetic path-way~ are u~ed by the cell~, which may ocour par~iculaxly ~ith very high cell den~ities and carbon deficiency.

Another disadvantage of the hee and Mohler procedure is that growth rapidly Rtop~ after exhau6tion of all ~he po3sibilitie~ for increa~ing the oxygen input in the particular ferment0r 3y8tem U ed, because the di~olvsd oxygen concentration decrea~Qs immediately. Thu~, after the growth phase at ~ - const ~ , no detectable prolongation o~ growth to increase the cell density further at ~ c ~3~ iq posYibla when ~ i3 continuou31y decreasing. Furtharmore, the di~solved carbon dioxide concentration depends ve~y greatly on ~he pH in the ad~usted p~ range; small pH change~ have adverse effect~
on the measured signal and may result in falsification of the ~alculation of ~
It is therefore al~o the intantion that this inYention de~criba a fermentation proces~ ab12 ~pecifi-cally to re~ult in ve~y high cell densitie~ of E. coli with ~table metabolism.

Ai~ o~ th~ i ven~io~
Tha aim of the invention compri~e~ de~cribing a proceoR for achi~ving high cell densitie~ of E. coli in an agitated tank fermanter, where srowth at the maximum specific growth rate is followed by forcing growth at ~ubmaximum ~pQCi~iC growth rate, and the dissolved oxygen concentration in the fer~entation msdium is maintained at or above a defined value.

Su~ary of thQ iN~e~ion ~ he invention ha3 the ob~ect of describing, with avoidance of the di3advantages of th~ prior art, a proc~ with which B. coli can be fermented in a gluco~e/
~i~eral ~alt m~dium in ~n agitated tank fermenter to high cQll dQnsities without further metering in of other nutrient 3ub~trate~, apart fro~ tha subsequent feeding with gluco~e ~upplemented with magnesium ~ulphate, ammonia and an~ifoam agent, being nece~ary and without oxygen-enrichment of ~he air which i3 introduced bein~
necessary. The intention in thi~ connection wa~ that formentation be carried out in ~uch a way that the growth ~- 6 - 2~ 7~
of the E. coli cells takes place initially at the ma~imu~
specific growth rat~ and then ater an appropriate interval at ~ubmaximum ~pecific growth rate.
The object on which the invention i~ bas~d i~
S achieved by a proce~ for the high call-density fexmenta-tion of Eschorichia coli in an agitated tank fermenter, which i~ characteri~ed in that fsrmhntation is carried out in a medium containing a nitrogen source, organic carbon source and mineral salt~, and ~he f~rmentation is carxied out in ections by providing - fir~tly a batch section with maxi~um specific growth ra~e and - then a fed-ba~ch section with ~ubmaximum specific growth rata, the latter being co~trolled and/or regulated 15via the o~ygen i~put~ using a ~lucose solution supple-mented with ma~ne~ium ulphate.
j. According to a spacific embodimen~, a time I profile with an approximat~ly con3tant ~ubmaximum ~pecific growth rate can be provided in the fed-batch 20~ction. It is furthermora pos~ible for the fed-batch s2ction to be comple~Qd, after the maximum oxygen input predete~mined by the agitated tank fermenter ha~ bee~
r~ached, by a sub~ection in which the ~pecific growth rate decre sa~.
25To carry out the proGess according to the in~e~-tion it i~ pos~ible to fermant E. coli R12 or a deriva-tive thareof and, preferably, the derivatiYe ~. coli TGl.
It i po~sible to carry out ~he ferment~tion in the pr~e~ce of ~upplin~ a~d thereby to compensate 30au~o~rophi~s. In the c~e of a thia~in~ auxotrop~y, the ~. coli ~rain can b~ ~e~m~nted in ~he ~re~once of thiamine. Fo~ exam~le, ~. coli ~Gl can be fermen~ed in the prRsen~e o a thihmine concentration s 4.5 m~/l.
It is pos~ible to introduce the gaseous o~yqen 35only with the aid of air throughout the ferm~ntation.
~owover, it i~ al~o pos~ible for the sa~eou~ oxygen ~irRt to be in~roduced with ~he aid of air and then to use air onriched with oxy~en, or oxygen.
~ono~accharides, di~accharide~ and/or polyols - 7 ~ J~J~ 7~
can be used a~ organic carbon source. ~xæmple8 are glucos~, ~ucrosa, lactosa and glycerol.
According to a qpe~if ic embodiment, a glucose ~olution which i~ ~upplemented with magnesiu~ ~ulphato S and contain~ an antifoam agent can be u~ed in tho fed-- batch section.
According to another ~pecific embodiment o~ the p~oce~s according to ~ha i~vention, gluco~e ~olution (for example of a concentration s 700 g/l) supplemented with 0 magnesium ulphate of a concen~ration 19.2 g of ~gS0~.7H20/l can be used in th~ fed-~atch section.
According to another ~pecific e~bodiment, the nutrient mediu~ ~mployed in the process according to the in~ention ean already contain all the nutrient 5ub~rate8 for the entire ferme~tation with the exceptions tha~
- aquaou3 ammonia solution (for example of a concentra-tion < 25%) i9 metered in f or pH ad~ust~ent, - gluco~e 301ution ( for example of a concentration ~ 700 g/l) ~upplemented with ma~ne~ium ~ulphate (for example of a conc~ntration < 19.2 g of MgSO~.7H20/l) i~
mete~ed in during the fed-~a~ch ~ection and - where appropriats an~ifoam age~t, for example Ucolub N115, iQ metered in.
The nutrient medium employ~d i~ the proce~s according to the inv~ntion can qualitati~ely compri~ tha following compono~t~, for example:
gluco~e, pota~sium dihydrogen pho~phate, diammonium hydrog~n phosphate, magnesiu~ ~ulphate, iro~ ~itrat~, cobalt chloride, mansane~e chloride, copper chloride;
borlc acid,-~odium ~olybdate, zinc ac~tate, Titriplex III
and/or citric acid and, wher3 appropriate, antifoa~
agent~ and, whare appropriata, ~uppline~ to compensate auxo~rophies.
In particul~r, the nut~i3nt ~edlum employed can compri~, for exa~ple, ~he following component~: gluco~e (< 50 g/l~, RH2PO, (s 13.3 g/l), (~H4)2HPb (S 4 g/1), NgSO,.7HzO (s 1.2 g/l), iron Gi~rate (5 60 mg/l), CoCl2.6H~O ( 2.5 mg/l), MnCl2.4H~O (s 15 mg/l), CuCl2.2~2O
(s 1.5 mg/l), H3BO3 (s 3 mg/l), Na~0O~.2H2O Is ~.5 mg/l), 2~a~9 7 ~ 8 ~
Zn(CH3C00)z.2H~O (s 8 mg/l), Titriplex III (s 8.4 mg/l) and/or citric acid ~ 1.7 g/l) and, where apprspriato, antifoan agent Ucolub ~115 (s 0.1 m~
To carry out the proce3s according to the inve~-tlon, the constituent~ o~ the nutrisnt medium can beintroduced into th~ fermenter in the following _equence:
- first potassium dihydrogen phosphate, dia~onium hydrogen phospha~o a~d/or citric acid;
- then a aolution, which i3 made up ~rom ~tock solution~
where appxopriate, of cobalt chloride, manganase chlor ids, COppGr chloride, boric acid, sodium molybdate9 zinc acetate andJor Titriplex III, - then iron citra~e, - then, wh~re appropriate, an antifoam agent, after which ~terili~ation i~ carried ou~.
Titriple~ III can be introduced fir~t for the ~olution mada up rom stock ~olution3.
After th& solution which ha~ been introduced initially into ths fermenter ha~ been 3terili~ed it is po~sible to intxoduce a ~terili~ed ~olu~ion o~ glucosa and/or magne3ium sulphatQ into the f~rmenter.
A lag pha3~ may occur after the inoculation o~
the nutrient medium and before ths batch section.
Farmentation in the batch ~ection can be carried out, ~or exampl~, at a p~ s 7 . 5, in particular in tha range fro~ 6.6 to 6.9 and preferably at ~bout 6.8. ~he pEI can b~ ad~u~ted u~ing aqlleous ammonia 8dUtiOIl ( of a concentration of, for exa~pl~, s 25~).
A~cording to a specific e~bod~men~ of ~he proceRs a~cording to the inven~ion, a P02 2 1%, praferably in the range ~rom 5 to 20% and, in p~rticular, about 10%, can be ad~u~ted in ~he ba~ch section by gradually increa~ng the sti~rar speed.
The transition to the ~ubmaximum ~peci~ic grow~h ~5 rate in th~ fed-batch sec~ion after the b~tch section i~
complete ca~ be effected by - ei~her ~educing the s~irrer 3peed, for example in the range fro~ 5 to 60 min, - or keeping th~ ~tirrer qpeed con~ant, for example in ~he ransa from 0.5 to 10 h.
A P02 2 1%, preferably in the range from 5 ~o 20~
and, in par~icular, of abou~ 10%, can be ad~usted in ~he fed-batch section by means of ~he glueose metQring.
It i~ also possible to ad~ust the requirad sub-- maximum ~pecific growth rata in the fed~batch section by inc~easing the o~ygen input. It i~ po-~ible for this purposa to increase the ~tirrer speed, the gas introduc-tion rate, the pres~ure and/or the oxygen content o~ the air, or to introduce oxygen gas.
Whon the ~ubmaximum sp~cific growth rate in the fed-batch ection i~ monitored, it is po~sible with its aid to ad~ust the oxygen inpu~ appropriately. ThQ 3ub maximu~ ~pacific growth rate can be monitored, for 1~ e~ample, by m~ans of the oxyge~ con~ent of the air l~aaving tho f~ nter.
~ccording to a speciic embodimant of the proce~3 according to the invention, after the maximum technically po~ible oxygen input has been reached in the fed-batch section it is po~sible to ferment the cultura further ~y meanY Of P02 control by me~ering in glucose while the su~maximum ~pecific ~rowth rate falls.
It qhould also be mentioned, for the ~ake of complet~ne~s, that th~ concentration of dry biomass can be altered by alteration of - the xatio of the duration o~ the ba~ch ~action to the duration of the fed-batch section andJor the level of the submaximum specific growth rate, which ; i3 to ba kept approx~ately constant, in the ~ed-batch ~ctionO Thu~, it is pos~i~le to ach~ve a~ least abou~
95 g of X/l, for examplo, on culti~ation of E. coli ~G1 at ~ = 0.107 l/h - con~ in thæ f~d-batch ec~ion.
After it ha~ cea3ed to be pos~ible to increa~e further the oxyge~ inpu~ effi~iency or the particular fermsnter ~y8~em, surprisingly ~he presen~ procas~
pe~lt8 further growt~, al~hough with ~alling ~ without the occur~nce of oxygen d~ficie~cy ~o~ the cell~ in the culture solution by means of the pre~e~ P02 control. A
dry bioma~3 concen~ration o~ at least 110 g o X/l ha~

- 10 ~ 2~ 3~',,J~
been obtained in the medium describad abov~ with ~. coli ~Gl in thi~ mannar, the final volu~e not e~cesding 76% of tha fermantar ne~ volume.
The invention i~ explained in detail hereinafter by a fi~ure and an exemplary embodimen~.

~xempl~ry ~bodiment ~he high cell-den~ity fermen~ation proces~ i9 to be described in datail taking the example of ~he cultiva-tion of E~ coli TGl in a 72 1 fermenter (B.Braun Mel~ungen A~, type 8015.1.01).
The 37 l o~ nutrient 501ution to ba inoculated cont~ined ~he following con~tituent~: glucoss (25 g/l), R~2PO4 (13.3 g/l), (NH4)2HP04 (4 g/l), MgSO4-7H20 (1.2 g~l), iron(III) citrate (60 mg/l), CoCl2 6H2O (2.5 mg~l), ~hC12 4HzO (15 mg/l), CuCl2-2H20 (1.5 mg~l), H3BO3 (3 mg/l), Na2~oO42H20 (2.5 mg/l), Zn(CH3COO)2-2H20 (~ mg/l)/ thiamine (4.s mg~l), Titriplex III (8.4 mg/l), citric acid (1.7 g/l) and Ucolub N115 (0.1 ml/l). ~his nutrlent ~olution wa~ prepared in th~ following way: firstly, 148 g of ~0 (NH4)2HP04/ 492.1 g of KH2PO~ and 62.9 g of citric acid a~
dry ~ubstance were dis~ol~ed in about 30 1 of deioni~ed watar i~ tha fermanter; this wa~ followed by ~ddition of 257.4 ml of trace mixture ~olution, addition o an iron(III) citrate solution (2.22 g of iron(III) citrate in 300 ml) and o 3.7 ml of ~colub NllS. The trace mixtur~ solu~ion wa~ prepar~d from ~tocX olutions in the following way: 62.2 ml o~ Titriplax III (5 g/l), 34O3 ml o~ CoClz-2HzO (2.7 g/13, 34.7 ~1 of CuCl8-2~2O (1.6 g~l), 34.7 of ml M~Cl2-4H20 (16 g/l), 27.8 ml of H3BO3 (4 g~l), 30.8 ml o~ Na2MoO~2H20 (3 g/l) a~d 32~9 ml of Zn(C~3COO)2-2~O (9 g/l). The ~olution in th~ fer~n~e~
wa~ terili~d in the u~ual way. Also added sub~eq~ently for c~m~letion-w~re sepaxately ~terili~ed glucose solu-` tion (1.0175 kg of glucose monohydrate in 2.5 1 of H20)/
sterili~ed magn~ium sulpha~e solu~ion (44.4 g of NgSO~-7H2O in 1 1 of ~2O), thia~ine ~olution ~terilised by filkration (166.5 mg in 100 ml of H2O), the pH was ad~usted t~ 6.8 with aqueous ammo~ia Rolution (25%), and finally the mixture wa~ made up to 36.8 l with ~eril~
H20. It wa subsequently inoculated with 200 ml of a thawed glycerol pre~erved cultur~ (20% v/v3 of E. coli TG1.
E. coli TG1 wa~ cultivated at a tempera~ure of 28C, a pre~sure of 1.5 bar, a pH of 6.8 and a gas intxoduction rate of 85 l/min. The pH was kept constan~
throughout the fermentation b~ controlled ~e~ering in of 25% NHa. Fig. 1 ~hows the tine cour~e~ of the stirrer speed, the concentration of dry biomas~ X, the glucose concentration and the di3solved oxygen concentration.
In the ba~ch ~ection of the high cell-den~ity farmentation, the culture grew, after the inoculation at time t ~ O and after a ~hort lag pha~e o about 2 h, at th~ maximum s~ecific growth rate (~ 0.456 1/h~. Ater tha PO2 had reached 10% it was subsequently kept con~tant by controlled increase of the ~irrer speed. The batch section wa~ co~pletQd after con~umption of the gluco~e întroduced at the start. The PO2 increased draxtically.
A~t2r thiC there was no further control of the PO2 via the ~tirrer speed. To reduce tho specific growth rate fro~ ~ = O.456 to ~ ~ 0.11 1/h, the ~tirrer ~psed wa~
reduced from 420 rpm to 250 rpm over a 30-minute period.
~he PO2 control was ubsequently operated via the meter-ing in of glucose ~olution (700 g/l) ~upple~ented with magnesium ulphate (lg.2 g of MgSO~-7~20/l). ~he main-tena~ce of ~ constant in pha~e 1 of ~he fed-batch ~ection wa~ effected by increa3ing the oxygen input. For thi~
purpo~e I the ~pead wa~ 9et via a proportional plu~
integral con~rollex ~i~h the control variable ~6. The ~peciic growth rate wa~ determined i~directly from the o:scyçle~ balance~ of tha 3y~tem by th~ followi3lg fonnula:
a,,~(t) t K~ao2(~

where: Qo2 i~ the o~ygen con~ump~ion rate and R i3 the quoti~nt of the dry bioma~

12 ~ 'al~
concentration at time to and ths yield cesf~ici0nt for oxygell.
A dry bioma3~ concen~ration of 95 g/l ~a~ reached at l:h~ ~d of the first pha~e of the fadbatch ~ection S Becau~e it wa3 no lon~er po~i~le to increase the oxys~er inpu~ into the culture ~olution ~fter the maaci;mum techni~
cally possi~le Rtlrrar ~peed had been ~ached, ths culture ~hen g:re~ in phase 2 of the ~ed batch section at a constantly falling ~. A dry bioma~s of 110 g of X/l wa~
obtained at the end of thi~ pha3e and thus at the comple-tion o~ the ~srmentation.

Claims (30)

1. Process the high cell-density fermentation of Escherichia coli in an agitated tank fermenter, characterized in that fermentation is carried out in a medium containing a nitrogen source, organic carbon source and mineral salts, and the fermentation is carried out in sections by providing - firstly a batch section with maximum specific growth rate and - then a fed-batch section with submaximum specific growth rate having an approximately constant time profile, the latter growth rate being controlled and/or regulated via the oxygen input, using a glucose solution supplemented with magnesium sulphate.

2. Process according to Claim 1, characterized in that the magnesium sulphate for supplementing the glucose solution is added separately into the agitated tank fermenter.

3. Process according to Claim 1, characterised in that an approximately constant submaximum specific growth rate is provided as time profile in the fed-batch section.

4. Process according to Claim 1, 2 or 3, charac-terised in that the fed-batch section is completed, after the maximum oxygen input predetermined by the agitated tank fermenter has been reached, by a subsection in which the submaximum specific growth rate decreases.

5. Process according to any of the preceding claims, characterised in that E. coli K12 or a derivative thereof and, preferably, the derivative E. coli TG1 is fermented.

6. Process according to any of the preceding claims, characterised in that fermentation is carried out in the presence of supplines, and auxotrophies are compensated.

7. Process according to Claim 6, characterised in that, in the case of thiamine auxotrophy of the E. coli strain to to fermented, fermentation is carried out in the presence of thiamine.

8. Process according to Claim 7, characterised in that E. coli TG1 is fermented in the presence of a thiamine concentration ? 4.5 mg/1.

9. Process according to any of the preceding claims, characterised in that - the gaseous oxygen is introduced throughout the fermen-tation only with the aid of air or - the gaseous oxygen is introduced at the start of fermentation with air and later with air enriched in oxygen or with pure oxygen.

10. Process according to any of the preceding claims, characterised in that monosaccharides, disaccharides and/or polyols are used as organic carbon source.

11. Process according to Claim 10, characterised in that glucose, sucrose, lactose and/or glycerol is used as organic carbon source.

12. Process according to any of the preceding claims, characterised in that a glucose solution which is supple-mented with magnesium sulphate and which contains an antifoam agent is used in the fed-batch section.

13. Process according to any of the preceding claims, characterised in that glucose solution (especially of a concentration ? 700 g/l) supplemented with magnesium sulphate of a concentration ? 19.2 g of MgSO4.7H2O/l is used in the fed-batch section.

14. Process according to any of the preceding claims, characterised in that the nutrient medium employed in the process already contains all the nutrient substrates for the entire fermentation with the exceptions that - aqueous ammonia solution (especially of a concentration ? 25%) is metered in for pH adjustment, - glucose solution (especially of a concentration ? 700 g/l) supplemented with magnesium sulphate (es-pecially of a concentration ? 19.2 g MgSO4.7H2O/l) is metered in during the fed-batch section and - where appropriate antifoam agent, for example Ucolub N115, is metered in.

15. Process according to any of the preceding claims, characterised in that the nutrient medium employed qualitatively comprises the following components:
glucose, potassium dihydrogen phosphate, diammonium hydrogen phosphate, magnesium sulphate, iron citrate, cobalt chloride, manganese chloride, copper chloride, boric acid, sodium molybdate, zinc acetate, Titriplex III
and/or citric acid and, where appropriate, antifoam agents and, where appropriate, supplines to compensate auxotrophies.

16. Process according to any of the preceding claims, characterised in that the nutrient medium used comprises the following components: glucose (? 50 g/l), RH2PO4 (? 13.3 g/l), (NH4)2HPO4 (? 4 g/l), MgSO4.7H2O (? 1.2 g/l), iron citrate (? 60 mg/l), CoCl2.6H2O (? 2.5 mg/l), MnCl2.4H2O (? 15 mg/l), CuCl2.2H2O (? 1.5 mg/l), H3BO3 (? 3 mg/l), Na2MoO4.2H2O (? 2.5 mg/l), Zn(CH3COO)2.2H2O (?
8 mg/l), Titriplex III (? 8.4 mg/l) and/or citric acid (? 1.7 g/l) and, where appropriate, antifoam agent Ucolub N115 (? 0.1 ml/l).

17. Process according to any of the preceding claims, characterised in that the constituents of the nutrient medium are introduced into the fermenter in the following sequence:
- first potassium dihydrogen phosphate, diammonium hydrogen phosphate and/or citric acid;
- then a solution, which is made up from stock solutions whers appropriate, of cobalt chloride, manganese chlor-ide, copper chloride, boric acid, sodium molybdate, zinc acetate and/or Titriplex III, - then iron citrate, - then, where appropriate, an antifoam agent and then sterilised.

18. Process according to Claim 17, charactarised in that Titriplex III is introduced first for the solution made up from stock solution

19. Process according to any of ths preceding claims, characterised in that, after sterilisation of the solu-tion initially introducad into the fermenter, a steri-lised solution of glucose and/or magnesium sulphate is introduced into the fermenter.

20. Process according to any of the preceding claims, characterised in that fermentation in the batch section is carried out at a pH ? 7.5, in particular in the range from 6.6 to 6.9 and preferably at about 6.8.

21. Process according to Claim 20, characterised in that the pH is adjusted using aqueous ammonia solution (of a concentration of, in particular, ? 25%).

22. Process according to any of the preceding claims, characterised in that a pO2 ? 1%, preferably in the range from 5 to 20% and, in particular, about 10%, is adjusted in the batch section by gradually increasing the stirrer speed.

23. Process according to any of the preceding claims, characterised in that, after the batch section is com-plete, the transition to the submaximum specific growth rate in the fed-batch section is effected by - either reducing the stirrer speed, in particular in the time period from 5 to 60 min, - or keeping the stirrer speed constant, in particular in the time period from 0.5 to 10 h.

24. Process according to any of the preceding claims, characterised in that a pO2 ? 1%, preferably in the range from 5 to 20% and, in particular, of about 10%, is adjusted in the fed-batch section by means of the glucose metering.

25. Process according to any of the preceding claims, characterised in that the required submaximum specific growth rate is adjusted in the fed-batch section by increasing the oxygen input.

26. Process according to Claim 25, characterised in that the stirrer speed, the gas introduction rate, the pressure and/or the oxygen content of the air is/are increased, or the gas introduced is oxygen.

27. Process according to any of the preceding claims, characterised in that the submaximum specific growth rate in the fed-batch section is monitored and, with its aid, the oxygen input is adjusted appropriately.

28. Process according to Claim 27, characterised in that the submaximum specific growth rate is monitored by means of the oxygen content of the air leaving the fermenter.

29. Process according to any of the preceding claims, characterised in that after the maximum technically possible oxygen input has been reached in the fed-batch section the culture is fermented further by means of pO2 control by metering in glucose while the submaximum specific growth rate falls.

30. Process according to any of the preceding claims, characterised in that the concentration of dry biomass is altered by alteration of - the ratio of the duration of the batch section to the duration of the fed-batch section and/or - the level of the submaximum specific growth rate, which is to be kept constant, in the fed-batch section.