Continuous Plant Cell Perfusion Culture Bioreactor Characterized & Enzyme Product

7/27/2019 Continuous Plant Cell Perfusion Culture Bioreactor Characterized & Enzyme Product

1/8

J OURNALOF BIOSCIENCE ND BIOENFINMRINGVol.95, No. 1,13-20.2003

Con tinuous Plant Cell Perfusion Culture:Bioreactor Characterization and Secreted Enzym e ProductionWE1 WEN SU* AN D RENEE ARIAS

Department of Molecular Biosciences & Bioengineering, University of Hawaii.Honolulu, HI 96822, USA

Received 16 January 2002iAccepted 0 August 2002Culture perfusion is widely p racticed in mam malian cell proc esses to enhance secreted anti-

body production. Here, we report the development of an efficient continuous perfusion processfor the cultivation of plant cell suspensions. The key to this proc ess is a perfusion biorea ctor thatincorpo rates an annular settling zone into a stirred-tank biorea ctor to achieve continuous cell/me-dium separation via gravitational sedimentation. From washout experiments, we found thatunder typical operating conditions (e.g., 200 rpm and 0.3 wm ) the liquid phas e in the entire per-fusion bioreactor was homogeneous despite the presence of the cylindrical baftle. Using secretedacid phosphatase (APase) produced in Anchusa officinulis cell culture as a model we have studiedthe perfusion cultures under comp lete or partial cell retention. The perfusion culture was oper-ated under phosphate limitation to stimulate AP ase production . Successful operation of the perfu-sion process over four weeks has been achieved in this work. When A. officinalis cells were grownin the perfusion reacto r and perfused at up to 0.4 wd with com plete cell retention, a cell dryweight exceeding 20 gl l could be achieved while secreted A Pase productivity leveled off at approx-imately 300 units/l/d. The culture became extremely dense with the maximum packed cell volume(PCV) surpassing 70%. In comparison, the maximum cell dry weight and overall secreted A Paseproductivity in a typical ba tch culture were lo-12 g/l and 100-1 50 units/Z/d, respectively. Ope ra-tion of the perfusion culture under extremely high PCV for a prolonged period, however, led todeclined oxygen uptake a nd reduce d viability. Subsequently, cell remov al via a bleed stream at upto 0.11 wd was tested and shown to stabilize the culture at a PCV below 60%. W ith culture bleed-ing, both specific oxyge n uptake rate and viability wer e shown to increase. This also led to ahigher cell dry weight exceeding 25 g/l, and further improvement of secreted APase productivitythat reach ed a plateau fluctuating around 490 units/l/d.[Key words: perfusion bioreactor, plant cells, protein secretion, high cell density]

Higher plants have been shown in recent years to be suit-able hosts for large-scale recombinant protein production(1). In fact, crop-based foreign protein production, or mo-lecular farming (2), has becom e a growing enterprise withdemo nstrated successe s in producing a variety of produc tsincluding an tibodies, therapeutic proteins and industrial en-zymes. For producing specialty protein products with highvalues, plant cell cultures have been proposed as an appeal-ing alternative to transgenic plants growing in open fields(24). The main advantages offered by plant cell culturesinclude faster grow th rates than their whole-plant counter-parts, their ability to grow using simple and inexpensivemedia in a well-controlled environment such as a bioreactor,and their capacity in comp lex posttranslational processing(2-4). Furthermore, GM 0 issues associated w ith field-growntransgenic crops are avoided w ith the plant-cell-culture pro-duction systems. In addition to these more genera1 points,de Wilde et al. (5) pointed out a less obvious distinctionbetween bioreactor-cultured plant cells and whole plants.

* Correspondin g author. e-mail: wsu@haw aii.eduphone: +I-808-956-3531 fax: +l-808-956-3542I3

These researche rs sugges ted that transgene silencing, w hichis triggered in a limited number of cells, might not spreadthrough out the bioreactor (5). In whole plants, the signal forsystemic acquired silencing is believed to be transmittedthrough the plasmodesmata and the vascular tissue, whereascultured plant cells are not interconnected in that way (5).de Wilde et al. (5) further reasoned that cultured plant cellsare commonly grown in a hemizygous state, which could beadvantag eous because in some cases transgene silencing istriggered only when the transgenes are present in a homo zy-gous state. In our study of green fluorescent protein (GFP)expression (6), as well as in a study of P-glucuronidase(GUS) expression (7), both using tobacco suspension cul-tures, the transgene silencing phenomenon was not ob-served.

To improve downstream recovery efficiency, it is desir-able to allow secretion of the target protein into the sur-rounding medium. Cultures operate d at a high cell densityusing a perfusion system can potentially increase the pro-ductivity of secreted produ cts (8, 9). In a perfusion system,cell-free spent medium containing secreted produc ts is con-stantly remov ed and harvested , while the culture is simulta-


2/8

14 SU AND ARIAS J. BIOSCI. BIOENG.,

neously replenished with fresh medium. There have beenonly a handful of published studies on perfusion plant cellcultures (10 -13). Findings from these studies, along withother aspects of perfusion plant cell cultures have beenreviewed by Su (8, 9). Previously, we reported a perfusionbioreactor design in which cell retention was accom plishedby incorporating an internal cell-settling zone into the down-come r region of an external-loop air-lift bioreactor (14). Us-ing this reactor, we were able to achieve a two-fold increasein cell concentration, over batch cultures, with continuousperfusion ( 14).

In an attempt to further enh ance the perfusion bioreactorperformanc e, the same concept for cell retention was ap-plied to a stirred-tank bioreactor in this study. Stirred-tanktype reactors are generally superior to air-lift reactors intheir mixing and mass transfer ca pacity, espe cially underhigh cell densities. For the perfusion bioreactor described inthis work , an annular settling zone was incorporated into astirred-tank reactor by inserting a cylindrical baffle. Here,we report on the liquid mixing characteristics of this stirred-tank perfusion bioreactor, and show that such a reactorcould be used for the continuous perfusion cultivation ofplant cell suspensions. Our ultimate goal is to use the sys-tem for the efficient production of secreted recombinantproteins. In this pilot study, the production of a non-recom-binant protein, the secreted acid phos phatas e (APase) fromA. officinalis cell suspension culture, was chosen as a modelto investigate the bioreactor performanc e and the perfusionprocess. APase was also selected because of the potentialbiotechnological applications of its promo ter for inducibleexpression of recombinant proteins. APase is the mostabundant enzyme found in the extracellular compa rtment ofA. officinalis cultures (Liang, Ph.D. thesis, Univ. of Hawaii,Honolulu, 199 8). The promo ter and putative signal se-quence of a member of the Anchusa APase multigene fam-ily have been cloned in our laboratory (Liang, Ph.D. thesis,Univ. of Hawaii, Honolulu, 1998 ). An APase prom oter sys-tem that is inducible by phos phate starvation has also beencloned from Arabidopsis by Hat-an et al . (15). By under-standing the kinetics of Anchusa APase synthesis and secre-tion in a perfusion process , this study also provides informa-tion useful for applying the APase promo ter in conjunctionwith high-density perfusion cultures to improve the produc-tion of high-value recombinant protein produ cts by trans-genie plant cells.

MATERIALS AND METHODSCell culture The routine culture of A. ofJicinalis cell sus-pensions ha s been described elsewh ere (14). Briefly, stock culturesof A. of$cinalis were maintained as cell suspensions in a liquidGamborg B5 medium supplemented with 1 mg/l2,4-dichlorophe-noxyacetic acid (2,4-D), 0.1 mg/l kinetin, and 30 g/Z sucrose (16).The suspension was maintained on a gyratory shake r at 120 rpmand 25C and subcultured every 10 d using a 10% inoculum. Cul-tured A. oJ?kinaZis cells formed tine suspensions with very fewlarge cell aggregates.Assays Analytical proced ures for cell dty weight, mediumosmolarity, and packed cell volume (PCV) have been described

elsewhere (17, 18). For the measurement of total extracellular pro-tein concentration, the culture sample was centrifuged, the super-

natant was filtered through a 0.2~pm mem brane filter and thenassayed using a protein assay reagent (Bio-Rad, Hercules, CA,USA) w ith bovine serum albumin as the standard (19). For intra-cellular phosphate measurement, plant cells were ruptured bysonification, followed by extraction of free ortho-phosphate intodeionized water. For extracellular phosphate measurement, theculture supernatant was collected by centrifugation. The molyb-date method (20) was used to determine phosphate concentrations.Oxygen uptake rate (yo) was measured on-line using the dynamicmethod (21). Acid phosphatase activity was measured by a colori-metric method usingp-nitrophenyl phosphate as the substrate (22).Cell viability was assessed by detecting the cell respiratory effr-ciency using 2,3,5triphenyltetrazolium chloride (TTC ) accordingto a modified protocol developed by Su and Arias (unpublished).By counting the formazan-producing cells, rather than measuringthe total formazan produced, the modified TT C method circum-vents the major drawbacks of the traditional TT C reduction assay(23), namely the heterogeneity of formazan production in a cellpopulation and the inconsistency in formazan extraction efficiency.Dissolved oxygen (D.O.) was measured using a polarographic oxy-gen electrode (Ingold, Lemexa, KS, USA). Northern analysis wasperformed using total RNA isolated from cultured A. oJkinaZiscells. Here, RNA extraction and gel blotting procedures similarto those described elsewhere (6) were used, except that a 1.2-kbAnchusa APase cDNA clone (AP32; Liang, Ph.D. thesis, Univ. ofHawaii, Honolulu, 1998 ) was used to synthesize the 32P-1abeledDNA probes. Sugars (sucrose, glucose, a nd fructose) w ere mea-sured by HPL C using a Supelcosil LC-N H2 column (i.d., 4.6 mm;length, 25 cm; cat no. 5833 8; Supelco, Bellefonte, PA, USA). Sep-aration was achieved using an aqueous solvent with 80% acetoni-trile a t a flow rate of 1 ml/min.

Perfusion bioreactor The perfusion stirred-tank reactor(PSTR ) system is presented schematically with the major dimen-sions in Fig. 1 A (note the reactor has a dished bottom). This glassbioreactor uses a New Brunsw ick Scientific BioFlo III head-plate

(4

15 cmt

E4 cm

13cm+

+_Cell Bleed

w

FIG. 1. (A) Schematic diagram of the PSTR with major dimen-sions; (B) A. officinalis cells cultured in the PSTR (note the wavy cellsediment in the bottom of the reactor).


3/8

VOL. 95,2003and it has a working volume of 3.3 1 (including the settling zone).A critical component of the reactor is a cylindrical baffle thatallows the creation of a stagnant zone within the reactor. Th is de-sign concept has been used in a perfusion reactor system for cultur-ing mam malian cells (24). Compared with plant cells, mamm aliancells are smaller in size and they are not as prone to cell aggrega-tion, and as a result, cell/medium separation with simple sedimen-tation was effective only under relatively low perfusion rates (24).Gravitational sedimentation nonetheless can be a simple and effec-tive way for achieving perfusion in plant cell culture, as indicatedin our previous study (14). The volume of the annular stagnant(settling) zone in the PSTR is 0.6 1. Pressure equilibration betweenthe well-mixed and the settling zones is ensured by connecting theheadspaces of the two compartments. Continuous cell-mediumseparation in the perfusion reactor is achieved by simple g ravita-tional sedimentation. The reactor was jacketed for temperaturecontrol. A six-bladed Rushton turbine (on top) and a three-bladedupward pum ping marine axial impeller (on the bottom, m odel E-0191 9-30; Cole Parmer, Vernon Hills, IL, USA) were used for agi-tation while a sintered glass tube with a mean pore op ening ofabout 140 pm was used as a sparger placed below the lower im-peller. The delivery of perfusion medium and culture bleedingwere done using ultra-low flow peristaltic pumps (model 77; Har-vard Apparatus, Holliston, M A, USA).

Cultivation of A. officinalis in the perfusion bioreactorThe bioreactor containing B5 medium supplemented with 1 mgil2,4-D, 0.1 mgil kinetin, and 30 g/l sucrose was inoculated with a1O-day-old culture of A. officinalis. For the perfusion runs, the cul-ture was allowed to grow in the batch mode until the extracellularAPase activity approached 1200-l 500 U/l (which is the maximumAPase activity encountered in a typical batch culture). The culturewas then perfused using B5 medium (containing 0.5 mM phos-phate) supplemented with 3% sucrose and reduced hormone con-centrations (0.1 mg/l 2,4-D an d 0.01 mgil kinetin). This perfusionmedium (except for the reduced phosphate concentration) wasused in our previous study to achieve high cell density in A. of$cci-nalis perfusion culture carried out in an external-loop air-lift bio-reactor (14). The agitation and aeration rates were set at 200 rpmand 0.3 vvm, respectively. To maintain D.O. at 30% air saturation,a PID algorithm was implemented using the LabVIE W program-ming language (National Instrument, Dallas, TX, USA) to adjustthe air/O? ratio in the gas inlet via two mass-flow controllers(Omega Engineering, Stamford, CT, USA) connected to a dataacquisition board (National Instrument) supervised by a PC. Foamcontrol was achieved using a silicone-based antifoam solution(Antifoam C emulsion, Sigma Chemical). The culture temperaturein the bioreactor was kept at 25C.

Charac terization of liquid-phase mixing in the perfusionbioreactor Mixing in the perfusion bioreactor was character-ized using washout experiments. Phosphate was chosen as thetracer in these experiments primarily because it was the limitingsubstrate in the perfusion culture medium and it can be measuredeasily. The reactor was initially charged with 1 mM aqueous phos-phate solution. At the onset of the experiment, a continuous waterstream was pumped into the reactor with or without the cylindricalbaffle at a preset rate using a peristaltic pump (model 77; HarvardApparatus) while another stream was continuously removed viagravitational overflow through a site port. In another set of experi-ments, the effluent stream was removed from within the well-mixed zone using a peristaltic pump identical to that used for feed-ing the input stream into the well-mixed zone. The two pumpswere carefully calibrated to ensure delivery of indistinguishableflows under the same setting within the time frame of each experi-ment. Agitation and aeration rates during these washout experi-ments were set at 200 rpm and 0.3 vvm, respectively. The tempera-ture in the bioreactor was kept at 25C. The decrease in phosphate

PERFUSION PLANT CELL CULTURE 15concentration at different locations in the reactor was followedduring the course of the experiments.

RESULTS AND DISCUSSIONLiquid-phase mixing in the perfusion bioreactor In

the PSTR , cell/medium separation during culture perfusionis achieved by incorporating a cylindrical baffle into the re-actor. By incorporating this baffle an annular stagnant re-gion is created inside the reactor to serve as a settling zone.It is thus necessary to characterize to what extent the baffleaffects the overall mixing in the bioreactor. While w e haveobserved from culture experimen ts that the annular settlingzone was essentially free of cells during the perfusion oper-ation (as long as the PCV w as kept below 60%), it was un-clear how liquid-phase mixing was affected by the presenceof the cylindrical baffle. To answer this question, we charac-terized the residence time distribution (RTD ) in the reactorby conducting a series of washo ut experimen ts. Mediu mperfusion was anticipated to cause a higher degree of bulkliquid mixing in the settling zone. As such, two sets ofwasho ut expe riments were perform ed - one for simulatingthe batch operation and the other for the perfusion opera-tion. To simulate reactor washou t during the batch phase ofthe bioreactor operation, the eMuent stream was withdrawnfrom within the well-mixed zone of the bioreactor, so thatthe settling zone was not disturbed. Typical w ashou t datafrom such experiments are presented in Fig. 2. To obtainthese washout data, liquid samples were withdrawn fromvarious locations in the reactor as indicated in the figureinsert. When collecting these samples, special care wastaken to minimize the disturbance to the liquid phase. T hesedata indicate a very uniformed concentration distributionthroughout the reactor, and the data match the ideal CSTRmodel, suggesting some degree of convective mass transferwould have to be considered in the settling zone even dur-ing the batch ph ase of the reactor operation. This findingwas not influenced by the washo ut rates within the rangetested (0.02-0 .5 h-). In fact, during routine culture e xperi-ments, w e have repeatedly observed wavy motion at the in-terface of the liquid medium and culture sediment at the en-

00 0.5 1.0 15 2.0 25VT(- )

FIG. 2. Tracer (phosphate) washout during simulated batch opera-tion (i.r., effluent stream withdrawn from the well-mixed zone). C,Tracer cont.; C,,, initial tracer cont.; t, time; r, residence time. Experi-mental data and ideal CST R model simu lation are represented by thesymbo ls and the solid line, respectively.


4/8

16 SU AND ARIAS J. BIOSCI. IOENG.,trance of the annular settling zone (note the bottom of thereactor as shown in the photo in Fig. 1B). During the perfu-sion phase o f the reactor operation, effluent exited from thesettling zone through an overflow tube. It is therefore notsurprising that the washo ut curves are essentially identicalin the presence or absence of the cylindrical baffle, and thewashout data match the ideal CSTR model, even at a wash-out rate as low as 0.02 h- (data not sh own). Results fromthese washout experiments suggest that the liquid phase inthe 3.3 I perfusion reactor could be considered homog ene-ous under normal operating conditions despite the presenceof the cylindrical baffle. This finding is useful for furtheranalysis of the bioreactor in which d issolved substrate/prod-uct concentrations in the liquid phase can now be assumedto be uniform through out the reactor.

m 0.5 Ink40 4.0 mM

DW APase

1500

31000QL$500 $

%0

Characteristics of acid phosphatase expression Inthis study, APase was chosen as a model secreted proteinprimarily for two reasons. First, APase is the major secretedprotein in the A. ojkinalis culture, as confirmed by elec-trophoresis and protein sequencing analysis (Liang, Ph.D.thesis, Univ. of Hawaii, Ho nolulu, 1998 ). Furthermore, theAPase pro moter may be used for effective inducible geneexpression. We have cloned the promo ter and putative sig-nal sequence of a memb er of the APase multi-gene familyfrom a cell culture of A. officinalis (Liang, Ph.D. thesis,Univ. of Hawaii, Honolulu, 1998 ). In addition, Haran et al .(15) recently dem onstrated inducible re porter gene expres-sion using an Arabidopsis APase promoter under phosphatestarvation.

FIG. 3. Effect of phosphate on cell dry weight, APase transcriptlevel, and intracellular APase activity. Insert: Northern blot of APasetranscript from 8-day-old A. officinalis cells cultured in medium con-taining either 0.5, 1 or 4 mM phosphate.

and 1 mM ) had a very strong hybridization signal, indicat-ing that the expression of the acid phosphatase genes in A.of$cinalis cell culture is stimulated by low pho sphate con-centration. Specific intracellular APase enzyme activity wasalso inversely correlated with the phos phate concentrationas shown in Fig. 3 while the biomass yield w as increased byabout 50% when the initial phosph ate concentration was in-creased from 0.5 to 4 mM.

Extracellular acid phosphatases appear to be ubiquitousin roots and plant cell cultures. These extracellular APasesmay be localized either within the cell wall or secreted tothe surrounding environment (25). In cultured tob acco cells,it was found that the synthesis and secretion of an APasewere regulated by pho sphate (26). In tomato suspension cul-tures, AP ase is also secreted and is sugges ted to functioneither as a phosphate transport agent or a phosphate-scav-enging agent that acts on the phosphorylated compounds inthe culture m edium (27). When linked to the ArabidopsisAPase promo ter and signal sequence, a GFP reporter wasshown to be secreted by the roots o f transformed Arabidop-sis plants under phosphate starvation (15).

To establish a nutrient (phos phate) regime th at suppo rtsAPase production in the perfusion bioreactor, we first char-acterized the effect of phosphate on APase expression inshake-flask cultures. We found that the Anchusa APase isencoded by a gene family the members of which share morethan 85% homology (Liang, Ph.D. thesis, Univ. of Hawaii,Honolulu, 1998). A 1.2-kb cDNA fragment that covered the3 part of the full-length gene enco ding one of the AnchusaAPase isoenzymes (AP32) was used to prepare a 32P-labeledDNA probe. Northern analysis w as perform ed using totalRNA isolated from 8-day-old A. of$cinaZis cells grown inmedium containing either 0.5, 1 or 4mM phosphate. Asshown in the insert of Fig. 3, the hybridization signals ob-tained with the AP32 probe showed a marked difference inthe mRNA levels for the acid phosphatase. The RNA fromcells growing in medium with higher levels of phosp hate(4 mM) h ad a very weak hybridization signal whereas theRNA from cells growing in lower levels of phosphate (0.5

Cell growth and acid phosphatase production in theperfusion bioreactor One of the main objectives in per-fusion cultivation is to achieve high cell density as a meansto increase reactor volumetric productivity. The first set ofperfusion reactor experiments was therefore conducted withtotal cell retention to minimize cell loss. We then examinedthe effect of culture bleeding in another set of experiments.To stimulate APase production, all cultures w ere operate dunder phospha te-limited growth . At an initial phosp hate con-centration of up to 1 mM, the Anchusa culture was shownto be under pho sphate limitation based on shake flask ex-periments (Smith, M. S. thesis, U niv. of Hawaii, Honolulu,1997 ). The initial pho sphate concentration in the mediumfor all perfusion culture experiments was thus set at 0.5mM . Typical results from perfusion experiments in the ab-sence of culture bleeding are presented in Fig. 4. In this andsubsequent figures, extracellular and intracellular phosp hateconcentrations, and extracellular and intracellular APaseactivities are denoted by P,,, P,,, APase,,, and APase,,, re-spectively. The goal here was not to achieve steady statesand study culture behavior under such conditions. Instead,our aim was to achieve the highest possible cell concentra-tion under phos phate limitation. Accordingly, the perfusionrate was adjusted empirically through out the culture withthe aim of sustaining cell grow th while k eeping a low in-tracellular phos phate level to stimulate APase production.In a typical batch culture using a standard B5 medium (con-taining 1 mM phosphate), the cell dry weight reached about12 g/l (Smith, M.S. thesis, Univ. of Hawaii, Honolulu, 1997 ).With culture perfusion, one can achieve and maintain ahigher cell density. As shown in Fig. 4, the perfusion culturebecame very thick (with a PCV exceeding 70%) after about17 d of cultivation. Once the culture reached such a highPCV , the cells began to enter the annular settling zone,


5/8

VOL.95, 2003 PERFUSION PLANT CELL CULTURE 17

c3 02253 20 0. 8 F. "g 15 s&IO 0.4 ho

50 0.0I O 15 20 25 30 3512

I I0 5 10 15 20 25 30 35

I Is- I2 1500~~1000% 500

00 5 15 20Time (d)

80

60g

40 j-g20

0

FIG. 4. Time course of perfusion culture with complete cell reten-tion. The first and second dashed lines are referring to onset of perfu-sion and switching to 2x medium , respectively (cf: Fig. 5A and seetext for details).which led to a gradual decline in the PCV from days 21 to25. From day 25, the perfusion rate was halved (from 0.4 to0.2 vvd) and we switched to a 2 x perfusion medium (Fig.5A). This was carried o ut to prevent cell washo ut while p ro-viding a similar level of nutrients to the cells. The upwardtrend of the PCV resumed after day 25. The cell dry weightfollowed a similar pattern to the PCV d uring the batchphase of the culture. How ever, during the perfusion phase ,the change in cell dry weight was subtler and the peak ob-served for the PCV around days 18-2 1 was not seen withthe dry weight. Cell dry weight reached and remained at ap-proximately 18 +2 g/ l from days 13 to 2 1 during which thesettling zone was practically cell free. A small but steady in-crease in cell dry weight was observed from day 25 on-wards, which is consistent with the increased PCV and oxy-gen uptake rate during th at period. The maximum cell dryweight reached approxim ately 22 g/l in the perfusion culturewithout cell bleeding. This is a little lower than thatachieved previously in a perfusion air-lift bioreactor inwhich the cell dry weight reached 27 g/l (14). It should benoted, ho weve r, in that previous study the culture w as notoperate d under any nutrient limitation.

The pho sphate concentration in the medium remained es-sentially undetectable from day 2 onward s d espite mediumperfusion being initiated on around day 11. The exceedinglyfast phosphate uptake by the cultured Anchusa cells corre-sponds to a very high maximum specific phosphate uptakerate of approximately 5 1 mg phosphate/g dry weight/d asdetermined in our previous study (28). The intracellular freephosphate content also stayed at a very low (below 1 mg P/g

- 0. 6 G5e,

- 0. 4 22 Z%- 0. 2 a2X perfusion medium

.o.oI O 15 20 25 30 35

F 600"c, g 500. >2 400ea5 300H2 200

100I O

B

r____-_---------------,J15 20 25

Time (d)30 :

0.8

z3. 6 v

aJ$B

0. 4 g38' Z0.2 Lga

0.0

FIG. 5. Time course of secreted APase productivity (circles) withcorresponding perfusion (solid line) and bleed rates (dashed line): (A)complete cell retention; (B ) perfusion with cell bleeding.DW ) and constant level througho ut the perfusion phase , in-dicating a rapid turnover of the absorbed free phosp hate intophosphorylated compounds (29). As such, the medium per-fusion strategy implemen ted met the pho sphate limitationrequirement. By comparing the phos phate profiles with theAPase production in Fig. 4, it is apparent that regulation ofAPase expression by phosphate starvation is governed byintracellular rather than extracellular phos phate levels. Theinitiation of medium perfusion caused a sharp decrease inthe extracellular enzyme activity due to the dilution effect.An increase in extracellular APase activity was neverthelessobserved about 2 d afterwards. Although no exact steadystate of APase prod uction was observed during the perfu-sion culture, by taking into account the changes in perfusionrates, the volumetric extracellular APase productivity fromday 20 onwards was found to fluctuate around 300 U/lid(f30 U/l/d; Fig. 5A), which is about tw ice the productivityseen in batch cultures (Smith, MS . thesis, Univ. of Hawaii,Honolulu, 1997). Note that between days 20 and 25, thereappe ared to be a metabolic shift that led to declined oxygenuptake and intracellular APase activity, despite stepwise in-creases in perfusion rates from 0.2 to 0.4vvd during thisperiod (Fig. 5A). Judging from the medium osmolarity (datanot shown) which dropped to 18 mOSM/l (regular freshmedium has an osmolarity reading of 220 mOSMII) on day18, the perfusion rate prior to day 19 might have been settoo low which led to excessive nutrient starvation. Toge therwith the possible stress caused by the extremely high PCV(the population crowding effect), it could have caused the


6/8

18 SU AND ARIASreduced metabolic activities seen from days 20 to 25. Alsoduring this period, an increased amount of cells began toenter the settling zone due to the high cell loading (a PCVexceeding 70%). W e noted that for the Anchusa culture, theminimum PCV that corresponds to a 100% settled cell vol-ume (the culture biotic volume fraction after the cells areallowed to settle for a sufficient period under g ravity) isabout 70% (14 ). Therefore, it is not surprising that as theculture PCV reached 70% in the well-mixed zone the cellsstarted to get pushe d into the settling zone.

The observation of an extremely high PCV and reducedmetabolic activity during the perfusion culture under com-plete cell retention prom pted us to look into ways to allevi-ate the potential cellular stresses created by the highlycrowd ed culture environment. In microbial chem ostats withcell recycling or in perfusion mammalian cell cultures, ma-nipulation of the cell bleed rate has been identified as a via-ble operating strategy to manipulate specific grow th ratesand to reduce dead cell accumulation (30). To study howcell removal affects the performanc e of the plant cell perfu-sion culture, the perfusion bioreactor was operate d with ableed stream . Th e results a re presented in Figs. 5-7. Thecell bleed rate was determined based on a conservative esti-mation of the cell specific grow th rate, and was initially setat 0.04 vvd and later increased to 0.11 vvd (Fig. 5B). Theperfusion rate was first set at 0.2 vvd, increased to 0.4 vvdon around day 20 and then adjusted back to 0.2vvd onabout day 25 (Fig. 5B). Under th ese conditions, improvedculture performance in terms of a higher cell dry weigh t,improved culture viability (data not shown), increased ex-tracellular phos phatas e production, as well as augmen tedoxygen uptake and a lower PCV were noted (Fig. 6). Simi-lar to the perfusion culture w ith complete cell retention, noapparent steady state with respect to extracellular APaseproductivity was reached ; rather the productivity fluctuatedbetween 400 and 600 U/l/d, and averaged around 490 U/l/dfrom day 20 onwards (Fig. 5B). This productivity is approx -imately 60% higher than that achieved in perfusion culturewith comp lete cell retention. With cell removal, the PCVcould be kept below 60% while the maximum cell dryweight exceeded 25 gll. The increase in oxygen uptake rela-tive to that seen in the case of comple te cell retention wasalso quite marked. A closer examination of the oxygen up-take rate data revealed that for both perfusion cultures thespecific oxygen uptake rate (SOUR, oxygen uptake ratedivided by cell dry weigh t) peak ed in early to mid exponen-tial growth during the batch phase of the cultures. TheSOU R then declined and fluctuated around a steady level(approximately 0.12 and 0.05 g 0,/g DW/d for partial andcomp lete cell retention, respectively) from day 14 onward sduring th e perfusion phase . Due to the relatively constantSOUR, the course of oxygen uptake closely paralleled thatof cell dry weight during the perfusion phase of the cultureunder either complete or partial cell retention. Oxygen up-take rate is therefore potentially useful as a metabolic in-dicator for estimating cell grow th online in a perfusion cul-ture. More over, since the SOU R did not increase after cul-ture perfusion was comm enced, it may be possible to furtherimprove the culture performa nce by initiating the perfusionearlier while the culture is metabolically more active (e.g.,

J.Brosa. BIOENG. ,35 5

e3 0 4325520 3g' g 15 2 BLpg10

5 1

0 0 5 10 15 20 25 30 35'

00 5 10 15 20 25 30 :1i5

12 - 60

g g - 6052 7@3s - 40 >

zD && -2 0

Ti me(d)FIG. 6. Time course of perfusion culture with cell bleeding. T hedashed lines represent changes in perfusion rates (cf: Fig. 5B and seetext for details).

as indicated by the high ATP production rate and SOUR).We also examined the carbon source utilization during

the continuous perfusion cultivation with cell bleeding.During perfusion, sucrose fed into the reactor w as rapidlyconverted to glucose and fructose by hydrolysis, while glu-cose was preferentially utilized by the Anchusa cells (Fig.7). By comparing Figs. 5B and 7, it is clear that the perfu-sion rates had a dominant effect on the sugar profiles. More -over, medium osmolarity closely mimicked the concentra-tion of the hydrolyz ed sugars (Fig. 7). This observation isconsistent with our previous study in which an A. officinalisculture was grown in a stirred tank bioreactor opera tedunder discontinuous perfusion via in situ filtration (31).Hence, medium osmolarity might serv e as a convenient

100 .otiz:o E6- 100,0 5 I O 15 20 25 30

Ti me(d)

FIG. 7. Ti me course of sugar concentrations during perfusion cul-ture with cell bleeding (cf Fig. 6).


7/8

VOL. 952003 PERFUSIONPLANTCELLCULTURE 19

process param eter fo r monitoring sugar concentrations inperfusion cultures. Also noted from these data was thepresence of a large amount of residual sugars (gluco se andfructose) in the medium. Thus decreasing the sugar concen-tration in the fresh perfusion medium might be considered,although operating at a very low level of residual sugar mayhave an adverse effect on the culture due to the low mediumosmotic pressure.

ACKNOWLEDGMENTThe authors are grateful to Hua Liang for conducting the North-em blot analysis, and to Charles Nelson for his excellent technicalassistance in constructing the reactor employed in this study. Thiswork was supported in part by the National Science Foundation(grant no. BES97-12916).

Cell retention and culture mixing in the perfusionbioreactor The cell retention efficiency of the bioreactorwas mainly affected by the cell loading, perfusion rate, andthe agitation and aeration rates (8). In our previous study(14) whe re we tested an external-loop air-lift perfusion bio-reactor, w e observed that it was especially important toalleviate gas bubble entrapment in the settling zone to as-sure prope r reactor performan ce. Air bubbles not only dis-turbed cell sedimentation but also carried cells upwa rds, re-sulting in excessive biomass accumulation and wall grow thabove th e settling zone. In air-lift rea ctors, aeration and agi-tation are coupled . As a result, as mixing becom es more de-manding at high cell loading, a higher aeration rate has to beused wh ich results in a higher gas holdup and the bubblesare more pron e to escape into the settling zone. Th is prob-lem is alleviated in a stirred-tank type reactor (such as thePSTR ) in which aeration and agitation are decou pled (i.e.,they can be controlled independently). More over, we incor-porate d two design considerations to minimize the presenceof bubbles in the settling zone o f PSTR . First, the cylindri-cal baffle was extended so that the lower edg e of the baffleleveled with the bottom impeller. Second , an upward-pum p-ing axial impeller was used as the bottom stirrer. In thisstudy, it was noted from the washout experiments that whilewe observed a homogeneous liquid ph ase in the perfusionbioreactor, mixing of the cell sediment below the cylindricalbaffle was clearly very sluggish. After extended operation,the cell sediment apparently becam e mo re comp act and itcould not be readily re-suspend ed without a substantial in-crease in the reactor agitation rate that could result in exces-sive turbulence in the bulk (the well-mixed zone) of the re-actor. This design shortcom ing could potentially be over-come by installing a separate mixer system for the bottompart of the reactor (e.g., using a magnetically driven stirrer)operated at a lower rpm than the stirrers used to mix the cul-ture in the bulk of the reactor.

REFERENCES

5.

6.

7.

8.

9.10 .

Il.

12 .

13 .

This study has demo nstrated the usefulness of a modifiedstirred-tank bioreactor for the continuous perfusion cultiva-tion of plant cells. This perfusion bioreactor system sh owsgreat potential as an effective alternative to immobilizedplant cell bioreactors. The results indicate that it is desirableto opera te the perfusion culture with cell bleeding. Furtherimprovement of the culture performance could be achievedby optimizing the perfusion and bleed rates. We suspect th atby operating the culture un der extremely high pack ed celldensity and prolonged nutrient starvation could lead to ex-cessive physiological stresses that cause culture deteriora-tion. Thus, to further extend and improve the operation ofperfusion culture, m ore study is necessary to better under-stand these cellular stresses, and ways to alleviate them.Such studies are currently underway in our laboratory.

14 .

15 .

16 .

17 .

18 .

19 .

Giddings, G.: Transgenic plants as protein factories. Curr.Opin. Biotechnol., 12,45 0454 (2001 ).Fischer, R. and Emans, N.: Molecular farming of pharma-ceutical proteins. Transgenic R es., 9,279- 299 (2000 ).Doran, P. M.: Foreign protein production in plant tissue cul-tures. Curr. Opin. Biotechnol., 11, 199-2 04 (2000 ).James, E. and Lee, J. M.: The production of foreign proteinsfrom genetically modified plant cells. Adv. Biochem. Eng.Biotechnol., 72, 127-1 56 (2001).de Wilde, C., van Houdt, H., ve Buck, S., Angenon, G., veJaeger, G., and Depicker, A.: Plants as bioreactors for pro-tein production: avoiding the problem of transgene silencing.Plant Mol. Biol., 43, 347-3 59 (2000 ).Liu, S., Bugo s, R ., Dharm asiri, N., and Su, W. W.: Greenfluorescent protein as a secretory reporter and a tool for proc-ess optimization in transgenic plant cell cultures. J. Biotech-nol., 87, I-16 (2001) .Gao, J., Lee, J. M., and An, G.: The stability of foreign pro-tein production in genetically modified plant cells. Plant CellRep., 10, 533-536 (1991).Su , W. W.: Perfusion bioreactors, p. 230-24 2. In Spier, R. E.(ed.), Encyclopedia of cell technology, vol. 1. Wiley, NewYork (2000).Su , W. W.: Bioprocessing technology for plant suspensioncultures. Appl. Biochem. Biotechnol., 50, 189-2 30 (1995).Pareilleux, A. and V inas, R. A.: A study on the alkaloid pro-duction by resting cell suspensions of Catharunth~s roseus ina continuous flow reactor. Appl. Microbial. Biotechnol., 19,316-320(1984).Kim, D. I., 00, G. H., Pedersen, H., and C hin, C. K.: Ahybrid bioreactor for high density cultivation of plant cell sus-pensions. Appl. Microbial. Biotechnol., 34, 726-7 29 (19 91).Shin, M . K., Park K., and Cho G. H.: Ultrasonic cell sepa-rator as a cell retaining device for high density cultures ofplant cell. Biotechnol. Bioprocess Eng., 4, 264-2 68 (1999 ).Su , W. W. and Hump hrey, A. E.: Production of plant sec-ondary metabolites from high density perfusion cultures,p. 266-2 69. In Furusaki, S., Endo, I., and Matsuno, R. (ed.),Biochemical engineering for 2001 . Springer-Verlag, Tokyo(1992).Su , W. W., He, B . J., Liang, H., and Sun, S. : A perfusionair-lift bioreactor for high density plant cell cultivation andsecreted protein production. J. Biotechnol., 50, 225-2 33(1996).Haran, S., Logen dra, S., Seskar, M., Bratanova , M., andRaskin, I.: Characterization of Arubidopsis acid phosphatasepromoter and regulation of acid phosphatase expression. PlantPhysiol., 124, 615-6 26 (2000) .De-Eknamku l, W. and Ellis, B. E.: Rosmarinic acid produc-tion and growth characteristics of Anchusa officinalis cell sus-pension cultures. Planta Med., 51, 346-3 50 (1984 ).Dixon, R.A.: Isolation and maintenance of callus and cellsuspension cultures, p. I-20. In Dixon, R. A. (ed.), P lant cellcultures: a practical approach. IRL Press, Oxford (19 85).Su , W. W., Lei, F., and Su, L. Y.: Perfusion strategy for ros-marinic acid production by Anchusa officinalis. Biotechnol.Bioeng., 42, 884-890 (1993).Bradford, M. M.: Rapid and sensitive method for the quanti-


8/8

20 SU AND ARIAS J. Broscr. BIOENG.,tation of microgram quantities of protein u tilizing the prin-ciple of protein dye binding. Anal. Biochem., 72, 248-2 54(1976).20 . Taras, M.: Standard methods for the examination of waterand waste water. American Public Health Association, NewYork (1971).21 . Taguchi, H. and Humphrey, A.: Dynam ic measurement ofthe volumetric oxygen transfer coefftcient in fermentationsystems. J. Ferment. Technol., 44, 881-8 89 (1966 ).22 . Nahas, E., Terenzi, H. F., and Ross i, A.: Effect of carbonsource and pH on the production and secretion of acid phos-phatase and alkaline phosphatase in Neurospora crassa. J.Gen. Microbial., 128,2017-2021 (1982).23. Towill, L. E. and M azur, P.: Studies on the reduction of2,3,5-triphenyltetrazolium chloride as a viability assay forplant tissue cultures. Can. J. Bot., 53, 1097-l 102 (1975 ).24 . Tokashiki, M., Hama moto, K., Takazawa, Y., andIchikawa, Y.: High-density culture of mouse-human hy-bridoma cells using a new perfusion culture vessel. K agakuKogaku Ronbunshu, 14,337-341 (1988).25 . Duff, S .M. G., Sarath, G., and Plaxton, W. C.: The role ofacid phosphatses in plant phosphorous metabolism. Physiol.

Plant., 90,791-800 (1994).26 . Ueki, K. and Sato, S.: Effect of inorganic phosphate on theextracellular acid phosphatase activity of tobacco cells cul-tured in vitro. Physiol. Plant., 24, 506-5 11 (1971 ).27 . Goldstein, A. H., Baerrlein, D.A., and McDaniel, R. G.:Phosphate starvation inducible metabolism in Lycopersiconesculentum. Plant Physiol., 87 , 711-7 15 (1988 ).28 . Zhang, J. N. and Su, W. W.: Estimation of intracellularphosphate content in plant cell cultures using extended Kal-man filter. J. Biosci. Bioeng., 94, 8-14 (2002 ).29 . van Gulik, W ., ten Hoop en, H., and Heijnen, J.: A struc-tured model describing carbon and phosphate limited growthof Catharanthus roseus plant cell suspensions in batch andchemostat culture. Biotechnol. Bioeng., 41, 771-7 80 (1 993).30. Hiller, G. W., Clark, D. S., and Blan ch, H. W.: Cell reten-tion-chemostat studies of hybridoma cells - analysis of hy-bridoma growth and metabolism in continuous suspensionculture on serum-free medium. Biotechnol. Bioeng., 42, 185-195 (1993).31. Su, W. W., Lei, F., and Kao, N. P.: High density cultivationof Anchusa officinalis in a stirred tank bioreactor with in situfiltration. Appl. Microbial. Biotechnol., 44,29 3-299 (1995 ).

Documents

Continuous Plant Cell Perfusion Culture Bioreactor Characterized & Enzyme Product