Effect of Methanol Feeding Strategies on
`Production and Yield of Recombinant
`Mouse Endostatin From Pichia pastoris
`
`L. B. Trinh, J. N. Phue, Joseph Shiloach
`
`Biotechnology Unit, National Institutes of Health, NIDDK Building 6, Room
`B1-33, Bethesda, Maryland 20892-2715; telephone: 301-496-9719; fax:
`301-451-5911; e-mail: yossi@nih.gov
`
`Received 21 April 2002; accepted 14 October 2002
`
`DOI: 10.1002/bit.10587
`
`Abstract: Pichia pastoris, a methylotrophic yeast, is an
`efficient producer of recombinant proteins in which the
`heterologous gene is under the control of the methanol-
`induced AOX1 promoter. Hence, the accepted produc-
`tion procedure has two phases: In the first phase, the
`yeast utilizes glycerol and biomass is accumulated; in the
`second phase, the yeast utilizes methanol which is used
`both as an inducer for the expression of the recombinant
`protein and as a carbon source. Since the yeast is sensi-
`tive to methanol concentration, the methanol is supplied
`gradually to the growing culture. Three methanol addi-
`tion strategies were evaluated for the purpose of opti-
`mizing recombinant endostatin production. Two strate-
`gies were based on the yeast metabolism; one respond-
`ing to the methanol consumption using a methanol
`sensor, and the other responding to the oxygen con-
`sumption. In these two strategies, the methanol supply is
`unlimited. The third strategy was based on a predeter-
`mined exponential feeding rate, controling the growth
`rate at 0.02 h−1, in this strategy the methanol supply is
`limited. Throughout the induction phase glycerol, in ad-
`dition to methanol, was continuously added at a rate of 1
`g L h−1. Total endostatin production was similar in all
`three strategies, (400 mg was obtained from 3 L initial
`volume), but the amount of methanol added and the bio-
`mass produced were lower in the predetermined rate
`method. This caused the specific production of end-
`ostatin per biomass and per methanol to be 2 times
`higher in the predetermined rate than in the other two
`methods, making the growth control strategy not only
`more efficient but also more convenient for downstream
`processing. © 2003 Wiley Periodicals, Inc. *Biotechnol Bioeng
`82: 438–444, 2003.
`Keywords: Pichia pastoris; growth strategies; methanol;
`endostatin
`
`INTRODUCTION
`
`Pichia pastoris is an efficient producer of recombinant pro-
`teins (Cregg et al., 2000). This microorganism is a methy-
`lotrophic yeast (Faber et al., 1995) that utilizes methanol as
`its carbon source and grows to very high cell density (above
`50% wet weight concentration). In many cases it secretes
`
`Correspondence to: Joseph Shiloach
`
`the produced protein to the culture media. The reason for the
`efficiency of this yeast as a producer of recombinant protein
`is the biosynthesis of alcohol oxidase (AOX). AOX is the
`first enzyme in the methanol oxidation pathway. The en-
`zyme is produced in response to the methanol presence in
`the growth media. Methanol initiates the enzyme biosyn-
`thesis by activating the AOX1 promoter, causing the enzyme
`accumulation to about 30% of the total yeast proteins.
`Therefore, the gene of choice is introduced downstream to
`the alcohol oxidase promoter and is expected to express the
`desired protein in an equivalent amount (Cereghino and
`Cregg, 2000).
`Three types of recombinant Pichia can be obtained (Hig-
`gins and Cregg, 1998): Mut+ where the two methanol oxi-
`dases genes AOX1 and AOX2 are intact; MutS, where only
`AOX2, which is responsible for 15% of the protein biosyn-
`thesis, is intact; and Mut− where both AOX1 and AOX2
`genes are disrupted. The Mut+ strain is the most responsive
`to methanol concentration and is the recombinant strain
`most commonly used.
`Alcohol oxidase promoter is activated by methanol, but it
`is also repressed by carbon source such as glycerol or glu-
`cose (Tschopp et al., 1987). In addition, the microorganism
`is sensitive to methanol, and concentrations above 0.4%
`inhibit growth and production (Katakuga et al., 1998; Zhang
`et al., 2000a). Hence, the accepted production procedure
`from the Mut+ strain has two phases. In the first phase, the
`yeast utilizes glycerol and biomass is accumulated. In the
`second phase, the glycerol supply is terminated or reduced
`(Brierley et al., 1990; d’Anjou and Daugulis, 1997), metha-
`nol alone or in combination with lower concentration of
`glycerol (mixed feed) is added and the biosynthesis of the
`recombinant protein is initiated.
`Since the yeast is sensitive to methanol concentration,
`methanol needs to be supplied continuously to the growing
`culture keeping its concentration below 0.4%. Several fed-
`batch strategies for methanol addition have been established
`(Zhang et al., 2000b). These strategies can be metabolism
`related, based on parameters such as methanol consumption
`(Curvers et al., 2001; Guarna et al., 1997; Hellwig et al.,
`
`© 2003 Wiley Periodicals, Inc. * This article is a US Government Work
`and, as such, is in the public domain in the United States of America.
`
`Motif Exhibit 1035, Page 1 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`

`

`2000; Katakuga et al., 1998; Kobayashi et al., 2000), oxy-
`gen consumption (Byrne et al., 2000; Chung, 2000; Min-
`ning et al., 2001), pH control or CO2 concentration. While
`implementing these strategies, the microorganism itself is
`controling the amount of methanol added as a response to its
`own metabolism, the methanol supply is practically unlim-
`ited and its concentration can have any value, usually not
`more than 4 g/L−1. Growth strategies can also be based on
`predetermined methanol addition schemes at constant, lin-
`ear, or exponential rates (Chauhan et al., 1999; Freyre et al.,
`2000; Inan et al., 1999; Marasugi et al., 2000, Zhang et al.,
`2000a). In these strategies the methanol feeding rate and not
`the microorganism is controling the growth, the actual
`methanol concentration is zero and its supply is limited.
`It is very likely that the growth and production charac-
`teristics of the culture such as growth rate, production yield,
`and methanol consumption would depend on the fed-batch
`strategy. It is therefore important to evaluate the various
`fed-batch methods when the production process of a recom-
`binant protein is being evaluated. In some cases, metabo-
`lism-related method is preferred (Curvers et al., 2001), and
`in other cases, nonmetabolism-related method, in which the
`growth rate can be controlled, is preferred (Zhang et al.,
`2000a). However, detailed comparisons of various fed-
`batch strategies for the production of a specific recombinant
`protein are very limited (Kupcsulik et al., 2001; Minning et
`al., 2001; Zhang et al., 2000b). Our purpose here is to evalu-
`ate several growth strategies for the production of extracel-
`lular murine endostatin. Endostatin, is a 20 KDa fragment of
`collagen XVIII, that had been shown to have an inhibitory
`effect on angionenesis and therefore can potentially be used
`as a tumor growth suppressor (Figg et al., 2002; O’Reilly et
`al., 1997). The compound is currently being tested in hu-
`mans and different delivery strategies are being evaluated in
`mice. Gram quantities of the compound were needed and
`therefore efficient production procedure was required.
`The growth strategies evaluation performed in this work
`can provide information not only on growth and production,
`but also on the specific production values of biomass in
`relation to methanol, and the specific production values of
`protein in relation to biomass and methanol.
`
`MATERIAL AND METHODS
`
`Yeast Strain and Fermentation Processes
`
`Pichia pastoris strain GS 115 His+ Mut+ expressing mouse
`endostatin was prepared as described earlier (Boehm et al.,
`1999).
`Bench-top fermentation was performed in a 7-L fermen-
`tor (New Brunswick Scientific, Edison, NJ) The fermentor
`is interfaced to an MD-Biostat System (B. Braun Biotech
`USA, Allentown, PA) furnished with an adaptive control
`algorithm (Hsiao et al., 1992) that maintained dissolved
`oxygen at 30% saturation by adjusting the agitation and air
`or oxygen. The fermentor was also equipped with a metha-
`
`nol sensor (Trinh et al., 2000; Wagner et al., 1997). Three
`liters BMGY media (peptone, 2%; yeast extract, 1%; YNB,
`1.34%; glycerol, 3%; biotin, 4 × 10−5%; and 100 mM po-
`tassium phosphate pH 6.0) was inoculated with 100 mL of
`overnight culture (O.D. ≈ 20 at 600 nm). The culture was
`grown in batch mode for approximately 16 h (O.D. ≈ 60 at
`600 nm), after which fed-batch mode was started by adding
`a 50% glycerol solution at a flow rate of 18 mL L−l h−l;
`during this period, a dose of 0.3% (w/v) methanol was also
`pumped into the fermentor to calibrate the methanol sensor.
`After stabilization of the sensor signal (around 1 h), the
`glycerol feed rate was reduced gradually over a 1-hour pe-
`riod to 2 mL L−l h−l, allowing the culture to adapt smoothly
`to methanol. The reduced glycerol feed rate was kept con-
`stant throughout the methanol induction period (mixed
`feed). After the adaptation period, methanol was added ac-
`cording to the selected fed-batch strategy. The duration of
`the methanol addition phase was 48 h. During the fermen-
`tation the pH was kept at 6.5 using 50% NH4OH solution,
`and foam was controlled by the addition of antifoam 289
`(Sigma Chemical Co., St. Louis, MO.)—this antifoam did
`not affect growth or endostatin production.
`Pilot-scale fermentation was performed in 80-L fermen-
`tor (Bioflo 5000, New Brunswick Scientific, Edison, NJ)
`equipped with the same control and data acquisition as the
`bench top.
`
`Methanol Addition Strategies
`
`Fed-Batch Fermentation Based on On-Line
`Methanol Sensor
`
`After the adaptation period, the addition of methanol con-
`taining 12 mL/L−l PTMI solution (per L: cupric sulphate-
`5H2O 6.0 gm; sodium iodide 0.08 gm; manganese sulphate-
`H2O 3.0 gm; sodium molibdate-2H2O 0.2 gm; boric acid
`0.02 gm; cobalt chloride 0.5 gm; zinc chloride 20.0 gm;
`ferrous sulphate-7H2O 65.0 gm; biotin 0.2 gm; sulfuric acid
`5.0 gm) was initiated. Methanol concentration was con-
`trolled at 3.0 g/L−l by an on-line monitoring and control
`device described previously (Trinh et al., 2000; Wagner et
`al., 1997).
`
`Fed-Batch Fermentation Based on Dissolved
`Oxygen Signal
`
`After the adaptation period, methanol solution containing
`12 mL/L−l PTMI was added into the fermentor when the
`dissolved oxygen signal spiked 10% above the set-point.
`Each time methanol was added to adjust the methanol con-
`centration to 0.3%.
`
`Fed-Batch Fermentation Based on Predetermined
`Exponential Feed Rate
`
`After the adaptation period, methanol was pumped to the
`culture by a ChemTec pump (Scilog, Middleton, WI) at a
`
`TRINH ET AL.: EFFECT OF METHANOL FEEDING STRATEGIES ON RECOMBINANT MOUSE ENDOSTATIN
`
`439
`
`Motif Exhibit 1035, Page 2 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`

`

`rate based on the following relationship (Zhang et al.,
`2000a), F ⳱ 0.962 ␮X0V0e␮t where F is the methanol flow
`rate (g/L−l), X0 is the biomass-wet weight at induction
`(gWCW/L−l), V0 is culture volume at induction (L), ␮ is
`culture growth rate under methanol (h−l).
`The methanol flow rate was ramped exponentially to
`keep the culture growth rate at 0.02 h−1.
`
`Analytical Methods
`
`Off-Line Measurements
`
`Off-line methanol measurements were made using YSI Bio-
`chemistry Analyzer (Yellow Spring Instruments, Yellow
`Spring, OH).
`
`Endostatin Determination
`
`Samples taken during the course of induction were centri-
`fuged at 4000g for 40 min. The supernatant was diluted to
`adjust the conductivity to 5.5 mS/cm and the pH to 6.0 and
`was loaded on a 5 mLHiTrap 娂 SP HP (Amersham Phar-
`macia Biotech AB, Piscataway, NJ) at a ratio of 20 mg total
`protein per mL packed resin. The column was washed with
`10 column volumes of 20 mM phosphate buffer pH 6.0
`containing 50 mM NaCl buffer, and the endostatin was
`eluted with 3 column volumes of 20 mM phosphate con-
`taining 1.5M NaCl. Endostatin was analyzed by loading the
`various samples and increasing amount (0–8.6 ␮g) of
`known endostatin standard on 4–20% Tris-glycine SDS
`PAGE under reducing conditions and running at 125 volts
`for 90 min. After staining with Simply Blue娂 SafeStain
`(Invitrogen, Carlsbad, CA) the amount of endostatin was
`determined by digitising and calculating with NIH image
`analysis software (NIH, Bethesda, MD).
`
`Dry Weight Determination
`
`Dry weight determination was done by drying cell paste in
`a moisture determination balance (Ohaus MB 200, NJ) for
`3 h at 95°C.
`
`Statistical Analysis
`
`Each of the methanol feeding strategies was tested in three
`repetitions. Standard deviation was calculated for the aver-
`age values of the three experiments from each feeding strat-
`egy.
`
`RESULTS
`
`Effect of the Various Fed-Batch Strategies on
`Growth Rate, Biomass Accumulation, and
`Methanol Supply
`
`Growth rate and biomass accumulation during the three dif-
`ferent fed-batch strategies are shown in Figure 1 and Table
`
`Figure 1. Pichia pastoris growth in response to different methanol sup-
`plies strategies. On line sensor (䊊), DO signal (䊉), predetermined rate (䉭).
`(Time zero indicates the initiation of the methanol induction phase). (A)
`Growth rate, (B) Biomass accumulation. Fermentation was conducted in a
`7-L fermentor with initial volume of 3.0 L. Methanol addition was started
`when the biomass concentration was between 100 and 120 gram per liter
`(wet weight).
`
`I. The growth rate during the predetermined methanol ad-
`dition strategy was kept constant at a value of 0.02 h−1, but
`the growth rates during the DO control strategy and the
`methanol sensor control strategy followed different pat-
`terns. The rate was 0.06 h−1 and 0.05 h−1, respectively 10 h
`after the induction, and the rate continuous to decline
`throughout the induction phase which lasted 48 h. The bio-
`mass accumulation rate was corresponded to the growth
`rate. It accumulated at a constant rate during the predeter-
`mined addition strategy reaching a final level of 434 g DCW
`(dry cell weight). During the DO control strategy and the
`methanol sensor strategy, the biomass accumulation was
`much faster at the beginning of the induction phase, and
`slowed down towards the end of the induction period, reach-
`ing a value of 986 and 988 gDCW, respectively, which is
`over 2 times as much as accumulated in the predetermined
`growth rate strategy.
`Methanol addition is described in Figure 2 and Table I.
`When the methanol concentration in the media was con-
`trolled by an online sensor, the total amount added was 3.23
`L, compared with 2.25 L added in response to the dissolved
`oxygen concentration and to less than 1 L when the metha-
`nol was added using the predetermined rate.
`
`440
`
`BIOTECHNOLOGY AND BIOENGINEERING, VOL. 82, NO. 4, MAY 20, 2003
`
`Motif Exhibit 1035, Page 3 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`

`

`Table I. Comparison of the various methanol feeding strategies. Fermentations were conducted in a bench top fermentor with initial volume of 3 L, the
`numbers reflect the values after 48 h induction.
`
`Methanol control strategy
`
`Final culture
`volume (L)
`⳱ 3.0 L
`V0
`
`Fed-batch using D.O. signal
`Fed-Batch using online gas sensor
`Fed-batch using predetermined feed rate
`
`6.09
`7.06
`4.25
`
`Endostatin
`
`Biomass
`
`(mg/L
`culture)
`
`66 ± 4
`67 ± 19
`96 ± 16
`
`Total
`(mg)
`
`402
`470
`408
`
`(gDCW/L)
`
`162 ± 2.6
`140 ± 2.7
`102 ± 1.7
`
`Total
`g DCW
`
`Methanol
`consumed (L)
`
`Base
`consumed (L)
`
`986
`988
`434
`
`2.25 ± 0.04
`3.23 ± 0.15
`0.90 ± 0.11
`
`0.84
`0.83
`0.34
`
`Effect of the Various Fed-Batch Strategies on
`Endostatin Production
`
`Total and volumetric endostatin productions during the dif-
`ferent fed batch strategies are described in Figure 3 and
`Table I. The total amount of endostatin produced was simi-
`lar in the fed-batch strategies using the dissolved oxygen
`signal and the predetermined feed rate. Approximately 400
`mg were accumulated using these two strategies. Higher
`endostatin, 470 mg, was accumulated using the on-line sen-
`sor fed-batch strategy. Endostatin concentration was higher
`in the predetermined methanol feeding strategy compared
`with the dissolved oxygen control and the methanol sensor
`strategies. Final concentration of 96 mg/L was detected with
`the predetermined feed rate, compare with concentrations of
`66 mg/L and 67 mg/L in the other two strategies.
`
`cific endostatin production on methanol with the predeter-
`mined feeding rate was much higher than it was with the
`two other fed-batch strategies. It was 489 mg/L of methanol
`with the predetermined feeding rate compared with 160 mg/
`L, and 116 mg/L with the dissolved oxygen controlled strat-
`egy and the methanol sensor strategy, respectively. 0.72 mg
`endostatin per gram dry cell weight were produced with the
`predetermined feeding rate strategy compared with 0.31 and
`0.33 mg endostatin per gram dry cell weight with the two
`other feeding strategies. The specific biomass production
`was 1.67 gram dry cell weight per gram carbon used
`(methanol and glycerol) with the predetermined feeding rate
`and 1.34 and 0.96 gram dry cell weight per gram carbon
`used, at the dissolved oxygen-controlled and the sensor-
`controlled strategies, respectively.
`
`Specific Production Values, Specific Consumption
`Values, and Production Rates in the Three
`Different Growth Strategies
`
`Specific production of endostatin based on methanol utili-
`zation, biomass production based on carbon utilization
`(methanol and glycerol), and specific production of end-
`ostatin based on biomass are summarized in Table II. Spe-
`
`Figure 2. Methanol consumption in response to different supply strate-
`gies. On line sensor (– – – –), DO signal (— - — –), predetermined rate
`(—-). (Time zero indicates the initiation of the methanol induction phase).
`
`Figure 3. Endostatin production in response to different methanol sup-
`plies strategies. On-line sensor (䊊), DO signal (䊉), predetermined rate
`(䉭). (Time zero indicates the initiation of the methanol induction phase).
`(A) Total endostatin production, (B) volumetric endostatin production.
`
`TRINH ET AL.: EFFECT OF METHANOL FEEDING STRATEGIES ON RECOMBINANT MOUSE ENDOSTATIN
`
`441
`
`Motif Exhibit 1035, Page 4 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`

`

`Table II. Specific endostatin and biomass production at different methanol feeding strategies. The
`yield constants reflect average values during the last 24 h of the induction phase.
`
`Methanol control strategy
`
`Specific endostatin production
`
`(mg Endostatin/
`1 methanol)
`
`(mg Endostatin/
`g DCW)
`
`Specific biomass
`production
`(g DCW/g carbon)a
`
`Fed-batch using D.O. signal
`Fed-batch using on-line gas sensor
`Fed-batch using predetermined feed rate
`
`160 ± 6
`116 ± 15
`489 ± 8
`
`0.31 ± 0.03
`0.33 ± 0.09
`0.72 ± 0.01
`
`1.34 ± 0.11
`0.96 ± 0.07
`1.67 ± 0.16
`
`aSpecific biomass production was based on the total utilization of carbon in the methanol and
`glycerol.
`
`Pilot production of endostatin using
`predetermined feeding rate
`
`Based on the above information, a pilot-scale production
`(10 times bigger) was conducted using the preferred growth
`conditions. Typical process is shown in Figure 4; after 48 h
`induction 6.8 L of methanol were added, producing 2.7 g
`endostatin and 9.3 kg biomass (wet weight). Specific end-
`ostatin production yield was 25% lower than the value ob-
`tained in the bench-top fermentor, total consumption of
`methanol and total biomass production were also 25%
`lower.
`
`DISCUSSION
`
`Due to the sensitivity of P. pastoris to methanol concentra-
`tion, production of recombinant proteins under the control
`of the AOX1 promoter is done by implementing different
`fed-batch strategies controling the methanol concentration.
`Growth characteristics can be different at the various strat-
`egies, and therefore, recombinant protein production can
`also be affected. In addition, different values of metabolic
`indicators (e.g., methanol concentration, dissolved oxygen
`
`Figure 4. Pilot scale production of endostatin using predetermined
`methanol addition strategy: Endostatin production (䊉), methanol addition
`(䊏), biomass production (䉱). (Time zero indicates the initiation of the
`methanol induction phase).
`
`concentration) and different methanol feeding rates and pro-
`files can be selected. These parameters can potentially affect
`growth and recombinant protein production within each
`strategy. Establishing optimal protein expression protocols
`should therefore include the evaluation of various growth
`and induction strategies in addition to other parameters such
`as pH and temperature. In this work, we examine the effect
`of three fed-batch strategies on the production of recombi-
`nant murine endostatin, evaluating not only the amount of
`endostatin produced but also the methanol consumption,
`biomass production, and the specific production values.
`Two strategies were based on the metabolic activity of the
`yeast (methanol and oxygen consumption) and one was
`based on supplying methanol at a predetermined exponen-
`tial rate (limited methanol supply). The total production of
`endostatin with the three growth strategies was similar. Ap-
`proximately 400 mg of endostatin were produced from a 3
`L initial volume, but the amount of methanol used and the
`biomass produced were different, affecting final volume
`and specific production values of both endostatin and bio-
`mass.
`Feeding the methanol at a predetermined exponential pro-
`file and controling the growth rate at 0.02 h−1 was more
`efficient than the other two strategies. The amount of metha-
`nol used and the biomass produced were lower, causing the
`specific production of endostatin per biomass unit to be over
`2 times higher than the dissolved oxygen controlled strategy
`or the sensor-based controlled strategy. Endostatin produc-
`tion per methanol used was almost 3 times higher in the
`predetermined strategy compared with the dissolved oxygen
`controlled strategy or the sensor strategy. A similar trend
`was observed with the specific biomass production. It was
`higher with the predetermined rate strategy than with the
`other two.
`Methanol, the main carbon source is used both for energy
`generation and anabolism (Jahic et al., 2002). The higher
`efficiency of methanol utilization at the predetermined ex-
`ponential rate suggests that the methanol is directed mainly
`towards energy generation, and that only a small portion is
`directed to biomass production. Since endostain biosynthe-
`sis is controlled by the same promoter responsible for the
`synthesis of the methanol oxidase, the amount of endostatin
`produced should be related to the methanol oxidase pro-
`duced and should not be affected by the low methanol uti-
`
`442
`
`BIOTECHNOLOGY AND BIOENGINEERING, VOL. 82, NO. 4, MAY 20, 2003
`
`Motif Exhibit 1035, Page 5 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`

`

`lization. It was also mentioned (Cregg, 1999) that when
`cells are fed methanol at a growth-limiting rate, the AOX1 is
`induced to levels 3 to 5 times higher than in cells growing
`in excess methanol. Methanol consumption and biomass
`production were lower in the dissolved oxygen-controlled
`strategy than in the sensor-controlled strategy. This fact
`supports the above explanation since the methanol supply is
`not continuously unlimited as it is in the sensor controlled
`strategy.
`Throughout the induction phase glycerol was supplied
`continuously to the culture at the rate of 1 gram per liter per
`hour. At this rate there was no glycerol accumulation. This
`supplemental carbon source is channelled directly to anabo-
`lism, (Jahic et al., 2002) and therefore compensate for pos-
`sible shortage in carbon source as a result of methanol uti-
`lization for energy. While the glycerol supply may not be
`essential for the culture grown at excess methanol, it may be
`essential for the culture grown at limited methanol supply.
`The results obtained from this comparison indicate that a
`growth strategy based on predetermined exponential feed-
`ing of methanol has an advantage for the production process
`of recombinant endostatin. The amount of methanol used
`and the biomass accumulation are lower than the values
`obtained using the dissolved oxygen or the methanol sensor
`methods, simplifying the endostatin recovery. In addition,
`the controlled growth rate limited the oxygen consumption
`and the heat production. It seems that these advantages are
`not necessarily specific for recombinant endostatin produc-
`tion. It is expected that any recombinant protein under the
`control of the AOX1 promoter should respond in the same
`manner.
`The high biomass concentration, above 50% wet weight,
`produced during the Pichia pastoris fermentation interferes
`with the recovery of the recombinant protein produced
`(Thommes et al., 2001; Trinh et al., 2000), especially when
`the protein is secreted to the media. In this aspect the pre-
`determined addition methanol strategy offers another ad-
`vantage, in addition to the high productivity, it simplifies
`the recovery process of the secreted endostatin due to the
`lower biomass produced and the lower overall volume.
`
`References
`
`Boehm T, Pirie-Shepherd S, Trinh LB, Shiloach J, Folkman J. 1999. Dis-
`ruption of the KEX1 gene in Pichia pastoris allows expression of
`full-length murine and human endostatin. Yeast 15:563–572.
`Brierley RA, Siegel RS, Bussineau CM, Craig WS, Holtz GC, Davis GR,
`Buckholtz RG. 1990. Mixed feed recombinant yeast fermentation. Pat-
`ent Cooperation Treaty No. WO 90/03431 US.
`Byrne MP, Titball RW, Holley J, Smith LA. 2000. Fermentation, purifi-
`cation, and efficacy of a recombinant vaccine candidate against botu-
`linum neurotoxin type F from Pichia pastoris. Prot Expr Puri 18:
`327–337.
`Cereghino JL, Cregg JM. 2000. Heterologous protein expression in the
`methylotrophic yeast Pichia pastoris. FEMS Microbiol Rev 24:45–66.
`Chauhan AK, Arora D, Khanna N. 1999. A novel feeding strategy for
`enhanced protein production by fed-batch fermentation in recombinant
`Pichia pastoris. Proc Biochem 34:139–145.
`
`Chung JD. 2000. Design of metabolic feed controllers: Application to high
`density fermentation of Pichia pastoris: Mathematical model for fed-
`batch high cell density culture. Biotechnol Bioeng 68:298–307.
`Cregg JM, Cereghino JL, Shi J, Higgins DR. 2000. Recombinant protein
`expression in Pichia pastoris. Molec Biotechnol—Part B 16:23–52.
`Cregg JM. 1999. Expression in the methylotrophic yeast Pichia pastoris.
`In: Fernander JM, Hoeffler JP, editor. Gene expression systems. San
`Diego: Academic Press. p. 157–191.
`Curvers S, Brixius P, Klauser T, Thommes J, Weuster-Botz D, Takors R,
`Wandrey C. 2001. Human chymotrypsinogen B production with Pi-
`chia pastoris by integrated development of fermentation and down-
`stream processing. Part 1. Fermentation. Biotechnol Prog 17:495–502.
`d’Anjou MC, Daugulis AJ 1997. A model-based feeding strategy for fed-
`batch fermentation of recombinant Pichia pastoris. Biotech Tech 11:
`865–868
`Faber KN, Harder W, AB G, Veenhuis M. 1995. Review: Methylotrophic
`yeasts as factories for the production of foreign proteins. Yeast 11:
`1331–1344.
`Figg WD, Kruger, EA, Price DK, Kim S, Dahunt WD. 2002. Inhibition of
`angiogenesis: Treatment potions for patients with metastatic prostate
`cancer. Inv New Drugs 20:183–194.
`Freyre FM, Vazquez JE, Ayala M, Canaan-Haden L, Bell H, Rodriguez I,
`Gonzalez A, Cintado A, Gavilondo JV. 2000. Very high expression of
`an anti-carcinoembryonic antigen single chain Fv antibody fragment in
`the yeast Pichia pastoris. J Biotechnol 76:157–163.
`Guarna MM, Lesnicki GJ, Tam BM, Robinson J, Radziminski CJ, Hasen-
`winkle D, Boraston A, Jervis E, Macgillivray RTA, Turner RFB, Kil-
`burn DG. 1997. On-line monitoring and control of methanol concen-
`tration in shake-flask cultures of Pichia pastoris. Biotechnol Bioeng
`56:297–286.
`Hellwig S, Emde F, Raven NPG, Henke M, Van der Logt P, Fischer R.
`2000. Analysis of single chain antibody production in Pichia pastoris
`using on-line methanol control in fed-batch and mixed-feed fermen-
`tation. Biotechnol Bioeng 74:344–352.
`Higgins DR, Cregg JM. 1998. Pichia protocols. Methods in molecular
`biology. Totowa, NJ: Humana Press Inc.
`Hsiao J, Ahluwalia M, Kaufman JB, Clem TR, Shiloach J. 1992. Adaptive
`control strategy for maintaining dissolved oxygen concentration in
`high density growth of recombinant E. coli. Ann NY Acad Sci 665:
`320–333.
`Inan M, Chiruvolu V, Eskridge KM, Vlasuk GP, Dickerson K, Brown S,
`Meagher MM. 1999. Optimization of temperature-glycerol-pH condi-
`tions for a fed-batch fermentation process for recombinant hookworm
`(Ancylostoma caninum) anticoagulant peptide (AcAP-5) production by
`Pichia pastoris. Enzyme MicrobTechnol 24:438–45.
`Jahic M, Rotticci-Mulder JC, Matinelle M, Hult K, Enfors S-O. 2002.
`Modeling of growth and energy meyabolism of Pichia pastoris pro-
`ducing a fusion protein. Bioprocess Biosyst Eng 24:385–393.
`Katakuga Y, Zhang W, Zhuang G, Omasa T, Kishimoto M, Goto Y, Suga
`KI. 1998. Effect of methanol concentration on the production of hu-
`man beta-2-glycoprotein I domain V by a recombinant Pichia pastoris:
`A simple system for control of methanol concentration using a semi-
`conductor gas sensor. J Ferment Bioeng 86:482–487.
`Kobayashi K, Kuwae S, Ohya T, Ohda T, Ohyama M, Tomomitsu K. 2000.
`High level secretion of recombinant human serum albumin by fed-
`batch fermentation of the methylotrophic yeast, Pichia pastoris, based
`on optimal methanol feeding strategy. J Biosci Bioeng 90:280–288.
`Kupcsulik B, Sevella B, Ballagi A, Kozma J. 2001. Evaluation of three
`methanol feed strategies for recombinant Pichia pastoris Muts fermen-
`tation. Acta Alimentaria 30:99–111.
`Minning S, Serrano A, Ferrer P, Sola C, Schmid RD, Valero F. 2001.
`Optimization of high-level production of Rhizopus oryzae lipase in
`Pichia pastoris: Enzyme production using plasmid pPICZ-alpha-A-
`ROL. J Biotechnol 86:59–70.
`Murasugi A, Tohma-Aiba Y, Asami Y. 2000. Production of recombinant
`
`TRINH ET AL.: EFFECT OF METHANOL FEEDING STRATEGIES ON RECOMBINANT MOUSE ENDOSTATIN
`
`443
`
`Motif Exhibit 1035, Page 6 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`

`

`human midkine in yeast, Pichia pastoris: Induction by methanol in
`high cell density fermentation. Biosci Bioeng 90:395–399.
`O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, Flynn E,
`Birkhead JR, Olsen BR, Folkman J. 1997. Endostatin: An endogenous
`inhibitor of angiogenesis and tumor growth. Cell 88:277–285.
`Tschopp JF, Brust PF, Cregg JM, Stillman CA, Gingeras TR. 1987. Ex-
`pression of Lac Z gene from two methanol regulated promoters in
`Pichia pastoris. Nucleic Acid Res. 15:3859–3876.
`Thommes J, Halfar M, Gieren H, Curvers S, Takors R, Brunschier R, Kula
`MR. 2001. Human chymotrypsinogen B production from Pichia pas-
`toris by integrated development of fermentation and downstream pro-
`cessing. Part 2. Protein recovery. Biotechnol Prog 17:503–512.
`Trinh LB, Noronha SB, Fannon M, Shiloach J. 2000. Recovery of mouse
`
`endostatin produced by Pichia pastoris using expanded bed adsorp-
`tion. Bioseparation 9:223–230.
`Wagner LW, Matheson NH, Heisey RF, Schineder K. 1997. Use of a
`silicone tubing sensor to control methanol concentration during fed
`batch fermentation of P.pastoris. Biotechnol Tech 11:791–795.
`Zhang W, Bevins MA, Plantz BA, Smith LA, Meagher MM. 2000a. Mod-
`eling Pichia pastoris growth on methanol and optimizing the produc-
`tion of a recombinant protein, the heavy-chain fragment C of botuli-
`num neurotoxin, serotype A. Biotechnol Bioeng 70:1–8.
`Zhang W, Inan M, Meagher MM. 2000b. Fermentation strategies for re-
`combinant protein expression in the methylotrophic yeast Pichia pas-
`toris. Biotechnol Bioproc Eng 5:275–287
`
`444
`
`BIOTECHNOLOGY AND BIOENGINEERING, VOL. 82, NO. 4, MAY 20, 2003
`
`Motif Exhibit 1035, Page 7 of 7
`
`Case No.: IPR2023-00322
`U.S. Patent No. 10,273,492
`
`