Trehalose suppresses antibody aggregation during the culture of Chinese hamster ovary cells

Journal of Bioscience and Bioengineering VOL. 117 No. 5, 632e638, 2014 www.elsevier.com/locate/jbiosc

Trehalose suppresses antibody aggregation during the culture of Chinese hamster ovary cells Masayoshi Onitsuka,1, * Miki Tatsuzawa,2 Ryutaro Asano,3 Izumi Kumagai,3 Akihiro Shirai,1 Hideaki Maseda,1 and Takeshi Omasa1, 4 Institute of Technology and Science, The University of Tokushima, Minamijosanjima-cho 2-1, Tokushima 770-8506, Japan,1 Advanced Technology and Science, The University of Tokushima, Minamijosanjima-cho 2-1, Tokushima 770-8506, Japan,2 Department of Biomolecular Engineering, Graduate School of Engineering, Tohoku University, 6-6-11-606 Aoba-yama, Aramaki, Aoba-ku, Sendai 980-8579, Japan,3 and Department of Biotechnology, Graduate School of Engineering, Osaka University, Yamadaoka 2-1, Suita, Osaka 565-0871, Japan4 Received 1 October 2013; accepted 29 October 2013 Available online 4 December 2013

The aggregation of therapeutic antibodies during the manufacturing process is problematic because of the potential risks posed by the aggregates, such as an unexpected immune response. One of the hallmark effects of trehalose, a disaccharide consisting of two alpha-glucose units, is as a chemical chaperone with anti-aggregation activity. In this study, Chinese hamster ovary (CHO) cell line producing a diabody-type bispecific antibody were cultured in medium containing trehalose and the aggregation of the secreted proteins during the culture process was analyzed. An analysis of the various forms of the antibody (monomeric, dimeric, and large aggregates) showed that trehalose decreased the relative content of large aggregates by two thirds. The aggregation kinetics indicated that trehalose directly inhibited the polymerization and aggregation steps in a nucleation-dependent aggregation mechanism. Moreover, both specific and volumetric antibody production were increased in CHO cells cultured in trehalose-containing medium. Thus, the addition of trehalose to recombinant CHO cell cultures would offer a practical strategy for quality improvement in the production of therapeutic antibodies. Ó 2013, The Society for Biotechnology, Japan. All rights reserved. [Key words: Chinese hamster ovary cell; Cell culture; Antibody production; Antibody aggregation; Chemical chaperone; Trehalose; Bispecific diabody]

In the manufacture of therapeutic proteins, aggregation is a common problem. The aggregates take on diverse forms with respect to their size, reversibility, solubility, covalent/non-covalent interactions, and native/non-native conformations (1e3). These structural changes are important because they can cause a loss of potency of the intact proteins. Moreover, aggregation and misfolding can induce a new and cryptic epitope presentation, resulting in an undesirable immune response (4,5). Stable biological activities must be ensured in therapeutic proteins. Additionally, harmful immune responses following the administration of aggregated proteins must be avoided. These views highlight the need to control and prevent aggregation in the manufacturing process. Chinese hamster ovary (CHO) cells are one of the most important industrial mammalian cell lines because of their use in GMPcertified recombinant protein production and in the development of industrial serum-free media as well as their reported production rates of more than 10 g recombinant antibody/L cultured cells (6e10). The aggregation of therapeutic proteins in cell culture has been reported in several studies (1e3). Aggregates are thought to

* Corresponding author. Tel.: þ81 88 656 7408; fax: þ81 88 656 9148. E-mail address: [email protected] (M. Onitsuka).

form within the cell interior and following secretion of the recombinant protein into the medium. The over-accumulation of polypeptides in the endoplasmic reticulum leads to misfolding and aggregation, resulting in the induction of the unfolded protein response (UPR). Indeed, cell engineering with UPR-related genes is a useful strategy for preventing aggregation and enhancing the production of recombinant proteins (11e13). Secreted proteins can aggregate in response to physicochemical stresses, such as changes in the pH and osmolality of the medium or in the cultivation temperature and time. It is therefore important to control the culture conditions to suppress aggregation. However, culture conditions are typically optimized for cell growth and protein production rather than to suppress aggregation, which therefore remains problematic. In contrast, the use of chemical chaperones is an attractive strategy to avoid aggregation. Osmolytes in the form of small organic additives, such as sugars, polyols, and amino acids, serve as chemical chaperones to stabilize proteins and inhibit aggregation (14e16). These properties suggest the use of these compounds to inhibit the formation of protein aggregates in culture media, as previously described in other studies (17,18). However, the mechanisms underlying aggregation and the ability of chemical chaperones to inhibit aggregate formation in cell culture are unclear. The osmolyte trehalose is widely distributed in many organisms,

1389-1723/$ e see front matter Ó 2013, The Society for Biotechnology, Japan. All rights reserved. http://dx.doi.org/10.1016/j.jbiosc.2013.10.022

VOL. 117, 2014

SUPPRESSED ANTIBODY AGGREGATION BY TREHALOSE

including bacteria, fungi, insects, plants, and invertebrates. A hallmark of this non-reducing sugar, in which two glucose units are joined by an a-1,1 linkage, is its chemical chaperone effects, as it protects biomolecules from dehydration, freezing, osmotic shock, oxidation, and heating (19e21). Trehalose is used as a protectant for several commercially available therapeutic antibodies and proteins, such as those marketed under the names Herceptin, Avastin, Lucentis, and Advate (22). Thus far, trehalose has been added only in the formulation process whereas its stabilization and anti-aggregation properties suggest its addition to the culture medium to protect secreted proteins against physicochemical stresses. In addition, trehalose should not be harmful to cell growth because, as noted above, it is a natural osmolyte with widespread occurrence. In this study, we examined the effects of trehalose on the cultivation of recombinant CHO Top-H cell line (23,24) producing Ex3-scDb-Fc. This humanized immunoglobulin G (IgG)-like diabody-type bispecific antibody retargets lymphokine-activated killer T cells to attack cells expressing the epidermal growth factor receptor (25,26). Bispecific diabodies are promising candidates as next-generation therapeutic antibodies because of their dual functionality. By applying trehalose to the CHO cell culture process, we were able to suppress antibody aggregation and to gain insight into the possible mechanism. MATERIALS AND METHODS Cell culture, antibody production, and purification The CHO Top-H cell line producing Ex3-scDb-Fc was cultivated as a suspension culture in serum-free ExCD medium [mixture of ExCell 302 (SAFC Bioscience, St. Louis, MO, USA) and IS CHO-CD (Irvine Scientific, Santa Ana, CA, USA), supplemented with 1 mM methotrexate and 1 mM G418]. The cells were adapted to trehalose-containing ExCD medium in 500-mL polycarbonate Erlenmeyer flasks (Corning Inc., Corning, NY, USA) containing a working volume of 80 mL of the same medium. The flask was incubated at 37 C and 80 rpm in an orbital Climo-shaker ISF1-X (Kuhner, Basel, Switzerland). Trehalose was kindly supplied by Hayashibara Biochemical Laboratories, Inc. (Okayama, Japan). The adapted cells were cultivated at a temperature of 37 C in a 1-L glass bioreactor (Biott, Tokyo, Japan) containing 750 mL of the abovedescribed medium. The agitation speed was 70 rpm, the headspace of the vessel was aerated with air supplied at a flow rate of 100 mL/min, and the pH was maintained at 7.1. The dissolved oxygen (DO) concentration was measured by a DO sensor (InPro 6880, Mettler Toledo, Zurich, Switzerland) and was always kept above 40% of air saturation. Cell concentration was quantitatively determined using the Vi-cell XR cell viability analyzer (Beckman Coulter, Fullerton, CA, USA). Antibody concentrations in the cultivation medium were determined by a sandwich enzyme-linked immunosorbent assay (ELISA) or a chemiluminescent assay based on AlphaLISA technology (PerkinElmer, MA, USA). In the former, goat anti-human IgG-Fc and horseradish-peroxidase-conjugated goat anti-human IgG-Fc antibodies (Bethyl Laboratories, Montgomery, TX, USA) were used as the capture and detection antibodies, respectively. The absorbance change at 405 nm was measured with an Infinite M200 microplate reader (Tecan, Grödig, Austria). In the latter, the concentration of Ex3-scDb-Fc in the culture medium was determined using an EnSpire Alpha plate reader with the human IgG kit (PerkinElmer). Glucose and lactate concentrations were measured with the BioFlow BF-7 biosensor (Oji Scientific Instruments, Hyogo, Japan). The kinetic parameters [specific growth rate (m), specific antibody production rate (rAb), specific glucose consumption rate (rGluc), specific lactate production rate (rLac)] were calculated as previously described (27). The Ex3-scDb-Fc antibody used in biophysical measurements was purified from the culture supernatant on a HiTrap protein A affinity chromatography with AKTA Prime Plus (GE Healthcare, Buckingham, UK). Eluted fractions, including the antibody, were dialyzed against equilibration buffer [25 mM Tris, 100 mM NaCl, 1 mM EDTA (pH 7.5)] and then analyzed by size-exclusion chromatography (SEC) using a Sephacryl S-300 column with AKTA Prime Plus system (GE Healthcare). The running buffer was the same as that in the dialysis step. The final purity was determined by SDS-PAGE. Purified Ex3-scDb-Fc concentrations were determined with the BCA protein assay kit (Thermo Scientific, Rockford, IL, USA). Quantitative real-time PCR Total RNA was isolated from 5-day suspension cultures of CHO Top-H cells using the High Pure total RNA isolation kit (Roche, Basel, Switzerland). The cDNA was synthesized using the Prime Script II first-strand cDNA synthesis kit (Takara Bio, Otsu, Japan). Real-time PCR was performed using the Thunderbird SYBR qPCR mix (Toyobo Life Science, Osaka, Japan) and the Step One Plus real-time PCR system (Applied Biosystems, Foster City, CA, USA). The PCR primers were as follows: 50 -AGGAGTACAAGTGCAAGGTCTCCAC-30 and 50 ACCTGGTTCTTGGTCAGCTCATCC-30 for Ex3-scDb-Fc; 50 -ACTCCTACGTGGGTGACGAG30 and 50 -AGGTGTGGTGCCAGATCTTC-30 for CHO ACTB (b-actin). The efficiency of

633

reverse transcription was verified by standardization with the housekeeping gene ACTB (b-actin), and mRNA levels were quantified from the Ct based on the standard curve method. Circular dichroism and fluorescence spectrum measurements Far-UV circular dichroism (CD) spectra were measured using a Jasco J-820 spectropolarimeter (Jasco, Tokyo, Japan) with a quartz cell of 1-mm path length. Antibody concentrations were prepared at 0.4 mg/mL in 25 mM Tris, 100 mM NaCl, and 1 mM EDTA (pH 7.5). The temperature of the samples was kept at 20 C by a Peltier temperature controller (PYC-347WI; Jasco). Fluorescence spectra were measured at an excitation wavelength of 445 nm using a Hitachi F-7000 fluorescence spectrophotometer (Hitachi, Tokyo, Japan). Each sample contained 0.1 mg antibody/mL and 10 mM thioflavin-T [(ThT) Wako Pure Chemicals (Osaka, Japan)] in the abovedescribed buffer. In aggregation kinetics experiments, concentrations of monomeric Ex3-scDb-Fc were prepared at 0.1 mg/mL in 25 mM sodium citrate, 100 mM NaCl, 1 mM EDTA and 10 mM ThT (pH 4.1) with and without 200 mM trehalose.

RESULTS Adaptation of recombinant CHO cells to trehalosecontaining medium and bioreactor cultivation The anti-aggregation effects of 200 mM trehalose during the culture process were examined in CHO Top-H cells cultured in serum-free medium. However, these cells, which were directly inoculated cells into the medium, were apparently subjected to abrupt hyperosmotic stress and cell growth was thereby prevented. We therefore attempted to adapt the cell line to trehalose-containing medium by increasing the concentration of trehalose stepwise in 50-mM increments. At each concentration, the time at which the specific growth rate and the passage interval became constant was defined as adaptation. Accordingly, for proliferating cell cultures we chose a final trehalose concentration of 150 mM, corresponding to a culture medium osmolality of 480 mOsm/kg compared with 319 mOsm/kg in the medium without trehalose. The adaptation to 150 mM trehalose required about 40 days; this concentration was the maximum concentration that resulted in adaptation. The time courses of viable cell density and antibody productivity in flask cultures are shown in Fig. 1A and B, respectively. The estimated kinetic parameters of the recombinant CHO cell culture, the specific growth rate (m; 1/h), and the specific production rate (r; pg/ cell/day) are listed in Table 1. The properties of cells cultured in 150 mM trehalose included: (i) a decrease in both the specific growth rate and the maximum cell density, (ii) a prolonged life span, and (iii) increased specific and volumetric antibody productivity. These are common characteristics of mammalian cells cultured in a hyperosmotic medium (18,28e30). We then cultured the same cells in a bioreactor and monitored the effects of trehalose on culture performance under controlled conditions. The viable cell density, Ex3-scDb-Fc concentration, and glucose and L-lactate concentrations are shown in Fig. 1CeE, respectively. The estimated kinetic parameters of the culture are summarized in Table 1. Bioreactor operation using a medium containing 150 mM trehalose qualitatively reproduced the properties of the flask cultures; that is, a decrease in the maximum cell density and enhanced antibody production. However, there were quantitative differences in the performance of the two culture systems. Thus, while in the bioreactor, the specific cell growth rate in the presence of 150 mM trehalose was almost identical to that without trehalose (Table 1); the rate of specific antibody production was enhanced to a lesser extent than in the flasks (Table 1). The reasons for the attenuated effects of trehalose on culture performance are unclear but may involve the pH and DO concentration of the controlled bioreactor conditions. Metabolic parameters were also changed by trehalose addition (Fig. 1E and Table 1). Specifically, the rates of glucose consumption and L-lactate production were enhanced, suggesting that CHO Top-H cells require a high energy supply under hyperosmotic conditions and that osmotic stress shifts cell metabolism to the increased production of L-lactate. In both the flask and the reactor cultivations, the presence of 150 mM trehalose greatly

ONITSUKA ET AL.

J. BIOSCI. BIOENG.,

B

60 50 40

5

30 20 10 0

60 50 40 30 20 10 0

0

100

200

300

0

400

100

200 Time (h)

Time (h)

D [Ex3-scDb-Fc] (mg/L)

60 50 40 30

5

(x10 cells/mL)

Viable cell density

C

20 10 0 0

50

100

150

200

250

20 15 10 5 0

300

0

50

100

4 3 2 1 0

F

3

5

L-lactate (g/L)

Glucose (g/L)

6

0

100

200

300

2 1 0

0

400

25

Time (h)

E

300

30

100

200

300

150

250

300

7 6 5 4 3 2 1 0 w/o

Time (h)

Time (h)

200

Time (h) Relative mRNA level

(x10 cells/mL)

Viable cell density

A

[Ex3-scDb-Fc] (mg/L)

634

150mM

FIG. 1. Culture profiles of CHO Top-H cells. In each panel, open and filled circles represent the profiles without and with 150 mM trehalose, respectively. (A) Viable cell densities and (B) antibody concentrations in flask cultures. Error bars show the standard error of the mean of three independent culture samples. (C) Viable cell densities and (D) antibody concentrations in bioreactor cultures. (E) Glucose concentrations and L-lactate concentrations in bioreactor cultures. (F) Real-time PCR analysis of Ex3-scDb-Fc mRNA. Relative mRNA levels were normalized to the levels of b-actin in CHO cells. Error bars show the standard error of three independent analyses.

enhanced Ex3-scDb-Fc production (Fig. 1B and D). Real-time quantitative PCR analysis of the mRNA level of the bispecific diabody at 5-day suspension culture (Fig. 1F) showed that in bioreactor cells the addition of 150 mM trehalose increased Ex3-scDb-Fc relative mRNA expression 6-fold and likely accounted for the enhancement of antibody production by trehalose. Anti-aggregation effects of trehalose during the culture of CHO cells The culture medium was harvested from flask and bioreactor cultures and Ex3-scDb-Fc was purified from the supernatants by protein A affinity chromatography. The aggregation state of the antibody was then examined by SEC analysis. It should be noted that antibody exposure to the low pH conditions of affinity chromatography can induce both a conformational change and aggregation (1,3) such that the results of the analysis are

ambiguous. To avoid this problem, 1 M Arg-HCl (pH 4.2) was used as the elution buffer in the protein A purification process (31). Arginine is an effective additive for protein refolding, increasing protein solubility and preventing protein aggregation in affinity purifications. We confirmed that the monomeric form of Ex3scDb-Fc re-purified by protein A chromatography using 1 M ArgHCl (pH 4.2) showed no tendency to aggregate and retained its monomeric form (Fig. S1). The SEC analysis of Ex3-scDb-Fc in flask and bioreactor cultures is shown in Fig. 2A and B, respectively. Under both culture conditions we observed three states of Ex3-scDb-Fc: monomeric, dimeric, and large aggregated forms, with retention peaks at 71, 58, and 42 mL, respectively. The presence of these three forms in the culture supernatants indicated that the dimers and large

TABLE 1. Kinetic parameters of CHO cell cultures with and without trehalose.

Without trehalose Flask Reactor þ150 mM trehalose Flask Reactor a b

Mean  S.D. (n ¼ 3). Not analyzed.

Growth rate (m; 102 1/h)

Antibody production rate (rAb; pg/cell/day)

Glucose consumption rate (rGluc; ng/cell/day)

3.07  0.18a 2.39

0.39  0.02a 0.42

N.A.b 0.31

N.A.b 0.30

1.51  0.04a 2.10

1.55  0.03a 0.83

N.A.b 0.49

N.A.b 0.52

L-Lactate

production rate (rLac; ng/cell/day)

VOL. 117, 2014

SUPPRESSED ANTIBODY AGGREGATION BY TREHALOSE

A

B

120

120

(mAU)

100

80

60

280

60

80

40

40

A

A

280

(mAU)

100

20

20

0

0 0

20

40

60

80

100

120

0

20

Retention volume (mL)

D

12 8 4 0 -4 -8 -12

200

210

220

230

240

40

60

80

100

120

Retension volume (mL) Fluorescence intensity (a.u.)

CD intensity (mdeg)

C

635

500 400 300 200 100 0 480

250

Wavelength (nm)

520

560

600

640

Wavelength (nm)

FIG. 2. Size-exclusion chromatography (SEC) elution profiles of Ex3-scDb-Fc purified from (A) flask cultures and (B) bioreactor cultures. The thin and thick lines correspond to the profiles of cells cultured without and with 150 mM trehalose, respectively. (C) Circular dichroism spectra and (D) thioflavin-T fluorescence spectra of monomeric (thick gray lines), dimeric (thin black lines), and large aggregates (thick black lines) of Ex3-scDb-Fc. Dashed line in panel D shows thioflavin-T fluorescence without an antibody.

buffer) differed from the neutral pH of the cell cultures, the kinetics of aggregate formation in the ThT-binding assay allowed us to understand the mechanism underlying the trehalose-mediated suppression of aggregation. Despite the higher trehalose concentration used in this experiment (200 vs. 150 mM), both concentrations exerted equivalent effects in heat-induced aggregation (data not shown). We confirmed that the solution structure of Ex3scDb-Fc immediately after preparation at pH 4.1 was almost identical to that determined at pH 7.4 (Fig. S2); that is, the pH change did not immediately induce the formation of large aggregates

300

Fluorescence Intensity (a.u.)

aggregates had formed during cell culture. However, in both the flask and the reactor cultures, the formation of the large aggregates was reduced approximately 3-fold in the presence of 150 mM trehalose, indicating the ability of this chemical chaperone to suppress the formation of large Ex3-scDb-Fc aggregates during cell culture. By contrast, the dimer content of the flask cultures was increased by trehalose, whereas in the reactor cultures it remained unchanged (Fig. 2A and B) but the area of the peak indicative of monomers was greater. Among the fundamental differences between the flask cultures and the reactor operation was the pH control of the culture medium. Long-term flask cultivation might have induced a slight shift of pH in response to L-lactate production, causing dimer formation in the flask cultures without trehalose. The solution structure of each state observed in the SEC analysis was further examined. The secondary structure of the dimeric Ex3scDb-Fc resembled that of the monomeric form (Fig. 2C), suggesting the assembly of two native monomers into the dimeric form without structural change. The large aggregates showed a significant increase in CD intensity at 218 nm, indicative of their misfolded state, which included a non-native b-strand structure. The bstrand-rich conformation is a typical property of misfolded proteins, such as amyloid fibrils (32,33). The fluorescent indicator ThT is commonly used to detect fibrils because binding to fibril-like oligomeric states causes a dramatic increase in fluorescence emission whereas the dye does not bind to proteins in the monomeric state (34). The ThT-binding assay results for the three states of Ex3scDb-Fc are shown in Fig. 2D. Florescence emission of ThT in the presence of Ex3-scDb-Fc monomers was identical to that without protein, and a slight enhancement was observed in the presence of dimers. The binding of large aggregates to ThT, however, produced a drastic enhancement in emission, suggesting that they consisted of fibril-like b-sheets. While our results indicate that trehalose suppresses the formation of large protein aggregates in CHO cell cultures, the underlying mechanism of this effect remained to be clarified. Although the acidic pH of the assay (carried out at pH 4.1 in citrate

250

200

200 150

150

100 50

100

0

1

2

3

4

5

6

50

0 0

10

20

30

40

50

60

Time (h) FIG. 3. Aggregation kinetics of Ex3-scDb-Fc monitored by thioflavin-T fluorescence (485 nm) at pH 4.1. Open and filled circles correspond to the data without and with 200 mM trehalose, respectively. Gray and black dashed lines are baseline intensities (fluorescence intensity without an antibody) without and with 200 mM trehalose, respectively. Continuous lines represent the fits of the regression analysis with a stretched exponential function (gray: without trehalose, black: with 200 mM trehalose).

ONITSUKA ET AL.

J. BIOSCI. BIOENG.,

TABLE 2. Kinetic parameters of aggregate formation with and without trehalose.a

Without trehalose þ200 mM trehalose a

DF (arbitrary units)

g

ksp (h1)

182  7 171  8

1.57  0.17 1.59  0.17

0.187  0.009 0.068  0.004

Errors are standard errors as determined in the regression analysis.

(Fig. S2). A plot of the time course of ThT fluorescence intensity at 485 nm (Fig. 3) showed that baseline ThT fluorescence intensity was slightly increased by trehalose addition (broken line). The time courses of assays carried out with and without trehalose did not show a monotonous increase of ThT fluorescence intensity upon binding of the dye to large aggregates, rather there was an apparent lag phase of up to 1 h (Fig. 3, inset). These sigmoidal time courses support a nucleation-dependent aggregation model in which the initial lag phase corresponds to the formation of an aggregation nucleus, which is the rate-limiting step in aggregation and polymerization (35,36). An analysis of the time courses by a nonlinear least squares curve fitting to a stretched exponential function, F ¼ FN þ DF exp([ksp,t]g), provided useful information in understanding the complex kinetic reaction (Table 2), although the parameters obtained by regression analysis of the equation are qualitative (37). Regardless of the presence or absence of trehalose, the DF values remained unchanged. Because DF represents the net formation of large aggregates, two unchanged values indicate that trehalose does not reduce the total amount of aggregate formation. The parameter g is related to the period of the lag phase and the cooperativity of the sigmoidal curve. The same two values of g suggest that trehalose does not influence aggregation nucleus formation, whereas trehalose did cause a large decrease in ksp, the rate of spontaneous formation of large aggregates. Trehalose shifted the resulting fitting curve to the right (Fig. 3), indicating that it delayed aggregate formation. Together, these results suggest that trehalose directly inhibits both spontaneous aggregate formation and polymerization but it does not reduce the total amount of aggregates formed or inhibit formation of the aggregation nucleus. DISCUSSION In serum-free CHO cell cultures engineered for antibody production, many environmental factors determine the subsequent aggregation of the secreted proteins, such as the cultivation temperature and pH, the osmolality, and the conductivity of the medium (1,3). Despite concern regarding the problems posed by aggregate formation, effective inhibitory strategies are lacking because culture conditions are optimized for cell growth and antibody production rather than for the suppression of aggregation. In this study, we considered all three aspects, which led us to explore the use of trehalose as an effective chemical chaperone to reduce antibody aggregation but without harmful effects on cell growth or recombinant protein production. Trehalose is used as a protectant in the formulation process, and is contained in commercially available therapeutic antibodies (22). Trehalose is an approved additive. The certified safety would promote the application of trehalose to CHO cell culture process. The addition of 150 mM trehalose to serum-free cultures of TopH cell line increased the osmolality of the medium to 480 mOsm/kg, vs. 319 mOm/kg in the absence of trehalose. Trehalose had a positive impact on the volumetric production of Ex3-scDb-Fc, even though cell growth was reduced. This behavior is the same as that previously reported for other cell lines engineered for recombinant protein production and cultured in hyperosmotic medium (18,28e30). The underlying mechanism responsible for the reduced cell growth and the observed increase in antibody production is unclear. Hyperosmotic stress might induce cell apoptosis. Anti-

apoptosis engineering such as supplementation with antiapoptotic additives (9) and over-expression of anti-apoptotic gene (38), would improve the cell growth. Differential transcriptome profiling showed that changes in the expression of the gene of interest in response to hyperosmotic conditions were dependent on the cell line (39). In the Top-H cell line used in this study, mRNA levels of Ex3-scDb-Fc were increased nearly 6-fold by the addition of trehalose to the culture medium, suggesting that a higher transcriptional level underlies the enhanced antibody production. The reason for increased mRNA levels of Ex3-scDb-Fc is not clarified. Trehalose may protect mRNA from degradation in vivo. Further investigations are necessary to elucidate the mechanism underlying the increase in mRNA levels by trehalose. Importantly, the transcriptional level could not be quantitatively matched to the enhanced productivity because rAbs with and without trehalose was 0.83 and 0.42 pg/cell/day, respectively. This 2-fold increase in antibody productivity implies that the rate-limiting step involved in the enhanced productivity by trehalose occurs after transcription, i.e., during translation, folding, or secretion. CHO cell engineering that includes an improvement of the UPR is effective in increasing cell-specific production rates by intensifying proteinfolding capacities (11e13). Secretory pathway engineering can likewise increase recombinant protein production (40). In either approach, the use of trehalose-containing medium may lead to significant increases in antibody productivity. Analyses of the aggregation kinetics and the solution structure of the large aggregates provided insight into the mechanism by which trehalose suppresses antibody aggregation (Fig. 2A and B). Kinetic analysis based on a nucleation-dependent aggregation model showed that trehalose inhibited the spontaneous polymerization of proteins (Fig. 3 and Table 2) into large aggregates, in which non-native b-strands are formed (Fig. 2C). This b-strand rich conformation has been shown to generally contribute to intermolecular association, leading to the formation of aggregates such as amyloid fibrils (32,33). A recent study demonstrated that trehalose is an amphiphilic molecule that in aqueous solution forms complexes with hydrophobic benzene (41). Accordingly, we propose that in CHO cell culture trehalose covers the solvent-exposed hydrophobic residues on the antibody that drive the intermolecular interactions between antibody molecules, thereby preventing the formation of large aggregates with non-native b-strand (Fig. 4). We also observed the formation of Ex3-scDb-Fc dimers during the cell culture process. The solution structure of these dimers was

Monomer

Amount of Large Aggregates

636

Precursor

Nucleus

Large aggregates

Trehalose

Time FIG. 4. Proposed model of the trehalose-induced suppression of antibody aggregation during CHO cell culture. According to a nucleation-dependent aggregation model (upper figure), trehalose targets the spontaneous polymerization step, such that the formation of large aggregates is delayed or inhibited (lower figure). Gray and black lines correspond to the aggregation kinetics in the absence and presence of trehalose, respectively.

VOL. 117, 2014

SUPPRESSED ANTIBODY AGGREGATION BY TREHALOSE

similar to that of monomers (Fig. 2C), except for the oligomeric state. A recent review provided insights into the aggregation mechanism, suggesting that subtle conformational changes in native-like monomers and oligomers are responsible for the nucleation step that under physiological conditions initiates the aggregation process (42). The conformation of therapeutic antibodies is influenced by heterogeneities in their glycosylation, lysine processing, and aspartate isomerization (43), all of which may cause dimer formation. Whether dimerization is also related to the formation of the aggregation nucleus during cell culture remains to be determined in further work. To the best of our knowledge, our study is the first to demonstrate the potential beneficial effects of trehalose in antibody production by mammalian cell cultures. The approach is simple to implement but must still be verified in further cell culture examples. Importantly, this study has broad implications for the use of chemical chaperones as a novel cell culture strategy to suppress antibody aggregation. Chemical chaperones show diverse effects, including the prevention of nucleation, which according to the nucleation-dependent aggregation model is the key step in aggregation. Elucidation of the driving force for nucleation and the use of chemical chaperones to block or limit this step would fundamentally contribute to resolving the problem of antibody aggregation in recombinant CHO cell cultures. Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jbiosc.2013.10.022.

enhances production of recombinant human antithrombin III in Chinese hamster ovary cells, J. Biosci. Bioeng., 106, 568e573 (2008). Arakawa, T. and Timasheff, S. N.: The stabilization of proteins by osmolytes, Biophys. J., 47, 411e414 (1985). Okanojo, M., Shiraki, K., Kudou, M., Nishikori, S., and Takagi, M.: Diamines prevent thermal aggregation and inactivation of lysozyme, J. Biosci. Bioeng., 100, 556e561 (2005). Matsuoka, T., Tomita, S., Hamada, H., and Shiraki, K.: Amidated amino acids are prominent additives for preventing heat-induced aggregation of lysozyme, J. Biosci. Bioeng., 103, 440e443 (2007). Rodriguez, J., Spearman, M., Huzel, N., and Butler, M.: Enhanced production of monomeric interferon-beta by CHO cells through the control of culture conditions, Biotechnol. Prog., 21, 22e30 (2005). Hwang, S. J., Jeon, C. J., Cho, S. M., Lee, G. M., and Yoon, S. K.: Effect of chemical chaperone addition on production and aggregation of recombinant flag-tagged COMP-angiopoietin 1 in Chinese hamster ovary cells, Biotechnol. Prog., 27, 587e591 (2011). Singer, M. A. and Lindquist, S.: Multiple effects of trehalose on protein folding in vitro and in vivo, Mol. Cell, 1, 639e648 (1998). Kaushik, J. K. and Bhat, R.: Why is trehalose an exceptional protein stabilizer? An analysis of the thermal stability of proteins in the presence of the compatible osmolyte trehalose, J. Biol. Chem., 278, 26458e26465 (2003). Saito, S., Hasegawa, J., Kobayashi, N., Tomitsuka, T., Uchiyama, S., and Fukui, K.: Effects of ionic strength and sugars on the aggregation propensity of monoclonal antibodies: influence of colloidal and conformational stabilities, Pharm. Res., 30, 1263e1280 (2013). Ohtake, S. and Wang, Y. J.: Trehalose: current use and future applications, J. Pharm. Sci., 100, 2020e2053 (2011). Kim, W. D., Tokunaga, M., Ozaki, H., Ishibashi, T., Honda, K., Kajiura, H., Fujiyama, K., Asano, R., Kumagai, I., Omasa, T., and Ohtake, H.: Glycosylation pattern of humanized IgG-like bispecific antibody produced by recombinant CHO cells, Appl. Microbiol. Biotechnol., 85, 535e542 (2010). Onitsuka, M., Kim, W. D., Ozaki, H., Kawaguchi, A., Honda, K., Kajiura, H., Fujiyama, K., Asano, R., Kumagai, I., Ohtake, H., and Omasa, T.: Enhancement of sialylation on humanized IgG-like bispecific antibody by overexpression of alpha2,6-sialyltransferase derived from Chinese hamster ovary cells, Appl. Microbiol. Biotechnol., 94, 69e80 (2012). Asano, R., Kawaguchi, H., Watanabe, Y., Nakanishi, T., Umetsu, M., Hayashi, H., Katayose, Y., Unno, M., Kudo, T., and Kumagai, I.: Diabodybased recombinant formats of humanized IgG-like bispecific antibody with effective retargeting of lymphocytes to tumor cells, J. Immunother., 31, 752e761 (2008). Asano, R., Ikoma, K., Kawaguchi, H., Ishiyama, Y., Nakanishi, T., Umetsu, M., Hayashi, H., Katayose, Y., Unno, M., Kudo, T., and Kumagai, I.: Application of the Fc fusion format to generate tag-free bi-specific diabodies, FEBS J., 277, 477e487 (2010). Omasa, T., Furuichi, K., Iemura, T., Katakura, Y., Kishimoto, M., and Suga, K.: Enhanced antibody production following intermediate addition based on flux analysis in mammalian cell continuous culture, Bioprocess Biosyst. Eng., 33, 117e125 (2010). Takagi, M., Moriyama, T., and Yoshida, T.: Effects of shifts up and down in osmotic pressure on production of tissue plasminogen activator by Chinese hamster ovary cells in suspension, J. Biosci. Bioeng., 91, 509e514 (2001). Zhu, M. M., Goyal, A., Rank, D. L., Gupta, S. K., Vanden Boom, T., and Lee, S. S.: Effects of elevated pCO2 and osmolality on growth of CHO cells and production of antibody-fusion protein B1: a case study, Biotechnol. Prog., 21, 70e77 (2005). Liu, C. H. and Chen, L. H.: Promotion of recombinant macrophage colony stimulating factor production by dimethyl sulfoxide addition in Chinese hamster ovary cells, J. Biosci. Bioeng., 103, 45e49 (2007). Arakawa, T., Philo, J. S., Tsumoto, K., Yumioka, R., and Ejima, D.: Elution of antibodies from a Protein-A column by aqueous arginine solutions, Protein Expr. Purif., 36, 244e248 (2004). Jaroniec, C. P., MacPhee, C. E., Bajaj, V. S., McMahon, M. T., Dobson, C. M., and Griffin, R. G.: High-resolution molecular structure of a peptide in an amyloid fibril determined by magic angle spinning NMR spectroscopy, Proc. Natl. Acad. Sci. USA, 101, 711e716 (2004). Nelson, R., Sawaya, M. R., Balbirnie, M., Madsen, A. O., Riekel, C., Grothe, R., and Eisenberg, D.: Structure of the cross-beta spine of amyloid-like fibrils, Nature, 435, 773e778 (2005). Biancalana, M. and Koide, S.: Molecular mechanism of Thioflavin-T binding to amyloid fibrils, Biochim. Biophys. Acta, 1804, 1405e1412 (2010). Harper, J. D. and Lansbury, P. T., Jr.: Models of amyloid seeding in Alzheimer’s disease and scrapie: mechanistic truths and physiological consequences of the time-dependent solubility of amyloid proteins, Annu. Rev. Biochem., 66, 385e407 (1997). Nilsson, M. R.: Techniques to study amyloid fibril formation in vitro, Methods, 34, 151e160 (2004). Hamada, D. and Dobson, C. M.: A kinetic study of beta-lactoglobulin amyloid fibril formation promoted by urea, Protein Sci., 11, 2417e2426 (2002).

14. 15.

16.

17.

18.

19. 20.

21.

22. 23.

24.

ACKNOWLEDGMENTS This study was supported by the Advanced Research for Medical Products Mining Programme of the National Institute of Biomedical Innovation (NIBIO). The authors thank Akiko Miyake (Hayashibara Co., Ltd.) for measuring the osmolality of the culture medium.

25.

26.

References 1. Cromwell, M. E., Hilario, E., and Jacobson, F.: Protein aggregation and bioprocessing, AAPS J., 8, E572eE579 (2006). 2. Philo, J. S. and Arakawa, T.: Mechanisms of protein aggregation, Curr. Pharm. Biotechnol., 10, 348e351 (2009). 3. Vázquez-Rey, M. and Lang, D. A.: Aggregates in monoclonal antibody manufacturing processes, Biotechnol. Bioeng., 108, 1494e1508 (2011). 4. Rosenberg, A. S.: Effects of protein aggregates: an immunologic perspective, AAPS J., 8, E501eE507 (2006). 5. Maas, C., Hermeling, S., Bouma, B., Jiskoot, W., and Gebbink, M. F.: A role for protein misfolding in immunogenicity of biopharmaceuticals, J. Biol. Chem., 282, 2229e2236 (2007). 6. Omasa, T., Onitsuka, M., and Kim, W. D.: Cell engineering and cultivation of Chinese hamster ovary (CHO) cells, Curr. Pharm. Biotechnol., 11, 233e240 (2010). 7. Kaneko, Y., Sato, R., and Aoyagi, H.: Evaluation of Chinese hamster ovary cell stability during repeated batch culture for large-scale antibody production, J. Biosci. Bioeng., 109, 274e280 (2010). 8. Kaneko, Y., Sato, R., and Aoyagi, H.: Changes in the quality of antibodies produced by Chinese hamster ovary cells during the death phase of cell culture, J. Biosci. Bioeng., 109, 281e287 (2010). 9. Liu, C. H. and Wu, K. W.: Synergistic effects of basic fibroblast growth factor and insulin on Chinese hamster ovary cells under serum-free conditions, J. Biosci. Bioeng., 107, 312e317 (2009). 10. Chen, F., Ye, Z., Zhao, L., Liu, X., Fan, L., and Tan, W. S.: Biphasic addition strategy of hypoxanthine and thymidine for improving monoclonal antibody production, J. Biosci. Bioeng., 114, 347e352 (2012). 11. Dinnis, D. M. and James, D. C.: Engineering mammalian cell factories for improved recombinant monoclonal antibody production: lessons from nature? Biotechnol. Bioeng., 91, 180e189 (2005). 12. Ohya, T., Hayashi, T., Kiyama, E., Nishii, H., Miki, H., Kobayashi, K., Honda, K., Omasa, T., and Ohtake, H.: Improved production of recombinant human antithrombin III in Chinese hamster ovary cells by ATF4 overexpression, Biotechnol. Bioeng., 100, 317e324 (2008). 13. Omasa, T., Takami, T., Ohya, T., Kiyama, E., Hayashi, T., Nishii, H., Miki, H., Kobayashi, K., Honda, K., and Ohtake, H.: Overexpression of GADD34

27.

28.

29.

30.

31.

32.

33.

34. 35.

36. 37.

637

638

ONITSUKA ET AL.

38. Liew, J. C., Tan, W. S., Alitheen, N. B., Chan, E. S., and Tey, B. T.: Overexpression of the X-linked inhibitor of apoptosis protein (XIAP) delays serum deprivation-induced apoptosis in CHO-K1 cells, J. Biosci. Bioeng., 110, 338e344 (2010). 39. Shen, D., Kiehl, T. R., Khattak, S. F., Li, Z. J., He, A., Kayne, P. S., Patel, V., Neuhaus, I. M., and Sharfstein, S. T.: Transcriptomic responses to sodium chloride-induced osmotic stress: a study of industrial fed-batch CHO cell cultures, Biotechnol. Prog., 26, 1104e1115 (2010). 40. Peng, R. W. and Fussenegger, M.: Molecular engineering of exocytic vesicle traffic enhances the productivity of Chinese hamster ovary cells, Biotechnol. Bioeng., 102, 1170e1181 (2009).

J. BIOSCI. BIOENG., 41. Sakakura, K., Okabe, A., Oku, K., and Sakurai, M.: Experimental and theoretical study on the intermolecular complex formation between trehalose and benzene compounds in aqueous solution, J. Phys. Chem. B., 115, 9823e9830 (2011). 42. Chiti, F. and Dobson, C. M.: Amyloid formation by globular proteins under native conditions, Nat. Chem. Biol., 5, 15e22 (2009). 43. Harris, R. J., Chin, E. T., Macchi, F. D., Keck, R. G., Shyong, B.-J., Ling, V. T., Cordoba, A. J., Marian, M., Sinclair, D., Battersby, J. E., and Jones, A. J. S.: Analytical characterization of monoclonal antibodies: linking structure to function, pp. 193e205, in: Shire, S. J., Gombotz, W. R., Bechtold-Peters, K., and Andya, J. (Eds.), Current trends in monoclonal antibody development and manufacturing. Springer, New York (2010).