Effects of amino acid additions on ammonium stressed CHO cells

Journal of Biotechnology 117 (2005) 277–286 Effects of amino acid additions on ammonium stressed CHO cells Peifeng Chen a , Sarah W. Harcum a,b,∗ a Department of Chemical Engineering, Clemson University, Clemson, SC 29634-0905, USA b Department of Bioengineering, Clemson University, Clemson, SC 29634-0905, USA Received 24 September 2004; received in revised form 3 February 2005; accepted 9 February 2005 Abstract Ammonium is a toxic and inhibitory byproduct of mammalian cell metabolism. At the end of a typical recombinant protein production campaign, the ammonium concentration can be as high as 10 mM, mainly due to glutamine metabolism. Intracellular pH (pHi ) levels are sensitive to ammonium, which negatively impacts both cell growth and recombinant protein productivity. Ammonium also negatively affects the recombinant protein glycosylation profile, thus altering quality. Many strategies have been adopted to reduce ammonium accumulation, with limited results. This study investigated the addition of amino acids to the growth media for Chinese hamster ovary (CHO) cell cultures as a means of mitigating the negative effects of ammonium. Threonine, proline, and glycine additions improved CHO cell growth and recombinant protein levels. Further, the threonine, proline, and glycine additions positively impacted important metabolic parameters, including glucose consumption, lactate production, glutamine utilization, and final ammonium levels. Additionally, threonine, proline, and glycine increased the level of ␣(2,3)-linked sialic acid, galactose-␤(1,4)-N-acetylglucosamine, and ␣(2,6)-linked sialic acid residues on the recombinant tissue plasminogen activator (t-PA). Thus, threonine, proline, and glycine can be used to mitigate some of the toxic effects of ammonium on cell growth, recombinant protein productivity, and protein quality. © 2005 Elsevier B.V. All rights reserved. Keywords: Amino acids; Tissue plasminogen activator; Ammonia/ammonium; Chinese hamster ovary cells; Glycosylation 1. Introduction Glutamine, a major nutrient in mammalian cell culture, is usually present at concentrations between 1 and ∗ Corresponding author. Tel.: +1 864 656 6865; fax: +1 864 656 0567. E-mail address: harcum@clemson.edu (S.W. Harcum). 0168-1656/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jbiotec.2005.02.003 7 mM. Glutamine metabolism and glutamine chemical decomposition are the primary sources of ammonium buildup in media and cell cytoplasm. Other media sources of ammonium include catabolism of some amino acids and fetal calf serum (McLimans et al., 1991). Ammonium concentrations can reach 10 mM at the end of a typical batch or fed-batch culture (Ozturk et al., 1992). Ammonium readily diffuses across the 278 P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 cell membrane, perturbing the intracellular pH (pHi ) and electrochemical gradients. Glutamine metabolism decreases pHi , while glutamine decomposition and ammonium-additions increase pHi (Schneider et al., 1996). The effects of ammonium or ammonia on cell growth, productivity, and glycosylation have been investigated. Specifically, elevated ammonium significantly inhibits cell growth and final cell densities (Gawlitzek et al., 1999; Hassell et al., 1991; McQueen and Bailey, 1991; Schneider et al., 1997; Singh et al., 1994; Yang and Butler, 2000b), and ammonium decreases recombinant protein productivity (Canning and Fields, 1983; Ito and McLimans, 1981; Reuveny et al., 1986). Tissue plasminogen activator (t-PA) production has been observed to be reduced under elevated ammonium conditions (Hansen and Emborg, 1994). Also, it has been observed that elevated ammonium levels result in lower terminal sialylation of all the glycans, lower tetraantennary and tetrasialylated oligosaccharide structures, and higher molecular heterogeneity (Andersen and Goochee, 1995; Borys et al., 1994; Gawlitzek et al., 1998; Thorens and Vassalli, 1986; Yang and Butler, 2000, 2002). These changes to the recombinant protein typically result in a less efficacious drug. Many strategies have been developed to overcome the toxic ammonium effects in mammalian cell culture. Some methods have attempted to reduce the production of ammonia from glutamine. For example, the substitution of glutamine for glutamate has been investigated; however, cell growth was lower with glutamate (Capiaumont et al., 1995; Hassell et al., 1991). Others have attempted to remove ammonium from the media by electrokinetic techniques (Gawlitzek et al., 1995); gas-permeable, hydrophobic porous membranes (Schneider, 1995); immobilized adsorbents (Jeong and Wang, 1992); and co-culturing with the cells that metabolize ammonia (Liu and Chang, 2003). However, no readily available, widely applicable method has been developed. Amino acids exist as zwitterions at physiological pH; thus, amino acids have the potential to act as buffers of pHi (Bavister and McKiernan, 1993; Gardner and Lane, 1997). Edwards et al. (1998) reported non-essential amino acids and glutamine had the greatest capacity to buffer pHi for early mouse preimplantation embryos. deZengotita et al. (2002) demonstrated that the amino acids glycine, threonine, and glycine betaine could protect hybridoma and CHO cells from elevated carbon dioxide and osmolality. The objective of this study was to develop a method to easily mitigate the negative effects of ammonium. Various amino acids were supplied to CHO cell cultures under ammonium stress. The effects of the amino acid additions on CHO cell growth, recombinant protein productivity, and glycosylation were examined. It was determined that some amino acids could mitigate the negative effect of ammonium, especially cell growth rates and productivity. 2. Materials and methods CHO cells (CRL-9606) were obtained from American Type Culture Collection (ATCC). This cell line constitutively expresses t-PA. All chemical reagents were purchased from Sigma (St. Louis, MO) unless otherwise specified. CHO cells were cultured in a basal media containing IS CHO-VTM (Irvine Scientific, Santa Ana, CA) with 0.1% fetal bovine serum (Invitrogen), 200 nM methotrexate, and 7 mM glutamine (Irvine Scientific, Santa Ana, CA). CHO cells were cultured in 1-L spinner flasks in humidified incubators at 37 ◦ C and 5% CO2 . NH4 Cl was used to simulate elevated ammonium levels. NH4 Cl and amino acids stock solution were prepared in Milli-Q water. The control spinner flask contained the basal media. The negative control contained the basal media supplemented with 10 mM NH4 Cl. The amino acid supplement spinner flasks contained the basal media plus 20 mM of the appropriate amino acid and 10 mM NH4 Cl. Milli-Q water was used to normalize culture volumes; this equalized basal media components and minimized the osmolality difference due to NH4 Cl and amino acids additions. NH4 Cl, amino acids, and water were added to the fresh media prior to the addition of cells. Each set of spinner flasks was inoculated from a common seed culture. The amino acid-supplemented cultures were run in duplicates. Cell densities and viabilities were determined by the Trypan blue exclusion method. Harvested cells were centrifuged at 1000 rpm for 5 min at 4 ◦ C. The supernatant and pellets were saved at −20 and −80 ◦ C, respectively, for later analysis. P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 The glucose concentrations were measured by One Touch glucose meter (Life Scan, Milapitas, CA). Glutamine concentrations were measured by a glutamine assay kit (Sigma, MI), as per kit instructions. Lactate concentration was measured with a lactate assay kit (R-biopharm, Germany), as per kit instructions. Ammonium and amino acids were analyzed with a Dionex D-300 amino acid analyzer and a Dionex 2000i ion exchange chromatography system with a dual wavelength detector (440 and 570 nm). Supernatants (2.8 mL) were mixed with 10% sulfosalicylic acid (1 mL) to precipitate the proteins prior to chromatography column. Samples were incubated at room temperature for 30 min and filtered prior to injection. The specific consumption rates of glucose and lactate for the metabolic parameters were calculated by the general formula q= dY 1 dt X (1) where Y represents each metabolic parameter, t represents culture time, and X represents viable cell densities. This method assumes that each specific consumption rate is constant. T-PA activity was measured by the Spectrolyse kit (American Diagnostica), as per kit instructions. The t-PA integrity was analyzed by Western blot. The Opti4CN kit was used (Bio-Rad). Rabbit anti-human recombinant t-PA (American Diagnostica) was used as primary antibody. The secondary antibody was GARHRP (Bio-Rad). The DIG glycan differentiation kit (Roche) was used to identify glycan structures on tPA. Equal volumes of culture supernatant (2 ␮L) were used for Western blots and DIG analysis. 3. Results To determine the effects of supplemented amino acids on cell growth for CHO cells exposed to elevated ammonium, various amino acids (alanine, aspartic acid, histidine, glycine, proline, serine, threonine, arginine, and lysine) were examined. These amino acids were selected based on at least one of the following properties: (1) potential metabolic effects related to ammonia or glutamine; (2) high secretion or consumption rates reported in other ammonia studies; (3) various pI values to determine the effect of pI on 279 ammonium toxicity; (4) examples of essential or nonessential amino acids. In order to determine the optimum amino acid concentrations to overcome ammonium toxicity, the ammonium inhibition for this cell line in the IS-CHOV media were first examined. The addition of 5 mM ammonium to the media prior to inoculation did not significantly inhibit the CHO cell growth rates (data not shown). The addition of 20 mM or greater levels of ammonium dramatically inhibited the CHO cells growth rates, and exceeded most industrially observed levels. Additionally, none of the amino acid additions were able to counteract the ammonium toxicity where 20 mM ammonium was used. The addition of 10 mM ammonium was observed to significantly inhibit the cell growth rates, and this has been observed for t-PA production and other proteins (Hansen and Emborg, 1994; Kurano et al., 1990; Schlaeger and Schumpp, 1989). Thus, the 10 mM ammonium level was selected for further study. The amino acid addition concentration was also examined, and it was determined that 20 mM amino acids additions had the greatest impact on the cell growth rates. Therefore, the set of conditions examined in detail used 10 mM ammonium and 20 mM amino acid additions. Most of the amino acids examined did not improve the cell growth rates, however, three amino acids (threonine, proline, and glycine) improved growth rates relative to the negative control cultures. The growth curves for the three amino acids that improved growth rates and an example of one amino acid (alanine) that did not improve the growth rates are shown in Fig. 1. For all amino acids examined, the control cultures had the highest growth rates and final cell densities. The negative control cultures had significant lower growth rates and final cell densities than the control culture. The threonine, proline, glycine-supplemented cultures had lower growth rates and final cell densities than the control cultures and significantly improved growth rates than the negative controls (ANOVA, p ≤ 0.05). Specifically, threonine addition improved the final cell density by 15% compared to the negative control culture; the proline addition improved the final cell density by 20%; the glycine addition improved the final cell density by 25%. However, the final cell densities for the three amino acid-supplemented cultures were approximately 20% lower than the control cultures. The alanine and serine additions in this study were selected as 280 P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 Fig. 1. CHO cell densities for amino acid supplemented cultures in the presence of elevated ammonium levels. Control (
), duplicate amino acid supplemented (䊉), (), and negative control () cultures. Error bars represent 95% confidence intervals. two examples that did not improve the growth rates. The alanine addition was observed to decrease cell growth rates to levels below the negative control. This inhibitory effect was proportional to the amount of alanine added to the media for concentrations between 5 and 20 mM. Serine also inhibited cell growth rates, but not as severely as alanine (data not shown). The glucose consumption rates and lactate production rates are indicators of cell health. The glucose consumption rates of the control and negative control cultures were nearly 13 and 20 nmol/105 cells h, respectively, as shown in Fig. 2. Glucose and lactate consumption rates for all conditions are shown in Fig. 2. The negative control culture utilized more glucose per cell in order to mitigate the ammonium stress. The lactate production rate for the control and negative control cultures were 8 and 12 nmol/105 cell h, respectively. High lactate production rates indicated the cells were more wasteful with respect to carbon utilization. The glucose consumption rates for the threonine, proline, and glycine treatments were between 13 and 17 nmol/105 cell h. The lactate consumption rates for the threonine, proline, and glycine treatments were between 9 and 11 nmol/105 cell h. The alanine addition increased glucose consumption rates and lactate production rates to 2.5-fold higher than the negative con- Fig. 2. (A) Glucose consumption and lactate production rates for the control, amino acid supplemented, and negative control cultures; (B) ammonium and glutamine concentration at the end of the cultures (61 h). P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 281 Table 1 Amino acid concentration in media for the basal media initially, control, negative control, and amino acid supplemented cultures at time 61 h Amino acid (mM) Alanine Arginine Aspartic acid Cystine Glycine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Proline Serine Threonine Tyrosine Valine Conditions Basal media Control Negative control Threonine Proline Glycine Alanine Serine 1.06 1.04 0.53 0.61 0.77 0.33 0.62 1.07 0.68 0.19 0.65 1.18 0.80 0.66 0.25 1.00 3.43 0.63 0.63 0.31 0.57 0.12 0.30 0.52 0.26 0 0.31 0 0.64 0.32 0.08 0.60 3.63 0.64 0.69 0.36 0.41 0.16 0.32 0.60 0.30 0 0.38 0 1.50 0.34 0 0.66 3.00 0.74 0.53 0.34 0.46 0.19 0.35 0.65 0.40 0 0.37 0 2.45 16.1 0.13 0.64 2.81 0.67 0.59 0.34 0.51 0.18 0.35 0.65 0.34 0 0.44 15.7 2.70 0.36 0.12 0.68 3.70 0.69 0.68 0.41 16.1 0.15 0.39 0.72 0.36 0 0.48 0 2.31 0.38 0.17 0.68 16.6 0.99 0.66 3.01 0.60 0.32 0.58 1.10 0.66 0.15 0.57 0 5.37 0.54 0.29 0.95 4.06 0.91 0.71 0.56 0.50 0.17 0.44 0.77 0.41 0.07 0.53 0 20.2 0.41 0.16 0.76 The amino acid concentrations corresponding to the supplemented amino acid are shown in italics. trol. The serine addition also increased glucose consumption rates and lactate production rates to levels above the negative control, although not as severely as the alanine addition. These results indicated threonine, proline, and glycine additions positively impacted carbon utilization under ammonium stress, while the alanine and serine additions negatively impacted carbon utilization. As noted previously, elevated ammonium levels are commonly observed near the end of mammalian cell cultures. The ammonium levels for the control and negative control cultures at the end of the cultures (61 h) were 4.1 and 11.7 mM, respectively, as shown in Fig. 2. End-of-culture ammonium levels for all conditions are also shown in Fig. 2. The ammonium levels for proline, glycine, alanine, and serine were approximately the same concentrations as the original ammonium additions; however, the threoninesupplemented cultures had significantly lower ammonium levels. Glutamine can be used as a carbon and nitrogen source by cells in culture. The glutamine concentrations for the control and negative control cultures at the end of the cultures were 0.8 and 2.1 mM, respectively (shown in Fig. 2). The initial glutamine concentration in the media was 7 mM for all cultures. The glutamine concentrations for all the amino acid-supplemented cultures were slightly-to-moderately higher than the negative controls (Fig. 2). Free amino acids in the media can be used by CHO cells directly as protein and nucleotide precursors, as well as carbon and nitrogen sources. The amino acid concentrations at the end of all the cultures were determined for the control cultures and the amino acidsupplemented cultures (threonine, proline, glycine, alanine, and serine), and these data are shown in Table 1. The amino acid concentrations for the basal media were also determined. For the control and negative control cultures at the end of the cultures, alanine and aspartic acid concentrations were higher relative to the basal media; thus, these amino acids are produced by the cells and secreted. The amino acids arginine, cysteine, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, threonine, tyrosine, and valine were all consumed approximately equally by the control and negative control cultures by the end of the cultures. Only proline and methionine were completely consumed by the control and negative control cultures. The only major differences in amino acid consumption between the control and negative-control cultures were observed for serine and tyrosine. Serine was consumed by the control cultures and generated by the negative control cultures. Tyrosine was completely consumed by the negative control cultures, whereas only approxi- 282 P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 Fig. 3. Western blots of t-PA from proline-, glycine-, and threonine-supplemented cultures in the presence of elevated ammonium levels. Control (C), amino acid (AA), negative control (N). Times corresponded to culture times. mately two-thirds of the basal media concentration was consumed by the control cultures. In general, the alanine-supplemented culture had the highest number of amino acids with significantly different amino acid concentrations at the end of the cultures when compared to the controls and other amino acidsupplemented cultures. The serine-supplemented cultures were more similar to the alanine-supplemented cultures in the end-of-culture amino acid profiles as well as in not protecting the cell growth rate. The amino acid concentrations of threonine-, proline-, and glycine-supplemented cultures were very similar to each other. The most striking difference between the amino acid-supplemented cultures and the control cultures was the serine concentration at the end of the culture. The serine levels for the amino acid-supplemented cultures were all significantly higher than the negative control cultures, which also produced serine, whereas the control cultures consumed serine. T-PA has three potential N-linked glycosylation sites. Two glycosylation variants of t-PA have been identified, type I and type II. Type I t-PA is glycosylated at amino acid positions 117, 184, and 448 and has a molecular weight of 68 kD. Type II t-PA is glycosylated at positions 117 and 448 and has a molecular weight 65 kD. Western blots were used to assess the effects of amino acid supplementation on the distribution of type I and II t-PA. In this study, both type I and II t-PA were produced by the control, amino acidsupplemented, and negative control cultures, as shown in Fig. 3. The t-PA activity was determined for the control, amino acid-supplemented (threonine, proline, and glycine only), and negative control cultures, as shown in Fig. 4. The control cultures had the highest t-PA activities at the end of the cultures. The negative control cultures had the lowest t-PA activities at the end of the cultures. For the threonine cultures, the observed t-PA activity was nearly identical to the control cultures. For glycine and proline, the amino acids increased t-PA activities; however, only approximately to the average of the control and negative control levels. The effects of the amino acid supplementation on the glycosylation profiles of t-PA were determined using lectins specific to oligosaccharide linkages commonly found on terminal glycan branches. T-PA samples from the amino acid-supplemented cultures with improved growth were compared (threonine, proline, and glycine) to the control and negative control cultures. The 49-h samples for the control, amino acidsupplemented, and negative control were compared. The 49-h samples were selected for comparison since the cells had reached late exponential phase, but were not nutrient limited, such as for proline. Therefore, the effect of nutrients on glycosylation were minimized. Five lectins were used to assess glycosylation. As expected, the Galanthus nivalis agglutinin (GNA) did not detect highly terminal mannose in any samples; and Peanut agglutinin (PNA) did not detect O-linked disaccharide galactose ␤(1,3)-N-acetylgalactosamine in any Fig. 4. Normalized t-PA activity for proline-, glycine-, and threonine-supplemented cultures in the presence of elevated ammonium levels at the end of the cultures (61 h). Control (C), amino acid (AA), negative control (N). P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 283 Fig. 5. Glycan differentiation of t-PA glycans for proline-, glycine-, and threonine-supplemented cultures in the presence of elevated ammonium levels. (A) Sialic acid terminally linked ␣(2,3) to galactose. (B) ␤(1,4)-N-acetylglucosamine linked to galactose. Control (C), amino acid (AA), negative control (N). samples. The Maackia amurensis agglutinin (MAA) detected ␣(2,3) sialic acid linked to galactose for both type I and II t-PA in all samples, as shown in Fig. 5. The Datura stramonium agglutinin (DSA) detected ␤(1,4)N-acetylglucosamine linked to galactose for both type I and II t-PA in all samples, as shown in Fig. 5. The Sambucus nigra agglutinin (SNA), which detects ␣(2,6) sialic acid linked to galactose, gave a faint signal for both type I and II t-PA in all samples (data not shown). This faint ␣(2,6) sialic acid signal was most likely due to the O-linked glycan (Harris et al., 1991; Moloney and Haltiwanger, 1999; Van den Steen et al., 1998) since CHO cells do not express the ␣(2,6) sialytransferase for N-linked glycans. For the lectin blots, all the amino acid supplemented cultures had intensity bands that were slightly higher than the negative controls, but lower than the controls, as shown in Fig. 5. The intensity differences could be attributed to higher ␣(2,3) sialic acid, higher ␤(1,4)N-acetylglucosamine, or higher ␣(2,6) sialic acid on the O-linked glycan, respectively. The intensity differences could also be attributed to higher levels of t-PA. Since lectin blots are only semi-quantitative, it was only possible to estimate the intensity difference; however, the differences in t-PA activity were less than the differences observed in the lectin blot band intensities. Thus, the observed increase in glycosylation was attributed to improved glycosylation and not just higher t-PA levels. ammonium, as demonstrated by lower growth rates. Normally, the main source of ammonium buildup is attributed to glutamine metabolism. Specifically, glutamine is hydrolyzed to glutamate and ammonia. Glutamate is further hydrolyzed to ␣-ketoglutarate via glutamate dehydrogenase, which releases another ammonia molecule, as shown in Pathway 1 in Fig. 6. Alternatively, glutamate and pyruvate can react to form alanine and ␣-ketoglutarate, as shown in Pathway 2 in Fig. 6. Alanine is hypothesized to be secreted into the media as a mechanism to reduce ammonia toxicity by redirecting the amine group from glutamate to alanine, instead of a free ammonia molecule (Hansen and Emborg, 1994; Miller et al., 1988; Ozturk et al., 1992; Schneider et al., 1996; Street et al., 1993) as shown in Fig. 6. Thus, the addition of alanine to the media probably inhibited the reaction of glutamate and pyruvate to alanine, therefore prevented the cells from secreting alanine. In this study, compared with the concentration in basal media, alanine was found to be secreted into the media at the end of the culture at high levels for the control, negative control, and other amino acid supplemented cultures. These results strongly support the hypothesis that alanine secretion reduces ammonium stress via this mechanism. Additionally, the alanine supplemented cultures had elevated glutamine 4. Discussion The objective of this research project was to determine if amino acids could protect mammalian cell growth, recombinant protein productivity, and glycosylation in the presence of elevated ammonium. Threonine, proline, and glycine had significant protective impact. Most amino acids, such as alanine, serine, and histidine, did not mitigate the negative effects of Fig. 6. Schematic diagram of amino acid metabolism for alanine, proline, glycine, and threonine, as related to ammonium. 284 P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 levels, even when compared to the negative control. The elevated glutamine levels were probably from a combination of lower consumption of glutamine from lower growth rates and a reduction in the glutamine/pyruvate to alanine/␣-ketoglutarate reaction. High levels of serine were observed in the media at the end of the cultures for the negative control and other amino acid (non-serine) supplemented cultures; however, the serine levels in the media at the end of the cultures for the control cultures were lower than the basal media. Elevated serine levels for the negative control and amino acid supplemented cultures could have been due to the reaction of pyruvate with ammonia to form serine. Additionally, this reaction in reverse (serine to pyruvate and ammonia) could have been responsible for the slightly higher levels of ammonium observed in the serine-supplemented cultures. The threonine-, proline-, and glycine- supplemented cultures had lower ammonium levels relative to the negative controls. Additionally, the ammonium levels in the threonine-supplemented cultures were lower than the amount of ammonium initially added to the media. Additionally, threonine protected the cells to the greatest extent based on growth rates and t-PA activity. The glutamine levels for the amino acid-supplemented cultures were higher than the control cultures and slightly higher than the negative control cultures. Interestingly, both the glucose consumption and lactate production rates for the threonine-, proline-, and glycinesupplemented cultures were between the levels observed for the control and negative control cultures. In contrast, the glucose consumption and lactate production levels for the alanine and serine supplemented cultures were both higher than the negative control cultures. These results indicate that the threonine, proline, and glycine additions may have provided both alternative carbon and nitrogen sources, which contributed to improved cell growth rates, productivity, and protein quality. The threonine, proline and glycine additions improved cell growth, t-PA productivity and glycosylation. Proline was slightly less effective than glycine and threonine at improving cell growth rates and tPA productivity. The protection provided by proline was probably related to glutamate synthesis, as shown in Fig. 6. Specifically, proline was probably converted to glutamate by the ATP-independent glutamate semialdehyde dehydrogenase. The proline addition allowed the cells to synthesize ␣-ketoglutarate via glutamate, which generates only a single ammonia molecule per proline. In contrast, the consumption of glutamine to ␣-ketoglutarate via glutamate generates two ammonia molecules per glutamine. Thus, proline metabolically substituted for glutamine and allowed the cells to have an active tricarboxylic (TCA) cycle with reduced ammonia production. This hypothesis is supported by the relatively high level of glutamine still present at the end of the proline supplemented culture (Fig. 2). Threonine and glycine additions had similar effects on cell growth, t-PA productivity, and glycosylation. It is probable that glycine and threonine used similar mechanisms to mitigate the ammonia stress. Glycine and threonine metabolism are connected, in that glycine is readily converted to threonine, as shown in Fig. 6. The increased pyruvate levels due to the glycine and threonine degradation mechanism may have increased the flux through the TCA cycle, which improved cell energetics. Specifically, threonine was converted to pyruvate via a pathway initiated by threonine dehydrogenase. This reaction yielded ␣-aminob-ketobutyrate. The ␣-amino-b-ketobutyrate molecule spontaneously degraded to aminoacetone, which was converted to pyruvate. The glycine mechanism was identical to threonine, except glycine was first converted to threonine by threonine aldolase. Thus, glycine and threonine may mitigate ammonia stress by increased pyruvate levels, which are then metabolized via the TCA cycle. Another possible protective mechanism that can be attributed to the amino acid supplements may be the zwitterions property of these amino acids. Glycosylation reactions are very sensitive to the Golgi pH (Valley et al., 1999). Ten mM ammonium has been observed to raise the Golgi pH from the normal range (between 6.5 and 6.75) to 7.0 (Andersen and Goochee, 1995). The pI values for threonine, proline, and glycine are 5.6, 6.3 and 5.97, respectively. Therefore, threonine, proline, and glycine have the ability to balance the Golgi pH due to zwitterionic properties. These results can be indirectly confirmed by the effect of histidine and arginine supplementation. These two basic amino acids, with pI values of 7.59 and 10.76, respectively, essentially prevented cell growth (data not shown). Although alanine has a pI of 6.02, it did not mitigate the negative effects of ammonia. This was probably due to its metabolic role in ammonia elimination. In this P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 study, alanine and serine did not protect the cells due to metabolic constraints, where as threonine, proline, and glycine probably mitigated the negative effects of ammonia via a combination of pI buffering effects and favorable metabolism. 5. Conclusions Various amino acids were screened to determine if the negative effects of ammonium stress on growth rates, productivity, and protein quality could be mitigated. Only a few amino acids were able to counteract the negative effects of ammonium on CHO cell growth rates. Threonine, proline, and glycine were able to partially recover CHO cell growth rates, improve t-PA, and improve glycosylation via metabolism and favorable pI values. Amino acid metabolism and the zwitterionic properties of the amino acids were both important properties for ammonium stress mitigation. For example, even with a favorable pI, alanine was unable to overcome its metabolic constraints, whereas threonine positively impacted metabolism and had a favorable pI value. In conclusion, these research results indicate that amino acid additions could be an effective and easy method to protect CHO cells from elevated ammonium stress. Acknowledgements The IS CHO-VTM media and l-glutamine were kindly donated by Irvine Scientific. This work is partially supported by the National Science Foundation under Grant No. 0303782. References Andersen, D.C., Goochee, C.F., 1995. The effect of ammonia on the O-linked glycosylation of granulocyte-colony-stimulating factor produced by Chinese-hamster ovary cells. Biotechnol. Bioeng. 47, 96–105. Bavister, B.D., McKiernan, S.H., 1993. Regulation of hamster embryo development in vitro by amino acids. In: Bavister, B.D. (Ed.), Preimplantation Embryo Development. Springer-Verlag, New York, pp. 57–72. Borys, M.C., Linzer, D.I.H., Papoutsakis, E.T., 1994. Ammonia affects the glycosylation pattern of recombinant mouse placental 285 lactogen-I by Chinese hamster ovary cells in a pH dependent manner. Biotechnol. Bioeng. 43, 505–514. Canning, W.M., Fields, B.N., 1983. Ammonium chloride prevents lytic growth of reovirus and helps to establish persistent infection in mouse L-cells. Science, 987–988. Capiaumont, J., Legrand, C., Carbonell, D., Dousset, B., Belleville, F., Nabet, P., 1995. Methods for reducing the ammonia in hybridoma cell-cultures. J. Biotechnol. 39, 49–58. deZengotita, V.M., Abston, L.R., Schmelzer, A.E., Shaw, S., Miller, W.M., 2002. Selected amino acids protect hybridoma and CHO cells from elevated carbon dioxide and osmolality. Biotechnol. Bioeng. 78, 741–752. Edwards, L.J., Williams, D.A., Gardner, D.K., 1998. Intracellular pH of the mouse preimplantation embryo: amino acids act as buffers of intracellular pH. Hum. Reprod. 13, 3441–3448. Gardner, D.K., Lane, M., 1997. Culture and selection of viable blastocysts: A feasible proposition for human IVF? Hum. Reprod. Update 3, 367–382. Gawlitzek, M., Conradt, H.S., Wagner, R., 1995. Effect of different cell culture conditions on the polypeptide integrity and Nglycosylation of a recombinant model glycoprotein. Biotechnol. Bioeng. 46, 536–544. Gawlitzek, M., Papac, D.I., Sliwkowski, M.B., Ryll, T., 1999. Incorporation of N-15 from ammonium into the N-linked oligosaccharides of an immunoadhesin glycoprotein expressed in Chinese hamster ovary cells. Glycobiology 9, 125–131. Gawlitzek, M., Valley, U., Wagner, R., 1998. Ammonium ion and glucosamine dependent increases of oligosaccharide complexity in recombinant glycoproteins secreted from cultivated BHK-21 cells. Biotechnol. Bioeng. 57, 518–528. Hansen, H.A., Emborg, C., 1994. Influence of ammonium on growth, metabolism, and productivity of a continuous suspension Chinese-hamster ovary cell-culture. Biotech. Prog. 10, 121– 124. Harris, R.J., Leonard, C.K., Guzzetta, A.W., Spellman, M.W., 1991. Tissue plasminogen-activator has an O-linked fucose attached to threonine-61 in the epidermal growth-factor domain. Biochemistry 30, 2311–2314. Hassell, T., Gleave, S., Butler, M., 1991. Growth-inhibition in animal-cell culture – the effect of lactate and ammonia. Appl. Biochem. Biotechnol. 30, 29–41. Ito, M., McLimans, W.F., 1981. Ammonia inhibition of interferon synthesis. Cell Biol. Int. Rep. 5, 661–666. Jeong, Y.-H., Wang, S.S., 1992. In-situ removal of ammonium ions from hybridoma cell culture media: selection of adsorbent. Biotechnol. Technol. 6, 341–346. Kurano, N., Leist, C., Messi, F., Kurano, S., Fiechter, A., 1990. Growth-behavior of Chinese hamster ovary cells in a compact loop bioreactor. 2. Effects of medium components and waste products. J. Biotechnol. 15, 113–128. Liu, Z.C., Chang, T.M.S., 2003. Coencapsulation of hepatocytes and bone marrow stem cells: in vitro conversion of ammonia and in vivo lowering of bilirubin in hyperbilirubemia Gunn rats. Int. J. Artificial Organs 26, 491–497. McLimans, W., Blumenson, L., Repasky, E., Ito, M., 1991. Ammonia loading in cell culture systems. Cell Biol. Int. Rep. 5, 653– 660. 286 P. Chen, S.W. Harcum / Journal of Biotechnology 117 (2005) 277–286 McQueen, A., Bailey, J.E., 1991. Growth-inhibition of hybridoma cells by ammonium ion – correlation with effects on intracellular pH. Bioproc. Eng. 6, 49–61. Miller, W.M., Wilke, C.R., Blanch, H.W., 1988. Transient responses of hybridoma cells to lactate and ammonia pulse and step changes in continuous culture. Bioproc. Eng. 3, 113–122. Moloney, D.J., Haltiwanger, R.S., 1999. The O-linked fucose glycosylation pathway: Identification and characterization of a uridine diphosphoglucose: fucose-␤ 1,3-glucosyltransferase activity from Chinese hamster ovary cells. Glycobiology 9, 679–687. Ozturk, S.S., Riley, M.R., Palsson, 1992. Effects of ammonia and lactate on hybridoma growth, metabolism, and serum concentration. Biotech. Prog. 6, 121–128. Reuveny, S., Velez, D., Macmillan, J.D., Miller, L., 1986. Factors affecting cell growth and monoclonal antibody production in stirred reactors. J. Immunol. Methods 86, 53–59. Schlaeger, E.-J., Schumpp, B., 1989. Studies on mammalian cell growth in suspension culture. In: Spier, R.E.G.J.B., Stephenne, J., Croopy, P.J. (Eds.), Advances in Animal Cell Biology, Technology for, Bioprocesses, Butterworths. Sevenoaks, Kent, UK, pp. 386–396. Schneider, M., 1995. Applications of hydrophobolic porous membranes in mammalian cell culture technology. Ph.D. Thesis No. 1412, Ecole Polytechnique Federal de Lausanne (EPFL), Switzerland. Schneider, M., ElAlaoui, M., vonStockar, U., Marison, I.W., 1997. Batch cultures of a hybridoma cell line performed with in situ ammonia removal. Enzym. Microb. Technol. 20, 268–276. Schneider, M., Marison, I.W., vonStockar, U., 1996. The importance of ammonia in mammalian cell culture. J. Biotechnol. 46, 161–185. Singh, R.P., Alrubeai, M., Gregory, C.D., Emery, A.N., 1994. Celldeath in bioreactors – a role for apoptosis. Biotechnol. Bioeng. 44, 720–726. Street, J.C., Delort, A.M., Braddock, P.S.H., Brindle, K.M., 1993. A H-1 N-15 NMR-study of nitrogen-metabolism in culturedmammalian-cells. Biochem. J. 291, 485–492. Thorens, B., Vassalli, P., 1986. Chloroquine and ammonium-chloride prevent terminal glycosylation of immunoglobulins in plasmacells without affecting secretion. Nature 321, 618–620. Valley, U., Nimtz, M., Conradt, H.S., Wagner, R., 1999. Incorporation of ammonium into intracellular UDP-activated Nacetylhexosamines and into carbohydrate structures in glycoproteins. Biotechnol. Bioeng. 64, 401–417. Van den Steen, P., Rudd, P.M., Dwek, R.A., Opdenakker, G., 1998. Concepts and principles of O-linked glycosylation. Crit. Rev. Biochem. Mol. Biol. 33, 151–208. Yang, M., Butler, M., 2000a. Effects of ammonia on CHO cell growth, erythropoietin production, and glycosylation. Biotechnol. Bioeng. 68, 370–380. Yang, M., Butler, M., 2000b. Effect of ammonia on the glycosylation of human recombinant erythropoietin in culture. Biotech. Prog. 16, 751–759. Yang, M., Butler, M., 2002. Effects of ammonia and glucosamine on the heterogeneity of erythropoietin glycoforms. Biotech. Prog. 18, 129–138.