The Effect of Limiting Phosphate Using the HMS174 Cell Line

ABSTRACT

The Escherichia coli recombinant cell line HMS174, when grown in a 2XYS media and fed with phosphate and glucose, reached a fairly high cell density. If phosphate was not added to the culture along with the glucose, the culture seemed to switch from glycolysis to the glyoxylate bypass metabolism of acetate, mimicking a glucose-limiting event. In this culture without phosphate, the level of acetate remained considerably lower, whereas glucose levels rose, suggesting a metabolic shift to using both acetate and glucose as a carbon source. The overall growth of this culture was approximately half that of the culture containing phosphate, but following induction with IPTG, the amount of recombinant protein produced per gram of wet cell weight was very similar for the two cultures. This suggests that even in a phosphate-limited state, HMS174 cells maintained their ability to produce expected amounts of recombinant protein by switching metabolic pathways and metabolizing acetate through the glyoxylate bypass pathway. Ultimately, the HMS174 recombinant host strain may be able to use both glucose and acetate in a mixed feed strategy to maximize recombinant product production.

High cell density fermentation of Escherichia coli hosted recombinant systems is increasingly being used for the production of protein products for vaccines, diagnostics, and therapeutic treatments. Understanding the metabolic needs of high cell density fermentation can help increase the productivity and efficiency of the fermentation as a whole.

Under aerobic respiration conditions, glucose is used as the main carbon source and is fed in a non-limiting fashion to reach high cell densities. Complications can arise when the culture maintains a high growth rate during the exponential phase of growth with the secretion of acetate into the surrounding media. At high enough concentrations, the acetate can inhibit cell growth or recombinant protein production.¹ Acetate also can decouple transmembrane pH gradients, affecting amino acid synthesis, osmotic pressure, and intracellular pH.² It has been shown that by adding yeast extract to the fermentation, either initially or during the feed, acetate formation and its negative effects on the culture can be lessened.³ Other nutrient additions to the growing culture also may have positive effects on growth and product formation.

(JOEY NICOLE PHOTOGRAPHY)

In this study, HMS174 competent cells containing a pET29a plasmid with a gene of interest were grown to a medium-high cell density in 2XYS media using a fed-batch fermentation method. Two cultures were fed 40% glucose in 2XYS, with and without the addition of phosphate at 20 g/L. Adding phosphate notably increased the culture's ability to sustain a high growth rate without the formation of inhibiting amounts of acetate. In contrast, the growth rate of the culture that was fed glucose without phosphate was half the rate, and after 3 h of growth, the phosphate concentration was undetectable.

Nevertheless, the cell culture maintained the recombinant protein expression fidelity of the culture that was fed glucose with phosphate. Both cultures produced the same amount of recombinant protein per gram of cell paste. In the absence of phosphate, the HMS174 strain seems to undergo the acetate switch and process acetate through the glyoxylate bypass pathway.⁴ Using an alternative carbon source in the absence of additional phosphate and the accumulation of glucose in the surrounding media may indicate an inhibition of the phosphotransferase system (PTS) for glucose⁵ or additional pathways that are dependent on a phosphorylated state. It also is noteworthy that the culture fed glucose with phosphate began metabolizing acetate upon induction of the culture with IPTG. These findings indicate that the HMS174 strain, in the absence or presence of excess phosphate, can metabolize acetate very efficiently and thus minimize the effects of acetate formation that would significantly affect other recombinant cell strains.

MATERIALS AND METHODS

Strain, Plasmid, and Media

The competent cell line HMS174 (DE3) was purchased through Novagen (EMD Chemicals, Darmstadt, Germany) and transformed with a pET-29a plasmid containing a recombinant gene of interest. The transformation was performed using the heat-shock method.

Inoculants and fermentations were grown in 2XYS media (16 g yeast extract, 10 g soytone, 5 g NaCl per L) with kanamycin (Sigma, St. Louis, MO) used as the antibiotic selection marker. Fermentation was carried out in 2.5 L vessels using a BioFlo 3000 controller (New Brunswick Scientific, Edison, NJ). The inoculating volume was 10% of the final fermentation volume. The dissolved oxygen (DO) level was held at 40% by sparging with 1 vvm of air and agitation was maintained at 500 rpms. If the DO fell below 40%, the air sparge was mixed with pure oxygen as needed to maintain 40% DO. NH₄OH and H2SO4 were used to maintain the pH at 7. Temperature was maintained at 37 °C. The growing culture was fed 40% glucose, and +20 g/L NaPO4 in 2XYS after the optical density of the culture reached 1–2 (A600). Feed rates were constant throughout culture growth and induction.

Analysis

Fermentation samples (20 mL) were taken every hour. Optical density (OD) was determined and 1 mL of culture was microfuged for 1 min at 14,000 rpm (Heraeus, DJB Labcare, UK). The corresponding pellet was used to determine dry cell weight (DCW) and the supernatant was analyzed for residual acetate, glucose, and phosphate concentrations. Pellets were dried in a rotovap for 2 h at medium heat. Growth curves were created from OD and DCW changes over time for both cultures. Pellets also were recovered to monitor induction of recombinant protein. The cell pellets were solubilized with a reducing sample buffer and run on an SDS-PAGE gel. The gel was then stained with Simply Blue (Invitrogen, Carlsbad, CA) and destained with water. The supernatants were analyzed using a Nova-Bio 300 bioanalyzer (Nova Biomedical, Waltham, MA). Final wet cell pellets were weighed for cell mass/L generated for both cultures. Inclusion bodies were isolated from both cultures by lysing cells in a microfluidizer at 18,000 psi and centrifuging inclusion bodies from cell lysate. Inclusion bodies were washed twice with 20% isopropyl alcohol (IPA). Their respective yields per gm of wet cell paste were calculated.

RESULTS

Lack of Phosphate in Feed Substantially Slows Growth of Culture

To evaluate the metabolic needs of a high cell density culture, phosphate was added to the 40% glucose feed for one of the cultures at 20 gm/L. Although the glucose feed rate of 0.16 g/min was the same for both cultures, the culture without phosphate feed had approximately half the growth rate (1.5 h) of the culture grown with phosphate in the feed (2.7 h). Furthermore, the cultures with and without phosphate feed metabolized 99% and 74% of the added glucose, respectively, while the optical densities for each culture reached 15 (–PO4) and 27 (+PO4) (Figure 1A). Recovered wet cell weights for the harvested cell pellets correlated well with final optical densities.

Figure 1a

Substantial Acetate Accumulation Absent in the Culture Without Phosphate Feed

Interestingly, the culture without phosphate feed showed a substantial residual glucose accumulation, but acetate levels were depressed (0.3 g/L). This may have something to do with the need for phosphate, in the form of phosphoenolpyruvate, to actively shuttle glucose from the outside to the inside of the cell.⁵ Accordingly, the culture with added phosphate showed no accumulation of glucose, with a much higher amount of acetate production per liter (1 g/L). In previous experiments under this feed strategy, acetate production was 5 g/L with no negative effects on cell growth or recombinant protein production (data not shown).

Figure 1b

The buildup of acetate is typical of a fast growing culture being fed glucose.^6,7,8 This suggests that the culture without phosphate feed may have switched to the glyoxylate shunt pathway (a 2-carbon source such as acetate) in the absence of an efficient metabolic state for the 6-carbon molecule glucose (Figure 1B). As shown in Figure 2A, the acetate concentration in the glucose culture without phosphate feed plateaus at ~4 h, whereas the culture with added phosphate feed continues to increase three-fold until about hour 5. After induction, the acetate levels decrease dramatically in the phosphate fed-culture, indicating the increased need for an additional carbon source and the ability of this strain to use acetate efficiently. The culture without phosphate feed maintained its extracellular acetate concentration while also maintaining the induction fidelity of the recombinant protein product.

Figure 2

Up-Regulation of Alkyl hydroxide Reductase Subunit C in Culture Without Phosphate Feed

During culture growth, samples were taken for media and cell pellet analysis. The pellets isolated after induction showed an increase in the production of a 22 kD protein in the culture without phosphate feed compared with the culture with added phosphate feed (Figure 3). This 22 kD protein was found to have strong sequence homology to alkyl hydroperoxide reductase (subunit C), an oxygen radical scavenger.⁹ The presence of such an oxygen radical scavenger indicates the highly oxidative state of the HMS174 culture fed only glucose. It is assumed that without additional phosphate, glycolysis is somewhat inhibited. With lower amounts of pyruvate being formed, converted to acetyl-CoA and metabolized in the citric acid cycle, the synthesis of NADH and FADH2 is compromised and thus inhibits the oxidative phosphorylation pathway. This would indeed be the case if the culture fed only glucose switched to the glyoxylate pathway that yields only 1 NADH molecule per acetate molecule.

Figure 3

Equivalence of Recombinant Protein Production on the Two Cultures

The production of recombinant protein product per gram wet cell paste is identical for the two cultures. As stated earlier, the growth rate of the HMS174 culture was substantially reduced when fed glucose only, compared to the culture being fed glucose and phosphate. However, this did not translate into a reduction of product formation per gram of cell paste when the cultures were induced with IPTG. The amount of product formation was determined after lysing the cell culture and isolating the inclusion bodies through centrifugation (Table 1). The expression of the recombinant protein in both cultures also was equivalent on SDS-PAGE stained with Coomassie blue (Figure 3). The amount loaded per lane was based on OD equivalents.

Table 1. Ratio of wet cell weight and inclusion body weight. The HMS174 culture with and without phosphate added to the glucose feed produced the same amount of inclusion body protein during a 2 h induction with IPTG.

DISCUSSION

Recombinant E. coli host strains such as K or B vary in their ability to produce sufficient quantities of recombinant proteins, to grow to high cell densities on different media, and most importantly, to show a robustness for changing metabolic demands throughout the fermentation process. Although the B strains produce less acetate because of an activated glyoxylate shunt,¹⁰ the host K strain HMS174 (DE3) containing the pET29a plasmid and gene of interest has not shown any negative effects from the production of acetate during high growth rates. During the exponential growth phase, the cultures fed glucose plus phosphate or glucose only cultures reached a maximum growth rate of 2.7 h and 1.5 h, respectively. Residual phosphate levels for both cultures decreased soon after the feed was started with the culture without phosphate feed reaching undetectable levels at the 3 h time point (Figure 2B). Phosphate in the culture with added phosphate feed was detectable throughout the growth, although decreasing steadily.

After induction with IPTG, not surprisingly, the culture with added phosphate feed produced twice the amount of recombinant product as the culture without phosphate feed because of the increased cell density. What was surprising was that both cultures produced identical amounts of recombinant product per g of wet cell paste. This was presented as a ratio of g of inclusion body per g of wet cell paste recovered. The equal production was confirmed by SDS-PAGE. For this to take place, the culture without phosphate feed must make use of other carbon sources besides glucose. Substances that contain 2 carbons, such as acetate, are an obvious choice. Post induction, the differently fed cultures showed markedly different residual profiles for glucose, acetate, and phosphate. Ironically, the culture with added phosphate feed immediately began uptake of acetate, glucose, and phosphate, whereas the culture fed only glucose maintained acetate levels and continued to increase residual glucose levels (Figure 2C).

The immediate uptake of acetate in the culture with added phosphate, after induction, suggests that this cell culture is already primed for the metabolism of a 2-carbon source, such as acetate. This contradicts the general assumption that when E. coli is grown on a carbon source such as glucose, the glyoxylate bypass is unnecessary and is shut down by the dephosphorylation of the enzyme isocitrate dehydrogenase.¹¹ This allows isocitrate to be forwarded through the tricarboxylic acid cycle. Moreover, when considering the two different post-induction metabolic states of these cultures, there is no noticeable reduction of product formation in the culture fed only glucose of HMS174. The reasons for this are not clear.

In further studies on the HMS174 host cell strain, the immediate use of the residual acetate after induction could indicate the possibility of a new mixed feed strategy using glucose or phosphate before induction, then adding acetate separately during induction. This may increase the recombinant protein product formation and further elucidate the metabolic pathways used in an HMS174 induced culture.

ACKNOWLEDGMENTS

This work was supported by Grant 39129 from the Bill & Melinda Gates Foundation.

Garner G. Moulton is a senior research associate and Thomas Vedvick is director of process sciences, both in the process sciences department at the Infectious Disease Research Institute, Seattle, WA, 206.330.2549, gmoulton@idri.org.

REFERENCES

1. Farmer W, Liao J. Reduction of aerobic acetate production by escherichia coli. Appl Env Microbiol. 1997; 63(8):3205–10.

2. Lin H, Castro N, Bennet G, San K. Acetyl-CoA synthetase overexpression in escherichia coli demonstrates more efficient acetate assimilation and lower acetate accumulation: a potential tool in metabolic engineering. Appl Microbiol Biotechnol. 2006;71:870–4.

3. Mare L, Cimander C, Elfwing A, Hagander P. Feeding strategies for e. coli fermentations demanding an enriched environment. Bioprocess Biosyst Eng. 2006;30:13–25.

4. Shinar G, Rabinowitz J, Alon U. Robustness in Glyoxylate Bypass Regulation. PloS Comput Biol. 2009;5(3):1–8.

5. Saier M. Bacterial phosphoenolpyruvate: sugar phosphatransferase systems: structural, functional, and evolutionary interrelationships. Bacteriol Rev. 1977;41(4):856–71.

6. Sooan S, Chang D, Pan J. Acetate consumption activity directly determines the level of acetate accumulation during escherichia coli W3110 growth. J Microbiol Biotechnol. 2009;19(10):1127–34.

7. Wolfe A. The acetate switch. Microbiol Mol Biol Rev. 2005;69(1):12–50.

8. Kleman G, Strohl W. Acetate metabolism by escherichia coli in high-cell-density fermentation. App Env Microbiol. 1994;60(11):3952–8.

9. Seaver L, Imlay J. Hydrogen peroxide fluxes and compartmentatlization inside frowing escherichia coli. 2001;183(24):7182–9.

10. Eiteman M, Altman E. Overcoming acetate in escherichia coli recombinant protein fermentations. Trends in Biotechnol. 2006;24(11).

11. Cozzone A. Regulation of Acetate Metabolism by Protein Phosphorylation in Enteric Bateria. Annu Rev Microbiol. 1998;52:127–64.