Knowledge

Fed-Batch Feeding Strategies: How Feed Rate Sets Cell Density and Soluble Yield

High-cell-density fermentation means growing a culture to a biomass concentration far above what a simple batch of medium supports, by adding the carbon source gradually instead of all at once. The technique is called fed-batch: feed medium is dosed into the vessel during the run according to the cells' consumption of nutrients, which changes how the cells grow and can raise both the yield and the quality of the recombinant protein they make [6].

It matters because volumetric yield — grams of product per litre of broth — is the product of two things that do not move together: how much biomass is in the tank and how much product each cell makes. Secretion of a heterologous protein depends both on biomass concentration and on the specific product secretion rate, and that secretion rate is itself not constant as the specific growth rate varies [4]. Feeding is the lever that sets both at once.

Feeding to limit growth, not to maximise it

The most common design is a carbon-limited feed: glucose is supplied slowly enough that the cells are always short of it, so the feed rate, not the cells' maximum capacity, sets the growth rate. In an E. coli shaken-culture system built around enzymatic glucose release, the process ran in three steps: controlled growth by glucose-limited fed-batch to an OD600 of about 10, then addition of growth boosters together with an inducer giving efficient protein synthesis over 3 to 6 hours, then a slow growth period of 16 to 21 hours during which the recombinant protein was slowly synthesised and folded [1]. Cell densities corresponding to 10 to 15 g l⁻¹ cell dry weight were achieved, and volumetric yields of soluble recombinant protein were typically over 10-fold higher than in standard cultures in LB, Terrific Broth or mineral salt medium [1]. The authors attribute the improvement to a well-controlled physiological state maintained throughout the process [1].

The quality point in that work is worth separating from the yield point. What improved most was the proportion of soluble — correctly folded — protein in the harvested cells [1]. Slow, controlled feeding does not only build biomass; it changes the ratio of folded to misfolded product you have to deal with downstream.

Glucose-limited fed-batch also scales. Four disulfide-bonded proteins were produced in stirred-tank bioreactors using an E. coli strain with the cytoplasmic reducing pathways intact, co-expressing a sulfhydryl oxidase and a disulfide bond isomerase, in glucose fed-batch fermentations on chemically defined minimal media [5]. Without process optimisation, purified human single chain IgA 1 antibody fragment reached 139 mg/L and hen avidin 71 mg/L, while purified yields of human growth hormone 1 and interleukin 6 were around 1 g/L [5].

The shape of the feed profile

One group modelled product accumulation iteratively in spreadsheet software and used an optimiser to search for the feed time course that maximises volumetric productivity [4]. The optimum feed phase they found consisted of an exponential feed at the maximum specific growth rate, followed by a phase with a linearly increasing feed rate and consequently a steadily decreasing specific growth rate [4]. Experimental evaluation in two independent fed-batch cultures correlated well with the optimised model data and gave a 2.2-fold improvement in volumetric productivity [4].

That result is a useful shape to carry into process development: grow fast while biomass is cheap to build, then deliberately let the growth rate fall while the cells keep producing. The modelling approach also accepts further biological and technological constraints, or different functions for specific productivity and biomass yield, as additional inputs [4].

Temperature as part of the feeding strategy

Feed rate is not the only variable under the operator's hand during the feed phase. A temperature-limited fed-batch technique for a Pichia pastoris Mut+ strain was compared with the traditional methanol-limited fed-batch for a secreted recombinant fusion protein [2]. The methanol-limited process produced about 1 g L⁻¹, but the product was considerably degraded by protease [2]. Under the temperature-limited process, the yield rose to 2 g L⁻¹ of full-length product with no proteolytic degradation observed [2].

Flow cytometry explained part of the difference. In the methanol-limited process, the percentage of dead cells rose rapidly during the initial methanol feed phase, reaching about 12% after some 40 to 70 hours of methanol feeding [2]. In the temperature-limited process the cell death rate was low and constant, reaching 4% dead cells after about 150 hours of methanol feeding, and this correlated with lower protease activity in the culture supernatant [2]. Specific alcohol oxidase activity in the temperature-limited process was 3.5 times higher [2]. The authors propose three contributing mechanisms: reduced proteolysis from the lower temperature, reduced proteolysis because fewer cells lysed and released protease, and a higher synthesis rate from the higher alcohol oxidase activity [2].

A separate P. pastoris study, on a Δoch1 strain expressing a plant peroxidase, examined temperature, pH and dissolved oxygen together against cell physiology, morphology, lysis, productivity and product purity [7]. Cultivation at 30 °C gave low specific methanol uptake during adaptation and a risk of methanol accumulation during the run, while 20 °C was highly beneficial for both productivity and product purity [7]. Cell cluster formation in that strain was a function of the carbon source rather than the process parameters, and went along with cell lysis [7].

Beyond microbial hosts

The same logic applies in mammalian and algal systems. Fed-batch culture is typically used for recombinant protein production in CHO cells, where feed medium containing multiple components affects cell growth characteristics and can improve both the yield and the quality of the protein [6]. Heterotrophic microalgae have reached cell densities of more than 100 g l⁻¹ cell dry weight with Chlorella , Crypthecodinium and Galdieria species by controlling the addition of organic carbon and energy sources in fed-batch mode, in inexpensive, well-defined mineral medium typically supplemented with glucose [3].

What can go wrong, and what to ask

Several of these studies show that the feed phase is also when product is lost. In an engineered E. coli process for geraniol, 81.4% of the volatile product was lost during the first 5 hours of fermentation in a solvent-free system, and most of the remainder was biotransformed by the host into geranyl acetate; adding isopropyl myristate to form an aqueous–organic two-phase system prevented the volatilisation, and under controlled fermentation conditions production reached 2.0 g/L [8]. The lesson generalises: an improved feed profile is worth little if the product is being consumed, modified or degraded in the tank at the same time.

For a manufacturer assessing a process or a process-transfer partner, the questions follow directly from the evidence. What limits growth during the feed — carbon, temperature, or both, and is that choice justified by data on cell lysis and product integrity, not just by titre [2][7]? Is the feed profile a fixed recipe or has it been optimised against a model of specific productivity versus specific growth rate [4]? Is the reported yield soluble, purified product or total expressed protein, since the two can diverge sharply [1][5]? And does the process run on chemically defined minimal medium, which makes the carbon and nitrogen balance explicit and transferable [5]? Higher volumetric yields also shrink the culture volume needed, which reduces the time and effort spent on downstream processing [1]. At scale, efficient and economically feasible large-scale bioprocesses remain the hard part even where the fundamental knowledge is available [3].

References

  1. Krause M, Ukkonen K, Haataja T, Ruottinen M, Glumoff T, Neubauer A, Neubauer P, Vasala A. A novel fed-batch based cultivation method provides high cell-density and improves yield of soluble recombinant proteins in shaken cultures — Microbial cell factories, 2010-02-19.
  2. Jahic M, Wallberg F, Bollok M, Garcia P, Enfors SO. Temperature limited fed-batch technique for control of proteolysis in Pichia pastoris bioreactor cultures — Microbial cell factories, 2003-06-18.
  3. Bumbak F, Cook S, Zachleder V, Hauser S, Kovar K. Best practices in heterotrophic high-cell-density microalgal processes: achievements, potential and possible limitations — Applied microbiology and biotechnology, 2011-05-13.
  4. Maurer M, Kühleitner M, Gasser B, Mattanovich D. Versatile modeling and optimization of fed batch processes for the production of secreted heterologous proteins with Pichia pastoris — Microbial cell factories, 2006-12-11.
  5. Gąciarz A, Khatri NK, Velez-Suberbie ML, Saaranen MJ, Uchida Y, Keshavarz-Moore E, Ruddock LW. Efficient soluble expression of disulfide bonded proteins in the cytoplasm of Escherichia coli in fed-batch fermentations on chemically defined minimal media — Microbial cell factories, 2017-06-15.
  6. Xu WJ, Lin Y, Mi CL, Pang JY, Wang TY. Progress in fed-batch culture for recombinant protein production in CHO cells — Applied microbiology and biotechnology, 2023-01-17.
  7. Gmeiner C, Saadati A, Maresch D, Krasteva S, Frank M, Altmann F, Herwig C, Spadiut O. Development of a fed-batch process for a recombinant Pichia pastoris Δoch1 strain expressing a plant peroxidase — Microbial cell factories, 2015-01-08.
  8. Liu W, Xu X, Zhang R, Cheng T, Cao Y, Li X, Guo J, Liu H, Xian M. Engineering Escherichia coli for high-yield geraniol production with biotransformation of geranyl acetate to geraniol under fed-batch culture — Biotechnology for biofuels, 2016-03-11.

Also on Instagram

All articles