Scaling up mammalian cell cultures and microbial fermentations involves similar, and familiar, considerations of cell expansion and density, mass and heat transfer, mixing, gassing, nutrient delivery, and generally how process operations affect product yield and quality. While the buzzwords remain nearly identical their influence on fermentation scaleup is much greater, for the simple fact that fermentations expand and metabolize much faster than mammalian cultures.
For Aaron Pilling, director of upstream process development at KBI Biopharma, the key driver when scaling up microbial fermentations is throughput, particularly at the bench scale. “Process developers are looking for experiments that provide high statistical power, through a large number of individual small-scale bioreactors.”
As noted by GE Healthcare’s Parrish Galliher, chief technical officer for upstream processing, microbial cultures grow as much as 75 times faster than mammalian cells, with some species doubling every half hour. “Such growth rates equal 20 to 75 times the heat evolution and oxygen demand of mammalian cells,” Galliher writes. Benchtop fermenters are therefore made of jacketed glass or stainless steel, whereas production-scale bioreactors are stainless steel vessels that can withstand overpressures of up to 5 bar.
From the smallest scale...
The small-scale bioreactors Pilling is referring to are the micro- and mini-bioreactors often used in scaledown mode for trouble-shooting, process development, and scaleup feasibility studies.
KBI Biopharma uses the ambr platform, which includes thambr® 15 Microbial Single-Use Advanced Micro Bioreactor System from Sartorius business unit TAP Biosystems. ambr 15 for fermentation is an automated microscale bioreactor system based on the standard ambr 15 platform used by virtually every bioprocess development company. The system includes disposable, 15 mL microbioreactor vessels, an automated workstation, and software. The ambr 250® system works similarly, but at 250 mL volumes.
“We apply ambr throughout process development, including for late-stage process design and characterization, for high-throughput DOE [design of experiment], process characterization, and even normal operating range studies requiring high-throughput statistical power,” Pilling explains.
When qualifying a larger-scale fermentation based on micro-scale experiments, developers apply industry-standard statistical approaches that need not duplicate large-scale results perfectly, Pilling notes.
“The processes are usually similar, with known offsets. You look at the output data and decide whether the scaledown model is an acceptable model. Most process developers would be concerned if the offsets were large and related to product quality, safety, or effectiveness. But if they involve small changes in titer or cell density, that’s often okay.”
Another microbioreactor platform to consider for fermentation is the Lector line from Mp2-labs, which is based on microplates with well volumes of between 800 and 2400 μL. Biolector I processes up to 48 samples; Pro and Robo models up to 32 samples, all under control of temperature, humidity, and atmosphere.
As with mammalian cultures, the knowledge transferable among experimental fermentation scales depends on many factors. “Comparing results between ten liter and two thousand liter fermenters is relatively easy,” Pilling says, but drawing conclusions from a larger number of scales becomes problematic. “That requires understanding the scalability of laboratory systems to other lab systems, and each of those laboratory systems to your representative manufacturing scale.”
...To the largest
“Mammalian cells are much slower-growing and more fragile than microbes such as bacteria and yeast,” says Brady Cole, vice president for commercial operations at ABEC. “The challenge with mammalian cells is obtaining good mixing and delivering oxygen without damaging the cells, especially for today’s high density cultures. Microbial cultures grow very quickly and are quite robust. Heat removal is the main issue, especially at larger volumes. Microbes generate more metabolic heat, and you’re also putting more energy in with higher agitation rates. If it gets too hot inside the fermenter, the bacteria die.”
Biocompare’s Bioreactors Search Tool
Find, compare and review bioreactors
from different suppliers Search
Single-use processing in fermentations has lagged behind disposable mammalian biomanufacturing for several reasons. Fermentations tend to span much larger volumetric scales, to the tens of thousands of liters, which are difficult to execute in plastic. Experts often cite bioreactor design, difficulty in achieving optimal mixing, and more robust requirements for gas exchange and heat removal to explain the lag in single-use adoption for fermentations. However, the true significance of bioreactor design (other than size considerations) with respect to these factors is unclear.
In early 2018 ABEC introduced a large-scale single-use fermentation product, within its Custom Single Run (CSR®) product line, that establishes a new volumetric benchmark, 1000 liters, for microbial fermentation. Part of ABEC’s Custom Single Run (CSR®) line of bioreactors, the device is an industry first.
The CSR bioreactor overcomes gassing, mixing, and heat removal limitations of previous-generation single-use microbial bioreactors with performance comparable to that of legacy stainless steel systems, according to the company. ABEC has similarly pushed the volumetric limits for mammalian single-use cultures as well, with a plastic bioreactor sporting 4,000-liter working volume.
SIngle-use fermentations
That 4,000 liters is the volumetric benchmark for cell culture, compared with 1,000 liters for fermentation, is a story in itself, namely that single-use bioprocessing is much further along for mammalian systems than for microbial processes.
To explain this rather large difference, commentators have cited difficulties in duplicating the robust mixing, oxygen transfer, and kLa achievable in stainless steel vessels within plastic bioreactors. These are indeed factors, but, according to Pilling, the fundamental difference between steel and plastic is that fixed tank bioreactors may be pressurized to facilitate mixing and mass transfer, but plastic bags cannot.
“Both cultures employ the same physical parameters but fermentation requires a higher load, higher oxygen flow, and higher mixing capabilities. A lot of work goes into the design of agitators, mixers, sparging systems, and pressurization to allow facile fermentation scaleup in single-use vessels.”
Through its HyPerforma™ Single-Use Fermentor product line, Thermo Fisher Scientific provides a scaleup path for microbial fermentations from 30 liters to 300 liters of working volume, or roughly from process development through small production runs depending on the product and titer. The HyPerforma bioreactors are vertically centered, use top-driven impellers and crossflow sparging, and feature a 5:1 turndown ratio. A high turndown ratio means the bioreactor can operate fully and optimally using a fraction of its usable volume, in this case 20%. High turndown helps to eliminate a seed train step when growing a culture out before the main act of fermentation and expression.
GE Healthcare’s Xcellerex™ XDR-50 MO fermenter system is a single-use 50 L stirred-tank system with the usual advantages of single use, comprising disposable fermentation bags housed within a jacketed, stainless steel vessel. XDR-50 includes essential process instrumentation, automation, and self-contained temperature control. Special features include a two-stage impeller supporting high oxygen transfer rates, dimple-jacketed heat transfer surface, and, according to the company, “performance comparable with hard-piped, stainless steel fermentor systems.”
This development activity points to a frontier of sorts—and a commercial opportunity—for single-use bioprocessing for microbial fermentations. But Eric Langer, managing partner at BioPlan Associates, is less excited. "My understanding is that single-use manufacturing is not taking off for microbial processes. The opportunities exist, but current microbial platforms based on stainless steel bioreactors are quite large and quite well established. It's a kind of 'if it ain't broke' situation. Although microbial makes up nearly twenty percent of biologic production platforms, we haven't seen as much interest as we might expect in single-use processing."