April 1, 2015 (Vol. 35, No. 7)

Optimize No Cell Line before Its Time. Delicate Balancing of the Cell-Line Terroir Will Not Overcome Fundamental Quality Issues

ll successful cell-line engineering projects begin with a robust host cell line, one characterized by good growth and expression capabilities. “If a line grows well, there is the potential that favorable characteristics will translate and be heritable for any recombinant cell line you make from them,” says Alison Porter, Ph.D., head of mammalian cell culture R&D at Fujifilm Diosynth Biotechnologies. “But if a line grows poorly, chances are you will wind up with cells that have growth difficulties.”

A cell line with “ideal” characteristics is achievable through cell-line engineering, but it is not always clear which genes should be targeted. To resolve the targeting question, it may be necessary to go after multiple targets, even though doing so, warns Dr. Porter, typically “has time and cost implications.” Besides practical impediments, the search for targets may have to clear intellectual property hurdles.

Dr. Porter’s group eventually took the approach of directed evolution to exploit favorable characteristics that were already present in a heterogeneous host cell line. The  group also considered vector and signal sequence, selectable marker, and promoter—optimizing them individually before combining them.
The difference between Fujifilm Diosynth’s approach and more conventional cell -line development is the number of factors considered in preparation. “We looked at so many different possibilities to optimize the vector,” Dr. Porter recalls.

“If you want to find a good host cell line and vector you have to work at it,” insists Dr. Porter. “You will have to examine multiple avenues and have a robust screening technique for selecting the best combination.”

This approach allowed Dr. Porter’s group to narrow down the best lines and vectors early in development. Screening began with a focus on basic growth and productivity characteristics, then assessed more complex attributes. All the way through, cells were grown under conditions as similar as possible to those needed for the eventual process. It was known, for example, that fed-batch conditions had to be accommodated. Only after reducing the candidates to two host cell lines and one vector did a full development program begin.

The take-home lesson: Use simple, straightforward screens early on to weed out undesirable characteristics such as poor growth, unsuitability to process conditions, and low productivity that subsequent lines are likely to inherit. Thereafter, apply more detailed characterization—but only after the likelihood of success is greatly improved.


If you are to find a good host cell line and vector, you should, Fujifilm Diosynth believes, examine multiple avenues and have a robust screening technique.

Engineering Viruses Out

During the production of biologics, the contamination of commercial bioreactors is a rare but devastating and costly event. Although infrequent, viral infection of a bioreactor can be catastrophic for the manufacturer, resulting in the loss of product, temporary withdrawal from the market, and extensive clean-down costs. Contamination events can also have an impact on patient safety and regulatory implications. Viruses can be introduced into biologic manufacturing processes through a variety of routes including contaminated raw materials, infected cell lines or via environmental factors during the manufacturing process.

One particularly prevalent and troublesome virus is the minute virus of mice (MVM). Even with extensive programs in place for the physical removal of these viruses, MVM has been detected in manufacturing systems. “The virus is exceptionally durable,” says Kevin Kayser, Ph.D., Director of Cell Culture Development at SAFC.

MVM is a parvovirus that infects a limited number of organisms, including rats, mice and hamsters (CHO). It can devastate a CHO culture system before anyone knows what is happening. The source? Rodents. Despite the best rodent control efforts, mice inevitably inhabit storage and process facilities and spread the virus through their urine and feces.

SAFC has “identified several techniques to reduce risk associated with viral contamination of biological drug manufacturing process,” such as high temperature short time treatments for the manufacture of critical raw materials and through cell line engineering. SAFC focused their initial efforts on the MVM virus, which is the most prevalent virus detected in animal component-free biological manufacturing processes.

MVM infects cells by binding to their outer membranes. Specifically, they attach to 2,3-linked sialic acid residues. This mechanism has been demonstrated in glycoarrays and validated by Kayser’s work.

Once MVM binds it enters the cell, migrates to the nucleus, hijacks the host’s DNA-replicative machinery and multiplies. The mechanism of viral infection event provides several potential intervention points. “You might go after the binding event, the receptor, viral-mediated uptake, transport from membrane to nucleus, or you might even block replication,” Kayser noted.

Knocking out sialic acid to prevent binding altogether was out of the question, because these sugars render glycoproteins more humanlike, improving both safety and efficacy. Instead, Kayser used gene-editing techniques to knock out several genes affecting viral entry into the cell and subsequent viral propagation in the cell.

Through collaborations with internal and academic groups, Kayser deleted surface sugar structures that he believed were responsible for binding, while keeping an eye on maintaining the cells’ ability to express therapeutic proteins with desirable protein quality attributes.

One significant change involved introducing 2,6-linked sialic acid structures that are native to humans but not to CHO. “MVM does not bind to 2,6-linked sialic acid,” Kayser explains, “which is why the virus does not infect humans.”

The approach to rendering CHO cells resistant to MVM infection is not that simple, however. Alternate uptake mechanisms might eventually arise that could thwart the glycan receptor mechanism. Kayser indicated his group might also employ additional genetic knockouts that combine redesign of receptor glycans, uptake and trafficking and/or the replication of viruses inside cells.

Transient to Permanent

With transient transfection coming into its own as a mid-scale production method, bioprocessors have acquired a powerful tool for identifying excellent candidate cells for later stable transfection. Previously considered a satisfactory technique for obtaining one hundred milligrams or so of protein, transient methods now generate tens of grams while helping to identify clones that are good candidates for stable transfection.

MaxCyte, for example, claims that its electroporation approach to transient transfection provides high transfection efficiency and viability, seamless scaleup or scaledown without reoptimization, and expeditious development of stable cell lines. Stable cell lines may be achieved less than two months after initial transfection, according to the company.

“Electroporation provides more than 1 g of protein for a 40 to 50 L culture,” says Weili Wang, Ph.D., MaxCyte’s principal scientist. By scaling up to 1,000 L, 50 g of protein is easily achievable. Dr. Wang points out that 50 to 100 g is usually enough for a toxicology study and that this quantity can be produced in a single batch. For highly potent vaccines or antibody-drug conjugates, 50 to 100 L may be sufficient for the world’s yearly supply. MaxCyte has published a white paper on its “vaccine in a box” idea.

Electroporation is, moreover, fast. Transfection takes just thirty minutes, and the process can be over in one day. “Electroporation is almost always superior to chemical transfection based on lipids,” asserts Dr. Wang.

Larger scale electroporation-transfections have been demonstrated. According to Dr. Wang, however, they lack product consistency.

How does MaxCyte make the leap from transient to permanent transfection? It does not do so directly, but through a winnowing process.

“The first step in stable cell-line generation is always transfection. That is how you uncover good candidates for stable cell-line screening,” elaborates Dr. Wang. “But if transcription is your bottleneck, you will not readily arrive at satisfactory stable cell-line candidates. If you can transfect only 30 or 50 cells out of 100, that is all you have to work with. If you obtain 98 viable, transfected cells, however, you have that many more healthy, viable, productive cells to screen.”

Low transfection rates force cell-line developers to use ancillary techniques for sorting viable cells and identifying high-producing clones. Through the law of larger numbers and the more-efficient transfection process, developers can begin cell optimization from a more “talented” pool of candidates, thereby cutting down on development time and cost.

Industry Perspective

The urgent needs of biomanufacturers are the same as those in other industries, says Craig Smith, Ph.D, vice president of bioproduction at Thermo Fisher Scientific: “They are not only looking for operational efficiency, they are also accelerating development of new products.”

Some emerging factors, however, make those needs more acute for biotherapeutics. Competing within a crowded biosimilars market drives the need for speed-to-market and efficiencies, both during development and in commercial launch. Many biosimilars companies are chasing a fairly defined set of molecules that are or will soon go off-patent. Reaching the market before anyone else and achieving cost-effective manufacturing are, in Dr. Smith’s view, serious competitive advantages: “Companies that can make more of a product for less money will use cost as a competitive advantage.”

Lower manufacturing costs also matter to innovator companies, which will lose revenue and market share when biosimilars are approved. With small molecule drugs, originator firms more or less abandon the brand in the face of 90% price reductions. With biosimilars, however, discounts are expected to be in the 20–40% range. In other words, brands still have a chance. Innovator firms will therefore re-examine production methodologies to improve yield in situations where they can compete with biosimilars.

Dr. Smith mentions the potential for next-generation sequencing to drive personalized medicine, which will generate more therapeutics—as well as more-targeted therapeutics.

“When companies realize the effects of more molecules/therapies that are more specific or targeted, drug companies will look for ways to drive more processes/molecules though their facilities,” Dr. Smith explains. “Single-use technology and yield optimization will become important factors. They will enable flexibility and, consequently, operational efficiency.” These factors will form the backdrop of biopharmaceutical development over the next decade, and they will be hotly debated at upcoming seminars and in the academic and trade press.

With these issues in mind, Dr. Smith’s team intends to concentrate on upstream biomanufacturing, particularly cell-line and media/feed development. Cell-line development may be speeded in several ways, but all lead to obtaining a clone of interest that produces protein at desired productivity and quality levels.

Many smaller biotech and biosimilar companies lack this expertise. “They have the molecule, but they lack the experience to produce it,” remarks Dr. Smith. “They don’t have legacy cell lines to work with, and almost certainly will lack media development skills. Those firms need external partners.” One such partner is, presumably, Thermo Fisher.

“Ultimately, the goal is to reduce the number of campaigns per year required to meet worldwide demand,” advises Dr. Smith. “For every campaign that can be eliminated, companies can realize savings in direct materials, manufacturing, and QC labor costs.”

On the media side, Dr. Smith notes how greater functionality can help processors combine or eliminate process steps. We are familiar with various dry versus liquid media and modern reconstitution technologies. Most of these require tweaking to maintain the proper pH and osmolality. Thermo Fisher, however, has been selling a media in granulated format which, according to Smith, provides media at the proper pH and osmolality: “It’s ready to use in hours instead of days like conventional dry powder format.”

* This story has been corrected from an earlier version.  The section pertaining to SAFC subtitled “Engineering Viruses Out”  has been updated.

Previous articleCommon Bacteria Poised to Become Drug-Resistant Superbugs
Next articleEmploying CRISPR Technology in Drug Discovery