Andrea Gough’s professional path began in medical microbiology, screening hospital samples for methicillin-resistant Staphylococcus aureus. A fascination with genetics soon pulled her toward biologics, first as an applications scientist helping customers integrate semi-solid screening and high-content imaging, then as a product manager guiding antimicrobial susceptibility platforms at Thermo Fisher.
Today, she is Senior Director for the Solentim portfolio at Advanced Instruments, leading a team that supplies single-cell dispensing, imaging, and clonality documentation tools to biopharma and CDMO clients worldwide. Years spent troubleshooting in customer labs have given her a panoramic view of cell line development (CLD) bottlenecks and emerging strategies that shorten timelines without compromising regulatory expectations.
This concept is discussed in greater detail in an episode of the Smart Biotech Scientist Podcast, hosted by David Brühlmann, founder of Brühlmann Consulting.
Why a Truly Optimal Cell Line Still Starts With DNA Choice
Gough emphasizes that every downstream efficiency hinges on the first decision: how the expression cassette is integrated.
Random Integration
- Works with well-established workflows and host cell lines
- Creates “needle in a haystack” variability because insertion sites and copy numbers are uncontrolled
- Requires large clone banks and extensive screening to find rare, stable, high producers
Semi-Targeted Transposon Systems
- Insert multiple copies into transcriptionally active regions
- Deliver higher average titers and better long-term stability
- Reduce the total number of clones that must be isolated and characterized
Selecting a semi-targeted system can take clone stability studies off the critical path, freeing teams to pursue clinical material from pools or early clones while confirmatory data accumulate.
Core Bottlenecks Across the CLD Workflow
Gough groups recurring delays into five categories and explains how each can be addressed.
Post-Transfection Recovery
Specific host–vector or molecule combinations recover slowly. She has seen pools that need four weeks before selection can begin. Process optimization at this step often repays itself later by increasing the percentage of viable clones.
High-Efficiency Single-Cell Seeding
Limiting dilution remains widespread, yet plating hundreds of microplates consumes time and consumables. Modern piezo, acoustic, or microfluidic dispensers now achieve >90 per cent single-cell efficiency, shrinking twenty or more plates to fewer than ten.
The critical bottlenecks is they're essentially all over the process, really. Starting with right at the beginning is the DNA. Whether customers are using random integration, that can be a bottleneck because ultimately if you're integrating your DNA randomly, then you don't know where it's going to go. You're going to have essentially this needle in a haystack aspect. You're trying to find that good clone. So essentially that's where it all starts is right at the beginning is your DNA and your choice of methodology and random integration is perfectly fine. And that's the way the market started out, is using that.
Proof of Clonality
Regulators expect visual or sequence-based proof that the production line originates from one progenitor. Early semi-solid pickers lacked high-resolution optics, forcing a second round of dilution. New imaging systems capture a good area immediately after dispensing, store raw files, and build automated audit trails that withstand BLA review.
Multi-Parameter Clone Ranking
Teams must compare growth kinetics, fed-batch productivity, and product quality attributes (PQA) before they advance candidates. Instruments that pair titer measurement with online viability or capture early glycan profiles allow faster attrition of poor performers.
Stability Passage Testing
Traditional 60–90-day protocols keep lines in shake flasks for 30+ generations. When semi-targeted integration is used, many developers now run abbreviated studies in parallel with tox or Phase I batches instead of waiting for completion.
The Changing Toolkit: From Fax Sorters to Integrated Workflows
Historically, fluorescence-activated cell sorting (FACS) served as the main single-cell isolation technology. Gough notes its strengths—precise gating and publication history—but also the drawbacks that have pushed many groups toward gentler dispensers: high shear stress, complex maintenance, and the need for specialist operators.
In response, vendors introduced:
- Semi-solid colony pickers that locate secreting clones by fluorescent halos
- Inkjet-style or piezoelectric droplet devices that document each dispensing event
- Modular incubators that automate feed, imaging, and transfer into scale-down fed-batch plates
Solentim’s VIPS instrument, for example, dispenses a nanoliter droplet, uses onboard AI to confirm cell numbers in real time, and then captures a full-well image without user intervention. The goal is to combine seeding, documentation, and expansion into one traceable pipeline.
Artificial Intelligence Enters Clone Selection
Gough sees AI primarily as a labour saver. During dispense, convolutional neural networks quickly decide whether an object is a single cell, doublet, or debris. After expansion, algorithms pre-screen thousands of well images and label them clonal or non-clonal, leaving humans only to inspect ambiguous cases.
Future directions under discussion among her customers include:
- Mining early morphology patterns to predict high producers or unstable lines
- Combining growth curves, titer, and image-derived features to prioritize clones before benchtop bioreactor runs
- Feeding historical production campaigns into models that recommend optimal media or feed strategies for new molecules
Although definitive genotype-to-phenotype prediction remains distant, Gough believes large annotated image sets will soon reveal subtle phenotypic markers invisible to the naked eye.
Beyond CHO: Alternative Hosts in Commercial Use
While Chinese hamster ovary cells still dominate monoclonal antibody manufacturing, Gough reports rising interest in specialized hosts:
- HEK293 remains the workhorse for adeno-associated virus and lentiviral vector production.
- HeLa is explored in certain viral platforms that demand high yields under serum-free conditions.
- Insect cells are adopted for baculovirus‐mediated vector manufacture, providing lower protease activity for fragile capsids.
- Mesenchymal stem cells are becoming standard for exosome-based delivery vehicles.
- Induced pluripotent stem cells underpin allogeneic cell therapy pipelines where autologous workflows prove too costly or slow.
Each new host introduces unique seeding densities, media supplements, and documentation requirements, reinforcing the need for flexible instrumentation.
Practical Advice for Scientists Launching Their First CLD Program
Gough distils her customer conversations into three recommendations:
- Secure experienced partners early, whether through CDMOs, platform cell line licensors, or instrument vendors offering on-site application support.
- Choose modular tools that adapt to your protocol instead of forcing conformity to a single vendor workflow. Dispense, imaging, and analytics should be integrated, allowing custom media or plate formats.
- Capture every piece of evidence from day one. Store certificates of analysis for all raw materials, raw microscopy files, instrument audit logs, and passage records in a searchable repository. Retrieving them months later during agency queries is far harder.
With the technology today, low productivity and stability issues can probably be targeted in one of two ways. One, you screen more clones and sometimes that is necessary, you screen more clones. It's quite labor intensive and that's the way you go. But at least if you're using random integration, you're finding that needle in a haystack. You're more likely to if you screen more. Or as I mentioned earlier, you look at those semi targeted integration technologies like transposons.
Systematic Troubleshooting of Low Productivity or Instability
When expression titers disappoint, Gough suggests two complementary strategies:
- Expand screening depth. If using random integration, adding more clones statistically increases the chance of finding a rare high-producer. High-throughput dispensers and miniaturized fed-batch assays can mitigate workforce costs.
- Adopt semi-targeted integration. Technologies such as transposons often raise the baseline productivity across the whole clone pool, reducing reliance on statistical luck and improving genetic stability.
Emerging Trends Likely to Reshape Development Timelines
- Screening early bulk pools produced with transposon methods to supply material for preclinical studies
- Leveraging AI image analysis to remove subjective steps and shorten documentation reviews
- Running parallel rather than sequential stability assessments once semi-targeted lines demonstrate historically robust behaviour
- Wider use of exosome or viral platforms derived from non-CHO hosts, demanding instrument platforms that handle multiple cell morphologies
Final Remarks
Andrea Gough’s experience in startup labs, diagnostic product management, and instrumentation development leads to a single overarching message: The fastest route to a licensed biologic is not a single technology but a mindset that unites flexible tools, meticulous recordkeeping, and integration strategies tailored to each molecule.
Random integration, limiting dilution, and manual image checks still work. Yet, they compete against semi-targeted vectors, automated dispensers, and AI classifiers that objectively prove clonality in a fraction of the time. As regulators continue to scrutinize cell line provenance, teams that invest in early evidence capture and scalable workflows will be best positioned to navigate future therapeutics from first DNA construct to commercial bioreactor.
About Andrea Gough
Andrea is the Senior Director for the Solentim Portfolio at Advanced Instruments (AI). Andrea has a first class Bachelor’s degree in Applied biochemistry and Molecule Cell Biology from the university of the West of England, UK and since leaving university has gained 15 years’ experience in the biologic drug development space; focused in the field of cell line development (CLD).
Andrea started in this field as a Field Application Scientist at Molecular Devices (formerly Genetix Ltd) supporting customers through their CLD challenges and their implementation of automation. She then progressed through several Global Product Management roles at Thermo Fisher Scientific, Solentim and Horizon Discovery until she moved to Advanced instruments where she initially took on the role of Senior Manager in Business Development for the Solentim Portfolio.
Andrea’s continued progression at AI, leading to Senior Director, have continued to focus on supporting customers enhance their CLD efficiencies with the advancements provided by the AI Solentim portfolio offering.
Connect with Andrea Gough on LinkedIn.
David Brühlmann is a strategic advisor who helps C-level biotech leaders reduce development and manufacturing costs to make life-saving therapies accessible to more patients worldwide.
He is also a biotech technology innovation coach, technology transfer leader, and host of the Smart Biotech Scientist podcast—the go-to podcast for biotech scientists who want to master biopharma CMC development and biomanufacturing.
Hear It From The Horse’s Mouth
Want to listen to the full interview? Go to Smart Biotech Scientist Podcast.
Want to hear more? Do visit the podcast page and check out other episodes.
Do you wish to simplify your biologics drug development project? Contact Us
