Detaching Vero, MDCK and HEK 293 at Scale: An Animal-Free Enzyme for Vaccine and Viral-Vector Production

Animal-free trypsin for vaccine production is no longer a niche preference — it is increasingly the default expectation of quality units reviewing an adherent cell process. This application note looks at what a recombinant, animal-origin-free trypsin-like enzyme does when it is used to release the three workhorse adherent lines of viral manufacturing — Vero, MDCK and HEK 293 — what the manufacturer has actually measured, what has not been measured, and how to design a short bridging test before you change anything in your own process.

Why animal-origin trypsin is a documentation problem, not just a preference

All three lines are established substrates for licensed products. Vero is the continuous cell line most widely accepted by regulators as a viral vaccine substrate, with more than three decades of use in polio and rabies vaccine production and a growing list of newer licensed products (Barrett et al., Expert Rev Vaccines 2009). MDCK is the workhorse for cell-culture-derived influenza vaccine, and it behaves differently enough from Vero that the two cannot be assumed interchangeable at process level (Genzel et al., Appl Microbiol Biotechnol 2010). HEK 293 underpins most transient AAV and lentiviral vector production.

Porcine pancreatic trypsin has been the standard dissociation reagent for adherent culture for decades. Its weakness is not enzymatic — it is provenance. A US FDA study screening vaccine-relevant cell substrates for porcine circovirus put the issue plainly: “animal-derived raw materials, such as cells, trypsin, and serum, can be a major source of introducing virus contamination in biological products” (Ma et al., Vaccine 2011).

That is not a theoretical concern. In 2010, porcine circovirus type 1 DNA was detected in a licensed live-attenuated human rotavirus vaccine, prompting a full manufacturer investigation and a temporary regulatory suspension of use in the United States (Dubin et al., Hum Vaccin Immunother 2013). Independent screening by NIBSC subsequently found PCV1 DNA in live poliovirus seeds from one manufacturer, although the finished inactivated product tested negative (Gilliland et al., Biologicals 2012). The downstream commercial effect was measurable: one analysis recorded a 93% drop in doses of the affected vaccine in the month after the suspension notice (Dore et al., Pharmacoepidemiol Drug Saf 2012). An industry-wide consortium later pooled real viral contamination events across biologic manufacturers and analysed which contaminants appeared, where they came from, and what recovery cost (Barone et al., Nature Biotechnology 2020).

Removing animal-derived material from the process does not eliminate adventitious-agent risk, but it removes one well-documented route in and shortens the supply-chain file you have to defend. That is the argument for a recombinant enzyme, and it is an argument about traceability and viral safety — not about detachment speed.

What CellTrypase is, and how it is specified

CellTrypase (c-LEcta GmbH, a Kerry company) is a recombinant trypsin-like serine protease. The gene originates from the fungus Fusarium oxysporum and is expressed in a Bacillus sp. host at GMO safety level S1. The mature enzyme is approximately 22 kDa and cleaves on the carboxyl side of lysine and arginine residues — the same specificity that makes mammalian trypsin useful for releasing adherent cells. It is manufactured without antibiotics, animal-derived raw materials or materials associated with TSE/BSE risk.

The specification a process development scientist actually needs:

Source: c-LEcta GmbH (Kerry Biotechnology Centre), PIS CellTrypase v4.0, valid 18 Aug 2026.

Two handling points matter for scale-up. First, no trypsin inhibitor is required — activity is quenched by dilution in buffer or medium, which removes a reagent, a hold step and an inhibitor-carryover question from the process. Second, the recommended dose is approximately 1 mL per 25 cm² of growth surface, with the enzyme pre-warmed to room temperature or 37 °C before addition.

Manufacturer release data for Vero, MDCK and HEK 293 — including the MDCK outlier

c-LEcta reports application data for four production lines. These are the only cell types for which performance figures exist, and they are reproduced here in full, including the result that is least flattering.

Source: c-LEcta GmbH (Kerry Biotechnology Centre), PIS CellTrypase v4.0, valid 18 Aug 2026. CHO-K1 is included for completeness. Values are the manufacturer's; BioHippo has not independently repeated these experiments.

Two honest caveats about this table. First, the product information sheet does not define the reference method behind the yield percentages, which is why values sit slightly above 100% — treat “106%” as an internal manufacturer metric rather than a validated comparison against a named commercial reagent, and confirm the reference definition before you carry these numbers into a comparability protocol. Second, all four datasets were generated in serum-containing medium. If your Vero or MDCK process runs serum-free, your detachment kinetics will not match these numbers; serum-free Vero culture is known to be considerably more sensitive to trypsin concentration than MDCK culture under otherwise comparable conditions (Genzel et al. 2010).

Why MDCK detaches ten times slower than HEK 293

At 23 minutes 32 seconds, MDCK release took roughly ten times longer than HEK 293 or CHO-K1 under the manufacturer's conditions. Viability and yield held up — 99% and 102% — so the cells came off intact, but they came off slowly. The figure is not an error, and it is worth understanding rather than hiding.

MDCK II forms a genuine polarised epithelium with functional tight junctions built from claudin family proteins, whereas HEK 293 is routinely used as a tight-junction-free counter-cell precisely because it does not (Inai, Methods Mol Biol 2011). A confluent MDCK monolayer therefore presents both cell–substrate attachment and a mature cell–cell junctional belt to be broken, while HEK 293 presents mostly the former. Any dissociation reagent, animal-derived or recombinant, has to work through that difference.

Four practical consequences:

• A 20-plus minute enzyme exposure changes the hold time of every vessel in a parallel harvest. Budget for it in your batch record.
• Do not set one global incubation time across a mixed-substrate facility. On the manufacturer's own data, the same reagent needs roughly ten times longer on MDCK than on HEK 293.
• Expect the 1.1 mM EDTA in the formulation to be doing real work on MDCK, since calcium chelation destabilises cadherin-dependent junctions — which is also why re-adding calcium-containing medium is what stops the reaction.
• The 10x concentrate exists and is an obvious variable to test for slow-releasing lines, and longer exposure at 37 °C is worth evaluating against room-temperature incubation, since MDCK sublines differ substantially in adhesion behaviour. But c-LEcta publishes no MDCK data at 10x, so treat that as an experiment to run, not a result to assume.

Trypsin has two jobs in an influenza process — this enzyme is documented for one

This distinction is missed often enough to be worth stating directly. In MDCK-based influenza manufacturing, trypsin appears at two separate points:

1. Cell dissociation — releasing the monolayer for passaging or harvest.
2. Haemagglutinin activation — exogenous protease supplied during infection to cleave HA0 into HA1/HA2 and permit multi-cycle replication.

The second role is a genuine process parameter with measurable effects on titre. Trypsin added during infection has been shown to interfere with the antiviral host response in MDCK cells, contributing to yield beyond simple HA cleavage (Seitz et al., Appl Microbiol Biotechnol 2011), and protease source and concentration have been optimised as independent variables for H1N1 replication in MDCK (Iskandar et al., J Virol Methods 2017).

CellTrypase is characterised and supplied as a dissociation reagent. c-LEcta publishes no data on HA activation, infection-phase protease performance or virus titre. Do not assume that replacing your dissociation enzyme also replaces your infection-phase protease — those are two qualification exercises, and only the first is supported by data today.

Dosing and cost from T-flask to ten-layer vessel

At the recommended ~1 mL per 25 cm², enzyme volume scales linearly with growth surface. The ready-to-use 1x format is recommended for most applications, with the 10x concentrate available where you want to dilute in-house:

Using BioHippo list prices at the time of writing, and diluting the 10x concentrate 1:10 to make working solution, cost per millilitre of 1x working solution works out at roughly $0.09 (R&D grade from 10x, 500 mL) to $0.29 (GMP grade, ready-to-use 1x, 500 mL). For a single ten-layer vessel that is approximately $23 to $74 of enzyme per harvest. This is simple arithmetic on published list prices, excluding shipping and duties — treat it as an order-of-magnitude planning figure, not a quotation. Two honest qualifications: diluting a 10x concentrate introduces an additional open manipulation and a diluent that you must qualify, which may be unattractive in a GMP suite; and pack size, not unit price, usually drives real cost at scale.

The handling detail that matters most at scale is inactivation. For a ten-layer vessel, dispensing with a trypsin inhibitor removes a soybean-inhibitor or serum-quench addition step, the associated volume, and one more component to qualify. It does not remove the wash step or the need to control contact time. Step-by-step handling, including wash volumes and stop conditions, is set out in the CellTrypase cell dissociation protocol.

What the data does not cover

This is the section a process development scientist should read first. Being specific about gaps is more useful than being vague about strengths. As of PIS v4.0 (valid 18 Aug 2026), c-LEcta (Kerry) reports no performance data for:

• Microcarrier, fixed-bed and packed-bed detachment. Vero and MDCK vaccine processes commonly run on Cytodex-type microcarriers or packed-bed carriers at high cell density (Yu et al., Hum Vaccin Immunother 2012; Sun et al., Appl Microbiol Biotechnol 2012), Vero and HEK 293T are routinely expanded on microcarriers in single-use bioreactors (Yang et al., AMB Express 2019), and Vero processes remain adherent even at production scale, whether on microcarriers or fixed-bed (Kiesslich et al., J Biotechnol 2020). Bead-to-bead transfer and carrier harvest are mechanically and enzymatically different from releasing a flat-flask monolayer. Treat carrier harvest as untested and run it yourself.
• Replating efficiency, proliferation rate or long-term culture performance after dissociation, in any cell type. Viability and yield at the point of harvest are not the same endpoints as recovery over subsequent passages.
• Any cell type other than CHO-K1, HEK 293, MDCK and Vero. Induced pluripotent stem cells, organoids, primary cells, neurons, mesenchymal stromal cells, T cells, cardiomyocytes and RPE are listed by the manufacturer as intended applications, but no viability, yield, recovery or marker-retention figures are published for them. We will not quote numbers that do not exist.
• Use as the exogenous protease in virus propagation medium. There is no data supporting CellTrypase in that role, and no data on virus titre.
• Head-to-head comparison against a named commercial reagent. No such study is on file, so this note makes no equivalence claim against any competing product.
• Long-term stability beyond the confirmed 21 months at 2–8 °C; c-LEcta states this study is ongoing. Brief warm excursions during shipping are reported not to affect activity.

A supplier also cannot resolve your change control. Swapping a raw material in a licensed or clinical-phase process is a regulated change on your side, and the required qualification and revalidation work is determined by your quality system and your filings, not by the supplier's data pack. No supplier can waive that on your behalf.

How to run a bridging test before you switch

Dosing translates one-to-one: the same volume per unit growth area you use today, pre-warmed. That is dosing guidance only — verify detachment time and viability in your own cell model before switching. Where the numbers do not exist, the sensible answer is a bench comparison rather than a claim.

A workable bridging test needs three passages, run side by side against your incumbent reagent, on the specific line and vessel format you actually use:

1. Fix cell density, passage number and confluence, and pre-warm both reagents identically.
2. Record time to complete release, viable cell yield and viability at harvest — and repeat for three consecutive passages rather than one. Attachment and growth over the following passages is the endpoint that actually matters, and the one with no vendor data behind it for any cell type.
3. Carry the harvested cells through your own critical quality attributes: virus titre, infectious-to-total particle ratio, vector genome titre, or whatever your process is actually judged on. Detachment performance that does not survive this step is not useful.

The R&D grade is available as a free 100 mL sample for exactly this purpose, so the trial costs you time rather than budget. For GMP-grade quantities, documentation packages or bulk pricing, request a quote. Related reading: the CellTrypase specification comparison guide and the wider cell line and cell culture catalogue.

References

1. Barrett PN, et al. Vero cell technology for rapid development of inactivated whole virus vaccines. Expert Rev Vaccines 2009. doi:10.1586/erv.09.19
2. Genzel Y, et al. MDCK and Vero cells for influenza virus vaccine production: a one-to-one comparison up to lab-scale bioreactor cultivation. Appl Microbiol Biotechnol 2010. doi:10.1007/s00253-010-2742-9
3. Ma H, et al. Investigations of porcine circovirus type 1 (PCV1) in vaccine-related and other cell lines. Vaccine 2011;29(46):8429–37. doi:10.1016/j.vaccine.2011.07.123
4. Dubin G, et al. Investigation of a regulatory agency enquiry into potential porcine circovirus type 1 contamination of the human rotavirus vaccine, Rotarix. Hum Vaccin Immunother 2013;9(11):2398–408. doi:10.4161/hv.25973
5. Gilliland SM, et al. Investigation of porcine circovirus contamination in human vaccines. Biologicals 2012;40(4):270–7. doi:10.1016/j.biologicals.2012.02.002
6. Dore DD, et al. Vaccine discontinuation and switching following regulatory interventions in response to rotavirus vaccine contamination with porcine circovirus DNA fragments. Pharmacoepidemiol Drug Saf 2012;21(4):415–9. doi:10.1002/pds.3217
7. Barone PW, et al. Viral contamination in biologic manufacture and implications for emerging therapies. Nat Biotechnol 2020;38(5):563–572. doi:10.1038/s41587-020-0507-2
8. Inai T. Claudin-based tight junctions in MDCK cells. Methods Mol Biol 2011;762:127–39. doi:10.1007/978-1-61779-185-7_8
9. Seitz C, et al. Trypsin promotes efficient influenza vaccine production in MDCK cells by interfering with the antiviral host response. Appl Microbiol Biotechnol 2011. doi:10.1007/s00253-011-3569-8
10. Iskandar A, et al. Optimization of trypsins for influenza A/H1N1 virus replication in MDCK SI-6 cells. J Virol Methods 2017. doi:10.1016/j.jviromet.2017.11.006
11. Yu M, et al. Production of inactivated influenza vaccine using Vero cells on microcarriers. Hum Vaccin Immunother 2012. doi:10.4161/hv.20985
12. Sun B, et al. High-density culture of Vero cells on microcarriers in a packed-bed bioreactor. Appl Microbiol Biotechnol 2012. doi:10.1007/s00253-012-4375-7
13. Yang Z, et al. Large-scale microcarrier culture of HEK293T cells and Vero cells in single-use bioreactors. AMB Express 2019. doi:10.1186/s13568-019-0794-5
14. Kiesslich S, et al. Serum-free production of rVSV-ZEBOV in Vero cells: microcarrier bioreactor versus scale-X hydro fixed-bed. J Biotechnol 2020. doi:10.1016/j.jbiotec.2020.01.015
15. c-LEcta GmbH (Kerry Biotechnology Centre). Product Information Sheet, CellTrypase v4.0, valid 18 Aug 2026.

Figures: Figures 1–4 are illustrative summaries of manufacturer specification data, published list prices and the qualification steps described in this article. They are visualisations prepared by BioHippo, not experimental data generated by BioHippo.
Data source: Performance figures are taken from c-LEcta GmbH (Kerry Biotechnology Centre), PIS CellTrypase v4.0, valid 18 Aug 2026. Literature identified via PubMed; citations link to the primary publications.
Research use statement: CellTrypase R&D grade (cat. 22103-1X-100, 100 mL, and related formats) is supplied For Research Use Only (RUO) and is not for use in diagnostic or therapeutic procedures. GMP grade is supplied as a processing aid and manufacturing raw material under EXCiPACT® GMP/GDP certification, and is not for direct administration to humans or animals.
Trademark notice: TrypLE™ is a trademark of Thermo Fisher Scientific Inc. Any reference is nominative and for identification or comparison only; no affiliation with, sponsorship by, or endorsement from the trademark owner is implied.
Pricing: Prices cited are BioHippo list prices at the time of writing and are subject to change.