Subculture of Cells (Passaging)

A culture vessel supplies a fixed growth area and a fixed volume of medium. As cells divide they consume glucose and glutamine, exhaust the buffering capacity of the medium, and accumulate lactate and ammonia. At the same time an adherent monolayer runs out of free surface. Once the surface is fully covered the culture is confluent, and from that point on the population is no longer in balanced exponential growth.

Subculturing resets both constraints at once. Cells are lifted (adherent) or diluted (suspension) and distributed into new vessels at a lower density with fresh medium, returning the population to the logarithmic phase of the growth curve where doubling time is constant and physiology is reproducible.

Three separate things go wrong if a culture is left past confluence:

• Growth arrest. Contact-inhibited lines such as NIH/3T3 stop dividing when the monolayer closes and are difficult to restart cleanly.
• Phenotypic drift. Transformed lines that are not contact-inhibited pile up in multilayers, detach in sheets, and can be selected for variants that tolerate crowding.
• Metabolic stress. Medium turns yellow as lactate accumulates and the bicarbonate buffer is overwhelmed. The pH shift alone reduces viability and plating efficiency at the next passage, even after the cells look fine again in fresh medium.

The subculture step is also the point at which most routine culture work happens: counting, viability checks, cryopreservation, mycoplasma sampling and setting up experimental plates all normally hang off a passage.

Confluence is the fraction of the growth surface covered by cells, assessed by eye on an inverted phase-contrast microscope at 4x or 10x. It is a visual estimate, not a measurement, so it needs a consistent operator convention.

The standard target is 70-80% confluence for continuous adherent lines. At that density cells are still in log phase, viability is high, and they recover quickly after splitting. Passaging much earlier wastes cells and can extend the lag phase; passaging late costs viability and reproducibility.

Practical cues that a flask is ready:

1. Coverage. At 70-80% the monolayer has closed over most of the surface but clear gaps between colonies are still visible. At 100% there are no gaps at all and cell borders form a continuous mosaic.
2. Cell shape. Well-spread, well-attached, refractile cells with clear borders. Cells that have started to round up, granulate or vacuolate are already stressed.
3. Medium colour. With phenol red, a healthy log-phase culture sits at orange-red. A shift to yellow-orange between feeds usually means the culture is dense enough to split. A shift to pink-purple in a sparse flask often means poor growth or a leaking incubator CO2 supply, not readiness.
4. Floating debris. A rising population of round, phase-bright floaters over the monolayer indicates the culture is past its best.

Some lines have their own rules. NIH/3T3 must be subcultured before it reaches confluence — allowing it to become confluent selects for transformed, focus-forming variants and permanently changes the line. Finite human diploid fibroblast strains such as MRC-5 are usually taken closer to 80-90%. Primary cultures and stem cells often have narrower windows still, set by the differentiation protocol rather than by confluence alone.

For suspension cultures, confluence does not apply. The equivalent trigger is cell concentration: count the culture at least twice a week and dilute back before the density reaches the top of the line's tolerated range.

Volumes below are for a T-75 flask (75 cm2). Scale roughly with growth area: a T-25 needs about one third, a T-175 about 2.3x. Warm medium to 37 degrees C before use; keep trypsin-EDTA at room temperature or warm it briefly, and do not leave it at 37 degrees C for long periods because trypsin autolyses.

1. Check the culture. Confirm 70-80% confluence and that the medium is clear, with no turbidity or unexpected pH shift. Note the passage number on the flask.
2. Remove the medium. Aspirate and discard all spent medium. Residual serum is the main reason trypsinisation fails, because serum contains alpha-1-antitrypsin and other protease inhibitors.
3. Wash. Add 5-10 mL of Ca2+/Mg2+-free DPBS or PBS, rock gently across the monolayer, and aspirate. Calcium and magnesium are required by the cadherin and integrin adhesions you are trying to break, so the wash buffer must be free of both. Wash twice if the line is grown in high serum.
4. Rinse with trypsin (optional but useful). Add 1-2 mL of trypsin-EDTA, rock to cover, and immediately aspirate. This removes the last of the inhibitor film.
5. Add trypsin-EDTA. Add 2-3 mL of 0.25% trypsin-0.53 mM EDTA (or 0.05% trypsin-EDTA for delicate or loosely attached lines) and rock so the whole surface is covered.
6. Incubate. Return the flask to 37 degrees C and check under the microscope after 2 minutes, then every minute. Most lines detach in 2-10 minutes; some tightly adherent lines take up to 15. Do not simply set a timer and walk away — over-trypsinisation is the single most common cause of poor recovery.
7. Release the sheet. When cells are rounded and the sheet begins to slide, rap the side of the flask sharply against the heel of your hand to dislodge it. Under the microscope you should see a suspension of single rounded cells, not sheets or strings.
8. Neutralise. Add 6-8 mL of complete growth medium containing serum and pipette gently over the growth surface to wash off residual cells and disperse clumps. For serum-free systems, use a defined trypsin inhibitor such as soybean trypsin inhibitor instead — serum-free medium alone will not stop the reaction.
9. Count if needed. Take an aliquot for a hemocytometer count with trypan blue. Counting at every passage is good practice and is required whenever a defined seeding density is specified.
10. Optional centrifugation. For lines sensitive to residual trypsin, or when the medium is being changed to a different formulation, spin at 100-200 x g for 5 minutes, discard the supernatant and resuspend in fresh medium. Skip this step for robust lines: the pellet-and-resuspend cycle costs viability.
11. Reseed. Distribute the suspension into new flasks at the chosen split ratio or seeding density, top up to the working volume (15-20 mL for a T-75) and label with the line, the new passage number and the date.
12. Return to the incubator. Rock the flask front-to-back and side-to-side — never swirl in a circle, which concentrates cells in the middle — and place it flat at 37 degrees C, 5% CO2.

Cells should be attached and beginning to spread within 4-24 hours depending on the line.

Suspension lines need no dissociation, so passaging is a dilution.

1. Count the culture and check viability. This is the whole basis of the decision, so it cannot be skipped.
2. Resuspend gently by swirling or by inverting the vessel, so the aliquot you take is representative.
3. Dilute back into fresh pre-warmed medium to the line's recommended seeding concentration — commonly 1 x 10^5 to 3 x 10^5 viable cells/mL for lymphoblastoid lines such as Jurkat.
4. Alternatively, partial medium exchange. For dense cultures, let the cells settle or spin at 100-200 x g for 5 minutes, remove a set fraction of the conditioned medium, and replace it with fresh. Some lines grow better with a proportion of conditioned medium retained.
5. Return to the incubator. Shaken or spinner cultures go back on the platform at the validated agitation rate; static suspension cultures go back flat.

The two failure modes are opposite and equally damaging. Letting the density climb too high depletes the medium and drops viability sharply — Jurkat cultures, for example, should not be allowed past 3 x 10^6 cells/mL. Diluting too far pushes the culture into an extended lag phase from which some lines never fully recover; if a line is being revived or is growing poorly, split it more conservatively.

A split ratio expresses how far the cells are diluted: a 1:4 split puts the contents of one confluent flask into four new flasks of the same size. It is quick, needs no count, and is fine for well-characterised robust lines on a fixed schedule.

A seeding density, in viable cells/cm2 or cells/mL, is the more precise instruction. It requires a count but is reproducible across operators, vessel types and confluence estimates, and it is what experimental protocols specify.

The two are linked by the confluent density of the line. If a line reaches confluence at roughly 1 x 10^5 cells/cm2 and you split 1:8, you are seeding about 1.25 x 10^4 cells/cm2.

Choosing a ratio comes down to doubling time and the interval you want:

• A line with a 24-hour doubling time gains about 3 doublings (8x) in 72 hours, so a 1:8 split on Friday returns it to confluence around Monday.
• The same line split 1:4 will be ready in roughly 48 hours.
• Slow-growing lines and finite strains tolerate only shallow splits — typically 1:2 to 1:4 — because a deep split leaves them below the density at which they grow well, and because every passage of a finite strain spends part of a fixed lifespan.

Two practical warnings. First, split ratios are not additive across vessel formats: splitting a T-75 "1:4" into four T-25s is really a 1:1.3 split by area. Always reason in growth area or in cells/cm2. Second, when a line is being recovered from liquid nitrogen, reduce the split depth for the first two or three passages; thawed cells have lower plating efficiency.

Passage number counts subculture events. It increments by one each time the culture is split, regardless of how deep the split was. It is a bookkeeping convention, not a measure of biological age, and its meaning depends entirely on the split ratios used along the way.

Population doubling level (PDL) is the biologically meaningful figure: the cumulative number of times the population has doubled. For each passage:

PD = 3.32 x [ log10(cells harvested) - log10(cells seeded) ]

and PDL is the running total. A 1:2 split adds one doubling; a 1:8 split adds three. Two labs both at "passage 20" can differ by more than 40 doublings.

This matters because cell behaviour changes with time in culture:

• Finite cell strains have a fixed replicative lifespan (the Hayflick limit) and then enter irreversible senescence. ATCC states that MRC-5 is capable of 42 to 46 population doublings before the onset of senescence. As the strain approaches that limit, doubling time lengthens, cells enlarge and flatten, and saturation density falls.
• Continuous cell lines are immortal but not stable. Karyotype, receptor expression, drug sensitivity, transfection efficiency and growth rate all drift over extended passaging. Published discrepancies between labs frequently trace back to high-passage stocks.

The standard control is a two-tier bank: a master bank frozen at low passage from the original vial, and working banks expanded from it. Routine culture runs from a working vial for a defined number of passages — commonly 15-20, or 10 for sensitive assays — and is then discarded and a fresh vial thawed. Record the passage number on every flask and in the culture log, and carry it forward into every experiment record.

The subculture logic is the same for both, but almost every practical detail differs.

Adherent cultures are limited by surface area, need a dissociation step, are counted only at passage, and are scaled by adding vessels or moving to roller bottles and multi-layer flasks. The dissociation step is where damage occurs: trypsin cleaves surface proteins including receptors, and prolonged exposure reduces viability and delays reattachment. Choosing the mildest reagent that works — 0.05% trypsin-EDTA, EDTA/Versene alone, or a recombinant or enzyme-free dissociation reagent — is usually worth the extra minute of incubation.

Suspension cultures are limited by medium volume and gas exchange, need no dissociation, must be counted routinely rather than only at passage, and scale straightforwardly by volume into shake flasks, spinners and bioreactors. Their failure modes are aggregation, shear damage at high agitation, and the density crash that follows an over-run culture.

Some lines exist in both forms. HEK293 and CHO are the common examples: the parental adherent line and a suspension-adapted derivative are not interchangeable, and adaptation between them takes weeks of stepwise transfer, not a single passage.

Cells will not detach. Residual serum is the usual cause — wash again with Ca2+/Mg2+-free buffer. Other causes are trypsin that has lost activity through repeated warming or freeze-thaw, an over-confluent monolayer with extensive cell-cell junctions, or a line that simply needs a stronger reagent (0.25% rather than 0.05%, or a brief EDTA pre-incubation).

Cells detach but will not reattach. Over-trypsinisation. Symptoms are a long lag, low attachment efficiency and a rising floating fraction. Shorten the incubation, drop to 0.05% trypsin, neutralise immediately and completely, and consider a non-enzymatic reagent.

Clumping after dissociation. Usually released DNA from lysed cells. Add a low concentration of DNase to the neutralising medium, or reduce the trypsinisation time. Clumps ruin cell counts as well as seeding uniformity.

Uneven monolayer, cells piled in the centre or at the edges. The flask was swirled rather than rocked, or it was moved before the cells settled. Rock front-to-back then side-to-side and set the flask down undisturbed.

Growth slows over successive passages. Check the PDL first — a finite strain may simply be approaching senescence. Otherwise suspect mycoplasma (which is invisible by eye and is the single most common cause of unexplained slow growth), a bad serum lot, incubator CO2 or humidity drift, or accumulated over-trypsinisation damage.

Culture reaches confluence far too fast or too slowly. The split ratio no longer matches the doubling time, which itself changes as a line drifts. Recount, recalculate the seeding density and adjust rather than keeping the historical ratio.

Contamination appears at a specific passage. Trace back to the shared reagent bottles opened at that session. Aliquot media and supplements into single-use volumes so a contamination event cannot propagate across a whole bank.