Passage Number and Population Doublings — Why P7 Means Nothing on Its Own

Passage number is a count of how many times a culture has been detached from its vessel and re-seeded into fresh vessels. Population doubling level (PDL) is a count of how many times the population has doubled in size. They are different quantities, they are not convertible without information the supplier usually withholds, and only one of them is comparable between two laboratories.

The confusion is not academic. Two vials labelled “P7” can differ by a factor of ten in remaining expansion capacity, and the buyer has no way to see it from the label.

What a passage actually is

A passage is a handling event. You dissociate the monolayer, count or do not count, dilute into a stated number of new vessels, and increment a number. Nothing about that procedure measures biological age. A laboratory that splits 1:2 every time and a laboratory that splits 1:10 every time will both call the fifth event “passage 5”, but the second culture has been through roughly 3.3 times as many divisions.

This is why the ICH Q5D guideline on cell substrates does not treat passage number as self-sufficient. It defines in vitro cell age as the measure of time between thaw of a master cell bank vial and harvest, and states that it may be measured by elapsed chronological time, by population doubling level, or by passage level of the cells when subcultivated by a defined procedure for dilution of the culture. The qualifying clause carries the whole meaning: passage level is a valid measure only when the dilution procedure is fixed and documented. Absent that, the number is a housekeeping tally.

The arithmetic

Population doublings for a single passage:

PD = log₂(harvested cell number ÷ seeded cell number)

Or, if you only know the split ratio and assume the culture was grown to the same confluence each time:

PD ≈ log₂(split ratio)

Cumulative PDL is the running sum across every passage since a defined zero point.

Split ratio	Population doublings per passage	PDL after 10 passages
1:2	1.0	10
1:3	~1.6	~15.8
1:4	2.0	20
1:5	~2.3	~23.2
1:6	~2.6	~25.8
1:8	3.0	30
1:10	~3.3	~33.2
1:20	~4.3	~43.2

Read the last column again. “P10” spans an eight-fold range of biological age depending on a number nobody printed on the vial. If the vendor’s characterisation was done at their P10 with a 1:3 habit and you have reached your P10 with a 1:10 habit, you are working with a materially older culture than the one the datasheet describes.

The split-ratio approximation assumes the culture is grown to the same density each passage and that plating efficiency is 100 percent. Neither is true. Cells lost at seeding inflate the true doubling count relative to the ratio estimate, because the population that survived had to divide more times to reach the same harvest number. This is why counting cells at seed and at harvest — and computing PD from the counts rather than the ratio — is the only method that produces a number worth transferring.

Why passage number is not comparable between suppliers

There is no convention. In our reading of commercial datasheets, at least five distinct conventions are in circulation:

Convention	Where the count starts	Consequence for the buyer
From primary isolation	The tissue digest is P0; first subculture is P1	The most informative convention, and the least common for derived products
From the master cell bank	The MCB vial is P0 regardless of its own history	Hides all upstream expansion. Common in industrial supply
From the working cell bank	Reset again at the WCB	Hides more still. Two resets between donor and vial
From differentiation start	The pluripotent starting material’s passage history is discarded	Standard for iPSC-derived terminal products, and defensible — a post-mitotic neuron has no passage number in any useful sense
From thaw of the shipped vial	Your P1 is the first split you perform	What most buyers actually track, and what the vendor’s number is not

None of these is wrong. The problem is that the same integer is printed under all five, with no label distinguishing them, and buyers compare across them as though the number were a measurement.

The practical rule: a passage number is only interpretable alongside the zero point and the split procedure. If a datasheet gives you a number without either, you have been given a serial number, not a specification.

PDL, senescence, and what the threshold actually looks like

The finite replicative lifespan of normal human diploid cells was established by Hayflick and Moorhead in 1961, working with human fibroblast strains. The observation was that normal diploid strains do not divide indefinitely in culture but enter a terminal non-dividing state after a characteristic number of doublings. That characteristic number is a property of the cell strain, the donor and the culture conditions — not a universal constant, and not a number a supplier can quote at you generically.

The more recent multiomics re-examination of the same WI-38 fibroblast system found something that matters more for procurement than any threshold value: the transition to senescence manifests early and increases gradually, rather than switching on at a boundary. The authors also found that senescent WI-38 cells acquire a striking resemblance to myofibroblasts, through a process similar to epithelial-to-mesenchymal transition.

Two consequences follow directly.

First, “below the senescence limit” is not a quality claim. A culture at 60 percent of its replicative lifespan is not a young culture with a small chance of being old. It is a population containing a measurable fraction of cells that have already drifted, with the fraction rising continuously. Any assay sensitive to that drift will show a passage-dependent trend long before the culture stops dividing.

Second, phenotype changes before proliferation stops. The myofibroblast-like shift is a change in what the cells are, not merely in how many times they can still divide. For anyone buying primary cells for a functional readout — contractility, secretion, barrier formation, drug response — the usable passage window closes earlier than the proliferative window, and the vendor’s stated maximum passage is usually the proliferative one.

The pluripotent case is different, and worse

Immortal and pluripotent lines do not senesce on the fibroblast pattern. Human pluripotent stem cells will expand for very large numbers of doublings. The risk changes from exhaustion to selection.

Recurrent genomic aberrations accumulate in human pluripotent stem cell culture, and the best-documented is gain of chromosome 20q11.21. The mechanism is a straightforward selective sweep: overexpression of the antiapoptotic gene BCL2L1 (Bcl-xL) from the amplified region confers a survival advantage, so the abnormal cells progressively out-compete the normal ones at every passage. The consequence is not neutral. The published analysis found that the altered signalling has a dramatic negative effect on neuroectodermal differentiation, while the cells retain the ability to differentiate to mesendoderm derivatives.

For a buyer, this is the passage-number risk that actually loses projects. Your cells look healthy, grow faster than they used to, pass every viability check, and have quietly lost the ability to make the lineage you bought them for. The observation of “faster growth after a few more passages” should be treated as a warning rather than a result.

Organoid systems occupy a third position again: derived from adult stem cell compartments, they can be propagated far beyond the fibroblast limit, which is the sense in which the literature describes them as defying the Hayflick limit. That does not mean they are stable — it means the constraint is drift and selection rather than exhaustion.

Cell class	Dominant passage risk	What actually fails first	What to monitor
Primary human cells (fibroblasts, myoblasts, endothelial)	Replicative senescence, gradual onset	Function and phenotype, before proliferation	PDL against the supplier’s stated capacity; doubling time trend; function assay at each bank
Human pluripotent stem cells	Selective sweep of culture-adapted variants, notably 20q11.21 gain	Differentiation capacity, lineage-selectively	Karyotype or CNV screen at defined intervals; differentiation efficiency; unexplained growth acceleration
iPSC-derived post-mitotic products (neurons, cardiomyocytes)	Not applicable after differentiation — they do not divide	The passage history of the parent line, which you cannot see	The parent line’s passage at differentiation, and whether it was karyotyped there
Immortalised lines	Clonal drift and, at the extreme, misidentification	Comparability with published data on “the same” line	STR authentication; a frozen low-passage reference stock of your own
Adult-stem-cell organoids	Drift and subclonal selection over long propagation	Regional identity and marker fidelity	Marker panel at intervals; early archiving

What to ask a supplier, and what a good answer looks like

Question	Why it matters	A good answer
At what passage was this lot’s characterisation performed?	The datasheet describes a specific culture age. If your vial is later, the datasheet does not describe your vial	A specific number, with the assay date
What is your zero point?	Determines whether the number means anything	“P0 is the primary isolate” or “P0 is the MCB, which was itself derived at donor-tissue P3”
What split ratio, or what seeded and harvested counts?	The only route from passage number to PDL	Actual counts. A ratio is second-best
What is the cumulative PDL of this vial?	The transferable unit	A number. If they cannot produce one, they are not tracking it
What is the maximum PDL or passage at which you have data that this product still performs?	Distinguishes proliferative capacity from functional capacity	A number with the assay that defines “still performs”
For pluripotent lines: at what passage was the karyotype or CNV screen done, and how often is it repeated?	20q11.21-class aberrations arise during culture, so a screen at the MCB says nothing about a vial ten passages later	A defined re-testing interval, not a single historical result

Practical procurement rules

Bank early and bank deep. The cheapest insurance against passage drift is a large low-passage working bank made immediately on receipt, before you have done anything to the cells. This is a one-off cost in labour and vials that removes the entire class of problem for the life of the programme. It is also what the supplier did, which is why they can supply consistent lots and you cannot.

Fix your split ratio and record counts. A programme with a fixed subculture procedure and recorded seed/harvest counts can convert to PDL retrospectively. One without it cannot, ever, and its passage numbers are not transferable to the next person who works on the project.

Record PDL on your own vials, alongside the vendor’s passage number. Two numbers. The vendor’s for traceability back to their record, yours for biology.

Treat a passage jump as a protocol change. If a comparison spans a wide passage range, that is a variable in the experiment. Bridge it deliberately or accept that a difference you observe may be culture age.

Do not accept “low passage” as a specification. It is not one. Ask for the integer, the zero point and the split procedure — and if the answer is a number with no context, you have learnt something about the supplier’s quality system rather than about the cells.

The one-line summary

Passage number tells you how many times someone opened the flask. Population doubling level tells you how old the cells are. Ask for both, record both, and treat any datasheet that gives you a P-number without a zero point and a split ratio as giving you a filename rather than a measurement.

Passage number vs population doublings: what a P-number actually tells you