Stable cell line generation
Killing everything that did not integrate the construct by using a drug-resistance gene carried on it, after establishing the right concentration with a kill curve on the parental line and waiting long enough for the marker to be expressed.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
Antibiotic selection
Killing everything that did not integrate the construct by using a drug-resistance gene carried on it, after establishing the right concentration with a kill curve on the parental line and waiting long enough for the marker to be expressed.
How it differs from the standard method
- Common markers: puromycin (1–10 µg/mL, kills in 2–4 days), blasticidin (2–10 µg/mL, fast), G418/neomycin (200–1000 µg/mL, slow, 1–2 weeks), hygromycin (50–500 µg/mL).
- Start selection 24–48 h after delivery and change the selection medium every 2–3 days.
- Always run an untransfected dish in the same drug: it is the control that makes the surviving colonies mean anything.
- Resistance proves integration of the marker, not expression of the gene of interest: screen every clone.
- A lentivirally transduced pool selects to a polyclonal population rather than to discrete colonies.
1From a transient population to a line that keeps the construct
A transfected plasmid is diluted away within a week, because it does not replicate with the chromosomes. A stable line is made from the rare cells in which the construct became part of the genome: integrated at a random site by the cell's own end-joining machinery after a transfection, or inserted by a retroviral vector that does this for a living.
Set the dose on your own line
The lowest concentration that kills every parental cell in 3–5 days; sensitivity differs several-fold between lines.
Wait before selecting
The resistance gene has to be expressed first, or the drug kills the cells that were about to become your clones.
Clone well below one
At 0.5 cells per well most wells stay empty and about 77% of the wells that grow came from a single cell.
2Why clones differ, and why that matters
An integration lands wherever the break was. The site decides how strongly the cassette is transcribed, whether it is gradually silenced over passages, and what it disrupts on the way in; plasmid integrations usually go in as tandem arrays of several copies, so copy number varies too.
3How it goes wrong
- Taking the antibiotic concentration from a paper rather than from a kill curve on your own line and passage.
- Starting selection a few hours after the transfection, before the marker is expressed.
- Never changing the selection medium, so dying cells shelter drug-tolerant parental cells.
- Picking a colony from a dish where the colonies have grown into each other: two clones in one tube.
- Treating drug resistance as proof of expression: the marker can be intact while the gene of interest is silenced or truncated.
- Calling every well that grows a clone, with no record of what was in it on day one.
- Characterising, and then publishing, a single clone.
- Using an inducible line in ordinary serum, which contains enough tetracycline to lift the off state.
4Routes to a clone
- Limiting dilution
- Serial dilution to well under one cell per well in 96-well plates; cheap, needs no instrument, and the clonality is a probability unless the wells are imaged on day one.
- Single-cell sorting
- A cytometer deposits one cell per well, and index sorting records what was deposited; faster, allows sorting on a marker, and harder on fragile cells.
- Cloning cylinders / colony picking
- Picking separated colonies from a selection dish with cylinders or a picker; only honest when the colonies are well separated.
- Semi-solid medium
- Growing clones in methylcellulose so each colony stays discrete, often with a fluorescent secretion assay for picking high producers.
Key terms
Practise these as flashcards →A titration of the selection drug on untransfected cells that identifies the lowest dose killing all of them in a few days.
A resistance gene (puromycin, neomycin/G418, hygromycin, blasticidin) carried on the construct, or a fluorescent protein used for sorting instead.
The mixed population of independent integrants left after selection, before any cloning.
The fraction of single cells that grow into a colony; often 10–50% and strongly dependent on medium and cell type.
Evidence, not an assumption, that a line descends from one cell: a day-1 image, index sorting records, or two rounds of cloning.
The influence of the integration site on how strongly and how stably the cassette is expressed.
The reverse tet transactivator and the tetracycline response element it binds in the presence of doxycycline: the Tet-On pair.
Expression from an inducible promoter with no inducer present; the denominator of the fold induction and the usual reason a clone is discarded.
When to use it
- When an experiment runs longer than a transient transfection lasts, or has to be repeated identically for months.
- To make a reporter line, a tagged knock-in or a line expressing a shRNA or a CRISPR component continuously.
- When every cell has to carry the construct, for example before a selection-free phenotypic assay or a co-culture.
- To produce protein or virus reproducibly from a defined, banked line rather than a fresh transfection each time.
- When the construct is toxic, so that expression has to be switched on at a chosen time with an inducible system.
Compared with alternatives
A transient transfection answers the question in two days when a few days of expression is all the experiment needs; a stable line is weeks of work before the first result.
Low-MOI transduction followed by selection is the usual way a stable pool is made, and the Poisson arithmetic that gives single-copy integrants is the same arithmetic as limiting dilution.
Editing a locus needs the same clone-isolation and genotyping workflow; the difference is that the clones are screened by sequence rather than by expression.
A clone is only worth making if it is banked at low passage: freeze a master bank as soon as the line is validated.
Single-cell sorting is the alternative to limiting dilution, and flow is also how expression and its homogeneity are checked in each clone.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
Killing everything that did not integrate the construct by using a drug-resistance gene carried on it, after establishing the right concentration with a kill curve on the parental line and waiting long enough for the marker to be expressed.
- Common markers: puromycin (1–10 µg/mL, kills in 2–4 days), blasticidin (2–10 µg/mL, fast), G418/neomycin (200–1000 µg/mL, slow, 1–2 weeks), hygromycin (50–500 µg/mL).
- Start selection 24–48 h after delivery and change the selection medium every 2–3 days.
- Always run an untransfected dish in the same drug: it is the control that makes the surviving colonies mean anything.
- Resistance proves integration of the marker, not expression of the gene of interest: screen every clone.
- A lentivirally transduced pool selects to a polyclonal population rather than to discrete colonies.