A cell-line label may name the wrong cells
Some cultured cell lines used in research are not the tissue or donor named on the vial. Faster-growing cells, often introduced by cross-contamination, can replace the intended culture while its label remains unchanged. Experiments may then be reproducible observations of the wrong biological model.
Continuous cell lines are shared, thawed, passaged and sometimes handled beside other cultures. A small contamination can take over if the intruder grows faster. DNA profiling has shown that many named lines are duplicates of another line or come from a different species. HeLa has been a frequent contaminant, but it is not the only one. Authentication by short tandem repeat profiling, species testing and provenance records can reveal the mismatch. Current registers continue to distinguish lines with no authentic stock from cases where a genuine stock still exists.
Replication cannot rescue a mistaken identity when every laboratory inherits the same mislabelled material. The episode separates procedural consistency from construct validity: an assay can be technically impeccable yet answer a question about different cells. Names in a supply chain require periodic measurement, not ceremonial trust.
Once a misidentified line becomes common, papers, repositories and collaborators can reinforce one another’s label. Similar results then look like independent confirmation even though the same contaminant is the shared ancestor. Citation networks preserve the semantic identity while cell division preserves the biological one. Correcting the record requires tracing both lineages. That boundary keeps the example informative without asking it to support a broader claim than the evidence can bear.
A DNA match can establish donor identity, but it does not prove that a heavily passaged line still models its original tissue well. Cells evolve in culture, acquire chromosomal changes and respond to media or handling. Good practice therefore combines identity testing with contamination checks, passage records and fit-for-purpose phenotyping. An authentic name is necessary evidence, not a warranty of relevance.
Are registries exaggerating old mistakes that modern repositories already fixed?
Repositories have improved authentication, but cultures also circulate outside repositories and can be contaminated after purchase. Living registers add new evidence and distinguish authentic stocks where possible. The durable claim is not a fixed percentage of all research; it is that label–identity mismatches are real, consequential and detectable enough to require routine controls. This distinction prevents a genuine caveat from being mistaken for evidence of the opposite conclusion.
A broad, independent authentication programme finding no mismatches in shared lines over time would show that the problem had become historical rather than current.