The contamination that ruins a Friday is not the one that ruins a career. A flask that turns cloudy and yellow overnight is annoying, expensive in lost time, and unmistakable. It gets discarded, the incubator gets cleaned, and the experiment restarts.
The contaminations that do lasting damage are the ones that never change the appearance of the culture. Mycoplasma grows freely in cell culture without turning the medium cloudy and without being visible under a routine microscope, while altering the behaviour of the cells it infects. A cell line that was quietly overgrown by a more vigorous line years ago looks entirely normal and grows beautifully. Both produce clean-looking experiments and results that nobody can reproduce.
Ranking contamination by how long it can hide is therefore a more useful framework than ranking it by how alarming it looks. This piece works through that ranking, from the problems that declare themselves immediately to the ones that have quietly compromised a substantial body of published literature.
Key takeaways
- Visible bacterial and fungal contamination is disruptive but self-announcing, and rarely affects results because affected cultures are discarded.
- Mycoplasma is undetectable by eye or routine microscopy yet alters growth, metabolism and gene expression, making it the most damaging common contaminant.
- Cross-contamination between cell lines is widespread historically and can only be excluded by genetic authentication.
- Chemical contamination from media components, plasticware and water systems produces effects easily mistaken for biology.
- Routine scheduled testing catches problems earlier and far more cheaply than investigating an inexplicable result months later.
Bacterial and Fungal Contamination Signs
Ordinary microbial contamination is fast, obvious and, from a scientific standpoint, the least dangerous kind, because it is caught before it can influence a result.
Bacteria in culture medium multiply rapidly under the conditions that suit mammalian cells, and the first sign is usually turbidity: medium that was clear becomes cloudy, sometimes within hours. Metabolic acid production overwhelms the buffering system, so the phenol red indicator in the medium shifts from red toward yellow. Under an inverted microscope, bacteria appear between the cells as small particles in constant motion, distinguishable from debris because they move under their own power rather than drifting.
Yeast contamination produces a slightly different picture. Individual yeast cells are considerably larger than bacteria, appearing as ovoid bodies that bud visibly. The medium becomes cloudy more slowly and often develops a distinct appearance rather than a smooth turbidity. Filamentous fungi are more distinctive still, growing as branching threads that may become visible to the naked eye as floating tufts, often after a delay of several days.
The practical response is discard rather than rescue. Antibiotic rescue of a contaminated culture is possible but almost always a mistake, since it selects for resistant organisms, may leave a low-level infection that persists, and consumes time better spent thawing a fresh vial. Where the contaminated culture is genuinely irreplaceable, treatment is attempted in strict isolation from every other culture in the laboratory.
The most useful thing about visible contamination is what it indicates about technique. A single contaminated flask is bad luck. A recurring pattern points at a specific cause: a failing filter in the cabinet, a contaminated bottle of medium or supplement, a water bath used to warm media, condensation in an incubator, or a habit at the bench such as passing a hand over an open vessel.
Mycoplasma as the Invisible Problem
Mycoplasmas are the smallest self-replicating organisms known, and everything problematic about them follows from their size and their lack of a cell wall.
They are small enough to pass through the filters that sterilise media. They have no cell wall, which makes them intrinsically resistant to the entire class of antibiotics that attack wall synthesis, including the penicillin routinely added to culture medium. Their size places them below the resolution at which they would be noticed under a standard inverted phase contrast microscope during a routine check, and they do not produce enough turbidity to cloud the medium even at very high densities.
They do, however, change the cells. Mycoplasma competes for nutrients, particularly certain amino acids, alters the metabolism of the host cell, and changes gene expression, growth rate, and responsiveness to stimuli. The alterations are not subtle at the level of data even though they are invisible at the level of appearance, and they vary between infections, which means an infected culture produces results that are internally consistent but not reproducible elsewhere.
Detection requires a deliberate test. Several approaches exist and they differ in sensitivity, cost and turnaround.
| Method | Basis | Turnaround | Practical notes |
|---|---|---|---|
| PCR-based kits | Amplifies conserved mycoplasma sequences | Same day | Fast, sensitive, most common routine choice |
| Fluorescent DNA staining | Stains extranuclear DNA in indicator cells | One to two days | Visual, requires a microscope and experience |
| Direct culture | Growth on specialised broth and agar | Several weeks | Reference standard, slow, misses some strains |
| Enzymatic assays | Detects mycoplasma-specific enzyme activity | Under an hour | Convenient screening, less sensitive |
The route into a laboratory is usually another culture rather than the environment, which is why mycoplasma is often described as spreading between laboratories along with shared cell lines. Serum and other biological supplements are a secondary route. Because a positive result in one flask usually implies broader exposure, laboratories treat a detection as a reason to test everything in the incubator, not just the affected line.
Prevention rests on quarantining incoming cultures until tested, avoiding shared media bottles between lines, and testing at intervals rather than only when something seems wrong. Many journals and funders now require a mycoplasma statement, which has improved practice considerably.
Cross-Contamination Between Cell Lines
The most consequential contamination in cell culture is not microbial at all. It happens when cells from one line end up in a flask labelled as another and outgrow the original.
The mechanism is mundane: a shared pipette, an aerosol generated by vigorous pipetting, a mislabelled vial, two flasks open in the cabinet simultaneously, or a stock that was already contaminated when it was received. Once a small number of a fast-growing line enters a slower-growing culture, ordinary passaging selects relentlessly for the faster grower, and within a few passages the original may be gone entirely.
Nothing about this is visible. The culture grows well, often better than before. Morphology may shift, but morphology varies with density and passage number anyway, so the change is easily attributed to normal drift. The line continues to be used, distributed to collaborators, and cited by its original name.
Databases of misidentified cell lines now list a substantial number of lines that are known to be something other than what their name implies, some of them widely used for decades before the substitution was recognised. The published work built on those lines is not necessarily wrong, but claims about the tissue or disease of origin certainly are, and reinterpreting the literature is not straightforward.
The only reliable defence is genetic authentication, discussed below, combined with disciplined handling: one line open in the cabinet at a time, dedicated media aliquots, and a properly maintained frozen stock established at a low passage number so there is always a verified point to return to.
Chemical Contamination From Media and Plastics
A third category involves no organism at all, and it is the easiest to misattribute to biology.
Media components vary. Serum is a biological product with lot-to-lot variation in growth factors, hormones and lipids, and a change of lot can shift proliferation rates and differentiation behaviour enough to look like a treatment effect. Laboratories running sensitive assays reserve a large batch of a tested serum lot for the duration of a project for precisely this reason.
Water quality matters more than it seems. Trace metals, endotoxin and organic residues in laboratory water all affect cells, and endotoxin in particular activates immune-responsive cell types powerfully at concentrations far below anything visible. Detergent residues on inadequately rinsed glassware are a classic cause of unexplained cell death that persists through every other troubleshooting step.
Plasticware leaches. Compounds used as plasticisers and stabilisers can migrate into medium, and some have measurable biological activity, including effects on hormone-responsive cells. Different manufacturers and different production lots of nominally identical plastics are not equivalent. Sterilisation residues are a related issue where gas-sterilised items have not been adequately aired.
Incubator problems complete the list. Carbon dioxide concentration drifting out of specification changes medium pH, which alters cell behaviour before it becomes visible as a colour change. Contaminated humidity trays disperse organisms and volatile antimicrobial agents alike. Volatile compounds from solvents or from disinfectants used carelessly inside an incubator affect every culture in the box simultaneously, which is a useful diagnostic clue: when everything in one incubator behaves oddly at once, the incubator is a better suspect than the cells.
Routine Testing Schedules That Work
A workable testing regime is one people actually follow, which usually means it is simpler than the ideal one.
Incoming material is the highest-yield checkpoint. Any culture arriving from outside the laboratory, whether from a repository, a collaborator or another group in the same building, is handled in quarantine until it has been tested for mycoplasma and, ideally, authenticated. Repository stocks are generally reliable; cultures passed between laboratories are the main route of transmission.
For established lines in continuous use, monthly mycoplasma testing is a common standard, with more frequent testing where cultures are shared between people or where a facility has a history of problems. Testing before freezing a batch of stock vials is essential, because freezing a contaminated culture preserves the problem indefinitely and reintroduces it every time the stock is thawed.
Frozen stock strategy does much of the work that testing cannot. Establishing a master stock at a low passage number, then working stocks derived from it, means a culture in use is never many passages from a verified starting point. Discarding working cultures after a defined number of passages and returning to frozen stock limits both genetic drift and the window in which an undetected contamination can spread.
Records make the system function. A log recording passage numbers, media lots, test dates and results turns an inexplicable result into a tractable investigation, because it becomes possible to ask what changed in the weeks before the anomaly appeared.
Authentication by Short Tandem Repeat Profiling
Cell line authentication uses the same principle as human identity testing. Scattered through the genome are regions where a short sequence repeats a variable number of times, and the number of repeats at each location differs between individuals.
Profiling a panel of these locations produces a set of numbers that functions as a fingerprint. The profile of a cultured line can be compared against reference profiles held in public databases and against the profile of the original donor material where it exists. A match confirms identity; a mismatch identifies the line as something else, and often identifies what it actually is.
The technique is not a universal solution. It distinguishes lines derived from different individuals but cannot separate two lines from the same donor, such as a tumour line and a matched normal line, without additional information. It works for human cells, and equivalent species-specific panels exist for common laboratory animals, but a separate species identification test is needed to detect contamination across species. Tumour lines can also show instability at these locations over long-term culture, so profiles drift somewhat, and interpretation guidelines allow for partial matches above a defined threshold.
Practice has shifted substantially. Many journals require an authentication statement, and many funders require it in applications. The realistic cadence for a laboratory is to authenticate on receipt, before establishing a frozen master stock, before starting a major project, and before submitting work for publication.
Decontaminating or Discarding a Culture
When contamination is confirmed, the decision is nearly always the same, and the exceptions are narrow.
For visible bacterial or fungal contamination, discard. The culture is replaceable from frozen stock, the time spent attempting rescue exceeds the time to thaw a new vial, and the risk of spreading the organism through the incubator while attempting treatment is real. The flask is sealed, removed from the incubator without being opened again, and disposed of appropriately, and the incubator is decontaminated if there is any suspicion of spread.
For mycoplasma, the calculation is similar but the reasoning differs. Treatment with agents active against organisms lacking a cell wall can clear an infection, but courses are long, the agents are toxic to the host cells to varying degrees, cells that survive treatment may have been altered by both the infection and the treatment, and clearance must be confirmed by repeated testing after the agent is withdrawn. Any treated culture must be handled in complete isolation throughout. This is worth doing for a line that genuinely cannot be replaced, such as a primary culture from an irreplaceable donor or a clone that took months to generate. For anything obtainable from a repository or a frozen stock, discarding is faster and safer.
For a line that fails authentication, there is nothing to decontaminate. The culture is whatever the profile says it is, and the questions become which experiments used it, when the substitution is likely to have occurred, and whether earlier frozen stocks are correct. A frozen stock predating the problem is the only route back.
The uncomfortable part of any of these decisions is the retrospective question. If a culture has been infected or misidentified for some period, work done during that period is in doubt, and establishing when the problem started is often impossible without dated frozen stocks that can be tested. That, more than anything, is the argument for scheduled testing: it converts an unbounded uncertainty into a bounded one.
Frequently asked questions
Do antibiotics in the medium prevent contamination?
They suppress some bacteria and mask others, which on balance makes routine antibiotic use a mixed proposition. Penicillin and streptomycin have no effect on mycoplasma, fungi or viruses, and their presence can allow a low-level bacterial contamination to persist undetected rather than declaring itself. Many experienced laboratories culture without routine antibiotics precisely so that lapses in technique become visible immediately, reserving antibiotics for genuinely high-risk work such as primary tissue isolation. Whichever approach a laboratory takes, antibiotics are never a substitute for aseptic technique.
How would mycoplasma have got into a culture that was never contaminated before?
The commonest route is an incoming culture from another laboratory, since transmission between cultures within an incubator or a cabinet happens readily once one infected flask is present. Biological supplements, particularly serum, are a second route, and laboratory personnel are a documented source for certain species. Because it spreads by aerosol during pipetting and through shared media bottles, a single infected culture in a shared facility can seed many others before anyone tests. This is why quarantine of incoming material is the highest-value single precaution.
Is a morphology change enough to suspect the wrong cell line?
It is a prompt to test, not evidence in itself. Cell morphology genuinely varies with confluence, passage number, serum lot, coating substrate and medium composition, so an apparent change often has an innocent explanation. The signs more suggestive of substitution are a marked change in growth rate, loss of a characteristic feature such as a specific marker or response, or a line behaving like a different well-known line. Since authentication is inexpensive relative to the cost of a compromised project, testing when suspicion arises is straightforward.
What should be done if contamination is found in a shared incubator?
Treat the incubator as affected rather than the single flask. Remove and discard obviously contaminated cultures without reopening them, test the remaining cultures, and decontaminate the incubator including the water tray, shelves and interior surfaces according to the manufacturer’s guidance. Check the obvious shared sources: media and supplement bottles in use, the cabinet’s airflow and filter service date, and water baths. Users should be told, since cultures moved elsewhere during the period of exposure may need testing too.
How long can a contaminated frozen stock stay undetected?
Indefinitely, because freezing preserves mycoplasma and any bacterial contaminant along with the cells, and thawing revives both. A stock frozen from an infected culture reintroduces the infection on every use, which is why testing before banking is more important than testing at any other point. Where records are incomplete, the practical approach is to test each stock ampoule after thawing and before it enters routine use, and to treat any untested legacy stock as suspect until it has passed.
The pattern worth internalising is that contamination becomes more damaging as it becomes less visible. Cloudy medium costs a week. Mycoplasma costs a set of experiments that nobody can repeat. A misidentified line costs the interpretation of everything built on it.
Because the damage is inversely related to how much attention the problem attracts, the defences have to be scheduled rather than reactive. Quarantine what arrives, test on a calendar rather than on suspicion, bank early and generously, and keep records detailed enough that a strange result can be traced backwards. None of this is intellectually interesting, and it protects more work than almost anything else done at the bench.
