Antibiotics and Antimycotics in Cell Culture

An antibiotic added to growth medium is a suppressant, not a sterilant. At the concentrations that cultured mammalian cells tolerate, these agents inhibit the growth of susceptible organisms; they do not reliably kill every organism present, and they have no effect at all on organisms outside their spectrum.

That distinction drives everything else on this page. Adding penicillin–streptomycin to a flask does not make it sterile. It makes contamination quieter. A culture on continuous antibiotics can carry a low-level bacterial population that never clouds the medium, and that population is still consuming nutrients, still shedding endotoxin, and still changing your results.

Three things antibiotics in cell culture cannot do, which are worth stating plainly because they are the source of most false confidence:

• They cannot touch mycoplasma with a cell-wall-acting agent. Mollicutes have no peptidoglycan wall, so penicillin has no target. A culture kept permanently on penicillin–streptomycin has exactly no protection against the most damaging contaminant in cell biology.
• They cannot prevent fungal contamination unless an antimycotic is present, and amphotericin B's usable window in culture is narrow.
• They cannot prevent cross-contamination between cell lines, which is arguably the most consequential contamination of all.

This page covers antibiotics and antimycotics used to suppress contamination. Selection agents — G418, hygromycin B, blasticidin, zeocin, puromycin — are a different class used for a different purpose and are covered in the selection antibiotics reference.

Penicillin G (sodium or potassium salt) is a β-lactam that blocks the transpeptidase cross-linking step of peptidoglycan synthesis. Its spectrum is essentially Gram-positive. Supplied in 100X solution at 10,000 U/mL, giving 100 U/mL at 1X.

Streptomycin sulfate is an aminoglycoside that binds the 30S ribosomal subunit and causes misreading of mRNA. It covers Gram-negative organisms and some Gram-positives, which is why it is paired with penicillin. Supplied at 10,000 µg/mL in 100X, giving 100 µg/mL at 1X.

The pairing is conventionally supplied together as penicillin–streptomycin 100X, added at 1% v/v (10 mL into 1 L, or 5 mL into a 500 mL bottle).

Gentamicin sulfate is a broad-spectrum aminoglycoside covering Gram-positive and Gram-negative bacteria and, at the upper end of its range, some mycoplasma. Working concentration is 5–50 µg/mL, commonly 50 µg/mL; supplied as a 10 mg/mL stock, so 50 µg/mL is a 1:200 dilution. Gentamicin's practical advantage over penicillin–streptomycin is stability: it is reported stable across pH 2–10 for 15 days at 37 °C in tissue culture medium, whereas penicillin has a short half-life at 37 °C and loses activity quickly at both acidic and alkaline pH.

Kanamycin sulfate covers Gram-positive and Gram-negative bacteria and some mycoplasma at 100 µg/mL, with reported stability of about 5 days at 37 °C in medium.

Amphotericin B is a polyene antimycotic. In the cell-culture context it binds ergosterol in fungal and yeast membranes and forms pores that cause leakage of intracellular contents. Working concentration is 0.25–2.50 µg/mL; it is supplied as a 250 µg/mL solution, which at 1:100 gives 2.5 µg/mL and at 1:1000 gives 0.25 µg/mL. Reported stability is about 3 days at 37 °C, so it does not survive long between medium changes. Note that mammalian membranes contain cholesterol rather than ergosterol, which is what gives amphotericin B its selectivity — but the selectivity is not absolute, and cytotoxicity to sensitive cell lines at the upper end of the range is real. Determine the tolerated concentration empirically for each line.

Antibiotic–antimycotic 100X combines all three: 10,000 units/mL penicillin, 10,000 µg/mL streptomycin and 25 µg/mL amphotericin B, giving 100 U/mL, 100 µg/mL and 0.25 µg/mL respectively at 1X. It is the convenient option when antifungal cover is genuinely needed, and it carries the lowest of the amphotericin B working concentrations, which is the sensible default.

This is the settled position among cell culture authorities — culture collections, the major reagent suppliers' own technical literature, and Good Cell Culture Practice guidance all say the same thing. We sell these reagents, and the position is still correct.

They mask low-level contamination. This is the primary objection. A sub-clinical bacterial population held below the turbidity threshold by continuous antibiotics is invisible but not absent. It consumes nutrients, produces endotoxin, and skews results. Worse, it removes the feedback loop: a lab that never sees a contaminated flask never discovers that its aseptic technique has drifted, until an antibiotic-free experiment or a QC test exposes months of accumulated problems.

They select for resistant organisms. Continuous sub-lethal selection pressure in a warm, nutrient-rich, 37 °C environment is close to a textbook enrichment protocol. Labs that run permanent antibiotics tend, over time, to acquire resistant contaminants that are far harder to eliminate than the sensitive population they started with.

They give no protection against the contaminants that matter most. Mycoplasma is intrinsically resistant to cell-wall-acting agents. Viruses are unaffected. Cross-contamination between cell lines is unaffected. Fungi are unaffected unless an antimycotic is included.

They can alter cell physiology. Aminoglycosides affect mitochondrial translation, since mitochondrial ribosomes resemble bacterial ribosomes. Published work has documented antibiotic effects on gene expression, differentiation capacity and drug response in mammalian cultures, and amphotericin B has been shown to alter viral replication in cell culture systems. If your medium contains a pharmacologically active compound throughout every experiment, that compound is an uncontrolled variable.

They encourage bad habits. The cost most often understated: antibiotics let a lab tolerate a hood that has not been certified, a waterbath that is never changed, and a technique that involves reaching over open vessels. The reagent absorbs the consequences until it cannot.

The argument above is against routine prophylactic use in established cell lines. It is not an argument that these reagents have no place. Legitimate uses:

• Primary isolation from tissue. Tissue arriving from surgery, biopsy or dissection carries its own flora, and the material is often irreplaceable. Antibiotics — frequently with an antimycotic — during the isolation and the first few passages are standard and sensible. Wean them off once the culture is established.
• Material from non-sterile sites — gut, skin, oral tissue, environmental or field-collected samples.
• Primary cultures that cannot be replaced, where the cost of losing the material exceeds the cost of the confounder.
• Short-term rescue of an irreplaceable culture that has become contaminated, as a bridge to re-deriving a clean stock — not as a permanent state.
• Some large-scale bioprocess runs, where the value at risk in a single bioreactor batch justifies the insurance. Even here, many manufacturers run antibiotic-free precisely so that a contamination event is detected rather than hidden.
• Selection of stably transfected lines. This is a different use of a different class of agents at cytotoxic concentrations, and it is not prophylaxis.

Note that some antibiotics are excluded from specific applications on scientific grounds. Media formulated for embryo and gamete work, for instance, are commonly supplied without penicillin and streptomycin.

Moving an established lab off routine antibiotics is straightforward, and the transition itself is diagnostic — whatever fails in the first fortnight tells you exactly where your technique is weak.

1. Fix the environment first. Certify the biosafety cabinet, replace the HEPA filter if it is due, decontaminate the incubators including shelves and water pans, and change the waterbath to a system that does not require immersion (bead baths or dry blocks remove one of the most reliable contamination routes).
2. Test the lines you intend to keep for mycoplasma before the transition, so you are not carrying an existing problem into the new regime.
3. Thaw fresh vials into antibiotic-free medium rather than weaning existing flasks. A culture that has been on antibiotics for years may be carrying suppressed contamination that will bloom on withdrawal — better to discover that in a dedicated test flask than in your main stock.
4. Keep an antibiotic-containing backup for one or two passages while you confirm the antibiotic-free culture is clean. Then discard it.
5. Tighten technique deliberately: one line open at a time, never reach over an open vessel, aliquot supplements into single-use volumes, spray and wipe every item entering the hood, and keep talking out of the cabinet.
6. Expect and welcome the first contamination. A visible contamination event in an antibiotic-free lab is information. In an antibiotic-dependent lab, the same event happened silently three months earlier.

• Store concentrates frozen and protected from light. Penicillin–streptomycin, antibiotic–antimycotic and amphotericin B solutions are typically stored at −30 to −5 °C, protected from light; gentamicin is stable at 2–30 °C.
• Aliquot on first thaw. Repeated freeze–thaw cycles degrade β-lactams in particular. Divide a 100 mL bottle into single-use aliquots sized to one bottle of medium.
• Do not assume the antibiotic lasts as long as the medium. Penicillin G and gentamicin are reported stable for around 3 days at 37 °C in medium and kanamycin for around 5 days; amphotericin B for around 3 days. Medium that has sat on cells for a week has, functionally, no penicillin activity left in it.
• Add after filtration, not before, if you are supplementing a medium you filter yourself, and never autoclave an antibiotic-containing solution.
• Do not exceed the recommended working concentration in the belief that more is safer. Cytotoxicity rises faster than antimicrobial benefit, and amphotericin B in particular has a narrow usable window.
• Withdraw antibiotics for at least two passages before mycoplasma testing. Residual antibiotic can suppress the organism below the assay's detection limit and produce a false negative.