RPMI 1640 Medium

RPMI 1640 was developed by George Moore, Robert Gerner and H. Addison Franklin at Roswell Park Memorial Institute in Buffalo, New York, and published in 1967 in a paper on the culture of normal human leukocytes. RPMI is the institute's initials; 1640 is the formulation number in a series. It began as a modification of McCoy's 5A (itself sometimes called RPMI 1630) and was aimed squarely at a problem the Eagle-derived media handled poorly: growing lymphoblastoid cells in suspension.

That origin explains almost every distinctive feature of the medium. Suspension cells do not need — and are inhibited by — the high calcium that promotes attachment. Leukocytes cultured over days need a buffering system that holds at the CO2 levels of a standard incubator. And the medium had to support cells that were fastidious about redox state, which is why it contains reduced glutathione, a component almost unique among classical basal media.

RPMI 1640 has since proven far more general than its name suggests. It supports many adherent lines perfectly well, and it is the standard base for the NCI-60 tumour panel, for hybridoma culture, for peripheral blood mononuclear cell work, and for most human leukaemia and lymphoma lines.

The standard RPMI 1640 formulation, as supplied by all major manufacturers:

Inorganic salts and bulk components

Amino acids (mg/L)

L-arginine 200; L-asparagine 50; L-aspartic acid 20; L-cystine·2HCl 65; L-glutamic acid 20; glycine 10; L-histidine 15; L-hydroxyproline 20; L-isoleucine 50; L-leucine 50; L-lysine·HCl 40; L-methionine 15; L-phenylalanine 15; L-proline 20; L-serine 30; L-threonine 20; L-tryptophan 5; L-tyrosine disodium salt dihydrate 29; L-valine 20.

Vitamins (mg/L)

i-Inositol 35; choline chloride 3; para-aminobenzoic acid 1; folic acid 1; niacinamide 1; thiamine·HCl 1; D-calcium pantothenate 0.25; biotin 0.2; riboflavin 0.2; vitamin B12 (cyanocobalamin) 0.005.

Four things in this list are worth pausing on, because they are what make RPMI RPMI:

• Calcium is very low. 0.42 mM, against 1.8 mM in DMEM and MEM. This is deliberate — low calcium discourages the cell–cell and cell–substrate adhesion that would defeat suspension culture.
• Phosphate is very high. About 5.6 mM from disodium phosphate alone, roughly six times DMEM's. This gives RPMI meaningful secondary buffering capacity beyond bicarbonate and contributes to its stability in routine handling.
• Reduced glutathione is present at 1 mg/L. No other classical basal medium includes it. It contributes antioxidant capacity, which matters for primary leukocytes.
• Biotin, vitamin B12, PABA and hydroxyproline are present, none of which appear in DMEM or MEM. RPMI's vitamin coverage is broader than the Eagle-derived media even though its amino acid concentrations are much lower.

Note also what is missing: RPMI contains no iron source and no trace elements. Serum supplies these.

RPMI 1640 contains 2,000 mg/L D-glucose — 2.0 g/L, or 11.1 mM. This is the standard formulation figure and it is the same across suppliers, because RPMI 1640 is a defined standard formulation rather than a proprietary one. If you are cross-checking a datasheet from any manufacturer against a protocol written for another, the glucose is 2.0 g/L in all of them unless the product name says otherwise.

The exceptions are explicitly labelled variants:

• ATCC-modified RPMI 1640 raises glucose to 4,500 mg/L (4.5 g/L, 25 mM) and adds 1 mM sodium pyruvate and 10 mM HEPES.
• Glucose-free RPMI 1640 contains none, and is used for metabolic work where you set the carbon source yourself.
• RPMI 1640 with HEPES keeps glucose at 2.0 g/L and adds the buffer.

At 11.1 mM, RPMI's glucose is close to physiological blood glucose and roughly half that of high-glucose DMEM. This is a genuine advantage for metabolic and immunometabolism work, where 25 mM glucose is a pharmacological concentration that suppresses oxidative phosphorylation in many cell types. It is a disadvantage for very dense cultures, which exhaust it sooner and need feeding more often.

The working pH of RPMI 1640 is 7.2 ± 0.2, measured on complete medium equilibrated in 5% CO2. As with any bicarbonate-buffered medium, a bottle measured cold on the bench reads higher because it has lost dissolved CO2, and returns to specification in the incubator. RPMI's high phosphate content gives it more CO2-independent buffering than DMEM, so it drifts more slowly during handling.

These are the two most-used basal media and they are not interchangeable. The headline differences are calcium, phosphate, glucose, amino acid concentration and bicarbonate — the comparison table at the foot of this page gives the full component-level view with the practical consequence of each difference.

The practical decision usually comes down to three questions.

Is the culture in suspension or attached? Suspension lymphoid and haematopoietic cells go in RPMI. Firmly adherent fibroblast and epithelial lines go in DMEM. RPMI's low calcium works against strong attachment; DMEM's 1.8 mM calcium supports it.

What does the line's published protocol say? For established lines this is not a matter of preference. Jurkat, K562, THP-1, Raji, U937, HL-60, MOLT-4, most NCI-60 panel members and most hybridomas have RPMI 1640 in their reference protocols. HEK293, HeLa, NIH/3T3, Vero, COS-7 and C2C12 have DMEM. Switching a line between them changes doubling time, saturation density and often protein expression, so it invalidates comparison with the literature unless you re-characterise.

What is the metabolic demand? DMEM at 25 mM glucose and roughly four times RPMI's amino acid concentration is built for dense, fast, high-demand cultures. RPMI at 11 mM glucose and modest amino acid levels is built for cells that grow at moderate rates in suspension over days. Pushing a high-demand culture in RPMI means feeding it more often.

One genuinely useful hybrid exists: the ATCC-modified RPMI 1640, which raises glucose to 4.5 g/L and adds 1 mM sodium pyruvate and 10 mM HEPES. It keeps RPMI's ionic character while giving the nutrient reservoir of a high-glucose medium, and is a good option for hybridomas and other cultures that are RPMI-dependent but grow to high density.

Established lymphoid and haematopoietic lines. Jurkat, K562, HL-60, THP-1, U937, Raji, Ramos, MOLT-4, Daudi, and most B- and T-lymphoblastoid lines. Standard supplementation is 10% FBS and 2 mM L-glutamine.

Primary human leukocytes. PBMC, isolated T cells, B cells, monocytes and dendritic cell cultures. RPMI's original design case. For long primary cultures, RPMI's glutathione content and broad vitamin coverage are meaningful advantages.

Hybridomas and antibody production. RPMI 1640, often with additional supplements, is the classical hybridoma base. Chemically defined hybridoma media are now the norm for production, but RPMI remains the standard for fusion and cloning.

Many adherent tumour lines. MCF-7, A549, PC-3, HT-29, SK-MEL and a great many others are routinely maintained in RPMI. There is no rule that RPMI is only for suspension cells; the NCI-60 panel is grown in RPMI across both adherent and suspension members.

Where RPMI is a poor choice: cells that need high calcium for junction formation or differentiation, very high-density fed-batch processes without a modified formulation, and clonal growth at low density without serum, which requires an F-12-derived or purpose-designed base.

RPMI 1640's 2,000 mg/L (23.8 mM) sodium bicarbonate is matched to a 5% CO2 atmosphere, which is what almost every incubator is set to. This is one of RPMI's quiet practical advantages over DMEM, whose 3,700 mg/L bicarbonate was formulated for 10% CO2 and therefore sits alkaline at 5%.

RPMI's high phosphate also gives it a secondary buffer that operates independently of CO2, so the medium tolerates time on the bench better than a low-phosphate medium of the same bicarbonate level. It still loses CO2 and drifts alkaline if left open, so keep bottle-open time short.

HEPES variants are supplied at 25 mM (or 10 mM in the ATCC modification). Use them for flow cytometry, live-cell imaging, extended sorting runs, or any protocol that keeps cells out of a CO2 atmosphere for more than a few minutes. Two caveats apply: HEPES generates hydrogen peroxide under light exposure, and at 25 mM it is toxic to some primary cells. Do not treat it as a free upgrade.

Phenol-red-free RPMI is available for fluorescence and absorbance assays, for flow cytometry where phenol red raises background, and for hormone-response work — phenol red has weak oestrogenic activity, which is a documented confounder in oestrogen receptor-positive lines such as MCF-7.

Serum. 10% FBS is standard; some protocols use 20% for primary or low-density cultures, and 5% for well-adapted lines.

Glutamine. RPMI contains 300 mg/L (about 2 mM) when supplied glutamine-containing. Because glutamine degrades in solution to ammonia, glutamine-free RPMI plus fresh addition at use — or the stable dipeptide L-alanyl-L-glutamine — gives more consistent results, particularly for cultures fed weekly rather than every two days.

2-Mercaptoethanol. 50 µM is standard for primary lymphocyte, hybridoma and mouse T cell culture. It supports cystine uptake and helps maintain intracellular glutathione. This is a genuine requirement for some primary cultures, not an optional extra.

Sodium pyruvate and non-essential amino acids. 1 mM pyruvate and 1x NEAA are common additions, particularly for the NCI-60 protocols and for cultures at reduced serum. RPMI already contains several non-essential amino acids, so NEAA supplementation adds less here than it does to DMEM.

Glucose-free RPMI is used for metabolic work — substituting galactose to force oxidative metabolism, or titrating glucose to a defined level. RPMI is a better base than DMEM for this because its baseline glucose is closer to physiological to begin with.

Powdered RPMI is the economical route at scale. It is supplied without sodium bicarbonate, which is added at reconstitution (2.0 g/L), the pH adjusted about 0.1–0.2 units below target to allow for the rise during filtration, and the solution sterile-filtered at 0.22 µm.

Liquid RPMI 1640 is stored at 2–8 °C protected from light, with a typical unopened shelf life of 12–24 months. Once glutamine has been added, plan on about four weeks of working life; serum-supplemented medium, two to four weeks.

Light exposure is the most underestimated storage problem. Riboflavin, folic acid and tryptophan photodegrade, and where HEPES is present the reaction generates hydrogen peroxide. Medium stored on an open shelf under laboratory lighting measurably loses plating efficiency. Keep bottles in a closed refrigerator, not a lit display fridge, and do not leave working aliquots on the bench between uses.

Warm only what you need. Repeatedly bringing a full 500 mL bottle to 37 °C accelerates glutamine hydrolysis and drives off bicarbonate. Room temperature is adequate for most feeding.

A faint pink deepening toward magenta after a bottle has been open is normal bicarbonate loss and reverses in the incubator. Cloudiness, particulates or a shift to yellow in an uninoculated bottle are not normal — discard, and check your filtration or aseptic technique rather than the medium.

Assuming RPMI is only for suspension cells. A large fraction of adherent tumour lines are maintained in it routinely.

Switching a line from DMEM to RPMI (or back) mid-project. Growth rate, saturation density, and expression of some genes change. If you must switch, adapt gradually over several passages and re-baseline your assays.

Forgetting 2-mercaptoethanol for primary lymphocyte or hybridoma culture. Its absence is a common reason primary T cell cultures underperform.

Using RPMI for cells that need calcium. At 0.42 mM, RPMI will not support the junction formation or calcium-driven differentiation that some epithelial protocols depend on.

Treating the ATCC modification as plain RPMI. It differs in glucose (4.5 versus 2.0 g/L), pyruvate and HEPES. Record which one you used.

Adding glutamine and not dating the bottle. Ammonia from glutamine decay accumulates and is toxic to lymphoid cells at lower concentrations than many adherent lines tolerate.