Iodixanol as a Density Gradient Medium

Iodixanol (CAS 92339-11-2, molecular weight 1550.2) is a non-ionic, highly water-soluble compound built from two tri-iodinated benzene rings joined by a hydroxylated propyl bridge, with multiple dihydroxypropyl substituents. It belongs to the same chemical family as non-ionic iodinated X-ray contrast agents, which is where the molecular design comes from: these compounds were engineered to be dense, extremely soluble, non-ionic and biologically well tolerated, and those are precisely the properties a good density gradient medium needs.

For laboratory use it is supplied as a sterile 60% (w/v) aqueous solution -- the format sold as OptiPrep and by other suppliers as a 60% iodixanol solution. The key specifications are:

• Density: 1.32 g/mL at 20 C
• Osmolality: approximately 290 mOsm/kg
• Non-ionic, so it contributes essentially nothing to ionic strength
• Sterile and endotoxin-controlled
• Metabolically inert and non-toxic to cells at working concentrations

The iso-osmotic property is the whole point. In a sucrose gradient, reaching a density of 1.25 g/mL requires a solution of several hundred milliosmoles above physiological -- typically well over 1000 mOsm/kg. Anything with a membrane placed in that solution loses water and shrinks, which changes its buoyant density, distorts its morphology and, for cells and organelles, damages it. Iodixanol reaches the same densities at roughly physiological osmolality because the molecule is large and heavy: you get many grams of mass per mole of dissolved particle. A single 1550 Da molecule contributes the same osmotic pressure as a single 342 Da sucrose molecule, but four and a half times the mass.

The practical consequence is that a particle's position in an iodixanol gradient reflects its true hydrated buoyant density rather than a density it acquired by being dehydrated during the run. For quantitative work -- comparing full and empty AAV capsids, characterising extracellular vesicle subpopulations, measuring organelle density -- this is not a minor convenience but the difference between a meaningful and a misleading result.

The 60% stock is water-based and contains no salts or buffer. Using it directly on biological material would expose the sample to a solution with no ionic support, so it is normally converted into a working solution first, or diluted directly into a buffer of choice.

The classical working solution. Mix 5 volumes of 60% iodixanol with 1 volume of a 6x diluent -- typically 0.85% NaCl with 60 mM Tricine-NaOH at pH 7.4, or an equivalent concentrated buffer. This gives a 50% (w/v) iodixanol working solution that is isotonic and buffered, with a density of approximately 1.27 g/mL. All further dilutions are then made from this working solution using 1x diluent (0.85% NaCl, 10 mM Tricine-NaOH pH 7.4), so ionic strength and pH remain constant across the entire gradient while only the iodixanol concentration -- and therefore the density -- varies.

This constant-buffer principle is what makes iodixanol gradients well behaved. Vary one parameter, hold everything else fixed.

Direct dilution into an application buffer. For virus work it is common to skip the classical working solution and dilute the 60% stock straight into the buffer the process uses, most often PBS-MK (phosphate-buffered saline with 1 mM MgCl2 and 2.5 mM KCl). This is what the standard AAV protocol does.

Calculating a dilution. To make volume V of a solution at concentration C% from the 60% stock:

Volume of 60% stock = V x C / 60 Volume of diluent = V - (V x C / 60)

For example, 10 mL of 40% iodixanol requires 10 x 40/60 = 6.7 mL of 60% stock and 3.3 mL of diluent.

Practical handling. The 60% solution is noticeably viscous, so pipette slowly and allow tips to drain fully -- casual pipetting is a real source of error in gradient composition. Warm to room temperature before use, since viscosity falls sharply with temperature. Store at room temperature protected from light; do not freeze. Iodixanol solutions are stable for long periods but should be kept sterile if the downstream product must be.

Adding phenol red as a visual marker to alternate steps is a small trick that pays for itself. In the standard AAV gradient, phenol red is added to the 25% and 60% layers, so the completed gradient shows visible coloured and colourless bands, the interfaces are obvious, and you can see immediately whether the layering was clean and whether the gradient was disturbed in transit to the rotor.

The four-step iodixanol gradient introduced by Zolotukhin and colleagues in 1999 remains the standard laboratory method for purifying recombinant AAV from crude cell lysate, and it is the single most common reason people search for iodixanol.

Why it works. AAV capsids containing a packaged genome are denser than empty capsids, because DNA is dense. A gradient tuned to that difference separates full from empty particles while simultaneously leaving behind the bulk of cellular protein and nucleic acid.

Gradient composition. Layered into a sealable ultracentrifuge tube, densest first, so that the sample sits on top:

The 1 M NaCl in the 15% step is not incidental -- it is doing specific work, stripping electrostatically bound host protein and nucleic acid from the capsid surface as the virus passes through. Omitting it noticeably reduces purity.

Centrifugation. In a fixed-angle rotor such as a Type 70 Ti, 350,000 x g for 90 minutes at 10 C is a standard condition; 200,000 x g for 2 hours at 18 C is an equivalent alternative. Sealed QuickSeal-type tubes are used because the tubes must be completely full.

Harvesting. Collect from the 40% layer, drawing from just below the 40-60% interface. Full capsids concentrate at the bottom of the 40% step. Puncture the tube side wall with a needle just below the interface and draw slowly, taking the 40% layer while avoiding the 25-40% interface above, where empty capsids and residual protein accumulate. Collecting too high is the most common way to carry empty capsids into the final product.

What to expect. Iodixanol-purified preparations typically still contain around 20% empty particles on average. The gradient is a strong enrichment step, not an absolute separation. Where a very low empty ratio is required, iodixanol is normally followed by an orthogonal step such as anion exchange chromatography, or replaced by analytical-grade caesium chloride banding, at the cost of throughput and capsid integrity respectively.

Removing the iodixanol. Iodixanol must be removed before most downstream applications, particularly anything in vivo or any assay sensitive to viscosity or refractive index. Buffer exchange by dialysis, desalting column, centrifugal concentrator or tangential flow filtration all work. Do not assume dilution alone is sufficient -- residual iodixanol interferes with protein quantification and with some titration methods.

Cell separation. Iodixanol supports both density-barrier and continuous-gradient cell separations at physiological osmolality, which is its main advantage over the polysucrose-diatrizoate media traditionally used for this purpose. Blood cell separations rely on well-established density cut-points -- mononuclear cells are conventionally recovered above a 1.077 g/mL barrier, while granulocytes and erythrocytes pellet through it. Iodixanol can be adjusted to any required density with a simple dilution, and because it is iso-osmotic, cells are not dehydrated during the separation, so their density during the run is their real density.

For cells, layer the sample over the barrier gently, use a swinging-bucket rotor, and centrifuge with the brake off or at minimum so that the interface is not disturbed during deceleration. Recover from the interface, dilute in buffer, and wash to remove the gradient medium.

Subcellular fractionation. Iodixanol separates mitochondria, lysosomes, peroxisomes, endoplasmic reticulum, Golgi and plasma membrane fractions, and it is particularly valuable for organelles whose densities are close together, since sucrose gradients distort them differentially by osmotic dehydration. Both pre-formed step or continuous gradients and self-generated gradients are used -- iodixanol will form a gradient in situ when a uniform solution is centrifuged at high speed in a vertical or near-vertical rotor, which removes the manual layering step and improves reproducibility between runs.

Extracellular vesicles. Density gradient flotation has become the reference method for separating extracellular vesicles from co-isolating protein aggregates and lipoproteins, which pelletable-by-ultracentrifugation methods cannot distinguish. A typical approach loads the crude vesicle preparation at the bottom of the tube in a dense iodixanol layer and lets vesicles float up into a gradient of lower-density steps -- for example 40%, 20%, 10% and 5% -- during a long spin. Floating rather than pelleting is deliberate: protein aggregates are dense and stay behind, while membrane-bounded vesicles rise to their buoyant density. Small extracellular vesicles typically band in the region of 1.08-1.19 g/mL, though the exact figure depends on the source material and should be determined for your system by measuring the refractive index of collected fractions rather than assumed.

Other viruses. The same principles used for AAV apply to lentivirus, adenovirus, influenza and many others, with the step percentages adjusted to bracket the buoyant density of the particle in question. Iodixanol is generally kinder to enveloped viruses than sucrose or caesium chloride, both of which can strip envelopes and reduce infectious titre.

Use iodixanol when you need physiological osmolality, you are working with intact cells, organelles, enveloped viruses or vesicles, you want buoyant densities that mean something, or you need a medium that can be removed by simple buffer exchange without dialysis against large volumes.

Sucrose is cheap, familiar and entirely adequate for robust, non-osmotically-sensitive material. Its problem is osmotic: high-density sucrose solutions are strongly hypertonic, dehydrating anything with a membrane, shifting apparent densities and damaging enveloped viruses and organelles. It is also viscous at high concentration, which lengthens runs.

Caesium chloride reaches far higher densities than any non-ionic medium, up to around 1.9 g/mL, and gives excellent resolution for nucleic acids and for separating full from empty AAV capsids. The costs are substantial: very high ionic strength, extended run times, mandatory dialysis to remove the salt, and documented loss of AAV infectivity relative to iodixanol -- which is exactly why the field moved to iodixanol for preparative work while retaining CsCl for analytical applications.

Percoll is a colloidal silica suspension coated with polyvinylpyrrolidone. It is iso-osmotic when adjusted with salts and is excellent for cell and organelle separations, but its maximum working density is limited to around 1.13 g/mL, which puts most viruses out of reach. Because it is a particulate colloid rather than a true solution it must be removed from the product, and it is not suitable where the preparation will go into an analytical method sensitive to particles.

Polysucrose-sodium diatrizoate media (the Ficoll-Paque family) are the traditional choice for peripheral blood mononuclear cell isolation at a 1.077 g/mL barrier. They are well validated for that specific job and remain widely used for it, but they are less flexible than iodixanol for building arbitrary densities and are not iso-osmotic across their full range.

Nycodenz, the non-ionic iodinated monomer that preceded iodixanol, reaches high density but becomes hyperosmotic above roughly 1.15 g/mL. Iodixanol was developed as the dimer specifically to fix that limitation, which is why it superseded Nycodenz for most high-density applications.

Layer carefully, from the bottom up or the top down consistently. Either underlay denser solutions beneath lighter ones with a long cannula, or layer lighter solutions gently onto denser ones down the tube wall. Mixing methods within one gradient produces disturbed interfaces. Use phenol red in alternate steps so you can see what you have built.

Fill sealed tubes completely. Ultracentrifuge tubes for fixed-angle rotors at these speeds must be full or they will collapse. Check the rotor manual for minimum fill.

Balance to within the rotor tolerance. At 350,000 x g an imbalance is not a minor matter.

Let gradients relax, but not too long. Pre-formed step gradients can be allowed to stand briefly to soften sharp interfaces, but a gradient left for hours will diffuse into something you did not design.

Deceleration. Use the brake off, or the lowest setting, for anything where an interface must survive the stop. This applies especially to cell separations.

Measure fractions rather than trusting the map. If a result matters, collect fractions and measure the refractive index of each to establish the actual density profile, then relate your material's position to a real number. Gradients do not always form exactly as planned, and refractive index measurement takes minutes.

Common problems:

• Material is spread across many fractions rather than banded. Usually insufficient run time, or a gradient whose density range is too wide. Narrow the range around the expected density and extend the spin.
• Nothing bands where expected. Confirm the actual gradient by refractive index before adjusting anything else. Check also that the sample was loaded at the right position -- flotation and sedimentation approaches load at opposite ends.
• Poor AAV purity. Check that the 15% step contained 1 M NaCl, and check where you drew the collection -- taking material too high in the 40% layer carries empty capsids across.
• Low recovery. Iodixanol is viscous; slow, careful harvesting matters. Also confirm the material was not left in gradient medium for an extended period before buffer exchange.
• Downstream assays misbehaving. Residual iodixanol interferes with protein quantification and refractive-index-based methods. Complete the buffer exchange before assaying.