Spheroid and organoid microwell plates: aggregates per well, and what each format costs

The question this category answers is “how many spheroids do I need, and how much do I care that they are the same size”. Those two requirements pull in opposite directions, and the plate formats on the market sit at different points along that trade-off — from one carefully handled spheroid per well to twelve thousand in a flask.

Buy from the assay backwards. If the endpoint is per-spheroid imaging with tracking, you need one per well in an addressable format. If the endpoint is a biochemical readout on pooled material, you need mass production and the addressing is wasted money.

The comparison

Two things fall straight out of that table.

The addressable formats cost roughly ten times more per vessel than the bulk formats, and they are worth it when you need addressing. An Akura 384 plate at $688 for five gives you 1,920 individually trackable spheroids at about $0.36 each in plasticware, with a flat imaging bottom and a defined pipetting geometry. An Elplasia 12K flask gives you twelve thousand spheroids you cannot tell apart. Neither is better. They answer different questions.

Microwell count per well spans a factor of 40 within one product line. AggreWell 400 in a 6-well plate gives 5,900 aggregates per well; AggreWell 800 in a 24-well gives 300. Same brand, same chemistry, entirely different experiment. Getting this wrong is the most common ordering error in the category.

Spheroid size is set by seeding density, not by the plate

Both the Kugelmeiers technical material and the AggreWell protocol literature state it plainly: spheroid size is controlled by initial seeding density, because every microwell receives an approximately equal share of the cells you put in the well.

The arithmetic is therefore fixed and worth doing before you order:

cells to load per well = (microwells per well) × (cells you want per spheroid)

The published worked example, from the AggreWell 400 protocol: to make spheroids of 1,000 cells each in a well containing 1,200 microwells, load 1,200 × 1,000 = 1.2 × 10⁶ cells, delivered in 0.4 mL — a density of 3 × 10⁶ cells per mL.

That equation is what makes microwell plates predictable, and it is also what constrains them. Published limits for AggreWell:

The source that publishes this table also states the numbers are approximations that vary with cell size and should be determined empirically. Take the ratios seriously and the absolute figures as a starting point.

Two asymmetries in that table are real design constraints:

• The small microwells have no lower size limit because they taper to a sharp point. Routine production of spheroids from an average of as few as 20 cells each is described as straightforward, though variability between spheroids becomes more significant at smaller sizes.
• The large microwells do have a lower limit, because the 800 µm well is not fully tapered. Attempts to form spheroids from small cell numbers can produce multiple smaller spheroids in each microwell — a failure mode that looks like poor uniformity and is actually the wrong plate.

The surface is doing the work

Every plate in this category depends on cells not sticking to the vessel, and the vendors implement that differently:

• Corning ULA — a proprietary, animal-free, covalently bonded hydrogel. Covalent bonding matters: it is the difference between a coating that survives 30 days of culture and one that does not. Corning states culture for 30 or more days is supported, cell-line dependent.
• Kugelmeiers AziGrip EX — a non-fouling coating applied only to the functional wells (A1–A6 and C1–C6), with the plate body in COC. The technical material emphasises sharp borders between microcavities to prevent cells settling anywhere other than in a microcavity, which is a geometric solution to the same problem the coating addresses chemically.
• AggreWell — relies on a surfactant rinsing step to prevent cell–surface interaction. Note the asymmetry: the protocol describes the surfactant step as generally advisable and comparable with and without for standard plates, but states it is not optional for AggreWell 400Ex due to slight differences in surface properties. Read the variant-specific protocol, not the family one.

A coating is a consumable property with a shelf life. Any custom microwell device you commission needs the same three questions answered as a catalogue one: what is the coating, how is it bonded, and how long does it last in culture. The surface treatment guide covers what answers are credible.

Bubbles: the failure mode nobody warns you about

Microwells trap air. Every published microwell protocol includes a debubbling step, and it is not optional garnish — a microwell with a bubble in it produces no spheroid, and a plate with a scattering of them produces a distribution with a spike at zero.

The published mitigation is specific: add liquid at one side of the well and let it spread across the surface rather than dropping it vertically onto the surface, then centrifuge for 2 minutes at 2,000 × g to clear residual bubbles, verifying under a low-magnification inverted microscope that they have gone. That centrifugation step requires a plate rotor properly balanced at 2,000 × g, which is a piece of equipment, not a technique. If your laboratory does not have one, choose a format with larger cavities or accept the loss.

Format selection: the decision in four questions

1. Do you need to track individual spheroids over time? If yes, you need one spheroid per addressable well — the Akura-style formats. Everything else pools identity.
2. How many spheroids does the endpoint consume? A biochemical assay on pooled material at several timepoints and several doses runs into thousands quickly, and that is where the Elplasia flask at ~12,000 per vessel and the SP5D at ~9,000 per plate earn their place.
3. What size spheroid does the biology need? Work backwards through the seeding equation. Large spheroids from many cells push you to 800 µm microwells; small uniform aggregates push you to 400 µm.
4. Does it have to fit an automated workflow? Then the footprint and the pipetting geometry matter as much as the biology. The Akura Twin is explicitly built to ANSI/SLAS 1-2004 with a published tip position; the SP5D is on the standard SBS footprint at 20.15 mm tall, which is above the SLAS 2 typical height and needs a z-clearance check. See plate format and instrument compatibility.

What the flask format changes

The Elplasia 12K is worth separating out because it is not a plate. It is an 80 cm² flask with 152 microcavities per cm², and its design consequences are different:

• One common medium reservoir across all ~12,000 spheroids, so every spheroid sees identical culture conditions — an advantage for consistency and a total loss of per-condition addressing. You cannot dose this vessel two ways.
• An internal liquid diverter to minimise spheroid disruption during medium exchange, plus a microcavity geometry that keeps spheroids in place during exchange while still permitting full recovery at harvest. Handling in a flask is otherwise the weak point of the format.
• Harvest by tilting. Add cells with the flask upright, move it to the incubation position so cells settle into the microcavities, then return to upright and rinse to collect.
• Working spheroid dimensions of 500 × 600 µm in an 850 × 650 µm cavity, which is a hard ceiling on final size that a plate with an open well does not impose.

If your process ends in a suspension of spheroids going into a downstream step — encapsulation, bioprinting, a bioreactor, an implantation study — the flask is the right shape. If it ends in a plate reader, it is not.

What this hardware does not do

Stated plainly, because it is regularly assumed otherwise:

• It does not make spheroids uniform on its own. It makes the number of cells per aggregate uniform, assuming an even distribution of a single-cell suspension. Incomplete dissociation, clumps in the suspension, or differential settling all propagate straight through.
• It does not supply cells, matrix or medium. Every product on this page is empty plasticware. If what you want is delivered, characterised spheroids rather than the vessel to make them in, that is a different purchase — see liver microtissues and spheroids, where the same vendor sells assay-ready microtissues at a different order of magnitude of price.
• It does not confer organoid identity. Aggregation is not differentiation. The plate produces an aggregate; the medium and the protocol produce the model.

How we can help

We do not manufacture plates and most of these are catalogue purchases you can make directly. Where we are useful: when the microwell geometry you need does not exist as a catalogue part, when you need a custom microcavity array in a specific footprint or material, or when you are deciding between buying the plasticware and buying finished microtissues — a comparison that turns entirely on your labour cost and the number of runs. Send us the specification, including the target spheroid size, the number you need per run and the endpoint, and we will tell you which side of that line you are on.