Experiment Design

Designing Cell-Activity Assays

We assume that a cell line and chemical chemotactic factor have already been selected as the focus of the assay, and that the effects of the chemical on the cells’ activity will be demonstrated using an instrument that includes sites at which the chemical is in contact with the under side of a membrane filter while a suspension containing the cells lies on top of the filter. For simplicity, we will also assume that chemotaxis (directional cell motion along a concentration gradient of a chemical) is the cell activity being detected and measured. (Neuro Probe instruments have many applications other than straightforward chemotaxis assays – for example experiments focusing on transmigration through an endothelial layer or metastasis through extracellular matrix proteins. Much of the following discussion can be adapted to such applications.)

At the simplest level, an assay must have “unknown” sites (with cell suspension above the filter and a solution containing the chemotactic factor below it) and “negative control” sites (with cell suspension above the filter and suspension media, but no chemotactic factor, below). Random migration of unstimulated cells will account for some of the cells that pass through the filter. Migrated cells at the negative control sites show the extent of unstimulated random migration, which can then be differentiated from chemotactic migration, or chemotaxis. The assay design will attempt to maximize the differential between the number of migrated cells at the negative and unknown sites. The following discussion is confined to this most basic assay goal, but the same kinds of considerations are implicit in assays of more complex design. (See controls for a discussion of various kinds of control and test sites commonly included in cell-activity assays.)

Many factors must be taken into account in designing a cell-activity assay. We can only point out the main types of considerations (see factors to consider) and describe two main strategies that can be used (see two design strategies).

Factors To Consider

Three main kinds of considerations are relevant in designing a cell-activity assay: (I) Characteristics of the cells, (II) Characteristics of the membrane filter, and (III) Characteristics of the detection system used to detect and measure the cells that pass through the membrane filter. We recommend a careful review of the literature to become as familiar as possible with these factors. Here are some of the considerations in each category:

  1. Characteristics of the cells.
    1. size of the cells when attached to a surface (The size of the cells as they adhere to the filter is the relevant factor, not their diameter in suspension or in other conditions.),
    2. size of the cell nuclei,
    3. deformability of the cell nuclei,
    4. how fast the cells migrate when stimulated and
    5. when unstimulated,
    6. whether close proximity or contact between cells affects cell activity, including chemotaxis, and
    7. whether the cell population contains multiple cell types.
  2. Characteristics of the membrane filter.
    1. Membrane material. (This discussion assumes a track-etch, capillary-pore membrane of polycarbonate or polyester material. Polycarbonate filters are currently available from Neuro Probe. Cellulose nitrate filters are used in some applications, but are not considered here.)
    2. Pore diameter. (Available diameters are given in the table below.)
    3. Pore density. (Available pore densities vary with pore diameter. See the table below.)
    4. Surface treatment. (Most diameter/density variations are available with the wetting agent polyvinylpyrrolidone (PVP). Filters without this treatment are best for assays involving quantifying cells adhered to the membrane, since some cells do not adhere as well when PVP is present. PVP may also inhibit extracellular matrix proteins from sticking to the membrane, and thus are not the best choice if such proteins must be applied to the filter.)
    5. Membrane thickness. (Thickness is 6 to 14 microns.)
    6. Pore angle: (Pores are cylindrical, straight, and within 34 degrees of being perpendicular to the surface.)

    Neuro Probe currently stocks polycarbonate membrane filters with the following pore-diameter/pore-density specifications. Framed microplate-size filters as well as smaller unframed filters are available. See filters for additional information.

    Pore Diameter1
    2 3 5 8 10 12 14
    Pore Density2
    20,000 20,000 4,000 1,000 1,000 1,000 500
    Pore Area/
    Unit Area
    6.28 14.14 7.85 5.03 7.85 11.31 7.70

    1Diameter specification is +0% to -20%. An 8 micron filter has pore diameters between 6.4 and 8 microns; a 2 micron filter has pore diameters between 1.6 and 2 microns. Nonstock pore diameters can be ordered but may be expensive and involve delays in delivery. Contact us for more information.
    2Pore density specification is +15% to -15%.

    Multiply the pore density (see the table above) by the exposed filter area to obtain the number of pores per site. The exposed filter areas for some of our most frequently used chambers are:

    • 101-series ChemoTx® system – 8mm2
    • 106-series or 116-series ChemoTx® system – 25mm2
    • 48-well chambers – 8mm2
    • MBA96 – 32mm2
    • MBB96 – 18mm2
    • MBC96 – 8mm2

    Please refer to the product page or protocol to find the filter area for chambers not listed above.

  3. Characteristics of the detection system.Normally one of the following five methods is used to count migrated cells.
    1. Adherent cells on the filter bottom or in the pores are stained and counted under a microscope by eye, or with an image analyzer. Although time and labor intensive, this method is highly sensitive, which allows good results with very low numbers of migrated cells at each site. This method is not appropriate for non-adherent cells (ones that drop off the bottom of the filter, as do lymphocytes).
    2. Migrated cells are read in a microplate with a densitometric or ELISA microplate reader. See automated readers for a description. This method counts all migrated cells, both adherent and non-adherent, but it is not sensitive and requires a minimum of 3,000 cells at each site.
    3. Cells are tagged with a fluorescent dye, such as Calcein AM from Molecular Probes, and read in a microplate with a fluorescence microplate reader. See automated readers for a description. The method works well only with cells that will retain a fluorescent tag, and if the tag is introduced before incubation it may influence cell function. Adherent cells on the filter and non-adherent cells in the microplate can be read separately or read together with the filter attached to the plate. This method is extremely sensitive, so requires relatively few cells.
    4. Cells treated with a radioactive tag are read in an opaque white microplate, using a scintillation counter. Radioactive tags may influence cell function. The number of cells required at each site will vary with the tag used and the quantity of tag taken up by the cells, but typically the method offers relatively low sensitivity.
    5. Cells treated with a reagent that causes light emission are read in an opaque white microplate, using a photoluminescence reader. The reagents are introduced after incubation, so do not influence cell migration. The method is not extensively documented in the literature, but appears to be quite sensitive.

Two Design Strategies

  1. The Small-Pore Strategy Suppose migrated cells will be counted using an ELISA microplate reader or other non-sensitive detection system. Large numbers of cells are called for, so this strategy maximizes cell density at each site: the cell suspension will contain enough cells to completely cover the filter surface (exposed filter area) at each site in a monolayer. (To calculate the number of cells needed one must know the cell size as described above. The graph below can be used to determine the number of cells needed to cover a square millimeter.) It is also advantageous to increase the number of cells at each site by choosing an instrument that maximizes the exposed filter area.

    Cell-size/Cell-density Graph

    With the entire filter area covered with cells, there will be a cell directly over every pore in the filter, which will increase the incidence of random migration through the filter. To minimize such migration, the size and nucleus characteristics of the cells are taken into account in order to select a filter pore size that will physically inhibit the unstimulated cells at negative control sites, but will not unduly slow down chemotacticly stimulated cells. If the pore size inhibits negative control cells enough, and the stimulated cells not too much, the differential cell count will be large.

    Clearly this strategy depends on how much is known about the size and deformability of the cell nuclei, and how the latter is influenced by the chemotactic factor. It also depends on availability of filters with the indicated pore size.

    With a cell over every pore, chemokinetic effects may be significant, and chemokinetic control sites are recommended. See controls.

  2. The Low-Pore-Density Strategy Suppose a fluorescence reader or other very sensitive detection system will be used to count migrated cells. Low numbers of migrated cells will give good results, so a sparse distribution of cells on the filter surface can suffice. With a sparse distribution of cells, most cells will not be over a pore and must travel to a pore prior to migrating through the membrane filter. At unknown sites a concentration gradient of chemotactic factor emanating from the pores will cause cells to do this, but cells at negative control sites will arrive at pores only as a result of random migration and are thus much less likely to move through the filter.In this strategy the optimum number of cells per mm2 of filter area is usually the same as the number of pores per mm2. For example, referring to the table above, an 8 micron filter has 1,000 pores per mm2 and these pores occupy only about 5% of the surface area of the filter. The optimum density of cells on this filter would be 1,000 cells per mm2 of filter area, and only one in 20 cells is likely to settle over a pore. At negative control sites 95% of the cells on the filter would be highly unlikely to arrive at pores at all, and the 5% that started out over pores would be as likely to migrate away from the pore as through it.In such an assay the differential cell counts between unknown and negative-control sites will be significant even though the pores are quite large, as long as they are few and far between.


ChemoTx® is a registered trademark of Neuro Probe, Inc.