Transfilter Cell Invasion Assays
The ability to invade surrounding extracellular
matrices and tissues is an important phenotype of
malignant tumor cells. To determine the invasive
properties of malignant cells, several invasion assays
have been developed. For example, organ fragments
(Nicolson et al.
, 1985), reconstituted tissue spheroids
(Mareel et al.
, 1988), membranous tissues (Nabeshima et al.
, 1988; Yagel et al.
, 1989), cultured cell monolayers
(Kramer and Nicolson, 1979; Waller et al.
, 1986), or
extracellular matrices (Albini et al.
, 1987; Schor et al.
1980) have been used as tissue or matrix for cell invasion
studies. Commercially available Matrigel, a mouse
EHS tumor extract consisting of major basement membrane
components polymerized into a gel, has been the
most commonly used material to determine the invasiveness
of various types of cells (Albini et al.
The apparatus of choice for measuring cell invasion has
been a modified Boyden chamber or Transwell. Briefly,
the apparatus contains two chambers (upper and
lower) that are separated by a microporous polycarbonate
filter, the upper surface of which is coated with
a thin layer of Matrigel. Tumor cells are placed into the
upper chamber where they settle by gravity onto the
Matrigel layer. In the lower chamber a chemotactic
agent can be placed to stimulate directional cell
migration of cells that invade the Matrigel layer. The
invasive ability of these cells can be expressed as the
number of cells invading through the Matrigel layer
and filter with time. For the most part, the invading
cells are found on the lower surface of the filter and
are not released into the fluid of the lower chamber.
Although there are some exceptions to this (Simon et al.
, 1992), the invasive abilities of tumor cells in the
transfilter invasion assay are usually related to their in vivo
invasion behavior (Albini et al.
, 1987; Repesh,
1989; Hendrix et al.
, 1987). This article describes the
basic protocol of the Matrigel transfilter invasion assay.
II. MATERIALS AND
Culture medium DME/FI2 (Cat. No. 11330-032),
fetal bovine serum (FBS), phosphate-buffered saline
(PBS) (Cat. No. 14040-133), and trypsin (2.5%, Cat No.
15090-046) are from GIBCO-BRL. Matrigel (Cat. No.
40234) and fibronectin (Cat. No. 40008) are from
Collaborative Biomedical Products. Bovine serum
albumin (BSA) (Cat. No. 810661, fraction V) is from
ICN. EDTA (Cat. No. 423-384) is from CMS. Hematoxylin
(Cat. No. GHS-1-16), eosin (Cat. No. HTIIO-3-
16), and 10% neutralized formaldehyde solution (Cat.
No. HT50-1-128) are from Sigma.
Transwells (Cat. No. 3421, 6.5-mm diameter, 5-µm
pore size) and 24-well culture plates (Cat. No. 25820)
are from Corning Costar Corporation. Additional
equipment includes general tissue culture supply and
equipment, forceps, cotton swabs, a reticle (1/10-mm
measurement), and a cell-counting device (Coulter
counter, hemocytometer, or equivalent equipment).
A. Matrigel Coating on the Microporous
1:30 diluted Matrigel solution (sterile)
: To make 900 µl
Matrigel solution for 12 Transwells (50 µl X the number of Transwells), thaw Matrigel at 4°C overnight and mix
30µl with 870µl of ice-cold PBS. Keep ice cold until
the solution is applied to the Transwell. Aliquot the
remaining Matrigel (e.g., 1 ml each), freeze, and store
1:100 diluted Matrigel solution (sterile)
: To make 2 ml,
mix 20µl of Matrigel with 2ml of ice-cold PBS. Keep
|FIGURE 1 Transwell invasion chamber system: (1) lid,
(2) Transwell, (3) 24-well culture plate, (4) lower
of 10µg/ml fibronectin in invasion buffer),
(5) upper chamber
(200µl), (6) Matrigel layer, (7)
microporous polycarbonate filter
(5-µm pore size), and
(8) lower surface of the filter coated with a thin
layer of Matrigel.
B. Setting up the Invasion Assay
2% BSA stock solution
- Place 500 µl of 1: 100 diluted Matrigel solution in
a small culture dish (i.e., 30mm diameter) on an icecold
- Use a forceps for the handling of Transwells.
The Transwell invasion chamber system is shown in
Fig. 1. Soak the lower surface of the polycarbonate
filter of each Transwell in the Matrigel solution. After
briefly removing any excessive amount of the solution,
place the Transwell into a 24-well culture plate
and allow it to dry in a hood overnight at room
- Pour 50µl of the 1:30 diluted Matrigel solution
into the upper chamber of the Transwell.
- Carefully overlay 200µl of sterilized, doubledistilled
H2O to each filter and allow it to dry completely
in a hood at room temperature under occasional
ultraviolet light. This usually takes about 2 or 3
- Proceed to the next step or store the coated Transwell
sets in a scaled 24-well culture plate at 4°C.
: To make 100ml, slowly dissolve
2 g of BSA in ice-cold PBS for 20min and sterilize
by filtration through a 0.22-µm filter.
Invasion buffer, 0.1% BSA in DME/F12
: To make
40ml, mix 2ml of 2% BSA stock solution with 38 ml of
EDTA stock solution (200 mM)
: To make 100 ml, dissolve
7.44g EDTA with double-distilled H2
adjust the pH to 7.4.
Trypsin-EDTA (0.25%): To make 100 ml, add 10ml of
2.5% trypsin solution and 1 ml of EDTA stock solution
into 89ml of PBS.
: To make 20 ml, mix 2 ml of FBS with
18 ml of DME/FI2 medium.
Fibronectin solution (10µg/ml)
: To make 10ml, mix
100µl of fibronectin stock solution (1 mg/ml) with 10ml
of invasion buffer. In the example here, fibronectin is
used as a chemoattractant.
C. Staining Invading Cells with
- These procedures should be performed under
sterile conditions. Culture (4 × 100-mm-diameter
dishes) B16-F10 cells in DME/FI2 medium supplemented
with 5% FBS at 37°C in an atmosphere of 5%
CO2-95% air and grow to about 80% confluency.
- Prewarm the Matrigel-coated Transwells in the
24-well culture plate to room temperature, and rehydrate
the Matrigel with 200 µl of invasion buffer for 1 h.
- Wash the B16-F10 culture dishes with 10 ml of
DME/FI2 medium and incubate with 3ml each of
trypsin-EDTA for 5 min. Place in a 50-ml tube and sediment
cells by centifugation at 1000rpm for 5 min.
- Wash the cells with 10ml of 10% FBS-DME
medium by centrifugation at 1000rpm for 5 min and
resuspend the cells in 10 ml of 10% FBS-DME medium.
Let the cells incubate for 15 min at room temperature.
- Wash the cells with 40ml of DME/FI2 medium
twice by centrifugation at 1000rpm for 10min. Resuspend
the cells in 5 ml of invasion buffer.
- Prepare 2.5 × 105 cells/ml in invasion buffer by
counting the cells with a Coulter counter, hemocytometer,
or equivalent equipment.
- Carefully discard the buffer in the upper chambers
of the Transwells, wash the chambers again
with 200 µl of invasion buffer, and carefully pour 800 µl
of the fibronectin solution (chemoattractant solution)
through the slit of the Transwell into the lower
- Add 200µl of the cell suspension into the upper
chamber. Avoid making bubbles in the upper chamber so that the cells can settle down evenly on the filter.
Carefully put the Transwell invasion chamber system
into an incubator and incubate for 48h at 37°C in 5%
D. Counting Invading Cells Using a Phase-
Contrast Microscope (or Coulter Counter)
- Prepare a 24-well culture plate with 400 µl of 10%
formaldehyde, hematoxylin, or eosin in each well and
fill four jars (about 500ml each) with distilled water.
- Place the Transwell in the 10% formadehyde
solution for 10min and then rinse the Transwell by
submerging it in a jar containing distilled water.
- Transfer the Transwell to the hematoxylin solution
for 10min and then rinse with water. Place the
Transwell into the well with warm distilled water
(about 40°C) for 10 min.
- Transfer the Transwell to the eosin well for 5 min
and then place it into a jar containing distilled water.
Gently rub the cells off the upper side of the filter using
a cotton swab and rinse the Transwell with distilled
1. Place the Transwell into a 24-well culture place.
Make sure that there is moisture on the inside of the
culture plate. Count the cells in the 16 fields indicated
in Fig. 2 at 200 × magnification.
2. Using a reticle with 1/10-mm guides (Bausch &
Comb Inc.), place the cell suspension onto the stage of
a phase-contrast microscope, and measure the diameter
of the field (r
; mm) with the magnification at 40×.
The total number of invading cells (1) can be calculated
with the formula
is the average number of invading cells per
×200 magnification field and R is the diameter of the
filter (6.5 mm). For example, the value of r
of a Diaphot
(Nikon) is 4.45. Therefore, (5 × R/r
= 53.3. If the cells
are too dense to count the entire field, a photo as
shown in Fig. 2 can be used. Using a Diaphot (Nikon),
the areas of a, b, and c in Fig. 2 are 47, 17, and 4.3% of
the total area of the field, respectively.
E. Reusing Transwells
|FIGURE 2 A Transwell filter in a microscopic field. (A) Spots
used to count the cells on the filter are shown as numbered circles.
The numbers of the spots are an example of the sequence recommended
for cell counting. (B) Photo frames of the Diaphot (Nikon).
Relative sizes of the frames to the field are 47 (a), 17 (b), and 4.3%
- After counting invading cells, wipe the cells off
the lower surface of the filter using a cotton swab and
soak the Transwells in a detergent solution (e.g., 1%
Contrad 70) for at least 1 day.
- Wipe both surfaces of the filter gently and thoroughly
using a cotton swab and place the Transwells
into the 24-well culture plate.
- Fill the entire Transwell plate with distilled
water, cover the plate with a lid, shake it several times,
and discard the water.
- Repeat step 3 at least three times. After the final
wash, let the Transwell plate sit for 1 h filled with distilled
- Wash the plate with double-distilled water
several times and rinse with 70% ethanol as in step 3.
Dry the Transwells in a tissue culture hood overnight
under ultraviolet light.
Although tumor cells in vivo
usually invade connective
or parenchymal tissue, a number of reports
indicate that the invasive and metastatic phenotypes
of cells correlate with their in vitro
capacity (Albini et al.
, 1987; Repesh, 1989; Hen&ix et al.
, 1987). This may be due to the fact that invasion in vivo
into connective or parenchymal tissue is limited
by the capability of tumor cells to undergo basement
membrane invasion, a key event in malignant progression.
The invasive process involves at least three
events (adhesion, degradation, and migration) (Liotta, 1986); for tumor cells to be invasive, they must be at
least somewhat proficient in each step (Nicolson,
The incubation time for the invasion assay is usually
24 to 72h. We feel that shorter incubation times are
better as long as a significant number of cells invade
the filter. Using longer incubation times, other parameters,
such as the growth of the cells on the Matrigel,
loss of cell viability, or the detachment of cells from the
underside of the filter, must be taken into account. The
incubation time can be varied by changing the amount
of Matrigel used in coating the filter, usually in the
range of 5 to 50µg (or 10µg in our case).
The very thin Matrigel coating on the lower surface
of the filter should be completely dried before adding
the 1:30 Matrigel solution in the upper chamber. This
will avoid leaks in the filter.
The overlay of 200 µl of double-distilled H2
O on the
Matrigel solution is necessary for an even coating of
Matrigel on the filer.
The number of cells applied onto the filter of the
Transwell should not exceed the number that will yield
100% confluency on the filter. Usually the cells are
50-80% confluent on the filter after settling.
The cell suspension volume in the upper chamber
should be 200µl instead of 100µl to reduce the meniscus
effect that can cause uneven settling of the cells
onto the Matrigel layer. Thus the volume of the lower
chamber should be increased to 800 btl (from the usual
The treatment with 10% FBS medium after harvesting
the cells by trypsinization is necessary for the
quantitative recovery of the cells.
Several other methods can be used to count the
invading cells. For example, cells can contain fluorogenic
substrates (Garrido et al.
, 1995), radioisotopes
(Garrido et al.
, 1995; Muir et al.
, 1993), or dyes
(Imamura et al.
, 1994). These labels are suitable when
dealing with a large number of samples; however,
visual cell counting is still necessary for optimizing
the methods or for checking reproducibility of the
methods. When the number of invading cells is large
enough (>1000), the cells can be counted using a
Coulter counter after collecting the invading cells by
trypsinization of the filter.
Transwells are normally reusable four or five times
or until their polycarbonate filters become cracked.
Albini, A., Iwamoto, Y., Kleinman, H. K., Martin, C. R., Aaronson,
S. A., Kozlowski, J. M., and McEwan, R. N. (1987). A rapid in vitro
assay for quantieating the invasive potential of tumor cells. Cancer Res
Garrido, T., Riese, H. H., Quesada, A. R., Barbacid, M. M., and Aracil,
M. (1995). Quantitative assay for cell invasion using the fluorogenic
substrate 2'. 7'-bis(2-carboxyethyl)-5(and-6)-carboxyfluorescein
acetoxymethylester. Anal. Biochem
Hendrix, M. J. C., Seftor, E. A., Seftor, R. E. B., and Fidler, L. J. (1987).
A simple quantitative assay for studying the invasive potential
of high and low human metastatic variants. Cancer Lett
Imamura, H., Takao, S., and Aikou, T. (1994). A modified invasion
assay for quantitating tumor cell invasion. Cancer Res
Kramer R. H., and Nicolson, C. L. (1979). Interactions of tumor cells
with vascular endothelial cell monolayers: A model for metastatic
invasion. Proc. Natl. Acad. Sci. USA 76
Liotta, L A. (1986). Tumor invasion and metastatic-role of extracellular
matrix: Rhoads memorial award lecture. Cancer Res
Mareel, M., Dragonetti, C., Travernier J., and Fiers, W. (1988). Tumorselective
cytotoxic effects of murine necrosis factor (TNF) and
interferon-gamma in organ culture of BIG melanoma cells and
heart tissue. Int. J. Cancer 42
Muir, D., Sukhu, L., Johnson, J., Lahorra, M. A., and Maria, B. L.
(1993). Quantitative methods for scoring cell migration and invasion
in filter-based assays. Anal. Biochem
Nabeshirna, K., Kataoka, H., Koita, H., Murayama, T., and Koono,
M. (1988). A new long-term in vitro
invasion assay using fibrous
connective tissue matrices maintaining architectural characteristics
of connective tissue. Invas. Metast
Nicolson, G. L. (1989). Metastatic tumor cell interactions with
endothelium, basement membrane and tissue. Curr. Opin. Cell
Nicolson, G. L., Dulski, K., Basson, C., and Welch, D. R. (1985). Preferential
organ attachment and invasion in vitro
by B16 melanoma
cells selected for differing metastatic colonization and invasive
properties. Invas. Metast
Repesh, L. A. (1989). A new in vitro
assay for quantitating tumor cell
invasion. Invas. Metast
Schor, S. L., Allen, T. D., and Harrison, C. J. (1980). Cell migration
through three-dimensional gels of native collagen fibres: Collagenolytic
activity is not required for the migration of two permanent
cell lines. J. Cell Sci
Simon, N., N6el, A., and Foidart, J.-M. (1992). Evaluation of in vitro
reconstituted basement membrane assay to assess the invasiveness
of tumor cells. Invas. Metast
Waller, C. A., Braun, M., and Schirrrnacher, V. (1986). Quantitative
analysis of cancer invasion in vitro
: Comparison of two new
assays and of tumour sublines with different metastatic capacity. Clin. Exp. Metast
Yagel, S., Khokha, R., Denhardt, D. T., Kerbel, R. S., Parhar, R. S., and
Lala, P. K. (1989). Mechanisms of cellular invasiveness: A comparison
of amnion invasion in vitro
and metastatic behavior in
vivo. JNCI 81