Cell Migration in 3D Matrices 177
and to facilitate excision of the tendons. In spite of carefully controlling all stages
of the protocol, there are differences amongst batches of collagen. Such differ-
ences are more apparent with some cell types than with others (14). Certain
batches appear to have toxic effects on ECs but not on fibroblasts or pericytes.
We have observed a similar, cell-specific, toxic effect for some batches of serum
and shown it to be caused by the presence of active TGF`-1 (see Note 4). The
reason for differences amongst batches prepared in the same way is not clear. It
may be due to variations in collagen crosslinking or even to the presence of
active TGF`-1. It is therefore important to test in a pilot experiment the collagen
batch to be used. For ECs it is better to test several batches before carrying out
the assay and retest the chosen batch if it has been stored for more than 3 mo.
Type I collagen preparations (bovine dermal and rat tail) suitable to make gels
for tissue culture are commercially available (Collagen Bio-Materials, Becton &
Dickinson, Sigma). We cannot comment on these, as we make our own collagen
preparations.
2. Analar or Aristar grade acetic acid is essential for the preparation of collagen; if
inferior grades are used, the collagen may set erratically or precipitate during
dialysis.
3. The concentration of the collagen working solution may vary from 2.0–2.6 mg/mL,
as long as the final concentration in the gels used for the assay is constant
(1.7–2.2 mg/mL). This is the concentration that allows maximal migration of
various cell types tested (14).
4. Either donor or fetal calf sera may be used, but batches should be tested for
growth-promoting properties on the ECs, even if they are to be used for the cul-
ture of fibroblasts or pericytes. Great variations may be found among batches of
sera, irrespective of their origin. We have found that some batches contain active
TGF`-1; these should not be used. However, TGF`-1 becomes activated by pro-
longed storage (e.g., 6 mo at –20°C), therefore, batches should be retested when
stored for more than 3–4 mo. Serum-free medium (containing growth factors) is
commercially available (e.g., TCS Biologicals, Buckingham, UK). If required, it
should be possible to use this, instead of serum-supplemented medium.
5. The use of ascorbic acid as an additive to the cell growth medium is also
optional, but it should be noted that the composition of the matrix synthesized by
the cells will be affected. We routinely add fresh ascorbic acid to endothelial and
pericyte medium, but not to that for fibroblasts. When used on the latter, it
becomes necessary to use collagenase, as well as trypsin, to bring the cells into
suspension.
6. Several types of inert markers are commercially available (e.g., New England
Nuclear Life Science Products; non-radioactive microspheres of 10-, 15-, 25-, or
50-µm diameter; Sigma, gel filtration cellulose beads from 30–50- to 250–500-µm
diameter; International Market Supplies dry ink, various colours). The prepara-
tion of the working suspension may vary accordingly, as long as the aim is
achieved. Namely, the particles should be sterile, immobile, easily visualized
and nonreactive (with the gel, the cells, or the medium). The use of a marker is