Biomedical Engineering Reference
In-Depth Information
has been used and shows a stencil mask containing ~100 × 150 μm squares and 10 to 20 μm lines
(
Figure 2.23
), suggesting that the features were made by photochemical machining. Foreseeing
the feasibility of massively parallel single-cell assays, he studied the spreading of single cells
conined to adhesive islands.
Carter's idea of patterning with a removable (and reusable) stencil mask is attractive for
its conceptual simplicity, compatibility with a variety of surfaces, low toxicity, and low cost.
However, metallic and oxide substrates, by virtue of being chemically foreign to in vivo cellular
environments, might arguably trigger unexpected, nonphysiological cell responses. Also, metal
stencils do not seal with the substrate, thus luids and cells might not obey the pattern as with
other stencil materials.
Carter's technique was so ahead of his times that it was used essentially unaltered by others
for more than twenty years. Here are some highlights:
1973: Albert Harris from Yale University used the same shadow-evaporation concept
in combination with iner stencil masks—transmission electron microscope (TEM)
grids (33-50 μm lines/square). He applied it to a variety of cell lines and substrates
and noticed that there is a materials hierarchy for cell adhesion, for example, cells
attached to palladium (Pd) on Pd/acetate patterns but attached to the polystyrene
background on Pd/polystyrene patterns.
1975: In his investigation of axon guidance in chicken embryo neuron cultures,
Paul Letourneau, then at Stanford University, studied the hierarchical preference of
growth cones for a set of surfaces and found that growth cones showed a poor prefer-
ence for the tissue culture plastic background over the Pd patterns but a strong prefer-
ence for the background when the background had been coated with polyornithine
or collagen before shadow-evaporation of Pd; this article marked an inlection in the
ield, for it became clear that natural proteins could be used as adhesive substrates (see
Figure 2.26
in Section 2.6). Interestingly, in glial cell cocultures, the growth cone-to-
polyornithine adhesion was roughly as strong as adhesion to glia.
1976: Guenter Albrecht-Buehler from the Cold Spring Harbor Laboratory in New York
used gold islands shadow-evaporated onto glass (through TEM copper grids) to study
the exploratory function of ilopodia during spreading of 3T3 cells. He observed that
cells did not extend lamellipodia unless some ilopodia, in a rapid scanning motion,
had contacted a gold area; 1 to 2 hours ater seeding, most of the cells had migrated to
gold-coated areas. his suggested that the directional preference of 3T3 cells during
migration is a response to the substrate adhesiveness as sensed by the ilopodia. He
also noticed that, 24 hours ater plating, the cells did not display any preference for
gold versus glass, which can be attributed to secretion of adhesive factors.
1978: Following Harris' studies, Bengt Westermark at the University of Uppsala
in Sweden further improved the attachment selectivity by shadow-evaporating Pd
onto a thin layer of air-dried agarose. When seeded on such surfaces, the attachment
and spreading of glial cells was entirely conined to the Pd islands, and their growth
stopped at conluence even though they had a free edge bordering the agarose “sea.”
1980: Jan Pontén and Lars Stolt, also at the University of Uppsala, adopted the same
Pd/agarose system as Westermark, but used a 14-μm-thick titanium mask with arbi-
trary patterns microfabricated by photolithography and chemical etch for the shadow
evaporation step. hey conirmed Westermark's indings on a variety of cell lines
down to single-cell, 26-μm-diameter islands, suggesting that cell contacts are not
necessary for the production of a quiescent growth state.
1986: Grenham Ireland and colleagues at the University of Manchester in England cre-
ated adhesive Pd islands onto substrates previously coated with poly-2-hydroxyethyl
