480
CLIFFORD GROBSTEIN
goes mucous metaplasia, and even ciliation, instead. Epidermis on heart
myoblasts becomes a single-layered, squamous epithelium. McLoughlin
speculates, in conformity with suggestions of other investigators (Weiss,
1945; Grobstein, 1955c, 1961a; Moscona, 1960; Wilde, 1961) that
intercellular mucoid material—and particularly basement membranes
—may play an important role in controlling epidermal differentiation.
Additional and concordant facts are available for systems other than
epithelio-mesenchymal ones which also illustrate heterotypic requirements for differentiation in vitro. One of these is metanephrogenic
mesenchyme of the mouse embryo, which has been shown not to continue its differentiation in vitro to epithelial tubules and glomeruli
except in the presence of a second tissue (Grobstein, 1955b). The
normal inductive source is ureteric bud, but embryonic dorsal spinal
cord, salivary epithelium, and to a lesser extent, pancreatic epithelium
(Golosow and Grobstein, 1962) also have proved effective. The dorsal
spinal cord effect has been analyzed most extensively. Like the salivary
epithelio-mesenchymal interaction the spinal cord effect can be exerted
across a membrane filter (Grobstein, 1956). By increasing the thickness,
or number of layers of the filter the transmission distance of the effect
has been measured (Grobstein, 1957). It is strong in the first 30/x,
declines rapidly in the next 20-30/x, and is undetectable beyond
60-80/x. Transmission is blocked by cellophane, but occurs through a
hole in the cellophane to produce a localized response over the hole.
Efforts to collect activity, or to demonstrate it in solution, so far have
been unsuccessful. A material can be detected in the filter, however,
following fixation with alcohol-formalin, which corresponds in distribution with the biological activity. It is digestible by trypsin, and certain
of its components are labelled when dorsal spinal cord is exposed to
tritiated amino acids before culturing (Grobstein, 1959b). Grain counts
over the filter made in autoradiograms at increasing distance from the
labelled spinal cord give a curve which corresponds reasonably closely
to that for biological activity, and no fixable label passes across cellophane (Grobstein, 1961b). Fixable label does, however, pass through a
hole in cellophane, is demonstrable first in the mesenchyme over it, and
spreads laterally in the mesenchyme from this point (Koch, 1961). The
whole body of information is consistent with the hypothesis that a largemolecular material including protein, normally associated closely with
cell surfaces but capable of limited mobility beyond them in active
form, promotes the differentiation of metanephrogenic mesenchyme
into characteristic tubules in vitro.
Another differentiating system requiring a heterotypic contribution
in vitro is somite mesenchyme from appropriate stages of mouse
CLIFFORD GROBSTEIN
goes mucous metaplasia, and even ciliation, instead. Epidermis on heart
myoblasts becomes a single-layered, squamous epithelium. McLoughlin
speculates, in conformity with suggestions of other investigators (Weiss,
1945; Grobstein, 1955c, 1961a; Moscona, 1960; Wilde, 1961) that
intercellular mucoid material—and particularly basement membranes
—may play an important role in controlling epidermal differentiation.
Additional and concordant facts are available for systems other than
epithelio-mesenchymal ones which also illustrate heterotypic requirements for differentiation in vitro. One of these is metanephrogenic
mesenchyme of the mouse embryo, which has been shown not to continue its differentiation in vitro to epithelial tubules and glomeruli
except in the presence of a second tissue (Grobstein, 1955b). The
normal inductive source is ureteric bud, but embryonic dorsal spinal
cord, salivary epithelium, and to a lesser extent, pancreatic epithelium
(Golosow and Grobstein, 1962) also have proved effective. The dorsal
spinal cord effect has been analyzed most extensively. Like the salivary
epithelio-mesenchymal interaction the spinal cord effect can be exerted
across a membrane filter (Grobstein, 1956). By increasing the thickness,
or number of layers of the filter the transmission distance of the effect
has been measured (Grobstein, 1957). It is strong in the first 30/x,
declines rapidly in the next 20-30/x, and is undetectable beyond
60-80/x. Transmission is blocked by cellophane, but occurs through a
hole in the cellophane to produce a localized response over the hole.
Efforts to collect activity, or to demonstrate it in solution, so far have
been unsuccessful. A material can be detected in the filter, however,
following fixation with alcohol-formalin, which corresponds in distribution with the biological activity. It is digestible by trypsin, and certain
of its components are labelled when dorsal spinal cord is exposed to
tritiated amino acids before culturing (Grobstein, 1959b). Grain counts
over the filter made in autoradiograms at increasing distance from the
labelled spinal cord give a curve which corresponds reasonably closely
to that for biological activity, and no fixable label passes across cellophane (Grobstein, 1961b). Fixable label does, however, pass through a
hole in cellophane, is demonstrable first in the mesenchyme over it, and
spreads laterally in the mesenchyme from this point (Koch, 1961). The
whole body of information is consistent with the hypothesis that a largemolecular material including protein, normally associated closely with
cell surfaces but capable of limited mobility beyond them in active
form, promotes the differentiation of metanephrogenic mesenchyme
into characteristic tubules in vitro.
Another differentiating system requiring a heterotypic contribution
in vitro is somite mesenchyme from appropriate stages of mouse