DIFFERENTIATION OF KIDNEY MESENCHYME
283
their subsequent development into tubules proceeds autonomously. Support for the view that the aggregation and probably the early maintenance of the aggregates is dependent on heterotypic induction has been
presented by Auerbach (1960b). If the early tubules are dissociated,
their recombination requires the presence of inductor. This finding indicates that although homotypic interaction is sufficient for the continuation of tubulogenesis at this stage, heterotypic interaction is still required
for reconstitution of the environmental state in which aggregation of
tubule cells occurs.
If the responding cell population in kidney mesenchyme is completely
homogeneous, what then determines which cells will be converted into
tubule cells and which will be left out of the tubules?
Histological observations and time-lapse cinematography have shed
light on this question. In the inductive action exerted by the spinal cord,
two characteristics are of importance. The process is time-consuming and
there seems to be a quantitative gradient, the effect being greatest at the
mesenchyme-inductor interface and decreasing toward the periphery of
the explant. The gradient effect is expressed in the sequence of the appearance of the condensates, first in the neighbourhood of the inductor
and later in the more remote parts of the explant. The induction time in
transfilter experiments can be divided into two periods. The first of these
is the time needed for transfer of the stimulus to the responding cells and
the second the time needed for the completion and expression of induction. The time-dependence of the transfer of the stimulus was also supported by the fact that the morphologically silent period could be shortened to less than half of the original time by preincubating the
filter-spinal cord combination (Section I,A), and repeated observations
indicate that in direct spinal cord-mesenchyme combinations without an
interposed filter, condensates form in an even shorter period of time.
When the cells at a given time have received a quantitatively sufficient
stimulus, they start forming condensates.
Time-lapse cinematography and electron microscopic analysis of the
formation of condensates (see Section II) show that loose mesenchymal
cells are trapped in the condensates, which soon form dense spherical cell
masses seemingly not accepting new cells seeking entrance. The immediate change in the cells participating in the condensates is their diminished
free motion. The cells which are left out of the condensates may not be
metabolically synchronized to mutual recognition and adhesion; it is
known, for example, that during the mitotic cycle the characteristics of
the cell periphery are altered (Mayhew, 1966). In the later stages of
283
their subsequent development into tubules proceeds autonomously. Support for the view that the aggregation and probably the early maintenance of the aggregates is dependent on heterotypic induction has been
presented by Auerbach (1960b). If the early tubules are dissociated,
their recombination requires the presence of inductor. This finding indicates that although homotypic interaction is sufficient for the continuation of tubulogenesis at this stage, heterotypic interaction is still required
for reconstitution of the environmental state in which aggregation of
tubule cells occurs.
If the responding cell population in kidney mesenchyme is completely
homogeneous, what then determines which cells will be converted into
tubule cells and which will be left out of the tubules?
Histological observations and time-lapse cinematography have shed
light on this question. In the inductive action exerted by the spinal cord,
two characteristics are of importance. The process is time-consuming and
there seems to be a quantitative gradient, the effect being greatest at the
mesenchyme-inductor interface and decreasing toward the periphery of
the explant. The gradient effect is expressed in the sequence of the appearance of the condensates, first in the neighbourhood of the inductor
and later in the more remote parts of the explant. The induction time in
transfilter experiments can be divided into two periods. The first of these
is the time needed for transfer of the stimulus to the responding cells and
the second the time needed for the completion and expression of induction. The time-dependence of the transfer of the stimulus was also supported by the fact that the morphologically silent period could be shortened to less than half of the original time by preincubating the
filter-spinal cord combination (Section I,A), and repeated observations
indicate that in direct spinal cord-mesenchyme combinations without an
interposed filter, condensates form in an even shorter period of time.
When the cells at a given time have received a quantitatively sufficient
stimulus, they start forming condensates.
Time-lapse cinematography and electron microscopic analysis of the
formation of condensates (see Section II) show that loose mesenchymal
cells are trapped in the condensates, which soon form dense spherical cell
masses seemingly not accepting new cells seeking entrance. The immediate change in the cells participating in the condensates is their diminished
free motion. The cells which are left out of the condensates may not be
metabolically synchronized to mutual recognition and adhesion; it is
known, for example, that during the mitotic cycle the characteristics of
the cell periphery are altered (Mayhew, 1966). In the later stages of