472
C L I F F O R D G R O B S T E I N
tion. If the piece is cut into fragments of critical size, however, the
degree of dispersion of the fragments proves decisive in controlling
differentiation. Fragments cultured in a close cluster, so as quickly to
reconstitute the original mass, show high frequency of neural differentiation on subsequent transfer to the chorio-allantoic membrane. Identical
fragments cultured in a dispersed arrangement show a sharply reduced
frequency of neural differentiation. The suggestion is strong that the
initiation, or the progression, or both, of neural differentiation in the
blastoderm cells requires interactions among the cells, and that these
are protected and continue in explants above a critical mass, but are
disrupted and inhibited in explants below this critical mass.
Zwilling (1960) has supplied additional data bearing on the point. He
found a difference in the behavior of dissociated and reaggregated cells
obtained from definitive streak (DPS) and later-stage head-fold to
early somite (HF) blastoderms. DPS reaggregates tended to spread
rapidly in culture and underwent little or no neural differentiation,
either in culture or in chorio-allantoic grafts. HF reaggregates remained cohesive and massive in culture and produced, among other
things, considerable neural tissue. HF reaggregates of smaller size,
however, showed lower frequency of neural differentiation, and reaggregates of both smaller and larger size were reduced in frequency of
neural differentiation when they were incorporated in culture into
DPS reaggregates. Zwilling interprets the latter result as due to the
altered spreading behavior of the HF reaggregates when incorporated
in the more rapidly spreading DPS tissue. He sees his results as supporting the conception of a liminal cell number or tissue mass for particular
differentiations.
Interpretation of data on the behavior of developmentally plastic
blastoderm cells is complicated somewhat by the possibility that the
rise of new cell types results in heterotypic interactions which are
involved in subsequent differentiation. It is significant, therefore, that
comparable observations have been made on later stage cells. Moscona
(1956) has described the culture behavior of intermixed suspensions of
chick mesonephros and chondrogenic mesenchyme. The intermixed cells
sort out to form aggregates which differentiate characteristically in
accordance with their tissue of origin. When the ratio of the two cell
types in the intermixed cell suspension is varied a minimum or liminal
proportion is found below which the minority type is unrepresented as
a tissue. The same absolute number of cells, aggregated homotypically,
undergoes typical differentiation. This suggests that the minority cell
type has been diluted beyond the point at which homotypic aggregates
of a minimal size necessary for characteristic differentiation are formed.
Recent studies of mouse pancreatic acinar differentiation (Grobstein
C L I F F O R D G R O B S T E I N
tion. If the piece is cut into fragments of critical size, however, the
degree of dispersion of the fragments proves decisive in controlling
differentiation. Fragments cultured in a close cluster, so as quickly to
reconstitute the original mass, show high frequency of neural differentiation on subsequent transfer to the chorio-allantoic membrane. Identical
fragments cultured in a dispersed arrangement show a sharply reduced
frequency of neural differentiation. The suggestion is strong that the
initiation, or the progression, or both, of neural differentiation in the
blastoderm cells requires interactions among the cells, and that these
are protected and continue in explants above a critical mass, but are
disrupted and inhibited in explants below this critical mass.
Zwilling (1960) has supplied additional data bearing on the point. He
found a difference in the behavior of dissociated and reaggregated cells
obtained from definitive streak (DPS) and later-stage head-fold to
early somite (HF) blastoderms. DPS reaggregates tended to spread
rapidly in culture and underwent little or no neural differentiation,
either in culture or in chorio-allantoic grafts. HF reaggregates remained cohesive and massive in culture and produced, among other
things, considerable neural tissue. HF reaggregates of smaller size,
however, showed lower frequency of neural differentiation, and reaggregates of both smaller and larger size were reduced in frequency of
neural differentiation when they were incorporated in culture into
DPS reaggregates. Zwilling interprets the latter result as due to the
altered spreading behavior of the HF reaggregates when incorporated
in the more rapidly spreading DPS tissue. He sees his results as supporting the conception of a liminal cell number or tissue mass for particular
differentiations.
Interpretation of data on the behavior of developmentally plastic
blastoderm cells is complicated somewhat by the possibility that the
rise of new cell types results in heterotypic interactions which are
involved in subsequent differentiation. It is significant, therefore, that
comparable observations have been made on later stage cells. Moscona
(1956) has described the culture behavior of intermixed suspensions of
chick mesonephros and chondrogenic mesenchyme. The intermixed cells
sort out to form aggregates which differentiate characteristically in
accordance with their tissue of origin. When the ratio of the two cell
types in the intermixed cell suspension is varied a minimum or liminal
proportion is found below which the minority type is unrepresented as
a tissue. The same absolute number of cells, aggregated homotypically,
undergoes typical differentiation. This suggests that the minority cell
type has been diluted beyond the point at which homotypic aggregates
of a minimal size necessary for characteristic differentiation are formed.
Recent studies of mouse pancreatic acinar differentiation (Grobstein