124
13. M. SHAFFER
posed that diffusible material from the mutants induced initiator cells
in the wild strain. This provided a complementary or alternative explanation of why different numbers of wild cells should have acted as
initiators when mixed with the different mutants. These mutants were
actually described as 'audiences'; but when 'audiences' participate to
this extent, it becomes rather difficult to distinguish the 'performers'.
Mixtures of wild strain with fruity. Having found that the mutant
fruity when developing alone at optimal density produced 100 times
more centres than the wild strain, Sussman (1955a, 1958) assumed that
the behaviour of mixtures of fruity and wild cells was merely additive
(though his earlier work (1952, 1954) made this unlikely), and accordingly subtracted from the total number of centres those that would
have been formed by the wild component when alone, attributed the
rest of them to the fruity, and then concluded that each of these was
started by a single mutant initiator cell. He found also that even
washings of fruity cells stimulated pure wild cells to yield many more
centres.
Comment. There was thus a radical difference in the interpretation of
the very similar results obtained with different mutants. If the increase
in centres when wild was mixed with aggregateless and with fruity was
due, respectively, to wild initiators and to mutant ones, the material
washed or diffusing from these two mutants that made wild populations
build more centres must in one case have produced initiators and not
come from them, and in the other case have come from them and not
produced them. However, extracts of aggregateless, fruity, and wild
cells had identical effects on the test system used—multiplying centres
without affecting optimal or minimal densities—and were all presumed
to induce initiators (Ennis and Sussman, 1958a).
It is difficult to agree that 'the implication of these findings for the
relationship between wild-type initiator and responder cells is obvious'
(Sussman, 1958).
When are initiators supposed to arise? And how are their numbers
controlled? The maximum number of centres produced by wild cells
alone was invariably 1 per 2,200 cells whatever the stage at which they
had been harvested from the growth plates (Sussman, 1956a). Thus the
proportion of initiators in the population appeared to be completely
independent of cell density, at least over a 1,000-fold range, and likewise
unaffected by vast changes in the environment. In contrast to this, the
increase in wild initiators supposedly induced by aggregateless mutants,
whether in actual contact with the wild cells or not, was proportional to
mutant density, in some cases linearly. Sussman (1958) tried to resolve
this conflict by suggesting that in pure wild populations the initiators
were so close together at the higher densities used that they competed
13. M. SHAFFER
posed that diffusible material from the mutants induced initiator cells
in the wild strain. This provided a complementary or alternative explanation of why different numbers of wild cells should have acted as
initiators when mixed with the different mutants. These mutants were
actually described as 'audiences'; but when 'audiences' participate to
this extent, it becomes rather difficult to distinguish the 'performers'.
Mixtures of wild strain with fruity. Having found that the mutant
fruity when developing alone at optimal density produced 100 times
more centres than the wild strain, Sussman (1955a, 1958) assumed that
the behaviour of mixtures of fruity and wild cells was merely additive
(though his earlier work (1952, 1954) made this unlikely), and accordingly subtracted from the total number of centres those that would
have been formed by the wild component when alone, attributed the
rest of them to the fruity, and then concluded that each of these was
started by a single mutant initiator cell. He found also that even
washings of fruity cells stimulated pure wild cells to yield many more
centres.
Comment. There was thus a radical difference in the interpretation of
the very similar results obtained with different mutants. If the increase
in centres when wild was mixed with aggregateless and with fruity was
due, respectively, to wild initiators and to mutant ones, the material
washed or diffusing from these two mutants that made wild populations
build more centres must in one case have produced initiators and not
come from them, and in the other case have come from them and not
produced them. However, extracts of aggregateless, fruity, and wild
cells had identical effects on the test system used—multiplying centres
without affecting optimal or minimal densities—and were all presumed
to induce initiators (Ennis and Sussman, 1958a).
It is difficult to agree that 'the implication of these findings for the
relationship between wild-type initiator and responder cells is obvious'
(Sussman, 1958).
When are initiators supposed to arise? And how are their numbers
controlled? The maximum number of centres produced by wild cells
alone was invariably 1 per 2,200 cells whatever the stage at which they
had been harvested from the growth plates (Sussman, 1956a). Thus the
proportion of initiators in the population appeared to be completely
independent of cell density, at least over a 1,000-fold range, and likewise
unaffected by vast changes in the environment. In contrast to this, the
increase in wild initiators supposedly induced by aggregateless mutants,
whether in actual contact with the wild cells or not, was proportional to
mutant density, in some cases linearly. Sussman (1958) tried to resolve
this conflict by suggesting that in pure wild populations the initiators
were so close together at the higher densities used that they competed