120
B. M. SHAFFER
Sussman and his associates believe that centres are started by single,
specialized initiator cells, and that this ratio is an expression of the fixed
proportion in which initiators occur in the population. This implies the
rather improbable coincidence that, with rising density, the point at
which all the initiators become able to start centres is immediately
below that at which these begin to inhibit one another (Jaffe, 1958;
Shaffer, 1958). In any case, in his recent experiments on large discoideum
populations, Bonner has obtained a ratio much higher than 1 : 2,200,
and a correspondingly smaller aggregation territory. Presumably the
intensity of inhibition depends on the precise environmental conditions.
Small-population experiments. Sussman and Noel (1952) found that a
considerable fraction of groups of about 1,000 cells, deposited at densities that were greater than the optimal in larger populations, failed
to aggregate at all. It was concluded that the special initiator cells
needed to start aggregation were absent in these cases, and therefore
that the population was heterogeneous by the end of the growth period.
By first assuming that the number of centres per group would have
shown a Poisson distribution had not the first centres to appear tended
to prevent others from developing later, and by using the fraction without any centres so as to avoid this bias, the authors calculated the number of cells per centre. The mean value from these and later experiments
was 2,250 (Sussman, 1958); and this was independent of cell density.
Comment. The agreement between the ratios in large and small populations was certainly remarkable, but neither value has been obtained
by other workers. In any case, this supposedly crucial experiment could
not establish the pre-existence of cellular heterogeneity: it only measured
the probability of centres' forming in certain conditions. As to these
conditions, it seemed possible that the method of preparing the cells
greatly reduced their capacity to aggregate. In all the experiments in
Sussman's laboratory on initiation, the cells were first freed from their
food bacterium by centrifuging (Runyon, 1942) up to 7 times in distilled
water, and then their performance was studied after they had been
deposited on distilled-water agar that had been washed up to 20 times.
It has been known for some time that the long pre-aggregation period
that follows centrifuging recurs if the cells are respun in distilled water
(but not isotonic buffer (Gerisch, 1959)) even when they have already
begun to aggregate, and that cytochemical change antecedent to aggregation is reversed (Bonner, Chiquoine, and Kolderie, 1955; Bonner,
1959a). Still more significantly, Sussman (1955a) could actually wash
out a material which promoted centres. Ennis and Sussman's (1958b)
description of the prepared cells as being hemispherical and very
sluggish emphasized their unhealthiness. Ennis and Sussman suggested
that washing the agar, which was done to prevent the cells dispersing,
B. M. SHAFFER
Sussman and his associates believe that centres are started by single,
specialized initiator cells, and that this ratio is an expression of the fixed
proportion in which initiators occur in the population. This implies the
rather improbable coincidence that, with rising density, the point at
which all the initiators become able to start centres is immediately
below that at which these begin to inhibit one another (Jaffe, 1958;
Shaffer, 1958). In any case, in his recent experiments on large discoideum
populations, Bonner has obtained a ratio much higher than 1 : 2,200,
and a correspondingly smaller aggregation territory. Presumably the
intensity of inhibition depends on the precise environmental conditions.
Small-population experiments. Sussman and Noel (1952) found that a
considerable fraction of groups of about 1,000 cells, deposited at densities that were greater than the optimal in larger populations, failed
to aggregate at all. It was concluded that the special initiator cells
needed to start aggregation were absent in these cases, and therefore
that the population was heterogeneous by the end of the growth period.
By first assuming that the number of centres per group would have
shown a Poisson distribution had not the first centres to appear tended
to prevent others from developing later, and by using the fraction without any centres so as to avoid this bias, the authors calculated the number of cells per centre. The mean value from these and later experiments
was 2,250 (Sussman, 1958); and this was independent of cell density.
Comment. The agreement between the ratios in large and small populations was certainly remarkable, but neither value has been obtained
by other workers. In any case, this supposedly crucial experiment could
not establish the pre-existence of cellular heterogeneity: it only measured
the probability of centres' forming in certain conditions. As to these
conditions, it seemed possible that the method of preparing the cells
greatly reduced their capacity to aggregate. In all the experiments in
Sussman's laboratory on initiation, the cells were first freed from their
food bacterium by centrifuging (Runyon, 1942) up to 7 times in distilled
water, and then their performance was studied after they had been
deposited on distilled-water agar that had been washed up to 20 times.
It has been known for some time that the long pre-aggregation period
that follows centrifuging recurs if the cells are respun in distilled water
(but not isotonic buffer (Gerisch, 1959)) even when they have already
begun to aggregate, and that cytochemical change antecedent to aggregation is reversed (Bonner, Chiquoine, and Kolderie, 1955; Bonner,
1959a). Still more significantly, Sussman (1955a) could actually wash
out a material which promoted centres. Ennis and Sussman's (1958b)
description of the prepared cells as being hemispherical and very
sluggish emphasized their unhealthiness. Ennis and Sussman suggested
that washing the agar, which was done to prevent the cells dispersing,