123island, which encodes Aur B, exhibit defects in furrow formation. However, in these
embryos, cytoskeletal structures associated with the furrow (see below) appear to be
induced relatively normally in the center of the blastomere, whereas they are entirely
absent in more distal regions of the furrow. As expected, Aur B protein in these
embryos is found both at the spindle midzone and the tips of astral microtubules that
contact the forming furrow (Fig. 4.2). Genetic analysis of the Aur B maternal-effect
cellular island mutant allele indicates that it retains some functional activity, as
homozygotes for complete loss of function alleles are zygotic lethal and do not
survive to adulthood, in contrast to homozygotes for the maternal-effect mutant
allele. A comparison of the maternal-effect cellular atoll phenotype, which allows
formation in the medial furrow region, with the effects of a specific Aur B inhibitor,
which cause furrow inhibition throughout the furrow, suggests that the partial activ-
ity in the maternal-effect cellular island allele is provided by Aur B function present
in the spindle. Such spindle-provided Aur B may be at a higher concentration or
have a higher functional activity than that present in astral microtubule ends.
Consistent with this interpretation, embryos maternally mutant for both futile cycle,
which fails to form spindle structures, and cellular island lack furrow-associated
structures throughout the length of the furrow, in both medial and distal regions.
Altogether, these observations suggest that, in large embryonic blastomeres, CPC
activity and potentially other signals from astral microtubules are essential for fur-
row induction in distal regions of the cell, which are presumably too distant to be
influenced by inducing signals from chromatin present at the spindle midzone (Figs.
4.2 and 4.4). Indeed, in several amphibians, early cleavage furrow advance may
depend on signals propagated solely through the cortex via the furrow’s distal ends
(Sawai 1974 , 1980 ; Mabuchi et al. 1988 ; Sawai and Yomota 1990 ). These spatial
functional specializations of furrow-inducing activity may be a necessary adapta-
tion to the small coverage of the spindle relative to the much larger embryonic
blastomeres.
Furrow-inducing activity from the spindle midzone and astral microtubules is
concentrated at the point halfway between the spindle poles, and in a plane perpen-
dicular to that of the spindle, thus functionally linking cleavage patterning to the
mechanistic determinants of spindle orientation within dividing blastomeres. This
linkage will be discussed in the next section.
4.3.2 Centering and Orienting Asters and Spindles
For over 125 years, scientists have known that mitotic spindles tend to align along
the longest axis of a dividing cell and that cleavage furrows tend to form perpen-
dicular to the mitotic spindle (Hertwig 1893 ). The tendency holds even when cell
shape is deliberately manipulated to change the orientation of a cell’s longest axis.
In the original manipulations, artificially elongating frog embryos by compression
generated a reorientation of the cleavage plane consistent with realignment of the
spindle (Pflüger 1884 ; Hertwig 1893 ; Black and Vincent 1988 ). This phenomenon,
4 Vertebrate Embryonic Cleavage Pattern Determination