Molecular Biology of the Cell by Bruce Alberts, Alexander Johnson, Julian Lewis, David Morgan, Martin Raff, Keith Roberts, Peter Walter by by Bruce Alberts, Alexander Johnson, Julian Lewis, David Morg

1178 Chapter 21: Development of Multicellular Organisms
retreats tailward, extending the embryo (Figure 21–38B). In the process, it deposits
a trail of somites formed from cells that group together into blocks as they emerge
from the anterior end of the presomitic region. The special character of the presomitic
mesoderm is maintained by a combination of fibroblast growth factor
(FGF) and Wnt signals, produced by a signaling center at the tail end of the
embryo, and the range of these signals seems to define the length of the presomitic
mesoderm. The somites emerge with clocklike timing, but what determines
the rhythm of the process?
In the posterior part of the presomitic mesoderm, the expression of certain
genes oscillates in time. Snapshots of gene expression taken by fixing embryos
for analysis at different times in the oscillation cycle reveal what is happening,
and the oscillations can now also be observed in time-lapse movies of embryos
containing fluorescent reporters of individual oscillating genes. One new somite
pair is formed in each oscillation cycle, and, in mutants where the oscillations fail
to occur, somite segmentation is disrupted: the cells may still break up, belatedly,
into separate clusters, but they do so in a haphazard, irregular way. The gene-expression
oscillator controlling regular segmentation is called the segmentation
clock. The length of one complete oscillation cycle depends on the species: it is 30
minutes in a zebrafish, 90 minutes in a chick, 120 minutes in a mouse.
As cells emerge from the presomitic mesoderm to form somites—in other
words, as they escape from the influence of the FGF and Wnt signals—their oscillation
stops. Some become arrested in one state, some in another, according to
the phase of the oscillation cycle at the time they leave the presomitic region. In
this way, the temporal oscillation of gene expression in the presomitic mesoderm
leaves its trace in a spatially periodic pattern of gene expression in the maturing
mesoderm; this in turn dictates how the tissue will break up into physically
separate blocks, through effects on the pattern of cell–cell adhesion (see Figure
21–38B).
How does the segmentation clock work? The first somite oscillator genes to
be discovered were Hes genes, which are key components of the Notch signaling
pathway. They are directly regulated by the activated form of Notch, and they code
for inhibitory transcription regulators that inhibit the expression of other genes,
including Delta. As well as regulating other genes, the products of Hes genes can
directly regulate their own expression, creating a remarkably simple negative
feedback loop. Autoregulation of certain specific Hes genes (depending on species)
is thought to be the basic generator of the oscillations of the somite clock.
Although the machinery has been modified in various ways in different species,
the underlying principle seems to be conserved. When the key Hes gene is transcribed,
the amount of Hes protein product builds up until it is sufficient to block
Hes gene transcription; synthesis of the protein ceases; the protein then decays,
permitting transcription to begin again; and so on, cyclically (Figure 21–39). The
period of oscillation, which determines the size of each somite, depends on the
delay in the feedback loop. This equals the sum of the gestation delays and accumulation
delays (that is, the molecular lifetimes) of the Hes mRNA and protein
molecules, according to the additive principle discussed earlier. Mathematical
modeling (see Chapter 8) allows us to relate these basic molecular parameters to
the cycle time of the segmentation clock: to a first approximation, the cycle period
is simply equal to twice the total delay in the negative feedback loop, and thus
twice the sum of the delays occurring at each step of the loop.
The feedback loop just described is intracellular, and each cell in the presomitic
mesoderm can generate oscillations on its own. But these oscillations at the
single-cell level are somewhat erratic and imprecise, reflecting the fundamentally
noisy, stochastic nature of the control of gene expression, as discussed in Chapter
7. A mechanism is needed to keep all the cells in the presomitic mesoderm that
will form a particular somite oscillating in synchrony. This is achieved through
cell–cell communication via the Notch signaling pathway, to which the Hes genes
are coupled. The gene regulatory circuitry is such that in this context Notch signaling
does not drive neighboring cells to be different, as in lateral inhibition, but
does just the opposite: it keeps them in unison. In mutants where Notch signaling