Since postmitotic cells do not need to grow in order to maintain cell size, it is to be expected that cellular growth rate might decline when terminal cell fates are specified and differentiation begins. Here, we focus on the transition that occurs in the eye disc between proliferating progenitor cells and specified, post-mitotic cells. The Drosophila
eye imaginal disc is convenient for such studies because a wave of differentiation sweeps across the tissue so that each preparation displays successive developmental stages across the anterior-posterior axis, and these can be compared directly. Of course postmitotic cells may also change size as part of specific differentiation processes, for example Drosophila
retinal cells enlarge in the pupa, a later stage that we have not examined 
The main finding is that nucleolus size, an indicator of ribosome biogenesis, is much reduced behind the furrow in comparison to proliferative cells ahead of the furrow, or compared to the proliferating antennal disc or peripodial epithelium. Much of the reduction in nucleolus size occurs at the same time that cells enter a prolonged G1 arrest ahead of the morphogenetic furrow, and is a cell-autonomous response to Hh and Dpp, the same signals that regulate the G1 arrest () 
. The Second Mitotic Wave, a synchronous cell cycle that affects only the subpopulation of cells that are still unspecified just behind the morphogenetic furrow, is associated with a small but statistically significant size increase in the nucleoli only of this cell population, and which is prevented by expression of human p21, which also prevents cell cycle entry () 
. These findings provide evidence that ribosome biogenesis, at least one component of cellular growth, is reduced with terminal differentiation and that this coordination is mediated by Hh and Dpp signaling.
Although mutant cells unable to respond to Hh and Dpp continue to progress through the cell cycle, previous measurements of mitotic figures and of S-phase DNA synthesis suggest that they do so rather slowly, and it is not clear that the mutant cells increase in size 
. It would be interesting to determine whether maintaining nucleolar size is sufficient to maintain cellular growth, as nucleolar size may not be the only factor limiting cell cycle progression as the morphogenetic furrow approaches. Other indicators of protein synthesis that also could be assessed would be the nascent transcription of rRNA or tRNA, or the distribution of mature ribosomes. Cellular growth also depends on biosynthesis of other molecules besides proteins.
Coordination of ribosome biogenesis with the cell cycle raises the question of whether one regulates the other. If cell cycle progression feeds forward on to growth, this might explain how human p21 interferes with the growth of nucleoli of cells that should enter the Second Mitotic Wave. Preventing cell cycle arrest by forced expression of cyclins did not lead to a clear conclusion. Levels of Cyclins sufficient to cause an extra round of cell divisions posterior to the morphogenetic did not increase nucleolar size. It is possible that cell cycle progression is necessary to increase nucleolar size but not sufficient, or that human p21 interferes with ribosome biogenesis independently of blocking S-phase entry.
There is also abundant evidence for the coupling of cellular growth to cell cycle progression, such that cell growth is sometimes considered the first step in the cell cycle, and its absence is a key distinction between G0 and cycling cells 
. Sensitivity of cell cycle protein translation to growth conditions is one mechanism that has been proposed 
. Cell cycle progression and growth may also be co-regulated in a parallel fashion, for example as common transcriptional targets of the DNA Replication Related Element binding Factor (DREF) 
. Previous studies have concluded that Cyclin D and Cdk4 promote cellular growth as a major part of their effect because ectopic cell division driven by Cyclin D and Cdk4 over-expression does not reduce cell size, and because these proteins can increase the size of postmitotic cells 
. The only indication of a possible effect on ribosome biogenesis in our experiments was in postmitotic, differentiating cells at the back of the disc (), however, not in the cells that divide in response to CyclinD and Cdk4. It is possible that Cyclin D and Cdk4 affect cellular growth through other pathways such as mitochondrial biogenesis 
. Another possibility is that division of the cell’s components at mitosis might transiently reduce the size of nucleoli, masking any increase due to Cyclin D/cdk4, and potentially reducing the size of nucleoli in cells over-expressing Cyclin E ().
A key target of Dpp and Hh with regard to eye differentiation is thought to be the proneural gene atonal
, which is required to specify the first photoreceptor cells but does not contribute to cell cycle arrest ahead of the furrow 
. In addition it is uncertain whether ato
is regulated directly by the Mad and Ci transcription factors that are the targets of Dpp and Hh signal transduction 
. Recently, it has been suggested that cell cycle arrest depends on repressing the Meis-family protooncogene homolog homothorax
. Another model posits that it is changes in Dpp signaling level, either in space or in time, that are required for proliferation 
. We hypothesize that differentiation, cell cycle arrest, and the attenuation of cellular growth are somewhat independent processes, coordinated by each sharing regulation from Hh and Dpp signaling in the eye. As has been noted before regarding the regulation of both differentiation and the cell cycle by the same extracellular signals, since so much developmental signaling is mediated by a small number of cell-cell signaling pathways, coordination (or antagonism) between developmental processes can be a natural consequence of regulation by common extracellular signals 
It would be no surprise if changes in ribosome biogenesis were accompanied by changes in other processes such as energy generation and protein turnover. Consistent with this idea, genome-wide studies point to large, antagonistic human gene networks whose expression changes as proliferation gives way to differentiation 
, and indicate that expression of a significant fraction of eukaryote genomes may be regulated by growth rate 
. These studies have been based on changes during aging or in nutrient availability, however, and it remains to be seen how regulation occurs during developmental patterning, when cellular growth rate can even differ within the same tissue at the same time, as seems to be the case in the eye imaginal disc.