Thoughts and observations on patterning in early mammalian development

The preimplantation mammalian conceptus shows an impressive ability to develop normally following the loss, gain or rearrangement of cells. This has prompted the view that, unlike in other species, patterning in mammals cannot depend on information that is already present in the zygote before it begins to cleave. However, various findings are hard to reconcile with this conclusion, including evidence that the incidence of monozygotic twinning is sensitive to the conditions to which eggs or very early concepti are exposed. Possible causes of early twinning are discussed, and it is argued that partial hatching of the conceptus through a hernia in the zona pellucida cannot account for all cases. Moreover, it remains questionable whether studies on aggregated morulae and isolated blastomeres really provide compelling evidence against the existence of indispensible patterning information in the egg. Finally, regularities in axial relationships between the blastocyst and zygote have been revealed employing strictly non-invasive techniques. These show that, at least in normal development, patterning begins before cleavage.