long periods of time without being used at all. 

 There are several kinds of developing and 

 differentiating systems to which the statements 

 I have just made are now known to apply. This 

 being so, we conclude that some agency of 

 control must exist in the cytoplasm to turn 

 on the reading of stored messages, since it is 

 demonstrable in all of the systems being studied 

 that the co-factors necessary for protein syn- 

 thesis are already available in the unfertilized 

 egg. Thus, there has emerged from studies of 

 macromolecule synthesis in development a need 

 to find out how translation-control systems 

 work. They clearly exist and they must be 

 concerned not only with the control of develop- 

 ment but with the control of decision-making 

 processes, in general, in differentiated higher 

 cells. 



As far as I know, there is no detailed 

 scheme that explains as yet how any translation- 

 control system works. Perhaps we will have 

 suggestions in the course of this week, as to 

 where to look for the agencies of control. 

 In the meantime, there are experiments that 

 led to the position I have just sketched, which 

 lead in turn to a closer study of the events of 

 macromolecule synthesis in early development. 

 I will discuss three lines of such experimenta- 

 tion briefly, relying mainly upon slides to sum- 

 marize the present position in each case. The 

 three problems with which we will be concerned 

 are (1) the pattern of synthesis of RNA during 

 early development, (2) the search for stored 

 maternal messages whose existence is sug- 

 gested although not proven by indirect evidence 

 and (3) a study of the proteins themselves, a 

 large fraction of which presumably are made on 



900 



700 



--300 



500 



--200 



--I00 



Fig. I. 



stable messages during the period of cleavage. 



The pattern of RNA synthesis is radically 

 different from what one might have expected 

 from the behavior of microbial systems. Figure 

 1 deals with a sucrose gradient and with RNA 

 labeled for 30 minutes at the blastula stage in 

 the sea urchin embryo. I have chosen this 

 pattern to start with because it is characteristic 

 of the pattern of synthesis of RNA throughout 

 the course of the period from cleavage to the 

 late blastula. The sea urchin has ordinary RNA 

 in bulk, with 28S, 18S and 4S species. (These 

 are the three major peaks of Fig. 1 from left 

 to right, respectively, shown by circles.) Radio- 

 activity incorporated in this case from labeled 

 uridine is distributed in gradients as shown by 

 the triangles. The circles are O. D. Such radio- 

 active material is non-coincident with the stable 

 pre-existing bulk RNA, except in the 4S region. 

 The radioactive product is highly heterogeneous 

 with respect to sedimentation constant. In the 

 4S region, where coincidence does occur, there 

 is also, throughout early cleavage, the most 

 rapidly labeled RNA. There is every reason 

 to suspect, on the basis of physical behavior 

 alone, that the non-4S material being labeled 

 is not ribosomal and is very likely, at least, 

 to be messenger RNA, or heterogeneous RNA 

 with possible template function. I should point 

 out that in these embryos there are no nucleoli 

 until long after the swimming blastula stage. 

 Figure 2 is a fortunate tangential cut through 

 the surface of a quite late blastula, already 

 ciliated and swimming. It shows nuclear profiles. 

 Note that there are no nucleoli. As long as there 

 are none, we see little or no ribosomal RNA 

 synthesis on gradients, no incorporation of label 

 that sediments in coincidence with the ribosomal 

 species and, as we shall see in a moment, base 

 compositions for the newly-synthesized mate- 

 rial that differ radically from those of the bulk 

 ribosomal RNA. When the nucleoli do appear 

 at the late gastrula stage, it becomes possible 

 to detect ribosomal RNA synthesis at a steadily 

 increasing rate. 



Figure 3 is another experiment like the one 

 represented in Fig. 1, but in this case the label- 

 ing was with radioactive phosphate. Four sets 

 of fractions were pooled, corresponding roughly 

 to the centers of gravity of the 28S, IBS, lOS 

 and 3-1/2S bulk RNA. Base-composition analy- 

 ses were performed. 



Table I will show what the compositions 

 are. The fractions shown here were indicated 

 in Fig. 3. This work was done with Arbacia. 

 DNA of this species has a GC content of slightly 



