REMODELING OF CHROMATIN LOOPS DOES NOT ACCOUNT FOR SPECIFICATION OF REPLICATION ORIGINS DURING XENOPUS DEVELOPMENT

We have investigated the possible relationship between replicons and c hromatin loops during Xenopus development. In early embryos, replicati on of the ribosomal RNA genes (rDNA) can initiate at apparently any se quence. Nevertheless, the need for a regular spacing of replication or igins suggests that some periodic chromatin folding might dictate whic h sites are actually used for initiation. After the midblastula transi tion, replication initiation is restricted to the rDNA intergenic spac ers. A remodeling of chromatin folding could account for this change i n origin usage. Here, it is reported that nuclear matrix anchorage of the Xenopus rDNA occurs at multiple, apparently random sequences, thro ughout embryonic development as well as in adult cells. In vitro matri x rebinding assays confirmed the lack of specific anchoring sequences in the rDNA, before as well as after specific replication origins are established. Thus, no change in loop attachment sites could explain th e change in origin usage at this locus. Nonspecific loop anchorage was a special feature of the rDNA locus, since the same nuclear matrices were able selectively to bind the scaffold attachment region (SAR) of the Drosophila histone gene cluster in vitro. Blastula and gastrula nu clear matrices bound a higher amount of SAR sequences than matrices fr om later stages or adult cells. This developmental change in SAR bindi ng might explain the increase in size of the bulk of genomic DNA loops that occurs after the gastrula stage. However, no change in chromatin loop organization that could explain the midblastula stage transition from small to large replicons was observed.