Is late replication of the inactive X chromosome irreversible in all cells of mammals?

The X chromosomes of the female bandicoot rat (Nesokia indica) were 3H-thymidine labeled during two consecutive cell divisions to determine if all of the same segments of the "triplicate-type" X chromosome of these animals always replicated late. In 87% of metaphases examined the findings were as expected. One entire X chromosome (X1) and the long arm of the other X (X2) synthesized DNA late in the S phase in both divisions. However, in the other 13% of the metaphases, the late-replicating and presumably genetically inactive short-arm segments of the X1 chromosome had completed DNA synthesis by the time it entered the late-S phase of the second cycle. Thus, in this species, some cells appear to have an X chromosome of which the facultative heteropycnotic segment condenses in one cell cycle but becomes euchromatic in the subsequent cell cycle. Although this appears at first to be inconsistent with the generally accepted pattern of X-chromosome condensation and genetic inactivation, it may represent an instance of evolutionary specialization for an as yet unexplained reason. It is also possible that closer analysis of other mammalian species with composite sex chromosomes or methods equally suitable for this type of analysis will reveal other instances where a minority of the somatic cells of females do not follow the predictions of the Lyons hypothesis completely.