For determination of whether DNA replication is initiated at the nuclear envelope, synchronizedChinese hamster ovary cells labeled with [3HIthymidine were examined by electron microscope radioautography. The cells were synchronized initially by mitotic shake-off and held at the G1-S border by 5-fluorodeoxyuridine plus amethopterin. Cells were fixed at 1, 5, 10, and 30 min after the inhibitors were counteracted with [3H]thymidine.Radioautographic silver grains in each case were present over the more central parts of nuclei and were generally absent from the region of the nuclear envelope. We conclude that neither initiation nor continuation of DNA replication is associated with the nuclear envelope.An important facet of the problem of DNA replication in eukaryotes is whether initiation of replication occurs at points of attachment of the chromosome to the nuclear envelope. The question arises in part because of the evidence that the replicons of prokaryotic chromosomes initiate and then continue replication at points of attachment to the plasma membrane (see, for example, refs. 1-3). Several kinds of experiments on prokaryotes suggest that the enzymes and other factors necessary for DNA replication may form a complex that remains associated with the plasma membrane during chromosome replication (see ref. 4
for review).In studies of human amnion cells, Comings and Kakefuda (5) concluded from electron microscope radioautography that DNA replication is initiated at the nuclear envelope at the beginning of the period of DNA synthesis (S period). Contrary to this conclusion, Williams and Ockey (6) have obtained evidence by electron microscope radioautography that DNA replication at the beginning of the S period is not initiated at the nuclear envelope in Chinese hamster cells. According to them and to Huberman et al. (7), DNA synthesis late in the S period is highly concentrated near the nuclear envelope; this result is to be expected because replication of DNA in heterochromatin dominates the later part of the S period, and heterochromatin tends to be tightly condensed against the nuclear envelope. The results of Williams and Ockey (6) are to some degree supported by work of Erlandson and de Harven (8) for HeLa cells and of Blondel (9) for KB cells. In another eukaryote, Amoeba proteus, DNA replication is probably not associated with the nuclear envelope at any time during the S period (10).To answer the question of association of initiation or continuation of replication with the nuclear envelope in mammalian cells requires a very high degree of synchrony of the cultured cells. Many of the agents used for synchronization are incompletely effective. Ockey (11), for example, showed that amethopterin does not block cells from entering the S period. A high thymidine concentration (thymidine block) also fails to block cells at the G1-S border (12).We describe experiments here that minimize the shortcomings of current methods of synchronization. Our results strongly indicate that DNA replication is not initiated...