Theoretical modes of DNA replication
There are three theoretical modes of DNA replication: semiconservative respiration, conservative replication, and dispersive replication. In semiconservative replication, each daughter DNA molecule has one parental DNA strand. In conservative replication, one daughter DNA molecule is made completely of newly-synthesized DNA, and the other daughter DNA molecule is made completely of parental DNA. In dispersive replication, each daughter strand has fragments of new DNA and fragments of parental DNA. The Meselson and Stahl experiment suggested that DNA usually replicates in a semiconservative manner.
Model of DNA Replication in Escherichia coli
DNA replication has been well characterized in Escherichia coli.
Initiation of Replication
The initiation of DNA replication occurs at a region called the replicator. Here, there are 13-bp AT repeats and 9-bp repeats. DnaA, an initiator protein, binds to the 9-bp repeats, which leads to the denaturation of the 13-bp sequences, which is the origin of replication, since this is where DNA replication will start. DnaC (helicase loader) and DnaB (helicase) then are recruited; DnaB loads DnaC onto both strands of the denatured 13-bp region and help separate the DNA strands further and unwind the DNA. This is an ATP dependent process. In order for actual replication to start, there need to be primers, since DNA polymerase must synthesize off of a preexisting 3'-OH end, so DnaG (primase) forms a complex with helicase (called the primosome) and synthesizes an RNA primer, which will be removed later. Replication can now begin.
While helicase is continually denaturing more and more DNA, SSB (single-strand DNA binding proteins) coat the denatured strands to prevent them from reannealing and disrupting replication. DNA polymerase III meanwhile continues replication on both strands. DNA synthesis ALWAYS occurs in the 5'-3' direction (there is a chemical reason for this). DNA polymerase III, consisting of 9 different polypeptides, adds nucleotides in the 5'-3' direction onto both DNA strands basically at the same time. However, the DNA strands are antiparallel, so only one strand can be synthesized continuously. This is why DNA replication is called semidiscontinuous. So, while the leading strand is synthesized continuously without interruption and without the constant need for primers, the lagging strand requires constant synthesis of primers, and short bouts of DNA polymerase III activity, synthesizing short bits starting from the 3'-OH of the primers, then detaching and reattaching itself to the lagging strand because if you consider the leading strand to be synthesized in the direction of helicase movement, the lagging strand is synthesized in the opposite direction. The "short bits" are called Okazaki fragments. The RNA primers are then removed via DNA polymerase I 5'-3' exonuclease activity and simultaneously replaced with DNA nucleotides. Then, the sugar-phosphate backbones between the DNA that filled in the primer and the DNA synthesized by DNA polymerase III are ligated together via DNA ligase.
In prokaryotes such as E. coli, provided that the DNA is circular, only one origin of replication is needed. However, for eukaryotes, such as humans, if we only had one origin of replication per chromosome, we would take a very long time to finish DNA replication. So, eukaryotes have multiple origins of replication per chromosome. The denatured area around one origin of replication is called a replication bubble.
Helpful animation for visualization of the process: