ADVERTISEMENTS:
It has now been over 30 years since J. D. Watson, F. H. C. Crick, and M. H. F. Wilkins established the double-stranded, helical nature of the DNA molecule and suggested how DNA serves in its own replication.
According to their original model for DNA replication, the two polynucleotide chains of the “parent” double helix separate and each serves as a “template” for the synthesis of a new, complementary polynucleotide chain.
During strand separation, the nitrogen bases of each original strand are exposed and establish sites for the association of free nucleotides.
ADVERTISEMENTS:
These nucleotides are then enzymatically linked together to form a new complementary strand. Because deoxyadenylic acid (dAMP) can form hydrogen bonds only with thymine of the template strand (and dGMP can bond only to cytosine, dCMP only to guanine, and dTMP only to adenine), the newly synthesized strand will be identical to the original complementary strand.
As a result, two new double helices are formed, each consisting of one polynucleotide strand from the parent double helix and a newly synthesized polynucleotide strand. Because each of the two double helices conserves only one of the parent polynucleotide strands, the process is said to be semiconservative.
Replication as a “Semiconservative” Process:
Although semiconservative replication of DNA was predicted by the original Watson-Crick model, it was not verified until the classic studies of M. S. Meselson and F. W. Stahl. At the time of their experiments, two other modes of replication were deemed equally feasible (Fig. 21-1):
ADVERTISEMENTS:
(1) Conservative replication, in which both strands of the parent double helix would be conserved and the new DNA molecule would consist of two newly synthesized strands; and
(2) Dispersive replication, in which replication would involve fragmentation of the parent double helix and the interspersing of pieces of the parent strands with newly synthesized pieces, thereby forming the two new double helices.
Meselson and Stahl verified the semiconservative nature of DNA replication in a series of elegant experiments using isotopically labeled DNA and a form of isopycnic density gradient centrifugation (see Chapter 12). They cultured Escherichia coli cells in a medium in which the nitrogen was 15N (a “heavy” isotope of nitrogen, but not a radioisotope) instead of the commonly occurring and lighter 14N.
In time, the purines and pyrimidines of DNA in new cells contained 15N (where 14N normally occurs) and thus the DNA molecules were denser. DNA in which the nitrogen atoms are 15N can be distinguished from DNA containing 14N because during isopycnic centrifugation, the two different DNAs band at different density positions in the centrifuge tube (Fig. 21-2).
Meselson and Stahl centrifuged DNA isolated from the cells for 2-3 days at very high rotational speeds in centrifuge tubes initially containing a uniform solution of CsCl. During centrifugation, density gradients were automatically formed in the tubes as a result of the equilibrium that was established between the sedimentation of CsCl toward the bottom of the tube and diffusion of the salt toward the top of the tube. This form of centrifugation, called equilibrium isopycnic centrifugation,
Depending on its content of 15N and 14N, the DNA bands at a specific position in the density gradient. Because the DNA synthesized by cells grown in 15N would be denser than 14N-containing DNA it would band further down the tube (Fig. 21-2).
Cells grown for some time in the presence of 15N- medium were washed free of the medium and transferred to 14N-containing medium and allowed to continue to grow for specific lengths of time (i.e., for various numbers of generation times). DNA isolated from cells grown for one generation of time in the 14N medium had a density intermediate to that of the DNA from cells grown only in 15N-containing medium (identified as generation 0 in Fig. 21-3) and that of DNA from cells grown only in 14N-containing medium (the controls of Fig. 21-3).
Such a result immediately ruled out the possibility that DNA replication was conservative, as conservative replication would have yielded two DNA bands in the density gradient for generation 1 (i.e., F,) cells. The single band of intermediate density (identified as “hybrid” DNA in Fig. 21-3) consisted of DNA molecules in which one strand contained 15N and the other contained 14N.
When the incubation in 14N medium was carried out for two generations of time (i.e., generation 2), two DNA bands were formed—one at the same density position as the DNA from cells grown exclusively in 14N medium (i.e., “light controls”) and one of intermediate density. Subsequent generations produced greater numbers of DNA molecules that banded at the “light” (14N- containing DNA) position in the density gradient. These results are consistent only with the model of semiconservative replication.
Dispersive replication would have produced a single band for each generation and the band would have been found at successively lighter density positions in the gradient. Studies using other prokaryotes as well as eukaryotes indicate that semiconservative replication of DNA is probably the universal mechanism.
Replication by Addition of Nucleotides in the 5′→3′ Direction:
ADVERTISEMENTS:
Each nucleotide of a DNA strand is joined to the next nucleotide by a phosphodiester bond that links the 3′ carbon of its deoxyribose to the 5′ carbon of the deoxyribose of the next nucleotide. At one end of the polynucleotide chain, there is a hydroxyl group attached to the 3′ carbon of the last nucleotide, and at the other end there is a phosphate group attached to the 5′ carbon.
The two chains of a double helix have opposite polarities and are said to be antiparallel, that is, each end of the double helix contains the 5′ end of one strand and the 3′ end of the other. During replication, attachment of a nucleotide to a growing strand always takes place at the terminal 3′ position of that strand. In other words, the forming polynucleotide chain “grows” from its 5′ end toward its 3′ end.
Unidirectional and Bidirectional Replication:
Replication starts at a point on the chromosome where the two parental strands begin to separate; this point is called the origin. Addition of complementary nucleotides to form two new strands takes place along both parent strand templates starting from that point (Fig. 21-4).
ADVERTISEMENTS:
In unidirectional replication, growth proceeds along both strands in the same direction leading from the origin. Along one of the parental template strands, synthesis of the new complementary strand takes place by the continuous addition of nucleotides to the available 3′ end of the forming strand. The growing strand is called the leading strand or continuous strand. The 5′ end of this strand is located at the origin and its 3′ end at the moving replication fork (i.e., the progressing point of separation of the parental strands).
The other polynucleotide strand being formed is called the lagging strand or discontinuous strand. The elongation of this strand takes place by a somewhat modified mechanism. In contrast to the leading strand, the lagging strand has its 3′ position at the origin and its 5′ position at the replication fork. If nucleotides were sequentially added to the end of the lagging strand at the replication fork, then this strand’s growth would proceed in a 3’→5′ direction.
This does not occur. Instead, growth takes place by the synthesis of a number of short polynucleotide chains between the replication fork and the origin. Each short chain is laid down in the direction 5′ to 3′ and these are later linked together and to the 5′ end of the lagging strand.
ADVERTISEMENTS:
As a result, the overall direction of growth of the lagging strand is the same as that of the leading strand. The unusual growth pattern that characterizes the synthesis of the lagging strand explains why it is also referred to as the “discontinuous” strand.
In bidirectional replication (Fig. 21-5), two replication forks are formed at the origin and these move away from the origin in both directions as the parental double helix is separated. The synthesis of the complementary strands also occurs in both directions. Behind each fork there is a set of leading and lagging strands. As in the case of unidirectional replication, elongation of the two leading strands is continuous, whereas elongation of the two lagging strands is discontinuous.
It is to be noted that regardless of whether replication is unidirectional or bidirectional, the addition of nucleotides always occurs in the direction from 5′ to 3′, as new nucleotides are added to available 3′ ends of either the continuous strand or the discontinuous strand. Discontinuous synthesis of lagging strands was first demonstrated by R. Okazaki. Okazaki incubated E. coli cells in a medium containing 3H-thymidine for very short periods of time (a pulse of only 15 seconds) and then examined the distribution of the radioisotope in newly synthesized DNA.
The radioisotope was found in a number of polynucleotides (1000-2000 nucleotides long), now referred to as Okazaki fragments (Figs. 21-4 and 21-5). When pulsed cells were transferred to unlabeled medium for varying lengths of time prior to analysis, the radioactive label was recovered in much longer stretches of DNA. This is because the Okazaki fragments produced during the short tritium pulse had been linked together and connected to the 5′ end of the lagging strand.
In eukaryotic cells, Okazaki fragments are usually smaller (about 100-200 nucleotides long). Bidirectional replication of DNA is the mechanism employed in all eukaryotic and most prokaryotic cells. Unidirectional replication is rare and appears to occur in only a limited number of prokaryotes.
ADVERTISEMENTS:
Visualization of Replication in E. coli:
In 1963, J. Cairns developed a procedure employing a combination of microscopy and autoradiography that made it possible to visualize the replication of the chromosome of E. coli. Cairns plaped E. coli cells in a medium containing 3H-thymidine for various periods of time so that the radioactive thymidine was incorporated into the DNA as the chromosome was replicated in successive generations of cells.
Cells were removed from the medium after various periods of incubation and gently lysed to release the chromosome from the cell (the shear forces created by harsh lysis break the chromosome into small pieces). The chromosomes were then transferred to glass slides and coated with a photographic emulsion sensitive to the low-energy beta particles emitted by the 3H-thymidine.
After exposing the emulsion to the beta rays, the emulsion was developed and examined by light microscopy. Wherever decay of labeled thymidine had occurred in a chromosome, the emulsion was exposed and created visible grains.
A chromosome not engaged in replication appeared as a circular structure formed from a close succession of exposure spots. Chromosomes “caught in the act” of replication gave rise to what are called theta structures because they have the appearance of the Greek letter theta (i.e., 0) (Fig. 21-6). The theta structures reveal the positions of the replication forks in the circular chromosome.
The Replicon and the Replication Sequence:
ADVERTISEMENTS:
The sequence of events that takes place during DNA replication is best understood for prokaryotes and appears to be as follows (Fig. 21-7). Parental strand separation begins at a site called the origin which contains a special nucleotide sequence and directs the association of a number of proteins. ATP-dependent unwinding enzymes (also called helicases) promote separation of the two parental strands and establish replication forks that will progressively move away from the origin (Fig. 21-7a); the helicases separate the parental strand at about 1000 base pairs per second.
Behind the replication fork, the single DNA strands are prevented from rewinding about one another (or forming double-stranded hairpin loops in each single strand) by the actions of a set of proteins called helix-destabilizing proteins or single-strand binding proteins (i.e., “SSBs”) (Fig. 21-7b). The action of a helicase introduces a positive supercoil into the duplex DNA ahead of the replication fork. Enzymes called topoisomerases relax the supercoil by attaching to the transiently supercoiled duplex, nicking one of the strands, and rotating it through the unbroken strand. The nick is then resealed.
Prior to DNA synthesis beginning at the origin, short RNA polynucleotides are formed that are complementary to the DNA template. These stretches of RNA are called primers and are also laid down in the 5′ to 3′ direction. DNA nucleotides are then added one at a time to the free 3′ ends of the RNA primers. Because growth of the lagging strand is discontinuous, several RNA primers and Okazaki fragments are formed. Note that an RNA primer must be formed for each Okazaki fragment to be laid down (Fig. 21-7c). The enzymes required for the synthesis of the RNA primers are a special class of RNA polymerases called RNA primases.
Elongation of the leading strand and synthesis of the Okazaki fragments are catalyzed by an enzyme called DNA polymerase III. The substrates of DNA polymerase III are the deoxynucleoside triphosphates (e.g., dATP, dGTP, dCTP, and dTTp). Addition of a nucleotide to the available 3′ position of the continuously growing leading strand or an Okazaki fragment of the lagging strand involves removal of pyrophosphate to yield a deoxynucleoside monophosphate (e.g., dAMP, dGMP, dCMP, and dTMP).
On completion of the Okazaki fragments, the RNA primers are excised by DNA polymerase I, which then fills the resulting gaps with DNA (Fig. 21-7d). After DNA polymerase I adds the final deoxyribonucleotide in the gap left by the excised primer, the enzyme DNA ligase forms the phosphodiester bond that links the free 3′ end of the primer replacement to the 5′ end of the Okazaki fragment (Fig. 21-7f).
DNA Polymerases and “Processivity”:
In general, three different DNA polymerase enzymes are found in cells. In prokaryotes, these are called DNA polymerase I, DNA polymerase II, and DNA polymerase III. As noted above, DNA polymerase I excises the RNA primers and fills the gaps with DNA, whereas DNA polymerase III adds nucleotides to the growing leading strand and to the 3′ ends of the RNA primers.
The function of DNA polymerase II remains unknown. In eukaryotic cells, the DNA polymerases are DNA polymerase a, DNA polymerase II, and DNA polymerase 7; their functions are compared with the pro- karyotic enzymes in Table 21-1. The rapidity and efficiency with which a DNA polymerase extends a growing chain is referred to as processivity. For example, the processivity of DNA polymerase III acting on the 3′ end of the leading strand is very high because the enzyme remains associated with the growing end of the chain and the template strand.
During unidirectional replication, the replication fork fully circles the chromosome and the resulting DNA molecules are separated. For prokaryotes in which replication is bidirectional, the replication forks proceed around the chromosome until they meet. In eukaryotic chromosomes, where there are many replicating units or replicons, all replicons are linked together before the chromatids can be -separated. The replicon consists of that segment of a chromosome that includes an origin and two termination points (i.e., points where replication ends).
The “Replisome”:
N. K. Sinha and A. Kornberg have suggested that the DNA polymerases, RNA primases, and helicases may be associated with each other to form a multienzyme complex, the replisome that carries out the synthesis of the leading and lagging strands in a coordinated fashion (Fig. 21-8). Such a complex would be highly processive and assure rapid replication of the DNA.
High Fidelity of Replication:
Despite the complexity of the process and the rapidity with which it proceeds, very few errors are made during DNA replication. For example, it is estimated that for every error that occurs, 109 base pairs are replicated faithfully. The high fidelity of DNA replication is attributable in part to the special properties of the DNA polymerases.
The DNA polymerases will add a nucleotide to the available 3′-OH end of a growing DNA strand (or RNA primer) only if the prior nucleotide is properly base-paired with the template nucleotide. If a mismatched nucleotide is present, growth of the strand is transiently halted while a segment of the strand containing the error is excised. With a corrected 3′ end reestablished, elongation by the DNA polymerase is resumed.
Much remains to be learned about the enzymes that catalyze the reactions of replication. So far, the major obstacle to such studies has been the difficulty of isolating and purifying the enzymes, either individually or in complexes. This is due in part to the fact that some of them may be associated with membranes. In prokaryotic cells, the replication forks are bound to the plasma membrane.
About two dozen different proteins have been shown to be involved with the replication of DNA in E. coli cells. Several of the E. coli proteins are involved with prepriming reactions, that is, the reactions that occur before the formation of RNA primers.