Biochemistry 3107 - Fall 2003

Bacterial DNA polymerase(s)

The discovery of an enzyme

In the 1950s, Arthur Kornberg recognized that the energetics of the biochemical synthesis of a polynucleotide chain would require deoxynucleoside triphosphates (dNTPs). He also realized that he needed a very sensitive assay for DNA synthesis and that if he radioactively labelled the a phosphate of dNTPs he could monitor the incorporation of the labelled dNTPs into DNA with a simple assay (actually precipitation with trichloroacetic acid - TCA). By choosing E. coli for his studies, he was also able to prepare large quantities of cells and to obtain large amounts of purified enzyme.

MVH24-2

In this way he discovered an enzyme in E. coli that could catalyse the synthesis of DNA. He called it DNA polymerase. We now know it as DNA polymerase I. We also now know that there are four other DNA polymerase enzymes in E. coli.

Kornberg was awarded the 1959 Nobel Prize in Physiology or Medicine jointly with Severo Ochoa, who had discovered an enzyme that catalyzed the synthesis of RNA. However, although both Kornberg's and Ochoa's enzymes are important, neither is the principal enzyme of either DNA or RNA synthesis.

 

Substrate requirements for synthesis of new DNA

DNA polymerase I catalyzes the addition of a complementary dNTP to the 3'OH end of a polydeoxynucleotide chain. The reaction mechanism is a simple nucleophilic displacement.

24-2 [S31-23]

There are FOUR essential requirements for the activity of DNA polymerase I:

 

1. TEMPLATE

   There must be a template strand to be copied.
   
   

2. PRIMER

   DNA polymerase I (and every other known DNA polymerase) cannot initiate DNA synthesis by itself. It can only extend a pre-existing DNA chain.
   
   

3. 3'OH END

   The reaction mechanism requires that the primer must have a free 3'OH end for synthesis to continue.

   The Sanger method of DNA sequencing actually makes use of this requirement by using low concentrations of dideoxy nucleoside triphosphates (ddNTPs) in a DNA synthesis reaction. ddNTPs are missing the 3'OH group. Once a ddNTP is incorporated into a DNA chain, synthesis will stop.

   Read about the Sanger method of DNA sequencing to see how this is done.

   View the animation of DNA Sequencing by the Enzymatic Method from Prof. Peter Chandler's course on Recombinant DNA.

   
   
   

4. deoxynucleoside triphosphates
   
   dNTPs must be present - usually as the Mg++ salt.
   

The following question asked on the midterm examination in Biochemistry 3102, Winter 1992, addresses these requirements:

 

Consider the following DNA molecules (shown schematically):

 

 

 

 

1. Which of the above are substrates for DNA polymerase I?
   
2. In the case of each molecule that is a substrate, draw the reaction product.

Energetics of DNA synthesis

Favourable energetic factors are:

 

• Hydrolysis of the pyrophosphate that is released.
  
• Formation of hydrogen bonds and base-stacking interactions.

 

Unfavourable energetic factors are:

 

• The phosphoric anhydride bond in the dNTP is replaced with a phosphodiester bond in the polynucleotide. The energy balance of this change is slightly unfavourable.
  
• Loss of entropy as the free dNTP is incorporated into the polymer.

Continuity of DNA synthesis

In thinking about the mechanism of synthesis of DNA polymerase I (or any DNA polymerase, for that matter), we must also consider what happens after each dNTP has been incorporated.

There are two possibilities:

 

• After each nucleotide addition, the enzyme-template-primer complex dissociates.
  
  A new complex will need to form between enzyme, template and primer - with the 3'OH end properly positioned for catalysis - before the next dNTP can associated and be joined to the growing polynucleotide chain.
  
  This mode of synthesis is called distributive synthesis.
  
  
• After each addition, the enzyme moves or slides one place along the template-primer.
  
  In this case, the enzyme can be viewed as sliding along the DNA after each dNTP has been incorporated. This mode of synthesis is called processive synthesis.
  
  

In reality, the processivity or distributivity of any DNA polymerase is an intrinsic property of each enzyme or enzyme complex and it will vary for each case. No enzyme is completely processive and none is completely distributive.

The length of Okazaki fragments depend primarily upon the processivity of the corresponding DNA polymerase.

DNA polymerase I

DNA polymerase I is a 928 amino-acid polypeptide (MW=103118) encoded by the polA gene. It has three distinct enzymatic activities:

 

• A 5' -> 3' polymerase activity
  
• A 3' -> 5' exonuclease activity [24-7] [S31-24]
  
• A 5' -> 3' exonuclease activity [24-8] [S31-25]

Structurally, the enzyme has two domains which can be separated by proteolysis with trypsin or subtilisin, both of which cleave at the amino acid 324. The smaller (34 kD) N-terminal fragment carries the 5' -> 3' exonuclease activity. The larger (76 kD) C-terminal fragment carries the polymerase and 3' -> 5' exonuclease activities. This fragment is often called Klenow fragment after Hans Klenow who first discovered this proteolytic cleavage.

NiceProt View of E. coli DNA polymerase I in the SWISS-PROT Database (entry P00582)

 

S31-26

24-10a 24-10b

       
       

The images show the crystal structure of Klenow fragment (top row) and Klenow fragment complexed with DNA (bottom row) solved to 2.8 and 3.2 Å, respectively. I have tried to present the same views in each pair of images. Note that the top images are missing part of the polypeptide backbone. However, you should still be able to make out the "thumb", "palm" and "fingers" of the protein. The image on the left shows the 3' -> 5' exonuclease proofreading domain on the right hand side; the images in the centre and on the right show the "thumb" and "fingers" of the DNA polymerase domain very clearly. The DNA molecule shown in the bottom images is complexed in the proofreading site.

View the The Bacteriophage T7 DNA Replication Complex from the Online Macromolecular Museum.

Klenow fragment is the most commonly used DNA polymerase in molecular biology labs (at least until Taq polymerase came along!). It is used in reactions to label DNA fragments for hybridization analyses.

The 3' -> 5' exonuclease activity of DNA polymerase is not simply due to the catalysis of the reverse polymerase reaction but is a separate and distinct enzymatic activity that has been mapped to its own active site in the enzyme.

figure from: http://www.biochem.ucl.ac.uk/bsm/xtal/teach/repl/klenow.html

 

The function of the 3' -> 5' exonuclease activity is that of PROOF-READING. Any nucleotides -- such as transiently pairing tautomers -- that are incorrectly incorporated are excised by this activity.

DNA polymerase I is not the replicative polymerase.

Until 1969, DNA polymerase I was the only known DNA polymerase in E. coli. However, as DNA polymerase I was characterized further, it became clear that the properties of this enzyme were not suitable for an enzyme with a central role in DNA replication:

 

• The enzyme is too slow!
  
  DNA polymerase I catalyzes the incorporation of dNTPs at a maximal rate of 20 nt/sec (approximately). At this rate, it would require approximately 460,000 sec (= 7667 min = 128 hr = 5.3 days) to replicate the entire E. coli chromosome! This is much too slow for an organism which can divide every 20 mins.
  
  
• The enzyme is too abundant
  
  There are approximately 400 molecules of DNA polymerase I per E. coli cell. This is excessive given that there are generally only 2 replication forks per cell.
  
  
• The enzyme is not processive enough
  
  DNA polymerase I dissociates after catalysing the incorporation of 20-50 nucleotides.
  
  
• DNA polymerase I cannot initiate DNA synthesis de novo.
  
  However, it must be noted that DNA polymerase I is not unique in this regard. It shares this particular problem with every other known DNA polymerase.
  
  
• Cells containing the polA1 mutation are viable.
  
  In 1969, De Luca and Cairns reported the isolation of a mutant of E. coli that was viable even though it lacked DNA polymerase I activity. The mutant was found to be an amber mutant; we now know from sequence information that the codon for tryptophan 342 is mutated to a stop codon in this mutant.
  
  S31-26
  
  Cells carrying this mutation grow at normal rates. However, they are more sensitive than normal to mutagenic agents such as UV light.
  

Because the polA1 mutants grow normally, they were used as in the search for other DNA polymerase activities in E. coli. Two were found: DNA polymerase II and DNA polymerase III.

 

DNA polymerase II

This enzyme is most likely involved in DNA repair systems. The enzyme is 89921 kD in size and is coded by the polB gene. Strains lacking the gene show no defect in growth or replication. Synthesis of PolII is induced during the stationary phase of cell growth. This is a phase in which little growth and DNA synthesis occurs. It is also a phase in which the DNA can accumulate damage such as short gaps, which act as a block to PolIII. Under these circumstances, PolII helps to overcome the problem because it can reinitiate DNA synthesis downstream of gaps. PolII has a low error rate but it is much too slow to be of any use in normal DNA synthesis.

 

DNA polymerase III

The DNA polymerase III holoenzyme is the principal replicative enzyme in E. coli. This enzyme is highly processive and catalyses polymerization at a high rate. The enzyme is a complex of 10 polypeptides.

MVH24-19

There are two forms of the enzyme.

 

Core enzyme

 

The core enzyme consists of only those subunits that are required for the basic underlying enzymatic activity. In the case of DNA polymerase III, the core enzyme consists of three subunits: alpha (a), epsilon (e) and theta (q).

• The alpha (a) subunit, which is coded by the polC gene, is a 129904 kD polypeptide that contains the 5' -> 3' DNA polymerase activity but no proofreading avtivity.
  
  
• The epsilon (e) subunit, which is coded by the dnaQ gene is a 27099 kD polypeptide that contains the proofreading 3' -> 5' exonuclease activity.
  
  
• The function of the 8846 kD theta (q) subunit, which is encoded by the holE gene is not known. It is not essential for growth and is generally not conserved in bacteria.
  
  

Although the core enzyme can catalyse DNA synthesis, it is not processive -- only about 10 -15 nt are incorporated at a time.

 

 

Holoenzyme

 

The holoenzyme is the fully functional form of an enzyme, complete with all of its necessary accessory subunits. The DNA polymerase III holoenzyme consists of the core enzyme (described above), the b sliding clamp and the clamp-loading complex.

 

the b sliding clamp

The beta (b) subunit is essential for processivity - as long as it is present, the enzyme has almost unlimited processivity. This 40.6 kD polypeptide, coded by dnaN, has been crystallized. It functions as a 'sliding clamp'. It assembles into a dimer with a circular structure through which a DNA double helix can pass. [24-14a 24-14b]

The following images show different views of the beta subunit dimer. The central hole is evident in all three images. The top left-hand image shows a detailed model of the two polypeptides. The top right-hand image shows the polypeptide backbones of the two subunits. The bottom left hand image shows the secondary structure of the polypeptides. The interior of the dimer is lined with alpha-helices; the exterior is formed by beta sheets. The bottom right hand image shows how the beta polypeptide can form a sliding clamp on DNA.

What kind of amino-acid residues would you expect to find in the alpha-helical portions of this structure? Would you expect any pattern to the occurrence of these residues?

 

the clamp-loading complex

The clamp-loading complex consists of the delta (d) (38.7 kD) and delta (d') prime (36.9 kD), chi (c) (16.6 kD) and psi (y) (15.2 kD) subunits and either or both of the gamma (g) subunit (68.4 kD) and the tau (t) subunit (71.1 kD).

Both the gamma and tau subunits are encoded by tthe dnaX gene. This gene is translated in two different ways. If the gene is translated completely then the tau (t) subunit is synthesized. If translation slips or frameshifts part way through the reading frame then a stop codon is encountered, and the gamma (g) subunit is synthesized.

Both the gamma (g) subunit and the tau(t) subunit are 'motor' ATPases.

The structures of many of these subunits have now been solved and they are beginning to reveal how the different subunits interact and function in the building of a replisome and in its ongoing function.

The d subunit binds to the b subunit and, in concert with the d' and g subunits and with ATP, it catalyses the opening of the b dimer to permit passage of DNA.

 

Image from: F.P. Leu and M. O'Donnell (2001) Interplay of Clamp Loader Subunits in Opening the Sliding Clamp of Escherichia coli DNA Polymerase III Holoenzyme. J. Biol. Chem., 276 (50): 47185-47194,

       

The following image shows models for two forms of DNA polIII that have been characterized in vitro. It is thought that the model containing two polIII core complexes is the active replicase while the model containing a single core complex may have a function in mismatch repair.

Image from: A.E. Pritchard, H.G. Dallmann, B. P. Glover and C. S. McHenry (2000) A novel assembly mechanism for the DNA polymerase III holoenzyme DnaX complex: association of ' with DnaX4 forms DnaX3'. EMBO J. 19(23): 6536-6545

View the E. coli DNA Polymerase III b subunit: the sliding DNA clamp from the Online Macromolecular Museum.

Although the model containing two polIII core complexes is thought to be the active replicase in E. coli, it has been shown that it is an asymmetric dimer. Recent (2001) evidence has shown that once the enzyme has been assembled at a replication fork, the clamp loader complex thereafter functions on the lagging strand only.

In Bacillus subtilis, it has also recently (2001) been found that there are two distinct core polymerase enzymes. One of these catalyses DNA synthesis on the leading strand while the other catalyses DNA synthesis on the lagging strand.

 

The following table summarizes the properties of three of the five DNA polymerases of E. coli.

DNA polymerase IV

Synthesis or PolIV is also induced in stationary phase cells, where it competes with PolII. PolIV is an error-prone DNA polymerase. Because of its error-prone mechanism, PolIV is thought to be responsible for 50% of the adaptive mutations that are observed to arise during stationary phase.

DNA polymerase V

When E. coli is exposed to high levels of radiation or of a mutagen, major damage to the bacterial DNA can occur. The cell responds by inducing a special "last-resort" repair pathway called the SOS repair pathway.

Among the SOS genes that are induced are umuC and umuD. The products of these two genes form DNA polymerase V. (Actually, UmuC and two copies of a truncated form of UmuD are required)

[Figure 10-17 from Snyder & Champness, Molecular Genetics of Bacteria]

PolV replicates past gaps in the DNA. DNA synthesis by PolV is error-prone; there are none of the proofreading functions of PolIII. It is a relatively poor polymerase that synthesises DNA distributively. PolV also requires the b subunit and g complex of PolIII for optimal activity. b, which functions as the sliding clamp, is required for processivity and g is the clamp loader.