Given the poor molar yields of typical next gen library construction methods, I have often wondered whether end-repair after shearing was an issue. There are a number of places along the phospho-deoxyribose backbone of a DNA strand where a shearing method might break the strand. Basically you have a repeating chain of (beginning at the 3' carbon and moving 5'):
C-C-C-O-P-O-C-C-C-O-P-O- etc.
So you could have C-C, C-O or P-O bonds "rupturing" to break the chain. I don't think any of the C-C breaks would be repairable by the typical end-polishing mixture of T4-DNA polymerase + T4-polynucleotide kinase. How about the others?
P-O rupture would likely result in one phosphoryl end and one hydroxyl end via the equivalent of hydrolysis. If it happens to be a 5' phosphate and a 3' hydroxyl, then no repair at all is needed--you have ends ligatable by T4-DNA ligase. If it is a 5' hydroxyl and a 3' phosphate then T4-polynucleotide kinase should kinase the 5' end and phosphatase the 3' end. The resulting ends, again, are substrates for T4-DNA ligase.
C-O rupture, as long as it leads to a hydrolysis-like result, would yield a similar prognosis. However, (see below) the result may not result in the restoration of the hydroxyl group via "solvolysis".
What type of ends do various shearing methods leave?
This question was addressed a paper that not only predates DNA sequencing of any kind, but also usage of agarose gel eletrophoresis to size DNA molecules:
Oliver C. Richards and P. D. Boyer, Chemical Mechanism of Sonic, Acid, Alkaline and Enzymic Degradation of DNA, J. Mol. Biol. (1965) 11, 327-340.
The take home was that sonication produced approximately 90% C-O bond rupture and 10% P-O bond rupture and no detectable C-C bond rupture, nor detectable removal of nitrogenous bases. The DNA sheared was phage T2 genomic. The authors give a MW of 1.6x10^8 daltons, or approximately 185 kbp, using 650 daltons as the MW of a basepair. They shear the DNA to an average of 492 +/- 92 bp in the first experiment and 492 +/- 15 bp in the second experiment. Here is there description of the sonication:
The analytic methods used in the paper seem a little arcane to the modern reader. All molecular weight determinations were done via sedimentation velocity. Sites of bond rupture were determined by amount of incorporation of 18O oxygen isotopes into terminal phosphates during shearing as assayed by mass spec on inorganic phosphate released by alkaline phosphatase. But once you get past that, the paper is approachable enough.
The authors evidence against C-C bond rupture was that this type of break would not result in a terminal phosphate that could be released by alkaline phosphatase. But alkaline phosphatase did release phosphate in molar quantities equal to the number of fragment ends after shearing. Further, P-O bond rupture would result in 18O capture into terminal phosphates via "solvolysis". C-O bond ruptures, whether they proceed through hydrolysis or not, would not result in 18O being incorporated into the terminal phosphate. Two trials with sonication resulted in (8%, 12%, respectively for experiments 1 and 2) of the cleavage occurring at a P-O bond.
The figures for acid, base and DNase hydrolysis of DNA were 20%, 33% and 127%, respectively.
Should all sonication-sheared ends be T4-DNA ligase ligatable after T4-poly/T4-PNK end polishing? Based on the evidence presented by the authors, the majority of sonication breaks would be produced at C-O bonds. The author write:
However, later in the discussion they continue:
They then consider and discard the possibility of a beta-elimination because it would result in nitrogenous base release--which they did not detect. They continue:
If I am interpreting this correctly, you could end up with a double-bond in the ribose backbone and no 3' hydroxyl. That, I would hazard, would not be ligatable by T4-ligase. Were it to form the terminus of a 3' overhang, it is conceivable that T4-polymerase's 3'-5' ssDNA exonuclease activity would remove it. But my guess is that it would not.
--
Phillip
C-C-C-O-P-O-C-C-C-O-P-O- etc.
So you could have C-C, C-O or P-O bonds "rupturing" to break the chain. I don't think any of the C-C breaks would be repairable by the typical end-polishing mixture of T4-DNA polymerase + T4-polynucleotide kinase. How about the others?
P-O rupture would likely result in one phosphoryl end and one hydroxyl end via the equivalent of hydrolysis. If it happens to be a 5' phosphate and a 3' hydroxyl, then no repair at all is needed--you have ends ligatable by T4-DNA ligase. If it is a 5' hydroxyl and a 3' phosphate then T4-polynucleotide kinase should kinase the 5' end and phosphatase the 3' end. The resulting ends, again, are substrates for T4-DNA ligase.
C-O rupture, as long as it leads to a hydrolysis-like result, would yield a similar prognosis. However, (see below) the result may not result in the restoration of the hydroxyl group via "solvolysis".
What type of ends do various shearing methods leave?
This question was addressed a paper that not only predates DNA sequencing of any kind, but also usage of agarose gel eletrophoresis to size DNA molecules:
Oliver C. Richards and P. D. Boyer, Chemical Mechanism of Sonic, Acid, Alkaline and Enzymic Degradation of DNA, J. Mol. Biol. (1965) 11, 327-340.
The take home was that sonication produced approximately 90% C-O bond rupture and 10% P-O bond rupture and no detectable C-C bond rupture, nor detectable removal of nitrogenous bases. The DNA sheared was phage T2 genomic. The authors give a MW of 1.6x10^8 daltons, or approximately 185 kbp, using 650 daltons as the MW of a basepair. They shear the DNA to an average of 492 +/- 92 bp in the first experiment and 492 +/- 15 bp in the second experiment. Here is there description of the sonication:
Sonic irradiation of DNA was performed with a Branson 20 kc/sec probe-type sonic oscillator on solutions in SSC/10 at a concentration of 0.2 to 0.5 mg DNA/ml . The container with the DNA solution was immersed in an ice-water bath and sonic treatment was usually performed for periods of from 2 to 15 min at peak power of the probe in an air atmosphere or in a nitrogen atmosphere (after prior bubbling with nitrogen for a period of 5 min). Batches of 50 ml. in a 100-ml. beaker, or batches of 7 ml. in the thimble attachment for the Branson sonic oscillator were used.
The authors evidence against C-C bond rupture was that this type of break would not result in a terminal phosphate that could be released by alkaline phosphatase. But alkaline phosphatase did release phosphate in molar quantities equal to the number of fragment ends after shearing. Further, P-O bond rupture would result in 18O capture into terminal phosphates via "solvolysis". C-O bond ruptures, whether they proceed through hydrolysis or not, would not result in 18O being incorporated into the terminal phosphate. Two trials with sonication resulted in (8%, 12%, respectively for experiments 1 and 2) of the cleavage occurring at a P-O bond.
The figures for acid, base and DNase hydrolysis of DNA were 20%, 33% and 127%, respectively.
Should all sonication-sheared ends be T4-DNA ligase ligatable after T4-poly/T4-PNK end polishing? Based on the evidence presented by the authors, the majority of sonication breaks would be produced at C-O bonds. The author write:
If, as seems likely, sonication breaks covalent bonds in the polynucleotide chain, C-O bond rupture must be predominant. Probably this occurs by solvolysis, with the introduction of water into the alcoholic group formed.
C-O cleavage may occur, however, without the uptake of medium oxygen by an elimination process.
Elimination not dependent upon a free carbonyl group remains possible, however. In this regard, Horwitz, Chua, Klundt, DaRooge & Noel (1964) have reported a base-catalysed elimination reaction of the pyrimidine nucleoside, 3'-O-mesyl-5' -O-trityl-2'-deoxyuridine, which results in the introduction of a 2',3'-double bond in the carbohydrate moiety. The uracilyloxy group behaves as the leaving group with the intermediate formation of bridge oxygen between the 3' position of the carbohydrate and the 2-position of uracil.
--
Phillip
Comment