As I am currently reading a chapter concerning the flow of genetic information, I found out how the cell is responsible for the maintenance of the DNA, an essential feature without whom the cells would suffer chaotic mutations and potentially hazardous transformations.
It turns out, that there are two ways through which the DNA gets damaged. First, there is the danger of "misspelling" a nucleotide during replication. For this, the cell has a system of post-replicative mismatch repair that corrects the eventual errors appeared in the replication of DNA by polymerase.
During cellular division, DNA replication takes place at a speed of about 50 nucleotides per second, leaving almost no room for error. In other words, post-replicative mismatch repair is the "proof-reading" technique used by the cell. Incompatible pairs of nucleotides are identified, the parent strand is kept, while the daughter strand is cut and re-synthesised by DNA Polymerase.
With this "proof-reading" technique, errors still slip at a rate of about 1/107. Still, there are other factors that contribute to the damaging of the chromosomal information, such as UV degradation, normal cellular activity, hazardous chemicals, etc.
For instance, UV damage causes the dimerization of adjacent pyrimidinic nucleotides in a process as follows:
The dimers are removed and replaced by excision repair using a special protein (Excision Repair Nuclease) that identifies the affected area (by geometric abnormal conformation), excises the malfunctioning DNA, fills the gap using a polymerase and then joins the bits together with a ligase.
Another common form of degradation of DNA is the transformation tautomerization of nucleotides of by their natural degradation in the cell.
In this case, the DNA strands are repaired by employing a technique called Base Excision Repair (B.E.R.). In the process the damaged base is first removed. Using the aid of an endonuclease, a gap is made into one of the two strands that are comprised within the helix structure. An exonuclease will enlarge the gap, also removing adjacent nucleotides in the process. Then, DNA polymerase fills the gap, and ligases bind the fragment into place.
In conclusion, given the 1/109 incidence rate of errors in DNA, every human has a rate of about 3.2 mutated or erroneous nucleotides per the entire genome, which is very little given the fact that all proof-reading is performed through only these three above mentioned mechanisms at incredible speeds.