Terminal Deoxyribonucleotidyl Transferase 99
i.e., temporarily melted 3' terminus. The melting (or breathing) process of
the primer end, in turn, is dependent on the relative base composition:
Ends that are rich in G:C base pairs are more stable and thus less accessible
to the elongation reaction than A:T-rich ends. As soon as one nucleotide
is inserted, the primer has a 3'-OH overhang, which will be elongated
more easily. As a consequence, a few blunt-end primers will be elongated
to a larger extent than initially expected. Furthermore, TdT binds about
10-fold stronger to a terminal dG or dA than to a dC or dT (18). This again
might lead to a bias in the expected distribution of elongated termini,
particularly if different primer ends are involved. All these effects lead to
more or less unpredictable results in both the fraction of primers elongated
and the number of nucleotides added.
For the extension of restriction fragments with subsequent insertion
into vectors, very long extensions are as harmful as very short ones.
Long stretches of homopolymers within a plasmid or phage vector are
not very stable, and tend to delete or otherwise hinder the vector propa-
gation. Very short stretches or a large fraction of primer ends without
any addition drastically reduce the cloning efficiency. Therefore, a
careful control of the lengths of tails by direct visualization from a gel
is required before the annealing reaction with a complementary tailed
counterpart is performed. This is not necessary for tailing with dGMP,
because this reaction ceases after the incorporation of about 20 nucle-
otides owing to aggregate formation of the newly generated oligo(dG)
stretches (4). Thus, by using dG/dC-tailing, only the lengths of the dC-
tail must be controlled and optimized. For the extension of restriction
fragments, the amount of enzyme relative to DNA ends is a critical
factor. It is conceivable that at a low ratio of enzyme to 3'-OH ends the
enzyme tends to elongate an already elongated primer in preference to
a new one, yielding few primers containing many nucleotides and
many primers without any addition. To ensure that most of the primers
will be elongated, particularly when the reaction is carried out with 3'-
recessive or blunt ends, a 10- to 100-fold molar excess of TdT over
primer ends is necessary (21). Terminal additions to blunt-ended DNA
and 3'-recessive-ended DNA also require a partial melting of such
ends. Therefore, conditions that destabilize DNA duplexes, such as
low salt buffers and/or the replacement of Mg 2÷ by Co 2÷ or Mn 2÷, are
favorable. Likewise, ~, exonuclease can be used to trim the 5' ends and
thereby converting a blunt or 3'-recessive end into a 3' overhang (5).