Biomedical Engineering Reference
In-Depth Information
this can be achieved is to inject mRNA fractions into
Xenopus
oocytes (p. 215) and test them for produc-
tion of the corresponding protein (Melton 1987). See
also the discussion of subtraction cloning on p. 115.
oligo-dT primer as usual, the cDNA is tailed with a
string of cytidine residues using the enzyme ter-
minal transferase. This artificial oligo-dC tail is then
used as an annealing site for a synthetic oligo-dG
primer, allowing synthesis of the second strand.
Using this method, Land
et al
. (1981) were able to
isolate a full-length cDNA corresponding to the
chicken lysozyme gene. However, the efficiency can
be lower for other cDNAs (e.g. Cooke
et al
. 1980).
For cDNA expression libraries, it is advantageous
if the cDNA can be inserted into the vector in the
correct orientation. With the self-priming method,
this can be achieved by adding a synthetic linker
to the double-stranded cDNA molecule before the
hairpin loop is cleaved (e.g. Kurtz & Nicodemus
1981; Fig. 6.7a). Where the second strand is primed
separately, direction cloning can be achieved using
an oligo-dT primer containing a linker sequence
(e.g. Coleclough & Erlitz 1985; Fig. 6.7b). An altern-
ative is to use primers for cDNA synthesis that are
already linked to a plasmid (Fig. 6.7c). This strategy
was devised by Okayama and Berg (1982) and has
two further notable characteristics. First, full-length
cDNAs are
preferentially obtained
because an RNA-
DNA hybrid molecule, the result of first-strand
synthesis, is the substrate for a terminal transferase
reaction. A cDNA that does not extend to the end of
the mRNA will present a shielded 3-hydroxyl group,
which is a poor substrate for tailing. Secondly, the
second-strand synthesis step is primed by nicking the
RNA at multiple sites with RNase H. Second-strand
synthesis therefore occurs by a nick-translation
type of reaction, which is highly efficient. Simpler
cDNA cloning strategies incorporating replacement
synthesis of the second strand are widely used (e.g.
Gubler & Hoffman 1983, Lapeyre & Amalric 1985).
The Gubler-Hoffman reaction, as it is commonly
known, is show in Fig. 6.8.
Preparation of cDNA for library
construction
The cDNA synthesis reaction
The synthesis of double-stranded cDNA suitable for
insertion into a cloning vector involves three major
steps: (i) first-strand DNA synthesis on the mRNA
template, carried out with a reverse transcriptase;
(ii) removal of the RNA template; and (iii) second-
strand DNA synthesis using the first DNA strand as a
template, carried out with a DNA-dependent DNA
polymerase, such as
E. coli
DNA polymerase I. All
DNA polymerases, whether they use RNA or DNA
as the template, require a primer to initiate strand
synthesis.
Development of cDNA cloning strategies
The first reports of cDNA cloning were published in
the mid-1970s and were all based on the homopoly-
mer tailing technique, which is described briefly in
Chapter 3. Of several alternative methods, the one
that became the most popular was that of Maniatis
et al
. (1976). This involved the use of an oligo-dT
primer annealing at the polyadenylate tail of the
mRNA to prime first-strand cDNA synthesis, and
took advantage of the fact that the first cDNA strand
has the tendency to transiently fold back on itself,
forming a hairpin loop, resulting in self-priming of
the second strand (Efstratiadis
et al
. 1976). After the
synthesis of the second DNA strand, this loop must
be cleaved with a single-strand-specific nuclease,
e.g. S1 nuclease, to allow insertion into the cloning
vector (Fig. 6.5).
A serious disadvantage of the hairpin method is
that cleavage with S1 nuclease results in the loss of
a certain amount of sequence at the 5
Full-length cDNA cloning
Limitations of conventional cloning strategies
end of the
clone. This strategy has therefore been superseded
by other methods in which the second strand is
primed in a separate reaction. One of the simplest
strategies is shown in Fig. 6.6 (Land
et al
. 1981).
After first-strand synthesis, which is primed with an
′
Conventional approaches to the production of cDNA
libraries have two major drawbacks. First, where
oligo-dT primers are used to initiate first-strand syn-
thesis, there is generally a 3
-end bias (preferential
recovery of clones representing the 3
′
′
end of cDNA
