Biomedical Engineering Reference
In-Depth Information
that cDNA libraries have been superseded? Despite
the advantages of RT-PCR, there are still reasons for
constructing cDNA libraries. The first reflects the
availability of starting material and the permanence
of the library. A sought-after mRNA may occur in a
source that is not readily available, perhaps a small
number of cells in a particular human tissue. A good-
quality cDNA library has only to be constructed
once from this tissue to give a virtually infinite
resource for future use. The specialized library is per-
manently available for screening. Indeed, the library
may be used as a source from which a specific cDNA
can be obtained by PCR amplification. The second
reason concerns screening strategies. The PCR-based
approaches are dependent upon specific primers.
However, with cDNA libraries, screening strategies
are possible that are based upon expression, e.g.
immunochemical screening, rather than nucleic
acid hybridization (see below).
As discussed above for genomic libraries, PCR can
be used to provide the DNA for library construc-
tion when the source is unsuitable for conventional
approaches, e.g. a very small amount of starting
material or fixed tissue. Instead of gene-specific
primers, universal primers can be used that lead to
the amplification of all mRNAs, which can then be
subcloned into suitable vectors. A disadvantage of
PCR-based strategies for cDNA library construction
is that the DNA polymerases used for PCR are more
error-prone than those used conventionally for
second-strand synthesis, so the library may contain
a large number of mutations. There is also likely to
be a certain amount of distortion due to competition
among templates, and a bias towards shorter cDNAs.
A potential problem with RT-PCR is false results
resulting from the amplification of contaminating
genomic sequences in the RNA preparation. Even
trace amounts of genomic DNA may be amplified.
In the study of eukaryotic mRNAs, it is therefore
desirable to choose primers that anneal in different
exons, such that the products expected from the
amplification of cDNA and genomic DNA would be
different sizes or, if the intron is suitably large, so
that genomic DNA would not be amplified at all.
Where this is not possible (e.g. when bacterial RNA
is used as the template), the RNA can be treated with
DNase prior to amplification to destroy any contam-
inating DNA.
Rapid amplification of cDNA ends (RACE)
Another way to address the problem of incomplete
cDNA sequences in libraries is to use a PCR-based
technique for the
rapid amplification of cDNA ends
(RACE) (Frohman
et al.
1988). Both 5
′
RACE and 3
′
RACE protocols are available, although 3
RACE is
usually only required if cDNAs have been generated
using random primers. In each case, only limited
knowledge of the mRNA sequence is required. A
single stretch of sequence within the mRNA is suf-
ficient, so an incomplete clone from a cDNA library
is a good starting-point. From this sequence, specific
primers are chosen which face
outwards
and which
produce overlapping cDNA fragments. In the two
RACE protocols, extension of the cDNAs from the
ends of the transcript to the specific primers is
accomplished by using primers that hybridize either
at the natural 3
′
poly(A) tail of the mRNA, or at a
synthetic poly(dA) tract added to the 5
′
end of the
first-strand cDNA (Fig. 6.11). Finally, after ampli-
fication, the overlapping RACE products can be
combined if desired, to produce an intact full-length
cDNA.
Although simple in principle, RACE suffers from
the same limitations that affect conventional cDNA
cloning procedures. In 5
′
RACE, for example, the
reverse transcriptase may not, in many cases, reach
the authentic 5
′
end of the mRNA, but all first-strand
cDNAs, whether full length or truncated, are tailed
in the subsequent reaction, leading to the amplifica-
tion of a population of variable-length products.
Furthermore, as might be anticipated, since only a
single
specific
primer is used in each of the RACE
protocols, the specificity of amplification may not be
very high. This is especially problematical where the
specific primer is degenerate. In order to overcome
this problem, a modification of the RACE method has
been devised which is based on using nested primers
to increase specificity (Frohman & Martin 1989).
Strategies for improving the specificity of RACE have
been reviewed (Schaefer 1995, Chen 1996).
′
Screening strategies
The identification of a specific clone from a DNA
library can be carried out by exploiting either the
sequence of the clone or the structure/function of its
