NucleicAcids Research, 1995, Vol. 23, No. 16 3355-3356
An improved method for generating subtracted cDNA
libraries using phage lambda vectors
Holger Hesse*, Wolf B. Frommer and Lothar WilImitzer
Institut fOr Genbiologische Forschung GmbH, lhnestraBe 63, 14195 Berlin, Germany
Received May 31, 1995; Accepted July 19, 1995
We have used a biotin-based subtraction procedure to enrich a
potato tuber cDNA library for sink-specific clones. The method
uses single-stranded phagemids with directional inserts as both
driver and target. We modified the LZAP II vector for the
directional cDNA cloning and for subsequent subtractive hybrid-
ization. This improved method of subtractive hybridization
circumvents many problems facing standard protocols. The
present method uses the advantage of X-phage cloning which
makes it possible to establish libraries from limited amounts of
tissue. In vivo excision of single-stranded phagemids containing
the cDNA inserts was performed to provide an unlimited source
of DNA for biotinylation, hybridization and subtraction. This
ability to produce unlimited quantities ofDNA for subtraction is
advantageous over other methods which require large amounts of
RNA as the starting point (2,7,8). The phage vector system is
more convenient than a plasmid/colony system (3,6,9) when the
original library is to be re-screened for additional clones. The
XZAP II vector was modified by introducing an adaptor (EXI:
5'-AAT TAT CTC GAG GGC CCG ATC GGC CGA ATT
CGT-3' annealed to EX2: 5'-T CGA ACG AAT TCG GCC GAT
CGGGCCCTCGAG AT-3'; TIB MolBiol, Germany, Berlin) that
destroys the original EcoRI and XhoI sites but contains both
restriction sites in inverted orientation flanking a centralSfilsite.
The resulting vector was designated XPAZ II. Total RNA from
sink (growing) and source (sprouting) tubers was isolated
according to (4) and poly A+RNA was purified by chromatogra-
phy on oligo-d(T)-cellulose (Pharmacia, Type 7). cDNA libraries
were constructed using Uni-ZAP cDNA synthesis system
(Stratagene). The cDNAs from sink and source tubers were
cloned into XZAP II and XPAZ II, respectively (Fig. 1). Both
libraries had a complexity of 3 x 106 p.f.u. with -90%
recombinant clones. The average size of the inserts were in the
range of 1 kb. From both libraries, single stranded (ss) circular
DNA were generated by in vivo excision and single-stranded
DNA was isolated (6). Single-stranded DNA from source tubers
was biotinylated with 1 ,ug/pl long-arm Photoprobe® biotin
(Vector Laboratories, Burlingame, UK; 5). Typically, 2.5 ,ug
ssDNA from the sink tuber library was hybridized to 20-30 jig of
biotinylated ssDNA from source tuber, in 20 ,ul 0.75 M NaCl, 50
mM HEPES pH 7.6, 10 mM EDTA and 0.1% SDS at 65°C for
20 h under mineral oil (Sigma). Subsequently, the mineral oil was
removed and the mixture was incubated with 100 ,ug vectrex-avi-
din (Vector Laboratories) for 5 min at room temperature.
Streptavidin-biotin-DNA complexes were precipitated by am-
Subtract with avidin
DNA Polymerase I
Figure 1. Schematic diagram of the
single-stranded; ds, double-stranded.
library subtraction procedure.
monium acetate to the final concentration of 2.5 M on ice.
Samples were extracted 3 times with avidin and ammonium
acetate. The ssDNA in the unbound fraction (enriched for sink
tuber specific ssDNA) was precipitated in the presence of
glycogen. The subtracted ssDNA was converted to dsDNA prior
to transfection into Escherichia coli using a poly A-tail specific
oligo (XA: 5'-(A)1ICTC GAG-3'; TIB MolBiol) and 10 U DNA
polymerase I, to increase the transformation efficiency. An
aliquot of the double-strand DNA was used to electrotransform
DH5a cells (BioRad Gene Pulser/Pulse Controller). The ratio of
blue versus white colonies is increased in the subtracted library
because ofthe simultaneous enrichment of non-insert containing
ssDNA molecules. Fifty randomly-picked white colonies from
the subtracted libraries were analysedby sequencingthe C-terminus.
*To whom correspondence should be addressed
k-KD 1995Oxford UniversityPress
3356 NucleicAcids Research, 1995, Vol. 23, No. 16
All analysed cDNAs contained a poly A-tail, indicating that the
cDNAs derived from the sink tuber cDNA library. Colonies
(-5000) were picked into microtiter plates and characterized by
differential colony hybridization and, in some cases, were
verified for differential expression by Northern analysis (Fig. 2).
Hybridization withcDNA probes encoding potato tuber proteins,
which were predominantly expressed in sink tubers (1), showed
that <10% of the subtracted cDNA library are represented by
these investigated genes. Almost 65% of the clones are not
detectable while only 1% of the cDNAs presented exclusively
expressed genes in source tubers (data not shown). This included
several novel sink-tuber specific cDNAs. While this work was in
progress, several reports (3,5,6,8,9) describing methods using
photobiotinylated ssDNA species were published. However, the
complementarity ofthe hybridizing species from two directional
cDNA libraries is an improvement over conventional hybridiza-
tion procedures. Use ofdifferent orientatedphagemids is suitable
for obtaining large amounts of biotinylated driver DNA and
conversion of ssDNA obtained in the unbound fraction further
improves the efficiency of the subtraction procedure described
here. The method represents a powerful tool for studying gene
expression even at the level of small cellular populations.
We would like to thank Sabine Hummel for excellent technical
assistance, JosefBergstein and Antje Voigt for photographic work.
We are grateful to Nicholas Provart for correcting our English.
Borgmann, K., Sinha, P. and Frommer, W.B. (1994) Plant Sci. 99, 97-108.
Davis, M.M., Cohen, D.I., Nielsen, E.A., Steinmetz, M., Paul, W.E. and
Hood, L. (1984) Proc. Natl. Acad. Sci. USA 81, 2194-2198.
Figure 2. Examples ofNorthern blots forcDNA clones isolated by subtraction.
Total RNA (50 gg) from sink and source tuber was fractionated by
electrophoresis. The resultant Northern blots were hybridized with random
primer labeled cDNA inserts from clones 29Dl 1, 24H4, 26D3 and 19F5 and
equilized with 18S rDNA probe.
Duguid, J.R., Rohwer, R.G. and Seed, B. (1988) Proc. Natl. Acad. Sci.
USA 85, 5738-5742.
4 Logemann, J., Schell, J. and Willmitzer, L. (1987) Proc. Natl. Acad. Sci.
USA 85, 1136-1140.
Rubenstein, J.L.R., Brice, A.E.J., Ciaranello, R.D. and Denney, D. (1990)
Nucleic Acids Res. 18, 4833-4842.
6 Schweinfest, C.W., Henderson, K.W., Gu, J.-R., Kottarides, S.D. and
Besbeas, S. (1990) Genet. Annal. Techn. Appl. 7, 64-70.
9 Swaroop, A., Xu, J., Agarwal, N. and Weissman, S.M. (1991) Nucleic
Acids Res. 19, 1954.
Sargent, T.D. and Dawid, J.B. (1983) Science 222, 135-139.
Sive, H.L. and St. John, T. (1988) Nucleic Acids Res. 16, 10937.