Acta Agronomica Hungarica, 56(4), pp. 435–441 (2008)
DOI: 10.1556/AAgr.56.2008.4.9
0238–0161/$ 20.00©2008 Akadémiai Kiadó, Budapest
LIMITATIONS OF USING DIFFERENTIAL DISPLAY RT-PCR
IN THE CHASE FOR SMOKE-RELATED GENES
V. SOÓS1, A. JUHÁSZ1, E. SEBESTYÉN1, M. E. LIGHT2, J. VAN STADEN2
and E. BALÁZS1
1DEPARTMENT OF APPLIED GENOMICS, AGRICULTURAL RESEARCH INSTITUTE OF THE
HUNGARIAN ACADEMY OF SCIENCES, HUNGARY; 2RESEARCH CENTRE FOR PLANT GROWTH
AND DEVELOPMENT, SCHOOL OF BIOLOGICAL AND CONSERVATION SCIENCES, UNIVERSITY
OF KWAZULU-NATAL PIETERMARITZBURG, SCOTTSVILLE, SOUTH AFRICA
Received: 30 May, 2008; accepted: 26 September, 2008
Differential Display RT-PCR was developed before the genomic era to serve as a
tool in hunting for genes. Nowadays, applications using state-of-the-art techniques to
obtain more information about the whole transcriptome or the genome have rapidly
overtaken DD-RT-PCR. This paper will discuss a few of the major drawbacks and
limitations of using this once highly valued method.
Key words: Differential Display RT-PCR, gene expression, germination, Lactuca
sativa L. cv. Grand Rapids, lettuce, smoke
Introduction
The characterization of regulated gene expression in eukaryotic cells is
essential for studying cell growth and differentiation, as well as for
understanding the molecular mechanisms of stresses and environmental impacts.
Differential display was developed for such comparative studies by allowing the
systematic and non-biased screening of molecular differences at the level of
mRNA expression between or within different treatments or tissues (Liang and
Pardee, 1992). The essence of the method is to amplify the messenger RNA 3'
termini using a pair of anchored oligo-dT primers and a short primer with an
arbitrary sequence. The amplified cDNAs, labelled with fluorescent dye, are
then run on a denaturing polyacrylamide gel and visualized. The side-by-side
comparison of mRNA species from two or more related samples allows the
identification of both up- and down-regulated genes of interest. Originally
described by Liang and Pardee (1992), Differential Display PCR (DD-RT PCR)
is a conceptually attractive technique to examine differential gene expression.
Encouraged by the promise of enhanced sensitivity, thousands of investigators

V. SOÓS et al.
Acta Agronomica Hungarica, 56, 2008
436
have applied this technology. However, few novel genes of interest have been
described, indicating that the method has failed to deliver on its promise.
Increased utilization has identified its limitations, which include: variable (low)
reproducibility (Liang and Pardee, 1992); a significant incidence of false
positives (Tiao et al., 1996; Yang et al., 1996); under-representation and
redundancy of mRNA signals (Linskens et al., 1995); frequent priming by the
G/C rich primer at both ends (Graf et al., 1997; Hadman et al., 1995); a bias
towards high copy number mRNAs (Tiao et al., 1996); and a highly labour-
intensive procedure. In efforts to unravel the mode of action of smoke on
germinating lettuce achenes, Differential Display RT-PCR was used to obtain
additional knowledge on smoke-induced genes. Aerosol smoke and smoke-water
can break dormancy and promote seed germination of many plant species in fire-
prone associations and crops (Adkins and Peters, 2001). To date, little is known
about the possible modes of action and the molecular background of the smoke
effect (Baxter et al., 1994). A series of experiments, including DD-RT-PCR,
microarray, gene and promoter functionalisation tests, were used to elucidate the
fundamental characteristics of smoke action. The present paper will discuss the
major limitations of DD-RT-PCR.
Materials and methods
Plant material and growth conditions
For RNA isolation, achenes (seeds) of Lactuca sativa L. cv. Grand Rapids (150 mg) were
germinated in an illuminated or dark environmental chamber (20°C) on tissue paper placed in Petri
dishes. One batch of seeds was treated with 3 ml water (control) and the dishes were wrapped in
non-transparent aluminium foil (dark). The other batch was treated with 3 ml 1000× diluted smoke
extract and then wrapped (dark). The smoke extract was prepared from burnt Themeda triandra
Forssk. (Poaceae), according to the method outlined in Baxter et al. (1994). Samples were
harvested 3, 5, 7, 9, 12 and 24 h after treatment. The germinated achenes were not removed from
the Petri dishes. Identical conditions were applied for the germination time course tests, except that
200 seeds were used as starting material.
RNA isolation
Total RNA was isolated from germinating Grand Rapids lettuce seeds (germinated seeds
were not removed from samples prior to extraction of RNA) using the Qiagen RNeasy Plant Mini
Kit (Qiagen). The RNA was then treated with RNase-free DNase I (Promega) and mRNA was
purified using an Oligotex mRNA Mini Kit (Qiagen) according to the manufacturer’s instructions.
The concentration of RNA was determined with a Nanodrop ND-1000 spectrophotometer
(NanoDrop).
Fluorescent differential display RT-PCR (FDD)
FDD was performed as described previously in the RNAimage protocol (GenHunter) with
some modifications. First-strand cDNAs were synthesized from each mRNA sample (200 ng)
using three different fluorescein-labelled 3'-anchored oligo(dT) primers (5'-fluorescein-GT12N-3',
n = G, C or A) and the Revertaid First Strand cDNA Synthesis Kit (Fermentas). Dilutions (10×) of
cDNAs were amplified by PCR using combinations of fluorescein-labelled anchored and arbitrary
primers (HAP 1-20, Genhunter). The conditions for PCR were as follows: 94°C for 3 min,
followed by 40 cycles of 94°C for 30 s, 40°C for 2 min and 72°C for 1.5 min, with an additional

CHASE FOR SMOKE-RELATED GENES WITH DD RT-PCR
Acta Agronomica Hungarica, 56, 2008
437
extension step at 72°C for 5 min. Electrophoresis was carried out on sequencing gel apparatus
(Thermo Electron Corporation). Samples were run for 2 h at constant 75 W and detection of the
PCR products was performed with a Typhoon Trio+ imager (Amersham Biosciences). In order to
facilitate the exact excision of bands of interest, the gel was stained with silver nitrate. The 6%
polyacrylamide gel was placed in 10% acetic acid for 20 min with gentle shaking. The gel was
washed with water three times, each for 2 min, and stained in a solution containing 0.1% silver
nitrate and 0.15% formaldehyde for 30 min. After a quick rinse with water the patterns were
revealed by adding the developer (3% sodium carbonate, 0.15% formaldehyde, 0.1% sodium
thiosulphate). The reaction was stopped with 10% acetic acid and the gel was washed with water
three times and dried at room temperature. Differentially displayed bands were excised and eluted
into distilled water, purified and then reamplified by PCR with the appropriate pairs of primers.
The products of reamplification were purified with a PCR purification kit (Qiagen) and subcloned
into the pGEM-T vector (Promega). For each reamplified fragment, several E. coli colonies were
chosen, and inserted fragments from these colonies were amplified by PCR. The sizes of inserts
were determined by a comparison of mobilities with the isolated band of the original FDD
samples. Several independent clones with inserts of the expected size were selected, gel-purified
(Qiagen) and sequenced (ABI 3100 Genetic Analyzer).
Real-Time PCR
Lettuce mRNAs (200 ng) were reverse transcribed with the RevertAid first-strand cDNA
synthesis kit (Fermentas). Real-time PCR was performed with an Applied Biosystems 7500 real-
time PCR system using SYBR Green detection chemistry (Applied Biosystems) and gene-specific
primers. The experiment consisted of three independent biological replicates and all reactions were
performed in quadruplicate. Specific product amplification was confirmed by Tm analysis using
the Dissociation Curve option. PCR efficiency (derived from the log slope of the fluorescence
versus cycle number in the exponential phase of each amplification plot) for all primer pairs
ranged from 95.5% to 98.0%. Lettuce actin (AY260165) and GAPDH (AF162202) were also
selected as potential internal controls and their expression was checked using PCR and microarray
data (data not shown). Based on the preliminary findings, actin was selected and used in further
experiments. The relative ratio of threshold cycle (Ct) values between the actin and the specific
genes and their standard deviations were calculated for each sample. The 3W (3 h water control)
samples were used as calibrators.
In silico analysis of the sequences
Sequence alignments were performed with the ClustalW module of the EBI server
(http://www.ebi.ac.uk).
Results and discussion
Isolation of smoke-induced and repressed genes
A differential display approach was used to isolate genes from
germinating Grand Rapids lettuce achenes whose transcription is affected by
smoke treatment. The cDNA fingerprints of smoke-treated germinating lettuce
seeds were compared with those of water-treated seeds kept in the dark. The
samples were harvested 3, 5, 7, 9, 12 and 24 h after treatment. First-strand
cDNA produced from the RNA extracted from the above-mentioned samples
was amplified with a combination of arbitrary primers (H-AP 1–16) and 12-
nucleotide anchor primers (H-T11-N, N = G, C, or A). A representative
differential display pattern is shown in Figure 1.

V. SOÓS et al.
Acta Agronomica Hungarica, 56, 2008
438
Fig. 1. A typical representation of a DD-RT-PCR gel. The arrow indicates a false positive band
The amplification products of the RNA preparations from one batch of
control seeds treated with water for 3, 5, 9, 12 and 24 h are shown in odd lanes
and the corresponding samples treated with smoke water are loaded in even
lanes. As in Figure 1, 200–250 bands can be seen in every lane, but only a few
of them can be regarded as differentially expressed according to the signal
intensity. In the smoke-water vs. water-treated controls placed in the dark a total
of 27 cDNA fragments with altered patterns were detected, cloned and
sequenced. Most bands were present in all twelve lanes, providing an important
check on the reaction in general, as major differences between the lanes probably
represent a fundamental problem with the reaction. Sixteen out of 27 fragments
showed no true altered gene expression pattern after smoke treatment. The false
positive bands were only present in one sample and no further occurrence was
detected in the repeated experiments, but these would have been classified as a
true positive if the experiment had not been performed in triplicate. All the real
false positives (8 out of 16) were short (150–200 bp) fragments with no
sequence similarity to known genes deposited in databases. The spurious false
positives (the other 8 fragments) were identical or highly similar to known
sequences and showed a treatment-dependent expression pattern, but not in the
triplicates. The importance of this phenomenon is that these genes are not false
positives, because they are truly differentially expressed, but not in response to

CHASE FOR SMOKE-RELATED GENES WITH DD RT-PCR
Acta Agronomica Hungarica, 56, 2008
439
the manipulation under investigation, being affected by other factors. In the
present system, about 20% of all the differentially expressed genes were found to
be of this type, i.e. they are differentially expressed from one sample preparation
to the other (Fig. 1). Real-time PCR confirmed these assumptions, as the
transcript abundance of these genes was not consistent in all the triplicates (Fig.
2). The Real-time PCR validation revealed that 11 fragments were true positives,
as they showed a parallel expression pattern with the DD-gel. Among these
clones, selected for further study (discussed in other publications), ten showed
significant homology to known genes or gene families, while the other showed
similarity to an unknown Arabidopsis gene. Performing all experiments in
triplicate, using different samples from independent experiments, has the
advantage of reducing the incidence of false positives by allowing the
identification of genes that are consistently regulated by the manipulation under
study. In addition, it avoids the isolation of genes that are truly differentially
regulated but whose regulation is not related to the manipulation under study but
to other factors specific to the individuals under study. This approach allows the
reliable identification of genes that are differentially expressed in a quantitative
fashion without being completely absent in any lane. Such quantitatively
regulated genes can be difficult to identify using the conventional single
preparation approach, because a quantitative difference in band strength is often
difficult to distinguish from loading differences between the lanes, even when
the strength of non-regulated bands is taken into account.
Fig. 2. Expression analysis of the fructose-1,6-bisphosphatase gene in response to smoke
treatment. Relative transcript abundance was calculated and normalized with respect to the actin
transcript level. The 3 h control was set as calibrator. Data shown represent mean values obtained
from four independent amplification reactions (n = 4). The experiment was repeated three times
with similar results. On the plots, W indicates water control (kept in dark), S indicates smoke-
treated samples (kept in dark)

V. SOÓS et al.
Acta Agronomica Hungarica, 56, 2008
440
As can be seen, only 11 fragments regarded as true positives were chosen
in connection with smoke action. It is anticipated that far more genes could be
expected to be smoke-related. Because of its limitations the DD approach only
allowed the most pronounced genes involved in smoke action to be separated,
and transcripts with lower abundance remained hidden, making it impossible to
plot possible regulatory pathways. Thus, the high occurrence of false positives
and the evidence that aberrant priming at both the 5′ and 3′ ends results in
competition in the PCR, precluding the detection of messages, even those which
are abundantly expressed (Ledakis et al., 1998), further narrows the number of
possible candidates. Not only the significant incidence of false positives but also
the under-representation and redundancy of mRNA signals are further problems
with the DD-RT-PCR approach. A microarray experiment was conducted on
germinating maize kernels to obtain the high density transcriptome of smoke-
treated kernels (discussed elsewhere). The experiment resulted in far more
possible candidate genes even after filtering the germination-related transcripts
(Table 1). While DD may be successfully applied in some settings, the
accumulating evidence indicates that even after extensive screening is
performed, only the tip of the iceberg has been explored.
Table 1
Up- and down-regulated genes in the maize transcriptome induced by smoke extract
Time (h)
3 6 9 12 24 27
Total 659 1549 1012 2479 1253 3321
Up-regulated 130 711 369 1280 371 1623
Down- regulated 529 838 643 1199 882 1698
References
Adkins, S. W., Peters, N. C. B. (2001): Smoke derived from burnt vegetation stimulates
germination of arable weeds. Seed Science and Research, 11, 213–222.
Baxter, B. J. M., Van Staden, J., Granger, J. E., Brown, N. A. C. (1994): Plant-derived smoke and
smoke extracts stimulate seed germination of the fire-climax grass Themeda triandra
Forssk. Environmental and Experimental Botany, 34, 217–223.
Graf, D., Fisher, A. G., Merkenschlager, M. (1997): Rational primer design greatly improves
differential display-PCR (DD-PCR). Nucleic Acids Research, 25, 2239–2240.
Hadman, M., Adam, B. L., Wright, G. L. Jr., Bos, T. J. (1995): Modifications to the differential
display technique reduce background and increase sensitivity. Analytical Biochemistry,
226, 383–386.
Ledakis, P., Tanimura, H., Fojo, T. (1998): Limitations of Differential Display. Biochemical and
Biophysical Research Communications, 251, 653–656.
Liang, P., Pardee, A. B. (1992): Differential display of eukaryotic messenger RNA by means of
the polymerase chain reaction. Science, 257, 967–972.
Linskens, M. H. K., Feng, J., Andrews, W. H., Enlow, B. E., Saati, S. M., Tonkin, L. A., Funk, W.
D., Villeponteau, B. (1995): Cataloging altered gene expression in young and senescent
cells using enhanced differential display. Nucleic Acids Research, 23, 3244–3251.

CHASE FOR SMOKE-RELATED GENES WITH DD RT-PCR
Acta Agronomica Hungarica, 56, 2008
441
Tiao, G. M., Hudson, K., Lieberman, M. A., Fischer, J. E., Hasselgren, P. (1996): Identification of
altered gene expression in skeletal muscle during sepsis using differential display. J.
Surgical Res., 64, 63–67.
Yang, X., Nakao, Y., Pater, M. M., Pater, A. (1996): Identification of two novel cellular genes
associated with multistage carcinogenesis of human endocervical cells by mRNA
differential display. Carcinogenesis, 17, 563–567.
Corresponding author: E. Balázs
Phone: +36 22 569 521
E-mail: balazs@mail.mgki.hu