Optimized Welsh DDRT-PCR protocol.

Mirta Grifman mirtag at leonardo.ls.huji.ac.il
Wed Nov 8 12:20:49 EST 1995

As many people asked me to send them the optimized Welsh protocol, I am=20
posting it to the newsgroup. Although it worked on almost everybody's=20
hands (including students in our lab course!), success is not guarantied!
This is a paper to be published in Neuroscience protocols (without the non-=
relevant parts)

 RNAzolTM B from BIOTECX Laboratories (Houston,TX), MicrocarrierTM from=20
Molecular Research Center, Inc. (Cincinnati, OH).
 Reverse transcription (RT) buffer from GIBCOBRL (Gaithersburg, MD),=20
deoxynucleotide triphosphate mixture (dNTPs) from Pharmacia Biotech=20
(Piscataway, NJ), dithiothreitol (DTT, GIBCOBRL), RNase inhibitor,=20
Boehringer (Mannheim), SuperscriptII reverse transcriptase, (GIBCOBRL).
 Taq DNA polymerase and buffer, (Boehringer), a [32P]-dATP=20
(3000Ci/mmol), Amersham International (Amersham).
 Loading buffer (90% deionized formamide, 10mM EDTA, pH 8.0, 0.025%=20
bromophenol blue and 0.025% xylene cyanol), acrylamide, bis-acrylamide,=20
urea (molecular biology grade), Whatmann 3mm paper, Tracker TapeTM=20
(Amersham International), x-ray films.
 STET buffer (8% w/v sucrose, 0.1% w/w Triton X-100, 50mM EDTA, pH 8.0, 50m=
Tris-HCl, pH 8.0), lysozyme, cetrimide (CTAB), agarose.

A. RNA preparation
=09Any type of RNA preparation may be used if the RNA obtained is=20
undegraded. RNA extraction with RNAzolTM B is rapid (2 hours) and simple=20
to perform. The yield can be improved by adding MicrocarrierTM to the=20
homogenization step. In order to overcome minor changes arising from
 differences between individuals, it is very=20
important to pool tissue samples from at least three experimental=20
animals. Special care should be taken to avoid RNA degradation, thus the=20
homogenization step should be carried out rapidly and external RNase=20
sources must be avoided. Other RNA extraction procedures such as the=20
guanidinium thiocyanate-cesium chloride method2 or cytoplasmic RNA=20
extraction4 are also recommended. Poly (A)+ mRNA purification from the=20
total or cytoplasmic RNA can also improve the results, as amplification=20
of mitochondrial RNA or of partially degraded transcripts is avoided.=20
However, It was noted by some researchers that the use of mRNA can cause=20
problems in the DD reaction, since poly dT can carry over into the PCR=20

B. Differential Display
I. First strand cDNA synthesis.
In this step the RNA is reverse transcribed using a primer whose sequence=
is arbitrarily chosen. The enzyme reverse transcriptase will extend those=
primers which have annealed to RNA molecules to yield single stranded cDNAs=
Incubate the RNA samples at 65=A1C for 10 min and transfer immediately to i=
Reaction mixture.=09=09=09=09
Stock solutions=09=09=09=09=09=09Final concentration
2microl 5X RT buffer=09=09=09=09=09=091X
1microl 100mM dNTPs=09=09=09=09=09=0910mM
1microl 100mM DTT=09=09=09=09=09=0910mM
1microl 10microM oligodeoxynucleotide (17-mer)* =09=091microM
0.25microl RNase inhibitor (40U/microl)=09=09               =091U/microl
0.25microl SuperscriptII RT (200U/microl)=09=09=095U/microl
1microl RNA (0.5mg/ml)=09=09=09=09=09=0950ng/microl
DDW (double distilled water) to 10microl
* The primer should be chosen so that its G,C content will not exceed=20
60%. Self complementarity must be avoided. It is advisable to test a set=20
of primers against any new RNA source and choose those that yield the=20
highest number of clearly distinct bands.
Place the reaction mixtures in the thermocycler.
Reaction conditions: 37=A1C, 45 min; 95=A1C, 5 min; 4=A1C until use.

II. Second strand cDNA synthesis and PCR amplification:
Reaction mixture:
Stock solutions=09=09=09=09=09=09Final concentration
5microl 10X Taq buffer=09=09=09=09=09=091X
5microl 10microM oligodeoxynucleotide (17-mer)=09=09=091micrM
0.5microl Taq DNA polymerase (5U/microl)=09=09=0950U/ml
10microl cDNA (from the first strand cDNA reaction)=09=09=09
0.5microll a [32P]-dATP (3000Ci/mmol)=09=09=09=090.1mCi/ml
DDW to 50ml
If needed overlay the reaction with two drops of mineral oil.
Place the reaction mixtures in the thermocycler.
Reaction conditions:=20
Second strand cDNA synthesis: 94=A1C, 5 min; 40=A1C, 5 min; 72=A1C, 5 min. =
PCR amplification: 94=A1C, 1 min; 55=A1C, 1 min; 72=A1C, 2 min. (30 cycles)=
72=A1C, 5 min.=20
Reaction mixtures can be stored at -20=A1C until the electrophoresis step,=
but it is suggested that this be performed no later than 72 hours after=20
the reaction to avoid band fading.

C. Gel electrophoresis:
It is advisable to run the samples together with size markers, for=20
example a sequencing reaction, which will yield a ladder of bands of=20
known lengths.
(1) Mix 2.5microl of each reaction mixture with 2.5microl loading buffer.
(2) Boil for 2 min, and transfer to ice.
(3) Spin down to collect the sample and load on a standard 4%=20
polyacrylamide, 50% urea, sequencing gel.
(4) Electrophorese at 2500V until the xylene cyanol dye reaches the=20
bottom of the gel.
(5) Dry the gel on a piece of Whatmann 3mm paper at 80=A1C, using a gel dri=
(6) Add orientation markers with either radioactive ink or Tracker TapeTM=
and expose the dried gel to autoradiography (overnight should suffice for=
the appearance of strong bands)
(7) Develop the film.
(8) Search for differentially displayed bands among the samples. We=20
consider a band that differs from control samples in the same manner in=20
both duplicate reactions from 3 experiments or animals to be reliable.

D. Isolation and identification of differentially displayed bands.
I. Reamplification (should be planned for the same day as the ligation):
(1) Align the x-ray film with the dried gel adhered to Whatmann paper,=20
cut out the bands of interest and incubate in 100microl DDW at room=20
temperature for 10 min.
(2) Boil for 15 min to allow the DNA to diffuse out of the gel.=20
(3) For DNA precipitation, mix:
100microl diffused DNA
250microl ethanol
11microll 3M Na acetate, pH 5.2
2mcrol MicrocarrierTM or 5microll of 10mg/ml glycogen.
(4) Centrifuge for 30 min at 4=A1C.
(5) Carefully decant supernatant .
(6) Add 500microll 75% ethanol and repeat steps (4) and (5).
(7) Dissolve in 10microll DDW.
Reamplification reaction mixture:
Stock solutions=09=09=09=09=09=09Final concentration
5microl 10X Taq buffer =09=09=09=09=09=091X
1microl 40mM dNTPs (10mM each)=09=09=09=09=09200microM each
5microl 10microM oligodeoxynucleotide (17-mer)=09=09=091microM
3microl DNA (from step 7)
0.5microll Taq DNA polymerase (5U/ml)=09=09=09=0950U/ml
DDW to 50microl
If needed overlay the reaction with two drops of mineral oil.
(9) Place samples in the thermocycler.=20
Reaction conditions: 94=A1C, 2 min (1 cycle); 94=A1C, 1 min; 55=A1C, 2 min =
72=A1C, 1 min (30 cycles) and 72=A1C, 5 min (1 cycle). Keep at 4=A1C until =
ligation step.

II. Cloning:
One efficient way to clone rapidly a PCR product is to use the TA-=20
Cloning Kit from Invitrogen. This kit takes advantage of the fact that=20
Taq polymerases add a single deoxyadenosine at the 3' ends of the PCR=20
products. The vector provided contains a 3' T-overhang ready for the=20
insertion of the PCR products. Note that ligation efficiency will sharply=
drop upon prolonged storage or freezing of the PCR product because the=20
A-overhand might be lost.
The kit also contains competent cells, and can be used virtually as=20
instructed. After the ligation and transformation steps, bacteria are=20
seeded on X-gal containing agar plates, allowing for blue-white=20
screening. Choose several (12-20) white colonies and extract DNA to=20
verify the insertion of the PCR product in the vector.

III. Sequencing:
In essence, any Sanger-based technique would do. We have employed the=20
"Hot Tub" cycle sequencing kit and have found it to be convenient for=20
obtaining fast sequence information. The following detailed method of=20
preparation of the template for sequencing is adapted for use with this=20
kit. If a kit other than "Hot Tub" will be used, the template should be=20
prepared to suit the specific kit requirements.
Template preparation:
(1) Grow a 1.5ml culture overnight at 37=A1C (shaking) from each colony to=
be analyzed.
(2) Spin the cells and discard supernatant.
(3) Dissolve the pellet in 200microl STET buffer.
(4) Add 4microl of 50 mg/ml lysozyme.
(5) Boil for 45 sec and centrifuge for 10 min at room temperature.
(6) Lift out the pellet with a toothpick and discard it.
(7) Add 8microl 5% CTAB and centrifuge for 5 min.
(8) Discard the supernatant and dissolve pellet in 300microl of 1.2M NaCl.
(9) Add 750 microl cold ethanol and centrifuge for 10 min at room=20
(10) Wash the pellet with 75% ethanol. Dry the pellet and dissolve in=20
20microl DDW.
(11) Electrophorese in a 1% agarose gel containing ethidium bromide to=20
verify the template concentration.
=09For the cycle sequencing reaction follow the kit instructions. It=20
is suggested to use a 100-500 fmol template. When used with the=20
Invitrogen cloning kit, the M13 (-20) forward primer can be used as the=20
sequencing primer.
=09The sequences obtained should be analyzed and compared to=20
databases such as the Genbank or the EMBL, using the University of=20
Wisconsin GCG software package. It is recommended to verify the=20
differential expression by an alternative method, such as RNA blot=20
hybridization, nuclear run-on or quantitative RT-PCR, due to the high=20
number of false positive bands obtained in DD. However, we found that in=20
many cases, less abundant transcripts seen after PCR amplification could=20
not be detected by less sensitive techniques. For this reason we=20
recommend to first clone and then sequence the PCR products of choice, so=
that verification can be done by standard quantitative PCR procedures.=20
Most importantly, the DD method is not quantitative, thus only all or=20
none band appearances should be further analyzed.

In order to identify genes which are expressed specifically in the=20
embryonic brain, total RNA was extracted from 50-100mg samples from whole=
brains obtained from mouse embryos (E14) or adults with the aid of=20
RNAzolTM.. The RNA extracted from three individual animals was pooled and=
subjected to the differential display reaction above using the following=20
primer 1: 5=D5 CCACAGGTCCTACCACT 3=D5 and=20
primer 2:  5=D5CCTCCGCGAGATCATCT 3=D5.=20
It can be observed in Fig. 2 that at least 30 bands were clearly observed=
in each lane. It was found that some bands were embryo specific , e.g.=20
expressed in the embryonic brain but not in the adult brain. Additional=20
bands, found to be differentially expressed in part but not all duplicate=
reactions, may also be observed in the same figure. These were not=20
considered for further study.

Differential display of eukaryotic mRNAs has by now been employed as a=20
research approach for almost three years in many laboratories5,7. It is a=
powerful technique which allows identification and cloning of those genes=
whose levels are subjected to changes upon specific circumstances. This=20
technique enables us to detect genes whose RNA levels were changed.=20
However, it is not possible to predict whether the levels of the=20
corresponding proteins are actually modified. Since only a part of the=20
transcript is amplified, the complete sequence data remains to be=20
determined. Transcript amplification is not absolutely random, some RNA=20
molecules will be preferably amplified over others according to their=20
relative concentration and to whether they contain or not a sequence=20
similar to the primer used for the DD.

A. Troubleshooting:
1. Film blank or nearly blank: RNA may be degraded. Extract RNA avoiding=20
degradation by RNAses (see section B.I), always check the quality of your=
RNA by gel electrophoresis and the concentration by spectophotometry.=20
Verify that the labeled nucleotide used is the correct one (a and not g)=20
and that it is not outdated. Some of the reaction components are=20
especially unstable: e.g primers and nucleotides are very sensitive to pH=
fluctuations , check the acidity of the DDW used. The Reverse=20
Transcriptase might be inactivated if kept for long periods on ice. Film=20
exposure time may not be sufficiently long.
2. Bands smeared
Improper gel preparation and running; gels should be cast using fresh=20
acrylamide solutions and should polymerize within 15 min of pouring,=20
should be run at 40-55=A1C.
3. Number of bands too high or too low; Try a different primer, the=20
pattern of obtained bands is variable (length and number) and depends on=20
primer sequence.=20
4. Results obtained are not reproducible; it is absolutely necessary to=20
run all the DD reactions in duplicate and select for further=20
characterization only those transcripts differentially displayed in both=20
samples. Another fact to be taken into consideration is that strain or=20
individual differences exist. Sample pooling helps to overcome this=20
artifact. Differences can also be found between cells grown in different=20
media or at different rates.
5. No detectable product is obtained after band reamplification; use a=20
sample from your PCR as template for a second round of PCR.
B. Alternative and Support Protocols:
=09The currently presented method is an optimization of the protocol=20
developed by Welsh and colleagues6. It presents the following features as=
compared to the original protocol. Primer length affects both, the number=
of bands obtained in the DD reaction and the length of these products. We=
found that one 17-mer oligonucleotide combined with a higher dNTPs=20
concentration, satisfied both requirements, the fragments obtained are=20
relatively large (typically up to 600bp), and the number of bands is=20
increased. Liang and Pardee3 have developed an alternative widely used=20
method which was further optimized by their group and also by Bauer et.=20
al.1 This protocol uses an oligo-dT anchored primer combined with a short=
(10-mer) arbitrary primer under low stringency conditions. However, using=
a single 17-mer primer allows transcript amplification at a random=20
distance from the 3' end of the untranslated part of the mRNA (because=20
the primer is not oligo-dT based), which increases the probability of=20
obtaining meaningful sequence information from coding regions. Moreover,=20
band length is typically higher. It permits, however the amplification of=
ribosomal and mitochondrial RNAs. The Welsh method is not sensitive to=20
genomic DNA contamination. Because it includes two annealing steps at low=
temperature, so that priming in opposite directions will occur, genomic=20
DNA will remain double stranded at this point, which will exclude it from=
further amplification.
=09In summary, this is a very sensitive and powerful approach for=20
the detection and isolation of known and novel genes whose levels of=20
expression are modulated in specific situations. Its importance is=20
stressed for the analysis of nervous system gene expression, whose=20
variability is enormous due to sequence complexity and the continuous=20
dynamic changes enabling the fulfillment of the highly sophisticated=20
functions, characteristic of nervous system cells.

THAT'S IT, HOPE IT HELPS, would love to get your feedback.
Mirta Grifman                      I  Tel: 972-2-6585450
Department of Biological Chemistry I  Fax: 972-2-6520258
Institute of Life Sciences         I  e-mail:mirtag at leonardo.ls.huji.ac.il
The Hebrew University of Jerusalem I  http://www.ls.huji.ac.il/~mirtag/home=

More information about the Methods mailing list