differential display methods and protocols

Given the fact that primers of nearly any length, with or without anchors, can generate cDNA fingerprints sufticrently reproducible to allow differentially expressed genes to be identifi

a variety of pathological alterations or disease states The rsolation and charac- terization of differentially expressed genes becomes one of the first steps toward the understanding of these important biological questions Differential display (1) and a related RAP-PCR method (2) were developed to more effi- ciently Identify and isolate these genes

The general strategy for differential display (Fig 1) IS based on a combma- tton of three techniques brought together by a concept:

1 Reverse transcrlptlon of mRNA from anchored primers (see Note l),

2 Choice of arbitrary primers for setting lengths of cDNAs to be amplified by the polymerase chain reaction (PCR), each corresponding to part of a mRNA (tags),

3 Sequencmg gels for high resolution of amplified cDNA

The objective IS to obtain a tag of a few hundred bases, which 1s sufficiently long to uniquely identify a mRNA and yet short enough to be separated from others by size Given the fact that primers of nearly any length, with or without anchors, can generate cDNA fingerprints sufticrently reproducible to allow differentially expressed genes to be identified, it may be hard to define what should be a standard protocol for differential display The followmg protocol using one-base anchored primer m combmatron with arbitrary 13-mers (3) IS given as an example to illustrate the methodology

From Methods m Molecular Bology, Vol 8.5 D/fferenf/a/ Dsplay Methods and Protocols

Edlted by P Llang and A B Pardee Humana Press Inc , Totowa, NJ

3

Liang and Pardee

L Reverse banswiption 5’.AAGC3’ (H-TUG)

dNTPs MMLV reverse tfanscnptase

Ampli’lhq DNA poiymelase

GWGAA

G-GU

IIL Denaturing polyacrylamide gel

Fig 1 Schemattc representation of one-base anchored differentral display

14 AmphTaq DNA polymerase, Perkin-Elmer Corporation (Norwalk, CT)

15 a-[33P]dATP (>2000 Wrnmole) or a-[35S]dATP (>l,OOO Ci/mmole) (see Note 2)

16 RNase-free DNase I (10 U/pL)

17 QIAEXrM DNA extraction kit (Qiagen, Chatsworth, CA)

18 pCR-TRAPTM clonmg system (GenHunter Corporation, Nashville, TN)

19 Thermocycler

20 6% denaturmg polyacrylamide gel

2 1 DNA sequencing apparatus

Although individual components may be purchased separately from varrous suppliers, most of them can be obtained in kit forms from GenHunter Corporation

3 Methods

3.1 DNase I Treatment of Total RNA

Purification polyadenylated RNAs is neither necessary nor helpful for dif- ferential display The major pitfalls of using the polyadenylated mRNAs are the frequent contammation of the oligo-dT primers, that give high background smearing in the display and the difficulty m assessing the integrity of the mRNAs templates (4) Total cellular RNAs can be easily purified with one-step acid-phenol extraction method (5) However, no matter what methods are used for the total RNA purification, a trace amount of chromosomal DNA contami- nation m the RNA sample could be amplified along with mRNAs thereby com- phcating the pattern of displayed bands Therefore removal of all contaminating chromosomal DNA from RNA samples is essential before carrying out differ- ential display

1 Incubate 10-100 pg of total cellular RNA with 10 U of DNase I (RNase free) in

10 mMTris-Cl, pH 8.3, 50 mMKC1, 1.5 mMMgC& for 30 mm at 37°C

2 Inactivate DNase I by adding an equal volume of phenolchloroform (3: 1) to the

sample

3 Mix by vortexing and leave the sample on ice for 10 mm

4 Centrifuge the sample for 5 min at 4’C m an Eppendorf centrifuge

5 Save the supernatant, and ethanol precipitate the RNA by adding 3 vol of ethanol

in the presence of 0.3MNaOAC, and incubate at -80°C for 30 mm

6 Pellet the RNA by centrifuging at 4°C for 10 min

7 Rinse the RNA pellet with 0.5 mL of 70% ethanol (made with DEPC-H20) and redissolve the RNA in 20 pL of DEPC-treated HzO

8 Measure the RNA concentration at ODS6s with a spectrophotometer by diluting

1 pL of the RNA sample in 1 mL of HzO

6 Llang and Pardee

9 Check the integrity of the RNA samples before and after cleanmg wtth DNase I

by runnmg 1-3 ~18 of each RNA on a 7% formaldehyde agarose gel

10 Store the RNA sample at a concentratton htgher then 1 pg/$ at -80°C before using for differential dtsplay

1 Set up three reverse transcription reactions for each RNA sample in three microfuge tubes (0 5-mL), each contammg one of the three dtfferent anchored ohgo-dT prrmers as follows For 20 pL final volume 9 4 p.L of dH,O, 4 pI of 5X

RT buffer, 1.6 pL of dNTP (250 I.&‘), 2 pL of DNA-free total RNA (freshly diluted to 0 1 pg/pL wtth DEPC-treated H,O), and 2 pL of AAGCT, ,M (2 $4) (M can be either G, A, or C)

2 Program your thermocycler to* 65°C for 5 mm, 37°C for 60 min, 75’C for 5 mm, 4°C (see Note 3)

3 1 pL MMLV reverse transcrlptase 1s added to each tube 10 mm after at 37°C and mix well quickly by finger tipping

4 Continue mcubation and at the end of the reverse transcription reaction, spm the tube briefly to collect condensation Set tubes on ice for PCR or store at -80°C for later use

3.3 PCR Amplification

1 Set up PCR reacttons at room temperature as follow* 20 p.L final volume for each primer set combmation 10 pL of dH,O, 2 $ of 10X PCR buffer, 1.6 l.iL of dNTP (25 pA4), 2 pL of arbitrary 13-mer (2 CUM), 2 pL of AAGCT, ,M (2 CIM), 2 pL of RT-mix from step 3 2 , 0 2 pL of a-[33P]-dATP (see Note 2), 0 2 p.L of AmpliTaq Mix well by pipetmg up and down (see Note 4)

2 Add 25 pI mineral oil if needed

3 PCR as follows 94°C for 30 s, 40°C for 2 mm, 72°C for 30 s for 40 cycles, 72°C for

5 mm, 4“C (For Perkm-Elmer’s 9600 thermocycler it is recommend that the denatur- anon temperature be shortened to 15 s and the rest of parameters kept the same )

1 Prepare a 6% denaturmg polyacrylamide gel m TBE buffer

2 Let it polymertze at least for more than 2 h before usmg

3 Prerun the gel for 30 mm

4 Mix 3.5 pL of each sample with 2 p.L of loading dye and incubate at 80°C for

2 mm immediately before loading onto a 6% DNA sequencmg gel (see Note 5)

5 Electrophorese for about 3 5 h at 60 W constant power (with voltage not to exceed

1700 V) until the xylene dye (the slower movmg dye) reaches the bottom Turn off the power supply and blot the gel onto a piece of 3M Paper Cover the gel with a plastic wrap and dry tt at 80°C for 1 h Do not fix the gel with methanol/ acetic acid (see Note 6)

6 Orient the autoradiogram and dried gel wtth radioacttve mk or needle punches before exposing to a X-ray film Figure 2 shows a representative differential dts-

Differential Display 7

H-T110 H-T110 H-TIIA H-WA H-TIIA H-TIIC H-TllC

Fig 2 Differential display using one-base anchored oligo-dT primers (7) Four RNA samples from non-transformed cell line Rat 1 and H-ras transformed cell lines rat 1 (ras), T101-4 and Al-5 (lanes from left to right, respectively) were compared by differential display using three one-base anchored oligo-dT primers, AAGCT, ,G, AAGCT, ,A and AAGCT, ,C in combinations with three arbitrary 13-mers, H-API (AAGCTTGATTGCC), HAP2 (AAGCTTCGACTGT) and HAP3 (AAGC’l’l’l’GGTCAG) The mob-l (ZO) and mob-7 cDNA fragments were marked by the right and let? arrowheads, respectively

play obtained with three one-base anchored oligo-dT primers in combinations with three arbitrary 13-mers (3)

1 After developing the film (overnight to 72-h exposure), orient the autoradiogram with the gel

2 Locate bands of interest (see Note 7) either by marking with a clean pencil from underneath the film or punching through the film with a needle at the four corners

Liang and Pardee

of each band of interest (Handle the dried gel with gloves and save it between two sheets of clean paper)

3 Cut out the located band with a clean razor blade

4 Soak the gel slice along with the 3M paper in 100 pL dH,O for 10 mm

5 Boil the tube with tightly closed cap (e.g., with parefilm) for 15 min

6 Spm for 2 mm to collect condensatron and pellet the gel and paper debris Trans- fer the supernatant to a new micromge tube

7 Add 10 pL of 3MNaOAC, 5 pL of glycogen (10 mg/mL) and 450 p.L of 100% EtOH

8 Let sit for 30 mm on dry ice or in a -8O’C freezer Spm for 10 mm at 4°C to pellet DNA

9 Remove supernatant and rinse the pellet with 200 pL we-cold 85% EtOH (you will lose your DNA if less concentrated EtOH is used!)

10 Spm briefly and remove the resrdual ethanol

11 Dissolve the pellet m 10 pL of PCR H,O and use 4 pL for reamplificatron

12 Save the rest at -20°C in case of mishaps

13 Reamplification should be done using the same primer set and PCR conditions except the dNTP concentrations are at 20 piV (use 250 @4 dNTP stock) instead

of 2-4 pA4 and no rsotopes added A 40-pL reaction is recommended for each reactron: 20.4 of pL dH,O, 4 pL of 10X PCR buffer, 3.2 pL of dNTP (250 @4),

4 & of arbitrary 13-mer (2 @?), 4 $ AAGCT, ,M (2 @4) (M can be either G, C,

or A), 4 pL of cDNA template from step 3.2 and 0.4 pL of AmphTaq (5 U/pL)

14 Run 30 pL of the PCR sample on a 1.5% agarose gel stained with ethidmm bro- mide (More than 90% probes should be visible on the agarose gel ) Save the remaining PCR samples at -2O’C for subclomng

15 Check to see rf the size of your reamplified PCR products are consistent with their size on the denaturing polyacrylamide gel

1 Extract the reamplified cDNA probe from the agarose gel using QIAEX kit

2 Use the extracted cDNA as a probe for Northern blot confirmation following the standard protocol (ref 5; see Note 8; Fig 3)

3 Clone the cDNA probe using the pCR-TR4PTM cloning system (see Note 9)

4 Confirmation of differentially expressed cDNA probes can be also carried out more efficiently by “Reverse Northern” dot blot or differential screening of cloned cDNA probes by colony hybrtdization (ref 6; Chapter 8 by H Zhang et

al m this book)

5 Clone the full-length cDNA by screening a cDNA library followmg the standard procedure (5)

4 Notes

1 The initial choice of usmg two-base anchored ohgo-dT primers (1) instead of one-base anchored primers (3) were owing to a historical rather than scienttfic reason The cloned marine thymidine kmase (TK) cDNA originally used as a

C 1234

Mob-7

rRNA

Fig 3 Nucleotide sequences of mob-l (A) and mob-7 (B) cDNA fragments cloned

by differential display The flanking primers are marked by arrow bars and the polyadenylation site is underlined Mob-7 differs from mob-l only by 6 base addition

at the 3’ end of the cDNA (see Note 10) Northern blot analysis with mob-7 cDNA probe (C) The 253 bp mob-7 cDNA was used as a probe to confirm the differential expres- sion of the gene using 20 pg of total RNA from Rat 1 and three transformed deriva- tives Rat 1 @as), T101-4 and Al-5 cells (lanes 1 to 4, respectively) The lower panel is ethidium bromide staining of ribosomal RNAs as a control for equal sample loading

10 Lang and Pardee

control cDNA template had only 11 As m its poly(A) tall It was found that one- base anchored prrmer Tl 1C fatled to amplify the TK 3’ termmus m combmatton wtth an upstream primer specific to TK Extension of one more base from the 3’ end instead of the 5’ end of the anchored primer was a logical Interestingly,

Tl 1CA started to work successfully m PCR to amplify the expected TK cDNA template (1) Later, longer one-base anchored primers that had mismatches at the 5’ ends of the prtmers were shown to be much more efficient for differential display m subdividing the mRNA populations mto three groups (3) One-base anchored primers have significant advantage over the two-base anchored primers

m that the former cuts down the redundancy of priming, elimmates the high back- ground smearing problem for two-base anchored pnmers ending with the 3’ “T” and reduce the number of reverse transcription reactions from 12 to 3 per RNA sample

2 It has been observed that 35S labeled nucleottde origmally used for differential display would leak through PCR reaction tubes (espectally when thin-walled tubes are used) and 33P labeled nucleotide was recommended as the best alterna- tive (9) 33P is not only safer to use but also gives better sensmvtty compared to 35S

3 For the reverse transcrtption reaction, the mmal 65°C incubation is intended to denature the RNA secondary structure The final incubation at 75°C for 5 minis

to inactivate the reverse transcrlptase without denaturing the cDNA/mRNA duplexes Therefore “hot start “PCR is neither necessary, nor helpful for the sub- sequent PCR reactions using cDNAs as templates

4 Make core mixes as much as possible to avoid ptpetmg errors (e g , aliquot RT-mix and AP-primer mdrvidually) Otherwise it would be difficult to pipet 0.2 pL of AmpliTaq Mix well by pipetmg up and down

5 It is crucial that the urea in the wells be completely flushed right before loading your samples For best resolution, flush every 4-6 wells each time during sample loading while trying not to disturb the samples that have been already loaded

6 DNA is acid labile, especially at high temperature when the gel is dried This will affect the subsequent PCR during the reamplificatton of the cDNA fragments to

be analyzed further

7 First tentatively identify those bands that appear to be differentially expressed on the initial display gel Then repeat the RT step and the PCR reactions for these lanes and see if these differences are reproducible before pursumg further It 1s recommended that bands bigger than 100 bp be selected It has been generally observed that shorter cDNA probes have higher probability of failing to detect any signals on the Northern blot

8 It IS recommended that the standard prehybrrdrzatron and hybridrzation condttton

at 42°C be used Wash with 1 X SSC, 0.1% SDS at room temperature for 15 min twice followed by washing with 0.25X SSC, 0 1% SDS at 55-6O”C for 15-30 mm

Do not go over 60°C Expose with intensifying screen at -80°C for overnight

to 1 wk

9 pCR-TRAP cloning system is by far the most efficient cloning method for PCR products that we have tested The pCR-TRAP clonmg system utilizes the third generation cloning vector that features postttve-selection for DNA mserts Only

10 It 1s known that the poly(A) tail of a rnRNA is not always added at a fixed posi- tion downstream of the AATAAA polyadenylatton signal This 1s why both mob-l and mob-7 correspondmg to the same mRNA were detected by the same arbitrary primer m combmatton with different anchored primers

Acknowledgment

We thank GenHunter Corporation for the permtsslon of adapting its proto- cols for Message CleanTM kit and RNAlmage TM kit for differential display The work was supported m part by a Natlonal Institute of Health grant CA61232 awarded to Arthur B Pardee and Peng Llang

3 Liang, P Zhu, W , Zhang, X., Guo, Z , O’Connell, R P., Averboukh, L , Wang,

F , and Pardee A B (1994) Differenttal display usmg one-base anchored ohgo dT primers Nucleic Acids Res 22, 5763,5764

4 Ltang, P Averboukh, L., and Pardee, A B (1993) Dtstrlbutton and cloning of eukaryottc mRNAs by means of differential display* refinements and optimiza- tion Nuclezc Acids Res 21, 3269-3275

5 Ausubel, F , Brent, R , Kingston, R E , Moore, D D., Seidman, J G , Smith, J

A, and Struhl, K (1988) Current Protocols In Molecular Biology, Greene and Wiley-Interscience, New York

6 Zhang, H , Zhang, R., and Ltang, P (1996) Differential screening of gene expres- sion difference enriched by differential display Nuclezc Acids Res 24,2454-2456

7 Trentmann, S M , Knaap, E., Kende, H., Ltang, P., and Pardee A B (1995) Alternatives to 35S as a label for the differential display of eukaryottc messenger RNA Sczence 267,1186,1187

Fingerprinting by Arbitrarily Primed PCR

1 Introduction

PCR using primers of arbitrary sequence can generate a reproducible tinger- print of products from DNA (1,2) Differences m the fingerprint of products generated from DNA of related organism are a result of polymorphisms These polymorphisms proved useful markers for genetic mappmg (3-7), population biology (8-13), epidemiology (Z&19), and even the discovery of the mutator phenotype m cancer (20) When applied to RNA the method is also capable of detecting polymorphisms in expressed transcripts (21) However, more impor- tantly, the method can detect a sample of differentially expressed genes (21,22) This chapter will concentrate on recent efforts in our laboratory to extend and improve the method and ascertain its limitations using relatively “low- tech” solutions available to most laboratories

There are three vital steps common to all RNA tingerprmting experiments: (1) the arbitrarily primed PCR amphfication of a sample of transcripts, (2) the isolation and characterization of differentially amplified products, and (3) the confirmation of differentially expressed products in the system of interest

We will discuss issues pertaining to each of these steps When performed with the correct controls, with efficient analysis protocols, and taking into account what it cannot do efficiently, the method is suitable for identifying a sample of regulated genes m a wide variety of experimental systems

It should be noted that our protocol (RNA arbitrarily primed PCR [RAP- PCR]) differs from the Liang and Pardee Differential Display protocol m that

we use an arbitrary primer m both steps of the PCR reaction, rather than an anchored oligo(dT) in the first step The use of arbitrary primers to define both ends of fingerprmt products allows internal RNA fragments to be sampled,

From Methods m Molecular Bfology, Vol 85 D/fferent/a/ Display Methods and Protocols

Edited by P Llang and A B Pardee Humana Press Inc , Totowa, NJ

13

14 McClelland et al including open reading frames In addition, mRNAs that are not polyadenylated can be sampled, such as some bacterial RNAs (23) We cannot be certain that all the properties of these two protocols are the same, nevertheless, there 1s reason to

believe that the issues we discuss here are equally applicable to both protocols

2 Materials

1 Multipipetor for 5- to 200~pL volumes

2 RNeasy total RNA purificatton kit (QIAGEN, Chatsworth, CA)

3 DNase stock (10 U/uL) (Boehrmger Mannhelm Biochemicals, Indianapolis, IN)

4 RNasm RNase mhibitor, 40 II/& (BMB)

5 2X RT Buffer 100 tiTris, pH 8 3, 100 mA4KC1, 8 mMMgC12

6 MuLV-reverse transcrtptase 1200 U/pL (Promega, Madison, WI)

7 Stocks of all four dNTPs (5 mM)

8 Stocks of primers (100 w (Genosys, Woodlands, TX)

9 [a-32P] dCTP (3000 Ci/mmol; ICN, Costa Mesa, CA)

10 2X Tuq polymerase mixture 20 mM Tris, pH 8 3; 20 mM or 100 mA4 KCl;

8 n-&f MgC12

I 1 AmphTaq polymerase, 5 U/pL (Perkm-Elmer-Cents, Norwalk, CT)

12 AmphTuq polymerase Stoffel fragment, 10 U/pL (Perkm-Elmer-Cetus)

13 GeneAmp PCR System 9600 thermocycler (Perkm-Elmer-Cetus)

14 Formamrde dye solution 96% formamide, 0.1% bromophenol blue, 0 1% xylene cyanol, 10 mA4 EDTA

15 Acrylamrde stock solutrons (40% 19.1 acrylamrde.bzs-acrylamide)

16 Urea

17 Ammonium persulfate (fresh 10% solution)

18 TEMED

19 10X TBE buffer: 90 mM Tris-Borate, 20 mMNa2EDTA, pH 8.3

20 Hydrolmk MDE gel solutions (J T Baker Inc , Phillipsburg, NJ)

21 NaOH dye solution 96% formamide, 0 1% bromophenol blue, 0.1% xylene cyanol, 10 mM NaOH

22 Fluroimager or Aquasol scintillation fluid and a scinhllation counter

3 Methods

3.1 Fingerprinting

1 Total RNA is purified usmg the RNeasy total RNA purification kit (QIAGEN, Chatsworth, CA) Typically, lo6 mammalian cells from cell culture yield 5 pg of RNA m 50 $ For bacterta, a hot phenol extraction may be used (23) This is treated with 0 08 U/pL DNase (plus 0 24 U/$ of RNasm, an RNase inhibitor) at 37°C for 40 mm m 1X RT buffer The RNA is repurtfied using the RNeasy kit The yield is estimated by spectrophotometry and the RNA is diluted to 200 ng/pL

in water If sufficient RNA is available, the quality and concentration of the RNA

is checked by agarose gel electrophoresis before bemg stored at -80°C (see Note

1 for comments on experrmental destgn)

Fmgerprin ting 15

Reverse transcription 1s performed on total RNA at three concentrations per sample (500, 250, and 125 ng per reactton) using an ohgonucleotrde primer of arbitrary sequence of 10-20 nt m length (see Note 2) 5 ,LIL of each RNA is mixed with the same volume of RT reaction mrxture for a 10 pL final reaction contam-

mg 50 mA4 Tris pH 8.3, 50 mA4 KCI, 4 mA4 MgCl,, 10 n&f DTT, 0 2 mM of each dNTP, 2 @4 of first primer, and 16 U of MuLV-reverse transcrrptase (see Notes 3-6)

The first strand cDNA synthesis reaction IS then ramped from room temperature

to 37°C over 5 min, held at 37’C for 1 h, then heated to 94°C for 5 mm to stop the reaction Finally, the resultant first strand cDNA IS diluted fourfold by the addt- tion of 3 vol of water

For second strand synthesis, the diluted cDNA (10 pL) 1s mixed with the same volume of PCR mixture for a 20-pL final reaction contammg 10 mMTrts pH 8.3,

10 rnA4KC1,4 mA4MgC12, 0.2 mA4of each dNTP, 4 pA4of a second primer, 1 &I [u-~~P] dCTP, and 4 U of AmphZrq polymerase Stoffel fragment (see Note 7) Thermocyclmg 1s performed using: 30 cycles of 94°C for 30 s, 35°C for 1 mm,

1 mm ramp to 72°C and 72°C for 2 mm

Amphficatron products (5 pL) are mixed with 15 ,rrL of formamide dye solution, denatured at 68°C for 10 mm, and 2 2 pL IS loaded onto a 5% acrylamrde-50% urea gel, prepared m 1X TBE buffer Electrophoresis 1s performed using a sequencing apparatus at 58 W constant power (about 1500 V) until the xylene cyan01 tracking dye reaches the bottom of the gel (approx 4 h)

After electrophoresrs the gel is transferred to Whatman 3MM paper, and dried under vacuum

The drred gel 1s autoradrographed using Bromax film (Kodak) for 12 h-4 d An mtensrfymg screen 1s not used because this tends to blur the fine details of closely packed PCR products An example IS shown in Fig 1,

One of the major bottlenecks m using RAP-PCR or Differential Display is the need to isolate and characterize the PCR fragments representmg differen- tially amplified PCR products The initial step employed has usually been the same as that used for AP-PCR of DNA (4), namely, cutting the band from the denaturing polyacrylamide gel and reamplifymg the resultmg product How- ever, there is quite often a problem with this approach The “Cot effect” (see Note 6) preferentially amplifies the other products of a stmrlar srze that are copurified with the band of interest Although this IS not a problem in many cases where the band of interest vastly predominates even after partial normal- rzatton during PCR, it can lead to a lot of wasted effort for a subset of products

this problem One way to increase the probability of cloning the correct band from the mixture is to simply mimmrze the number of cycles of PCR, thereby limiting the mass of DNA made and thus the Cot effect Alternatively, the

fingerprinting 17

reamplified material can be separated by single-strand conformation polymor- phisms (SSCP) resolved on a native polyacrylamlde gel (24) This method resolves the denatured reampllfied PCR product by virtue of secondary struc- ture Because the product of interest and the contaminants are of entirely dif- ferent sequence, they migrate to different points on the gel The level of contammatton can be assessed from these gels allowing problematic reampiificatlon mixtures to be rejected for further study

Finally, if enough PCR product is loaded in the initial fingerprmtmg gel then SSCP can be performed on the eluted product of interest without reamplification For this procedure we have found a 2-4 d autoradiography with an intensifying screen is needed to vlsuahze the strands However, this strategy entirely avoids the problems associated with reampllficatlon of the initial PCR product until the product of interest is highly purified on the SSCP gel After removal from the SSCP gel the product(s) are reamplified for 20 cycles or less and either used directly for sequencing, or for cloning then sequencing This latter protocol is presented here

1 When products of Interest are identified the RAP-PCR reactions of interest and suitable control RAP-PCR reactions are loaded in multiple adJacent lanes fol- lowing the fingerprinting protocol m Section 3.1 , steps 6-8 The purpose of this preparative gel is to resolve large quantities of the PCR products of interest (see Note 8) RadIoactive ink or fluorescent markers are attached to the dried gel to allow reorlentatlon of the film with the gel

2 The band of interest IS excised from this preparative gel, and the gel re-exposed

to X-ray film to reveal a clear swathe where the band was excrsed

3 The piece of dried acrylamide containing the band of Interest IS placed in 100 &

TE, and heated to 68°C for an hour then left overnight at room temperature to allow the PCR product to diffuse out

4 The eluted material IS ethanol precipitated The incompletely polymerized acrylamide acts as an effective carrier

5 The radioactive pellet is directly dissolved m 4 pL 10 mA4 NaOH dye solution, denatured for 2 min at 94”C, placed on ice for 5 mm, then loaded on an MDE gel for resolution of single strand conformation polymorphisms

6 After electrophoresls, the SSCP gel IS dried (see steps 6, and 7 in Section 3.1,) and autoradlographed using an intenslfymg screen Because the band of interest

IS of higher intensity than background bands, the darkest band on the SSCP gel corresponds to the desired product Frequently, this band will resolve mto two strands Background bands, of relatively low intensity and of an entirely dlffer- ent sequence, will resolve to many positions on the gel, and are often not visible after short autoradiographic exposure Usually double stranded product runs much faster than the single stranded products of interest An example is shown

in Frg 2

7 Once resolved by SSCP, the product can be excised and eluted from the gel, PCR amplified for 20 cycles or less, and used for direct sequencing or for cloning

We have developed streamlined ways to confirm differential expression One method that has proved both effective and relatively simple IS based on RT-PCR

A patr of spectfic primers with melting temperatures of 60°C or more (typically

18 bases or longer) are derived from the sequence of the RAP-PCR product

be easrly detected, regardless of mtttal abundance

Electrophorests and autoradtography are as m Section 3.1 , steps 6-8

Thts protocol generates a product of the expected size for the transcript of mter- est However, it also generates other products from arbttrartly prtmed PCR Most

of these arbitrary products are derived from RNAs that are not dtfferenttally expressed Such products act as a convenient internal control for the level of amphfication and the quality of the PCR reaction The desired PCR products and

a swathe of control products are each cut from each lane of the gel and counted m scmtillation fluid If a fluortmager is available this would be the preferred tool When the mass of the PCR product from the gene of interest is normahzed against the mternal controls an estimate of the relative expresston of the gene m the dif- ferent samples is obtained One caveat m this method IS that the Cot effect (see Note 6) causes differences observed between products to underestimate the true differences between the startmg RNA samples However, this discrepancy is reduced at lower cycle numbers so the amplificatton IS best sampled at a number

of PCR cycles or at a number of starting RNA concentrations The most useful data is then derived for the lowest number of cycles or lowest RNA concentratton that 1s compatible wtth the method employed for measuring product formation

20 McClellana et al

4 Notes

1 Experimental design: There are multiple steps needed to detect and characterize

a differentially expressed transcript and sampling is biased towards more abun- dant transcripts (see Note 4) Thus, for many questions regarding differential gene expression there may be other methods that are better suited, such as differ- ential screening or subtractive hybridization However, unlike these methods, RAP-PCR allows many RNA samples to be compared in parallel (25) If eight different RNA samples are compared and the levels of each transcript is unchanged or up- or downregulated in each RNA sample then there are almost 3* (more than 6000) possible permutations of gene expression that are surveyed This calculation does not take into account the further division of expression profiles to account for large versus small increase or decreases in gene expres- sion Most of the vast number of observable regulatory categories will not exist but, nevertheless, examples of genes that fall into rather sophisticated regulatory categories can be searched for Furthermore, genes that fall into totally unex- pected categories may be found

2 Longer primers can be used For an 1%mer primer the annealing temperature during PCR is changed to 45°C and Tuq polymerase Stoffel fragment is replaced

by AmpliTaq One advantage of longer primers is that they can easily accommo- date restriction sites for subsequent cloning or can encode “motif’ sequences that may direct priming to transcripts encoding conserved amino acid sequences (26) Even 1 0-mer primers can encode short motifs An example is the Tryp 1- primer (5’-GTGGCGTTGAT) in Fig I that is a conceptual translation of a tyrosine phos- phatase conserved amino acid motif

3 Fingerprinting multiple RNA concentrations: As is true for DNA fingerprints, a few products in each RNA fingerprint can be sensitive to RNA concentration and quality For this reason we perform RNA fingerprints using at least two RNA concentrations that differ by twofold This strategy allows those PCR products that are not reproducible at both RNA concentrations to be eliminated from further consideration, Fingerprinting duplicates of each RNA from each experi- mental condition at the same concentration is less effective because this does not necessarily control for variation that is concentration- and quality-dependent Fin- gerprinting separate RNA preparations for each sample condition is also less effective because differences in RNA quality between different experimental con- ditions are not controlled

4 Sampling efficiency: One of the primary limitations of arbitrarily primed RNA fingerprinting methods is the fact that the probability of being able to visualize a PCR product derived from a particular transcript is a function of both the quality

of the match of the primer with two sites in the template and the abundance of the transcript For example, if two templates have the same match and are equally efficiently amplified but differ by loo-fold in abundance, the ratio of these prod- ucts will remain 100: 1 during most of the reaction As a consequence, less abun- dant transcripts will be more difficult to see on a fingerprint or may not be visible

at all Calculations that assume each transcript can be sampled with equal effi-

Fingerprinting

ctency do not take into account the fact that rarer transcripts will be on average much more difficult to vtsualize Thus, a calculation of the number of finger- prints needed to sample, say, 95% of the transcripts that assumed normalized sampling and equal visibility of products derived from rare transcripts would be

a considerable underestimate of the actual number of fingerprints needed One logical response to these considerations is to ensure that any expertment that

is to be performed using this method does not require efficient coverage of rarer transcripts

5 One method we have used in an attempt to improve sampling of rarer transcripts

is to reamplify the RNA fingerprint with a primer that contains an extra arbi- trarily selected base at the 5’ end (or more than one such base) (26) The Idea is to select a subset of the initial fingerprint for reamplification, Including PCR prod- ucts, which are not visible on the initial fingerprint because they originate from the complex class of rare transcripts When primers of 18 bases m length are used for the initial fingerprinting and nestmgs, this method can generate a new finger- print when the “nesting” includes up to three bases However, the fingerprint becomes less reliable as the nesting length is increased, presumably because prod- ucts of extension from poorer and less efficient matches are also amplified from the background of the fingerprint Interestingly, lo-mer primers are too short for nesting presumably because annealing must take place at so low a temperature that extension occurs from all of the initial fingerprint products, regardless of the 3’ match The method is also useful if the RNA used for fingerprmtmg is so limited that the generation of further fingerprints from the first fingerprint would

be desirable to conserve precious RNA

6 The “Cot effect”: In later cycles of PCR the concentration of the more abundant PCR products is sufficiently high that there is considerable product self-anneal- ing that occurs in the lower temperature steps of the PCR cycles This has the effect of preferentially slowing the amplification efficiency of the more abundant PCR products Two consequences can be expected First, less abundant products have a chance to gain some ground and become visible This advantage of rarer products may slightly mitigate against the fact that rarer products are otherwise less easily seen in fingerprints Second, differences visible between different RNA samples may be partially erased, particularly for the more abundant prod- ucts Thus, surprismgly, differences visible between lanes may actually underes- timate the true differences in abundances of the transcripts We have called this phenomenon of product self-annealing the Cot effect in reference to the depen- dence of annealing on the initial concentration and time

It is interesting to note that intentionally increasing the Cot effect could be desirable in some circumstances For example, if one is interested in “normahz- ing” a mtxture of PCR products Possible application would include sampling the entire complexity of genotypes for an environmental sample A class of genes of interest would be PCR amplified under conditions where homologs would rehybndize, allowing other amplified genes from rarer organisms m the sample

to become relatively enriched in the sample Another example might be the nor-

22 McClelland et al

maltzatton of a cDNA library, although crossover PCR might be a concern In both cases the PCR reaction would be held at 60-85°C for many mmutes, or perhaps even hours, at each cycle to block amphficatton of the more abundant products

7 About 0 5 l&f of the first primer 1s carried over into the second reaction Adding

4 @4 of a second primer for the second strand syntheses results prtmartly m PCR products that have the first primer at the 3’ end of the sense strand of the tran- script and the second primer at the 5’ end Thus, the sense ortentatton of the prod- ucts 1s generally known Thts phenomenon is accentuated by the observatron that PCR products that have the same primer at each end seem to be out-competed by products that have different primers at each end (27)

The use of an arbitrary primer m the RT reaction allows PCR products to be derived from internal parts of a transcript, mcludmg the protein coding region This can help later when determinmg the nature of the transcript Also, transcripts that are not polyadenylated, such as bactertal RNAs, can be sampled by thus method

8 If the origmal fingerprmt IS no longer radioactive then It can be made radioactive again by reamphficatron of 1 pL of the reaction for a further five PCR cycles m

10 & of fresh RAP-PCR amplification mixture

Acknowledgments

Thts work was supported in part by grants A134829, NS33377, and CA68822 from the National Institutes of Health and by a generous gift from Sidney Kmnnel

Rerter, R S , Willrams, J G., Feldmann, K A., Rafalskt, J A , Tingey, S V , and Scolmk, P A (1992) Global and local genome mappmg m Arabzdopszs thalzana

by using recombinant inbred lines and random amphfied polymorphic DNAs Proc Nat1 Acad Sa USA 89,1477-1481

Welsh, J , Petersen, C , and McClelland, M (1991) Polymorphtsms generated by arbttrartly prrmed PCR in the mouse application to strain rdenttficatton and genetic mappmg Nuclezc Aczds Res 19, 303-306

Al Janabt, S M , Honeycutt, R J , McClelland, M., and Sobral, B W (1993) A genetic linkage map of Saccharum spontaneum L ‘SES 208’ Genetics 134,

1249-1260

Birkenmeter, E H , Schneider, U , and Thurston, S J (1992) Fingerprmtmg genomes by use of PCR with prrmers that encode protem mottfs or contam sequences that regulate gene expression [published erratum appears m Mamm Genome 1993,4(2) 1331 Mamm Genome 3,537-545

Fingerprinting 23

7 Mlchelmore, R W , Paran, I , and Kesseh, R V (1991) Identtficatlon of markers lmked to disease-resistance genes by bulked segregant analysts* a rapid method to detect markers m specific genomic regions by usmg segregating populattons

Proc Natl Acad Scz USA 88,9828-9832

8 Mathteu-Daude, F , Stevens, J., Welsh, J., Tibayrenc, M , and McClelland, M (1995) Genetic diversity and population structure of Trypanosoma brucet clonality versus sexuality Mol Blochem Parasltol 72, 89-101

9 O’Rourke, M and Sprat& B G (1994) Further evidence for the non-clonal popu- lation structure of Nezsserza gonorrhoeae extensive genettc dtverstty within ISO- lates of the same electrophoretlc type Mzcrobzology 140, 1285-1290

10 Fukatsu, T and Ishtkawa, H (1994) Dtfferenttatlon of aphid clones by arbttrarlly primed polymerase cham reaction (AP-PCR) DNA fingerprmtmg Mel Ecol 3, 187-192

11 Levttan, D R and Grosberg, R K (1993) The analysts of paternity and mater-t-my

m the marme hydrozoan Hydvactznla symbzolonglcarpus usmg randomly ampll- tied polymorphic DNA (RAPD) markers MoZ Ecol 2,3 15-326

12 Tamate, H B., Shlbata, K , Tsuchtya, T , and Ohtalshi, N (1995) Assessment of genettc variations within populations of Sika deer m Japan by analysts of ran- domly amplified polymorphic DNA (RAPD) Zoolog Scz 12,669-673

13 Chapco, W., Ashton, N W , Martel, R K , Antomshyn, N., and Crosby, W L (1992) A feaslblllty study of the use of random amplified polymorphtc DNA m the population genetics and systematics of grasshoppers Genome 35, 569-574

14 Fang, F C , McClelland, M , Gumey, D G , Jackson, M M , Hartstem, A I, Morthland, V H , Davis, C E , McPherson, D C , and Welsh, J (1993) Value of molecular epldemlologrc analysts in a nosocomtal methlclllm-resistant Staphylo-

coccus aureus outbreak [see comments] JAMA 270, 1323-1328

15 van Belkum, A , van Leeuwen, W , Kluytmans, J., and Verbrugh, H (1995) Molecular nosocomtal epidemiology* high speed typmg of mlcroblal pathogens

by arbitrary primed polymerase chain reaction assays Infect Control Hasp Epldemzol 16,658466

16 Tang, Y J., Houston, S T , Gumerlock, P H , Mulhgan, M E , Gerdmg, D N , Johnson, S , Fekety, F R., and Silva, J J (1995) Comparison of arbitrarily primed PCR with restrmtton endonuclease and lmmunoblot analyses for typmg

Clostrldlum d&f?clle isolates J Clan Mlcroblol 33, 3169-3173

17 Coelho, A., Vicente, A C , Baptista, M A , Momen, H , Santos, F A , and Salles,

C A (1995) The distinction of pathogenic Vtbrto cholerae groups using arbl- trarlly primed PCR fingerprints Res Mlcroblol 146, 67 l-683

18 van Belkum, A , Kluytmans, J , van Leeuwen, W., Bax, R , Qumt, W., Peters, E , Fluit, A , Vandenbroucke Grauls, C , van den Brule, A , Koeleman, H , et al (1995) Multtcenter evaluation of arbitrarily primed PCR for typmg of Staphylo-

coccus aureus strains J Clwz Mlcroblol 33, 1537-1547

19 Madico, G., Akopyants, N S , and Berg, D E (1995) Arbitrarily primed PCR DNA fingerprmtmg of Escherzchza co11 0157.H7 strams by using templates from boiled cultures J Clrn Mzcroblol 33, 1534-1536

24 McClelland et al

20 Perucho, M., Welsh, J., Peinado, M A., Ionov, Y., and McClelland, M (1995) Fingerprinting of DNA and RNA by arbitrarily primed polymerase chain reac- tion: apphcatlons m cancer research Methods Enzymol 254,275-290

21 Welsh, J., Chada, K., Dalal, S S., Cheng, R., Ralph, D., and McClelland, M (1992) Arbitrarily primed PCR tingerprmting of RNA Nucleic Acids Res 20,

24 Hayashi, K and Yandell, D W (1993) How sensitive is PCR-SSCP? Hum Mutat 2,338-346

25 McClelland, M., Ralph, D., Cheng, R., and Welsh, J (1994) Interactions among regulators of RNA abundance characterized using RNA tingerprintmg by arbi- trarily primed PCR Nuclezc Acids Res 22,44 19-443 1

26 Ralph, D., McClelland, M., and Welsh, J (1993) RNA fingerprmtmg usmg arbl- trarlly primed PCR identifies drfferentially regulated RNAs in mink lung (Mv 1 Lu) cells growth arrested by transforming growth factor beta 1 Proc Nut1 Acad Sci USA 90, 10,71O-10,714

27 Welsh, J and McClelland, M (1991) Genomic tingerprmting using arbitrarily primed PCR and a matrix of painvlse combinations of primers Nuclex Acids Res 19,5275-5279

3

Differential Display

Motif Primers, and Agarose Gel Electrophoresis

Patrick J Donohue, Debbie K W Hsu,

and Jeffrey A Winkles

I, Introduction

Polypeptide growth factors stimulate cellular proliferation by bmdmg to the extracellular domam of transmembrane receptors and thereby activating mtra- cellular signal transduction pathways One cellular response to mitogenic stimulation is the sequential transcriptional induction of specific nuclear genes encoding proteins of diverse functions (reviewed in refs 1 and 2) Many of these proteins are likely to be required for DNA replication and cellular division

As an approach to identify novel gene products involved in polypeptide growth factor signaling, we are studying tibroblast growth factor (FGF)- 1 -inducible gene expression m murine NIH-3T3 cells FGF-1 is a member of a family of structurally-related mitogens than can promote cellular proliferation, migra- tion and differentiation (reviewed in refs 3’ and 4) Its biological effects are mediated by protein tyrosine kinase cell surface receptors present on most cell types FGF-1 is likely to be involved in the pathogenesis of several human diseases, including atherosclerosis and cancer

We reported in 1993 that a differential display approach using agarose gel electrophoresis could be used successfully to identify genes expressed follow- ing FGF-1 treatment of serum-starved NEH-3T3 fibroblasts (5) This approach

is conceptually similar to the mRNA differential display and RNA fingerprmtmg methods described by other groups (reviewed in refs 6 and 7) However, in comparison to these approaches, in our method (1) cDNA is synthesized usmg random hexamer primers; (2) polymerase chain reaction (PCR) assays are per- formed using sense and antisense oligonucleotide primers, usually degenerate

From Methods m Molecular Bology, Vol 85 D/fferentral D/splay Methods and Protocols

Edlted by P Llang and A B Pardee Humana Press Inc , Totowa, NJ

25

26 Donohue, Hsu, and Winkles

m sequence, which are designed to amplify cDNA templates encoding proteins with particular structural motifs; and (3) ampllficatlon products are displayed using agarose gel electrophoresls and ethldmm bromide staining (reviewed m ref 8) This method does not require the use of radlolsotopes nor the potent neurotoxms, acrylamlde and bu-acrylamlde However, It is hmlted by the rela- tively low resolution of agarose gels and the Inability of fluorescent dye staining

to detect amplification products derived from relatively rare cDNA templates In our mltlal series of experiments, 30 cDNA fragments were isolated and 25 of these were successfully reamphfied and cloned mto a plasmld vector When used

as probes m Northern blot hybridization experiments, 15 of the 25 cDNAs detected transcripts that were expressed at an increased level m FGF- 1 -stimulated cells DNA sequence analysis revealed that 13 of the 15 cDNAs were unique and that four of the 13 cDNAs were amplified when a single ollgonucleotlde func- tioned as both a sense and antisense primer Furthermore, although our initial goal was to use the motif primers to enrich for differentially expressed members

of particular gene families, the majority of the FGF- l-inducible genes character- ized to date do not encode protems with the targeted motifs ($9-11) This 1s because, under the PCR condltlons used, many of the motif primers were able to anneal to and prime cDNA templates with a relatively low degree of sequence identity However, m at least one case, the targeting aspect of the approach descrtbed here was successful (12) In this report, we outline our basic differential display tech- nique and note vanatlons performed by us and also described by others

2 Materials

2 I RNA Isolation from Tissue Culture Cells

1 RNA STAT-60 (Tel-Test “B,” Friendswood, TX) This solution should be stored at 4°C up to 9 mo and 1s light-sensitive It contains phenol and guamdlmum thlocyanate and should be handled wearing gloves and a lab coat Avold breathmg vapor

2 Chloroform This should also be handled using the precautions described m item 1

4 dNTP mix: 1 25 mM of each dNTP (Boehringer Mannhelm, Indianapolis, IN)

5 RNasm ribonuclease inhibitor (Promega, Madison, WI), 33 U/@ Store at -2O’C (not In a frost-free freezer)

6 Random hexamer (pd[N]& primers (Boehringer Mannhelm), 50 ng/&

Differential D/splay Vauatlons 27

Table 1

Motif Oligonucleotide Primers Used for PCR Amplification

Motif

Ammo acid sequence

DNA strand Primer sequencea b Protein tyrosme IHRDL Sense CGGATCCACMGNGAYYT

kmase DVWSFG Antisense GGAATTCCAWAGGACCASACRTC

Zmc linger GQKPYEC Sense GGNGAGAARCCCTWYGARTG

HQRIHTG Antisense CCHGTGTGARTCCTCTGRTG Leucme zipper LEEKATQL Sense CTGGAGGAGAAGGYGRCCCAGCT

LEEKATQL Sense CTGGARGMNVAGRHSRMSMMGCT LEEKATQL Antisense AGCTGGGYCRCCTTCTCCTCCAG LEEKATQL Antisense AGCKKSKYSDYCTBNKCYTCCAG

Src homology-2 FLVRESET Sense TTCCTGGTGCGGGAGTCTGAGACC

VKHYKIR Sense GTGAAGCACTACAAGATCCGG FLVRESET Antisense GGTCTCAGACTCCCGCACCAGGAA VKHYKIR Antisense CCGGATCTTGTAGTGCTTCAC

aPrlmer sequences are 5’ to 3’, addltlonal nucleotldes used for restrxtlon enzyme recogmtlon are m bold

6Degenerate bases m the primers are abbreviated as recommended by a nomenclature committee (21)

2.3 PCR

1 Taq DNA polymerase (Boehrmger Mannhelm), 5 U/pL Store at -20°C (not m a frost-free freezer)

2 10X PCR buffer 100 mMTris-HCl, pH 8 3,500 mM KCI, 15 mMMgC1,

3 dNTP mix 1 25 mA4 of each dNTP (Boehrmger Mannhelm)

4 Motif ohgonucleotide primers (Table 1); 0 5 ug/pL

2.4 Agarose Gel EIectrophoresis

1 Agarose (Life Technologies)

2 1 OX Tris-acetate (TAE) buffer 400 mA4 Tris-acetate, 10 mM EDTA

3 10X DNA gel loading buffer 50% glycerol, 0 2% bromophenol blue, 0.2% xylene cyan01

4 Ethtdmm bromide (Sigma, St Louis, MO), 10 mg/mL This fluorescent dye is stored at 4°C and is light-sensitive It is a mutagen and may be carcmogemc/ teratogemc, therefore, it should be handled wearing gloves and ethidmm bromide-contammg solutions should be disposed of properly

1 PCR and agarose gel electrophoresis reagents described above m Sections 2 3 and 2.4

2 Plasmid pCRI1 (Invitrogen, San Diego, CA), 25 ng/pL

28 Donohue, Hsu, and Winkles

3 1 OX Ligation buffer 600 mM Trts-HCI, pH 7 5,600 mA4 MgCI,, 500 mM NaCf ,

10 mg/mL bovine serum albumin, 700 mM P-mercaptoethanol, 10 mM ATP,

200 mM dtthtothrettol, 100 mA4 spermidme

4 T4 DNA ligase (Invitrogen), 4 UIuL Store at -20°C (not in a frost-free freezer)

5 E colz DH5a competent cells (Life Technologies) Store in ahquots at -70°C

6 SOC medra 2% bacto-tryptone, 0 5% bacto-yeast extract, 10 mA4NaCl 2 5 mM KCl, 10 mM MgCl,, 10 mM MgSO,, 20 mM glucose This media is prepared by first combmmg the tryptone, yeast extract, NaCl and KC1 and then autoclavmg

A glucose stock solutton (2M) and a Mg2+ stock solution (2M, comprised of 1M MgCl, and 1M MgSO,) are then prepared and sterilized by filtration through a 22-pm membrane Finally, the media, glucose and Mg2+ are combined and steril- ized by filtration

7 Luria broth* 1% bacto-tryptone, 0 5% bacto-yeast extract, 1% NaCl Sterilize by autoclaving

8 Bacto-agar (Dtfco, Detroit, MI)

9 Ampicillm (Sigma), 50 mg/mL

10 5-Bromo-4-chloro-3-mdolyl-p-n-galactopyranostde (X-gal, Boehrmger Mannhelm),

40 mg/mL in NJ-dtmethylformamtde

3 Methods

3.1 RNA Isolation from Tissue Culture Cells

1 Lyse frozen cell pellets (-5 x lo6 cells; see Note 1) m 1 mL of RNA STAT-60 by repetmve plpeting Incubate at room temperature for 5 mm

2 Add 0 2 mL of chloroform, mix and place samples at room temperature for 5 mm

3 Centrifuge the homogenate at 15,OOOg for 15 mm

4 Transfer 0.5 mL of the upper, colorless aqueous phase to a new tube, add 0 5 mL tsopropanol, vortex, and place samples at room temperature for 10 min

5 Centrifuge at 15,OOOg for 10 min and discard supematant

6 Wash RNA precipitate once with 1 mL 70% ethanol by vortexmg Centrifuge at 15,OOOg for 10 mm Discard supematant

7 Dry RNA pellet briefly for l-2 mm m Savant Speed-Vat Concentrator

8 Dissolve RNA m 20 & autoclaved distilled H20 An mcubatlon at 55°C for 15 min may be required for complete resuspenston

9 Determine the RNA concentration by measurmg the A260 of an ahquot using a

Different/al Display Variations 29

3 Incubate at 37’C for 1 h

4 Terminate reactton by heating at 95°C for 10 mm

5 Add 200 pL of dtstilled H,O and store at -20°C (see Note 4)

3.3 PCR

1 Set up the followmg 50 ~JL reaction by adding the reagents in this order: 5.0 pL cDNA,

5 0 p.L 10X PCR buffer, 8.0 pL dNTP mix, 30.5 l.tL H,O, 0.5 pL sense motif primer (see Notes 5-7), 0 5 l.tL antisense motif pnmer, and 0.5 pL Tuq DNA polymerase

2 PCR 1s performed using the following condittons (see Note 8) and a Perkm-Elmer Model 9600 thermocycler:

Imtial Denaturation 94”C, 5 mm

Stage I(8 cycles) 94°C 30 s denaturation

46°C 30 s annealing 68’C, 30 s extension Stage II (24 cycles) 94”C, 30 s denaturation

58“C, 30 s annealing 72”C, 30 s extension Final primer extension 72”C, 10 mm

3 Store samples at 4°C

3.4 Agarose Gel Hectrophoresis

1 Add 1.8 g agarose per 100 mL 1X TAE buffer (see Note 9) Dtssolve by heating

m mmrowave oven, cool, and pour mto gel tray

2 Prepare 1X TAE running buffer containing 1.5 pg/mL ethidium bromide

3 An aliquot (18 pL) of each amplification mixture is then combined with 2 $ of

1 OX DNA gel loading buffer and carefully pipeted into a gel lane

4 Gel electrophorests is performed at 50-60 V constant voltage, usually for 0.5-2 h

5 The gel is then briefly destamed by soaking in H,O with agitation and differenttal display products are visualized by UV light illumination (see Notes 10 and 11)

1 DNA fragments of interest are excised from the gel using a razor blade, placed m 1.5-mL Eppendorf tubes and frozen at -7O’C for 15 min

2 Centrifuge at 15,000g for 15 mm Collect supematant and place in new tube

3 A PCR product reampllfication step is then performed to generate a sufficient amount of DNA for subsequent cloning The reaction is set up as described m Section 3.3 using 5 pL of the gel-recovered DNA solution and the primer pair used in the imttal PCR PCR IS performed using the same conditions as m the initial amplificatton except Stage I is omitted

4 The DNA (18 pL) is then subjected to agarose gel electrophorests and recovered

as described prevtously m steps 1 and 2 (see Note 12)

5 Set up the following IO-@., reaction by adding the reagents m this order: 6.0 p.L PCR product, 2.0 pL Plasmtd (pCRI1) DNA, I 0 uL 10X ligation buffer, and

1 O pL T4 DNA ligase

30 Donohue, Hsu, and Winkles

6 Incubate at 12°C overnight

7 Set up the bacterial transformatton by combmmg 2-3 pL of the ligation reactton wtth 50 & of competent cells

8 Incubate on tee for 30 mm Heat shock at 42°C for 1 mm Place on ice

9 Add 450 pL of SOC media to cells Incubate at 37’C with agitation for 1 h

10 Prepare four bactertal plates (100 x 15 mm) per transformation by combmmg

100 mL Lurla broth with 1.5 g Bacto-agar and autoclavmg

11 Place agar solution m 60°C water bath until cooled Add 150 p.L amptctllm and 100 p.L X-gal, mix and pour 25 mL per plate Place at room temperature unttl sohdrfied

12 Spread 125 & of transformed cells per plate Incubate at 37°C overnight

13 Blue and white colomes should be visible Pick several white colonies, grow m Lurta broth/ampicillin media and prepare plasmid DNA by standard techniques (see Note 13)

4 Notes

1 In our specific case, we were attemptmg to tdenttfy genes that were expressed followmg FGF-1 sttmulatton of quiescent murme NIH-3T3 fibroblasts Thus, RNA was Isolated from quiescent cells as well as from cells treated with FGF- 1 for either 2 or 12 h

2 It 1s important to confirm that the RNA is undegraded prior to its use An altquot (5-10 pg) should be subJected to electrophorests on a 1 2% agarose gel contam-

mg 2.2M formaldehyde and then 28s and 18s rRNA integrity examined by ethtdium bromide stammg/UV light tllummatton usmg standard procedures

3 Reverse transcription with random hexamers will generate cDNA molecuies rep- resenting several regions ofmost mRNA species However, unless poly(A)+ RNA 1s used as the template, most of the cDNA populatton wtll be copied from rRNA, which comprises the majority of the cellular RNA mass In some cases, we have also made cDNA using the RACE ohgo d(T)-double adapter prtmer described by Frohman (13)

4 We always determme whether the cDNA synthesis reaction was successful by performmg PCR using 2% of the product and either p-actin or FGF receptor-l oligonucleotrde primers as descrtbed (14)

5 Twelve motif oltgonucleottde primers are listed m Table 1 The degenerate pro- tem tyrosme kmase domain primers are identical to those described by Walks

(15) The degenerate zmc finger and leucme zipper domam prtmers were destgned

by alignmg various mouse DNA sequences and establishmg a consensus sequence The SK homology-2 domam primers were designed usmg the ammo acid sequence altgnment figure presented in Koch et al (16) and codon utiltza- tton data (27) PCR assays were performed usmg one pan of primers at a time (sense and antisense) and all of the possible combmattons As menttoned in Secton 1 , often these motif primers behaved as arbttrary primers under the PCR condtttons used in our experiments Several examples of the primer cDNA mter- actions that we have noted are shown m Ftg 1,

Differential Display Varia Cons 31

1 Low sequence IdentIty between pnmer and cDNA

2 High sequence Identtty, but sense primer acttng as antisense primer (or we versa)

i!F sense 5'-GG?F+Q?AGCCCTTCGAGTG3' 11111,1111,1

Fig 1 Nucleotlde sequence identity between several of the motif primers and the correspondmg regions of the FR cDNA clones DNA sequence analysis has Indicated that the majority ofthe genes that were identified using the motif primers do not encode proteins contammg the targeted structural domains Examples illustrating several of the most common explanations for this finding are shown here In general, under the PCR conditions employed, the selectivity of the motif primers was poor and thus they were actually functlonmg m a similar manner as long arbitrary primers

6 Motif primers have also been used by two other groups attemptmg to identify differentially expressed members of known gene families Stone and Wharton (18)

used several different degenerate protein kmase primers or zmc finger primers in combmation with degenerate arbitrary primers in their PCR assays They were able

to identify cDNAs encodmg a novel casein kmase I lsoform and several zinc finger domain-contammg proteins More recently, Bayarsalhan et al (19) used inosine-containing SPC homology-2 and homeobox domain primers m combination with an arbitrary pnmer m their PCR assays They isolated several cDNA clones but did not report whether they encoded proteins with the targeted structural mottfs

7 In those cases when cDNA samples were prepared using the RACE oligo d(T)-double adaptor primer (see Note 3), individual motif sense primers are used

m combination with an antisense RACE outer adaptor primer (I 3) One of the 13 cDNAs that we ldentlfied by differential display was amplified using this RACE primer approach (5)

8 Some of our PCR assays have been performed using the followmg Stage I and II condltlons

Stage I (8 cycles) 94”C, 30 s denaturation

4O”C, 30 s annealing 68”C, 30 s extension Stage II (3@-40 cycles) 94”C, 30 s denaturatlon

.54”C, 30 s annealing 72”C, 30 s extension

32 Donohue, Hsu, and Winkles

As a control for the cDNA synthesis, PCR was also performed using sense and antisense FGF receptor (FGFR)- 1 primers (bottom panel) Amplification products were displayed using agarose gel electrophoresis and ethidium bromide staining DNA size markers (M; in bp) are shown on the left The pattern of amplified cDNAs obtained using RNA from quiescent or FGF-l-stimulated cells was similar except for an approx 700~bp DNA fragment, noted with an arrow in the top panel, that was present in the 12-h poststimulation lane This fragment, termed FR- 1, was excised from the gel, reamplified, and cloned

9 We have also displayed PCR amplification products on either 2% agarose gels or 4% agarose gels prepared using a 3:l ratio of NuSieve GTG low melt agarose (FMC Bioproducts, Rockland, ME) and standard agarose

10 A photograph depicting a typical differential display gel (1.8% agarose) is shown

in Fig 2

11 Two other laboratories have also successfully identified differentially expressed genes by differential display using agarose gel electrophoresis and ethidium bro- mide staining (Z9,20)

12 We have had poor success using this DNA as either a template for DNA sequence analysis or as a probe for Northern blot hybridization analysis Therefore, we

Differential Display Variations 33

FR-1

rRNA

Fig 3 Northern blot hybridization analysis of FR- 1 mRNA expression in FGF- l- stimulated flbroblasts Northern blot analysis was performed to confirm the differential display results shown in Fig 2 indicating that FR-1 mRNA levels were elevated in FGF- 1 -stimulated cells Quiescent NIH-3T3 cells were either left untreated (NT, no treat- ment) or treated with FGF- 1, FGF- 1 and actinomycin D (Act.D), or actinomycin D alone for 8 h RNA was isolated and equivalent amounts of each sample, analyzed by Northern blot hybridization The bottom panel is a photograph illustrating the relative amounts of 28s rRNA in each gel lane FGF-1 induced FR-1 mRNA expression; furthermore, this effect is due, at least in part, to increased FR-1 gene transcription, since it does not occur

if RNA synthesis is inhibited by actinomycin D

routinely clone cDNAs isolated by differential display using the TA Cloning System (Invitrogen)

13 We always perform Northern blot hybridization analysis with purified cDNA insert as our first experiment in order to confirm that the gene under study is in fact differentially expressed Our Northern blot hybridization protocol is described in Hsu et al (I 1) and a representative result is shown in Fig 3

Acknowledgments

This work was supported in part by National Institutes of Health Research Grants HL-39727 and HL-547 10 P J D performed this work in partial fulfill- ment of the requirements for the degree of Doctor of Philosophy from the Graduate Genetics Program, George Washington University, Washington, DC

D K W H was supported in part by National Institutes of Health Training Grant HL-07698 We thank W H Burgess for providing the FGF-1, K Wawzinski and D Weber for excellent secretarial assistance, and G F Alberts for reviewing the manuscript

34 Donohue, Hsu, and Winkles References

1 Willtams, G T., Abler, A S., and Lau, L F (1992) Regulation of gene expression

by serum growth factors, m Molecular and Cellular Approaches to the Control of Prolzferation and Dtfferentzatton (Stem, G S and Lian, J B , eds ) Academic, Orlando, FL, pp 115-161

2 Muller, R , Mumberg, D , and Lucibello, F C (1993) Signals and genes m the control of cell-cycle progression Brochtm Bzophys Acta 1155, 15 l-1 79

3 Burgess, W H and Wmkles, J A (1996) The fibroblast growth factor famtly multifunctional regulators of cell proliferation, in Cell Proltferatton zn Cancer Regulatory Mechanisms ofNeoplasttc Cell Growth (Pusztai, L , Lewis, C E , and Yap, E , eds ), Oxford University Press, Oxford, pp 154-2 17

4 Fernig, D G and Gallagher, J T (1994) Fibroblast growth factors and their receptors: An information network controllmg tissue growth, morphogenests and repair Prog Growth Factor Rex 5, 353-317

5 Hsu, D K W , Donohue, P J., Alberts, G F , and Winkles, J A (1993) Fibro- blast growth factor-l induces phosphofructokmase, fatty acid synthase and Ca2+-ATPase mRNA expression in NIH 3T3 cells Bzochem Bzophys Res Commun 197,1483-1491

6 McClelland, M., Mathieu-Daude, F , and Welsh, J (1995) RNA fingerprmtmg and differential display using arbitrarily primed PCR Trends Genet 11,24 l-246

7 Liang, P , and Pardee, A B (1995) Recent advances m differential display Curr Open Immunol 7,274-280

8 Winkles, J A, Donohue, P J , Hsu, D K W , Guo, Y , Alberts, G F , andpeifley,

K A (1995) Identification of FGF-1 -mducible genes by differential display, in Cardtovascular Dtsease 2 Cellular and Molecular Mechanums, Preventton and Treatment (Gallo, L L , ed ) Plenum, New York, pp 109-120

9 Donohue, P J., Alberts, G F., Hampton, B S., and Winkles, J A (1994) A delayed-early gene activated by tibroblast growth factor-l encodes a protein related to aldose reductase J Biol Chem 269,8604-8609

10 Hsu, D K W., Guo, Y., Alberts, G F., Petfley, K A , and Winkles, J A (1996) Fibroblast growth factor-l -inducible gene FR-17 encodes a nonmuscle a-actmm isoform J Cell Physiol 167,261-268

11 Hsu, D K W , Guo, Y., Alberts, G F., Copeland, N G., Gilbert, D J., Jenkins, N A., Peifley, K A, and Winkles, J A, (1996) Identification of a murme TEF-

1 -related gene expressed after mitogemc stimulation of quiescent fibroblasts and during myogemc dtfferentiation J Btol Chem 271, 13,786-13,795

12 Donohue, P J., Alberts, G F., Guo, Y , and Winkles, J A (1995) Identification

by targeted differential display of an immediate-early gene encoding a putative serinelthreonine kinase J B1o1 Chem 270, 10,351-10,357

13 Frohman, M A (1993) Rapid amphfication of cDNA ends (RACE)* User-friendly cDNA cloning Amplzficatzons 5, 1 l-l 5

14 Brogi, E., Winkles, J A., Underwood, R , Clmton, S K., Alberts, G F., and Libby,

P (1993) Distinct patterns of expression of fibroblast growth factors and their receptors m human atheroma and non-atherosclerotic arteries Assoctatton of

Differential Display Variations 35

acidic FGF with plaque microvessels and macrophages J Clin Invest 92,

17 Lathe, R (1985) Synthetic oligonucleotide probes deduced from ammo acid sequence data* Theoretical and practical considerations J Mel Bzol 183, 1-12

18 Stone, B and Wharton, W (1994) Targeted RNA fingerprintmg* The cloning of differentially-expressed cDNA fragments enriched for members of the zmc finger gene family Nuclezc Acids Res 22,2612-2618

19 Bayarsarhan, D., Enkhmandakh, B., Farrell, C , and Lukens, L (1996) Rapid tden- trticatton of a novel chondrocyte-specific gene by RNA differential display

Blochem Blophys Res Commun 220,449-452

20 Sokolov, B P and Prockop, D J (1994) A rapid and simple PCR-based method for tsolatton of cDNAs from differentially expressed genes Nuclex Acids Res 22,4009-40 15

21 Dixon, H B F , Blelka, H , Cantor, C R , Llebecq, C , Sharon, N , Velrck, S F , and Vhegenthart, J F G (1986) Nomenclature for mcompletely specified bases

m nucleic acid sequences J Bzol Chem 261, 13-17

4

Fluorescent Differential Display

Takashi Ito and Yoshiyuki Sakaki

1 Introduction

Since their first mtroduction in 1992 (1,2), differential display (DD) and its relatives have been used extensively in various fields of biology where the identification of differentially expressed messages are of particular interest and importance (for review, see ref 3) These new techniques have several unique advantages over the conventtonal methods such as differential and/or subtrac- tive hybrtdization-based ones They can compare a number of samples m par- allel to reveal transcripts of various behaviors, unlike subtractive hybridization techniques that essentially compare two samples unidirectionally They can detect transcripts of low abundance as well as those showing subtle changes, that would be overlooked by conventional hybridization-based methods Also, the techniques require only a tiny amount of RNAs to start the analysis so that they can be applied to biological samples that are difficult or impossible to prepare

in large amounts

Despite these attractive features, the methods have several drawbacks For instance, since they are essentially random sampling approaches, considerable numbers of reactions (i.e., primer combinations) have to be tested to statisti- cally cover the complex transcript population If one assume the sampling 1s a completely stochastic event, the number of reactions required (N is calculated as

N=ln(l -P)lln(l -B/E) where P is the probability desired, E is the number of expressed transcript species, and B is the number of bands per each reaction Therefore, if each DD reaction amplifies 100 cDNAs (B = loo), one has to perform about 450 reac- tion to cover all the transcript in typical mammalian cells (E = 15,000) with

From Methods m Molecular B/ology, Vol 85 D/fferent/al D/splay Methods and Protocols

E&ted by P Llang and A B Pardee Humana Press Inc , Totowa, NJ

37

38 Ito and Sakakl

To run this scale of analyses with multiple samples, one has to Increase the speed of each analysis, m parttcular, that of electrophoresis, postrun gel pro- cessing and signal detection Furthermore, the safety problem inherent to radioactive DD would not be negligible m that scale (4) To overcome these drawbacks, we established two DD protocols compatible with fluorescent detection, termed fluorescent differential display (FDD) protocols S and L (S-7) Protocol S uses short arbitrary primers (1 0-mer) and low stringency PCR condition as the origmal DD protocol, but the design of anchor primers were modified so that one can obtain signals with enough mtensmes as well as satis- factory reproducibility The anchor primers currently used m protocol S are GT,,N (N = A, C, or G) They contam additional dGs at then S-ends, that are necessary for the generation of signals intense enough for fluorescent detection (Addition of further nucleotides did not improve the signal anymore.)

The protocol L uses primers of usual lengths (-20-mer) as upstream arbi- trary primers The anchor primers for protocol L are CCCGGATCCTt5N (N = A,

C, or G) In the mittal cycle of PCR in protocol L, a low strmgency condition was used so that even a longer primer can nutlate second strand synthesis on multiple cDNAs In the second and later cycles, the molecules tagged at their both ends with anchor and arbitrary primers are specifically amplified by means

of high stringency PCR The sequences added to the S-ends of the anchor prim- ers are to increase their T, to be used m the high stringency PCR steps Here we describe both FDD protocols using fluorescence image scanner, that ensure highly reliable DD analysis with much Improved speed and safety

2 Preamplification Kit (BRL, ME)* This kit contains Superscript 11 reverse tran-

scriptase and other soluttons for reverse transcription

3 Thermal cycler: Since reverse transcription step uses vanous temperatures, we

are using tube-type thermal cycler for the mcubatton

4 TE buffer 10 mA4 Tris-HCl, pH 8 0, and 1 mM EDTA

1 Arbitrary primers: Arbitrary lo-mers were obtained from Operon (Alameda, CA)

or synthesized in our laboratory (see Note 4), whereas various primers of 20 nt

Fiuorescent Different/al Display 39

long, that had been origmally designed for mdivtdual purposes m our lab, were recruited as arbitrary primers for protocol L They were adJusted to 10 fl and stored at -20°C

2 Taq DNA polymerase* We use GeneTaq DNA polymerase (Nippon Gene, Toyama, Japan), an N-terminally truncated Tuq DNA polymerase supplied with

1 OX buffer and dNTP solution To improve ampltfication of htgher mol-wt spe- cies, conventional Taq DNA polymerase from Perkm-Elmer or BRL 1s used with GeneTaq

3 Thermal cycler For a large number of PCR, those with 96-well microtiter plate format are convenient We routmely use Techne PHC-3 thermal cycler (Techne, UK) for PCR

1 Heat a thermal cycler to 70°C

2 Mix 8 0 l.tL of DEPC-treated water, 1 0 pL of Anchor Primer (50 CUM) and 1 0 pL

of total RNA solution (2 5 pg/uL) m a 0 5-mL tube (see Note 1)

3 Place the tube m the thermocycler at 70°C for 5-10 mm

4 Durmg the mcubatton, prepare 2X RT solution by mixmg 2.0 pL of DEPC-treated water, 2.0 &of 10X PCR buffer, 2 0 pL of25 mMMgCl,, 2.0 &of 100 mMDTT,

1 0 cls, of 10 mM dNTP and 1 0 clr, of Superscript II (200 II/$) using a BRL preampltfication kit (see Note 2) Prepare enough amount, taking the pipeting loss mto account Followmg thorough but gentle mixing, place the tube on Ice

5 Place the heated tubes containing RNA and primers mto ice-water bath for a few minutes, and cool the thermal cycler to 25°C Following the brief spin, add 10 $

of 2X RT solutton mto each tube and mtx thoroughly by gentle pipetmg

6 Place the tubes m the thermal cycler, and incubate them for 10 mm at 25’C,

50 mm at 42°C and 15 mm at 70°C

7 Following the above mcubatton, spm the tube briefly and add 80 & of TE Store

at -20°C until use

1 Mtx 13.0 (x N) $ ofdistilled water, 2 0 (x N) & of 10X GeneTaq buffer, 1.6 (x N) & of 2.5 mA4 dNTP, 0.2 (x N) pL o anchor primer and 2 0 (x N) $ of f cDNA solution (PCR mtx I) Calculate iV, taking into account the loss during repeated ptpetmg

40 Ito and Sakaki

2 Dispense 1 O pL of arbitrary primers ( 10 pmol) m each well of 96-well thermal plate

3 Make PCR mix II by mixmg 18.8 (x N) pL of PCR mix I, 0.1(x N) pL of GeneTaq DNA polymerase and 0 1 (x N) pL of AmphTaq DNA polymerase (see Note 3)

4 Dispense 19.0 pL of PCR mix II to each well containing 1.0 pL of arbitrary primers

5 Overlay each well with a drop of mineral oil

6 Place the 96-well plate in thermal cycler, and run the followmg program (see Note 4)

1 Prepare 6% polyacrylamide (or LongRanger) gel (200 x 330 x 0.35 mm)

2 Followmg polymerization, prerun the gel for 1 h at 1000 V

3 Mix PCR products with isovolume of formamide dye solution, and heat at 90°C for 3 mm

4 Load 5-6 pL of sample to each well of shark-tooth comb and run the gel at 1500 V for 1 h until bromophenol blue dye reaches to the bottom of the gel

5 Remove the gel slab from the electrophoretic chamber, clean the glass plate and scan it by FluorImager SI using the high sensitivity mode This scan is to obtain images focusing on lower mol-wt bands (see Note 5 and Fig 1)

6 Following the first scanning, place the gel slab to the electrophoretic chamber again and run for an additional hour until xylenecyanol dye migrates near the bottom of the gel Scan the gel again to visualize higher mol-wt bands

1 Repeat the experiment using RNA batches different from those used m the first experiments to confirm the reproducibility of the behavior of the bands of your interest (see Note 6)

2 Run the gel of appropriate concentration and for an optimized duration to maxi- mize the separation of the bands of interest Remove the upper glass plate, and scan the gel on FluorImager

3 Prmt the image usmg “actual size mode” and place the gel precisely onto the prmtout

4 Excise the band of your mterest using a razor blade Following the excision, scan the gel again to see how precisely the band was excised

5 Rinse the excised gel piece with distilled water

6 Put the half of the gel piece mto a PCR tube, and cut tt into several smaller pieces using a flat loading tip just like a blade

7 Add 100 pL of the PCR reaction mix II as described above except for the lower primer concentration (0.25 @4each) and subjected to the following thermal cycling

Fluorescent Differential Display

Fig 1 Identification of a transcript induced during the Alzheimer’s beta-peptide- induced apoptosis of neuroblastoma cells by the two step scanning procedure Total RNAs were isolated from neuroblastoma LA-N-5 cells (lane 1) and those treated with Alzheimer’s beta-peptide for 0, 1, 6, 12,24, 30, and 48 h in lanes 2-8, respectively The anchor primer used was GT&, and the arbitrary primers used from left to right were CTCACCGTCC, AAGCCTCGTC, GACGGATCAG and TTCCCCCCAG, respectively While the left image was obtained following -1 h of gel running to focus

on lower mol-wt species, the right one was taken from the same gel after an additional running for -1 h to put emphasis on higher mol-wt species Note that the band shown

by the arrowhead, that had been poorly resolved at the first scanning, was clearly separated in the right panel

/to and Sakakr

42

3.5 Selection of Correct Clones

1 Suspend colomes of the transformants m 40 pL of L-broth

2 Use 1 0 pL of above suspenston m 20 pL of PCR using the same conditton described m Section 3 4 , step 7

3 Run 0 l-l pL of the product m parallel with the ortgmal FDD reactton (“comigration test”) Select clones bearing inserts that precisely comtgrate wrth the band of Interest

4 Sequence the selected clones

5 Search restrtctron enzyme sites m the nucleottde sequences of the candidate clones Digest the amplified mserts of the clones and the origmal FDD reaction with these restrtction enzymes, and run them m parallel to compare the digestion pattern of the target band and the mserts (“restrtctton test”) Select clones that are digested stmilarly to the band m FDD reaction for further analysrs

6 Confirm the expression pattern of the correspondmg transcrtpt by Northern blot hybndrzation, RNase protectton or quantitative RT-PCR assays

4 Notes

Total RNAs prepared by various methods can be used for FDD We routinely use total RNAs prepared by a modified actd-guamdmmm-phenol-chloroform method usmg TRIzol reagents (BRL) accordmg to the supplier’s recommendatton As the contaminatron of genomtc DNA has severe effects on the DD pattern, m par- ticular, m protocol S, we usually treated total RNA with RNase-free DNase (Promega) m the presence of placental RNase inhibttor or vanadyl-ribonuceoside complex to remove the residual contammatmg genomtc DNAs Although most RNA samples prepared by this method can be used for FDD, some may requrre further treatment For Instance, we experrenced that RNAs from Xenopus early embryo requrres LiCl precipitation step for successful FDD (6) It should be noted that the use of RNAs of similar qualities is crtttcal for FDD We recommend to discard low quality RNA samples and re-prepare RNAs for the FDD

We are using a kit from BRL for the reverse transcrtptton step, but others will give comparable results Smce the first strand syntheses reqmres various tem- peratures, tt IS convement to use thermal cycler rather than preparing several water baths Anchor primers used in protocols S and L are GT,,N and CCCGGATCCTlsN (N = A, C or G), respectively As they are incubated with RNA, special attention not to contammate with RNase has to be paid when pre- paring these primers

We routinely use GeneTaq DNA polymerase (NIpponGene, Japan) for FDD analysis This Tuq polymerase has a large deletron m its N-termmal portion and gives much stronger signals in lower mol-wt size range but weaker signals m higher mol-wt range than conventtonal Taq DNA polymerase Thus, we use a cocktail of this enzyme and usual full-length Tuq polymerase to cover wider size range with higher signal mtensities We prefer 96-well plate-type thermocycler for large-number of PCR As the fingerprmtmg pattern IS affected by the make of thermal cycler, one must not compare results obtamed by drfferent themal cyclers