AP-PCR: Combination of Differential Display and Arbitrarily Primed PCR of Oligo(dT) cDNA

ANALYTICAL BIOCHEMISTRY ARTICLE NO.

245, 48–54 (1997)

AB969936

DD/AP-PCR: Combination of Differential Display and Arbitrarily Primed PCR of Oligo(dT) cDNA Cynthia B. Rothschild, Catherine S. Brewer, and Donald W. Bowden Department of Biochemistry, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, North Carolina 27157

Received August 19, 1996

In this report we describe the first direct comparison of differential display (DD) and arbitrarily primed PCR (AP-PCR) amplification of oligo(dT)-primed cDNA. Our results indicate that both of these widely used RNA fingerprinting techniques have their respective advantages and limitations. DD produces profiles specific to the anchored oligo(dT) primer used for cDNA synthesis. AP-PCR displays significant redundancy of profiles generated from different oligo(dT) cDNA pools, but is not as biased to the isolation of A/T-rich or 3* sequences. It was found that both techniques can utilize cDNA synthesized using a generic anchored oligo(dT) primer (dT12VN; equimolar amounts of dT12VA, dT12VC, dT12VG, and dT12VT, where V is dA, dC, or dG); this efficiently selects for poly(A)/ sequences from total RNA, and significantly reduces the number of cDNA preparations required per experiment. Using dT12VN cDNA pools generated from rat liver, spleen, and brain, the two approaches (AP-PCR and DD) were used in combination. Several known mRNAs were identified; some were unique to either technique and some were common to both. Since it is the RNA which is usually the limiting resource, maximum utilization may be achieved by generating a single pool of dT12VN-primed cDNA and performing both AP-PCR and DD (DD/AP-PCR). q 1997 Academic Press

The detection of differentially expressed genes is of interest in molecular analysis of biological processes including differentiation, development, and carcinogenesis. Previously, techniques such as subtractive or differential hybridization have identified mRNAs expressed in specific cell populations. In general, these approaches are laborious and require significant amounts of material. Recently, two PCR-based techniques, differential display and arbitrarily primed

PCR, have been developed as methods to identify and isolate mRNA specific to a cell type. Arbitrarily primed PCR (AP-PCR)1 generates cell specific fingerprints from DNA (1) or RNA (2, 3). Traditionally, in RNA fingerprinting by arbitrarily primed PCR (2, 3), cDNA is synthesized using a specific but arbitrarily chosen primer (Ç20-mer). The cDNA is then PCR amplified using the same arbitrary primer, with one low stringency cycle to promote incorporation of the primer into the second strand of cDNA, followed by several cycles of high stringency amplification. AP-PCR products are radiolabeled during PCR and resolved by gel electrophoresis, generating a fingerprint which is specific to the RNA population and arbitrary primer used. Differential display (DD) (4, 5) uses 3* anchored oligo(dT) primers for the synthesis of cDNA subsets representative of the poly(A)/ RNA population. Oligo(dT) primers, which are anchored with two specific nucleotides (e.g., dT12AG) (4), degenerate at the penultimate base (e.g., dT12VG; where V is equimolar dA, dG, and dC) (5) or anchored with only one base (e.g., dT12G) (6), have been used successfully for DD. The individual cDNA pools are then amplified in the presence of radiolabel using the same anchored oligo(dT) primer and a 10-mer. For both AP-PCR and DD, bands specific to the cell population of interest can be excised from the gel and the DNA eluted and reamplified for subsequent cloning and characterization. In this report, we directly compare AP-PCR and DD as methods to identify differentially expressed sequences. Our results indicate that both methods can be used to generate RNA fingerprints from oligo(dT)generated cDNA. Both techniques are straightforward 1 Abbreviations used: AP-PCR, arbitrarily primed PCR; DD, differential display; BSA, bovine serum albumin, IAPE, intracisternal A particle element.

48

AID

0003-2697/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AB 9936

/

6m26$$$$$1

12-30-96 04:26:00

abas

DIFFERENTIAL DISPLAY AND ARBITRARILY PRIMED PCR

and reproducible, each with its respective advantages and limitations. DD produces profiles which are specific to the oligo(dT) primer used (e.g., dT12VG vs dT12VC) for amplification and is suited for isolation of A-rich and 3* mRNA sequences. AP-PCR displays significant redundancy in profiles generated from various oligo(dT) cDNA pools, but is not as biased to the isolation of A/T-rich or 3* sequences. With selection of expressed sequences as oligo(dT)-primed cDNA, both techniques are complementary and can be employed in a combined approach (DD/AP-PCR) for isolation of differentially expressed mRNA. MATERIALS AND METHODS

Primers synthesized by Operon Inc. (Alameda, CA) were MFD95GT (5*-CTTTATCTTCACACAGCTTC-3*) (D21S171), PPGB (5*-GATGAGAAGACCCTTCAACC3*) (7), and OL4 (5*-AGATGAGCATAGATACGAGA-3*) (8); these primers were arbitrarily chosen from primers available in our laboratory. Primers synthesized by the DNA Synthesis Core Laboratory of the Comprehensive Cancer Center of Wake Forest University were MFD-3* (5*-ACACAGCTTC-3*), PPGB-3* (5*-CCCTTCAACC-3*), and OL4-3* (5*-AGATACGAGA-3*). The anchored oligo(dT) primers were dT12VT, dT12 VA, dT12VG, and dT12VC (V Å equimolar dA, dC, and dG). Isolation of RNA and cDNA Synthesis Rat tissue (liver, spleen, and brain) was snap-frozen in liquid nitrogen and stored at 0707C. Total RNA was isolated using TRIzol Reagent (Gibco BRL/Life Technologies, Gaithersburg, MD) and treated with DNase I (Gibco/BRL) as recommended by the manufacturer. Poly(A)/ RNA was isolated using the Micro-FastTrack kit (Invitrogen, San Diego, CA). cDNA was synthesized from mRNA (100 ng mRNA/20-ml reaction) or DNAfree total RNA (1 mg/20-ml reaction) using the cDNA Cycle kit (Invitrogen) or SuperScript II RNase H0 reverse transcriptase (Gibco BRL); 2 ml of cDNA was used in AP-PCR or DD reactions as described below. Arbitrarily Primed PCR (AP-PCR) cDNA was synthesized using random hexamers (100 ng/ml) or oligo(dT) (20 ng/ml) (cDNA Cycle Kit, Invitrogen), anchored oligo(dT) primers (20 ng/ml), or, in some experiments, the specific 20-mers MFD95GT, PPGB, or OL4 (20 ng/ml). cDNA pools were amplified by AP-PCR using MFD95GT, PPGB, or OL4 with two low-stringency cycles (5 min at 947C, 5 min at 407C, and 5 min at 727C) for incorporation of primer into initial amplification products), followed by 40 highstringency cycles (1 min at 947C, 1 min at 587C, and 2

AB 9936

/

6m26$$$$$2

min at 727C). Amplification reactions (15 ml) contained 2 ml cDNA, 50 mM KCl, 10 mM Tris–HCl, pH 8.3, 0.01% BSA, 2.5 mM MgCl2 , 250 ng specific 20-mer (3.3 mM), 100 mM dNTPs, 2.5 mCi [a-32P]dCTP (3000 Ci/mM; 10 mCi/ml), and 0.75 units Taq polymerase (BoehringerMannheim). In some experiments, a two-step amplification protocol (1) was employed in which cDNA was amplified for 2 low-stringency and 10 high-stringency cycles using increased primer (10 mM) and MgCl2 (4 mM), followed by 30 high-stringency cycles with reduced primer (1 mM) and MgCl2 (1.5 mM), and 5 mCi [a32 P]dCTP. Results are essentially the same using either protocol. Differential Display (DD)

Primers

AID

49

12-30-96 04:26:00

cDNA was synthesized using oligo(dT), random hexamers, or the anchored oligo(dT) primers as described for AP-PCR. DD reactions (15 ml) contained 2 ml cDNA, 50 mM KCl, 10 mM Tris–HCl, pH 8.3, 1.5 mM MgCl2 , 0.01% BSA, 5 mM dNTPs, 5 mCi [a-32P]dCTP, 100 ng of the primer used for cDNA synthesis, 50 ng of a specific 10-mer (e.g., PPGB-3*), and 2.5 units Taq. In some experiments DD products were radiolabeled by incorporation of an anchored oligo(dT) primer (e.g., dT12VC) which had been end-labeled using [g-32P]ATP; in these amplifications 100 mM dNTP was used. Samples were heated to 957C for 1 min, followed by 40 cycles of 947C for 45 s, 407C for 90 s, and 727C for 30 s with 2 s extension per cycle, followed by 5 min at 727C. Cloning of AP-PCR and DD Products AP-PCR and DD products were resolved by electrophoresis on a 6% DNA sequencing gel. Bands of interest were excised from the gel and incubated in 100 ml TE (10 mM Tris–HCl, pH 7.5; 1 mM EDTA) for 2 h at 607C. An aliquot of the elution was then used for reamplification using the same primers and thermocyling conditions used for original amplification (high stringency for AP-PCR). Reamplification reactions (100 ml) contained 7.5 ml eluted DNA, 10 mM Tris–HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2 , 0.01% BSA, 200 mM dNTPs, 500 ng primer(s), and 2.5 units Taq. Successful reamplification of the fragment was verified by agarose gel electrophoresis, and products were ligated into pCRII TA cloning vector (Invitrogen) for DNA sequencing. Sequences were compared by electronic mail (BLAST) to current GenEMBL and dBEST databases (National Center of Biotechnology, Washington, DC). RESULTS AND DISCUSSION

RNA Fingerprinting by Arbitrarily Primed PCR (AP-PCR) and Differential Display (DD) The overall strategy for direct comparison of AP-PCR and DD is diagramed in Fig. 1. Rat liver was excised

abas

50

ROTHSCHILD, BREWER, AND BOWDEN

FIG. 1. Strategy for comparison of AP-PCR and differential display. Total RNA or mRNA was isolated from rat liver, and cDNA was generated by priming with oligo(dT), random hexamers (N6), anchored oligo(dT) primers (dT12VA, dT12VC, dT12VG, or dT12VT), or a combination of the four anchored oligo(dT) primers (dT12VN) as indicated. Individual cDNA pools were then amplified by AP-PCR using a specific 20-mer (top) or by differential display (bottom) using both the primer used for cDNA synthesis and a 10-mer corresponding to the 3* half of the 20-mer used for AP-PCR.

and immediately frozen in liquid nitrogen. Both total RNA and poly(A)/ RNA (mRNA) were isolated. cDNA pools from either total RNA or mRNA were generated using oligo(dT), random hexamers, or the anchored oligo(dT) primers dT12VA, dT12VC, dT12VG, dT12VT (where V is degenerate as dA, dC, or dG), or dT12VN (equimolar amounts of dT12VC, G, A, and T). In an effort to directly compare AP-PCR and DD as methods for generating RNA fingerprints, 10-mers used for DD corresponded to the 3* half of 20-mers used for APPCR. For example, the cDNA pool generated with dT12VC was amplified either with MFD95GT (5*-CTTTATCTTCACACAGCTTC-3*) for AP-PCR or dT12VC and MFD-3* (5*-ACACAGCTTC-3*) for DD. AP-PCR and DD profiles generated from rat liver total RNA are shown in Fig. 2. cDNA was synthesized using oligo(dT), random hexamers, or the anchored oligo(dT) primers, and then aliquots were used for AP-PCR or DD. AP-PCR profiles generated with PPGB and MFD95GT are shown in Fig. 2A. It can be seen that AP-PCR products generated from oligo(dT) (lanes 1), random hexamers (lanes 2), and dT12VN (lanes 7) cDNA pools have some overlap, although bands specific to each lane can be detected. cDNA generated with the anchored oligo(dT) primers (dT12VC, G, A, or T) (lanes 3, 4, 5, and 6, respectively) generated profiles similar to each other (and dT12VN). DD profiles generated using the oligonucleotide that was used for cDNA synthesis (e.g., dT12VC for dT12VC-primed cDNA) and either PPGB-3 * or

AID

AB 9936

/

6m26$$$$$2

12-30-96 04:26:00

MFD-3 * are shown in Fig. 2B. DD profiles generated from oligo(dT) cDNA (lanes 1) resulted in a smear as expected, since the 3 * end of the DD products will not be fixed. For random hexamers (lanes 2) and the anchored oligo(dT) primers (dT12VC, G, A, or T; lanes 3 – 6), however, the resulting profiles were specific to both the 10-mer and the anchored oligo(dT) primer used for amplification. DD of dT12VN primed cDNA with dT12VN and a specific 10-mer resulted in a profile which contained some (but not all) of the bands seen in the individual anchored primer reactions, suggesting that a subset of the DD products was preferentially amplified. Overall, although the complexity of the pattern generated in an individual lane was about the same with either technique (Ç20 – 40 bands/lane), redundancy between cDNA pools was greater with AP-PCR than DD. Similar results were found with the OL4/OL4-3 * primers. It appears then, that reverse transcription with the anchored primers is not completely specific; therefore, AP-PCR gives redundant profiles, whereas DD enables an extra level of specificity due to the use of anchored oligo(dT) primers during the amplification step. Overall, both AP-PCR and DD were qualitatively reproducible between experiments, although efficiencies of amplification for individual reactions were found to vary (e.g., Fig. 2B; dT12VG/MFD-3). In Fig. 3, AP-PCR and DD fingerprints generated from mRNA and total RNA are compared. cDNA was generated from rat liver total RNA or mRNA using the

abas

DIFFERENTIAL DISPLAY AND ARBITRARILY PRIMED PCR

51

FIG. 2. Comparison of AP-PCR and differential display (DD) amplification of different cDNA pools. cDNA was generated from rat total RNA using oligo(dT) (lanes 1), random hexamers (lanes 2), anchored primers dT12VC (lanes 3), dT12VG (lanes 4), dT12VA (lanes 5), dT12VT (lanes 6), or dT12VN (equimolar mix of the four anchored primers) (lanes 7). (A) AP-PCR with PPGB and MFD95GT. (B) DD with PPGB3* or MFD-3 * and the same primer used for cDNA synthesis. Approximate sizes are given in nucleotides (nt) and were determined by reamplification of bands and agarose gel electrophoresis. A representative experiment of two separate determinations is shown.

anchored primers dT12VC, dT12VG, or dT12VA amplified using PPGB for AP-PCR or PPGB-3* and dT12VC, G, or A, respectively, for DD. Both approaches (AP-PCR and DD) result in product bands of similar sizes and overall complexity. For both AP-PCR and DD there is significant overlap between fingerprints generated from mRNA versus total RNA, indicative that selection of poly(A)/ sequences by oligo(dT)-primed cDNA synthesis is fairly efficient. Use of dT12VN-Primed cDNA Typically, both AP-PCR and DD rely on the use of primer specific cDNA synthesis for each analysis. For DD, individual cDNA pools are generated with each anchored oligo(dT) primer; each pool is then used for PCR with the same anchored oligo(dT) and a variety of different 10-mers. For AP-PCR, the conventional method is to synthesize cDNA using the same 20-mer that is subsequently used for amplification. For both techniques, to ensure adequate representation of the transcripts present, multiple cDNA pools must be generated for each sample.

AID

AB 9936

/

6m26$$$$$2

12-30-96 04:26:00

We evaluated whether the composite set of anchored oligo(dT) primers (dT12VN) could be used to generate a representative cDNA pool suitable for DD. In the experiment shown in Fig. 4, cDNA synthesized using dT12VN is compared to cDNA synthesized using a specific anchored oligo(dT) primer (e.g., dT12VC), to support DD amplification using the same specific anchored oligo(dT) primer (e.g., dT12VC) and MFD-3*. This is distinct from the experiment shown in Fig. 2, where dT12VN was used in the amplification step as well. It was found that cDNA synthesized with the mixed oligo(dT) set is an efficient substrate for DD. Thus, virtually identical DD profiles were generated using dT12VN-primed cDNA and the cDNA primed with a specific anchored primer; this can reduce the number of cDNA syntheses required per DD experiment. To evaluate the efficiency of various cDNA pools to support AP-PCR, cDNA was generated from rat liver mRNA and total RNA using either MFD95GT or PPGB (20-mers), and AP-PCR was performed using the same primer. It was found that cDNA synthesized with specific 20-mers resulted in RNA fingerprints of similar complexity as cDNA synthesized using dT12VT (not

abas

52

ROTHSCHILD, BREWER, AND BOWDEN

ers dT12VN. These cDNA pools were amplified by APPCR using the 20-mer MFD95GT (set A) (bands 1–8) or by DD. For the DD amplifications, two approaches were compared. For set B (bands 9–14) PCR products were labeled by incorporation of [a-32P]dCTP. Set C (bands 15–22) employed end-labeled dT12VC; thus only sequences containing A-rich (or T-rich) regions should be labeled. Representative AP-PCR and DD bands (i.e., of differing sizes and intensities) were excised from the gel and reamplified, and products were subcloned into pCRII. Negative controls (no RNA, no reverse transcriptase) were included to assess for potential false positives resulting from PCR artifacts or amplification of residual genomic DNA. Four subclones were isolated for each reamplified band. Subclones were analyzed by digestion with Sau3A (9) to determine the number of unique subclones isolated for each band. Each unique subclone was sequenced from both ends, and the sequences compared to current GenEMBL and dBEST

FIG. 3. Comparison of AP-PCR and DD profiles from mRNA and total RNA. cDNA pools were generated from rat liver total RNA or mRNA with dT12VC, dT12VG, or dT12VA as indicated. cDNA was then amplified by AP-PCR using PPGB or by differential display (DD) using PPGB-3 * and the anchored dT12V primer used for cDNA synthesis as indicated. Shown is a representative experiment of two separate determinations.

shown). Thus for some applications, (e.g., when only total RNA, as opposed to when mRNA is available), it may be preferable to generate a representative cDNA pool using a composite set of anchored oligo(dT) primers (to select for poly(A)/ RNA), which can then be used for AP-PCR or DD amplification. Comparison of Clones Isolated by AP-PCR and DD Amplification of dT12VN-Primed cDNA cDNA was generated from rat liver, spleen, and brain total RNA and mRNA using the mixed anchored prim-

AID

AB 9936

/

6m26$$$$$2

12-30-96 04:26:00

FIG. 4. Use of dT12VN-primed cDNA for DD. cDNA was synthesized from rat liver total RNA using either dT12VN or the specific anchored oligo(dT) (i.e., dT12VA, dT12VC, or dT12VG) as indicated directly above the lanes. The cDNA was then amplified by DD using the specific anchored oligo(dT) indicated (i.e., dT12VA, dT12VC, or dT12VG) and the 10-mer MFD-3*. Boxed bands were eluted and the products reamplified (not shown). Shown is a representative experiment of three separate determinations.

abas

53

DIFFERENTIAL DISPLAY AND ARBITRARILY PRIMED PCR TABLE 1

Comparison of Clones Isolated by AP-PCR (Set A) and DD (Sets B and C) from Rat Liver, Spleen, and Brain Banda

Sizeb

Originc

Clonesd

Identitye

1 2 3 4 5 6 7 8

800 600 425 500 300 275 600 425

L S B L S L B S S S B L S B

1 3 2 1 1 1 2 0

Set A None Rat dimethylglycine Dh (X55995); cDNA (D62589) Poly(A) binding protein (YOO345) None None None Rat Na//K/ ATPase b (J02701); cDNA (N49076) —

9 10 11 12 13 14

325 250 225 425 350 220

L S L S B L S B S L S B L S B

2 3 3 2 2 1

Set B Rat lecithin cholesterol acyltransferase (X54096) Rat IAPEf (U23776); cDNA (T08245) Rat IAPE; rat factor X (X79807); cDNA (F05295) Poly(A) binding protein — Rat IAPE

15 16 17 18 19 20 21 22 23

200 225 200 — 450 300 175 225 200

L L S B L S B S B L S B S B S L S B L S B

2 1 1 0 1 1 2 1 1

Set C Mouse membrane glycoprotein (J02700) None Rat IAPE — Alu-like None Multiple cDNA clones Rat calcineurin (M29275) Rat IAPE

a Bands 1–8 correspond to AP-PCR (Set A); bands 9–14 correspond to DD using incorporation of [a-32P]dCTP (Set B); and bands 15–23 are from DD using end-labeled dT12VC (Set C). b Approximate size in base pairs of amplified sequence by re-PCR and subcloning. c Tissue (Liver, Spleen, Brain) in which AP-PCR or DD band of the same size detected; letter in bold is lane from which AP-PCR or DD band was isolated. d Number of unique subclones (out of four isolated) as determined by sequencing and digestion with Sau3A. e Identity (and Accession No.) of AP-PCR or DD product as determined by Blast search of GenEMBL and dBEST databases. f IAPE, intracisternal A particle element.

databases. The results of this experiment are summarized in Table 1. There were two bands for which no data was obtained. Band 8 (AP-PCR) was reamplified as a 425-bp product but could not be subcloned, and band 18 (DD; ú500 bp) did not reamplify. Of seven AP-PCR bands from which subclones were isolated, three (bands 2, 3, and 7) correspond to sequences from known proteins, and four (bands 1, 4, 5, and 6), could not be identified (Table 1). For example, as assessed by Sau3A digestion and sequencing, three of the four subclones isolated from band 2 were unique; one subclone was homologous to rat liver dimethylglycine dehydrogenase (X55995), one subclone was homologous to a human aorta cDNA (D62589), and one was not identified (i.e., no homology to sequences in database). Band 3 subclones were homologous to human poly(A) binding protein mRNA, and one band 7 subclone corresponded to rat Na//K/

AID

AB 9936

/

6m26$$$$$2

12-30-96 04:26:00

ATPase b subunit mRNA. The finding that more than one unique clone may be isolated from a band is a common finding for both AP-PCR and DD (e.g., 9–11) and is due to the fact that several amplification products may be contained in the area excised from the gel. Clones from DD band 9 corresponded to rat lecithin cholesterol acyltransferase. Several DD bands identified the same exact region of rat intracisternal A particle element (IAPE) (U23776; bases 482–658), including clones isolated from bands 10, 11, 14, 17, and 23. This duplication was expected for bands 17 and 23, which appeared to be the same product but from different tissue cDNAs (liver and brain, respectively). The other IAPE bands, however, although similar in size, appeared (by electrophoresis of DD reactions) to be different products. Band 11 also identified rat factor X, band 12 subclones corresponded to the same region of poly(A) binding protein mRNA as band 3, band 15 subclones

abas

54

ROTHSCHILD, BREWER, AND BOWDEN

were homologous to a mouse membrane glycoprotein, and band 22 identified rat calcineurin. Thus, using a common cDNA pool (generated with dT12VN), complementary mRNAs were detected using AP-PCR and DD. Poly(A) binding protein mRNA was detected using both AP-PCR and DD, whereas the Na// K/ ATPase b subunit and dimethylglycine dehydrogenase were detected using AP-PCR. IAPE, factor X, calcineurin, and a membrane glycoprotein were detected by DD. For all AP-PCR subclones, potential flanking MFD-3* amplification sites were identified; for DD subclones, flanking MFD-3* amplification sites as well as internal or flanking T-rich sites were identified. Overall, three previously identified mRNAs were detected by AP-PCR (37.5% efficiency) and six previously identified mRNAs were detected by DD (40% efficiency). In summary, it was found that both DD and AP-PCR generate RNA fingerprints of similar complexity and product size. By generating a cDNA pool with a mixed anchored oligo(dT) primer set (dT12VN), both approaches can be employed to select for poly(A)/ sequences even when total RNA is used as the template. DD selects for A/T-rich sequences and thus preferentially identifies 3* regions, although the extent to which 5* sequences are identified by either approach is practically limited by the quality of the cDNA. For many applications, for example, the analysis of clinical biopsy specimens, it is difficult or impossible to isolate sufficient quantities of high-quality mRNA. In these cases, maximum utilization of the sample may be achieved by isolating total RNA, generating a large pool of cDNA using the degenerate anchored oligo(dT) primer

AID

AB 9936

/

6m26$$$$$2

12-30-96 04:26:00

(dT12VN), and performing both AP-PCR and DD as a combined strategy (DD/AP-PCR). ACKNOWLEDGMENTS This work was supported by an Individual Allocation to C.B.R. from an American Cancer Society Institutional Grant to Bowman Gray School of Medicine and NIH Grant CA46806.

REFERENCES 1. Welsh, J., and McClelland, M. (1990) Nucleic Acids Res. 18, 7213–7218. 2. McClelland, M., Ralph, D., Cheng, R., and Welsh, J. (1994) Nucleic Acids Res. 22, 4419–4431. 3. Welsh, J., Chada, K., Dalal, S. S., Cheng, R., Ralph, D., and McClelland, M. (1992) Nucleic Acids Res. 20, 4965–4970. 4. Liang, P., and Pardee, A. B. (1992) Science 257, 967–971. 5. Liang, P., Averboukh, L., and Pardee, A. B. (1993) Nucleic Acids Res. 21, 3269–3275. 6. Liang, P., Zhu, W., Zhang, X., Guo, Z., O’Connell, R. P., Averboukh, L., Wang, F., and Pardee, A. B. (1994). Nucleic Acids Res. 22, 5763–5764. 7. Rothschild, C. B., Akots, G., Hayworth, R., Pettenati, M. J., Rao, P. N., Wood, P., Stolz, F.-M., Hansmann, I., Serino, K., Keith, T. P., Fajans, S. S., and Bowden, D. W. (1993) Am. J. Hum. Genet. 52, 110–123. 8. Economou, E. P., Bergen, A. W., Warren, A. C., and Antonarakis, S. E. (1990). Proc. Natl. Acad. Sci. USA 87, 2951–2954. 9. Shoham, N. G., Arad, T., Rosin-Abersfeld, R., Mashiah, P., Gazit, A., and Yaniv, A. (1996) Bio Techniques 20, 182–184. 10. Bauer, D., Muller, H., Reich, J., Riedel, H., Ahrenkiel, V., Warthoe, P., and Strauss, M. (1993) Nucleic Acids Res. 21, 4272– 4280. 11. Callard, D., Lescure, B., and Mazzolini, L. (1994) Biotechniques 16, 1096–1103.

abas