ABSTRACT
Differential display of mRNA is a simple, sensitive and powerful method to identify differentially expressed gene fragments. The main drawback of differential display is the lack of reproducibility and the inability to read and compare complex gels. This issue results from employing unoptimized primer combinations and non-specific amplification, most likely due to unavoidable low annealing temperatures. In order to display most of the expressed transcripts (80-120 bands/lane), 26 different 5' primers were used in conjunction with nine different 3' poly (dT) primers. These primer combinations, used with the optimized annealing temperature for each set of primers, produced highly reproducible bands. BSA has a direct effect on the number of bands resolved. Variations in ramping time (9-40 s) had little or no effect on the resolution and reproducibility of differential display. The application of mRNA differential display (DDRT-PCR) to gene discovery operations has found broad application since its original description by Liang and Pardee (1). The method is PCR-based and facilitates the identification and cloning of differentially expressed mRNAs. Other methods similar to DDRT-PCR have been described (2,3), although infrequently employed. mRNA differential display has successfully been used to identify and clone many differentially expressed cDNAs; examples include genes expressed in cancers (4), heart disease (5), diabetes (6) and during embryogenesis (7).
To create a catalog of genes expressed during human fetal blood development we performed DDRT-PCR with RNA samples derived from 10 and 24 week fetal livers, representing the beginning and the decline of liver hematopoiesis, respectively. Genes expressed at 10 and 24 weeks gestation and not in adult, or genes only expressed at 10 weeks gestation were further studied within this developmental period. This characterization is the subject of a manuscript to be published elsewhere (8). A problem of DDRT-PCR is the recovery of cDNA fragments that are not differentially expressed, i.e., false positives. This occurs for various reasons including sequence content of the primers and the annealing temperature of the PCR. In an attempt to reduce the frequency of false positives, we describe the interaction of anchored and arbitrary primer sequences, and the optimization of PCR annealing temperature for a set of primers that should encompass most of the estimated 15 000 expressed genes (9). We also compared the cDNA fragment pattern on two thermal cyclers that differ significantly in their design.
Humanfetal and adultliver samples were obtained from a non-profit research tissue bank (AGF, Woodbine, GA) in accordance with the guidelines of the Department of Health and Human Services regulations. Fetal livers were frozen in liquid nitrogen with no longer than 8-15 min warm ischemia time and the age determined by multiple methods (10).
Total cellular RNA was extracted from frozen liver tissue (10 and 24 week fetal and adult liver) by using the RNeasyr protocol (Qiagen) and were treated with DNase I (Boehringer Mannheim). mRNA differential display was performed as described (9) with some modifications. cDNAs were synthesized using MMLV reverse transcriptase (Gibco-BRL), and one of nine poly (dT) primers. To increase the reproducibility and specificity of cDNAs, nine different specific primers of the type T11VV were used, where V can be A, G or C but not T because of reduction in local specific priming. With a T in the terminal 5' or 3' base position, the primer drifts at the cDNA string, and a smear appears on the gel. The 26 arbitrary 5' primer sequences are the same as described in ref. 9. The poly(dT) primer was radiolabeled as described (11), except the unincorporated nucleotides were removed by using a Microcon 3 column (Pharmacia). For PCR amplification, 9.9 µl ddH2O, 2 µl 10* PCR Buffer (Boehringer Mannheim), 1.2 µl 100 µM dNTPs, 2 µl arbitrary primer (2 µM), 4 µl 1 µM 33P-labeled downstream primer, 0.4 µl 5 U/µl Taq DNA polymerase (Boehringer Mannheim) and 0.5 µl of reverse transcriptase reaction were mixed. In the Perkin-Elmer 9600 thermal cycler, the amplification was done at 94°C for 4 min followed by 40 cycles of 94°C for 30 s, 40°C or 42°C for 2 min and 72°C for 30 s and an additional extension at 72°C for 10 min. The cycling parameters for the Idaho Technologies Rapidcycler (ITR) thermal cycler were the same as above except that the annealing temperature was 40°C. The amplified cDNAs were separated on a 6% DNA sequencing gel containing 7 M urea. The BSA was nuclease and protease free (Sigma).
The estimated number of different mRNAs in a eukaryotic cell is in the range of 15 000-20 000 (12). In total we have performed 286 PCRs using the 26 arbitrary primers and the nine poly (dT) primed cDNA fractions for each of three liver developmental time points. We surveyed ~24 000 cDNAs for each time point which is more than the predicted number of expressed genes in a specific cell type. Based on 286 differential display reactions, 20 differentially expressed cDNAs were chosen for further analysis by northern and cDNA Southern blotting (present at 10 and 24 week fetal liver but not adult liver). We found 14 truly differentially expressed and three not differentially expressed genes. Three genes were not detected by the above techniques (8).
As the PCR annealing temperature is increased, the stringency of primer annealing is also increased leading to more specific and reproducible amplification. For all the different combinations of poly (dT) and arbitrary primer, we first tested the 42°C annealing temperature to determine which primer sets work well (yield 100 cDNAs) at this temperature. We found that the T11CG, T11CC, T11GG and T11GC primers generate many intense bands with all arbitrary primers except nos 5, 11, 20 and 25 at the 42°C annealing temperature. The T11CA, T11GA and T11AC primers overall generated fewer cDNA bands of lower intensity at 42°C and the number of bands was increased with all arbitrary primers (except nos 5, 11, 20 and 25) by more than 2-fold when 40°C was used as the annealing temperature. Reactions with the T11AA and T11AG primers resulted in few cDNAs at either the 40°C or 42°C annealing temperatures (Table 1). These data indicate that the cDNA banding pattern is, in part, dependent on the combination of the poly (dT) and arbitrary primer. The data also suggest that the primers containing one or two A residues at the 3' end might not anneal or extend efficiently at 42°C. Overall, 162 of 286 PCR reactions reproducibly generated 80-120 bands at either the 40°C and/or 42°C.
Table 1.
Using a portion of the same primer set, Chapman et al. (13) found that at a PCR annealing temperature of 40°C, T11GC, T11GG, T11AC and T11CA generated the largest number of cDNAs. Since no quantitative data on band number per lane or information with respect to arbitrary primer combinations was reported, it is difficult to make direct comparisons. Again at 40°C only, Mou et al. (14) found using a different set of six arbitrary primers that T11AG, T11GA, T11GG, T11AC, T11CG and T11GC produced many intense bands. A clear discrepancy is the T11AG primer which in our hands performed almost as poorly as T11AA with all 26 arbitrary primers tested (Table 1) and at both annealing temperatures. This could be a function of the differences in the arbitrary primer sets used. It has been generalized (14) that anchored primers with at least one G residue were superior to those which had one C residue and those ending in A or T were the least efficient. Our data indicate that the anchor primers T11AG, T11CA and T11GA are an exception to this generalization when tested with these 26 arbitrary primers at both 40°C and 42°C.
Since DDRT-PCR uses low annealing temperatures we were interested to see what effect a shorter ramp time between the annealing and elongation step would have on the number and pattern of amplified bands. We compared the Perkin-Elmer 9600 and ITR thermal cycler. The ITR changes the reaction temperature by blowing hot air from a light bulb onto the sample (15) instead of heating a sample block. Since the physical thermal mass is smaller, heating and cooling is faster and this results in a shorter ramping time between the annealing and elongation temperatures; 9 s versus ~40 s for the ITR and the Perkin-Elmer 9600, respectively. As annealing or transition time increases, there might be an increased risk of non-specific primer annealing and undesirable amplification. The cDNA banding pattern is highly comparable for the two thermal cyclers regardless of the ramp time (Fig. 1).
Nucleic Acids Research
Pages
Acknowledgements
References
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 23 Jan 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. W. Tregear, F. Morcillo, F. Richaud, A. Berger, R. Singh, S. C. Cheah, C. Hartmann, A. Rival, and Y. Duval Characterization of a defensin gene expressed in oil palm inflorescences: induction during tissue culture and possible association with epigenetic somaclonal variation events J. Exp. Bot., June 1, 2002; 53(373): 1387 - 1396. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Sturtevant Applications of Differential-Display Reverse Transcription-PCR to Molecular Pathogenesis and Medical Mycology Clin. Microbiol. Rev., July 1, 2000; 13(3): 408 - 427. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. R. Miles, J. P. Sluka, R. F. Santerre, L. V. Hale, L. Bloem, G. Boguslawski, K. Thirunavukkarasu, J. M. Hock, and J. E. Onyia Dynamic Regulation of RGS2 in Bone: Potential New Insights into Parathyroid Hormone Signaling Mechanisms Endocrinology, January 1, 2000; 141(1): 28 - 36. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||