Verification of differential gene transcription using virtual northern blotting

doi:10.1093/nar/27.11.e3

This Article

	Abstract
	Alert me when this article is cited
	Alert me if a correction is posted

Services

	Email this article to a friend
	Similar articles in this journal
	Similar articles in PubMed
	Alert me to new issues of the journal
	Add to My Personal Archive
	Download to citation manager
	Request Permissions

Citing Articles

	Scopus Links
	Commercial Re-use Guidelines for Open Access NAR Content

Google Scholar

	Articles by Franz, O.
	Articles by Roeder, T.
	Search for Related Content

PubMed

	PubMed Citation
	Articles by Franz, O.
	Articles by Roeder, T.

Related Collections

	Nucleic acid amplification
	Miscellaneous/other
	Monitoring gene expression

Social Bookmarking

	What's this?

Nucleic Acids Research Article e3

Verification of differential gene transcription using virtual northern blotting
Introduction
Materials And Methods
Results And Discussion
Primers used for LA-PCR
Acknowledgements
References

Verification of differential gene transcription using virtual northern blotting

Oliver Franz, Iris Bruchhaus¹ and Thomas Roeder^*

Universität Hamburg, Zoologisches Institut, Neurophysiologie, Martin-Luther-King-Platz-3, D-20146 Hamburg, Germany and ¹Bernhard-Nocht-Institute for Tropical Medicine, Bernhard-Nocht-Straße 74, D-20359 Hamburg, Germany

Received January 13, 1999; Revised and Accepted April 19, 1999

ABSTRACT

We introduce here an alternative to conventional northern blotting that requires only minute amounts of RNA. This has been achieved by modification of methods currently used for the mapping of mRNA 5[prime]-terminal ends. The terminal desoxynucleotidyl transferase-mediated G-tailing, cap finder, ligation-anchored and RNA ligase-mediated approachesfollowed by polymerase chain reaction protocols all produced high quality cDNAs in large amounts. These cDNAs could be separated by electrophoresis to obtain virtual northern blots that could replace conventional northern blots. All the essential information, including transcript length and the expression pattern, are preserved in these cDNAs, even if the transcripts are long or GC-rich. In addition, minute amounts of material (less than 100 cells) are sufficient to produce more than 100 virtual northern blots, making this approach extremely versatile.

INTRODUCTION

Differential gene expression either in time or space is a fundamental phenomenon found in all biological systems. Knowledge of those genes that are differentially expressed should enable us to understand the molecular rationale underlying phenomena as diverse as tumorigenesis, development or tissue specification. To tackle these problems, methods such as subtractive hybridisation, polymerase chain reaction (PCR)-based mRNA fingerprinting and the differential display-PCR (DD-PCR) are preferred. DD-PCR and mRNA fingerprinting in particular have the potential, due to the inherent PCR amplification step, to start with minimal amounts of tissue sample. Unfortunately, these methods have a high probability of so-called false positives. Verification of differential expression, which is usually done by northern blot analysis or RNase protection assay, is the bottleneck in all these protocols. They are, in some cases, difficult to perform and require large amounts of material, counteracting one advantage of the DD-PCR approach, the requirement of small amounts of RNA.

To circumvent these problems, methods with intrinsic amplification steps are required. These approaches are usually used to map unknown 5[prime]-regions of mRNAs. To achieve this, it is necessary to add a known sequence to the 3[prime]-end of the first-strand cDNA. This allows PCR amplification of cDNAs enclosed by the new tag at their 3[prime]-end and a specific primer derived from the already known sequence. It is obvious that these approaches could also be used for the amplification of complex cDNA libraries, simply by replacing the specific 5[prime]-primer by an oligodT primer that tags the 5[prime]-region of all first-strand cDNAs. At least four different approaches for the characterisation of 5[prime]-mRNA ends are currently available. The first approach introduced was the homopolymeric tailing using terminal desoxynucleotidyl transferase and dGTPs (TdT) (1). This GTP tail could be recognised by an oligodC primer, thus tagging the 3[prime]-end of the first-strand cDNA with a sequence known to be suitable for PCR amplification. More recently, two other approaches have been introduced, where either an oligonucleotide is added to the first-strand cDNA or an oligoribonucleotide is added to the 5[prime]-end of the mRNA and both are catalysed by the T4 RNA polymerase (ligation-anchored and RNA ligase-mediated; LA and RLM, respectively) (1). The fourth approach, the so-called cap-finder (CF) cDNA depends on the ability of reverse transcriptase to add nucleotides to the 3[prime]-end of the newly synthesised cDNA after completion of first-strand synthesis (2). As the M-MLV RT preferentially adds cytosines, a second primer that contains a G-stretch at its 3[prime]-end is added to the reaction mixture. It could anneal to this complementary region and serves itself as a template for the reverse transcriptase, thus delivering cDNA with known sequences at the 5[prime]- and 3[prime]-terminal end. This approach was introduced by Clontech and termed `virtual' northern blotting (3). First attempts to replace conventional northern blots using this technique were reported recently (4). The main goal of this study is to evaluate whether these approaches are suited to produce high quality cDNA by LA-PCR (5) and whether these DNAs could be used to verify differential expression of cDNAs isolated from DD-PCR experiments by virtual northern blots, thus replacing conventional northern blots.

MATERIALS AND METHODS

To perform virtual northern blots, we started with equal amounts of mRNA. RNA quantities ranging from a maximum of 1 µg to the amount obtained from a few cells were used for the first-strand cDNA synthesis. RNA was isolated either from different parts of the locust (Schistocerca gregaria) brain (6) or from a few individuals of the human protozoan parasite Leishmania donovani. We isolated the RNA and mRNA according to standard protocols.

First-strand synthesis was primed by an anchored oligodT primer (oligodTSalIA) at elevated temperatures (48°C, 50 mM Tris-HCl pH 8.3, 8 mM MgCl₂, 10 µM dithiothreitol, 1 mM dNTPs, 2 pmol/µl oligodT primer, 100 U RNasin, 200 U SuperscriptII⁺; Life Technologies, Karlsruhe, Germany) and proceeded for 1 h. For the CF approach, tagging of the 3[prime]-end of the first-strand cDNA is simply achieved by addition of the CF primer (CFBI). Experimental details regarding the last three approaches are summarised by Schaefer (1). The cDNAs are now ready for LA-PCR amplifications with 5% of the first-strand cDNA reaction volume per 50 µl PCR mixture. Primers used for the LA-PCR reaction are, in the CF approach, similar to those used for the first-strand synthesis, but without the homopolymeric stretches. (LA-PCR reaction conditions: 95°C, 1 min; 60°C, 1 min; 68°C, 12 min for one cycle; 95°C, 30 s; 60°C, 30 s; 68°C, 12 min for seven cycles; seven cycles with an extension of 14 min; and seven additional cycles with an extension step of 16 min; 1 U TaKaRa ExTaq, 200 µM dNTPs, 2.5 mM MgCl₂ and 20 pmol of each primer using a Stratagene Robocycler). Five microlitres of these reaction mixtures (2-5 µg of amplified CF cDNA) were run on a 0.7% agarose gel, denatured and subsequently blotted onto a nylon membrane in a conventional Southern transfer [downward with 10× standard saline citrate (SSC) for 1-2 h] and UV-crosslinked. Radioactive as well as non-radioactive hybridisations were performed according to standard protocols.

RESULTS AND DISCUSSION

As outlined above, four different methods for the PCR-mediated amplification of cDNA libraries are currently in use. To evaluate whether the four different approaches are able to produce high quality cDNA in reasonable amounts, we amplified cDNA derived from the locust thoracic ganglia (6) with all four methods. Invertebrate mRNAs are from ~300 to 8000 bases in length. The tissue-specific appearance of highly abundant transcripts, seen as stronger extra bands, remained almost constant for all four cDNAs, which indicates that these cDNAs should be equally well suited for the construction of virtual northern blots. Care has to be taken that the PCR amplification is in the linear range. This is required to ensure that differential expression is mirrored in the corresponding amplified cDNAs. Pilot amplifications with different numbers of PCR cycles should be performed to choose the optimal conditions. The DNA smear produced should not be in the saturation phase, which is the case if an additional cycle yields about twice the amount of cDNA.

These virtual northern blots were either hybridised with a probe of a ubiquitous house-keeping gene (cytochrome oxidase C), or a transcript exclusively expressed in the thoracic ganglia. The probes were labelled with alkaline phosphatase, hybridised, washed and subsequently developed with CDPstar. As seen in Figure 1, the different virtual northern blots revealed signals of the expected length (~1.5 and 0.33 kb, respectively). This shows that these four methods, the TdT-, the CF-, the LA- and the RLM-cDNA, are almost equally suited for the production of virtual northern blots. These methods differ significantly in their experimental complexity. The CF-cDNA approach outscores the other approaches because only a single experimental step, the first-strand cDNA synthesis precedes the LA-PCR process. All other approaches consist of a chain of enzymatic reactions, which might introduce numerous failures. Nevertheless, cDNA produced by these procedures could be used for the detection of their unknown 5[prime]-ends by 5[prime]-RACE-PCR, which is almost impossible using CF-cDNAs.

Nucleic Acids Research