Improved differential screening approach to analyse transcriptional variations in organized cDNA libraries
Introduction
In recent years, numerous genome sequencing projects and systematic analysis of cDNA libraries (Adams et al., 1991, Adams et al., 1992, Adams et al., 1995) have led to the identification of an increasing number of new gene sequences. However, these approaches do not usually permit functional characterization of the genes.
The temporal, developmental, topographical and physiological patterns in which a gene is expressed provide clues to its biological role (Schena et al., 1995). Hence, it seems possible to obtain hints on functional information about genes by identifying those which are differentially expressed under distinct biological conditions (Calvet, 1991). The differences in gene expression may be measured by analysing the corresponding variations in mRNA expression levels.
Several experimental approaches can be used to identify genes based on their expression pattern: differential PCR display (Liang and Pardee, 1992; Wan et al., 1996); subtractive libraries (Sargent, 1987; Nedevi et al., 1993), suppression subtractive hybridization (Diatchenko et al., 1996), serial analysis of gene expression (Velculescu et al., 1995) and differential screening on organized cDNA libraries (Okubo et al., 1992; Zhao et al., 1995; Takahashi et al., 1995; Nguyen et al., 1995). The techniques based on PCR, such as differential PCR display, are potentially the fastest methods for identifying differentially expressed genes, using relatively small amounts of mRNA and without the necessity for constructing cDNA libraries. However, these methods produce a high level of false positives and artifactual responses (Sompayrac et al., 1995; Bertioli et al., 1995).
Despite the successful identification of numerous important genes using subtraction techniques, these methods require a great amount of poly A+ RNA. In addition, they involve a high number of controls and often, multiple subtraction steps, which render reproducibility nearly impossible.
Among the nucleic acid techniques, the hybridization of arrayed cDNA libraries by complex probes enables one to visualize the expression level of many thousands of genes corresponding to a given biological state. Therefore, differential screening analysis permits the comparison of several biological conditions and the identification of transcripts which show significant variations, thus allowing for the accumulation of biological information for the whole set of clones present in the arrayed library.
The success of the differential screening approach mainly depends on: (1) the quality and quantity of clones present in the cDNA reference library; (2) the possibility to rapidly organize, replicate and store a large number of clones; (3) the use of complex probes with similar cDNA patterns, capable of detecting weakly expressed genes; (4) the availability of an analytical framework to quantitatively correlate information resulting from the hybridization of the library with different probes.
We present here a strategy that fulfils most of the abovementioned conditions. We have constructed a cDNA library from mRNA extracted from different rat brain tissues. The library contained approx. 10 000 clones, with a low number of redundancies and approx. 60% of full-length cDNA sequences. It was organized into 1536-well dishes, using a fluorescence activated cell sorter (FACS) acting as a single cell deposition system (Magazin et al., 1992). High density colony filters were prepared using a modified BIOMEK 1000 workstation. The identical replicates were screened with different probes capable of detecting weakly expressed genes. The probes were specially prepared to have similar cDNA size profiles and the same specific activity.
We also show how a standard software package (Visage 2D Investigator Database of BioImage) can be adapted to analyse digital images corresponding to the hybridization of the filters with different probes. In addition, we propose a normalization method, based on the available quantity of plasmid material on the filters, that greatly improves the performance of the differential screening approach.
Altogether, the above technical improvements open the possibility to compare a great number of different probes and, in consequence, to accumulate biological information for each clone present in an organized cDNA library.
Section snippets
Construction of a cDNA reference library
60 male, 8–10-week-old Sprague–Dawley rats were used; brains were dissected for the isolation of cortex, septum, striatum and hippocampus and stored at −80°C. RNA was extracted from the isolated tissues by the acid phenol/guanidine thiocyanate method of Chomczynski and Sacchi (1987). RNA poly A+ was prepared from rat brain using oligo (dT)-magnetic beads (Dynal) and quantified using 3H-poly(U). mRNAs were pooled and used for the preparation of cDNA using SuperScript II reverse transcriptase
Quality of the reference library
Successful differential screening depends on the quality and quantity of clones present in the reference library. In our case, the total content of the library was approx. 106 clones. A fraction of them was used to obtain a cDNA library array containing approx. 104 clones (ten plates of 1536 wells), following the above described experimental strategy (Fig. 1).
The quality of the reference library can be extrapolated from the analysis of 298 cDNA clones that were selected as described in Section
Discussion
Systematic sequencing of human cDNA libraries is generating lists of genes expressed in many different tissues (Adams et al., 1995). It is difficult and time-consuming to determine the biological function of a gene merely from knowledge of its sequence. However, useful functional information may be obtained by identifying gene expression variations in cells or tissues under different biological conditions. Therefore, the use of the differential screening approach to analyse transcriptional
Acknowledgements
We thank D. Shire and M. Magazin for valuable comments and corrections to the manuscript.
References (38)
- et al.
Basic local alignment search tool
J. Mol. Biol.
(1990) - et al.
Single-step method of RNA isolation by acidic guanidinium thiocyanate–phenol–chloroform extraction
Anal. Biochem.
(1987) - et al.
Hybridization analyses of arrayed cDNA libraries
Trends in Genetics
(1991) - et al.
Differential gene expression in the murine thymus assayed by quantitative hybridization of arrayed cDNA clones
Genomics
(1995) - et al.
High-density cDNA filter analysis of the expression profiles of the genes preferentially expressed in human brain
Gene
(1995) - et al.
High-density cDNA filter analysis: a novel approach for large-scale quantitative analysis of gene expression
Gene
(1995) - not in ref....
Complementary DNA sequencing: expressed sequence tags and human genome project
Science
(1991)- et al.
Sequence identification of 2,375 human brain genes
Nature
(1992) Initial assessment of human gene diversity and expression patterns based upon 83 million nucleotides of cDNA sequence
Nature
(1995)
Multiplex messenger assay: simultaneous, quantitative measurement of expression of many genes in the context of T cell activation
Nucl. Acids Res.
An analysis of differential display shows a strong bias towards high copy number mRNAs
Nucl. Acids Res.
Molecular approaches for analysing differential gene expression: differential cDNA library construction and screening
Pediatr. Nephrol.
Identification of a common nucleotide sequence in the 3′-untranslated region of mRNA molecules specifying inflammatory mediators
Proc. Natl. Acad. Sci. USA
Use a cDNA microarray to analyse gene expression patterns in human cancer
Nature Genetics
A comprehensive set of sequence analysis programs for the VAX
Nucl. Acids Res.
Suppression subtractive hybridization: a method for generating differentially regulated or tissue-specific cDNA probes and libraries
Proc. Natl. Acad. Sci. USA
Discovery and analysis of inflammatory disease-related genes using cDNA microarrays
Proc. Natl. Acad. Sci. USA
Cited by (10)
Statistical issues with microarrays: Processing and analysis
2002, Trends in GeneticsCitation Excerpt :Systematic errors (bias) are controlled experimentally as far as possible, although additional statistical correction is invariably necessary with current microarray technology. A partial list of extraneous factors that contribute to random error and/or to bias of microarray expression values includes: the time of day that arrays are processed [13]; target accessibility, which is affected by variations in the absorbency of discrete nylon membranes [14]; target fixation to glass slides [2]; and variations in washing procedures [15]. Random error is a measure of uncertainty in the measurement and is therefore central to statistical inference.
Single cell sorting and cloning
2000, Journal of Immunological MethodsFrom the molecular biology of prolactin and its receptor to the lessons learned from knockout mice models
1999, Genetic Analysis - Biomolecular EngineeringRNA expression profiles and data mining of sugarcane response to low temperature
2003, Plant Physiology