ReviewProfiling brain transcription: neurons learn a lesson from yeast
Introduction
DNA microarrays have become an integral part of the research tools used in the laboratory, and have been avidly adopted by molecular neurobiologists. Boosted by the completion of genome sequencing projects in various model organisms and in humans, DNA chips hold great promise for the systematic measurement of complete transcriptional programs in any tissue or cell type, and at any stage of a given physiological, developmental or pathological process. As we have learned recently from studies on yeast, expression profiling provides access to entire regulatory mechanisms and networks in any given physiological process by revealing concerted and genome-wide changes in transcription 1., 2., 3., 4.. In doing so, large-scale expression analysis reveals interrelated processes, for example the upregulation of transcripts involved in wound healing during the physiological response of human fibroblasts to serum [5]. Furthermore, by inferring that the relative abundance of specific transcripts is a response to specific cellular needs, the function of many previously uncharacterized genes, whose expression level tightly correlates with the time course of a specific event and transcriptional changes in known genes, has been tentatively predicted. Finally, the global transcription profile itself is a direct representation of cellular phenotype. By providing a complex, but accurate cellular identification, gene expression profiling leads the way to the generation of new diagnostic and prognostic tools, as recently demonstrated for tumor classification 6••., 7••., 8••..
These characteristics make biological discovery with microarray analysis extraordinarily tantalizing for molecular neuroscientists. The ability to monitor in parallel the expression of tens of thousands of transcripts in each biological sample is likely to provide insightful information into issues of extreme complexity: the cellular commitment to various neuronal cell fates; the formation and maintenance of appropriate neural connections; the identity of neuronal circuits involved in cognitive processes and behavioral arrays; and the identity and progression of neurological diseases, being some examples.
When, then, will the neurochip revolution occur? Following a relatively late start compared to other fields in biology [9], a small but increasing number of publications report the transcriptional profiling of neurons. As we discuss in this review, data extracted from these initial microarray analyses represent useful proof of principles that enlighten the specific challenges and unique promises of large-scale profiling of brain transcription. For a more general description of microarray technology, its method of analysis, and its fields of application outside neuroscience, the reader is referred to a number of recent reviews 10., 11., 12..
Section snippets
Molecular characterization of brain units
The vertebrate brain is subdivided into anatomically and functionally distinct regions and nuclei. Stereotaxic lesions and, more recently, functional imaging have correlated specific cytoarchitectonical units with defined physiological functions. The molecular characterization of these cellular units is now a critical step for further functional analysis. Knowledge of the entire molecular composition of a given brain compartment—its specific set of neurotransmitters and associated processing
Developmental time points
Brain development is regulated by intricate signaling cascades and involves ubiquitous neural maturation events, as well as differentiation processes specific to a given neuronal cell type. Direct comparison of the transcriptional profiles of a particular neural structure, at different developmental time points, may permit the characterization of sets of genes that specify its unique identity, and the direct visualization of the coordinated developmental events that are essential for its proper
A step towards the study of complex phenotypic traits and neurological diseases
The ability to follow orchestrated and genome-wide gene expression has led several groups to identify a complex issue at the core of brain function: can one identify transcriptional differences associated with specific behavioral or neuropathological traits?
Are existing microarrays adapted to the needs of neuroscience?
An immediate problem in neurogenomic studies arises from the fact that public databases are biased toward most abundant genes and ESTs but, because of the complexity of the nervous system, neuronal-specific genes are either less abundant or only expressed in specific cell types. For mammalian species, in contrast to organisms like yeast and Caenorhabditis elegans, most commercial microarrays are designed on the basis of known genes and ESTs, and not on whole-genome information, and therefore
Switching from multi- to single-unit molecular recording
One of the biggest challenges for microarray technologies in the neurosciences is how to deal with the marked cellular heterogeneity and low expression levels in many genes of the nervous system. Existing techniques require large amounts of starting materials. In general, a few million cells are needed to obtain enough RNA for a single array experiment. However, nervous systems are often composed of heterogeneous cell populations that are difficult to distinguish on the basis of their location
Conclusions: complex system, complex tools, complex message
A survey of the literature on mammalian brain transcriptional profiling together with comparisons with advances in other experimental fields and organisms underlines some specific challenges for the use for microarray technologies in neuroscience. Not surprisingly, when studying a complex experimental system such as the brain, with a genome-wide analytical tool, the message provided appears extraordinarily complex. The question now is whether we have the right means to decode this message, and
Acknowledgements
We acknowledge Lubert Stryer, Andrew Murray, Ian Tietjen, Jason Rihel and Lisa Stowers for critical reading of the manuscript. We thank Cecilia Lee for help in preparing the text. C Dulac is supported by the National Institute on Deafness and Other Communication Disorders (Grant 3903-01), the Searle Scholar Program and the Howard Hughes Medical Institute.
References and recommended reading
Papers of particular interest, published within the annual period of review, have been highlighted as:
•of special interest
••of outstanding interest
References (44)
- et al.
A genome-wide transcriptional analysis of the mitotic cell cycle
Mol Cell
(1998) - et al.
A screen for genes required for meiosis and spore formation based on whole-genome expression
Curr Biol
(2001) Of neurons and gene chips
Curr Opin Neurobiol
(1999)Biomedical discovery with DNA arrays
Cell
(2000)- et al.
Microarray analysis of the transcriptional network controlled by the photoreceptor homeobox gene Crx
Curr Biol
(2000) - et al.
Functional discovery via a compendium of expression profiles
Cell
(2000) - et al.
A genetic analysis of neural progenitor differentiation
Neuron
(2001) - et al.
Upregulation of calcineurin Aβ mRNA in the Alzheimer's disease brain: assessment by cDNA microarray
Biochem Biophys Res Commun
(2001) - et al.
Altered expression of a-type but not b-type synapsin isoform in the brain of patients at high risk for Alzheimer's disease assessed by DNA microarray technique
Neurosci Lett
(2001) - et al.
Molecular characterization of schizophrenia viewed by microarray analysis of gene expression in prefrontal cortex
Neuron
(2000)