Sometimes the whole is not equal to the sum of its parts. Changes in intracellular Ca²⁺ concentration affect whole transcript abundance. However, Ca²⁺ also affects transcript architecture, which is not measured by traditional high-throughput genomics techniques. How can researchers detect alternative splicing and alternative transcription initiation and termination across the genome? McKee et al. use exon-centric microarrays to profile Ca²⁺-regulated exon expression in a recent article in Genome Biology.

Alternative splicing is an important form of gene regulation in the brain, which shows the highest rate of alternative splicing relative to other tissues. Exon-centric microarrays use exon-specific probes to detect stimulus-induced changes in individual exons or whole transcripts, indicated by coordinated changes across a gene's exons.

The authors treated human neuroblastoma cells with depolarizing concentrations of KCl, which induces an influx of extracellular Ca²⁺, or thapsigargin, which induces the release of Ca²⁺ from the endoplasmic reticulum. Using the exon-centric arrays, they identified 1505 and 3489 whole transcripts that were regulated by KCl and thapsigargin treatment, respectively. Although nearly all of the transcripts regulated by KCl and thapsigargin contained regulated exons, not all regulated exons belonged to regulated transcripts. KCl and thapsigargin respectively regulated 2979 and 1983 exons without associated changes in their transcripts.

The authors focused on transcripts regulated by KCl. They clustered transcripts that showed similar expression changes over time. Gene ontology (GO) analysis indicated that KCl treatment downregulated transcripts involved in metabolism and upregulated transcripts involved in Ca²⁺ ion binding, retrograde vesicle trafficking and transcription.

PCR with exon-spanning primers verified a subset of exon-specific expression changes. The exon dataset was enriched in exons at the terminal end of transcripts and at known alternative splice sites. GO analysis showed enrichment of similar gene categories in the exon-specific and whole transcript datasets. However, whereas Ca²⁺ ion-binding transcripts were upregulated soon after KCl treatment and declined thereafter, genes involved in Ca²⁺ ion binding remained elevated 24 hours following KCl treatment in the exon dataset, and the exon dataset contained Ca²⁺ ion-binding genes that were not in the transcript dataset. Different functional classes of gene exons changed over time. Immediately after KCl treatment, kinase and kinase-regulating proteins were over-represented in the exon database. Then exons from genes encoding proteins involved in metabolism and cell cycle control were upregulated, and genes encoding proteins involved in lipid, carbohydrate and nucleic acid synthesis and cyclin and cyclin-dependent kinases were downregulated. Exons from genes involved in RNA processing were regulated several hours after KCl treatment.

Several genes in the exon dataset are ion channel components. Ca²⁺ induced a C-terminal truncation of the brain-specific voltage-gated potassium channel KCNH4. This truncation would delete most of KCNH4's cytoplasmic region, which is important in protein-protein interactions and may affect channel conduction or gating. Ca²⁺ also regulated alternative splicing of H⁺ channels in the V_o vacuolar proton pump, which are important in recycling neurotransmitters.

Therefore, exon-specific microarrays identified stimulus-induced changes in genetic architecture and might be used to identify unwanted genome-wide side effects during the drug discovery process.

1. McKee, A. E. et al. Exon expression profiling reveals stimulus-mediated exon use in neural cells. Genome Biology 8, R159 (2007). | Article |