Beyond the Identification of Transcribed
Functional and Expression Analysis
11th Annual Workshop
November 9-12, 2001
John G. Conboy
Lawrence Berkeley National Laboratory
1 Cyclotron Road
Berkeley, CA 94720
presenter: John G, Conboy
Victor C. Hou, Robert Lersch, Sherry L. Gee, Marilyn Parra, Michael Wu, Chris W. Turck, Mark Koury, Adrian R. Krainer, Akila Mayeda, and John G. Conboy
Lawrence Berkeley National Laboratory, Life Sciences Division/ Mailstop 74-157, 1 Cyclotron Road, Berkeley, CA 94720; University of California, Berkeley, Dept. of Molecular and Cellular Biology, Berkeley, CA 94720; University of California, San Francisco, Howard Hughes Medical Institute, Dept. of Medicine and Cardiovascular Research Institute, San Francisco, CA 94143; Dept. of Medicine, Vanderbilt University, Veterans Affairs Medical Centers, Nashville, TN 37232; Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724; University of Miami School of Medicine, Dept. of Biochemistry and Molecular Biology, Miami, FL 33136.
The single gene for erythroid skeletal protein 4.1R is an extraordinarily complex locus on human chromosome 1 that encodes a large family of isoforms varying in size, subcellular localization, and functional interactions with other proteins. Much of this diversity is generated through the mechanism of tissue-specific alternative pre-mRNA splicing and, preliminary evidence suggests, alternative transcriptional events at multiple alternative first exons. The magnitude of the complexity is such that 15 of the 26 known exons comprising this large gene (>220kb) are alternatively expressed via these processes, and several are regulated in tissue- and developmental-specific patterns. Similarly complex regulation of RNA processing also characterizes expression of three paralogous genes, 4.1G, 4.1N, and 4.1B. In the 4.1 gene family alone, there are multiple alternative exons with tissue-specific or tissue-restricted patterns of expression in erythroid cells, epithelial cells, muscle, and brain. Given recent estimates that a majority of human genes exhibit some form of alternative splicing, it seems clear that regulation of RNA processing is a major mechanism of cellular differentiation, and that identification of genomic regulatory sequences is of considerable importance.
We are studying the molecular mechanisms regulating three regulated splicing events in the 4.1R gene: activation of exon 16 in differentiating erythroid cells; activation of exon 17B in mammary epithelial cells; and regulation of exon 2' splicing that controls expression of an alternative translation initiation site. Here we will focus on recent experiments that provide new insights into the regulation of exon 16 (E16). E16 expression is controlled by a physiologically important alternative splicing "switch" that operates at a specific stage of erythroid differentiation. E16, 63nt in length, encodes a critical region of the spectrin-actin binding domain and is essential for normal membrane mechanical properties. E16 is skipped during 4.1R pre-mRNA splicing in early erythroid progenitors but efficiently included in 4.1R mRNA in later progenitors. A major goal of these studies is to identify both the regulatory sequences in 4.1 pre-mRNA and the splicing factors that bind to these sequences to mediate this splicing switch.
For these studies we constructed model 4.1R pre-mRNAs containing wild type
or mutated E16 sequences, assayed the splicing of these pre-mRNAs in vitro (using
HeLa cell nuclear extracts) or in vivo (in transfected cells), and characterized
the spliced products using RT/PCT techniques. Recent experiments have identified
two elements within exon 16- a 5' purine-rich element (PRE16), and 3' evolutionarily
conserved element (CE16)- that are critical to the repression of E16 splicing
in early erythroid cells. A combination of experimental approaches including
site specific mutagenesis of exon 16, RNA-protein binding assays, splicing factor
depletion and add-back experiments with recombinant proteins, and Western blot
analysis of splicing factors in differentiating erythroid cells, all support
the following model. In early erythroid progenitors, high nuclear concentrations
of hnRNP A1 protein, a known splicing inhibitor protein, represses exon 16 splicing
by binding to two "silencer" elements within exon 16. One site is
located in PRE16 and a second site is in CE16; both silencers are required for
optimal repression. As erythroblasts mature, there is a several-fold decrease
in cellular A1 levels that temporally correlates with the activation of exon
16 splicing. Control experiments suggest that these effects are specific, i.e.,
that other known splicing factors do not correlate in activity or expression
patterns with the E16 splicing switch. Remarkably, the correlated changes in
A1 expression and E16 splicing can be observed in culture using a mouse erythropoiesis
model system. These findings demonstrate that natural developmental changes
in hnRNP A protein expression can effect physiologically important switches
in pre-mRNA splicing through interactions with silencer elements within E16.
We speculate further that detailed biochemical studies of this type, in combination
with computational analysis of nucleotide sequences near other regulated alternative
exons, will eventually reveal the rules governing tissue-specific regulation
of splicing in higher eukaryotes.