In order to understand how genes are regulated in the heart, many gene promoters have been isolated and characterised with regard to the key regulatory DNA sequences they harbour and the transcription factors that bind them. A widely used method of measuring promoter function is to insert (clone) the promoter and various lengths of upstream sequence into an artificial plasmid-based construct in front of a reporter gene whose expression can be easily monitored. Typically the firefly gene luciferase or the bacterial genes chloramphenicol acetyl transferase (CAT) and P-galactosidase (LacZ) are used for this purpose. Constructs can be introduced into cells in culture by a variety of transfection techniques, or into whole animals using transgenic approaches. By careful choice of expression constructs containing progressively smaller deletions, the positions of DNA sequences responsible for high level transcription, tissue specificity or specific responses to stimuli (for example, stretch or agonist induced hypertrophy) can be found. The binding sites for candidate transcription factors can then be pinpointed accurately through mutagenesis of individual nucleotides to test the validity of those sequences. In this way, we and others have dissected the promoter of the cardiac troponin I gene, which is only ever expressed in cardiac muscle.
Deletion analysis of the human gene promoter and upstream sequences has revealed several important transcription factor binding sites within 100 nucleotides upstream of + 1 (the conventional notation for promoter sequence is negative numbering running upstream from the transcription start site). Among these is an A/T rich element centred around -30 that binds the TATA-box factor, TBP, the octamer protein Oct-1, and several MEF2 proteins. There are two binding sites in the human TnIc gene for GATA-4 and a C-rich sequence around -95 that encompasses both binding sites for the zinc finger factor Sp1 and a CACC-box with the sequence CCCACCCC.20 Mutation of each site in TnIc promoter-CAT reporter constructs results in a 50-95% reduction in transcriptional activity when transfected into cultured cardiac myocytes, suggesting that each site serves to bind proteins involved in maximal transcriptional activity. The identity of these proteins was characterised using the electrophoretic mobility shift assay (EMSA), also known as band shift assay. In this assay, a radiolabelled double stranded DNA fragment containing a putative binding site for a transcription factor (usually referred to as a cassette and generated by annealing short synthetic oligonucleotides corresponding to the two complementary strands of DNA) is mixed with an extract of nuclear proteins prepared from cells or tissue. If a protein binds the cassette, it will retard its migration compared to unbound cassette when subjected to electrophoresis through a non-denaturing polyacrylamide gel on account of the added mass of the protein.
Mutagenesis of individual nucleotides in the cassette or titration of unlabelled competitor cassettes in molar excess enable us to analyse the specificity of interaction between factor and DNA. By incubating the protein-DNA complex with antibody to putative factors, a "supershift" complex can be obtained (due to the added mass of the antibody) and the identity of bound proteins thereby confirmed (fig 28.3). Furthermore, by using nuclear extracts from different cells, an indication of the distribution of the factor can be determined. For example, our group has recently found shown that of four proteins binding the human TnIc CACC-box, two are members of the widely expressed Sp family of zinc finger factors while the other two appear to be expressed only in cardiac myocytes.21 These experiments therefore identify the regions involved and the factors they bind. Similar experiments in mouse have been taken further by showing that only 230bp of the mouse TnIc promoter are necessary to drive cardiac restricted expression of a LacZ reporter gene in transgenic mice.22
The role of specific factors in regulating a promoter can be assessed by simultaneous introduction (co-transfection) of suitable promoter-reporter constructs with an expression construct encoding the factor(s) in question. For example, the role of GATA-4 in regulating cardiac specific expression has been examined in a number of contexts. In an elegant experi
Was this article helpful?