In the ten years since the first sequencing of the human genome, much has been made of the need to look to gene regulation, and not gene number or DNA sequence, for the evolution of organismal diversity and complexity - an issue that rose to prominence, with the realization first, that the number of human genes is about the same as the number required to specify a nematode worm; and second, that the DNA of H. sapiens is roughly 96% identical to that of the chimpanzee.
But the realization that the secret of evolution lies in changes in gene regulation considerably predates the revelations of genomics. Allan Wilson and colleagues, in a paper published in 1974 , drew attention to the simple and striking fact that morphologically homogeneous frog species also have relatively homogeneous karyotypes, whereas mammalian species, which are markedly diverse morphologically, also show major differences in chromosome number and organization; changes in proteins, by contrast, are much the same for both groups. They concluded that genome organization, and by implication gene regulation, is more important for metazoan evolution than protein sequence (and cite earlier publications of EB Ford and Susumu Ohno for the same insight). The following year, Mary-Claire King and Wilson published a more detailed examination of the chromosomal distinctions between human and chimpanzee , arguing compellingly, without benefit of high-throughput anything, that changes in the organization of the genome, and not changes in protein-coding sequence, must account for the crucial differences between the two primates.
In those pre-genomic days, the protein data were in large part immunological and electrophoretic; the analysis of genome reorganization depended on chromosome banding patterns (Giemsa banding, not FISH); and almost nothing was known of the mechanism of gene regulation in eukaryotes. The ground between then and now is covered in a recent review by Sean Carroll , who acknowledges Emile Zuckerkandl and Eric Davidson as early proponents of the importance of gene regulation in morphological evolution and charts the remarkable history of the development of ideas consequent on the discovery of the homeobox genes, with a strong emphasis on the evolution of cis-regulatory elements - that is to say, DNA binding sites for gene regulatory proteins - as the basis for morphological change. The argument is that DNA regulatory elements and the proteins that bind to them, often combinatorially, constitute regulatory networks that can evolve rapidly through changes to the regulatory elements, which are often modular, different modules binding different proteins characteristic of distinct differentiated states of a cell. The gene regulatory proteins can also change, of course, but are generally more highly conserved than their binding sites. Tuch et al.  have published a short and pellucid overview of the essential points and principles of this schema, in the context of recent evidence on how such regulatory circuits can become rewired in yeast.
In our video Q&A published today , John Mattick gives a personal account of his arguments for the view that the regulatory potential of proteins and their binding sites is not sufficient to account for the evolution of complex higher organisms, and explains his case for invoking a largely uncharted universe of regulatory RNAs.