FIGURE I.I Complexity of egg RNA plotted against genome size for different species. Data are from RNA excess hybridization against single-copy DNA or cDNA hybridization kinetics, and are expressed in terms of length of single-copy DNA sequence represented in the egg RNA. Echinoderms are shown in red: 2, Arbacia; 4, Strongylocentrotus; 6, Lytechinus; 7, Tripneustes (all sea urchins); I0, Pisaster (a starfish). Dipteran insects are in black: I, Drosophila; 3, Musca (the housefly). Amphibians are in blue: 8, Xenopus; 9, Triturus (a newt). 5, Urechis (a nonsegmented worm) is in green. Points connected by dashed lines indicate two independent measurements. [Adapted from Davidson (1986) "Gene Activity in Early Development." Academic Press, Orlando, FL, where original references to both genome size and RNA complexity measurements can be found.]
of these two creatures are generated during the lampbrush chromosome stages of oogenesis, in which the transcription complexes are spectacularly displayed in the thousands of lateral chromosomal loops. Transcriptional processes in the large lampbrush chromosome loops of Triturus and the much smaller ones of Xenopus were thoroughly studied by cytological and biochemical methods, particularly by O. Miller and J. Gall and their colleagues (see Davidson, 1986, for review). As a result, we understand something of how such differently sized genomes generate cytoplasmic RNAs of such similar complexity. A very short summary is that both the noncoding transcribed sequence (i.e., intergenic and intronic sequences) and the amount of nontranscribed sequence is far greater in Triturus lampbrush chromosomes than in the equivalent Xenopus chromosomes. Thus, as one might expect, in the two species about the same amount of genetic information indeed ends up being used for the equivalent jobs of making an egg that will develop into an amphibian embryo.
Global estimates of gene content per genome have now become available for several species of different genome size as a result of genome sequencing, and these are summarized in Fig. 1.2 (reviewed by Rubin et al., 2000; see legend for detailed references). In this book we are to be concerned almost exclusively with bilaterally organized animals, sometimes referred to as the "triploblasts," here as the "bilaterians;" this excludes protozoans, sponges, cnidarians, and ctenophores. Only the Caenorhabditis elegans, Drosophila, and human values can be taken literally, since the others are all based on small samples of genomic sequence. However, the general import is not very dependent on the exact accuracy of the estimates. Even though the data in Fig. 1.2 pertain to only a few species, we can see from them the size of the basic "package" of protein coding genes needed in the genome of a bilaterian animal. Figure 1.2 shows that the C. elegans genome contains about 18,400 genes, and the invertebrate chordate Ciona and the arthropod Drosophila, apparently a slightly smaller number. Estimates for Strongylocen-trotuspurpuratus, a sea urchin, are about twice the value for Drosophila, which is estimated at 13,600 genes (Rubin et al., 2000). Two vertebrates, i.e., ourselves and the puffer fish (Fugu rubripes, the genome size of which lies at the lower limits known for vertebrates) probably have 70,000 genes. It is widely believed that early in the evolution of the vertebrates at least two whole genome duplications occurred, mainly on the basis of the larger number of copies of genes of given gene families in vertebrate genomes. That is, vertebrate genomes include more genes of these families than do lower deuterostomes, i.e., ascidians, echinoderms, and cephalochordates (amphioxus). The clearest argument for this is afforded by the box gene complexes: amniotes have four box gene complexes located on four different chromosomes (Ruddle et al., 1994; McGinnis and Krumlauf, 1992), while the invertebrate chordate amphioxus has one (Garcia-Fernandez and Holland, 1994), as does S. purpuratus (Martinez et al., 1999). If we divide the estimate of 80,000 human genes (Rubin et al., 2000) by four, we end up with a number slightly larger than the "basic package" value, that is, the 2 x 104 genes implied by the first three entries in Fig. 1.2.
Was this article helpful?