Genetic Significance of Microbial DNA Composition

GENETIC SIGNIFICANCE OFMICROBIAL DNA COMPOSITION FRANK LANNI, Ph.D.* Now that deoxyribonucleic acid (DNA) in some microorganisms and ribonucleic acid (RNA) in others are widely accepted as the material carriers ofgenetic information, and now that both the functional genes and the nucleicacid macromolecules are conceived as linear structures minutely differentiated along their lengths, attention has focused sharply on details ofmicrostructure (1-5). The root idea, which is still only a brilliant hypothesis , is that an intimate correspondence exists between the sequentially ordered mutational sites that make up a functional gene and the similarly ordered nucleotides, or nucleotide pairs, that make up the carrier molecule ofnucleic acid or the relevant segment ofsuch a molecule. There is ajustified hope that experimental validation or invalidation ofthe idea will soon be forthcoming. There is also a danger. The intense and necessary preoccupation with the fine structure ofgenes and molecules can easily divert attention from significant relations at higher structural levels. In fact, so great is the devotion to fine structure that one can sense in some quarters a skepticism about the value ofa panoramic view. This essay proposes first that near-identity in overall DNA composition is a highly correlative and heuristic feature ofgenetically interacting microbes (6). The later part of the discussion deals with somewhat more speculative aspects ofthe genetic significance ofDNA composition. Occasional reference is made to microbes with RNA genomes. * Department ofBacteriology, Emory University, Atlanta 22, Georgia. TTie author's research is supported in part by grant E-837 from the National Institute ofAllergy and Infectious Diseases, U.S. Public Health Service. The author wishes to express his appreciation to numerous colleagues, at Emory University and elsewhere, for their valuable suggestions and criticisms, and especially to Dr.John M. Reiner for his inspiration and guidance. 418 Frank Lanni · Microbial DNA Perspectives in Biology and Medicine · Spring i960 I. Correlation between DNA Composition and Genetic Homology Associated closely with the root idea is a second familiar idea: namely, that genotypic similarities and differences between two organisms reflect similarities and differences in the nucleotide fine structure of their DNA. A testable consequence should be that organisms which show extensive phenotypic similarity—as judged conveniently for bacteria by their close association in a determinative scheme such as Bergey's—ought to resemble one another in DNA base composition. The monumental studies ofbacterial DNA by Lee, Wahl, and Barbu (7) and later studies by Spirin and colleagues (8) havelargely confirmed this expectation. Thereare, however, many notable exceptions, as might be expected from the notoriously arbitrary character ofphenotypic classifications. To propose DNA composition itself as a taxonomic criterion may be valuable for taxonomy but would merely beg some ofthe questions that we have in mind. Accordingly, we look for a more reliable sign ofactual genetic homology . We find it easily in the ability of two microorganisms to undergo specific genomic interactions (matings), as evidenced by genetic recombination and certain other phenomena. In current thinking, these interactions involve specific molecular pairings, which themselves imply considerable similarity in the nucleotide fine structure of the genomes. If, moreover, the homologous portions oftwo genomes were to encompass a large number ofnucleotides, it would be reasonable to expect a similarity in overall base composition of the appropriate nucleic acids. For DNA genomes, it seems immaterial for our purpose whether the genomes mate directly or whether the actual pairing occurs between accurately coded RNA intermediaries (9). A survey ofthe literature reveals a striking correlation. Taken in pairs, all microorganisms whose genomes are known or reasonably suspected to be at least partially homologous (through direct or indirect signs ofinteraction ), and whose DNA composition is known or may be confidently surmised , are effectively indistinguishable as to quantitative DNA base composition . Conversely, microorganisms for which there is no positive sign or reasonable suspicion ofgenetic homology may, and frequently do, differ appreciably in base composition. The relevant data are gathered in Table 1, which lists DNA analyses for the well-knownTphages ofEscherichia coli B, several temperate phages or 419 their virulent mutants, and selected bacteria. The tabulation neglects trace constituents in the DNA ofseveral of the listed organisms. In the following discussion the characteristic measure AT/GC— (adenine + thymine)/(guanine + cytosine)—is used for briefspecification of...