Basic History
Plasmids: Histories of a ConceptA Basic History of Plasmid Research
Plasmid Early History Time-Line:
Particulate HeredityThe early history of the concept of a plasmid is rooted in the concept of particulate determinants of inheritance. In the first decade of the 20th century, the chromosome theory was developed and two key papers are usually cited: Johannsen and Boveri. These workers argued from diverse observations that the cytologically observable structures in the cell nucleus are the physical units that determine the Mendelian characters. Of course, it was very unclear just what a "Mendelian character" was. In the second decade of this century, Thomas Hunt Morgan and his group, in experiments with fruit flies (Drosophila melanogaster) presented evidence for the formal agreement of the behavior of several "Mendelian factors" and the behavior of the physical structures known as chromosomes. Morgan and his school generalized these results into a broad "Theory of the Gene" which held that the Mendelian factors (genes) were arranged linearly on the visible structures (chromosomes) that resided in the nucleus of every cell and which were duplicated and partitioned equally to the daughter cells during cell division. Thus was solved (at one level, at least) the age-old problem of "how like begets like". Many biologists took up this approach and gathered much evidence to supports its validity and universality.
Cytoplasmic ContributionsAt the same time, other biologists, working on problems of embryology and morphology, saw genes as determinants of the way an organism developed from a fertilized ovum into a mature adult. For them, genetics was not about transmission of characters across the generations, but about how gene action worked to make the organism a nearly exact copy of its parents, that is, a different version of the age-old problem of "how like begets like." For many of these biologists, the determinants of the characters involved in development and differentiation seemed to be neither obviously nuclear, nor chromosomal. Some of these genes seems to be passed on through cytoplasmic transfer. For example, in 1937 the eminent biologist Ross Harrison wrote (Science 85:372)"The prestige of success enjoyed by the gene theory might easily become a hindrance to the understanding of development by directing our attention solely to the genome, whereas cell movements, differentiation and in fact all developmental processes are actually effected by the cytoplasm. Already we have theories that refer the process of development to genic action and regard the whole performance as no more that the realization of the potencies of the genes. Such theories are altogether too one-sided." By the mid 1930s, these cytoplasmic determinants came to be known as "plasmagenes." Plasmagenes, however, were often invoked to explain the possible mechanisms of "inheritance of acquired traits" and played directly into the schemes of the Michurinist/Lysenkoist genetics in the Soviet Union. At the time, then, plasmagenes acquired the extra baggage of Cold War ideology (See Chapter 6: The Cold War in Genetics: Jan Sapp: Beyond the Gene, Oxford, 1987).
Genes in BacteriaThe existence of genes in bacteria was much debated in the first half of the 20th century. Without a visible nucleus, without visible chromosomes, without a known dimorphic sexual phase, and without many distinguishing charcteristics, it was easy to believe that bacteria were altogether different from organisms which reproduced sexually. In 1942 (Evolution: The Modern Synthesis, p. 131-132) the famous British biologist Julian Huxley wrote:"Bacteria (and a fortiori viruses, if they can be considered to be true organisms), in spite of occasional reports of a sexual cycle, appear to be not only wholly asexual but pre-mitotic. Their hereditary constitution is not differentiated into special parts with different functions. They have no genes in the sense of accurately quantized portions of hereditary substance; and therefore they have no need for the accurate division of the genetic system which is accomplished by mitosis.... We must, in fact, expect that the process of variation, heredity, and evolution in bacteria are quite different from the corresponding processes in multicellular organisms. But their secret has not yet been unraveled." By 1946, however, the experiments of J. Lederberg and E.L. Tatum (Nature 158:558) challenged and clarified the understanding of genes in bacteria. Without dealing with the physical nature of the genetic structures in bacteria (there was considerable debate about the existence of a bacterial nucleus), Lederberg and Tatum obtained clear support for a mating system in a bacterium ( Escherichia coli, strain K-12) and in 1947 Lederberg employed Morgan's paradigm of genetic linkage, established a genetic map in E. coli based on the frequency of recombination of genetic determinants observed in standardized "matings" (Genetics 32:505). At this time, the dominant model was based on the sexual processes in higher cells: cell fusion with zygote formation, recombination, and marker segregation and cell division. In 1949 L.L. Cavalli and H. Heslot (Nature164:1057), and in 1951, J. Lederberg (Science114:68) surveyed other strains of E. coli for their ability to mate with Lederberg's strain K-12, and found that only 9 of 140 isolates could mate with K-12. Thus, there seemed to be something peculiar about mating in E. coli. In London, William Hayes started to study the kinetics of the mating process in 1950 and at the suggestion of D. Mitchison, conceived of bacterial mating as an asymmetric process involving a gene donor and a gene acceptor. This model for bacterial mating fitted the data Hayes was obtaining in various bacterial matings much better than a classical cell fusion model, and in 1952 he published this work (Nature169:118) and presented it at meetings in the summer of 1952: James D. Watson described the event (The Double Helix. Atheneum,1968, pp. 141-142): "Bill's appearance was the sleeper of the three day gathering; before his talk no one except Cavalli-Sforza knew he existed. As soon as he had finished his unassuming report, however, everyone in the audience knew that a bombshell had exploded in the world of Joshua Lederberg!"
The F-factorThe directionality and polarity of the bacterial mating process, first suggested by Hayes, greatly clarified the understanding of bacterial genetics as studied by mating experiments. The problem of sexual compatibility, however, remained. The rather rare property of a given E. coli strain to mate was a puzzle. In 1952 J. Lederberg, L.L. Cavalli, and E.M. Lederberg (Genetics 37:720-730) and in 1953 Hayes (J. Gen. Microbiol. 8:72-88) independently reported that the ability to act as a donor in a bacterial mating was a property controlled by an "factor" designated "F" (fertility) that seemed to behave as "an infectious particle."
Lambda Bacteriophage and ColicinesIn the mid-1950s two other anomalous hereditary "factors" were discovered to behave as "infectious particles" as well. One was the bacteriophage lambda, a lysogenic phage found in E. coli K-12 by Esther Lederberg, and the other was the factor determining the production of colicine, a killer substance, produced by some strains of E. coli and studied by Pierre Frédéricq (originally discovered by André Gratia).
Plasmids and EpisomesIn a broad review of "Cell Genetics and Hereditary Symbiosis" (Physiol. Rev. 1952, pp. 403-430) Joshua Lederberg proposed that all "extrachromosomal hereditary determinants" be subsumed under the designation "plasmid." [To view excerpt from this reference Click Here] He did not distinguish nuclear or cytoplasmic location, nor the possibility of association of such determinants with the chromosome on some occasions. In a more limited review of bacterial genetic systems, F. Jacob and E. Wollman in 1958 ( Compt rend. acad. sci. 247:154) suggested that genetic elements which were optionally associated with the chromosomes of the cell be termed "episomes." [To view an excerpt from this reference Click Here]. They used the F-factor, the colicinogenic factor, and bacteriophage lambda as prototypic episomes. By this time it was known that the F- factor could become associated with the bacterial chromosome and result in the transfer of chromosomal genes with high frequency in mating experiments (Hfr strains).By the end of the 1950s the recognition of genetic determinants which were not able to be located on the genetic map in standard crosses established the concept of "plasmid" (episome was used rather interchangeably with plasmid by some, but William Hayes, for one, calling the F-factor "a small, supernumerary chromosome," stated (The Genetics of Bacteria and their Viruses, 2nd ed. Blackwell, 1964, pp. 747-748): "We think the word 'episome,' although an excellent substitute for 'plasmid,' has become a source of confusion because the existence of alternative chromosomal and cytoplasmic states was central to its original usage. .... It now seems to us that the most meaningful biological distinction is between plasmids which promote conjugation, which we will classify as sex factors, and those which do not."
Chromosomal AssociationsThe understanding of the possibility of the attachment (by some unknown mechanism, often diagramed as a "bump" on a linear diagram of a chromosome) of the F factor to the chromosome in Hfr strains probably helped the understanding of the linage of the fertility property and the genetic determinant for lactose fermentation (lac) in the work of F. Jacob and Edward A. Adelberg in 1959 (Compt. rend. acad. sci. 249:189). They concluded that the F-factor could become associated with cell genes which then became part of the "infectious hereditary particle" that was the F-factor. Soon these "augmented" F-factors became known as "F-prime" factors. Soon many variant F-primes were found and it was realized that F-prime plasmids carrying any desired part of the bacterial chromosome could be constructed. Elie Wollman recalled the history of F-prime factors (quoted in Thomas Brock, The Emergence of Bacterial Genetics, Cold Spring Harbor Press, 1990, p. 104):"Adelberg had brought back to Berkeley some of our Hfr strains. I spent the year 1958-59 in Berkeley -- finishing the writing of our book [Wollman and Jacob, 1959]. Once Ed Adelberg came to me telling me that one of the Hfr strains had changed: the frequency of recombinants was less than the expected, but all were donors of intermediate frequency. I suggested that, by comparison with HFT phage the sex factor had left its site accompanied by neighboring genetic fragments. This was verified experimentally. Lwoff, who had come to visit, brought the news back to Francois Jacob who immediately used it for making partial Lac diploids. This is the history of F prime factors."
The Physical Nature of Plasmids and Episomes: DNABy 1960 it was clear, of course, that "the genetic material is DNA", but the identification of cytoplasmic DNA was still questionable. Likewise, the structure of DNA in genes and in chromosomes was debated. The sizes of DNA "molecules" seemed to increase each year as the methods of preparation improved, and as the techniques for study of large, linear polymers got better. The recognition that shear forces could easily break large DNA molecules was especially important. Since bacteriophage were believed to be simple models for the genetic material of the cell, the nature of the DNA in phage was thought to be relevant. The sizes of the DNA in phages was rather ingeniously and indirectly determined by a technique known as "star gazing." This method compared the amount of DNA radioactivity (32-P) in a single phage particle, with the amount of radioactivity in the isolated DNA molecules released from the same phages under very gentle conditions. The radioactivity was detected by counting (under a microscope) the tracks in photographic emulsion in which the phages and the DNA were embedded. Each phage particle and each DNA molecule formed a "star" of such tracks. Since the number of tracks was the same for the intact phage particle and released DNA molecule, it was concluded that the DNA was present in the phage particle as one long piece (possibly held together by some non-DNA linkages). From the chemical composition of the phage and the bulk specific activity of the DNA, it was possible to calculate the molecular weight of the phage chromosome.That plasmids are DNA was rather conclusively demonstrated in physical experiments, first reported in 1961 by J. Mamur, R. Rownd, S. Falkow, L.S. Baron, C. Schildkraut and P. Doty (Proc. Natl. Acad. Sci. USA47: 972) who used the CsCl buoyant density separation of DNA based on nucleotide base composition to show that "light density E. coli-like DNA" appeared in Serratia marcescens (which has a somewhat "heavier" DNA) after transfer of the F-factor to Serratia. In 1962 S. Silver and H. Ozeki provided evidence for the same conclusion based on labeled DNA transfer of the colicine factor (Nature 195:875).
The "Campbell Model"Most experiments on the chemistry of DNA confirmed that DNA molecules were very long, linear, non-branched structures. How, then, to envision the attachment of episomes to the chromosome? In 1962, in a review on episomes (Adv. Genetics 11:101), Allan Campbell proposed a beautifully simple solution to this problem: the recombinational interaction of one circular molecule with another. [To view an excerpt from this reference Click Here]. "The Campbell Model" as it came to be known, explained the reversible association of some episomes with the chromosome, the inversion of the genetic map of lambda bacteriophage upon lysogeny as recently reported in 1960 by E. Calef and G. Licciardello (Virology 12:81), and the formation of double lysogens and defective heterogenotes in lambda phage (J. Whitfield and R. Appleyard, Virology 5:275, 1958). The apparently crucial insight of Campbell was that the episome must exist as a physical circular DNA structure. Interestingly he reasoned from the genetic map of phage T4 (there were, of course, no physical structures established for genomes).(p. 112) "Detailed linkage studies lead to the conclusion that the genome of one phage (T4) is indeed circular (Streisinger, Edgar and Harrar, quoted by Stahl (1961). If circularity is a property of phages in general, the equivalent of the insertion hypothesis is to the one circle out of two.[....] If the phage genome [now referring to lambda] is circular rather than linear, the lambda chromosome need not be split into parts [to account for map inversion in lysogens] but rather could be cut at a specific point on the circle when it lysogenizes. It is actually very simple (on paper) to insert a circular phage chromosome into a linear bacterial chromosome by reciprocal crossing over (Fig. 2)." It is, of course, interesting to note that while Campbell based his argument on the T4 genome, which turns out to be linear although it has a circular map, and he applied it to lambda which turns out to have a linear map, but a circular intracellular form.
Circular DNAWhile the genetics of plasmids pointed the way to circular forms, the chemistry of DNA was just becoming clearer. The key step in the modern concept of the plasmid was the confirmation that DNA molecules can, and often do, exist as circular structures. The first confirmation of this fact came again, from the study of phage biology. In an attempt to study the smallest life form, biologists had been studying bacteriophages, and the smallest known phages were two related phages phi-X174 and S13. Robert Sinsheimer had shown that phi-X was unusual in that it contained the single-stranded form of DNA rather than the double helical DNA of the Watson-Crick model. Using the recently characterized nucleases with specificity for exonucleolytic attack coupled with hydrodynamic studies, W. Fiers and R. Sinsheimer asserted in 1962 (J. Mol. Biol.5:408- 434) that phi-X DNA was in the form of a small circular, double stranded DNA molecule.This precedent for circular DNA molecules was soon followed by the discovery in 1963 of:
(1) the cohesive ends of the DNA of bacteriophage lambda and its ability to form circles (called
"folded molecules" at the time) by A. D. Hershey, E. Burgi, and L. Ingraham (Proc. Natl.
Acad. Sci. USA 49:748). Even though these studies with phage and viral DNAs provided the methods and concepts to characterize circular DNAs, the study of the physical nature of most plasmids was complicated by the difficulty in separation of the plasmid DNA from the mass of chromosomal DNA. This problem was solved in 1967 by the introduction of the dye-buoyant density method by R. Radloff, W. Bauer, and J. Vinograd (Proc. Natl. Acad. Sci. USA 57:1514). This method depended on the restriction on binding of a DNA-intercalating dye such as ethidium bromide by covalently closed circular DNA molecules in comparison to linear and nicked circular molecules. These dye-DNA complexes could be separated in density gradients of the dense salt CsCl formed in the ultracentrifuge. Plasmid DNAs were easily isolated by this method for detailed characterization and in 1968 M. Bazaral and D.R. Helinski applied this method to colicine factors E1, E2 and E3 and showed that these factors were circular DNA molecules of homogeneous molecular weights (J. Mol. Biol. 36:185). Beginning in 1959 (A.K. Kleinschmidt and R.K. Zahn, Zeitsch. Naturforsch. 14b:770-779, 1959) electron microscopic visualization began to be applied to DNA molecules spread in protein films, and this technique soon allowed "direct" visualization of both phage and plasmid DNAs and provided dramatic confirmation of the circular nature of plasmid DNAs.
R-factorsAnother important class of plasmids which were discovered in relation to their pathogenesis is the R-factor. In the early 1950s it was observed in Japan that multiple antibiotic resistance was developing in a single step in patients with enteric infections. In 1960 T. Akiba, T. Koyama, Y. Isshiki, S. Kimura, and T. Fukushima (Jap. med Wochschr. 1866:45) described these phenomena and in 1961 T. Watanabe and T. Fukasawa reported that this multiple drug resistance was being transferred by a plasmid (? an episome) which they called a resistance transfer factor (RTF or R-factor) (J. Bact. 81:679).
Organelle GeneticsAs difficult as it was to elaborate an understanding of the genetic and physical basis for non- chromosomal heredity in bacteria, the parallel history of eucaryotic cells is even more tortuous. While many observations in eucaryotes (mainly yeast and protozoans, single-cell organisms more amenable to genetic analysis than many multicellular organisms) suggested that non-chromosomal, especially cytoplasmic heredity exists, the acceptance of this conclusion and evidence for its physical basis were long in coming.Mitochondrial genetics, pioneered in the 1950s by B. Ephrussi and P. Slonimski in their studies of the respiratory-deficient petite mutants of yeast (such mutants grow slowly, depending as they do on glycolysis, and give small colonies, hence the designation, petite), became well-established only in the late 1960s when many additional mutants were identified that were associated with the mitochondria. Also, as early as 1954 some mutations in Chlamydomonas were found by R. Sager (Proc. Natl. Acad. Sci. USA 40:356-363) to behave in non-Mendelian fashion and were attributed to mutations in the chloroplasts. Finding the physical basis of organelle heredity (that is, the DNA in these structures) proved difficult as well. Cytochemical, electron microscopical, and biochemical evidence was offered, but until the techniques for study of DNA based on sequence comparisons (first nucleic acid hybridization and more recently direct nucleotide sequence analysis), the existence of cytoplasmic genes in eucaryotes was controversial. As R. Sager noted in 1972 (Cytoplasmic Genes and Organelles, Academic Press, p. 2): "The pendulum of opinion had swung from one extreme -- cytoplasmic genes do not exist because we do not see cytoplasmic chromosomes to the other extreme -- cytoplasmic DNA's exist, and therefore there must be cytoplasmic genes."
The "Modern Period" of Plasmid ResearchBy the end of the 1960s, then, both the genetic and physical understanding of plasmids and cytoplasmic heredity had reached a level of detail which was to allow the subsequent massive exploitation of these genetic elements as tools to study key cellular processes such as DNA replication as well as to manipulate and engineer the genetic contents of cells at will by means of the newly devised methods of in vitro recombinant DNA chemistry.
Main Page / History / Bibliography / Search |