Microbial Genomics: Standing on the Shoulders of Giants
Professor Stanley Falkow, part II
Described as the ‘father of molecular microbial pathogenesis’ and 2004 Marjory Stephenson Prize winner. Professor Falkow discusses his own research, which by his own account, is divided into three sections: firstly, antibiotic resistance/plasmid biology; secondly, plasmids that contributed directly with bacterial pathogenicity, and finally, his work away from virulence plasmids to investigate 'what is a pathogen?’.
This is the second installment from ‘Standing on the Shoulders of Giants’ featured giant, Professor Stanley Falkow, and focuses on the discovery and characterisation of R-factors in the 1960s–70s.
Read The Early History of Plasmids: Part I
The Early History of Plasmids: Part II – The R-Factors
The R-factors were originally isolated in Japan. Clinicians were alarmed to discover that Shigella flexneri isolated from patients ill with dysentery were multiply resistant to antibiotics. Prior to the late 1950s, isolates of the dysentery Bacillus were uniformly susceptible to most antibiotics available for treatment. Suddenly the bacteria were simulataneously resistant to streptomycin, tetracycline, sulfonamide and chloramphenicol. Some of the E. coli isolated from these patients were also multiply resistant. The Japanese investigators also noted that some patients at the beginning of treatment had only antibiotic-sensitive bacteria, but that following treatment they excreted multiply resistant Shigella and E. coli as well. They knew there was something unusual happening in the intestinal tracts of their patients. In the 1950s we didn’t have that many antibiotics to employ against infection so the discovery that bacteria were becoming multiply resistant was frightening.
Trivial explanations to account for this multiple resistance, like a single mutation that made the bacteria impermeable to antibiotics, were quickly eliminated. Mutants might be resistant to a single antibiotic, but based on conventional genetic theory, not four simultaneously. Also, the probability that this might occur at the same time in both Shigella and E. coli in the same patient within days seemed unlikely.
Because both bacterial species exhibited this curious (and potentially deadly) genetic change in the gastrointestinal tract of the same patient, the Japanese clinical scientists, aware of the experiments of Adelberg and Jacob, wondered if the multiple resistance might be on a genetic element analogous to F-lac and was transferred from bacteria to bacteria in the patient’s intestinal tract under the selective pressure of antibiotic usage. This perceptive insight proved to be correct. R-factors were discovered and announced to the world in scientific papers (in Japanese) in 1958 and were followed by several other papers in English appearing in the early 1960s. R-fators were not unique to Japan and were quickly confirmed in Salmonella isolated by Naomi Datta in 1962 and Gerhardt Lebek in Germany in 1963. There was a veritable flood of publications confirming that transferable antibiotic resistance was common in clinical isolates from animals and humans around the world. Transferable resistance was documented in microbes isolated from the environment; in humans that had never been hospitalized; and, importantly, in populations of humans living in Borneo and Australia who had no evidence of exposure to contemporary medicine, and certainly not to prescription antibiotics. Perhaps there was a clue in the finding that resistance to mercury and silver were also carried on extrachromosomal elements.
The implications for human and veterinary medicine were profound. However, most early investigators had to concede, as did Naomi Datta who wrote: “The present findings emphasize the need for bacteriological diagnosis and sensitivity testing: resistance is not at present a serious clinical problem because there remains a choice of suitable therapeutic agents, active against the most common resistant strains.” She warned: “We should, however, watch out for further spread of resistance and, particularly, for further accumulation of multiple, linked resistance genes, all of which will be favored by the use of only one drug, such as tetracycline.” She was quite right.
I organized the first international meeting on R-factors in 1967 which was sponsored by the U.S. Food and Drug Agency. This meeting emphasized the medical and veterinary importance of the promiscous genetic elements. As well, these discussions brought sharply into view the differences between the European attitude towards antibiotic supplements in animal feed as compared with the attitude of the United States. There was international concern; however I recall that about this time there were pronouncements in the United States that infectious diseases were no longer an important factor in societal well-being.
Basic scientists were faced with a more compelling fundamental question that overshadowed all the rest. What is the origin of these R-factors and how closely were their properties compared to the well studied temperate phages and the paradigm episome, F? Between, 1963 and 1969 the focus in the plasmid/episome field was on R-factors and their relationship to the prototypic F-factor. There was also extraordinary interest in the temperate bacteriophage, lambda, fueled by Allan Campbell’s brilliant model explaining lambda insertion into the chromosome. Much of the initial debate, however, revolved around whether R-factors were like F and lambda, and readily inserted into the bacterial chromosome and, if so, whether it occurred at a fixed location.
It is quite impossible to sieve through the large amount of data and their interpretation in those early days. I will be forgiven, I hope, if I select out what seems to me to be the major findings as they emerged over the next decade. I was an active participant in this research arena and therefore while I try to be objective, I am undoubtedly guilty of bias. Below is a modified version of a Table published in 1963 by Wantanabe in his landmark review of the R-factors, arguably the first major review of the topic in the English language.
From: Watanabe T. Infective heredity of multiple drug resistance in bacteria. Bacteriol Rev. 1963;27:87–115. PMCID: PMC441171.
This table attempts to show the comparative properties of the mobile genetic elements known at this time. Watanabe clearly understood that R-factors possessed a transfer element totally anaolgous to the classical F-factor but he reasoned that the resistant genes had a chromosomal origin. He suggested chromsomal interaction must have occurred and, thus, he deemed R-factors to be likely episomes because he emphasized that episomes have two properties: automonous replication and integration. The discovery of R-factors brought about a reinterest in the findings of Pierre Fredericq who first described Colicin I, and his work was extended by Royston (Roy) Clowes and his co-workers as a genetic element capable of conjugation and the transfer of chromosomal material but decidedly different than the F-factor. At the time, classification was solely on the basis of phenotypic chracterteristics. F transferred chromosomal markers, ColI synthesized an antibiotic – like activity on certain other closely related bacterial species and R-factors were associated with the transfer of one or more antibiotic resistance traits.
There was a major breakthrough in 1961 when Norton Zinder and Timothy Loeb, a graduate student in his laboratory, discovered seven new bacterial viruses specific for E. coli F-containing bacteria, and again in 1963 when Charlie Brinton reported the presence of a F-specific pilus that served as the receptor for the same phage. Although the ‘sex’ pilus, as it was called, was not thought to be the actual conduit for DNA transfer, it gave for the first time a structure and an antigen which could characterize F other that the phenotype of ‘chromosome’ transfer. The Meynells (Guy and Elinor) characterized the pilus associated with the ColI transfer factor and isolated a specific phage that recognized it. They began to characterize transferable factors into two groups, f+ and fi-, largely based on susceptibility to one or the other sex pilus-specific phages.
A formal scheme of classification based on incompatibility was developed in the early 1970s, mainly by Naomi Datta and Bob Hedges; incompatibility was defined by inability of two plasmids to be propagated stably in the same cell line. Plasmids incompatible with each other were assigned to the same incompatibility group. Conjugative plasmids were classified according to their incompatibility, a property directly related to replication. My laboratory worked closely with Datta and Hedges to define the molecular basis for incompatibility. We showed that the plasmids of any particular compatibility class were found to be relatively homogeneous with respect to their general molecular size, G + C content, and common polynucleotide sequences. Plasmids of different compatibility classes showed less than I5% sequence similarity. What was surprising to me was that there was very little polynucleotide similarity between members of the different incomptibility groups – a finding that was repeatedly confirmed. There appeared to be many different ‘species’ of plasmids in nature. Yet, it was puzzling that these plasmids encoded for similar antibiotic resistance traits and often used identical enzymatic mechanisms to inactivate antimicrobial activity.
The early R-factor research scientists soon discovered that bacterial strains in nature could carry a considerable number of genetic elements simultaneously. It was also soon discovered that there was no common universal host making incompatibility testing not feasible anymore (of course afterwards, an alternative classification scheme was widely accepted that used similarity of plasmid-encoded replication region(s) for classification). Meanwhile, an ever increasing number of resistant determinants were being identified on mobile genetic elements. It seemed that an antibiotic resistant gene appeared on a plasmid soon after the antibiotic came into use in general medical practice.
As the studies on plasmids expanded during the 1960s there was a concurrence of genetics and an increasing number of molecular methods to characterize DNA brought to bare on the nature of extrachromosomal elements. I had extended my research to show that the Proteus miriabilis could be used as a host organism to easily biologically fractionate the extrachromosomal element from the chromosomal DNA, based on a striking difference in G+C composition. Thus there was a means to physcially fractionate a number of genetic elements and analyze its (usually radioactively-labeled) DNA. In this way, T.F. Roth and Donald (Don) Helinski first determined that the ColE1, could be found as covalently closed circular forms. Circular DNA became a central question that was addressed experimentally both from a genetic analysis of gene transfer during conjugation and the radioautographic study of the chromosome of Escherichia coli, and from biochemical and electronmicroscopic analyses of the bacteriophage phiX 174. The development of several techniques, especially dye-buoyant density centrifugation and alkaline sucrose centrifugation taking advantage of some of the special properties of circular DNA, eventually led to a deluge of information that all sorts and sizes of extrachromosomal elements were readily found in bacterial cells. What’s more, these experimental methods could be applied directly to the bacterial strains isolated from nature harboring these genetic elements. Ron Citarella and I had used biological fractionation in Proteus to measure homology of sex factor DNA and reported that the size of F represents about 1.9% of the E. coli genome and about 40% of the F DNA was found to be homologous to the chromosome of an of E. coli K-12. Now using the selective labeling methods developed by the Friefelders, F could be labeled and directly examined in E. coli; indeed it became possible to label virtually any extrachomosomal element and perform the physical characterization of their DNA including the direct examination of molecules under the electron microscope.
Acknowledgement: Professor Falkow composed these thoughts when preparing for Cold Spring Harbor Symposium Plasmids, History and Biology (held 21–24 September 2014) where he gave an oral presentation that can be found here. When told about the aspirations, aims and scope of Microbial Genomics he provided this article to help launch 'Standing on the Shoulders of Giants'. The journal would like to thank Professor Falkow for his contribution.
One of the most important observations on the nature of extrachromosomal elements was the finding of a class of elements that were found to be nonconjugative. That is, they were incapable of self-transmission but they could be ‘mobilized’ by a self-transmissible genetic element in the same cell. This was an extraordinary finding that sounded the death knell for the term ‘episome’. This will feature in the next section of Professor Falkow’s account of The Early History of Plasmids.