From a modern microbiological viewpoint the motivation behind the experiments that led to the discovery of bacteriophage is hard to understand. What was the purpose of filtering a bacterial culture to remove the bacteria, then remixing the filtrate with a fresh bacterial culture? We all know the outcome of this classic experiment, but why was it done in the first place? To fathom the intent of the original investigator, Félix d'Herelle, one has to put aside twenty-first century ideas and recover the context of microbiology and disease in the early twentieth century.
D'Herelle was an enigmatic personality, born in 1873 just outside of Montreal, raised by his mother in Holland and France, and essentially self-educated after the age of 17. By his own admission he was a restless world traveler who had a passionate interest in the new field of microbiology. In later life he claimed to have modeled his career on that of Louis Pasteur, first studying fermentations, then diseases of insects, and finally human infections.
His path to the discovery of bacteriophage, as reconstructed from later memories as well as contemporary records, was tortuous. While he was working on fermentation studies in Mexico, he observed a massive and devastating locust plague. Curiously, however, some of the locusts were afflicted with bacterial enteritis of some sort, similar to human dysentery. D'Herelle reasoned that if he isolated the enteric pathogen, he might be able to deliberately spread this lethal infection and thus suppress the locust infestation. This approach worked well, and he is remembered in the field of entomology as a key pioneer of biological pest control. The pathogen he isolated was a rather fragile organism which has now been replaced by the sturdy spore-former, Bacillus thuringiensis. Occasionally, however, his cultures of the enteric pathogen were subject to a mysterious and apparently spontaneous lysis. Based on some complex and ultimately incorrect ideas about “interacting factors” that controlled pathogenicity, based on research on the hog cholera bacillus, d'Herelle speculated that the variability in the potency of his cultures was due to a second, unknown “factor.” Later, in 1916, while investigating an outbreak of dysentery among French soldiers during WWI, this hypothesis led him to remove the dysentery bacteria from the culture by filtration through a porous filter (a terra cotta Chamberland filter), and then test the bacteria-free filtrate for biological activities. One such test was to reconstitute the original sample by mixing portions of the filtrate with a pure culture of the original bacteria. It was in such a test that the first plaques were observed. Because the filter was already a tool for classifying infectious agents as “filterable” (i.e., filter-passing) or “non-filterable” viruses (a term used more broadly at the time to mean only a disease-causing agent; it was not until the 1930s that “filterable virus” because synonymous with the modern concept of a virus), d'Herelle interpreted his plaque-forming agent as an “ultravirus” that infected the bacteria. A prescient interpretation, indeed.
In a creative leap, d'Herelle noted that the bacteriophage titers in the stool samples of his dysentery patients seemed to peak just as their recovery started. From this observation, he proposed that recovery from infectious disease was mediated by bacteriophage. This essentially ecological view of the balance between two microbes, bacteria and bacteriophage, suggested to him that phage, in his words, are the “agents of natural immunity.” He even went so far as to challenge the developing orthodoxy which established cellular and humoural immunity as the key mechanisms of disease resistance and recovery.
While the viral interpretation seems so natural to the modern reader, d'Herelle faced substantial opposition to this idea. Some of this opposition came in response to his personal style and some from real confusion over conflicting experimental results. His direct challenge to the work of Jules Bordet, who had just won the Nobel Prize in 1919 for his work on immunity turned Bordet and his school into staunch opponents. Bordet tried to repeat some of d'Herelle's work, but obtained conflicting results. Much later it was realized that Bordet had used a lysogenic strain of bacteria which led him to conclude that phage was produced by the cell as a physiologic response, similar to other cell products such as enzymes and excreted carbohydrates. The concept of phage as a serially transmitted infection was all but overwhelmed by opposition from Bordet with his Nobel Prize and his own Pasteur Institute in Brussels, all arrayed against the irascible, unpaid, self-educated, volunteer researcher at the Pasteur Institute in Paris.
Adding to the complexity of the evaluation of d'Herelle's work was the competing claim for priority advanced on behalf of Anthony Twort, a British microbiologist who published his observations of “transmissible lysis” of bacteria two years before d'Herelle's initial report. Bordet and his colleague seized upon Twort's work to attack d'Herelle, not on his concepts of the viral nature of phage, but on his claim of priority. Like d'Herelle, Twort was studying something else (growth of vaccinia virus on cell-free agar media) and he noted that the contaminating micrococci in his vaccinia preparations sometimes exhibited a patchy dissolution of the colonies. This effect could be transmitted to fresh colonies by transfer with a sterile needle. Twort's paper, “An investigation on the nature of ultra-microscopic viruses” (Lancet 1915; 1241–3), was all but ignored and he was unable to follow up on this observation. After only one other paper describing some of the properties, such as heat sensitivity, of this lysis factor, Twort moved on to other work. It was Bordet, and especially his protégé, Andre Gratia, who championed Twort in their battle against d'Herelle and thus developed the “Twort-d'Herelle Controversy.”
While d'Herelle was engaged in disputes with Gratia and Bordet over issues of both priority and the nature of the lytic principle, he also built up his observations on the relationship of phage titers to recovery from infectious diseases, especially enteric diseases, to develop approaches to exploit bacteriophages as therapeutic agents. In the era before effective antimicrobial drugs, phage offered an attractive new way to treat infections. There was no doubt that phages killed bacteria, no matter what their biological nature or who discovered them. Soon the therapeutic potential of bacteriophage was a major topic. Even the American author Sinclair Lewis used phage to control a fictional plague in his Nobel Prize novel, Arrowsmith. Major pharmaceutical houses sold phage preparations, clinicians and basic scientists studied phage as remedies for everything from the common cold to flatulence. Needless to say, because the basic biology of phage was poorly understood, many conflicting results were obtained. In 1934 the AMA's Council on Pharmacy and Chemistry undertook a review of the phage literature to determine the validity of the claims made for phage therapies. The confusion in the literature, however, confounded all attempts at consensus, and this report simply recommended additional research. Soon, however, the antibiotic era set in and phage therapy with its inherent complexities was eclipsed by the effectiveness and simplicity of penicillin, streptomycin and the new chemotherapies of sulfa and isoniazid. While in the US, the UK and most of Western Europe physicians gave up on phage therapy, in the USSR, Poland and to some extent, India, the use and study of therapeutic phage continued. It is from these sources that much of the current resurgence of interest and knowledge of phage therapy emanates.
Bacteriophage research took a new and important direction based on the debates on the biological nature of these “ultraviruses.” D'Herelle's view that phage were particulate microbes, analogous to other filterable viruses slowly gained a following. The phenomenon of plaque formation was central to this point of view. D'Herelle recruited no less an “expert” than Albert Einstein who agreed that the phenomenon of serial dilution of plaque-forming agents could only be explained by a particulate (“quantized”?) agent. At the same time, the observation of step-wise bursts of phage production that accompanied bacterial lysis suggested to d'Herelle that the phage were undergoing intracellular reproduction. In the late 1930s the process of reproduction, especially by simple organisms, was of major interest. The faithful reproduction of both organisms and genes was seen as essential to understanding life processes, and phage seemed ideal as a representative of a simple organism, which, as Muller had suggested in 1922 might even be considered “naked genes.” Emory Ellis at Caltech decided that phage would be a good model for the study of viral reproduction in his work on carcinogenesis (actual animal virus experiments were deemed too expensive). His work on phage reproduction intrigued Max Delbrück, who was working on Drosophila gene reproduction, and soon Delbrück abandoned fruit flies, joined Ellis, and together they carried out a series of classic experiments that soon convinced others that understanding phage reproduction was the key to understanding gene duplication. Soon a small cadre of phage workers coalesced around Delbrück, Salvador Luria, and Alfred Hershey, two other phage workers who saw the promise of this experimental system, and an active “American Phage Group” was formed. At the same time, phage research originally started by d'Herelle at the Pasteur Institute in Paris was carried on by the Wollmans, parents and son, and later Andre Lwoff and Francois Jacob, and a European tradition of phage research developed somewhat in parallel to that in America.
Bacteriophage research had its start with a short note in Comptes rendus in 1917, but the brevity of d'Herelle's first account belies its rich content. Not only did he recognize the nature of the agent he discovered, he appreciated its ecological role in the microbial world, and saw both therapeutic and theoretical roles for these amazing little microbes.
Commentary to: d'Herelle F. Sur un microbe invisible antagoniste des bacilles dysentériques. CR Acad Sci Paris. 1917;165:373–375.