by George O'Toole
I have studied bacterial biofilms since I joined Roberto Kolter's lab at Harvard Medical School in 1995. I "discovered" biofilms as I was finishing my graduate work at Wisconsin and continued tracking the literature backwards after I joined Roberto's lab. The oldest reference I could find was from the 1920s – a Navy report on hull biofouling cited in another paper I had read. Several calls to the naval archives led me to a dead end – I never found that report. That exercise in tracking the literature revealed that it was great fun to read a paper, find references describing earlier work, then getting those papers and looking at the references, and so forth. It was like going back in time. I decided to do a bit more digging and put together this "Brief History of Biofilm Research." I am not a historian and I more than likely have some facts wrong. Even so, I hope this post might crowdsource a more accurate history and also serve as a useful introduction to new biofilm researchers and curious microbiologists everywhere.
Before They Were Biofilms. New knowledge is built on work done by generations of scientists that came before. Biofilm studies are no exception. Early work on biofilms goes back to the 1930s. These were simple but elegant experiments – a clean glass slide placed in a pond at the University of Minnesota by Henrici or off the coast of La Jolla, California by Zobell & Allen. These studies resulted in similar conclusions – microbes rapidly colonized naïve surfaces. While the word "biofilm" did not appear for ~35 years, in 1943 Zobell wrote "we have observed sessile bacteria which form a film on the walls of test tubes and flasks without perceptibly clouding dilute nutrient solutions."
Descriptions of biofilm formation as a multistep process were clear, even without the term biofilm. Zobell and Allen wrote "firm attachment [of bacteria] after coming into contact with a surface is not immediate, as it seems to take several minutes for bacteria to cement themselves to glass… But let the slide remain submerged for an hour or two… they will be found profusely, so firmly glued to the slide that running water will not detach them." This is what we refer today as "reversible" and "irreversible" attachment, two key early steps in biofilm formation. Microcolony formation came afterwards: "micro-colonies seldom appear on glass slides exposed to sea water for only a few hours..." Henrici made similar observations: "That the cells are actually growing upon the glass [slide] is indicated by their occurrence in microcolonies of steadily increasing size."
Early descriptions of the benefits for bacteria living on surfaces, particularly in oligotrophic environments, abound (for examples see Zobell, Heukelekian & Heller). Investigators concluded that bacteria are more active on surfaces when nutrients are scarce because the surface can concentrate the nutrients. There is a report from 1913, cited by Zobell, describing increased degradation of hydrocarbons by surface-attached bacteria. Thus, knowledge that bacteria attach to surfaces, and once there develop new traits, is nearly 100 years old.
Biofilms are Here to Stay: From the Environment to the Clinic. Only relatively recently were such microbial communities given their modern name and recognized as a clinically relevant problem. The word "biofilm" first appeared in a 1975 paper. The term caught on rather quickly, particularly in the field of water treatment and other areas of environmental engineering.
The evolution of the field of microbial biofilms can be followed through the lens of impactful papers by Bill Costerton and colleagues. While Costerton was not the first to use "biofilm" in print, he was arguably its most enthusiastic advocate. In the late 1970s Geesey and Costerton published two papers describing the role of biofilms in alpine streams, making the case for the existence of biofilms and their associated "slime" in natural environments (see here and here). Ten years after these reports of environmental biofilms the landmark review "Bacterial Biofilms in Nature and Disease" appeared. It reviewed the work to-date in environmental and industrial biofilms. Importantly, it presented the first diagrammatic model that still looks familiar today (Figure 1). The figure effectively synthesized the findings of the earlier biofilm researchers into a single, coherent, easy-to-digest scheme. This article also painted a picture of a generic definition of a biofilm: surface-associated bacteria, encased "slime" matrix (or glycocalyx), composed of one or more different types of bacteria, and expressing traits distinct from their planktonic counterparts. Interestingly, there is no frank (or simple) definition of a biofilm presented in the text! Nor were there any no such clear definitions offered in several other articles by Costerton and colleagues; the definition of a biofilm is encompassed over several pages of text or simply reduced to "sessile bacteria." Perhaps Costerton understood that the lack of a formal definition allowed for a nuanced understanding of what makes a "biofilm."
The 1987 review and others published a little later (here, here & here) united disparate observations from the environment, industry and clinic into a unified view, which argues that microbes in many settings exist in biofilms. Such communities were not an oddity of particular environments, but rather the rule for most microbes. By the late 1990s and early 2000s investigators recognized the associations between biofilms and chronic infections such as otitis media and the infections in the airways of persons with cystic fibrosis (pwCF). Of particular importance, the general tolerance of biofilms to environmental insults now included their increased ability to withstand high levels of antibiotics (here & here). Thus, the concept of biofilm evolved from a key mode of microbial growth in environmental and industrial settings, to clinical settings in the context of chronic and medical implant infections, and eventually encompassed a mode of bacterial growth in infections lacking any obvious abiotic surface on which to colonize. Thus, our modern view of a biofilm.
To Build a Biofilm. Three broad approaches are used to study biofilms. The first uses direct sampling and visualization of the biofilm, whether it is the surface of a submerged glass slide or rock in a stream or pond, a pipe, a colonized medical device, or sputum from the airway of a pwCF. These direct visualization approaches include using a variety of stains paired with bright-field microscopy, scanning or transmission electron microscopy (SEM/TEM), or fluorescent microscopy including confocal scanning laser microscopy (examples here, here & here). Such direct visualization approaches provide morphological data, and when combined with fluorescent in situ hybridization (FISH), can provide important information about the bacteria comprising such communities. The second general approach uses flow devices to build a biofilm in vitro. These include the Robbins device, a variety of other flow devices that can mimic an industrial process, drip reactors, CDC biofilm reactor, model flow cells, and increasingly, microfluidic devices. Such devices allow one to control the microbes forming the biofilm, tuning of environmental conditions (e.g. nutrients, pH, oxygen, temperature), and a means to readily image the community over time. The third approach, plastic dishes, allow one to investigate the earliest steps in biofilm formation, and importantly, to do so in a high throughput fashion conducive to performing genetic screens.
Each approach has relative strengths and weaknesses, can be used to address particular aspects of biofilm biology, and in some sense, can be used to operationally define the biofilm. For example, the SEM/TEM of an endocarditis infection may be defined as being biofilm-based because this nidus of infection is not cleared by conventional antibiotic therapy. As another example, a flow cell-grown community can earn the biofilm moniker because the community formed stains with a lectin targeting an extracellular polysaccharide. Alternatively, an operational definition can be used in the context of a microtiter plate assay, as "bacteria that are attached to a surface in sufficient numbers to be detected macroscopically", meaning that growth is robust, more than a monolayer of cells. Combining the microtiter assays with a flow cell allows one to define discrete developmental steps that a microbe undergoes as it transitions from single cells to a "mature biofilm," including changes in morphology, physiology or gene/protein expression (see a review here). Thus, all of these approaches are valid models to study biofilms, but none alone capture all aspects or nuances of how these communities form, their traits and functions, and/or their impact on the local environment.
Conclusions. Since I attended my first biofilm meeting in the late 1990s, a meeting dominated by speakers from the Center for Biofilm Engineering at Montana State University, and by labs from England, Denmark, and Germany, research in biofilms has exploded. It has been both exciting and fun to watch investigators from a broad range of fields – microbiology, engineering, ecology, physics, chemistry, and more – bring their expertise to study these complex and fascinating communities. As we move forward in our understanding of how biofilms form, function and disperse, don't forget to take a moment to look back to see where you came from!
George O'Toole is Professor in the Department of Microbiology & Immunology in the Geisel School of Medicine at Dartmouth in Hanover, NH.