Research

Biofilms and Health

A bacterium is able to switch between living as a free-floating planktonic cell and living as a member of a multicellular, matrix-encased community called a biofilm. Generally, planktonic cells are much easier to control because disinfectants, antibiotics, and our immune system are more effective at killing and/or neutralizing them. So when harmful bacteria (such as pathogens) form biofilms in environments where we don’t want them, it can lead to problems.

Contextualizing Biofilm Formation

Many bacterial pathogens, namely opportunistic pathogens (those that can cause disease in susceptible people but typically occupy a non-host environmental niche) can form multiple different types of biofilms. For example, environmental cues that a bacterial cell would encounter in the soil (such as low carbon) might trigger it to form a biofilm that is made mostly out of polysaccharides, while environmental cues that a bacterial cell would encounter in the human body (such as a temperature of 37°C) might trigger it to form a biofilm made mostly out of protein. These two types of biofilms, although made by the same species of bacteria, would likely demonstrate differences in susceptibilities to stressors (like antibiotics) and might require very different strategies in order to trigger dispersal. Therefore, it is imperative that we make sure that the types of biofilms we are growing in a lab match the types of biofilms in our environment of interest. To achieve this goal our lab uses a three-pronged approach when we study biofilms:

In situ Characterization

The first step is to obtain as much detail as possible about a bacterial community in the environment of interest. This can be difficult — biofilms are 3D structures and characterizing them in a native tissue presents many challenges:

1.) Sites of bacterial colonization/infection are often unstable and difficult to process without disrupting the structure

2.) Tissues are opaque, making it difficult to image deeply into them

3.) Bacteria are small and fluorescent labels are often obscured by tissue autofluorescence

To overcome these challenges we utilize MiPACT-HCR, a tissue clearing and microbial imaging technique.

MiPACT (Microbial Identification after Passive CLARITY Technique) combines two technologies, a tissue-clearing technique called PACT (read more here) and a fluorescent label amplification technique called HCR (read more here). The first step in MiPACT-HCR is to fix the tissue with formaldehyde and embed it in a gel-like bis-acrylamide matrix to provide stability and structural support (This solves imaging problem #1). Then the sample is heated in detergent for varying periods of time — depending on the tissue size and composition — to remove lipids, which are a major culprit for light diffraction (This, along with RIMS, makes the tissue transparent and solves imaging problem #2). After the clearing step, HCR is used to label bacterial RNA, with rRNA probes providing species-specificity and mRNA probes providing a readout of specific cellular activities. Because HCR amplifies the fluorescent signal from any given RNA target many times over, this technique produces a very bright signal and solves imaging problem #3. We also use various fluorescent dyes to stain components of the tissue, providing a structural context while imaging. Finally, the tissue is placed in RIMS (refractive index matching solution) and imaged via confocal microscopy. When we perform this technique on sputum samples from Cystic Fibrosis (CF) patients, the end result can look like the following video —

— Magenta corresponds to an rRNA HCR tag that labels all bacteria, and green is an rRNA HCR tag specific to Staphylococcus aureus. In this particular sample, the magenta objects that don’t also stain green are most likely Achromobacter xylosoxidans. The blue color that makes up most of the area — turned off at first so as not to obscure the bacteria — is DAPI, a fluorescent dye that stains DNA. In this case DAPI is mostly labeling immune cells called neutrophils that are attempting to fight off the bacterial invaders.

MiPACT-HCR is powerful because it provides us with quantifiable 3D spatial information. We can ask things like: What size aggregates do nontuberculous mycobacteria (NTM) tend to form during a lung infection? or What biofilm-related genes are expressed in an S. aureus community within a skin abscess? or What immune cells are most often associated with intermediate-sized biofilms during any given bacterial infection?

We are also very excited to expand MiPACT-HCR to other systems involving host microbe interactions. We are working with collaborators to survey bacterial aggregation during intestinal infections in mice and determine patterns of bacterial colonization on coral, among other projects. If you have an interesting system that might benefit from analysis with MiPACT-HCR, let us know!

Basic in vitro Studies and NTM

Once we know some information about in vivo aggregation patterns from MiPACT-HCR (such as average aggregate size and what biofilm matrix genes are expressed during infection), we can then start to dissect those specific pathways in vitro in order to acquire mechanistic insight into pathogenesis-relevant biofilm formation and dispersal processes. We need this level of detailed understanding if we hope to intelligently develop therapies to control biofilm formation in any given context. These type of in vitro studies require a tractable laboratory model with which to study biofilms.

NTM are emerging pathogens that persist in household water systems and in animal models of disease at least in part because of their ability to aggregate into biofilms. CF patients are particularly susceptible to NTM lung infections, and the current treatment regime for NTM is long and involves antibiotic cocktails that have undesirable side effects. Our main research interest, therefore, is understanding how nontuberculous mycobacteria (NTM) aggregation affects pathogenesis during CF lung infections. In parallel to deploying MiPACT-HCR to study in situ aggregation patterns of NTM, we are working to understand how certain environmental cues trigger NTM to switch between planktonic and biofilm states.

Informed Model Systems

This brings us to the last box in our three-pronged approach, informed model systems. These help us bridge in situ data and in vitro data and test their compatibility in an environmentally-relevant context. An informed model system will recapitulate at least one facet of the infection environment that our basic in vitro system does not. For example, when surveying sputum samples with MiPACT-HCR, it becomes evident that most bacterial communities are embedded in thick mucous and/or DNA (see the MiPACT video above). Bacterial communities embedded in a 3D matrix likely behave differently than aggregates shaking in a liquid culture, so we can set up an experimental system to model this parameter. For this purpose we utilize ABBA (agar block biofilm assay — see examples here and here), which involves embedding bacteria in soft agar and determining how the resultant 3D communities grow and react to stressors.

Here’s a video of M. smegmatis in an ABBA, the embedded bacterial aggregates are stained blue and the space between the aggregates is the gel-like agar (fluorescently invisible).

We suspect that our ABBA aggregates do a better job of mimicking the microbial communities we see in sputum with MiPACT-HCR. Moving forward, we can test whether the gene-expression patterns in ABBA match those we see in situ and in vitro, and how ABBA communities react to stress.

That’s the lab in a nutshell — thanks for making it to the end! If you’d like to know more, see our publications and feel free to contact us.