Vanderbilt Institute of Chemical Biology



Discovery at the VICB







Dissecting the Structure of Biofilms Formed by Uropathogenic Escherichia coli


By: Carol A. Rouzer, VICB Communications
Published:  March 20, 2015



Imaging mass spectrometry reveals the regional proteome of uropathogenic E. coli biofilms, providing insight into the mechanism of biofilm formation.


In nature, bacteria frequently grow in multicellular structures called biofilms that can coat the internal or external surface of another living organism as well as abiotic surfaces. In vertebrate hosts, biofilms formed within the gastrointestinal tract from mixtures of nonpathogenic bacteria provide a protective barrier that prevents colonization by pathogens. However, biofilm formation by pathogenic bacteria serves as a mechanism by which they evade host defense mechanisms and resist antibiotic therapy. Regardless of the number of bacterial species present, cells in different regions of a biofilm tend to specialize, taking on specific metabolic functions that are appropriate for their immediate environment. This specialization is, at least in part, the result of gene expression changes that occur in response to local conditions, such as the presence or absence of nutrients and/or oxygen. Knowledge of how a biofilm is formed, how bacteria acquire specialized functions within the biofilm, and how these functions change in response to environmental stressors is key to understanding how pathogenic bacteria exploit biofilm formation to survive in the host environment. This led Vanderbilt Institute of Chemical Biology investigators Maria Hadjifrangiskou and Richard Caprioli to investigate the spatial distribution of bacterial proteins within biofilms formed by uropathogenic Escherichia coli (UPEC) (Figure 1) [K. A. Floyd, et al. (2015) PLoS Pathogens, 11, e1004697].


Figure 1. Cluster of E. coli bacteria at 1,000-fold magnification. Reproduced from Wikimedia Commons. Produced by the U.S. Department of Agriculture. Public domain.


UPEC are the primary cause of urinary tract infections. These bacteria are very proficient at forming biofilms on urinary catheters, on the surface of urinary bladder epithelium, and within bladder epithelial cells. Their ability to form biofilms contributes significantly to their pathogenicity, including their ability to induce life-threatening urosepsis. Key to understanding biofilm organization and function is a knowledge of how bacteria in different regions of the biofilm vary in terms of protein expression and metabolic adaptation. Prior attempts to address these questions have mostly depended on studies of individual proteins that can be tagged with a fluorescent marker or an antibody. This approach is biased, limited in scope, and requires selection of the protein based on previously existing information. In their quest for a broader, unbiased approach to the assessment of the spatial organization of the bacterial proteomes within a biofilm, the Hadjifrangiskou laboratory turned to matrix-assisted laser desoprtion/ionization time-of-flight imaging mass spectrometry (MALDI-TOF IMS), a specialty of the Caprioli laboratory. For MALDI-TOF IMS, the investigator first coats the sample with a matrix that absorbs ultraviolet light and then scans the sample with a laser. The laser light ionizes and volatilizes analytes of interest, which are then captured by a TOF mass analyzer for identification by mass-to-charge ratio. Individual ions are associated with a specific location in the sample on the basis of the position of the laser at their time of detection.


Having selected MALDI-TOF IMS as their analytical method, the Hadjifrangiskou lab created UPEC biofilms by placing a glass slide into a conical tube containing a suspension of bacteria in liquid culture medium. Under these conditions, the bacteria form a biofilm on the surface of the slide at the air-liquid interface (Figure 2). The investigators thoroughly washed the slides to remove small molecules and salts that could interfere with the analysis and then subjected the biofilms to MALDI-TOF IMS, focusing on small molecular weight proteins in the range of 2,000 to 50,000 Daltons (Figure 3).

Figure 2.  Preparation of a UPEC biofilm. A glass slide is placed into a 50 mL conical tube containing a culture of UPEC bacteria in liquid medium. The bacteria establish a biofilm on the slide at the air-liquid interface. Scanning electron micrographs show the structure of the biofilm at the interface in comparison to bacteria from the medium or air that have attached to the slide. Figure reproduced under the Creative Commons Attribution License (CC BY 3.0) from K. A. Floyd, et al. (2015) PLoS Pathogens, 11, e1004697.

Figure 3.  MALDI-TOF IMS analysis of a UPEC biofilm prepared as in Figure 2. The slide is coated with a matrix that absorbs ultraviolet light. A laser is passed over the slide systematically, targeting discrete regions. Ions released by the laser are captured by the mass spectrometer. The mass-to-charge ratios of the ions are reported and correlated with the position of the laser on the biofilm. Figure reproduced under the Creative Commons Attribution License (CC BY 3.0) from K. A. Floyd, et al. (2015) PLoS Pathogens, 11, e1004697.



The analysis identified 60 protein ions that were observed reproducibly in five biological replicates. These proteins exhibited four patterns of distribution: diffuse across the entire biofilm, concentrated at the air-exposed surface, concentrated at the liquid-exposed surface, and concentrated at the air-liquid interface. Thus, MALDI-TOF IMS clearly distinguished proteins on the basis of their location within the biofilm.


Enzymatic digestion of the bacteria in the biofilm followed by tandem mass spectrometry provided positive identification of six of the proteins detected by MALDI-TOF IMS. These proteins included two histone-like global transcription factors, HU-α and HU-β, found predominantly in the air-exposed regions, an acid stress-response chaperone, HdeB, localized to the air-liquid interface, and an uncharacterized protein, YahO, which was diffusely distributed across the biofilm. Of particular interest were two proteins that play a role in biofilm formation by serving as subunits of adhesive organelles in the bacteria. These included CsgA, a subunit of the curli amyloid fibers found at the air-liquid interface, and FimA, a subunit of type1 pili found primarily in the air-exposed region.


Prior studies had shown that the physical structure of UPEC biofilms is dependent upon an extracellular matrix comprising curli fibers and cellulose. The investigators confirmed that they had correctly identified CsgA in their biofilms by immunofluorescence, using super-resolution illumination microscopy (Figure 4). Treatment of cells with FN075, a small molecule inhibitor of curli biogenesis resulted in a reduction in total number of bacteria in the biofilm accompanied by an increase in CsgA as detected by MALDI-TOF IMS. The investigators concluded that FN075 prevented the aggregation of CsgA subunits into mature curli fibers, resulting in poor biofilm formation, and an accumulation of monomeric subunits. These results also demonstrated that MALDI-TOF IMS can capture biofilm responses to a perturbing stressor.




Figure 4. Expression of curli fibers as detected by immunofluorescence using super-resolution illumination microscopy. Curli fibers (red) are rare in the air-exposed (A) and liquid-exposed (C) regions of the biofilm but highly expressed at the air-liquid interface (B). Nucleic acid is stained blue. (Figure reproduced under the Creative Commons Attribution License (CC BY 3.0) from K. A. Floyd, et al. (2015) PLoS Pathogens, 11, e1004697.



While curli fibers are absolutely required for biofilm formation under the conditions tested, type 1 pili serve an accessory role, perhaps by increasing the tensile strength of the biofilm. Previous studies have shown that the absence of type 1 pili leads to formation of biofilms with holes at the air-exposed surface. This is consistent with the MALDI-TOF IMS results that localized FimA to the air-exposed region of the biofilm, and led the investigators to search for the environmental triggers that lead to FimA expression in the air-exposed region.


Type 1 pili (fim) gene expression is phase-variable. This means that the promoter that controls expression of the fim genes can exist in two states: a transcription-competent state (ON), driving expression, or a transcription-incompetent state (OFF) that cannot support transcription. This phase variation is brought about by the action of three recombinase enzymes—FimB, FimE, and FimX—which regulate the position of the element. Depending on the growth conditions of the bacteria, the fim operon is either ON or OFF. The Hadjifrangiskou lab started with UPEC bacteria in the fimOFF state, and cultured them under six different culture conditions, varying medium (YESCA versus Lysogeny Broth), temperature (25oC versus 37oC) and oxygen (semi-aerobic versus anaerobic). They discovered that, when cultured in the absence of oxygen, the bacteria remained fimOFF, while the presence of oxygen induced the bacteria to switch to fimON. Addition of nitrate to serve as an alternative electron acceptor to oxygen did not result in a switch to fimON in these bacteria. In contrast, if the bacteria started in the fimON state, this state was maintained by culture in the presence of oxygen or nitrate, though culture under anaerobic/fermentative conditions induced the bacteria to switch to fimOFF (Figure 4). These results suggested that anaerobic/fermentative conditions induce a switch to fimOFF, while the presence of oxygen has the opposite effect. Nitrate can help maintain the fimON condition in the absence of oxygen, but is not capable of switching bacteria from fimOFF to fimON. These conclusions were complicated somewhat by the observation that bacteria that had been genetically engineered to remain permanently in the fimON state exhibited decreased formation of type 1 pili under anaerobic/fermentative conditions (Figure 5). Thus, it appears that in addition to regulation of the phase of the fimA promoter, oxygen also regulates type 1 pili expression by a mechanism that is independent of this promoter. Further studies using an assay that detects the presence of type 1 pili by the ability of the bacteria to agglutinate red blood cells gave consistent results. This assay also showed that cells not producing type 1 pili tend to produce S pili, another adhesive pilus known to be expressed inversely with type 1 pili.



Figure 5. Expression of type 1 pili in wild-type UPEC (WT UTI89) and UPEC genetically engineered to maintain a permanent fimON state (UTI89_ON). Type 1 pili observed as fine fibrils surrounding the cells are present under semi-aerobic conditions, but nearly absent under anaerobic/fermentative conditions. Anaerobic conditions with the inclusion of nitrate results in reduced, but not eliminated, expression of type 1 pili. Figure reproduced under the Creative Commons Attribution License (CC BY 3.0) from K. A. Floyd, et al. (2015) PLoS Pathogens, 11, e1004697.



Together, the results demonstrate the power of MALDI-TOF IMS to carry out an unbiased survey of the biofilm proteome, while also providing spatial information within the three-dimensional structure. Further refinement of this approach should allow the identification and localization of a broader sample of proteins over a wider range of molecular weights. In addition to serving as proof-of-concept for this approach, the MALDI-TOF IMS results inspired new studies into fimA gene regulation, demonstrating how oxygen induces FimA biosynthesis in air-exposed locations of the biofilm. These results are consistent with the expectation that bacteria will only produce these energy-requiring structures when plentiful oxygen, or in some cases an alternative electron acceptor, allows efficient energy generation by respiration.


The “division of labor” among bacteria in biofilms results in the adoption of a metabolically inactive state by some cells. These cells are often highly resistant to antibiotics, serving as “persister cells” that survive and grow after therapy is discontinued. Clearly, an understanding of the mechanism of biofilm formation is important if such resistance mechanisms are to be overcome. We look forward to the next interesting discovery to arise from the MALDI-TOF IMS analysis of UPEC biofilms.





The Vanderbilt Institute of Chemical Biology, 896 Preston Building, Nashville, TN 37232-6304, phone 866.303 VICB (8422), fax 615 936 3884
Vanderbilt University is committed to principles of equal opportunity and affirmative action. Copyright © 2014 by Vanderbilt University Medical Center