Biofilms, sticky conglomerations of microorganisms and extracellular polymers, are among the Earth's most common life forms. One component for their survival is an ability to withstand external mechanical stress. Measurements indicate that biofilm elastic relaxation times are approximately the same (about 18 min) over a wide sample of biofilms though other material properties vary significantly. A possible survival significance of this time scale is that it is the shortest period over which a biofilm can mount a phenotypic response to transient mechanical stress.

Biofilms of various Pseudomonas aeruginosa strains were grown in glass flow cells under laminar and turbulent flows. By relating the physical deformation of biofilms to variations in fluid shear, we found that the biofilms were viscoelastic fluids which behaved like elastic solids over periods of a few seconds but like linear viscous fluids over longer times. These data can be explained using concepts of associated polymeric systems, suggesting that the extracellular polymeric slime matrix determines the cohesive strength. Biofilms grown under high shear tended to form filamentous streamers while those grown under low shear formed an isotropic pattern of mound-shaped microcolonies. In some cases, sustained creep and necking in response to elevated shear resulted in a time-dependent fracture failure of the "tail" of the streamer from the attached upstream "head." In addition to structural differences, our data suggest that biofilms grown under higher shear were more strongly attached and were cohesively stronger than those grown under lower shears.

A mathematical model describing the constitutive properties of biofilms is required for predicting biofilm deformation, failure, and detachment in response to mechanical forces. Laboratory observations indicate that biofilms are viscoelastic materials. Likewise, current knowledge of biofilm internal structure suggests modeling biofilms as associated polymer viscoelastic systems. Supporting experimental results and a system of viscoelastic fluid equations with a linear Jeffreys viscoelastic stress-strain law are presented here. This system of equations is based on elements of associated polymer physics and is also consistent with presented and previous experimental results. A number of predictions can be made. One particularly interesting result is the prediction of an elastic relaxation time on the order of a few minutes-biofilm disturbances on shorter time scales produce an elastic response, biofilm disturbances on longer time scales result in viscous flow, i.e., nonreversible biofilm deformation. Although not previously recognized, evidence of this phenomenon is in fact present in recent experimental results.

Staphylococcus aureus is a leading cause of catheter-related bloodstream infections and endocarditis. Both involve (i) biofilm formation, (ii) exposure to fluid shear, and (iii) high rates of dissemination. We found that viscoelasticity allowed S. aureus biofilms to resist detachment due to increased fluid shear by deformation, while remaining attached to a surface. Further, we report that S. aureus microcolonies moved downstream by rolling along the lumen walls of a glass flow cell, driven by the flow of the overlying fluid. The rolling appeared to be controlled by viscoelastic tethers. This tethered rolling may be important for the surface colonization of medical devices by nonmotile bacteria.Biofilm infections are increasingly associated with a variety of medical prostheses (e.g., vascular and orthopedic implants) and delivery devices (e.g., percutaneous catheters), as well as with diseases such as native-valve endocarditis and infectious kidney stones (1,2,3,13). Staphylococcus aureus is a key pathogen, feared for its high mortality rate and antibiotic resistance (10). Endovascular biofilm infections caused by S. aureus also carry a high risk of metastasis (4, 10). These biofilms are exposed to a broad range of fluid shears, ranging from the continuous, laminar flow in infected infusion lines to the highly variable, turbulent flow in the case of bacterial vegetations on a heart valve. Therefore, the continuous adaptation to mechanical stresses is an important feature of biofilm physiology (7). The responses of biofilms to shear stresses may determine biofilm dissemination in general and the rate of formation and size of biofilm emboli in particular. In previous work, we showed that S. aureus biofilm emboli ranged in size from single cells to microcolonies containing thousands of cells in an in vitro catheter infection model (6). Large emboli expressed antibiotic (oxacillin) tolerance similar to that observed in attached biofilms. In the present study, we used a glass capillary system previously used to mimic physiological shear in both catheter (6) and vascular (9) infection models to quantify the viscoelastic response and motion of S. aureus biofilm microcolonies in response to varying or steady fluid shear over short (seconds) and sustained (minutes) periods. Biofilms were grown from an S. aureus strain (ATTC 25923) on 1/10-strength brain heart infusion broth at 37°C. One milliliter of a 24-h broth culture was inoculated into a 1-mm 2 , 140-mm-long glass capillary flow cell (model FC91; BioSurface Technologies, Bozeman, Mont.) integrated into a once-through flow system. Biofilms were monitored with a Cohu (San Diego, Calif.) model 4910 camera mounted on an Olympus BH2 microscope using Scion Image software (Scion Inc., Frederick, Md.). After an attachment period of 30 min, a continuous laminar flow rate (Q) of 60 ml h Ϫ1 was established to approximate the hydrodynamic conditions typically used in a central venous catheter (6). The average velocity was 1.67 cm s Ϫ1 , the Reynolds number was 17, and the wall ...

The mechanical properties of mixed culture biofilms were determined by creep analysis using an AR1000 rotating disk rheometer. The biofilms were grown directly on the rheometer disks which were rotated in a chemostat for 12 d. The resulting biofilms were heterogeneous and ranged from 35 microns to 50 microns in thickness. The creep curves were all viscoelastic in nature. The close agreement between stress and strain ratio of a sample tested at 0.1 and 0.5 Pa suggested that the biofilms were tested in the linear viscoelastic range and supported the use of linear viscoelastic theory in the development of a constitutive law. The experimental data was fit to a 4-element Burger spring and dashpot model. The shear modulus (G) ranged from 0.2 to 24 Pa and the viscous coefficient (eta) from 10 to 3000 Pa. These values were in the same range as those previously estimated from fluid shear deformation of biofilms in flow cells. A viscoelastic biofilm model will help to predict shear related biofilm phenomena such as elevated pressure drop, detachment, and the flow of biofilms over solid surfaces.