Rapid formation of bioaggregates and morphology transition to biofilm streamers induced by pore-throat flows

Edited by Howard Stone, Princeton University, Princeton, NJ; received March 14, 2022; accepted February 28, 2023
March 29, 2023
120 (14) e2204466120

Significance

Biomass in porous media is commonly found as aggregates, which strongly affect the physical and biochemical environment of porous media systems. However, the underlying biophysical mechanisms that cause bioaggregation remain unclear. Here, we elucidate the formation and shaping mechanisms of bioaggregates in porous media. First, pore throat structure is shown to cause unique flow structures that lead to an enhanced biomass attachment at the pore throat. Further, bioaggregates show a change in mechanical behavior from elastic to viscous when critical shear stress is reached. This critical shear stress is shown to determine the morphology transition from rounded to streamer-like bioaggregates. Our findings provide a critical understanding for engineering the morphology of bioaggregates and biofouling.

Abstract

Bioaggregates are condensed porous materials comprising microbes, organic and inorganic matters, and water. They are commonly found in natural and engineered porous media and often cause clogging. Despite their importance, the formation mechanism of bioaggregates in porous media systems is largely unknown. Through microfluidic experiments and direct numerical simulations of fluid flow, we show that the rapid bioaggregation is driven by the interplay of the viscoelastic nature of biomass and hydrodynamic conditions at pore throats. At an early stage, unique flow structures around a pore throat promote the biomass attachment at the throat. Then, the attached biomass fluidizes when the shear stress at the partially clogged pore throat reaches a critical value. After the fluidization, the biomass is displaced and accumulated in the expansion region of throats forming bioaggregates. We further find that such criticality in shear stress triggers morphological changes in bioaggregates from rounded- to streamer-like shapes. This knowledge was used to control the clogging of throats by tuning the flow conditions: When the shear stress at the throat exceeded the critical value, clogging was prevented. The bioaggregation process did not depend on the detailed pore-throat geometry, as we reproduced the same dynamics in various pore-throat geometries. This study demonstrates that pore-throat structures, which are ubiquitous in porous media systems, induce bioaggregation and can lead to abrupt disruptions in flow.
Microbes in natural and engineering systems are often found as aggregates consisting of cells, organic and inorganic matters, and water (1, 2), known as bioaggregates. The term “bioaggregate” is not uniquely defined in the literature (37). In this study, we broadly define bioaggregates as either suspended or partially surface-attached aggregates of biomass. In their surface-attached form, bioaggregates are only partly attached to surfaces. This characteristic allows to distinguish them from biofilms, which are sessile surface-attached bacterial colonies (8, 9). Bioaggregates play essential roles in biogeochemical processes in subsurface and marine environments (10, 11), bioremediation of contaminations (12, 13), biofilm formation (14), clogging (15), biomineralization (16), and infection of human lungs (17, 18). Moreover, aggregated cells display enhanced protections against external stresses such as antibiotics (19), nutrient starvation (20), oxidative stress (21) facilitating the adaptation to environmental changes. Thus, bioaggregation has been an active area of research in various fields, including evolutionary biology and environmental and biomedical sciences. Further, bioaggregates have recently been recognized as another major form of biomass in addition to biofilm and planktonic cells. As a result, scientists have proposed a new conceptual framework to include aggregates into the widely accepted biofilm life cycle model (22). Despite such recognition, the underlying mechanism that leads to bioaggregation is still largely an open question.
Pore-scale flows and channel geometry have been shown to play a critical role in shaping biofilm morphology (2328). For example, converging secondary flows generated by corners (29) and pillars (30) are known to rapidly accumulate extracellular polymeric substances (EPS) and bacterial cells in specific locations of curved surfaces. Further, the viscoelastic properties of biomass make the formation and shaping of biofilms sensitive to flow conditions: Biomass exhibits elastic properties at low stress, but it behaves like viscous fluids when the applied stress exceeds its yield point (31). This viscoelastic property of biofilms allows them to adapt their architecture based on the flow conditions: Biofilms grow as surface-attached layers under stagnant conditions; however, they may take the form of biofilm streamers with extruded filaments under fluid flowing conditions (32, 33). Thus, the bioaggregation process may also be strongly influenced by pore structure and pore-scale flow conditions.
Indeed, bioaggregates are often found in systems that involve porous media flows (3436). Column-scale experiments reported that bioaggregates are frequently found in channel contraction and expansion regions (i.e., pore throats) (37, 38). However, most previous pore-scale studies considered channels with a constant width or 2D porous media systems with loosely spaced pillars, which did not exhibit bioaggregates (23, 30, 32). In nature, both pore structure and flow conditions can vary widely, but their effects on bioaggregation are yet to be clarified. Consequently, there is a lack of fundamental understanding regarding the formation mechanism of bioaggregates at the pore scale, and questions about the impact of pore-throat structures and flow conditions on bioaggregate formation remain open.
In this study, we elucidate the pore-scale mechanisms driving bioaggregation at pore throats by combining microfluidics experiments and direct numerical simulations of fluid flow. A microfluidic flow channel with a constriction followed by an expansion was used to model a pore-throat segment in porous media systems (Fig. 1A); Escherichia coli was used as the model bacterium because of its tendency to readily form biofilms under diverse conditions (39). Upon the injection of E. coli suspensions, we found that the cells rapidly form bioaggregates at the throat and clog the pore. By systematically varying pore-throat geometries and flow rates, we show that unique flow structures at a pore throat lead to the rapid aggregation of cells in throats of various shapes. Furthermore, we identify the critical shear stress below at which cells form aggregates and clog a throat and above at which aggregates transition to a streamer-like morphology.
Fig. 1.
E. coli bioaggregates are rapidly formed at a pore throat. (A) Schematics of the microfluidics platform used in this study. The pore throat is 35 μm wide and 55 μm deep. Inset: velocity field and streamlines at the channel midplane under a mean flow velocity of 1.7 mm/s at the pore throat. Constant flow rate of 0.2 μL/min, corresponding to an average shear rate of 24.7 s−1 at the throat, is imposed with a syringe pump and fluorescently tagged E. coli bacterial suspension is injected. Fluorescent images are captured every minute. (B) Fluorescence images, focused on the channel middepth show the formation of bioaggregates near the pore throat at different time points. (C) 3D reconstruction of bioaggregates generated from z-stack images (dz =0.5 μm) acquired with a confocal laser scanning microscopy (CLSM). The main figure shows bioaggregates after 0.5 h of injection under the same experimental condition as in B. Insets show the top view of the 3D reconstructed images of bioaggregates at 0.1 and 0.5 h, respectively. Yellow arrows indicate the flow direction. (D) Biomass accumulation at the pore-throat region (circles) and on the surface of the straight channel (triangles) in time, measured from the fluorescent intensity of the E. coli cells. Biomass quantification is performed by summing the fluorescence intensity of pixels where the fluorescent bacterial cells are present in a selected area of 680 μm wide by 155 μm high. The centers of the selected areas are located 120 μm downstream of the center of the pore throat for the pore-throat curve and 5 mm upstream of the pore throat for the straight-channel curve. The color of the filled symbols represents the data points corresponding to the experimental images in the Inset and panel B. Inset: Fluorescence images of E. coli cells attached to the surface of the straight portion of the channel, taken after 0.1 h (black triangle) and 4 h (orange triangle) during the same experiment reported in B.

Results

E. coli Bioaggregates Are Rapidly Formed at a Pore Throat.

We observe a rapid formation of bioaggregates at the pore throat (Fig. 1B) after the injection of an E. coli cell suspension in a lactose minimal media (OD600 =0.1) into the polydimethylsiloxane (PDMS) microfluidic channel (Fig. 1A) at a constant flow rate of 0.2 μL/min or an average shear rate of 24.7 s−1 at the throat. The flow is laminar with a Reynolds number (Re) of 0.07, and the streamlines smoothly follow the shape of the pore throat (Fig. 1A, Inset). Bacteria cells preferentially attach to the channel walls at the throat region within a few minutes (black circle, Fig. 1B), and round-shaped bioaggregates with a diameter of approximately 50 μm are formed within 0.5 h of continuous flow (red circle, Fig. 1B). Particle size distribution analyses are performed to quantify the presence of preformed aggregates in the injected solution (SI Appendix, Fig. S1A). Although 1% of the particle volume had sizes larger than 3.5 μm and the maximum particle size was 5 μm, their sizes are still significantly smaller than the pore throat size (35 μm). In addition, we confirmed that E. coli cells attach at the pore throat primarily as a single cell (SI Appendix, Movie S1 and Fig. S1B). Even if preformed aggregates attach at the pore throat, they will not contribute significantly to the blockage of a throat due to the large throat-to-particle size ratio (40). As more cells aggregate and clog the throat, the aggregates are pushed downstream (blue circle, Fig. 1B) and eventually fully detach at around 2.5 h, leaving the pore throat open (green circle, Fig. 1B). The 3D reconstruction of bioaggregates at 0.1 and 0.5 h from images acquired using a confocal laser scanning microscopy (CLSM) highlights the preferential attachment of biomass at the pore throat and the rapid development of bulky bioaggregates at the downstream region (Fig. 1C). The bioaggregates span the entire channel depth at 0.5 h, and the central part of the channel remains open with a relatively uniform opening width.
Movie S1.
Timelapse video showing the aggregation of E. coli cells at the throat under lactose minimal medium condition. OD600 of 0.1 suspension was injected at a flow rate of 0.2 μL/min. Fluorescence images were taken every 1 minute, and the time on the videos shows hours and minutes.
In contrast, a negligible cell attachment and accumulation are detected in the upstream, straight portion of the same channel during the same interval of time (black and orange triangles and Fig. 1D, Inset). To quantitatively compare the biomass accumulation at the pore throat and on the surface of the straight region of the channel, we measured the fluorescence intensity, which is proportional to the biomass amount, at both locations. The intensity measured at the pore-throat region was significantly higher (up to two orders of magnitude) than that at the straight region (Fig. 1D). The abrupt decline at around 2.5 h is attributed to the detachment of the aggregates. The result confirms that rapid bioaggregation occurs at the pore-throat region.

Bioaggregation Is Controlled by Flow Near a Pore Throat and Viscoelasticity of Biomass.

We identify three major stages that lead to bioaggregation through a series of microfluidics experiments and direct numerical simulations of flow: the flow-induced preferential attachment of cells at a pore throat (Fig. 2A), shear stress-induced displacement of attached biomass to the expansion region (Fig. 2B), and biomass accumulation and aggregation through the continuous cycle of the attachment and displacement (Fig. 2C). Here, we summarize key aspects of each stage and then discuss each stage in detail in subsequent sections.
Fig. 2.
Three stages of bioaggregation controlled by the interplay between the flow near a pore throat and the viscoelasticity of the biomass. Schematics (Top half of each panel) and experimental images color-coded with respect to time (Bottom half of each panel) show the three stages of bioaggregate formation. (A) Enhanced cell attachment at a pore throat induced by the combined effects of flow-line interception, secondary flow, and shear-enhanced collision. (B) Shear-induced displacement of the biomass maintains a throat-opening width, as shown by the red dotted lines. As cells attach and narrow down the throat, the shear stress exceeds the critical value at which the viscoelastic biomass becomes fluidized, and the biomass gets displaced downstream. Red dotted lines represent the boundaries where the critical shear stress is reached. (C) The aggregates accumulate in the expansion region where the shear stress drops below the critical value. The cycle of cell attachment and displacement by shear leads to the formation of bioaggregates in the expansion region.
First, unique flow structures caused by a pore-throat structure lead to the enhanced attachment of biomass near the pore throat (Fig. 2A). The densification of streamlines induced by the channel constriction brings cells closer to the channel wall, which promotes cell-to-wall collisions through flow-line interception or direct interception (4144). Further, the attachment is facilitated through shear-enhanced cell collisions and enhanced cell captures caused by converging secondary flows. The elevated shear rate at the throat enhances the cell-to-cell and cell-to-wall collision frequencies, whereas the secondary flows accumulate EPS near the middepth of the channel and capture the flowing cells.
The width of the open flow channel (red dotted lines in Fig. 2 B and C) becomes narrow as the biomass accumulates near the pore throat. For a fixed injection rate, the shear stress on the flow channel wall increases with a decrease in the opening width of the flow channel. Eventually, the shear stress on the surface of the narrowed channel reaches a critical value at which the viscoelastic E. coli biomass becomes fluidized (Fig. 2B and Movie S1). We use the term fluidization to describe the state at which the biomass behaves as viscous fluid exhibiting nonelastic properties, and we performed creep tests (45, 46) to confirm the viscoelasticity of biomass (SI Appendix, Fig. S2). Consequently, the biomass flows to the downstream expansion region where the shear stress drops below the critical value (Fig. 2C). The fluidization and displacement of the biomass at the critical shear stress maintains an open flow path along the center of the throat (red dotted lines in Fig. 2C). This cycle of cell attachment and displacement continues until large aggregates are formed, which eventually get discharged.

Unique Flow Structures at a Pore Throat Enhance Biomass Attachment.

To test the role of microbial growth on bioaggregation, we performed microfluidic experiments with E. coli cells suspended in carbon source-deprived medium that limits microbial growth. As shown in Movie S2, a very similar bioaggregation pattern was observed at the pore throat. In addition, the estimated doubling time of 2.8 ± 0.7 h from the growth experiment (SI Appendix, Fig. S3) in the lactose minimal medium is significantly larger than the timescale of bioaggregation (< 30 min from Fig. 1B). On the other hand, the advective and diffusive transport timescales of cells were estimated to be 0.02 s and 6 s, respectively (SI Appendix for details). In addition, the experiments with abiotic particles show no aggregation of particles at the pore throat even after 2 hrs of injection (SI Appendix, Fig. S4). This indicates that the observed bioaggregation cannot be easily reproduced by abiotic particles and that unique properties of EPS and extracellular appendages such as pili and fimbriae aid E. coli attachment and aggregation (47). These results collectively indicate that microbial growth is not playing a meaningful role in the observed bioaggregation and confirm the relevance of hydrodynamics conditions at the pore throat and surface properties of cells on the initial attachment.
Movie S2.
Timelapse video showing the aggregation of E. coli cells at the throat under carbondeprived medium (no growth) condition. OD600 of 0.1 suspension was injected at a flow rate of 0.2 μL/min. Fluorescence images were taken every 1 minute, and the time on the videos shows hours and minutes.
We identified three flow features in the pore-throat region that enhance cell attachment: streamline densification, secondary flows, and shear-enhanced collision. First, streamline densification attributed to the constriction (Fig. 3A) brings cells closer to the throat surface, which increases the chances of cell attachment. This particle deposition mechanism on a surface is called flow-line or direct interception; it has been widely exploited in venturi scrubbers that have a constriction in a pipe for capturing airborne particulates (41, 42).
Fig. 3.
Unique flow structures at a pore throat enhance cell attachment. 3D images are produced from CLSM images obtained at t =0.1 h and 3D numerical simulations of flow. All experiments are performed with E. coli (OD600 = 0.1) at 0.2 μL/min flow rate (24.7 s−1 shear rate at the pore throat); yellow arrows represent the flow direction. Streamlines, velocity profile, shear rate, and secondary flow magnitudes are computed from direct numerical simulations of flow. (A) E. coli attachment via flow-line interception. 3D reconstructed image from CLSM shows the enhanced attachment of E. coli biomass at the throat as the streamline densification occurs. (B) Converging secondary flows in the z-direction facilitates the capture of cells in the middepth of the channel. The pronounced accumulation of biomass at the middepth corresponds to the expansion region of the throat where the z-directional flow convergence occurs. The cross-sectional z-component velocity field at 10 μm downstream of the throat was obtained from the numerical simulation, and it is superimposed on the 3D distribution of biomass. (C) Velocity profiles at the middepth of the straight and pore-throat regions. (D) Shear rate magnitude profiles across the lines ysys′ (black line) and yy′ (red line) shown in C. The gray dotted lines represent the channel walls. The shear rate is significantly higher near the pore-throat region. (E) Fluorescence intensity of the biomass, normalized shear rate, and normalized secondary flow magnitudes are plotted as a function h/H. The magnitudes of the shear rate at the channel middepth and the z-directional secondary flow are obtained at 10 μm downstream from the throat. The amounts of biomass are estimated by summing up the fluorescence intensity in the span of 680 μm by 155 μm centered at the pore throat. The average values from four independent experiments are plotted, and the error bars indicate SDs.
Second, the 3D flow simulations revealed the existence of z-directional secondary flows caused by the constriction and expansion (Fig. 3B). The z-component flow velocities in the expansion side are directed toward the middle xy-plane, which results in the flow convergence to the middepth of the channel. Such converging secondary flows around corners concentrate EPS in the half-depth of the channel onto which cells are captured (23); this promotes cell adhesion. Indeed, we observe the preferential attachment of cells near the middle plane (Fig. 3B).
The third cause for the preferential attachment of cells at the throat is the elevated shear rate (i.e., velocity gradient, du/dy) at the throat. The particle-to-particle collision frequency is proportional to the shear rate (48, 49). Similarly, the increased shear rate enhances the mixing of flow parcels in porous media flows (50, 51). The shear rate increases and maximizes at the pore throat as the channel is narrowed at the constriction; this is shown in Fig. 3 C and D. Further, an increased shear can bring cells close to each other and to walls overcoming the Derjaguin–Landau–Verwey–Overbeek (DLVO) repulsive force, which promotes the attachment (52, 53).
We studied channels with five different throat widths (h) while keeping the channel width (H) fixed to further confirm the relationship between the biomass attachment and the pore-throat structure (Fig. 3E). The results from quadruplicate microfluidic experiments show a dramatic increase in the amount of attached biomass as the ratio of the pore-throat width to channel width (h/H) decreases (Fig. 3E). The numerical simulations of flow are performed to quantify the magnitudes of shear rate and secondary flows (blue squares and red triangles in Fig. 3E), and the results show a strong correlation between the magnitude of the aforementioned flow structures at the throat and biomass attachment. The degree of streamline densification is also inversely proportional to h/H. From the mass conservation, the number of streamlines passing through a cross-sectional area should be constant (54). Hence, when the cross-sectional area of a flow channel (which is proportional to h/H) reduces, the interstreamline distance also decreases proportionally (i.e., the streamline density increases), as illustrated in Fig. 1A, Inset. The strong correlation between the unique flow structures and the enhanced biomass attachment is confirmed, but further study is required to quantify the relative importance of each flow feature on biomass attachment.

Critical Shear Stress at which Biomass Becomes Fluidized Exists.

We observe that a fairly constant throat opening width is maintained (Fig. 2 B and C) during the cycle of attachment and displacement of the biomass at the pore throat. The constant throat-opening width for a given flow rate implies that there is a critical shear stress value that triggers the mobilization of E. coli biomass to the downstream, which prevents the complete clogging of the throat. Further, from a series of experiments at varying flow rates, we observe an increase in the throat-opening width at increasing flow rate (Fig. 4A). To estimate the critical shear stress value at which biomass becomes fluidized, we estimate throat-opening widths from fluorescence intensity profiles as shown in Fig. 4A. First, three linear lines are fitted to each profile: the flat valley region in the middle corresponding to the biomass-free region and the two nearby linear regions where biomass is present. Then, the distance between the intersecting points of the fitted linear lines is estimated as the throat-opening width as shown in SI Appendix, Fig. S5. The opening width at the throat increases from 10.4 ± 1.4 to 15.7 ± 1.0 and 20.1 ± 1.1 μm with an increase in the flow rate from 0.1 to 0.2 and 0.3 μL/min, respectively. Then, we estimated the shear stress values at the respective channel-opening widths.
Fig. 4.
Critical shear stress controls biomass fluidization, morphological change, and clogging. (A) Fluorescence intensity profiles along the crossline yy’ at three different flow rates after 0.5 h of injection. The channel has an h/H value of 0.23. The throat-opening widths are obtained from the fluorescence intensity profiles. The width values are reported as the average values with standard deviations as error bars from five replicate microfluidic experiments at each flow rate of 0.1, 0.2, and 0.3 μL/min with corresponding average shear rates of 24.7, 49.5, and 74.2 s−1, respecitvely. (B) Calculation of maximum shear stress at the throat at different flow rates through the direct numerical simulations of flow. The biomass is incorporated as cylindrical impermeable and no-slip surfaces of which the degree of intrusion into the channel is determined from experiments. The maximum shear stress values at the throat with intrusions are represented by green, blue, and red asterisks for flow rates 0.1, 0.2, and 0.3 μL/min, respectively. Black circles represent the maximum shear stresses at a clean throat wall. In a clean throat, the critical shear stress of 1.8 Pa is reached at 0.75 μL/min. (C) Bioaggregates with different morphologies from experiments at various flow rates (Top) and the shear stress map at the middepth of the channel (Bottom) obtained from numerical simulations. The white area (noted by white arrows) represents the regions with critical shear stress values (1.8 ± 0.2 Pa), whereas the red and blue areas represent regions with shear stress values higher and lower than the critical value, respectively. (D) Flow rate profile from experiments at a constant pressure of 31 mbar, which corresponds to an average initial flow rate of 0.3 μL/min and max shear stress at the clean throat of 0.7 Pa. Fluorescent images in the plot denotes E. coli accumulation at the corresponding time points. The fluctuation in the flow rate is attributed to the sway of biomass at the throat. Inset: a snapshot of the throat after 2 h of operation at 40 mbar, which corresponds to an average flow rate of 1.9 μL/min and max shear stress at the clean throat of 4.6 Pa.
Shear stress values along the biomass–fluid boundary at the throat under different flow rates are calculated from the numerical simulations of flow. Bacterial transport and attachment of cells are not explicitly simulated. However, instead, we incorporated biomass attached at the throat as no-slip and impermeable, as described in ref. 55, cylindrical surfaces in the flow model (Fig. 4B, Inset) as the biomass subject to shear stress below the critical value is expected to be immobile (56) and compact because of the increased pressure at the pore throats (57, 58). The degree of extrusion into the channel is determined by the throat-opening width obtained experimentally at different flow rates (Fig. 4A). Such an approach to flow simulation allowed us to estimate shear stress values at the biomass–fluid interfaces. At all flow rates, the computed shear stress values at the biomass–fluid interfaces at the middepth were close to 1.8 Pa with an error range of ±0.2 Pa. Green, blue, and red asterisks in Fig. 4C corresponds to flow rates 0.1, 0.2, and 0.3 μL/min, respectively. Hence, we can conclude that the value of 1.8 ± 0.2 Pa is the critical shear stress value that determines the yield point of the biomass considered in this study; this value is in the range of the yield point of diverse bacterial biomass (59). Note that this value of critical shear stress should be specific to the microbial strain and the growth conditions tested in this study. The critical shear stress value is expected to be dependent on the bacterial strain, growth phase, surface properties, and chemistry of suspension solution (31, 60). Our creep test results confirm that the attached biomass at the pore throat behaves as an elastic material at a low shear stress while it behaves as viscoelastic fluid upon exposure to a shear stress value that is close to the critical value (SI Appendix, Fig. S2). The result confirms that the aggregation of attached biomass occurs through the interplay between the viscoelasticity of biomass and varying shear stress at the throat (Fig. 2 B and C). Biomass becomes fluidized when the shear stress reaches the critical value of 1.8 Pa and then accumulates and aggregates in the expansion regions where shear stress drops below the critical value.

Critical Shear Stress Controls the Morphological Change of Bioaggregates and Pore-Throat Clogging.

Interestingly, the morphology of bioaggregates transition from that of the rounded shape to the elongated streamer-like shape when the flow rate that induces the critical shear stress at the clean pore-throat wall (0.75 μL/min) is imposed (Fig. 4C). Round-shaped bioaggregates are formed in the expansion region at the flow rate of 0.65 μL/min or average shear rate of 161 s−1, at which the shear stress at the clean throat is below the critical value of 1.8 Pa. On the other hand, the suspended filaments of biomass with a streamer-like morphology are formed when the clean throat is exposed to shear stress equal to or slightly higher than the critical value that corresponds to the flow rates of 0.75 and 0.85 μL/min or 185 and 210 s−1, respectively. The timescale of streamer formation was in the order of 10 min (Movie S3). Similar to bioaggregate formation, the transport (advection and diffusion) timescales are more relevant to the streamer formation than cell growth timescales. Hence, the streamer formation is driven by the interplay between the viscoelasticity of the biomass and the hydrodynamic conditions, and not by microbial growth. This result suggests that attached cells are almost immediately fluidized and stretched downstream, forming biofilm streamers, when a throat experiences sufficiently high shear stress. That is, above the critical shear stress, the rate of biomass stretching is higher than the rate of biomass accumulation, which leads to biofilm streamers. In addition, we observed that cells attach directly onto the streamers as reported by Drescher et al. (32), contributing to the thickening of the streamers. When the flow rate is raised by one order of magnitude (flow rate > 10 μL/min), the shear stress in the entire throat region is higher than the critical value (Fig. 4C red area), and neither aggregates nor streamers are formed. At this high flow rate, high shear stress in the pore-throat region prevents the attachment and accumulation of biomass on the wall. The result confirms that the critical shear stress also controls the morphological change of bioaggregates from rounded- to streamer-like shapes.
Movie S3.
Timelapse video showing the formation of streamer-like bioaggregate of E. coli cells at the throat under lactose minimal medium. OD600 of 0.1 suspension was injected at a flow rate of 0.75 μL/min. Fluorescence images were taken every 1 minute, and the time on the videos shows hours and minutes.
The concept of critical shear stress can be related to that of critical pressure, which determines the clogging of the pore throats. Imposing a constant flow rate implies that the shear stress will continue to increase as the throat becomes narrower until it reaches the critical value that fluidizes the attached biomass. On the other hand, the pore throat may be completely clogged when a constant pressure boundary condition is imposed because the flow rate will decrease continuously as the pressure drop increases owing to cell attachment. To test this, we imposed the flow of an E. coli suspension (OD600 = 0.1) using a constant pressure pump set at 31 mbar, which initially produced an average flow rate of 0.3 μL/min and the maximum shear stress of 0.7 Pa. The result shows the rapid accumulation of biomass at the pore throat and the abrupt reduction of flow rate to near zero within 1 h of the experiment (Fig. 4D). The throat remained clean throughout the entire experiment duration of 2 h when the pressure is raised to 40 mbar, which yields the maximum shear stress at the throat of 4.6 Pa (Fig. 4D, Inset). The result indicates that the clogging of a throat can be controlled by tuning the shear stress around a pore throat. One can maintain a clean throat by making the shear stress around the pore throat above the critical value. If not, the throat will be rapidly clogged.

Bioaggregation at Pore Throats Occurs in Diverse Flow Systems.

Pore throats are ubiquitous in natural and engineered systems such as soil-like porous media (Fig. 5A), hydrogels (Fig. 5B), filtration membranes (61), and blood flows in the brain (62, 63). Pore throats in such diverse porous media systems have various pore-throat geometries. We conducted microfluidics experiments with different pore-throat geometries to demonstrate the generality of our findings. Instead of the sinusoidal throat, we incorporated semicircular and rectangular throats in a channel. The throat width and flow rate are 35 μm and 0.2 μL/min, respectively, which are identical to those in the experiments with the sinusoidal throat shown in Fig. 1. After 30 min of flow, the results show similar flow-induced bioaggregation at both rectangular and semicircular throats (Fig. 5C). This result indicates that the bioaggregation process is not specific to a particular geometry and can occur in various types of pore throats when the maximum shear stress at the throat is below the critical shear stress value.
Fig. 5.
Bioaggregation at pore throats occurs in diverse flow systems. (A and B) Pore throats are commonly found in porous media, as we show in two examples: the velocity field of a 2D soil-like media showing the effects of pore throats on local flow fields (A) and the scanning electron microscopy image of a pore throat in a polyvinyl alcohol-alginate hydrogel bead (B). The velocity field in A was obtained from the direct numerical simulation of flow with an inlet velocity of 3.5 mm/s. The bead in B was prepared via extrusion-dripping technique followed by ion gelation crosslinking (64). (C) Bioaggregation in a rectangular (Left) and semicircular (Right) pore throats at the flow rate of 0.2 μL/min after 0.5 h of flow. The width and height of the rectangle are 120 μm and 60 μm, respectively, and the radius of the semicircle is 60 μm. The pore-throat width h is set to 35 μm (h/H =0.23). (D) Flow through pore throats can lead to the formation of multispecies bioaggregates. After the injection of the 1 : 1 mixture of E. coli (OD600 = 0.2 and shown in green) and Methylobacterium extorquens (OD600 = 0.05 and shown in red), multispecies bioaggregates rapidly formed at a throat. The image was captured when the aggregate was detached from the sinusoidal pore-throat channel. All experiments are performed at a constant flow rate of 0.2 μL/min with initial cell concentrations of OD600 = 0.1 at 25 °C.
Furthermore, we observed that the bioaggregation process promotes the formation of multispecies bioaggregates (Fig. 5D). We injected a 1 : 1 mixed culture bacterial suspensions of E. coli (OD600 = 0.2), shown in green, and Methylorubrum extorquens (OD600 = 0.05) cells, shown in red, into the channel with the sinusoidal pore throat at a flow rate of 0.2 μL/min. Like in the previous experiments with E. coli cells alone, E. coli cells rapidly attached and aggregated near the pore throat while M. extorquens cells were trapped amid E. coli aggregates. The resulting mixed culture injection eventually developed into multispecies bioaggregates and detached downstream, wherein the image was captured. The result shows that bioaggregates can be formed in diverse systems and that pore throats can facilitate the formation of multispecies bioaggregates promoting active interspecies interactions.

Discussion and Conclusion

There has been no clarity on how bioaggregates are formed in response to fluid flow and channel geometry, although numerous packed-column experiments (37, 38) and modeling studies (65, 66) have stressed the importance of aggregates in the biofouling of porous media. To the best of our knowledge, this is the first study to report the underlying bioaggregation mechanisms and visualize the flow-induced bioaggregation process at the pore scale. We first elucidated how flow structures formed at the pore throat enhance the initial attachment of E. coli cells. The constriction densifies the streamlines bringing cells closer to the throat walls, which facilitates the interception of cells at the throat. In addition, the reduced cross-sectional area increases the shear rate at the throat, which enhances the collision frequency between the wall and the cells. The subsequent expansion of the channel generates the z-directional converging secondary flows reported to facilitate cell attachment. The shear stress increases to a critical value as the cells attach and narrow down the throat; at this point, the biomass is fluidized and displaced downstream, where the cells become aggregated. The findings also clarify the seemingly counterintuitive phenomenon of the cells being attached at pore throats where the flow velocity is highest. At low to moderate flow rates, the shear stress is not high enough to offset the cell attachment caused by the flow structures at the pore throat, and the increase in shear stress contributes positively to the cell attachment (67). On the other hand, the large shear inhibits cell attachments under a very high shearing condition (well above the critical shear stress, Fig. 4C) (56). Such duality in the role of shear stress in cell attachment has also been evidenced in other studies (68, 69).
Our results suggest that the observed critical shear stress is the yield point of the viscoelastic biomass, and such a concept is crucial for explaining the biomass morphology and clogging of throats. We observed the morphological change in E. coli bioaggregates from the rounded shape to the elongated streamer-like shape; the transition was explained by the concept of critical shear stress. Previous studies have shown that biofilm morphology change could also be induced by the preferential orientation of cells, controlled by the local shear rate (70, 71). However, our results demonstrated that an increased shear triggers the streamer extrusion (72), and no preferential alignment in the cells forming the streamers is detected (SI Appendix, Fig. S6). Further, throats remained open when the shear stress exceeded the critical value while they became rapidly clogged (i.e., significant reduction in flow rate) when the maximum shear stress at the throat was below the critical value. The concept of critical shear stress was related to that of critical pressure, which implies the presence of a minimum pressure for anticlogging. Such a concept of critical pressure in clogging was alluded to in the previous simulation study with abiotic particles flowing through a constriction (73); however, this is the first experimental demonstration. Moreover, it is noteworthy that aggregates can detach from the throat and plug downstream throats, leading to an abrupt decline in flow rate and rise in pressure drop.
Studies have reported that biofilms of different morphologies can significantly alter local hydrodynamics which subsequently influence the mixing and reaction dynamics (32, 7476) and microbial community structures (77). Especially, it has been reported that significantly less biomass is needed to clog a porous media system through aggregates as opposed to wall-attached biofilms (78). Flow-induced bioaggregates can be detached and instantly clog downstream pore throats, while it takes significantly more biomass and a longer period for a surface-grown biofilm to clog a throat. Thus, the results of this study highlight that biomass-induced clogging of a porous medium can proceed rapidly by flow-induced cell attachment and aggregation. As such, the ‘chunky’ bioaggregates at pore throats under the low shear stress can lead to a dramatically different mixing, reaction, and microbial population dynamics compared to those when surface-attached biofilms or streamer-like bioaggregates are formed (79, 80). These findings have direct implications on the upscaling and prediction of flow, reactive transport, and biomass accumulation in porous media (8185).
As evidenced in our experiment, flow-induced bioaggregation promotes the formation of multispecies bioaggregates. The proximity of microbes within the bioaggregate will allow active intra- and inter-species interactions (86) that can drive the aggregate population structure and dynamics, which can affect its surrounding environments (87). Furthermore, the flow-induced bioaggregation at pore throats can be exploited by nonaggregate-forming microbes to be integrated into bioaggregates formed by aggregate formers because bioaggregates offer advantages in resistance to external stresses such as antibiotics and oxidation. This study shows that multispecies bioaggregates can be effectively formed at pore throats, and they can be created in lab settings.
In summary, the flow-induced bioaggregation reported in this study sheds light on the biofouling process in microchannels and porous media systems containing constriction-expansion structures. The existence of critical shear stress has direct implications on the operation of various engineered porous media flow systems aimed at minimizing biofouling. In addition, the observed relationship between the throat geometry and the attachment of microbial cells can provide insights into the design of porous media structures that are less prone to biofouling. This study will motivate future studies investigating the relationship between viscoelasticity and bioaggregation. For example, transcriptome profiling of the flow-induced bioaggregation and rigorous analysis of the viscoelastic behavior (e.g., varying viscoelasticity by adding biofilm stiffening agents such as calcium ions or polyethylene glycol) are expected to elucidate further the mechanisms of the bioaggregation and morphology transition and how the aggregated cells behave differently from biofilms grown from a surface.

Materials and Methods

Bacterial Strains and Culture Condition.

The bacterial strains used in this study are E. coli K12 wild type tagged with a cyan fluorescent protein (88) and M. extorquens AM1 wild type tagged with mCherry fluorescent protein (89). E. coli and M. extorquens solutions are prepared by inoculating a frozen stock in a lactose and acetate minimal medium, respectively, overnight at 25 °C under shaking and reinoculating 50 μL of the culture to a fresh medium until the OD600 reached values between 0.3 and 0.5. The culture was diluted with a fresh medium to yield the final OD600 of 0.1 before it was injected into microfluidic chips. The lactose minimal media was prepared following the recipe reported in Harcombe et al. (90). The composition of the acetate minimal media is as follows: 2.53 g/L K2HPO4, 2.25 g/L NaH2PO4, 0.54 g/L Na2SO4, 0.10 g/L MgSO4, 0.003 g/L CaCl2, 0.68 g/L acetate, 0.8 g/L methylamine, and 1 mL of metal mix in a 1 L of the medium. The composition of the metal mix is, in g/L, 0.34 of ZnSO4 ⋅ 7H2O, 0.20 of MnCl2 ⋅ 4H2O, 5.0 of FeSO4 ⋅ 7H2O, 2.5 of (NH4)6Mo7O24 ⋅ 4H2O, 0.25 of CuSO4 ⋅ 5H2O, 0.48 of CoCl2 ⋅ 6H2O and 0.11 of Na2WO4 ⋅ 2H2O. Media are autoclaved at 121 °C for 30 min upon preparation, and all chemicals are purchased from Sigma-Aldrich (Missouri, U.S.).

Microfluidics Experiments and Microscopy.

Poly-dimethylsiloxane (PDMS) (Sylgard 184, Dow Corning) microfluidic chips of a single straight channel (width=155 μm, length=1 cm, and depth=55 μm) having a sinusoidal pore-throat feature 5 mm from the inlet (Fig. 1E) were prepared from a SU-8 patterned silicon wafer mold fabricated from the photolithography process. Channels with rectangular and semicircular pore throats were prepared by using the same method. The width and height of the rectangle are 120 μm and 60 μm, respectively, and the radius of the semicircle is 60 μm. The pore-throat width of all throats was set to 35 μm except for experiments shown in Fig. 3E. E. coli suspension was injected into the chip using a pulsation-free syringe pump (neMESYS 290N, Cetoni). Re is defined as U0Hν where U0 is the average flow velocity through a channel, H is the aperture of the channel, and ν is the kinematic viscosity of the fluid. Average shear rate at the throat, γ˙, is defined as Q/(Ah), where Q is the flow rate, A is the cross-sectional area at the throat, and h is the throat size. The range of flow rates and respective Reynolds (Re) numbers and shear rates are 0.1 to 10 μL/min, 0.03 to 3.4, and 24.7 to 2470 s−1, respectively. Bacterial motility was assumed to be negligible as the minimum mean velocity at the throat (1 mm/s) in this study was significantly higher than the typical velocity of E. coli cell by two orders of magnitude (91, 92). For flows under the constant pressure condition, a pressure pump system (Flow-EZ Module, Fluigent, Massachusetts, U.S.) with a flow meter (FLOW UNIT XS, Fluigent, Massachusetts, U.S.) was used. The cells were monitored through either a scientific CMOS camera (Orca-Flash4.0, Hamamatsu, Shizuoka, Japan) connected to a fully motorized epifluorescence inverted microscope system (TI2-E Nikon, Tokyo, Japan). A Confocal laser scanning microscope (CLSM) system (A1Rsi Nikon, Tokyo, Japan) is used for 3D imaging of bioaggregates. S Plan Fluor ELWD 20x Ph1 ADM objective and fluorescence filters of peak excitation and emission of 470 and 525 nm, respectively, for cyan fluorescence and 560 and 630 nm for red fluorescence were used for epi-fluorescence microscope while Plan Apo VC 20x DIC N2 and fluorescence filter cube having peak excitation at 488 nm and peak emission at 525 nm were used for CLSM imaging with z-step of 0.5 μm.

Direct Numerical Simulation of Flow.

The 3D fluid flow simulation was performed in COMSOL Multiphysics (ver. 5.4). Flow channel geometry identical to that of the PDMS microfluidic chip is used, and water was injected by setting fixed inlet flow rates equal to those from the experiments. The density and kinematic viscosity of water at 25 °C of 997 kg/m3 and 8.9 × 10−7 m2/s, respectively, were used. The flow field was obtained by solving the continuity equation and the Navier–Stokes equations with the Finite Element Method (FEM) using a steady-state solver with self-adaptive simulation time step assuming Newtonian and incompressible fluid and no-slip and impermeable boundaries on surfaces. The flow channel domain was discretized into 4.1 × 106 mixture of tetrahedral, pyramid, prism, triangular, quadrilateral, edge, and vertex elements for 3D simulation via a built-in Physics-controlled meshing algorithm. Shear stress was obtained by computing dudyμ where du is the change in x-component velocity, dy is y-component displacement, and μ is the dynamic viscosity of water. The 2D velocity field of soil-like media was produced from COMSOL via the same method described above. Reported images are either directly produced from COMSOL or imported to Paraview for superimposition with CLSM results.

Image Postprocessing and Analysis.

Acquired fluorescent and CLSM images were postprocessed in either Matlab (R2017a, ver. 9.2.0) or NIS-Elements (ver. 5.11.01), Paraview (ver. 5.8.1), and BiofilmQ (ver. 0.2.2) (93). Built-in green colormap was applied to fluorescent images of E. coli in NIS-Elements software for improved visibility of the biomass. Fluorescence intensities of images obtained from epifluorescence microscopy were analyzed in Matlab through background subtraction and thresholding at 350 of intensity. The quantification of biomass (for Fig. 1C) was done by summing the fluorescence intensity of pixels where the fluorescent bacterial cells are present in the span of 680 μm by 155 μm. In the case of channel with pore throat, the center of the span was located at 120 μm downstream of the center of the pore throat, whereas it was 5 mm upstream of the pore throat for estimation of biomass attachment in the straight channel. The throat opening width was obtained by fitting trendlines in the three linear sections (flat valley region in the middle and two inclinations on both sides corresponding to biomass-free region and regions transitioning to where biomass is present, respectively) of the fluorescence intensity profile obtained after 30 min of injection. The distance between the intersecting points between the middle fitted line and each of the slanted lines on both sides was calculated.
3D reconstruction of biomass from CLSM images was proceeded as follows. Raw images from CLSM were deconvolved via Richardson–Lucy algorithm (100 iterations) in NIS-element, and the resulting images were voxelized in BiofilmQ. Operations involved in voxelization of biomass in BiofilmQ involved denoising (suppression of floating cells via median filtering along z and application of top-hat filter by 11 voxels), manual thresholding by 134, clumping of objects into cubes of 1.02 μm side length, and removal of voxel clusters less than 0.85 μm3. The final product in VTK format was imported to Paraview for visualization.

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.

Acknowledgments

This work was supported by MnDRIVE Advancing Industry, Conserving Our Environment at the University of Minnesota, and preparation of PDMS microfluidic devices were conducted in the Minnesota Nano Center, which is supported by the NSF through the National Nano Coordinated Infrastructure Network (NNCI) under Award Number ECCS-2025124. We would like to thank Prof. William Harcombe and Dr. Jeremy Chacon in the College of Biological Sciences at the University of Minnesota for kindly donating the E. coli and M. extorquens strains.

Author contributions

S.H.L. and P.K.K. designed research; S.H.L. performed research; S.H.L., E.S., and P.K.K. analyzed data; and S.H.L., E.S., and P.K.K. wrote the paper.

Competing interests

The authors declare no competing interest.

Supporting Information

Appendix 01 (PDF)
Movie S1.
Timelapse video showing the aggregation of E. coli cells at the throat under lactose minimal medium condition. OD600 of 0.1 suspension was injected at a flow rate of 0.2 μL/min. Fluorescence images were taken every 1 minute, and the time on the videos shows hours and minutes.
Movie S2.
Timelapse video showing the aggregation of E. coli cells at the throat under carbondeprived medium (no growth) condition. OD600 of 0.1 suspension was injected at a flow rate of 0.2 μL/min. Fluorescence images were taken every 1 minute, and the time on the videos shows hours and minutes.
Movie S3.
Timelapse video showing the formation of streamer-like bioaggregate of E. coli cells at the throat under lactose minimal medium. OD600 of 0.1 suspension was injected at a flow rate of 0.75 μL/min. Fluorescence images were taken every 1 minute, and the time on the videos shows hours and minutes.

References

1
R. L. Wilpiszeski et al., Soil aggregate microbial communities: Towards understanding microbiome interactions at biologically relevant scales. Appl. Environ. Microbiol. 85, e00324-19 (2019).
2
K. U. Totsche et al., Microaggregates in soils. J. Plant Nutr. Soil Sci. 181, 104–136 (2018).
3
H. Aqeel et al., Drivers of bioaggregation from flocs to biofilms and granular sludge. Environ. Sci. Water Res. Technol. 5, 2072–2089 (2019).
4
Z. Ding et al., Role of extracellular polymeric substances (EPS) production in bioaggregation: Application to wastewater treatment. Appl. Microbiol. Biotechnol. 99, 9883–9905 (2015).
5
T. R. Neu, J. R. Lawrence, “Extracellular polymeric substances in microbial biofilms” in Microbial Glycobiology (Academic Press, 2009), pp. 733–758.
6
R. Xu, S. Zhang, F. Meng, Large-sized planktonic bioaggregates possess high biofilm formation potentials: Bacterial succession and assembly in the biofilm metacommunity. Water Res. 170, 115307 (2020).
7
V. Gordon et al., Physics of biofilms: The initial stages of biofilm formation and dynamics. New J. Phys. 16, 045005 (2014).
8
C. D. Nadell, J. B. Xavier, K. R. Foster, The sociobiology of biofilms. FEMS Microbiol. Rev. 33, 206–224 (2009).
9
F. Beroz et al., Verticalization of bacterial biofilms. Nat. Phys. 14, 954 (2018).
10
A. Ebrahimi, D. Or, Microbial community dynamics in soil aggregates shape biogeochemical gas fluxes from soil profiles - upscaling an aggregate biophysical model. Glob. Change Biol. 22, 3141–3156 (2016).
11
E. A. Pelve, K. M. Fontanez, E. F. DeLong, Bacterial succession on sinking particles in the ocean’s interior. Front. Microbiol. 8, 2269 (2017).
12
A. Akbari, S. Ghoshal, Bioaccessible porosity in soil aggregates and implications for biodegradation of high molecular weight petroleum compounds. Environ. Sci. Technol. 49, 14368–14375 (2015).
13
B. E. Rittmann, P. L. McCarty, Environmental Biotechnology: Principles and Applications (Tata McGraw-Hill Education, 2012).
14
K. N. Kragh et al., Role of multicellular aggregates in biofilm formation. mBio 7, e00237-16 (2016).
15
P. C. Baveye, C. Darnault, Microbial competition and evolution in natural porous environments: Not that simple. Proc. Natl. Acad. Sci. U.S.A. 114, E2802–E2803 (2017).
16
J. Jimenez-Martinez, J. Nguyen, D. Or, Controlling pore-scale processes to tame subsurface biomineralization. Rev. Environ. Sci. Biotechnol. 21, 27–52 (2022).
17
M. Alhede et al., Phenotypes of non-attached pseudomonas aeruginosa aggregates resemble surface attached biofilm. PLoS One 6, e27943 (2011).
18
W. H. Depas, M. Bergkessel, D. K. Newman, Aggregation of nontuberculous mycobacteria is regulated by carbon-nitrogen balance. mBio 10, e01715-19 (2019).
19
J. Haaber, M. T. Cohn, D. Frees, T. J. Andersen, H. Ingmer, Planktonic aggregates of staphylococcus aureus protect against common antibiotics. PLoS One 7, e41075. (2012).
20
P. Bossier, W. Verstraete, “Triggers for microbial aggregation in activated sludge?” (Tech. Rep. 1-2, 1996).
21
J. J. Tree et al., The multicopper oxidase (CueO) and cell aggregation in Escherichia coli. Environ. Microbiol. 9, 2110–2116 (2007).
22
K. Sauer et al., The biofilm life cycle: Expanding the conceptual model of biofilm formation. Nat. Rev. Microbiol. 20, 608–620 (2022).
23
R. Rusconi, S. Lecuyer, N. Autrusson, L. Guglielmini, H. A. Stone, Secondary flow as a mechanism for the formation of biofilm streamers. Biophys. J. 100, 1392–1399 (2011).
24
F. Recupido et al., The role of flow in bacterial biofilm morphology and wetting properties. Colloids Surf. B: Biointerfaces 192, 111047 (2020).
25
E. Secchi et al., The effect of flow on swimming bacteria controls the initial colonization of curved surfaces. Nat. Commun. 11, 1–12 (2020).
26
G. L. Miño et al., E. coli accumulation behind an obstacle. Adv. Microbiol. 8, 451–464 (2018).
27
E. Altshuler et al., Flow-controlled densification and anomalous dispersion of E. coli through a constriction. Soft Matter 9, 1864–1870 (2013).
28
N. Autrusson, L. Guglielmini, S. Lecuyer, R. Rusconi, H. A. Stone, The shape of an elastic filament in a two-dimensional corner flow. Phys. Fluids 23, 063602 (2011).
29
R. Rusconi, S. Lecuyer, L. Guglielmini, H. A. Stone, Laminar flow around corners triggers the formation of biofilm streamers. J. R. Soc. Interface 7, 1293–1299 (2010).
30
A. Valiei, A. Kumar, P. P. Mukherjee, Y. Liu, T. Thundat, A web of streamers: Biofilm formation in a porous microfluidic device. Lab Chip 12, 5133–5137 (2012).
31
B. W. Peterson et al., Viscoelasticity of biofilms and their recalcitrance to mechanical and chemical challenges. FEMS Microbiol. Rev. 39, 234–245 (2015).
32
K. Drescher, Y. Shen, B. L. Bassler, H. A. Stone, Biofilm streamers cause catastrophic disruption of flow with consequences for environmental and medical systems. Proc. Natl. Acad. Sci. U.S.A. 110, 4345–4350 (2013).
33
D. Scheidweiler, H. Peter, P. Pramateftaki, P. de Anna, T. J. Battin, Unraveling the biophysical underpinnings to the success of multispecies biofilms in porous environments. ISME J. 13, 1700–1710 (2019).
34
V. Chaignon, B. S. Lartiges, A. El Samrani, C. Mustin, Evolution of size distribution and transfer of mineral particles between flocs in activated sludges: An insight into floc exchange dynamics. Water Res. 36, 676–684 (2002).
35
A. Carminati et al., Water flow between soil aggregates. Transp. Porous Med. 68, 219–236 (2007).
36
C. N. Trumbore, Shear-induced amyloid formation in the brain: I. Potential vascular and parenchymal processes. J. Alzheimer’s Dis. 54, 457–470 (2016).
37
P. C. DeLeo, P. Baveye, W. C. Ghiorse, Use of confocal laser scanning microscopy on soil thin-sections for improved characterization of microbial growth in unconsolidated soils and aquifer materials. J. Microbiol. Meth. 30, 193–203 (1997).
38
P. Vandevivere, P. Baveye, Saturated hydraulic conductivity reduction caused by aerobic bacteria in sand columns. Soil Sci. Soc. Am. J. 56, 1 (1992).
39
M. A. Anderson, J. E. Whitlock, V. J. Harwood, Diversity and distribution of Escherichia coli genotypes and antibiotic resistance phenotypes in feces of humans, cattle, and horses. Appl. Environ. Microbiol. 72, 6914–6922 (2006).
40
N. Delouche, B. Dersoir, A. B. Schofield, H. Tabuteau, Flow decline during pore clogging by colloidal particles. Phys. Rev. Fluids 7, 034304 (2022).
41
J. B. Joshi et al., “Design of passive safety systems for advanced reactors using CFD” in Advances of Computational Fluid Dynamics in Nuclear Reactor Design and Safety Assessment (2019), pp. 387–485.
42
C. Tien, B. V. Ramarao, Granular Filtration of Aerosols and Hydrosols (Elsevier Ltd., 2007).
43
N. Delouche, A. B. Schofield, H. Tabuteau, Dynamics of progressive pore clogging by colloidal aggregates. Soft. Matter 16, 9899–9907 (2020).
44
N. Delouche et al., The contribution of colloidal aggregates to the clogging dynamics at the pore scale. J. Membr. Sci. 635, 119509 (2021).
45
C. T. Mierke, Viscoelasticity, like forces, plays a role in mechanotransduction. Front. Cell Dev. Biol. 10, 101 (2022).
46
D. Humphrey, C. Duggan, D. Saha, D. Smith, J. Käs, Active fluidization of polymer networks through molecular motors. Nature 416, 413–416 (2002).
47
R. M. Goulter, I. R. Gentle, G. A. Dykes, Issues in determining factors influencing bacterial attachment: A review using the attachment of Escherichia coli to abiotic surfaces as an example. Lett. Appl. Microbiol. 49, 1–7 (2009).
48
M. Smoluchowski, Mathematical theory of the kinetics of the coagulation of colloidal solutions. Z. für Phys. Che. 19, 129–135 (1917).
49
S. Wangsoub, G. R. Mitchell, Shear controlled crystal size definition in a low molar mass compound using a polymeric solvent. Soft Matter 5, 525–528 (2009).
50
J. Heyman, D. R. Lester, R. Turuban, Y. Méheust, T. Le Borgne, Stretching and folding sustain microscale chemical gradients in porous media. Proc. Natl. Acad. Sci. U.S.A. 117, 13359–13365 (2020).
51
S. Yoon, M. Dentz, P. K. Kang, Optimal fluid stretching for mixing-limited reactions in rough channel flows. J. Fluid Mech. 916, A45 (2021).
52
A. J. Perez, J. E. Patiño, M. Soos, V. L. Morales, Morphology of shear-induced colloidal aggregates in porous media: Consequences for transport, deposition, and re-entrainment. Environ. Sci. Technol. 54, 5813–5821 (2020).
53
A. Zaccone, D. Gentili, M. Morbidelli, Shear-induced reaction-limited aggregation kinetics of Brownian particles at arbitrary concentrations. J. Chem. Phys. 132, 134903 (2010).
54
F. M. White, Fluid Mechanics (McGraw-Hill, ed. 7, 2011).
55
J. S. B. Tai et al., Social evolution of shared biofilm matrix components. Proc. Natl. Acad. Sci. U.S.A. 119, e2123469119 (2022).
56
M. R. Nejadnik, H. C. Van Der Mei, H. J. Busscher, W. Norde, Determination of the shear force at the balance between bacterial attachment and detachment in weak-adherence systems, using a flow displacement chamber. Appl. Environ. Microbiol. 74, 916–919 (2008).
57
J. Yang, S. Cheng, C. Li, Y. Sun, H. Huang, Shear stress affects biofilm structure and consequently current generation of bioanode in microbial electrochemical systems (MESS). Front. Microbiol. 10, 398 (2019).
58
E. Paul, J. C. Ochoa, Y. Pechaud, Y. Liu, A. Liné, Effect of shear stress and growth conditions on detachment and physical properties of biofilms. Water Res. 46, 5499–5508 (2012).
59
P. Stoodley, Z. Lewandowski, J. D. Boyle, H. M. Lappin-Scott, Structural deformation of bacterial biofilms caused by short-term fluctuations in fluid shear: An in situ investigation of biofilm rheology. Biotechnol. Bioeng. 65, 83–92 (1999).
60
T. Shaw, M. Winston, C. J. Rupp, I. Klapper, P. Stoodley, Commonality of elastic relaxation times in biofilms (2004).
61
H. H. Tseng, J. C. Wu, Y. C. Lin, G. L. Zhuang, Superoleophilic and superhydrophobic carbon membranes for high quantity and quality separation of trace water-in-oil emulsions. J. Memb. Sci. 559, 148–158 (2018).
62
C. N. Trumbore, Shear-induced amyloid formation in the brain: III. The roles of shear energy and seeding in a proposed shear model. J. Alzheimer’s Dis. 65, 47–70 (2018).
63
F. Goirand, T. Le Borgne, S. Lorthois, Network-driven anomalous transport is a fundamental component of brain microvascular dysfunction. Nat. Commun. 12, 1–11 (2021).
64
S. H. Lee et al., Effect of permeate flux and backwashing on quorum sensing and quenching within biocake layer in membrane bioreactor. J. Environ. Eng. 146, 04020025 (2020).
65
D. Seifert, P. Engesgaard, Sand box experiments with bioclogging of porous media: Hydraulic conductivity reductions. J. Contam. Hydrol. 136–137, 1–9 (2012).
66
A. Brovelli, F. Malaguerra, D. A. Barry, Bioclogging in porous media: Model development and sensitivity to initial conditions. Environ. Modell. Softw. 24, 611–626 (2009).
67
S. Lecuyer et al., Shear stress increases the residence time of adhesion of Pseudomonas aeruginosa. Biophys. J. 100, 341–350 (2011).
68
J. M. Meinders, H. C. Van Der Mei, H. J. Busscher, Physicochemical aspects of deposition of streptococcus thermophilus B to hydrophobic and hydrophilic substrata in a parallel plate flow chamber. J. Colloid Interface Sci. 164, 355–363 (1994).
69
L. M. Nilsson, W. E. Thomas, E. V. Sokurenko, V. Vogel, Elevated shear stress protects Escherichia coli cells adhering to surfaces via catch bonds from detachment by soluble inhibitors. Appl. Environ. Microbiol. 72, 3005 (2006).
70
Q. Zhang et al., Morphogenesis and cell ordering in confined bacterial biofilms. Proc. Natl. Acad. Sci. U.S.A. 118, 2021 (2021).
71
P. Pearce et al., Flow-induced symmetry breaking in growing bacterial biofilms, 7126 (2019).
72
S. Das, A. Kumar, Formation and post-formation dynamics of bacterial biofilm streamers as highly viscous liquid jets. Sci. Rep. 4, 7126 (2014).
73
M. Samari Kermani, S. Jafari, M. Rahnama, A. Raoof, Direct pore scale numerical simulation of colloid transport and retention. Part I: Fluid flow velocity, colloid size, and pore structure effects. Adv. Water Res. 144, 103694 (2020).
74
W. Deng, M. B. Cardenas, M. F. Kirk, S. J. Altman, P. C. Bennett, Effect of permeable biofilm on micro- and macro-scale flow and transport in bioclogged pores. Environ. Sci. Technol. 47, 11092–11098 (2013).
75
M. Carrel et al., Biofilms in 3D porous media: Delineating the influence of the pore network geometry, flow and mass transfer on biofilm development. Water Res. 134, 280–291 (2018).
76
S. H. Lee, P. K. Kang, Three-dimensional vortex-induced reaction hot spots at flow intersections. Phys. Rev. Lett. 124, 144501 (2020).
77
C. D. Nadell, D. Ricaurte, J. Yan, K. Drescher, B. L. Bassler, Flow environment and matrix structure interact to determine spatial competition in Pseudomonas aeruginosa biofilms. eLife 6, e21855 (2017).
78
B. E. Rittmann, The significance of biofilms in porous media. Water Resour. Res. 29, 2195–2202 (1993).
79
G. Savorana, J. Słomka, R. Stocker, R. Rusconi, E. Secchi, A microfluidic platform for characterizing the structure and rheology of biofilm streamers. Soft. Matter 18, 3878–3890 (2022).
80
E. Secchi et al., The structural role of bacterial eDNA in the formation of biofilm streamers. Proc. Natl. Acad. Sci. U.S.A. 119, 1–10 (2022).
81
T. L. van Noorden, I. S. Pop, A. Ebigbo, R. Helmig, An upscaled model for biofilm growth in a thin strip. Water Resour. Res. 46, 6505 (2010).
82
M. Peszynska, A. Trykozko, G. Iltis, S. Schlueter, D. Wildenschild, Biofilm growth in porous media: Experiments, computational modeling at the porescale, and upscaling. Adv. Water Resour. 95, 288–301 (2016).
83
A. J. C. Ladd, P. Szymczak, Reactive flows in porous media: Challenges in theoretical and numerical methods. 12, 543–571 (2021). https://doi.org/10.1146/annurev-chembioeng-092920-102703.
84
M. Thullner, Comparison of bioclogging effects in saturated porous media within one- and two-dimensional flow systems. Ecol. Eng. 36, 176–196 (2010).
85
D. L. Kurz et al., Competition between growth and shear stress drives intermittency in preferential flow paths in porous medium biofilms. Proc. Natl. Acad. Sci. U.S.A. 119, e2122202119 (2022).
86
M. Whiteley, S. P. Diggle, E. P. Greenberg, Progress in and promise of bacterial quorum sensing research. Nature 551, 313–320 (2017).
87
O. X. Cordero, M. S. Datta, Microbial interactions and community assembly at microscales. Curr. Opin. Microbiol. 31, 227–234 (2016).
88
L. Fazzino, J. Anisman, J. M. Chacón, R. H. Heineman, W. R. Harcombe, Lytic bacteriophage have diverse indirect effects in a synthetic cross-feeding community. ISME J. 14, 123–134 (2020).
89
N. Klitgord, D. Segrè, Environments that induce synthetic microbial ecosystems. PLOS Comput. Biol. 6, e1001002 (2010).
90
W. R. Harcombe, J. M. Chacón, E. M. Adamowicz, L. M. Chubiz, C. J. Marx, Evolution of bidirectional costly mutualism from byproduct consumption. Proc. Natl. Acad. Sci. U.S.A. 115, 12000–12004 (2018).
91
L. J. Perez, T. Bhattacharjee, S. S. Datta, R. Parashar, N. L. Sund, Impact of confined geometries on hopping and trapping of motile bacteria in porous media. Phys. Rev. E. 103, 012611 (2021).
92
T. Bhattacharjee, S. S. Datta, Bacterial hopping and trapping in porous media. Nat. Commun. 10, 2075 (2019).
93
R. Hartmann et al., Quantitative image analysis of microbial communities with BiofilmQ. Nat. Microbiol. 6, 151–156 (2021).

Information & Authors

Information

Published in

Go to Proceedings of the National Academy of Sciences
Proceedings of the National Academy of Sciences
Vol. 120 | No. 14
April 4, 2023
PubMed: 36989304

Classifications

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.

Submission history

Received: March 14, 2022
Accepted: February 28, 2023
Published online: March 29, 2023
Published in issue: April 4, 2023

Keywords

  1. bioaggregate
  2. critical shear stress
  3. pore-throat flows
  4. microfluidics
  5. clogging

Acknowledgments

This work was supported by MnDRIVE Advancing Industry, Conserving Our Environment at the University of Minnesota, and preparation of PDMS microfluidic devices were conducted in the Minnesota Nano Center, which is supported by the NSF through the National Nano Coordinated Infrastructure Network (NNCI) under Award Number ECCS-2025124. We would like to thank Prof. William Harcombe and Dr. Jeremy Chacon in the College of Biological Sciences at the University of Minnesota for kindly donating the E. coli and M. extorquens strains.
Author Contributions
S.H.L. and P.K.K. designed research; S.H.L. performed research; S.H.L., E.S., and P.K.K. analyzed data; and S.H.L., E.S., and P.K.K. wrote the paper.
Competing Interests
The authors declare no competing interest.

Notes

This article is a PNAS Direct Submission.

Authors

Affiliations

Department of Earth and Environmental Sciences, University of Minnesota, Minneapolis, MN 55455
Institute of Environmental Engineering, ETH Zürich, Zürich 8093, Switzerland
Department of Earth and Environmental Sciences, University of Minnesota, Minneapolis, MN 55455
Saint Anthony Falls Laboratory, University of Minnesota, Minneapolis, MN 55455

Notes

1
To whom correspondence may be addressed. Email: [email protected].

Metrics & Citations

Metrics

Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.


Citation statements




Altmetrics

Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

View Options

View options

PDF format

Download this article as a PDF file

DOWNLOAD PDF

Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Personal login Institutional Login

Recommend to a librarian

Recommend PNAS to a Librarian

Purchase options

Purchase this article to access the full text.

Single Article Purchase

Rapid formation of bioaggregates and morphology transition to biofilm streamers induced by pore-throat flows
Proceedings of the National Academy of Sciences
  • Vol. 120
  • No. 14

Media

Figures

Tables

Other

Share

Share

Share article link

Share on social media