ABSTRACT
Biofilm formation renders sessile microbial populations growing in continuous-flow systems less susceptible to variation in dilution rate than planktonic cells, where dilution rates exceeding an organism's maximum growth rate (μmax) results in planktonic cell washout. In biofilm-dominated systems, the biofilm's overall μmax may therefore be more relevant than the organism's μmax, where the biofilm μmax is considered as a net process dependent on the adsorption rate, growth rate, and removal rate of cells within the biofilm. Together with lag (acclimation) time, the biofilm's overall μmax is important wherever biofilm growth is a dominant form, from clinical settings, where the aim is to prevent transition from lag to exponential growth, to industrial bioreactors, where the aim is to shorten the lag and rapidly reach maximum activity. The purpose of this study was to measure CO2 production as an indicator of biofilm activity to determine the effect of nutrient type and concentration and of the origin of the inoculum on the length of the lag phase, biofilm μmax, and steady-state metabolic activity of Pseudomonas aeruginosa PA01 (containing gfp), Pseudomonas fluorescens CT07 (containing gfp), and a mixed community. As expected, for different microorganisms the lengths of the lag phase in biofilm development and the biofilm μmax values differ, whereas different nutrient concentrations result in differences in the lengths of lag phase and steady-state values but not in biofilm μmax rates. The data further showed that inocula from different phenotypic origins give rise to lag time of different lengths and that this influence persists for a number of generations after inoculation.
Microbial growth in batch cultures has been studied for a long time, and the observed phases have been designated the lag phase, the acceleration phase, the exponential phase, the retardation phase, the stationary phase, and the phase of decline although not each culture displays all of the mentioned phases (16). In contrast to batch cultures and static (no flow) biofilms (e.g., those that form in 96-well plates), the increase in biofilm cells in a flowing environment is a net process that is dependent on the irreversible adsorption rate of cells to the surface, the growth rate of the microorganisms, and the removal rate of cells lost to the bulk flow (18). There are numerous benefits for the cells in biofilms, e.g., protection against antimicrobials and the opportunity for and proliferation by continuous cell dispersion. There is also a possible competitive advantage if cells colonize surfaces at multiple sites and grow in such a manner that the resulting three-dimensional architecture exposes the maximum biofilm surface area to surrounding nutrients. The most successful colonizers would therefore be the cells with the ability to adhere to the surface (and stay adhered) and to start multiplying at maximum rate. The process of events from being free-floating cells to the so-called permanently surface-attached phase involves early steps including reversible attachment and a phenotypic change in the cells from a planktonic state to a sessile state, with concomitant changes in gene expression; these steps contribute to a lag phase that will occur before maximal growth/biofilm development can take place (23). Clearly, the ability to progress from the lag phase to a fast-growing phase, as well as the duration of the lag phase, is an important determinant of biofilm function and has an impact in a diverse range of environments, often with implications for infection or contamination control, as well as in industrial processes.
At the cell level, an extended lag phase and slower growth create the risk that the cells will be displaced by faster-growing microcolonies, as was demonstrated by Klayman et al. (13) in dual species biofilms. A microorganism's competence in dominating a surface area can therefore be evaluated by comparing the lag phases and maximal growth rates (μmax) of a biofilm growth curve. Knowing a bacterial population's specific growth rate is a requirement for its cultivation at optimum rates in a chemostat or other continuously fed bioreactor. A key assumption for this type of cultivation is that wall growth has a negligible effect, which is in stark contrast to systems where surface-associated growth dominates. Indeed, while dilution rates exceeding an organism's μmax results in cell washout in a conventional chemostat setting, biofilm formation enables microbial populations to persist at dilution rates much higher than the organism's μmax.
Biofilm growth rates have been determined by various techniques, such as fluorescence in situ hybridization (FISH) (12, 31), measuring the incorporation of radioactive substances like [3H]thymidine and 32P (8, 9), microscopy (3, 13, 17), measuring the total increase in biofilm mass (both cells and extracellular polymeric substances) (24, 28), colorimetric 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-[(phenyl-amino)carbonyl]-2H- tetrazolium hydroxide (XTT) assays (26), or measurement of amide II bands, as determined by attenuated total reflectance-Fourier transform infrared spectroscopy (6). Some of the above-mentioned techniques suffer the drawback that the biofilms have to be sacrificed with sampling or that the measured increases do not distinguish between live and dead matter in the biofilm (i.e., increases measured might not represent an accurate increase in viable cell numbers).
In this study a carbon dioxide evolution measurement system (CEMS) (15) was used to track the biofilm development rate in real time. The advantage of using this system is that the measured rates represent the metabolic activity of the active cell mass and can be done nondestructively for any biofilm-forming microorganism. In the past, measurement of oxygen uptake rates has been used for determination of growth rates in batch cultures (19) and of the localized growth rate in biofilms (32). CO2 measurements by a gas chromatograph have been used to determine growth rates in batch systems (4), but to our knowledge this is the first time that CO2 measurements have been used to determine whole-biofilm specific growth rates. We applied this technique to compare the biofilm μmax values for two well-described pseudomonads and a mixed microbial community when the organisms are grown on different nutrients and to test the premise that the origin of the inoculum has an impact on early biofilm development.
MATERIALS AND METHODS
Strains and culture conditions.An environmental isolate, Pseudomonas fluorescens CT07 (in this study, containing the gfp gene; GenBank accession number DQ777633) (30) was used for the majority of the experiments. Pseudomonas aeruginosa PA01 (in this study, containing the gfp gene) was used for comparison and validation of the approach. Plasmid construction and the protocol followed were described by Bester et al. (1). The two strains were maintained on defined medium agar plates with 5 mM citrate as the carbon source. A mixed microbial community was obtained from the drain of a sink in a public washroom. This community was maintained in 3 g/liter tryptic soy broth (TSB) (EMD Chemicals) at room temperature. In related research (11) it was shown that multidrug-resistant Pseudomonas biofilms persist in drain sinks and demonstrated in the case of hospital intensive care units that poor facility design led to the dissemination of the bacteria from these drain biofilms during hand washing, causing nosocomial infections and fatalities. We further demonstrated that P. aeruginosa PA01 effectively integrates into these biofilms (M. Ghadakpour et al., unpublished results), and we were therefore interested in the current study to compare biofilm development rates of the communities with those of the test strains. Routine cultivation of the organisms was carried out on a modified AB medium with final concentrations of 1.51 mM (NH4)2SO4, 3.37 mM Na2HPO4, 2.20 mM KH2PO4, 179 mM NaCl, 0.1 mM MgCl2·6H2O, 0.01 mM CaCl2·2H2O, and 0.001 mM FeCl3 and with different concentrations of sodium citrate as the sole carbon source (5). TSB was used as growth medium in various concentrations, as indicated below. Precultures from the stationary growth phase (grown overnight) or the exponential growth phase (confirmed by the optical density at 600 nm [OD600]) measurements) used for inoculation of the biofilm reactors were incubated in either defined medium with 5 mM citrate or 3 g/liter TSB at 32 ± 3°C and shaking at 250 rpm. The biofilm effluent cells used for inoculation were harvested from 1-week-old biofilms grown at room temperature in 50-cm-long silicone tube (inside diameter, 0.16 cm) reactors fed with 1 mM citrate minimal medium at 15 ml/h. The cell numbers for each inoculation were determined by spread plate counting.
Biofilm cultivation.Biofilms were cultivated in a continuous-flow reactor system with growth medium supplied at 15 ml/h by a Watson Marlow 205U peristaltic pump with a corresponding residence time of 12 min, given the following reactor dimensions: inside diameter, 0.16 cm; outside diameter, 0.24 cm; length 150 cm. The reactor was a CEMS (15) that allows the continuous online measurement of CO2 in the gaseous phase. It was shown that the amount CO2 measured in the gas phase has a direct correlation with the amount of dissolved CO2 in the liquid phase (at the physiological CO2 concentrations under investigation) and can therefore be used as a reliable indication of active cell mass in the reactor during the early stages of biofilm development (15). In essence, the reactor consists of a CO2- and O2-permeable silicone tube housed in an outer shell of gas-impermeable Tygon tubing. CO2 produced by the biofilm growing on the inside of the silicone tube crosses the silicone tube wall into the annular space between the Tygon and silicone tubes. The annular space is connected to an absolute, nondispersive, infrared LI-820 CO2 gas analyzer (Li-Cor Biosciences, NE), and compressed CO2-free air (total organic carbon [TOC] grade; CO2 of < 0.5 ppm, CO of < 0.5 ppm, O2 at 20 to 22%, and total hydrocarbons [THC] of < 0.1 ppm; Linde, Canada) was used as the sweeper gas at rates that varied from 1.5 to 2.3 liters/h. Gas flow rates were determined by volumetric displacement. The CEMS was submerged in a water bath that was kept at 27 °C for all experiments. The CEMS was inoculated with 200 μl of preculture under no-flow conditions. The cells were allowed to attach to the reactor tube wall for 30 min, after which growth medium flow was resumed.
Growth rate measurements in CEMS.Since during the early stages of biofilm development the environmental conditions in the reactor are not changing (with regard to pH, nutrient concentration, and temperature), the CO2 production rate can be used as an indication of active cell numbers at a specific time (in contrast with respiration rates of a fixed number of cells that may vary due to environmental changes). A biofilm growth rate (μbiofilm) was obtained by determining the slope of the natural logarithm of the CO2 production rate plotted against time: μbiofilm = (lnXt2 − lnXt1)/(t2 − t1), where Xt2 is the CO2 production at time t2, and Xt1 is the CO2 production rate at time t1. Often the biofilm exponential growth phase did not occur as one smooth continuous line (as found in batch cultures), possibly due to sloughing events, and so different time intervals (t2 − t1) of 1, 2, and 4 h were used for comparison purposes. For all subsequent discussion, μbiofilm is considered the net growth rate comprised of adsorption, cellular growth, and loss of active biomass to the bulk flow via sloughing and erosion processes. The length of the lag phase was determined as the time where the tangent to the maximum slope intersected with the x axis.
Correlating cell numbers to protein concentration and CO2 production.In order to verify the validity of CO2 measurements to quantify biofilm growth rate, protein concentrations as an indication of increase in biofilm biomass (22) were measured in conjunction with the metabolic activity of the biofilm.
Cell lysis and protein concentration determination.Cells were lysed in 0.1 N NaOH (final concentration) with incubation at 70°C for 1 h. The solubilized proteins from the lysed cells were measured with a Pierce bicinchoninic acid (BCA) total protein determination kit according to the manufacturer's instructions.
Protein concentration as a measure of cell numbers.Different volumes of an overnight culture of P.fluorescens CT07 were centrifuged at 12,000 × g for 5 min at 4°C. The supernatant was discarded, the pelleted cells were lysed in 0.1 N NaOH with incubation at 70°C for 1 h, and the protein concentration was determined. The protein concentration showed a high correlation with cell numbers (R2 of 0.996).
Protein concentration as an indication of cell numbers in biofilms.Twelve silicone tubes of the same dimensions as the inner tube of the CEMS were inoculated with P. fluorescens CT07 from cells in the exponential growth phase, and cells were grown at room temperature under continuous-flow conditions (15 ml/h) with growth medium containing 1 mM citrate. The biofilms growing in these tubes were sacrificed in duplicate, for a total of six protein concentration data points, by first draining the entire volume of bulk fluid (∼3 ml) into a receptacle and adding NaOH (for a final concentration of 0.1 N) to this liquid containing free-floating cells and loosely associated biofilm cells. Three milliliters of prewarmed (60°C) 0.1 N NaOH was injected into the emptied silicone tube, and the tube was rinsed by pipetting the liquid back and forth to loosen and dissolve the attached biofilm. Subsequently, the ends of the silicone tube were sealed off, and then the tubes were incubated (with the 0.1 N NaOH still inside) at 70°C, similar to the procedure for the drained bulk fluid. After 1 h of incubation, the contents of the silicone tube (removed and dissolved biofilm) were further rinsed and stored as fraction 1 at −20°C until the protein concentration could be determined. The rinsing step (with 0.1 N NaOH) was done a second time, and after incubation at 70°C this sample was stored as fraction 2. Two CEMSs were inoculated at the same time as the 12 silicone tubes with the same inocula and grown under the same conditions to compare CO2 production with biofilm cell numbers.
Evaluation of the effect of various parameters on early biofilm development.The experimental system and conditions described above were used to compare (i) the two test strains and the mixed community, (ii) the effect of nutrient concentrations, and (iii) the effect of the origin of inoculum on early biofilm development.
RESULTS
Correlation of biofilm growth determined by protein and CO2 measurement.Biofilm development as determined by protein measurement was compared with the results obtained by CEMS for CO2 production (Fig. 1). The duration of the lag phase as determined by CO2 measurements and protein determination were 22 h and 24 h, respectively, with biofilm μmax values of 0.27 ± 0.01 h−1 and 0.25 ± 0.04 h−1, respectively.
Comparison of P. fluorescens CT07 biofilm growth during the exponential phase as determined by protein and CO2 measurements.
Protein concentrations for the sample that was collected earlier than the first data point shown were below the detection limit of the assay used. Biofilm growth was measured with the CEMS, and rates were compared in terms of lag phase, maximum growth rate, and the steady-state plateau phase, in which the increase in the CO2 production rate had leveled off.
Different microorganisms have different lag phase lengths and maximum biofilm growth rates.As expected, under identical growth conditions (27°C with a growth medium of 1 mM citrate minimal medium), different organisms exhibited differences in lag phases. As shown in Fig. 2, P. aeruginosa PA01 had a shorter lag phase than P. fluorescens CT07, which indicates an enhanced ability to irreversibly attach to the surface and convert to a biofilm phenotype, despite the fact that P. aeruginosa PA01 was growing at suboptimal temperatures while the growth temperature was within the optimum range for P. fluorescens CT07 (25 to 30°C as reported in the literature). The maximum biofilm growth rates also differed between P. aeruginosa PA01 and P. fluorescens CT07. PA01 had an average maximum biofilm growth rate of 0.37 ± 0.01 h−1 while CT07 had an average maximum growth rate of 0.26 ± 0.03 h−1 (both growth rates taken with a time interval of t2 − t1 = 2 h), rates which differ significantly from one another (P < 0.05). The same average steady-state values were obtained for the two Pseudomonas species, with relatively minor sloughing events even ∼50 h after inoculation.
Growth curves of P. aeruginosa PA01 (solid line) and P. fluorescens CT07 (dashed line) biofilms inoculated from citrate medium overnight precultures and grown on 1 mM citrate.
When the behaviors of the mixed culture and P. fluorescens CT07 grown on 1.5 g/liter TSB at 27°C were compared, the mixed culture had a much shorter lag phase than CT07 (Fig. 3). Interestingly, the comparison with growth on TSB showed similar maximum growth rates (0.39 ± 0.05 h−1 and 0.39 ± 0.04 h−1 for mixed cultures and P. fluorescens CT07, respectively) and steady-state values. However, the sloughing events occurring between 40 and 50 h of incubation were much more pronounced than in the biofilms grown on 1 mM citrate minimal medium.
Growth curves of mixed culture (solid line) and P. fluorescens CT07 (dashed line) biofilms inoculated from TSB medium overnight precultures and grown on 1.5 g/liter TSB.
Different nutrient concentrations result in different lengths of lag phase and steady-state values but not in differences in maximum growth rates.When P. fluorescens CT07 was grown in minimal growth medium with 1, 2, and 4 mM citrate as the sole carbon source, the average lag phases decreased with increasing citrate concentrations. The average maximum growth rates increased with citrate concentration (0.26 ± 0.03 h−1, 0.26 ± 0.03 h−1, and 0.31 ± 0.01 h−1 for 1, 2, and 4 mM citrate, respectively), but these changes were not significantly different. The steady-state values and intensity of sloughing events increased with increasing citrate concentrations. Remarkable recovery was observed after these sloughing events (Fig. 4). P. fluorescens CT07 grown on 0.75 and 1.5 g/liter TSB showed similar trends, with lag phases decreasing with increasing TSB concentrations, similar maximum growth rates (0.39 ± 0.04 h−1 and 0.34 ± 0.03 h−1 for 0.75 and 1.5 g/liter TSB, respectively), and increasing steady-state values and sloughing with increasing TSB concentrations (data not shown).
Growth curves of P. fluorescens CT07 biofilms inoculated from citrate medium overnight precultures and grown on different concentrations of citrate medium. Notice that steady-state values and the severity of sloughing increase with increasing citrate concentrations. Inset A shows replicate runs of biofilms grown on 4 mM citrate, and inset B shows replicate runs of biofilms grown on 1 mM citrate to illustrate the differences and reproducibility.
Different origins of inocula result in different lag phases but not in different maximum growth rates.When biofilm effluent cells were used as the inoculum, the average lag phase was considerably shorter than when overnight cultures (from batch) were used as the inoculum (Fig. 5). The maximum growth rates were not significantly different when the slopes were calculated over a period of 2 h (0.31 ± 0.01 h−1 for effluent inoculum and 0.26 ± 0.03 h−1 for overnight inoculum), but they were significantly different (P < 0.05) when the slopes were calculated over 4 h (0.30 ± 0.01 h−1 for effluent inoculum and 0.24 ± 0.02 h−1 for overnight inoculum). The steady states stabilized at similar values. When batch-grown exponential phase cells were used as the inoculum and compared to inocula from overnight cultures, similar lag phase durations, maximum growth rates, and steady-state values were obtained. Interestingly, we repeatedly observed that the growth of biofilms inoculated from exponentially growing cells exhibited almost identical behaviors for replicate runs (Fig. 6). To further evaluate the effect that the origin of inoculum may have on subsequent biofilm development, P. fluorescens CT07 biofilms were grown with 1.5 g/liter TSB as growth medium, but one set of precultures was grown in 3 g/liter TSB while the other set of precultures was grown in 5 mM citrate minimal medium. The average lag phase for the biofilms originating from TSB preculture was significantly shorter than that of the citrate preculture biofilms (Fig. 7). The maximum growth rates were almost the same (0.37 ± 0.04 h−1 and 0.39 ± 0.04 h−1) for the TSB preculture and the citrate preculture, while the biofilms from the TSB preculture reached much higher steady-state values than the biofilms from the citrate precultures.
Growth curves for P. fluorescens CT07 biofilms grown on 1 mM citrate. Solid line, cultures inoculated from biofilm effluent; dashed lines, cultures inoculated from overnight precultures.
Growth curves for P. fluorescens CT07 biofilms inoculated from exponentially growing cells and grown on 1 mM citrate. Notice the almost identical growth behavior for replicate runs inoculated from the same inoculum on each respective day.
Growth curves for P. fluorescens CT07 biofilms grown on 1.5 g/liter TSB with precultures from 3 g/liter TSB medium (solid line) and 5 mM citrate minimal medium (dashed line).
DISCUSSION
Measuring biofilm development with CEMS allowed determination of the length of the lag phase, maximum biofilm growth rate, and steady-state values with predictability as well as the monitoring of dynamic behavior such as biofilm sloughing events and recovery. A shorter lag phase and higher maximal growth rate are both indicators of the competitive advantage that an inoculum may have to colonize a new surface. The process that takes place when free-floating cells leave the bulk liquid to settle permanently on a surface encompasses the steps of reversible attachment and irreversible attachment (18, 23), coupled with the expression of a cascade of different genes in converting from a planktonic to a biofilm phenotype. All of these steps contribute to the lag phase that occurs before attached cells can start to reproduce maximally. Not all cells that attach irreversibly start to multiply at maximum rates. Rice et al. (23) observed that the first cells to attach irreversibly (what they referred to as the primary biofilm cells) and their progeny (the secondary biofilm cells) behaved in markedly different ways in that the latter replicated at a much higher rate and migrated more readily than the primary biofilm cells. Klayman et al. (13) also observed a base layer of cells uniformly colonizing the surface with a zero net accumulation of cells when they coinoculated two P. aeruginosa PA01 strains labeled with different fluorescent proteins into glass capillary flow cells.
A major benefit of faster growing cells is that they can displace nongrowing attached cells (13) while the development of a three-dimensional structure by these fast growers that protrude into the bulk liquid may result in faster metabolism because of improved exposure to nutrients and oxygen, as was the case where stalk formers displayed lower metabolic activity than the cap formers of mushroom-type microcolonies in P. aeruginosa biofilms (21). It appears probable that higher steady-state metabolic values are an indication of the stability of the biofilm structure, i.e., the ability of the biofilm cells overall to remain active at a maximum rate. Our results suggest that higher nutrient concentrations led to more biofilm biomass accumulation and a higher overall activity by the accumulated biomass. However, the larger biomass was also prone to more sloughing, which is likely a form of self-regulation in order to prevent stagnation due to the development of excessively large inactive zones in the biofilm. It is assumed here that the significant fluctuation in net metabolic activity as measured by CO2 production rates has a direct relationship to sloughing of the biofilm biomass; the bases for this assumption are material balances done in our previous studies, which showed that up to 10% of the total carbon input to P. fluorescens CT07 in a similar system over an 8-day period was released as aggregates, presumably following sloughing events (compared to 4% that remained as biofilm biomass) (15). Furthermore, using a large-area photometer to measure biofilm biomass, Bester et al. (3) showed highly variable values of biofilm biomass by this strain after the first sloughing event, the timing of which correlates with the results presented here. Different nutrient sources result in different biofilm architectures. For instance, P. aeruginosa developed three-dimensional microcolonies and mushroom-type shapes in 10% TSB (13), whereas flat, densely packed biofilms formed in 0.1 mM citrate minimal medium (10). In a similar way P. fluorescens CT07 displayed different biofilm morphologies when grown in 10% TBS (30) or 1 mM citrate minimal medium (2). It is conceivable that a three-dimensional architecture with more pronounced protrusion forming at higher nutrient concentrations would be more readily prone to sloughing. P. aeruginosa and P. fluorescens biofilms cultivated under almost any condition that allows growth (20), and it is therefore doubtful that settlement on a surface is primarily a safety mechanism for protection from harsh environments by these organisms. Biofilms can, however, act as agents of continuous proliferation by sending offspring (as single cells or clumped in small aggregated clusters) into the environment. Bester et al. (1) reported that Pseudomonas biofilms yielded a considerable number of cells to the effluent as early as 6 h after inoculation, and one of their major conclusions was that biofilm formation is a mechanism for proliferation in addition to the role in survival typically mentioned in the literature. Interestingly, for the experimental system used, the planktonic population contributed only ∼1% of the total CO2 (unpublished data). Telgmann et al. (29) concluded that sloughing and erosion should not be seen as disturbance events but, rather, as integral parts of biofilm development.
In a batch culture, repeatable growth behavior is obtained when a particular microorganism is grown under similar environmental conditions (nutrient composition, temperature, etc.); i.e., the environment dictates the outcome of the growth curve, as is also suggested by the maxim that men resemble the times more than they resemble their fathers. The latter would suggest that the decade when people are born (reflecting the prevailing political, economic, and cultural status of an era) and, therefore, their environment have a more pronounced influence on the generalized behavior of a generation than the people who raised them. With respect to microorganisms, this would be true for batch cultures, but results presented in this study indicate that this seems not to be the case with early biofilm formation (specifically with regard to lag phase and maximum specific growth rate) in P. fluorescens CT07. Rice et al. (23) pointed to a critical need for understanding the nature of the inoculum used to generate the initial biofilm. Our results show that the environment from which the cells originate prior to incorporation into a biofilm has a dramatic effect on the behavior (and potentially the architecture) of the ensuing biofilm. When the inoculum originated from biofilm effluent, the cells were able to adapt much more quickly to growth in a biofilm, which is in agreement with observations by Rollet et al. (25), who tested the biofilm-forming capabilities of cells from different origins, i.e., sessile cells, cells detached from biofilm, and batch-grown cells. Their observation that the cells detached from a biofilm (effluent cells in our context) exhibit a “greater capacity to form biofilms” led to the postulation that there exists a transitional phenotype (detached cells) between sessile and planktonic states. This observation provides support to the statement by Rice et al. (23) that the biofilm phenotype might already be expressed when the cells approach a new surface, which would account for the decreased length of the lag phase. Bester et al. (1) commented on the large number of freely swimming single cells in close association with microcolonies of P. fluorescens CT07 biofilms grown on citrate minimal medium. The freely swimming cells that appear to be minimally influenced by the much higher flow rates in the bulk liquid phase could be the origin of the mainly nonaggregated cells encountered in the effluent of both P. fluorescens CT07 and P. aeruginosa PA01 biofilms (1). This observation could, indeed, suggest that planktonic cells closely associated with the surface (or microcolonies in biofilms) have a biofilm-phenotypic advantage to settle on new surfaces. On the other hand, it could be that these surface-associated planktonic cells are starved for nutrients, which may influence their attachment behavior. Mueller (17) showed that starved P. fluorescens cells increased their ability to adsorb to a surface 4-fold compared to nonstarving cells. Effluent cells could also be in the form of small aggregates (27), with extracellular polymeric substances (EPS) aiding in surface attachment. The data show that although cells from different phenotypic origins do not constantly give rise to different biofilm specific growth rates, the lag phases differ markedly; given that the primary colonizers go through a number of cell divisions before there is a measurable change in overall biofilm activity, it appears that preculture conditions may exert an influence on the resulting biofilm for a number of generations after inoculation. Precultures grown in citrate medium showed extended lag phases and lower steady-state values than biofilms originating from precultures grown in TSB medium. Such differences in lag phases may reflect both the adaptation period and the resulting biofilm architecture. Kraigsley and Finkel (14) demonstrated a heritable ability of cells from older biofilms to outcompete younger cells in the presence of a preexisting biofilm.
For the different nutrient compositions and concentrations tested, no significant difference in specific growth rates could be found, which is in agreement with work from Ellis et al. (7), although higher concentrations led to higher biomass build-up and also more severe sloughing events. In the current work the ability of CEMS to use CO2 measurements as a way to determine biofilm growth curves was demonstrated in real time and in a nondestructive manner. Growth rates determined were comparable to values found in literature (13) and with good repeatability. When the observed differences in lag time are considered, together with the fact that biofilm growth in flow systems is a net process that balances irreversible attachment, cellular division, and loss of cells to the bulk phase via sloughing and erosion followed by colonization of new surfaces, it appears that the origin of the inoculum may have a profound effect on biofilm function. The approach described here should be applicable to efforts to model biofilm kinetics in bioreactors and provide a method to evaluate the effect of genetic modifications and external factors such as chemical treatment on biofilm function.
ACKNOWLEDGMENTS
Financial support was provided by the Canada Research Chair Program, NSERC, Bioshield Technologies, AFMNet, and a Ryerson Graduate Award (O.K.).
FOOTNOTES
- Received 11 January 2010.
- Accepted 9 July 2010.
- Copyright © 2010 American Society for Microbiology