3.1 Structural differences in OCT-imaged biofilms
Four growth conditions were applied to biofilms that differed in volumetric flow rate \(Q\) and the Fe2+ concentration\(c\) (see Table 1). The aim of the study was the identification of a dependency of biofilm development and structure in terms of inflowing Fe2+ as well as of different flow regimes. Figure 1 illustrates height maps showing the topography of the developed biofilms at day 10 for each flow cell under each applied condition. Images of all flow cells on each day are given in Supplementary Information Figure 1.
Figure 1 reveals biofilm growth in all flow cells. To be noticed, FC 9 of E4 had a different appearance among the other flow cells for these experimental conditions. The same is visible from E3 in FC 5 among FCs 8 and 10. Experiment E1 (\(c\) = 2.5 mg/L Fe2+, \(u\) = 0.75 cm/s) depicts biofilms with a mean biofilm thickness\({\overset{\overline{}}{L}}_{F}\) = 75 µm in all replicates. Additionally, under condition E1, biofilm aggregates of \(L_{F}\) up to 200 µm established and the substratum was covered with heterogeneous aggregates, ranging from smaller and spherical colonies to longer (> 2 mm) and elliptic colonies. In experiment E2 (\(c\) = 0.25 mg/L Fe2+, \(u\) = 0.75 cm/s) less biofilm developed and merely FC 1, FC 2 and FC 5 to 7 are covered with biofilm (\({\overset{\overline{}}{L}}_{F}\) = 50 µm). In comparison to E1, biofilm growth within E2 started delayed and only small colonies of minor widths and lengths, as well as of minor heights accumulated. In E3 (\(c\) = 2.5 mg/L Fe2+, \(u\) = 3.75 cm/s) in turn, biofilm growth seemed to be more dense at day 10 of the growth course. Likewise E2, bacterial colonization was first visible around days 3 and 4. Here, accretion of the flow chamber took place from the walls to the center of the FC, probably due to the higher flow velocity. In E4 (\(c\)= 0.25 mg/L Fe2+, \(u\) = 3.75 cm/s), coverage of the substratum was again low as in E2 and only a few colonies randomly developed with high biofilm thicknesses of\({\overset{\overline{}}{L}}_{F}\) = 200 µm.
Figure 1 already provides visible differences in growth patterns for different experimental conditions. However, for quantification of the influence of flow velocity (shear stress) and iron (Fe2+) dosage on biofilm structure and development several structural parameters have been calculated (confer Materials & Methods). Those are presented in Figure 2 and discussed in the following.
In Figure 2, the development of 10 different structural biofilm parameters for all conditions E1-E4 are illustrated. As already visible in Figure 1, the effect of the high (E1 + E3, \(c\) = 2.5 mg/L Fe2+) and the low iron(II) concentration (E2 + E4,\(c\) = 2.5 mg/L Fe2+) is visible. While mean biofilm thickness \({\overset{\overline{}}{L}}_{F}\) and substratum coverageSC of E1 and E3 show a steady increase until the end of the experiment, those parameters stay at a minimum level in E2 and E4. A similar trend is distinct for the parameters textural entropyTE, average horizontal AHRL and average vertical run length AVRL. Additionally, stated parameters of E3 (\(c\)= 2.5 mg/L Fe2+, \(u\) = 3.75 cm/s) exceed those of E1 (\(c\) = 2.5 mg/L Fe2+, \(u\) = 0.75 cm/s) at the end of the experiment (from day 8 to day 10). The same trend is shown for biofilms grown with \(c\) = 0.25 mg/L Fe2+, whereas mentioned structural parameters in E4 (\(u\) = 3.75 cm/s) exceed those of E2 (\(u\) = 0.75 cm/s). While mean biofilm thickness\({\overset{\overline{}}{L}}_{F}\ \)and substratum coverageSC indicate an early accumulation of biofilms with high biofilm volume in E1 and E3 (\(c\) = 2.5 mg/L Fe2+), higher values of textural entropy TE exhibit more heterogeneous biofilms. Thus, biofilms grown with 2.5 mg/L Fe2+ show more differentiated and partially distributed structures. Furthermore, average run lengths AHRLand AVRL of biofilms in E1 and E3 display longer and wider biofilm aggregates resulting in the largest aggregates regarding biofilm volume in E3 (\(c\) = 2.5 mg/L Fe2+, \(u\) = 3.75 cm/s) (compare Figure 1).
As seen from the structural parameters skewness \(R_{\text{SK}}\) and kurtosis \(R_{\text{KU}}\), biofilms in E4 (\(c\) = 0.25 mg/L Fe2+, \(u\) = 3.75 cm/s) displayed random distributed and high biofilm hills with a low substratum coverage SC at the beginning of the experiment. From day 5 to the end of the experiment (day 10), values of \(R_{\text{SK}}\) and \(R_{\text{KU}}\) under all conditions approximate and stay near zero. Thereby, a homogeneous distribution of biofilm as well as an equalized ratio of colonies with either low or high biofilm thickness \({\overset{\overline{}}{L}}_{F}\), except for E2 (\(c\) = 0.25 mg/L Fe2+, \(u\) = 0.75 cm/s) where only several high and small biofilm aggregates occurred, is confirmed (compare Figure 1). Again, structural biofilm parameters fractal dimension FD, angular second moment ASMand inverse difference moment IDM prove a lucid differentiation between biofilms grown with \(c\) = 0.25 mg/L Fe2+ and \(c\) = 2.5 mg/L Fe2+. Higher values of FD up to 1.8 in E1 and E3 (\(c\) = 2.5 mg/L Fe2+) explain more irregular surfaces of the aggregates (compare Table 2). Here, the parameter ASMdescribes a change in growth direction (additional growth in y-direction (width)) and IDM proves additional growth in width, since distances between cell clusters are minimizing to the end of the experiment, compared to E2 and E4 (\(c\) = 0.25 mg/L Fe2+).
As can be seen, biofilm thickness \({\overset{\overline{}}{L}}_{F}\)and/or surface coverage SC, which are typically used for characterization of biofilm structure, do not provide the complete information on biofilm development. Therefore, only the joint consideration of all evaluated structural parameters leads to a complete overview of biofilm development over the cultivation period of 10 days. Figure 2 shows, that at the beginning, biofilms cultivated at \(u\) = 3.75 cm/s (E3, E4) developed more slowly due to the higher shear stress. However, towards the end of the experiment (days 8 to 9) those biofilms became more stable in terms of \({\overset{\overline{}}{L}}_{F}\),SC, ASM and IDM in relation to biofilms in E1 and E2 (\(u\) = 0.75 cm/s). With regard to the different iron(II) concentrations, more unambiguous differences in mostly all structural parameters could be identified (correlation of E1 + E3 and E2 + E4). The results confirm the positive influence of Fe2+ on biofilm accumulation and differentiation. Likewise (Körstgens, Flemming, Wingender, & Borchard, 2001) showed, that bivalent cations (e.g., Ca2+) enhance the stability of the biofilm matrix in Pseudomonas aeruginosabiofilms which could explain the increased adhesion and biofilm accumulation in E1 and E3 (\(c\) = 2.5 mg/L Fe2+) of the presented study. Additionally, in the study by (Möhle et al., 2007), increased concentrations of iron sulfate in the nutrient medium (\(c\) = 10 mg/L) prevented sloughing of microbial biofilms (activated sludge) in a rotating disk reactor. Furthermore, the authors documented the dependence of the biofilm thickness from the substrate concentration\(c\) as well as from the shear stress on the biofilm surface. Beyond, in studies with iron complexing agents it was found that a minimal concentration of soluble iron is necessary for the formation of P. aeruginosa biofilms in flow cells (Renslow, Lewandowski, & Beyenal, 2011; Singh et al., 2002; L. Yang et al., 2007). Thereby, one theory is, that iron regulates the surface motility of the bacteria and again promotes the biofilm formation by stabilizing the EPS matrix, which mainly consists of negatively charged polymers (Berlutti et al., 2005; Lin et al., 2012; Singh, 2004; Weinberg, 2004).
These studies further show, that an excess of iron concentrations inhibits biofilm formation, too, since the release of DNA from deadP. aeruginosa cells is suppressed. This release is an important structural component of biofilms (Lin et al., 2012; L. Yang et al., 2007). However, an ideal iron concentration in culture medium cannot easily be determined. While (Berlutti et al., 2005) define “high” iron concentrations in a range of 0.55 – 5.5 mg/L as positive in terms of aggregation and manipulation of biofilm development and structure in different reactor systems and tests, (L. Yang et al., 2007) reported an inhibition of biofilm growth in microtiter plates and flow cells in this concentration range. In the present study, this inhibition by high iron concentrations cannot be proven. Presumably, an addition of iron (Fe2+) does not stimulate every bacterial biofilm system or possibly optimum iron amounts can vary among different biofilm species. Nevertheless, (Weinberg, 2004) confirmed that zinc, manganese and iron have key functions in biochemical as well as in morphological conversion of pro- and eucaryotes, respectively. Since soil carries high amounts of iron, a positive influence on growth of the used soil bacterium Bacillus subtilis could be demonstrated in the present study (Kolodkin-Gal et al., 2013; Pelchovich, Omer-Bendori, & Gophna, 2013; Rizzi, Roy, Bellenger, & Beauregard, 2019).
Likewise, the influence of hydrodynamics on biofilms is well-known and documented in several studies (Manz, Volke, Goll, & Horn, 2003; Park et al., 2011; Paul et al., 2012; Purevdorj, Costerton, & Stoodley, 2002; P. Stoodley et al., 1999; P Stoodley et al., 2002; Teodósio, Simões, Melo, & Mergulhão, 2011; Weiss et al., 2016). These studies verify the formation of streamers at higher flow velocities meaning increased growth in length, which is best visible in E3 for FC 8 and E4 for FC 9 (\(u\) = 3.75 cm/s, see Figure 1 and Figure 2 AHRL). Statements about the viscoelastic properties and strength of the cell clusters characterize the positive influence of the hydrodynamics furthermore (Allen et al., 2018; Peterson et al., 2015; Rupp, Fux, & Stoodley, 2005; Safari, Tukovic, Walter, Casey, & Ivankovic, 2015; P Stoodley et al., 2002; Paul Stoodley, Lewandowski, Boyle, & Lappin-scott, 1999; Towler, Rupp, Cunningham, & Stoodley, 2003).