Interdependence between bacterial EPS and early grain coat development

Bacteria are the most abundant forms of life we know on our planet, able to survive in a variety of habitats, that play an important role in mineral formation and transformation processes. Here, we present laboratory experiments in which unconsolidated quartz grains were seeded with Geobacter sulfurreducens cells and exposed to a mineral medium solution for 96 hours at temperatures of between 60°C and 120°C. Experimental data show the interdependence between extracellular polymeric substances (EPS) and the early formation of grain‐coating material. The occurrence of EPS promotes the development of web and bridging structures binding the quartz grains and creating EPS‐coated surfaces. With increasing temperature, an amorphous mineral phase grows preferentially on these surfaces suggesting that EPS can act as a template for mineral nucleation. At temperatures >100°C, the order of crystallinity of the amorphous authigenic phase increases, transitioning to poorly‐ordered rosette‐like textures.


| INTRODUC TI ON
Grain-coating minerals in sandstones have been demonstrated as effective barriers preventing quartz cementation. Such minerals result from the transformation of precursor phases (Aagaard et al., 2000;Charlaftis et al., 2021) formed during deposition or soon afterward (Bloch et al., 2002). Detrital clay aggregates, potential sources for diagenetic coats, have been found on sand grains in modern estuarine environments (Dowey et al., 2017;Virolle et al., 2019) where bio-sediment interaction is prominent.
Microbial activity at the sediment-water interface promotes biofilm development. Biofilm is defined as an assemblage of microbial cells of various species, such as bacteria and microbial eukaryotes, attached to a surface in a matrix consisting of extracellular polymeric substances (EPS). EPS comprise organic molecules, mainly polysaccharides (Stoodley et al., 2002), are structurally and functionally essential for biofilm growth and involved in microbial adhesion to surfaces (Costa et al., 2018).
In estuarine and lagoonal settings, most sediment is composed of non-cohesive, unconsolidated sand grains. Biofilms provide important ecosystem functions, including sediment stabilization (Paterson et al., 2018;Vignaga et al., 2013), transportation, and bedform stabilization Schindler et al., 2015). For example, diatom motility and adhesion in near-surface sediments are facilitated by sticky EPS strands bridging grain surfaces (Higgins et al., 2003). EPS contributes to sediment fabric by enhancing the cohesive forces between the sedimentary components (Van Colen et al., 2014).

Numerous studies document interactions between microbes
and clay minerals and their importance in geological processes (Cuadros, 2017;Dong, 2012;Douglas & Beveridge, 1998;Folk & Lynch, 1997). A two-step process proposed for biogenic clay authigenesis includes the initial encrustation of iron that surrounds cells or filaments followed by Al-Si complexion (Konhauser et al., 1993).
This initial cellularly-bound iron phase acts as a ligand that promotes the precipitation of an amorphous, poorly crystallized aluminosilicate phase engulfing the cell wall or EPS. Further mineralization results in more crystalline phases (Sánchez-Navas et al., 1998) tangentially oriented to the substratum (Konhauser & Urrutia, 1999). Ueshima and Tazaki (2001) showed that EPS of biofilms could serve as a template for layer-silicate synthesis and catalyse the precipitation of iron-rich smectite. Wooldridge et al. (2017) used chlorophyll-a as a proxy to show a positive correlation between sediment biofilm abundance and claycoat coverage in intertidal estuarine sediments. Duteil et al. (2020) performed laboratory experiments producing clay-coated quartz sands by mixing EPS derived from intertidal diatom biofilms with clay minerals and quartz. These experiments are associated with microorganisms living in aquatic environments. However, close occurrence of biofilms and minerals can also be found in the soil microbiome.
In this study, we mixed single-species bacterial biofilms with high purity quartz sand samples, which were subsequently used in temperature-controlled laboratory experiments to address the following questions: (1) how do EPS of biofilms coat the sand grains, (2) do biofilms act as a growth surface, (3) is the stickiness of the biofilms the main reason for clays to coat the surface of quartz grains, and (4) does mineral authigenesis continue on biofilm surfaces only or expand on clean grain surfaces with increasing temperature. An artificial solution was used to grow and preserve the bacterial population and promoted mineral precipitation during each experiment.
Decoding the interdependence between bacterial EPS and early formation of grain-coating material has identified their importance for acting as a template for authigenic mineral growth.

| MATERIAL S AND ME THODS
The experimental process followed here is based on the following requirements: (1) to ensure the production of EPS that bridge and coat the quartz grains, (2) the artificial solution is enriched with the correct geochemical components to sustain biofilm growth and, with progressive temperature increase, allow mineral nucleation, and (3) besides their use as substrate material, quartz grains act as a silica source for authigenesis.

| Sand grains and microorganism cultivation
Unconsolidated high-purity quartz sand grains were used as a template for biofilm development. Biofilm growth was achieved using Geobacter sulfurreducens, selected for their wide distribution in soils, freshwater and marine environments Lonergan et al., 1996;Lovley et al., 2004). Geobacter sulfurreducens is an anaerobic microorganism capable of reducing e.g. Fe(III) minerals, elemental sulphur or fumarate with acetate as electron donor (Lovley et al., 2004). The bacterial strains were obtained from the laboratory culture collection of the Tuebingen Geomicrobiology Group.
The strains were cultivated in the dark under anoxic conditions (N 2 :CO 2 ; 90:10 headspace) at 35°C in a modified mineral medium (Tables S1 and S2). Growth and cell concentration were monitored by optical density measurements at 600 nm (OD600). Stationary phase cultures were harvested by centrifugation at 2800 g at 10°C three times for 20 min and washed with bicarbonate buffer (30 mM, pH 6.8) to prepare a cell suspension with a final cell density of 10 9 cells/mL. Subsequently, 10 mL of the cell suspension was dispensed in a newly prepared mineral medium (pH 6.8). 50 g of pure quartz grains and 50 mL of the new medium containing the strains were placed in 250 mL serum bottles and left in the dark at 35°C for incubation.
The bottles were manually rotated daily, keeping the quartz sand below the media surface to ensure biofilm development between sand grains (Figure 1c).

| Experimental setup
Hydrothermal reactor experiments were carried out in a Parr© Series 4 560 Mini Reactor comprised of a cylindrical 316 stainless steel pressure vessel (600 mL) under strictly closed conditions ( Figure 1). Four experiments were performed at temperatures between 60°C and 120°C using 5 g of the pure sand grains seeded with Geobacter cells and 100 mL of the mineral medium as pore fluid (see supplementary material). Although most bacteria thrive at lower temperatures in natural environments, the experimental temperature range was selected to include shallow burial (30-70°C) and mesodiagenetic (>70°C) temperature conditions (Morad et al., 2000), ensuring mineral growth within the experimental timeframe. Based on a geothermal gradient of 30°C km −1 , the experimental temperature range equates to a burial depth of between 2-4 km. All samples were held at the desired F I G U R E 2 Pure sand (quartz) sample (starting material) before and after being seeded with Geobacter sulfurreducens cells forming biofilm structures. Diverse types of quartz grains were present. For the associated EDS analysis (e.g., spectrum 1-6) refer to Table 1. (a, b) SE image of the unconsolidated quartz grains (c) Binding of quartz grains by the extracellular polymeric substances (EPS) matrix of the biofilm structures. d) EPS web linking quartz grains. (e, f) Bridging structure between two quartz grains due to EPS strands. experimental temperature for 96 h. At the end of each experiment, the samples were removed from the vessel and air-dried.

| Analytical techniques
A Hitachi SU-70 field emission gun scanning electron microscope (SEM) equipped with an energy-dispersive detector (EDS) was used for rapid elemental identification of the new, due to the experimental treatment, mineral precipitates. The mineralogy of the samples was also evaluated by bulk and clay fraction (<2 μm fraction) X-ray powder diffraction (XRD) analysis.
Aqueous bulk solutions were analysed, using triple quadrupole inductively coupled plasma mass spectrometry (TQ ICP-MS), to establish the elemental composition before and after the experiments at 60°C, 80°C, and 100°C. Liquid fraction could not be obtained from the hydrothermal vessel after the experiment at 120°C as the solution was consumed by the reaction. For more information regarding each technique, see supplementary material.

| Sand grains pre-and post-biofilm growth
The initial unconsolidated pure sand was composed of quartz grains ( Figure 2) of subangular to subrounded shape and size ranging between 150 and 400 μm. SEM imaging indicated that some grains showed smooth surfaces while others were etched and displayed conchoidal fractures.

| Hydrothermally treated biofilm-rich sand samples
At 60°C, the EPS strands were preserved (Figure 3a After the hydrothermal experiment at 100°C, the loose quartz grains transformed into a more rigid framework due to a change in the EPS texture associated with mineral growth (Figure 4a). The EPS strands were still preserved, and a thin layer of EPS coating several grains was observed (Figure 4b). Mineral precipitation occurred mainly in regions surrounding grains bounded by EPS (Figure 4c).
The new precipitates were low-ordered structures with chemical compositions differing from the detrital material (Figure 4d).
At 120°C, quartz grains were partially coated by the neo-  Note: All data are reported in atomic percent normalized to oxygen by stoichiometry. All errors are standard deviations.
( Figure 5a,f). Sharp contacts between clean quartz surfaces and the regions covered by the grain-coating material were prominent.
Rosette-like textures, indicative of an increase of the order of crystallinity, were observed (Figure 5c).

| Sample mineralogy and aqueous solution
Whole-rock XRD and <2 μm fraction analysis of the pre-experiment samples validated that the starting material was composed solely of quartz. Clay-separate analysis could not identify the morphologically clay-like material observed under the SEM. This is likely due to the lack of abundance of the newly precipitated phase either because the amount precipitated was insufficient to be captured by this technique or due to the nature of the phase (e.g., grain-coating) that, even after the ultrasound and centrifugation separation treatment, remained attached to the grains and did not make it into the analysis. However, traces of dolomite were identified at temperatures >60°C, occurring as rhombic-shaped crystals observed under the SEM ( Figure S1).  by creating bonds between them (Decho, 2000). Our initial experiment of seeding pure sand grains with Geobacter cells demonstrates that biofilm colonization preferentially happens on quartz substratum with abraded surfaces and points of contact between grains. After the successful adhesion and attachment, further biofilm growth is controlled by the adsorption of dissolved constituents and sufficient nutrient provision from the artificial aqueous environment.
At near-neutral pH, EPS have a net negative charge on their surface produced by carboxylic, hydroxyl and phosphoryl groups so that they can attract and retain cation species (Cuadros, 2017). The presence of monovalent (Na + ) and divalent (Ca 2+ , Mg 2+ ) cations in the mineral medium (pH = 6.8) used potentially helped in bridging negatively charged groups within polymers and between the different surfaces of bacteria and sand particles, leading to biostabilization (de Brouwer et al., 2002;Flemming & Wingender, 2010). Interactions between EPS and the surface of quartz grains could result from electrostatic bonds formed between charged groups of EPS and silanol groups (Tang et al., 2015). Additionally, the biocolonization of the F I G U R E 4 SEM photomicrographs after hydrothermal treatment at 100 °C. For the associated EDS analysis (e.g., spectrum 11-13) refer to Table 1  This allowed the bacteria to produce and sustain the observed cohesive bio-coating and bridging structures.

| Association of EPS with mineral formation
Bacteria biomineralization is dependent upon the chemical composition of the waters in which the microorganisms are growing (Konhauser et al., 1998). Biofilms, owing to their chemical reactivity and large surface area, are considered ideal scavengers of metal cations (Geesey et al., 1988). In our experiments, Mg and Ca cations were bound by the surface of the polymers. The continued adsorption results in surface site saturation, and once critical supersaturation has been reached, critical nuclei become established.
With increasing temperature and appropriate geochemical components provided by the mineral medium, mineral growth continues abiogenically transitioning from an initial amorphous phase to a low order crystalline material with rosette-like textures (Figure 5c).
Except for the morphologically clay-like phase, rhombic-shaped dolomite crystals were also observed; however, their abundance was volumetrically less significant. The limited abundance of the dolomite crystals can be attributed to the fact that amorphous phases have lower interfacial free energies, with a faster nucleation rate than those of more stable, crystalline phases. As a result, Mg is primarily used by the clay-like phase, thus preventing extensive calcium carbonate precipitation on the EPS-coated surfaces. This could also explain why dolomite is sporadically found on EPS-free quartz grain surfaces (Figure 3d).

| Limitations of the present study and future work
Natural early diagenetic conditions cannot be fully reproduced in the laboratory due to the short experimental time compared to a F I G U R E 5 SEM photomicrographs showing the interrelation of biofilm structures and neoformed mineral phases after hydrothermal treatment at 120°C. For the associated EDS analysis (e.g., spectrum 14-16) refer to 4. These initial findings enhance our understanding on the formation of thinly layered grain-coating material resembling precursor phases that can potentially govern later diagenetic processes.

ACK N O WLE D G E M ENTS
DC thanks Durham University for sponsoring this project through a Durham Doctoral Studentship. SJJ and AK thank the Universities of Durham and Tuebingen for the Seedcorn funding.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that supports the findings of this study are available in the supplementary material of this article.