Different Impact of Suspended Al2O3 Nanoparticles on Microbial Communities: Formation of 2D-Networks (Without Humic Acids) or 3D-Colonies (With Humic Acids)

The use of metal-based and, particularly, Al2O3 nanoparticles (Al2O3-NP) for diverse purposes is exponentially growing. However, the growth of such promissory market is not accompanied by a parallel extensive investigation related to the impact of this pollution on groundwater and biological systems. Pseudomonas species, ubiquitous, environmentally critical microbes, frequently respond to stress conditions with diverse strategies that generally include extracellular polymeric substances (EPS) formation. The aim of this study is to report that changes in the aqueous environment, particularly, the addition of Al2O3-NP without and with humic acids, induce different adaptive strategies of Pseudomonas aeruginosa early biofilms. To this purpose, early biofilms were incubated in diluted culture media without (control) and with Al2O3-NP, and with humic acids (HA-control, HA-Al2O3-NP) for 24 h. 3D colonies with EPS strings and isolated bacteria in their surroundings were detected in the control biofilms. Unlikely, an unusual adaptive behaviour was developed in the presence of Al2O3-NP. Bacteria opt to disassemble the 3D arrangements and to implement a 2D network promoting morphological and size changes of bacterial cells (small coccoid shapes). Remarkably, this strategy allows their temporarily non-EPS-depending survival without decreasing the number of cells. This behaviour was not observed with ZnO-NP, HA-Al2O3-NP, or HA-ZnO-NP. Physicochemical analysis revealed that HA were adsorbed on Al2O3-NP and promoted the Al(III) ions complexation. This supports the hypothesis that the reduction of toxicity of Al ions and the 3D colony formation in the presence of HA-Al2O3-NP is promoted by the complexation of the metal ions with HA components.


Introduction
Production of the engineered nanoparticles (NP) is growing exponentially due to their multiple applications. However, our understanding on the environmental fate and effects of these nanoproducts is far from being complete [1]. There are several important factors to be considered in relation to groundwater contamination by NPs, among them, release of metal ions by NPs, surface reactions with organic components of aquatic media and NP impact in the biological environment [1][2][3][4][5]. Although most studies have been made considering the effect of high NP concentrations, latest investigations also report sublethal effects detected in some species over various generations [6].
Among metal oxide NPs, aluminium oxide NPs (Al 2 O 3 -NP) are widely used and this market is expected to have an annual growth rate of 7.00% [7]. They are characterized by their thermal barrier and superhydrophobic properties as well as by their chemically stability, mechanical strength enhancement and electrical insulating. For these reasons Al 2 O 3 -NP are extensively used in different coatings and composites for diverse applications with the aim of providing resistance from corrosion, scratch and/or wear among other advantages. However, the growing commercial applications of these NPs have turned them into a toxicological concern [8]. It has been shown that they are cytotoxic towards freshwater algae [9] and also affect plant growth, cell morphology, gene expression and metabolic processes in a manner dependent on the physicochemical properties of Al 2 O 3 -NP and environmental conditions [10]. Considering the possible future expansion of this particular NP market, their impact under different natural exposure scenarios where Al-containing NPs can accumulate deserves a special attention since they may affect biological ecosystems [11,12].
Like Al 2 O 3 -NP, ZnO-NPs are other extensively used nanostructures (third highest annually produced nanoparticles, depicting 550 tons per year) [13]. ZnO-NPs have numerous applications due to their advantageous properties appropriate for gas sensing, optoelectronic, photovoltaic, biomedical applications (antibacterial, anticancer) among others [14] and impact in different ecosystems. Thus, it is interesting to compare Al 2 O 3 -NP vs ZnO-NP environmental impact [15].
When the effects of NP disposal is investigated, the presence of dissolved organic matter and, particularly, humic acids (HA) should be taken into account [16] since they include -C = O, -COOH, -OH, -NH-, -NH 2 , -N functional groups capable of interacting with NP and, in turn, conditioning the NP interaction with the microbial community. Additionally, in case of metal oxide NPs like Al 2 O 3 -NP and ZnO-NP, significant challenges may be imposed to the biological medium since the released metal ions may be toxic and create stress conditions such as pH and osmotic changes.
Among microbial communities, mature biofilms are generally more resistant to aggressive environments than the planktonic counterpart. Their resilience is frequently associated to the development of persisters (specialized survivor cells) able to survive to hard stress condition and to the generation of the protective extracellular polymeric matrix (EPS) [17,18]. The EPS matrix is composed of DNA, bacterial polysaccharides, and proteins [19]. Extracellular DNA (eDNA) is a highly anionic polymer that functions as a structural support to maintain biofilm structure. However, at specific concentrations, eDNA may cause cell lysis by chelating metal cations (such as Mg 2+ and Ca 2+ ) that stabilize lipopolysaccharides and the outer membrane, inducing the release of the cytoplasmic content [20,21]. Under specific conditions, intercellular nanoscale connections (known as nanotubes) between neighbouring isolated bacteria are formed [22].
Regarding environmental microbial contaminants, some microbial species such as Pseudomonas are critical because they employ many resistance mechanisms and show extensive metabolic diversity that allow them to thrive in a wide variety of environments and nutrient sources [21,23]. They live as planktonic cells or attached to surfaces and are able to produce large amounts of EPS [24]. Interestingly, in case of exposures to metal ions, unlike antibiotics, both, planktonic and attached bacteria of Pseudomonas species may be tolerant to high concentrations of specific metal ions [21]. Particularly, they display different strategies to survive at high levels of Al ions. One of them is the generation of complexation agents like citrate [7]. Thus, with the aim of overcoming the stress produced by high concentrations of Al ions, metabolic shifts are displayed to convert malate to citrate [25].
The aim of this work is to report an atypical adaptative strategy developed by Pseudomonas aeruginosa early biofilms to survive in aqueous suspension of the widely used Al 2 O 3 -NP. The survival strategies were compared with those displayed in case of ZnO-NPs. The effect of dissolved organic matter (HA) and their impact on the physicochemical properties of the system are also considered.

Materials
The assays with NPs were made at 50 ppm concentration. The selected NP concentration corresponds to a level that was slightly below the concentration able to inhibit the growth of P. aeruginosa cultures (subinhibitory concentration) measured by the agar disk-diffusion method (see Supplementary Information (SI) for details). Commercially available Al 2 O 3 -NP (GetNanoMaterials, purity: 99.97%) or ZnO-NP (GetNanoMaterials, purity: 99.8%) were used. They were suspended in simulate natural waters inorganic matrix (SIM) (see experimental details in SI).

Characterization Assays and Spectroscopic Analysis
Dynamic light scattering (DLS), FTIR-ATR spectroscopy analysis and inductively coupled plasma (ICP) atomic emission spectroscopy were used to physicochemical characterization of the NPs and their suspensions (experimental details are included in SI).

Bacterial Culture and Biofilm Formation
P. aeruginosa (PAK wt), a laboratory reference strain [26], was used in the assays. The isolate was obtained from General Hospital for Acute Illnesses, La Plata, Argentina. It was inoculated in 100 mL of nutrient broth (NB; BritaniaTM, Argentina) and grew overnight at 31 °C in a rotary shaker (250 rpm). As NB is characterized by a carbon content several times higher than that of wastewater, the assays in the current study were made with a diluted culture medium (DCM = 1:10 diluted NB, OD 590 = 0.5) to be closer to the level found in water flows, while ensuring the minimum carbon content needed for bacterial growth). Thus, bacterial inoculum was diluted with fresh nutrient broth (DCM, OD 590 = 0.5) to get 1 × 10 8 (CFU)/mL DCM also used in previous works [27,28,30].
Biofilm assays were performed as described in our previous works with minor modifications [29,30]. Accordingly, P. aeruginosa biofilms were formed during 1 h in DCM on sterilized glass slides (P. aeruginosa suspension OD 590 = 0.5; 1 × 10 8 CFU/mL) and then were introduced into a multiwell plate (24 wells) to form 24 h-biofilms under different conditions: with 10 ppm HA, and 50 ppm of either (a) Al 2 O 3 -NP or HA-Al 2 O 3 -NP; (b) ZnO-NP or HA-ZnO-NP to evaluate the possible antimicrobial effect on the nanoparticles (Al 2 O 3 -NP or ZnO-NP) with and without HA. After the incubation period, loosely adherent bacteria of the biofilms were washed off two times with sterile PBS. The enumeration of sessile bacteria on the surface of the glass slides was carried out by serial dilution method after their detachment by sonication. Thus, the biofilmed glass slides were individually placed in glass tubes with 1 mL of sterile PBS and the irreversibly adherent bacteria were detached by sonication for 15 min using a Testlab ultra-sonic bath (40 kHz, 160 W). Finally, the quantification of the bacteria in the sonicated suspension was performed by appropriate dilutions, onto nutrient agar plates.
The experiments were performed in duplicate, and the assays were repeated at least three times. Student's t-test was used for comparing means on the data sets and p values were indicated according to p < 0.01 (**) and p < 0.05 (*).

SEM Observations
A scheme of biofilm formation like that described in the literature [30] with additional pretreatment to preserve biological material was used to perform SEM observations (see SI for details). EDS analysis was also performed (ESEM, FEI, Quanta 200 with EDX).
The analysis of the distribution of the lengths of the bacteria growing in biofilms was made using the images of SEM microscopy. The percentage of bacteria of each size for the different conditions assayed was calculated by measuring the size of more than 80 bacteria for each condition using SEM microphotographs at 10000X magnification.

Results and Discussion
Early biofilms were formed in the absence and in the presence of the NP (Al 2 O 3 -NP and ZnO-NP). Results of SEM observations and bacterial enumeration were presented in the following sections to investigate possible changes of the bacterial morphology and/or the structure of the aggregations under each condition. The evaluation of the physicochemical changes of the aqueous media associated to the release of metal ions from NP and the presence of HA was also made and subsequently a comparative analysis to find possible relationships between the microbiological response and the physicochemical properties of the media was performed.

P. aeruginosa Early Biofilm Formation
Epifluorescence microscopy of early biofilms formed on glass (control, Fig. S1) showed that isolated bacteria were attached on the surface and small microbial agglomerates were formed. The glass samples with the early biofilms were incubated for 24 h and SEM observations (Fig. 1A, centre) showed that 3D colonies were formed on the control surface with isolated bacteria around the colonies (Fig. 1A (a)). EPS strings can also be identified. They connect bacteria and stick them to generate a 3D matrix, similar to those previously reported [30] (Fig. 1A, control (b) yellow arrows).

Attached Bacterial Growth and Nanotubes Formation in Al 2 O 3 -NP and ZnO Suspension
The behaviour of bacteria in contact with Al 2 O 3 -NP and ZnO was comparatively investigated by contrasting the number of attached bacteria under the different conditions and analysing biofilm structure by SEM microscopy. The addition of Al 2 O 3 -NP did not imply the reduction in the number of attached bacteria in relation to the control ( Fig. 2A and  B). However, SEM observations showed morphological changes of the rod-shape bacteria (Fig. 1B (b)) to coccoid and the reduction in the size (Fig. 2C), revealing a stress response to the toxic environment. Additionally, 3D colonies were not formed and a 2D bacterial net could be distinguished. Remarkably, bacteria were connected by flagella/ nanotubes (Fig. 1B (b) blue arrows) that formed a biological meshwork without producing EPS matrix. Considering that the initial biofilm development involves the formation of small microbial agglomerates (Fig. S1) with isolated bacteria that were not connected with the others in the surroundings, it can be inferred that during the subsequent 24 h period, after the addition of Al 2 O 3 -NP a peculiar restructuration occurred to originate the network (Fig. 1B, left). This nanotubes network was absent in the control samples ( Fig. 1A (control (a)).
The assays performed in the presence of 50 ppm ZnO-NP showed (Fig. 1C) similar allocation of cells and distribution of sizes on the surface to those of the control. However, significantly higher number of living attached bacteria than the control was found (Fig. 1C, centre). Within the colonies, high amounts of EPS strings can be noticed ( Fig. 1C (b), yellow arrows). Nanotubes can also be distinguished structuring the colonial matrix, some of them are in the polar position and are probably flagella (blue arrows in Fig. 1C) while others seem to be generated to provide a connection between the bacterial cytoplasm within the colony.

Changes of Bacterial Size and Morphology by the Addition of Al 2 O 3 -NP, HA-Al 2 O 3 -NP, ZnO-NP, and HA-ZnO-NP
SEM microphotographs of attached bacteria showed that in the presence of Al 2 O 3 -NP the cells were smaller than those of the control (average lengths of 0.8 µm and 1.4 µm, respectively). The frequency distribution of the lengths (Fig. 2C) revealed the marked influence of the environment in the bacterial size and morphology; the smaller sizes correspond to coccoid shapes. A detailed analysis revealed that ranges of size for more than 90% of the control attached cells corresponding to HA-control, ZnONP and HA-ZnO-NP are between 0.6 and 2.2 µm while ranges less wide are related to Al 2 O 3 -NP and HA-Al 2 O 3 -NP conditions (in the 0.4-1.2 µm and 0.4-1.6 µm respectively). Thus, aluminium ions may be toxic for bacteria since they cause morphological and size changes in response to this stress condition.

Physicochemical Changes of the NP Suspension by the Addition of Humic Acids
In view that the complexation of the cations may reduce their toxic effect [16] and that the components of humic acids (HA), frequently present in natural aqueous environments, may function as chelator agents, HA was added to the Al 2 O 3 -NP suspension to investigate their influence in the physicochemical and biological response of the microbial system. Figure S2 shows that the E465/E665 ratio in the liquid filtrate of a Al 2 O 3 -NP suspension containing HA is close to 3, indicating that the high molecular weight aromatic structures remain in the liquid filtrate and the low molecular weight aromatic compounds containing carboxylic and/or carbonyl groups are adsorbed on the Al 2 O 3 -NP surface. DLS results (Table S1) showed that the distribution of the diameters of the Al 2 O 3 -NP aggregates changes if HA is present indicating the influence of the composition of the medium on the size of aggregates.

Adsorption of Humic Acids on Al2O3-NP and Release of Al Ions
FTIR-ATR results presented in Fig. 3 complement this information and reveal the formation of carboxylate groups that seem to promote Al ions complexation (see SI for additional details). Particularly, the spectra corresponding to HA,  [31]. On the other hand, in the case of the liquid filtrate HA-Al 2 O 3 -NP(l), the shift of the band at 1586 to 1637 cm −1 can be attributed to the carboxylate groups with covalent coordinated bonds formed when HA and Al(III) are present. Additionally, a shift from the 1090 cm −1 (HA) to 1114 and 1124 cm −1 for HA-Al 2 (SO 4 ) 3 and HA-Al 2 O 3 -NP(l), respectively, also indicates the formation of coordination complexes between Al(III) and the organic aliphatic molecules with -OH groups. All in all, the formation of carboxylic complexes seems to be promoted in the presence of aluminium and HA.
To determine the release of Al(III) by suspended Al 2 O 3 -NP, the concentration of Al-containing species in the liquid phase was evaluated. To that purpose, Al 2 O 3 -NP(l) was analysed by ICP. The results showed that the Al-containing species released by Al 2 O 3 -NP depend on the composition of the medium in line with previous results [31,16]. Thus, in SIM suspensions, the concentration was 0.993 ppm, while in the presence of HA (HA-Al 2 O 3 -NP(l)), the NPs were able to release 2.77 ppm. These results are in accordance with FTIR-ATR findings supporting that the presence of HA favours the formation of less toxic Al-carboxilic complexes.

Less Toxic Effect of HA-Al 2 O 3 -NP than Al 2 O 3 -NP on Bacterial Cultures
In view of previous results, the addition of HA-Al 2 O 3 -NP suspension to the bacterial culture medium resulted in the increase of the number of attached bacteria in relation to the control. SEM observations revealed that 3D colonies with EPS strings were formed and, consequently, HA addition probably led to the reduction of the toxicity by the chelation of Al-containing ions.
In HA-Al 2 O 3 -NP-containing medium, isolated attached bacteria did not form a network. The average size was slightly larger with less coccoid shape bacteria than in the case of Al 2 O 3 -NP without HA but significantly smaller that the control (Figs. 1D and 2C). As expected, the complexation of the Al ions by HA reduced the toxicity caused by these ions; however, the wellbeing growth condition could not be achieved since their size was smaller. EDS analysis after the biofilm formation showed a decrease of Al/C ratio from 0.088 in the presence of Al 2 O 3 -NP to 0.070 with HA-Al 2 O 3 -NP, revealing a decrease of the relative Al content related to the organic components on the surface after the addition of these HA-containing NPs. According to FTIR analysis, these results are related to the complexation process that induces the release of less toxic Al-carboxilic complexes and favours the growth of bacteria with the consequent increase of carbon content.

Adaptive Behaviour of Bacteria to Al2O3-NP Toxicity
Unlike current antibiotic agents, metal ions under certain conditions can eradicate biofilms at similar metal ions concentrations than the more labile planktonic cells [18] revealing different action mechanisms. It was reported recently that to overcome the stress condition caused by metal ions, Pseudomonas species developed different survival strategies. One of them is the generation of complexation agents such as citrate for Al ions [11]. In this case, a metabolic shift in order to convert malate to citrate is produced [25]. In this context, results reported here complement previous results and show that the addition of Al-containing NPs in the case of P. aeruginosa induces complex atypical transformations in early biofilms that can be reverted in the presence of HA. Among them, morphological and size changes of attached cells take place in line with previous reports [32]. Additionally, peculiar results were found related to biofilm structure showing that the original 3D bacterial aggregates disappear and a 2D network is formed with nanotubes as connectors of small coccoid bacteria when Al 2 O 3 -NP are present. It could be speculated that the advantages of this new structure could be related to the ability of nanotubes to scavenge and deliver molecules inside them, providing a network for communication and exchange of cellular molecules [22] and also mediating the transfer of cytoplasmic molecules between adjacent cells and the transiently acquired nonhereditary resistance to antimicrobials. Besides, the transference of plasmids may also occurred granting hereditary features to recipient cells [33], 34. Results also reveal that, although EPS formation is frequently interpreted as an adaptative behaviour of P. aeruginosa to survive in toxic environments, our results show that, under the in vitro conditions analysed, abundant production of EPS is not the way that P. aeruginosa uses to adapt to Al 2 O 3 -NP toxicity.
The detailed observation of Fig. 1B (left, red arrow) also reveals interesting features since the direction of the nanotubes of several bacteria is focused on a particular cell that according to previous reports may act as a nutrient donor. It was suggested that weak bacteria die under unfavourable conditions, and the surviving cells live on their expense (strategy known as "bust-and-boom") [35]. eDNA that may be produced through explosive cell lysis events could facilitate the survival in toxic or non-nutrient stress condition [36]. Thus, bust-and-boom strategy may also be used by P. aeruginosa in the networks.
To our best knowledge, it is the first time that such changes in bacterial behaviour are reported as a consequence of the presence of NP. In fact, the atypical stress response described here for Al 2 O 3 -NP was not observed in case of other NP such as ZnO or after the addition of HA that contains ligands for Al ions, and it may be assumed that it is caused by the presence of free Al ions.
Results reported here contribute to the understanding of how bacteria cope with environmental changes that Fig. 4 Scheme of the transformation of early biofilm architecture after being exposed for 24 h to different media: culture medium (3D colony formation of normal rod-shape bacteria with EPS strings), Al 2 O 3 -NP suspension (network of smaller bacteria without EPS strings), and HA-Al 2 O 3 -NP (3D colony formation of smaller rod-shape and coccoid bacteria with EPS strings) challenge survival, although further studies are being designed to achieve a detailed information about metabolic changes in different environmental conditions and exposure periods. The atypical P. aeruginosa strategy described above reveals the diverse mechanisms developed by bacteria to overcome stress conditions that should be considered for the design of the bioremediation processes and medical treatments for infections.

Conclusions
Al 2 O 3 -NP and, particularly, Al ions release, induces transformations in early biofilms (Fig. 4). The initial small aggregates of bacteria disassemble, and a network is developed without significant EPS production. During the meshwork generation, isolated bacteria joint with the neighbouring cells by using intercellular connections, known as nanotubes. These nanotubes may facilitate communication and exchange of cellular molecules between cells, reducing their interchanges with the environment.
In the case of HA-Al 2 O 3 -NP suspension, physicochemical studies revealed that the adsorption of HA on Al 2 O 3 -NP and the complexation of metal ions with HA components seem to reduce the toxicity level of the metal ions with beneficial consequences for the microbial system.
2D network formation was associated to the presence of Al(III) free ions since it was observed neither with ZnO-NP nor when HA was added to the Al 2 O 3 -NP, due to complexation of Al(III). Remarkably, to the best of our knowledge, this 2D-network strategy survival of smaller bacteria without reducing the number of cells related to the control or increasing EPS production has not been previously reported for P. aeruginosa NPs/Al(III) systems.

Supplementary Information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s00248-022-01961-6. Data Availability Data will be available for reviews and editors as required.
Code Availability Not applicable.