Silver nanoparticles are the most widely used in commercial applications such as personal care products. They are highly stable and can persist in the natural environment resulting in chronic exposure threats. The common Ag nanoparticles' physicochemical transformations include aggregation, dissolution, oxidation, sulfidation, and organic coating; sulfidation being the key mechanism of transformation (Levard et al., 2012). E. coli chronically exposed to silver sulfide nanoparticles accumulated phenotypic and genotypic changes that contributed to its tolerance and resistance to various antibiotics. Chronic exposure to silver sulfide nanoparticles causes an upregulation of stress-related genes belonging to electron transport, cellular respiration, indicated respiration pathway disturbances, and oxidative stress. Upregulation of the multicomponent copper transport system (cusFCBA) which belongs to the P-type ATPases correlated well with the increase in energy requirement. The authors reported that this nanoparticle-induced intracellular stress also caused an upregulation of the multidrug-resistant genes. These changes triggered decreased sensitivity to antibiotics such as penicillin, kanamycin, ciprofloxacin, and gentamycin. Thus, oxidative stress started a domino effect affecting the overexpression of membrane porin genes and MDR efflux genes such as marA and acrAB-tolC, thus promoting adaptive pathogenic evolution (Li et al., 2019). This study highlighted the necessity to evaluate the sulfidation of nanoparticles or any modifications to the surface chemistry of nanoparticles for improved bactericidal activity, as it can backfire.

Nanoparticles first encounter the microbial cell membrane. Therefore, the nanoparticle-induced biophysical and biochemical changes need to be understood to determine toxicity or tolerance. The lipids and proteins in the cell membrane are highly dynamic and are altered in response to any environmental stress. These changes can serve as a fingerprint to understand the biophysical mechanism of nanoparticles (Kamat et al., 2022a). The role of bacterial cell membrane composition and the toxicity mechanisms of metal oxide nanoparticles are currently debated. Recent studies attempted to decipher this relation using transcriptomics, nanomechanics, electrodynamics, and fluorescence assays (Pagnout et al., 2021). The effect of commercial nanoscale zerovalent iron on nanoparticle–membrane interaction was evaluated in P. putida F1. Beyond a certain concentration, the tolerance to nanoparticles increased substantially. The nanoparticles accumulated onto and penetrated the bacterial cell membrane, making it rigid due to cis/trans isomerization of fatty acids. The tolerant bacterial cells have a higher charge density and lower electrophoretic softness possibly due to modification in outer membrane components (Kotchaplai et al., 2017). Since metal oxide nanoparticles have effective antibacterial properties, their toxicity mechanisms are studied in detail. For a long time, the antibacterial activity of TiO₂ nanoparticles was associated with its photocatalytic property that could induce ROS, membrane alteration, and lipid peroxidation. However, the same observations were also recorded even in the absence of light for TiO₂ nanoparticles and other nanomaterials such as ZnO, MgO, CeO₂, and fullerene nanoparticles. The toxicity mechanisms in dark were associated with osmotic stress and cell membrane alterations as a result of electrostatic attachment of metal nanoparticles as well as mechanical membrane disruption (Pagnout et al., 2012; Leung et al., 2016). Such studies highlighted the necessity to unravel the interaction of nanoparticles and membrane proteins, lipopolysaccharides, and membrane composition that could govern the possible entry of metallic nanoparticles into the cell. The differential sensitivity of bacteria to TiO₂ nanoparticles was studied by Pagnout et al. (2021), wherein they specifically focused on the much-neglected membrane vesicles. Since the production of membrane vesicles is utilized as a defense in mitigating oxidative or osmotic stress, they could have an implication in nanoparticle resistance. The authors exposed E. coli K-12 mutant strains with truncated lipopolysaccharide (LPS) and rfaC gene mutation to TiO₂ nanoparticles and recorded the resistance or sensitivity. These bacteria with altered membranes could efficiently resist the toxicity of TiO₂ nanoparticles its hypervesiculation capacity and regulation of osmotic stress. With an increasing concentration of TiO₂ nanoparticles, the membrane integrity is compromised coupled with oxidative stress and lipid peroxidation. There is also an increase in the secreted membrane vesicles to cope with the toxicity. They could be mediating membrane stress relief through the shuttling of proteins, LPS, and other harmful peroxidation products that accumulate in the periplasmic space or cause modifications on the outer membrane (Pagnout et al., 2021). This study warrants the consideration of membrane biophysics and composition to evaluate the sensitivity of nanoparticles.

Zhang et al. (2018) evaluated the effect of ZnO nanoparticles in E. coli. The resistant bacteria could grow at high concentrations of ZnO (1,000 μg/mL). When the nanoparticles were removed from the growth medium for several days, the antibacterial sensitivity was restored. The development of this transient resistance is adaptive and not due to genetic changes but basically due to changes in the shape of bacteria. The resistant bacteria were oval shaped and switched to the typical rod-like shape when turned sensitive. The authors suggested this change in shape could be due to the lack of protein RodZ or envZ that provides binding sites on the cell membrane for cytoskeletal protein MreB. It is anticipated that the bacteria could have adapted to the low expression of RodZ and the high concentration of ZnO led to the oval phenotype. Another possible mechanism could be related to the porin proteins OmpC and OmpF. While OmpC is more restrictive than OmpF, it is the ratio of these porins that is influenced by the nanoparticles. ZnO nanoparticles could be binding to cAMP, thereby blocking the formation of the cAMP–CRP complex that would otherwise bind to the DNA for the production of OmpR. The resulting low expression of OmpR results in limited phosphorylation of OmpR which, in turn, will increase the expression of OmpF. This results in a more permeable outer membrane as OmpF is an open porin complex (Forst et al., 1988). Because the number of porins per cell volume increases when the cell shape changes to oval, it provides an additional route for nano resistance by mere change in the expression of OmpF when stressed by antibacterial nanoparticles. The authors conjecture that the oval shape during prolonged exposure to nanoparticles reduces the possibility of the nanoparticles entering the bacteria and therefore contributes to unstable nanoparticle resistance (Zhang et al., 2018). We infer that it is the change in membrane proteins such as the porins and RodZ that renders the bacteria momentarily resistant to ZnO nanoparticles. This is because, the porins, especially the OmpF allow the passive transport of small solutes. The nanoparticles could be crossing the bacterial membrane but a different mechanism allows for survival. Pagnout et al. (2021) observed non-monotonous OmpF levels with varying concentrations of TiO₂ nanoparticles. The increased membrane permeability compensates for the Turgor stress and allows the survival of bacteria. In a recent study, the biocidal effect of nanostructured anatase rutile and carbon coating was observed to induce adaptive resistance in E. coli BW25113 (Wasa et al., 2022). The bacteria were observed to utilize their innate response to toxic environments that included changes in efflux proteins and membrane permeability. An increase in the levels of outer membrane protein TolC and superoxide stress protein SoxS was recorded; however, a whole transcriptome analysis is necessary to trace a detailed map of the mechanism.

Bacteria can adapt to environmental conditions by regulating cell division and external conditions too can modulate cell division. This is a well-orchestrated event and involves the transduction of environmental signals to control cell division proteins. In most bacteria, FtsZ is the central tubulin-like protein that marks the site of cytokinesis. Nanoparticle-induced oxidative stress initiates an SOS response and an upregulation of protective proteins such as YpsA and UgtP. Overproduction of such proteins leads to impaired FtsZ ring assembly and ultimately cell division inhibition (Brzozowski et al., 2019). Prolonged exposure to nanoparticle stress could induce an irreversible change in the tightly regulated cell division cycle “Z-ring component assembly” (Gunawan et al., 2013). In bacteria, several regulatory proteins render the Z-ring assembly/disassembly dynamic. During an SOS response, a regulatory protein ugtP delays cell division by destabilizing the Z-ring (Adams and Errington, 2009). This halts the cytokinesis and therefore the cell division cycle stops. This mechanism could be an indirect protective mechanism developed by bacteria against nanoparticles and needs to be further investigated.

Stabryla et al. (2021) evaluated whether Ag nanoparticles or the released Ag(I) ions or their combination drives the resistance in E. coli. Bacterial resistance to Ag(I) ions through efflux pumps or reduction to its lesser toxic oxidation state Ag(0) is well established. The resistance is also attributed to E. coli deficiency in outer membrane proteins, resulting in the limited uptake of Ag. The sil operon in gram-negative bacteria is also well studied and encodes for efflux pumps: SilP (P-type ATPase) and SilCBA (antiporter) along with SilE and SilF (two periplasmic Ag binding chaperones. These components function synergistically to drive out Ag(I) of the gram-negative bacteria (Silver, 2003). A hypermotile E. coli strain demonstrated specific resistance to Ag nanoparticles as compared to the non-motile phenotypic strain. Both strains demonstrated the same degree of particle aggregation, indicating that aggregation alone could not be responsible for the differential resistance. However, an alternative mechanism that could be responsible is that the motile bacteria relocate away from the aggregated nanoparticles, thus reducing exposure to nanoparticles and increasing nutrient availability (Stabryla et al., 2021). Flagellin also triggers the aggregation of nanoparticles and contributes to resistance in P. aeruginosa CCM 3955 and E. coli CCM 3954. The evolution of this non-genetic change reduces the colloidal stability of nanoparticles and their antibacterial property. Flagellin forms the extracellular flagellar filaments necessary for bacterial motility. It also has adhesive properties and together with flagellum is critical for biofilm formation. However, the addition of pomegranate rind extract (which inhibits flagellin production) eliminated nanoparticle aggregation and bacterial resistance to Ag nanoparticles (Panáček et al., 2018).

Valentin et al. (2020) reported the development of heritable adaptation to nanosilver (NAg) in S. aureus in an extensive study. The bacterium was serially cultured in the presence or absence of nanosilver for 50 days as performed for antibiotic resistance development (Ling et al., 2015). Whole-genome sequencing of nanoparticle-adapted S. aureus, the unevolved wild-type S. aureus and the wild-type S. aureus revealed two critical clues that explained the heritability of the evolved resistance adaptations. The first difference observed in Nag-resistant S. aureus was a single-nucleotide variation (TCT to TAT mutation) in the purR gene that encodes PurR, a putative purine operon repressor. The mutation resulted in serine to tyrosine at position 169 in the binding motif of PurR. PurR binding to the operon generally downregulates the enzymes responsible for the synthesis of inosine monophosphate (a precursor of the adenosine and guanosine monophosphates) (Cui et al., 2013). The evolved mutation reduces the binding affinity of the repressor resulting in elevated purine biosynthesis. Since NAg induces bacterial toxicity by DNA and protein damage, the evolved resistance mechanism compensates for survival by the overproduction of purines necessary for DNA replication, repair, and amino acid synthesis. The second difference observed was a single-nucleotide deletion in the tcyA gene of the Nag-resistant S. aureus. The gene encodes TcyA, a putative L-cystine binding protein that imports cystine from the extracellular environment into the cytoplasm for reduction to cysteine. In the bacterium, cysteine contributes to the iron–sulfur active site of thioredoxin and many enzymes of the tricarboxylic acid cycle. The oxidative stress induced by NAg exposure is mitigated by the antioxidant action of thioredoxin. The ATC to -TC mutation in the tcyA translates into a truncated protein. This results in a reduced import of cystine and low cellular cysteine (Even et al., 2006). This could reduce the cysteine-mediated redox imbalance linked to NAg toxicity in bacterial cells. These mutations were continually detected in daughter cells even upon removal of NAg exposure (Even et al., 2006; Valentin et al., 2020).

The most important parameters which affect the antimicrobial properties of nanoparticles are their type, shape, size, and the type of surface corona (Kumari et al., 2016, 2017). The maximum number of cases of microbes developing nanoparticle resistance is against metal nanoparticles, specifically silver and copper nanoparticles (Amaro et al., 2021). However, researchers are emphasizing the surface properties and size of nanoparticles to overcome the resistance developed by microbes. Zheng et al. (2021) developed 4,6-diamino-2-pyrimidine thiol (DAPT)-capped gold nanoparticles (AuDAPTs), against which E. coli gradually developed resistance after long-term use. Though it was observed that the resistance developed was size specific, and it can be reverted by simply modifying the size of AuDAPTs. Similarly, silver covalently bound to cyanographene (GCN/Ag) effectively killed the silver nanoparticles-resistant bacteria by inducing strong interaction of GCN/Ag with the bacterial membrane (Panáček et al., 2021).

The most important parameter responsible for resistance development is the concentration of nanoparticles used and the aggregation and stabilization pattern of nanoparticles in the microbes. Recent studies found that bacteria can produce flagellin upon exposure to silver nanoparticles which can aggregate and thus deactivate the antimicrobial potential of silver nanoparticles (Panáček et al., 2018, 2021). Sub-lethal concentrations of metal nanoparticles are known to induce mutagenic effects and the emergence of selection pressure in bacterial genome (Amaro et al., 2021; McNeilly et al., 2021), though contradictory results have been reported. McNeilly et al. (2021) reported that the continuous use of AgNO₃ and silver nanoparticles against E. coli resulted in Ag+ resistance development by the induction of endogenous mutation. In another study, Wu et al. (2022) reported a negligible role of sub-lethal doses of silver nanoparticles on the mutation rate and resistance development in E. coli.

Many bacteria have R plasmids and metal exclusion capacity, which provides them an additional edge over metallic nanoparticles such as silver, copper, and alumina. It has been observed that the sub-lethal concentration of metallic nanoparticles can promote the conjugational transfer of multi-resistance plasmid RP-4 (Qiu et al., 2012, 2015) in certain bacterial strains and thus increases the chances of acquiring resistance against metal nanoparticles. The metal efflux complexes located in R plasmids or bacterial genome can encode proteins that are responsible for metal ion exclusion outside the bacterial cells (Salas Orozco et al., 2019).

The ability to form biofilms by microbial strains is another crucial factor that decides the susceptibility or resistance to nanoparticles. Though antimicrobial nanoparticles can effectively eradicate biofilms, bacterial biofilms can also cause aggregation of nanoparticles making them ineffective as antimicrobials (Joshi et al., 2020). The components of biofilm, such as protein, lipids, nucleic acid, and polysaccharides can interact and modify the nanoparticles corona, making them lesser effective as antimicrobials (Joo and Aggarwal, 2018). Furthermore, the role of the size and surface corona of nanoparticles cannot be neglected while battling biofilms (Nallathamby et al., 2010).

Both exogenous and endogenous molecular determinants are responsible for developing nanoparticles specific resistance in microbes. As discussed in the earlier section, mechanisms of nanoparticle resistance include multiple genes involved in oxidative stress, membrane disruption, adaptive resistance, cell division, motility, and biofilm formation. However, Panáček et al. (2021) reported that no genetic change in silver nanoparticles aggregation was involved by flagellin secretion in E. coli. The consecutive use of sub-lethal concentrations of metal nanoparticles in different environments and lab settings are continuously spreading the horizontal gene transfer and increases the co-selection of metallic and antibiotic resistance (Amaro et al., 2021). The studies have elucidated that DNA uptake by damaged membranes, increased mutation frequencies, and SOS DNA repair are some of the mechanisms employed by microbes after exposure to metallic nanoparticles which promote nanoparticle resistance (Crane et al., 2018; Amaro et al., 2021). It was observed that membrane alteration can act as double-edged sword for killing or developing nanoparticle resistance in bacteria. The disrupted membrane can increase the frequency of foreign DNA uptake inside the bacterial cells, which can promote the increase in conjugation frequencies and SOS DNA repair responses. The increased SOS DNA repair responses can further lead to horizontal gene transfer by increasing recombination between the foreign DNA and the recipient's DNA (Amaro et al., 2021).

It is quite evident that one single factor is not responsible for microbe resistance to nanoparticles; rather, the mechanism of microbes–nanoparticle interactions, properties of nanoparticles and microbes, and their surrounding environment decides the winner of this important battle.