Contrasting responses of motile and non-motile Escherichia coli strains in resuscitation against stable ultrafine gold nanosystems

Global public health confronts a pressing challenge in antimicrobial resistance (AMR), necessitating urgent intervention strategies due to the low success rate of new antibiotic development. Bacterial motility, beyond conventional antibiotic usage, significantly influences resistance evolution and ecological dynamics. Our recent study marks a breakthrough, revealing the unexplored ability of ultrafine gold nanosystems (UGNs) to inhibit bacterial resuscitation using a motile Escherichia coli ( E.coli ) K12 strain. We aim to deepen our comparative understanding of UGNs’ efficacy and resuscitation propensity against a non-motile E. coli K12 strain to assess the role of motility. Through UGN application, we identified heritable resistance in both strains, with motile strains exhibiting notably higher mutation rates. Resuscitation experiments unveiled faster recovery in motile strains, attributable to virulence factors, compared to non-motile strains. Additionally, our investigation into aggregation dynamics highlighted the role of protein-mediated aggregation in resistance development to nano-antimicrobials. Overall, the study reveals that the non-motile strains are more susceptible against UGNs, which shows promise in combating AMR.


Introduction
Microorganisms started developing resistance to antibiotics much before humans began manufacturing these drugs to combat infectious diseases [1][2][3].Microbes' innate ability to survive in hostile environments contributes to antimicrobial resistance (AMR), which caused 1.27 million deaths worldwide in 2019, with numbers expected to rise sharply in the coming years [4].The staggering death rate and economic impact projections further highlight the urgent intervention in this domain to avoid preventable deaths.Though active research is being pursued to develop new antibiotics, the success rate from the hit-to-lead phase to the clinical trial, followed by market approval, is grim since only one out of ~ 10,000 molecules is reported to qualify with a substantial investment of ~ 100 million Euros over ~ 14 years [5].Yet, the possibility of microbes developing resistance against them cannot be ruled out [6,7].
Current studies suggest the heterogeneous use of antibiotics at varied concentrations is not the sole factor accelerating the evolution of antibiotic resistance [3,8].
A variety of cell types exhibit a certain level of motility, which regulates their ecology and physiological processes [9,10].Motility holds significant importance for a broad range of unicellular organisms, from bacteria to amoeba and algae [10,11].It enables them to locate nutrients, light, or a host, while also aiding in escaping toxic compounds, predators, or parasites [10,12,13].Bacterial reproduction and disease-causing ability depend on their motility, which also influences the emergence of resistance in environments with antimicrobial compounds [9,14].The most well-known form of motility in bacteria involves the utilization of a specialized rotating organelle known as the flagellum [15,16].The bacterial type III secretion system consists of an inner membrane molecular motor, a hook-basal body complex forming an export channel, a 'hook' connector, and a rotating filament.All of these components contribute to the virulence factors and consequently to the development of resistance [17].Bacterial motility contributes significantly to the emergence of resistance through several mechanisms.For instance, motile bacteria can more effectively disperse in their surroundings, facilitating the colonization in new environments and the dissemination of resistance genes within bacterial populations [18].Flagella, integral to motility, can also influence efflux pump activity, a key mechanism in resistance development [19].
Understanding the intricate relationship between bacterial motility and resistance is crucial for devising effective strategies against antimicrobial resistance.Targeting pathways associated with motility, such as flagellar biosynthesis and chemotaxis, offer novel approaches to disrupt bacterial colonization, biofilm formation, and the spread of resistance genes.Recent studies involving silver nanoparticles have shed light on the crucial role of protein-mediated nanoparticle aggregation and flagellar motility as the key fitness traits in conferring resistance to Gram-negative bacteria [20,21].
Recently, our group reported the use of stable ultrafine gold nanosystems (UGN) to overcome microbial resuscitation [22].These UGNs were synthesized using a pseudo solid-phase synthetic approach by employing citric acid (CA) as the reducing as well as capping agent.Later, the CA was neutralized with NaOH to eliminate the acid stress caused by CA.However, these UGNs were found to be labile and hence susceptible to aggregation in a few hours duration.Another category of UGN was also developed, which were capped with glutathione (GSH), whose thiol end was tagged with the gold surface and thereby providing stability.By varying the amount of GSH to 50%, 100%, and 200% relative to the gold atomic percentage, stable UGNs having either fully or partially GSH-capped nano-antimicrobials were synthesized.While the labile UGNs exhibited burst activity against the bacteria, the stable ones exhibited steady-but-sustained activity.Nevertheless, the activity of UGNs was only studied in Escherichia coli K12 (DH5α), which is a motile strain.Given the factors affecting bacterial motility, our current study focuses on utilizing a non-motile strain of E. coli (BL21(DE3)pLysS) as a model system.We investigate the resuscitation of treated populations and the development of resistance in response to UGNs.Additionally, we examine the aggregation dynamics of UGNs when in contact with both motile and non-motile strains.Whole genome sequencing analysis of the mutational profiles of resistant strains reveals significantly higher mutations in multiple metabolic pathways in the motile strain compared to the non-motile strain.Finally, we demonstrate the ability of PGRE treatment to sensitize resistant populations against UGNs.Therefore, our study focuses on understanding the role of motility in the revival of persister cells and its impact on the development of resistance.This allows us to delve into and contrast the ligand-decoupled antibacterial effects of UGNs compared to their impact on motile strains [23,24].The non-motile strain with chloramphenicol selection marker was devoid of the fliC gene, responsible for flagellin production for motility.

Synthesis of UGNs
The synthesis of weakly-capped gold nanoclusters (Au NCs) was performed using a pseudo-solid-state approach, as described in our previous report [25].Briefly, the synthesis involved taking 150 mg of citric acid and adding ~ 6 mg of AuCl 3 .For the entire batch, the amounts of GSH used were 3 mg, 6 mg, and 12 mg for UGN(+)GSH50, UGN(+)GSH100, and UGN(+)GSH200, respectively.Surface ligand modulation of the synthesized Au NCs was achieved by taking 10 mg of the Au NCs and adding it to 1 mL of an aqueous NaOH solution, where the NaOH was three equivalents of citric acid, resulting in UGN(-)GSH.For UGN(+)GSH50, UGN(+) GSH100, and UGN(+)GSH200, an additional 0.2 mg, 0.4 mg, and 0.8 mg of GSH were added to the original NaOH solution, respectively.

Determination of inhibitory concentration-50 (IC 50 ) by Alamar blue
100 µL of E.coli culture with a cell density of 10 7 CFU/ mL was added into each well of a 96-well plate containing 100 µL of varying concentrations of antimicrobial aqueous solutions of UGNs ranging from 5 to 55 µg/mL and ampicillin ranging from 0.35 to 33.25 µg/mL [26].After incubating for 6 h, 20 µL of 0.02% Alamar blue dye solution was added to all cultures, including the blank and control.The fluorescence intensity of the sample solutions was recorded and normalized against the blank [27,28].The average of triplicate measurements was calculated, and the corresponding standard deviation was reported.The antimicrobial concentration at which approximately 50% inhibition occurred was determined as IC 50 .

Determination of minimum inhibitory concentration (MIC)
The MIC values of UGNs against E. coli K12 (DH5α) and (BL21(DE3)pLysS) were determined using the standard broth dilution method.Overnight cultures were measured for absorbance at 600 nm and then diluted with sterile LB broth to achieve a cell population of 5 × 10 5 CFU/mL.Subsequently, 100 μL of this diluted culture was added to each well of a 96-well plate containing 100 μL of the twofold diluted antimicrobial compound, followed by incubation for 18 h at 37 °C.The MIC was defined as the lowest concentration of antimicrobial solution resulting in clear, transparent liquid with no visible bacterial growth in the LB broth medium.

Determination of minimum bactericidal concentration (MBC)
The MBC of the antimicrobials was determined using the standard LB agar plating method.Initially, 100 μL of bacterial cultures, with a concentration of 10 7 CFU/mL, were inoculated into 1.5 mL Eppendorf tubes containing 100 μL of antimicrobial solutions with varying concentrations ranging from IC 50 to 16 × IC 50 .These solutions were then incubated at 37 °C for 6 h.Following incubation, 10 µL of solution from each concentration was diluted with 20 µL of autoclaved MilliQ water and spread onto a LB agar plate, which was subsequently incubated for 24 h, after which the bacterial colonies were counted.In the time-kill MBC analysis, the initial incubation times with the antimicrobials were varied to 6, 12, 24, and 48 h.After each time interval, plating was performed from the treated cultures.The MBC was determined as the lowest concentration at which no bacterial colonies were observed.

Studies on microbial resuscitation
Microbial resuscitation studies were performed using freshly grown bacterial cultures having an initial cell density of 10 7 CFU/mL and the antimicrobials varying in concentration between IC 50 and 10 × IC 50 .The experiment was conducted by monitoring the changes over a time of 6, 12, 24, and 48 h.These changes were determined by assessing the bacterial viability using the Alamar Blue dye as discussed in "Determination of inhibitory concentration-50 (IC 50 ) by Alamar blue" along with the changes at OD 600.

fluorescence-activated cell sorting (FACS) analysis
FACS analysis for the determination of live and dead cells was done for both motile and non-motile strains with ampicillin, UGN(−)GSH and UGN(+)GSH50.The concentrations administered include IC 50 , 2 × IC 50 , 4 × IC 50, and 8 × IC 50.This range was chosen deliberately as attempts beyond it resulted in minimal cell pellet recovery, presumably due to extensive cellular damage.The treated bacterial cultures were centrifuged at 5000×g for 10 min and then washed with 0.9% NaCl.The pelleted cells were then stained with a mixture of SYTO9 and PI in a ratio of 2:1 and incubated in dark for 1 h [6].
After staining, the cells were centrifuged again at 5000×g for 10 min, and the pellet obtained was washed once with 0.9% NaCl before analysis.Three controls-live, dead, and unstained cells-were used as references, and the dead cells were obtained by treating them with 70% isopropanol.

Quantifying resistance in response to UGNs
To assess the development of resistance, both motile and non-motile bacterial strains underwent 30 passages with intermittent exposure to IC 50 concentrations of both categories of UGNs.The passaging study was executed as follows: 1.5 mL of bacterial culture (with a concentration of 10 7 CFU/mL) with an equal volume of the antimicrobial substance was incubated for 3 h at 37 °C [20,29].After treatment, the cells were centrifuged at 5000×g for 10 min at 4 °C, and the pellet obtained was resuspended in 1 mL of fresh media.Subsequently, 500 μL of the suspension was transferred to 2.5 mL of fresh media and incubated overnight until reaching an OD 600 beyond 2. The overnight cultures were then diluted to a cell density of 10 7 CFU/mL for use in the next passage.The untreated cells served as controls.A separate lineage was established for each antimicrobial (UGN(-)GSH, UGN(+) GSH50, and ampicillin) for both strains, and the IC 50 was determined using the Alamar blue assay, with duplicate measurements performed in each case.
Pomegranate rind extract (PGRE) is acknowledged for its capability to inhibit the production of bacterial flagellin.Therefore, the sensitization effect of PGRE on the resistant populations of both strains was performed using the extract obtained similarly to the literature report [20,30].For this, the populations obtained after the 34th passage were subjected to the treatment with sub-inhibitory concentrations of PGRE.Briefly, 10 7 CFU/mL were incubated with 0.3% concentration of PGRE for 24 h at 37 °C, after which the cells were centrifuged at 5000×g for 10 min at 4 °C.The pellet was processed similarly as mentioned above to obtain the culture for the Alamar blue assay.

UV-visible, DLS measurements, and gold quantification
The absorbance and particle size analyses of the as-synthesized UGNs were initially characterized using UVvisible spectroscopy and dynamic light scattering (DLS), respectively [31][32][33].The same analyses were also performed on the solutions containing the bacterial cultures treated with the antimicrobials before and after filtering the same using a 0.2 μm PVDF membrane filter.
Gold quantification was done by utilizing inductively coupled plasma optical emission spectroscopy (ICP-OES).The gold present in the pellet and supernatant solutions were taken for the quantification.About 200 μL of the ancestor and resistant bacterial populations of both strains were treated with their respective 2 × IC 50 concentrations (200 μL) of UGN(-)GSH and UGN(+)GSH50 for 6 h.Following treatment, the cells were centrifuged at 5000×g for 10 min.The pellets and supernatant solutions obtained were then dissolved in 1 mL and 600 μL of aqua regia, respectively.Additionally, before gold quantification, 5 mL of 10% HCl was added to both solutions [34,35].

Whole genome sequencing analysis
The motile and non-motile strains subjected to passaging studies with antimicrobials for up to 30 days were employed for genomic DNA isolation using the phenolchloroform method [36].A control population was also included in the study which underwent a similar passaging protocol but without exposure to any antimicrobial substances.
The process involved creating paired-end sequencing libraries from the DNA samples using the Illumina TruSeq Nano DNA Library Prep kit to conduct whole genome sequencing on the extracted genomic DNA.Subsequently, BWA MEM (version 0.7.17) was utilized to align the highquality reads of the obtained variant sequences to the reference genome, (E. coli K-12 MG1655 database accessible via the National Center for Biotechnology Information (NCBI)), aiming to identify mutations.Comparisons were made between the coding sequences of each strain and those of the ancestor and control populations to identify mutations occurring with 100% frequency.Following this, KEGG enrichment analysis was performed using Partek ® Genomics Suite ® (version: 7.18.0723) to assess the impact of mutant genes on biological, cellular, and molecular processes in each category.The process involved performing GO enrichment analysis to identify statistically enriched categories in the mutant genes, with a significant threshold p-value of < 0.05.Additionally, each mutation was functionally classified by manually determining gene functions using the NCBI gene database (http:// www.ncbi.nlm.nih.gov/ gene).Furthermore, a dataset containing altered genes from each strain was separately uploaded to the STRING database (version 11.0b, https:// string-db.org/) to conduct specific pathway analysis.Default parameter values were used, including a complete STRING network, a medium confidence score greater than 0.4, and a false discovery rate (FDR) stringency of less than 5%, to identify substantially enriched KEGG pathways.

Determination of inhibitory concentration-50 (IC 50 ) by alamar blue
Following the successful assessment of the non-toxicity of the NaOH-neutralized CA and GSH incorporation (Fig. S1), Alamar blue assay studies were undertaken using the UGNs-based compositions to obtain their IC 50 values.Alamar Blue, also known as resazurin dye, is a non-fluorescent and non-toxic compound.It is reduced to resorufin, a pink and highly fluorescent substance, through cellular metabolic activity, facilitated by NADH and NADPH in the presence of NADPH dehydrogenase or NADH dehydrogenase enzymes in the cytoplasm.This reduction provides a quantitative measure of cell viability [37,38].The Alamar blue assay revealed an identical IC 50 of 8.3 μg/mL with both UGN(-)GSH and UGN(+)GSH50 for motile and non-motile strain, though the percentage killing at higher concentrations with the latter was found to be lesser.As can be seen from Table S1, while the IC 50 of UGN(−)GSH against both strains was found to be similar to the MIC determined by the conventional broth dilution method, the MIC of both strains against UGN(+)GSH50 was found to be double (16.6 μg/mL).With further increase in the GSH content as in UGN(+) GSH100, the IC 50 was increased to 55 μg/mL, whereas, even at this higher concentration, no significant bactericidal activity was elicited by UGN(+)GSH200, as shown in Fig. 1a, b.These results revealed the decreased bactericidal performance with the decrease in gold surface availability in both motile and non-motile strains.Therefore, in most of the subsequent experiments, UGN(-)GSH and UGN(+)GSH50 were employed, representing labile and stable categories of UGNs, respectively.Ampicillin was employed as a representative control antibiotic whose IC 50 was found to be 0.35 μg/mL against both microbes (Fig. 1c).

Determination of minimum bactericidal concentration (MBC)
The bacterial cultures of both strains underwent treatment with both categories of UGNs for 6 h of treatment duration, aligning with the period consistent with the Alamar blue assay.Subsequently, the MBC was determined through a 24 h incubation using the standard plating technique (Fig. S2) [29].Similar to the findings with the motile strain from our previous study, the MBC of the non-motile strain also reached at 8 × IC 50 for UGN(-)GSH, while no MBC was achieved for UGN(+) GSH even at a higher concentration of 16 × IC 50 .Despite similar MBC values for both motile and non-motile strains, as shown in Table 1 and Fig. S2, the inhibition of bacterial colonies was slightly higher in the non-motile strain compared to the motile strain for both types of UGNs.Therefore, the higher bacterial growth in the case of the motile strain can be attributed to factors leading to the aggregation of UGNs mediated by flagellar activity.

FACS analysis
Fluorescence-activated cell sorter (FACS) analysis was conducted in a manner similar to that of the motile strain with ampicillin, UGN(-)GSH, and UGN(+)GSH50 at concentrations ranging from IC 50 to 8 × IC 50 (Fig. 2a,  b).This analysis utilizes the most common fluorescent markers in the LIVE/DEAD BacLight stain (SYTO9-PI), which have been extensively tested with microscopy [39,40].The sorting of the cells was as follows: Q1 represented the dead cell population, while Q2 represented the viable, lethally injured, persister, and VBNC cells [29].Both strains exhibited almost identical results, indicating a decrease in viability with an increase in antimicrobial concentration.The viability of the strains when treated with ampicillin at a concentration of 8 × IC 50 was 13.6% and 15.1% for the motile and non-motile strains, respectively.However, the viability was less than 5% in the case of UGNs, signifying an elevated level of cell disruption compared to the conventional antibiotic.There is also a significant decrease in population in the Q2 group when treated with UGNs compared to ampicillin treatment.This indicates that the rate of revival of persister cells is slower after UGN treatment at the given concentration compared to conventional antibiotics.This assumption is further supported by the resuscitation study conducted on both strains.

Resuscitation
A resuscitation study was conducted to understand and differentiate the aggregation of nanosystems and bacterial growth between motile and non-motile strains.Absorbance at 600 nm (OD 600 ) was measured, and bacterial viability was quantified using the Alamar blue assay with varying concentrations of antimicrobial compounds (Fig. 2c, h).In the case of antibiotics, there is no significant difference observed in the resuscitation propensity between the motile and non-motile strains.However, significant variations were noticed when UGNs were employed as the antimicrobials.A near identical behavior was observed between the motile and non-motile strains against UGNs till 12 h.Interestingly, a gradual resuscitation of the cells was observed in the viability study at 24 h with lower concentrations of UGNs, particularly with non-GSH-capped, up to 8 × IC 50 in the case of motile strain, while a near-complete inhibition was still maintained in the non-motile strain starting from as low concentration as 2 × IC 50 .The OD 600 values were particularly helpful in the case of UGNs to discern the viability from the aggregation of UGNs.After 48 h, the bacterial resuscitation behavior was observed to be similar for both motile and non-motile strains against both categories of UGNs.Resuscitation studies help in understanding the probability and rate of revival from an apparently dead quiescent phase, often elicited by dormant phenotypes such as persistent and viable but non-culturable cells (VBNCs).Slow-growing and regulated cells, particularly under adverse environmental conditions, need mechanisms to maintain adequate levels of macromolecules necessary for metabolic activities.This ensures they can perform de novo protein synthesis and regenerate when conditions become favorable [41,42].As a result, mechanisms have evolved to prevent the total depletion of these vital metabolites during dormancy.Our current study shows the early onset of resuscitation by the motile cells against both categories of UGNs can be attributed to the higher aggregation of UGNs mediated by flagella and flagellin protein, but the same did not prevail with the non-motile strain that lacked flagellar activity.Consistent with our previous findings with the motile strain, similar results were obtained for the non-motile strain when treated with UGN(+)GSH100, whose surface gold atoms are covered more with GSH that inherently elicited slower antibacterial activity only at higher concentrations (Fig. S3).Overall, the motile strain exhibited a faster resuscitation than the non-motile strain, particularly with the non-GSH capped ones at lower IC 50 values.This indicates that appropriate nature (stable vs. labile) as well as appropriate concentration of the antimicrobial are critical factors in preventing resuscitation.

Resuscitation through MBC
MBC via the plating method was further done to validate the results obtained from the time-point studies for 6, 12, 24, and 48 h, Table 1 [43].The degree of difference between the motile and non-motile strains followed a similar trend as that of the Alamar blue assay for both types of UGNs.While 16 × IC 50 of non-GSH-capped UGN was completely unsuccessful in preventing the growth of the motile strain from 24 h onwards, the same was found to be successful against the non-motile strain till 48 h.UGN(+)GSH50 exhibited greater effectiveness in preventing resuscitation compared to UGN(-)GSH, corroborating our previous findings with the motile strain.Moreover, this effectiveness was further enhanced in the case of the non-motile strain compared to the motile strain.Consequently, the study indicates that the prevention of resuscitation in the non-motile strain is more achievable than in the motile strain, as summarized in Table 1.

Passaging
The non-motile strain underwent a similar passage regimen as described in our previous study for the motile strain [22].Briefly, both strains were passaged for 30 cycles with intermittent exposure to an inhibitory concentration of UGNs to induce resistance development (Fig. 3a, c).To mitigate the environmental factors, a control group of cells was passaged without any treatment.Results from the non-motile strain were found to follow a similar trend to the motile strain.A delayed resistance development was observed in the UGN(−) GSH group due to burst activity, while early resistance onset was noted from the second passage onwards in the ampicillin and UGN(+)GSH50 groups.The fold increase in inhibitory concentration (~ 3.5 to fourfold) compared to the non-motile strain was similar to that of the motile strain when exposed to UGNs [22].Experimental evidence supported the heritable transfer of resistant genes across generations through vertical and horizontal transfer mechanisms, as demonstrated by the removal and reintroduction of antimicrobial pressure over three subsequent passages [21].
The resistant cell populations were then analyzed for the development of cross-resistance or collateral sensitivity and the results are presented in Fig. 3d, f [44].For instance, the BL21 strain resistant against UGN(-) GSH (UGN(-)GSH-R BL21) is analyzed against UGN(+) GSH-50 and vice versa.Results indicated similar fold increase in cross-resistance and collateral sensitivity with the non-motile and motile strains, when tested against the other UGN composition.Compared to the ancestor strain, these resistant populations exhibited ~ 1.5 to twofold cross-resistance; however, considering their resistant nature to a given UGN category (labile or stable), the values can be regarded as collateral sensitivity.Another intriguing observation is that the UGN-resistant strains overall exhibited a high degree of cross-resistance against ampicillin.Specifically, the motile strain resistant to UGN(+)GSH50 showed a 1.5-fold higher inhibitory concentration compared to the non-motile strain.Though the motile strain resistant to UGN(−)GSH did not exhibit such a behavior, it is presumed that excessive production of virulence proteins like flagellin generated with the stable UGNs may contribute to increased crossresistance via enhanced multimodal resistance pathways.
Many studies have highlighted the antimicrobial effectiveness of PGRE against a wide range of bacteria, including both Gram-positive and Gram-negative strains.Recent research highlights the effectiveness of PGRE in enhancing the susceptibility of E. coli K12 to silver nanoparticles [20,30].This effect is achieved by inhibiting the production of flagellin protein, leading us further to examine the impact of PGRE on resistant bacterial populations from the 34th passage, as the gene sequencing studies showed evidence on the inheritance of resistant genes across generations through both vertical and horizontal transfer mechanisms.The results presented in Fig. S4a, b showed that the antimicrobial effectiveness of non-GSH-capped UGNs significantly improved after a single 24 h treatment with PGRE, while the efficacy of GSH-capped UGNs modestly increased by 40-45%.Interestingly, the IC 50 of ampicillin was found to be decreased than the initial state, signifying the high degree of sensitization caused by the treatment with PGRE.The The statistical significance of the IC 50 study on the 30th passage was obtained from the student's t-test and presented as follows: p ≤ 0.001 is most significant (***), p ≤ 0.01 and > 0.001 is more significant (**), and p ≤ 0.05 and > 0.01 is significant (*)).d-f Cross-resistance evaluation using resazurin assay: The alphabet 'R' in the code represents the resistant population to the particular antimicrobial.A representative code, Amp-R-BL21-UGN(−) GSH, depicts that the ampicillin-resistant BL21(DE3)pLysS population is treated with UGN(-)GSH aforementioned results indicate that PGRE treatment suppressed the production of proteins responsible for evading the antimicrobial threat(shown later).

Gold quantification
The details of the expression levels of selected genes are provided in the Supplementary Material (Fig. 4a, d).ICP-OES analysis was conducted on both the motile and non-motile strains to investigate the distribution of gold in the pellet and supernatant, respectively (Fig. 4e, f ).Despite minor variations, the percentage distribution of gold in the pellet and supernatant exhibited a similar pattern in both strains.Further, a higher gold content was found in the pellet of the resistant populations of both strains, which is suggestive of mutations being responsible for the increased need for antimicrobials to elicit similar antibacterial efficacy.Strikingly, both motile and non-motile strains showed the same pattern in the gold distribution studies.

Aggregation dynamics
UV-visible and DLS measurements were performed with the ancestor and resistant populations to investigate the UGNs aggregation dynamics and the results are presented in Fig. 4g, n [21].The absorbance spectra of the ancestor populations treated with both categories of UGNs showed a symmetrical double-humped profile in the wavelength range of 290 nm to 360 nm, that remained unchanged after filtering the bacterial solution.However, treatment with the resistant population caused a substantial distortion of the double-humped feature, characterized by a red shift in the hump at higher wavelengths and a noticeable broadening.Such an observation is typical of particle aggregation, which results in light scattering across a broad range of wavelengths, leading to the loss of transmittance [45][46][47].The filtration process decreased peak intensity, indicating the removal of UGNs that were previously bound to the microbes.
The hydrodynamic radius of bacterial solutions treated with antimicrobials was dominated by the bacteria (Table S3).However, after filtration, the ancestor population had a lesser size of 20-40 nm, while the resistant population exhibited a size of > 800 nm, indicating microbe-induced aggregation of the UGNs.In brief, a distinct trend in the absorbance profiles and particle sizes was observed between the ancestor and resistant populations; however, these properties were almost similar among the two different UGNs and strains.The increased particle size and absorption in the resistant populations shown earlier suggest that the aggregation of UGNs is responsible for the development of resistance in these microbes, likely due to adhesive proteins, including flagellin [21].

Gene sequencing
Identifying precise mechanisms of resistance development is a complex task owing to the stochastic nature of the evolutionary adaptation of the microbes [48].This is especially true when a resistant subpopulation emerges as a result of antimicrobial pressure, resulting in heteroresistance.In the first report of heteroresistance, Napier et.al. identified an inadvertent cross-resistance of colistin in Enterobacter cloacae against lysozyme, which is the first line of defense mechanism of the host innate immune system [49].Thus, it is evident that single colony passaging strategies run the risk of overlooking a potentially resistant subpopulation, thereby limiting the identification of probable mechanistic pathways among a vast array of possibilities [20,21].Besides, considering the total bacterial population in studies ensures more representative results, as all bacteria are exposed to antimicrobial treatment in practical settings.In our latest study, we employed passaging with the entire bacterial cell population subjected to intermittent treatment, mirroring the passaging approach utilized in our recent publication involving the motile strain.We investigated the genetic basis of bacterial resistance using a next-generation sequencing (NGS) approach across all three categories of cell populations.Notably, within the coding sequences, the genomes of the populations that exhibit resistance, both types of UGNs revealed a significant presence of single nucleotide polymorphisms (SNPs), numbering in the hundreds (~ 160-200), along with a small number of insertions/deletions (in-dels).On the contrary, the number of mutations in the ampicillin-resistant population was in the range of a few tens (~ 30-40).The mutations were sorted according to their frequency of incidence, specifically distinguishing between those associated with UGN(-)GSH or UGN(+)GSH50 and those that were common to both types.To gain a comprehensive understanding, the mutations were subjected to Gene Ontology (GO) analyses to identify the processes that were predominantly affected.The enrichment analysis based on biological processes is given in Fig. 5, while that of cellular component and molecular function analyses are presented in Fig. S6a, b.It can be noted that only the top 10 processes identified based on the enrichment score have been represented in the graphical format, whereas, the complete details are presented in Supplementary excel files.The enrichment score as well as the number of processes impacted were found to be higher in the case of UGNs when compared to ampicillin.This can be attributed to a significantly higher number of gene mutations observed with UGNs, indicating a severe, multi-faceted attack by the antimicrobials.
The enrichment score as well as the number of processes impacted were found to be higher in the case of UGNs when compared to ampicillin.This can be attributed to a significantly higher number of gene mutations observed with UGNs, indicating an intense, multifaceted attack by the antimicrobials.In addition, the categorization of these activities based on gene function and location was manually performed as well as depicted in a  c, d) in the ancestor and resistant populations after treatment with the antimicrobials in comparison to the untreated ancestor populations.The statistical significance of each analysis is obtained from the student's t-test and presented as follows: p ≤ 0.001 is most significant (***), p ≤ 0.01 and > 0.001 is more significant (**), and p ≤ 0.05 and > 0.01 is significant (*).e, f ICP-OES measurements in duplicate for gold quantification in the cell pellet (represented by 'P') and supernatant (designated as 'S') of the populations treated with UGNs.The numbers 0 and 30 in the sample codes represent the bacterial cultures from the 0th and 30th cycles of the passaging studies, respectively.Particle size analyses through DLS measurements g-j and UV-visible absorbance profiles k-n of the ancestor and resistant populations.The zoomed in DLS profiles below 200 nm are presented as inset.The plots (g, i, k, and m) represent DH5α, while (h, j, l, and n) correspond to BL21(DE3)pLysS pie-chart format, revealing a significant influence on genes engaged in cytosolic metabolic and transmembrane modification processes, Fig. 5. Significantly, the number of transmembrane transformations is notably more pronounced with nanosystems in contrast to ampicillin, suggesting that membrane constituents (like membrane proteins, porin channels, and efflux pumps) are actively engaged in resistance.development.While the mutations in the cytosolic, metabolic, and energy metabolism pathways were noticed to be the predominant ones against UGNs in both strains, the effect was significantly higher in the motile strain (43-51%) as opposed to (38%) in the non-motile strain.Though the percentage of mutations in cytosolic pathways in the case of ampicillin was found to be closer to 60%, the net lesser number of mutations in combination with higher resistance indicates an easy resistance development by the microbes against the conventional antibiotic for both strains.
To delve deeper into the particular pathways primarily influenced by the mutations, further analysis was carried out utilizing the STRING software tool, Figs.S7, S13.In this analysis, the mutations common to both categories of UGNs were processed by the software tool and hence are prioritized based on the false discovery rate (FDR) Table S4.The major pathways affected in the non-motile strain included metabolic, glycolysis/ gluconeogenesis, fatty acid degradation, biosynthesis of secondary metabolites, carbon metabolism, RNA degradation, RNA polymerase, aminoacyl-tRNA biosynthesis, 2-oxocarboxylic acid metabolism, and citrate Fig. 5 Gene Ontology (GO) and pathway analyses.The pattern of mutations in antimicrobial-resistant populations categorized through GO analyses (bar chart diagrams) for the biological processes in motile (left panel) and non-motile (right panel) strains.The analyses are sorted based on mutations that are common to both UGN(−)GSH and UGN(+)GSH50) as well as those specific to each.These mutations have been broadly classified into 7 types, namely, transcription, translation, replication, transmembrane modification, oxidative stress, cytosolic metabolic pathways, motility, and others.The corresponding pathways are represented in the pie chart format, in which the color code is followed clockwise cycle while in the motile strain, the affected pathways were detailed in our recent publication [22].

Conclusions
To summarize, the study sheds light on the differential behavior between motile and non-motile strains in their ability to resuscitate and develop resistance when exposed to two categories of UGNs.Our present study demonstrates the antibacterial effects of UGNs against non-motile strains which align consistently with those observed in experiments involving motile strains performed in our previous study.The motile strain in the presence of certain virulent factors like flagellin protein and flagella, exhibits an added advantage by increasing the likelihood of becoming persister/VBNC cells, facilitating resuscitation in stress-free environments.Conversely, non-motile strains, lacking such virulence factors, experience delayed resuscitation.Passaging studies with both motile and non-motile strains of E. coli K12 revealed their capability to develop heritable resistance against UGNs through periodic exposure to IC 50 levels.UV-visible and DLS experiments indicated that protein-mediated aggregation of UGNs significantly contributed to microbial resistance against nano-based antimicrobials with both strains.However, treatment with PGRE showed promise in sensitizing resistant populations to UGNs by suppressing adhesive protein production.Notably, our findings suggest that flagellar motility in motile strains confers fitness advantages, leading to higher resuscitation propensity compared to non-motile strain.Gene sequencing highlights significantly higher mutation rates in motile strains, particularly in cytosolic, metabolic, and energy metabolism pathways.Further studies on cross-resistance confirmed that motile strains resistant to UGN(+) GSH50 exhibit a 1.5-fold increase in IC 50 when subjected to ampicillin compared to non-motile strain.Overall, our study underscores the ease of preventing resuscitation and resistance development in non-motile strain compared to motile strain with the appropriate properties of UGNs.Nevertheless, further detailed studies are required with several more bacterial strains to ascertain the generality in the differential behavior between motile and non-motile strains.

Fig. 1
Fig. 1 Alamar blue assay.The viability of bacteria studied in triplicate for the motile (a) and non-motile (b) strains after treatment with GSH-capped and non-GSH-capped UGNs.c The viability study of both the strains against ampicillin

Fig. 2
Fig. 2 FACS and resuscitation studies.a-b Live-dead analysis (duplicate) of the motile (a) and non-motile (b) strains treated with IC50 to 8 × IC50 of the antimicrobials.c-h Resazurin assay and OD600 measurements of both the strains against different concentrations of ampicillin (c, f), UGN(-)GSH (d, g), and UGN(+)GSH50 (e, h).The codes (D) and (B) given in the parenthesis of the figure legend represent DH5α and BL21(DE3)pLysS, respectively

Fig. 3
Fig. 3 Evolution of resistance.a-c Fold increase in IC 50 of the strains after each passage with UGN(−)GSH (a), UGN(+)GSH50 (b), and ampicillin (c).The statistical significance of the IC 50 study on the 30th passage was obtained from the student's t-test and presented as follows: p ≤ 0.001 is most significant (***), p ≤ 0.01 and > 0.001 is more significant (**), and p ≤ 0.05 and > 0.01 is significant (*)).d-f Cross-resistance evaluation using resazurin assay: The alphabet 'R' in the code represents the resistant population to the particular antimicrobial.A representative code, Amp-R-BL21-UGN(−) GSH, depicts that the ampicillin-resistant BL21(DE3)pLysS population is treated with UGN(-)GSH

Fig. 4
Fig. 4 Reverse transcription-quantitative polymerase chain reaction (RT-qPCR), gold quantification, and aggregation dynamics studies.a-d RT-qPCR analyses in triplicates on the expression levels of genes associated with oxidative stress (a, b) and motility (c, d) in the ancestor and resistant populations after treatment with the antimicrobials in comparison to the untreated ancestor populations.The statistical significance of each analysis is obtained from the student's t-test and presented as follows: p ≤ 0.001 is most significant (***), p ≤ 0.01 and > 0.001 is more significant (**), and p ≤ 0.05 and > 0.01 is significant (*).e, f ICP-OES measurements in duplicate for gold quantification in the cell pellet (represented by 'P') and supernatant (designated as 'S') of the populations treated with UGNs.The numbers 0 and 30 in the sample codes represent the bacterial cultures from the 0th and 30th cycles of the passaging studies, respectively.Particle size analyses through DLS measurements g-j and UV-visible absorbance profiles k-n of the ancestor and resistant populations.The zoomed in DLS profiles below 200 nm are presented as inset.The plots (g, i, k, and m) represent DH5α, while (h, j, l, and n) correspond to BL21(DE3)pLysS

Table 1
Time-kill MBC studies with motile and non-motile strains against different concentrations of UGN(−)GSH and UGN(+)GSH50