Green synthesis of gold-silver nanoparticles using plant extracts and their modulatory effects on the metabolism of human gut symbiont Lactobacillus

2.1. Reagents and plant material preparation

Silver nitrate (AgNO₃, ≥99.9%) and hydrogen tetrachloroaurate (HAuCl₄·3H₂O, ≥99.9%) were purchased from Sinopharm Chemical Reagent Co., Ltd. Methanol, acetonitrile, formic acid, and other solvents for chromatographic analysis were high-performance liquid chromatography (HPLC)-grade and sourced from Shanghai Aladdin Biochemical Technology Co., Ltd. The plant material employed for nanoparticle synthesis was Camellia sinensis (green tea) leaves, harvested from a certified organic farm in Anxi County, Fujian Province. Fresh leaves were air-dried at 40°C for 48 h and subsequently ground to a fine powder using a stainless-steel mill. For extract preparation, 10 g of the powder was refluxed in 200 mL of deionized water at 80°C for 1 h. The mixture was filtered through Whatman No. 1 filter paper, cooled to room temperature, and stored at 4°C for no more than 48 h prior to use. The extract served both as a reducing agent and a stabilizing matrix for nanoparticle formation. The total phenolic content (TPC) of the working extract was determined using the Folin-Ciocalteu colorimetric method following ISO 14502-1; results are reported as mg gallic acid equivalents (GAE) per mL of extract based on a gallic acid calibration (A765). Where indicated, total catechin content was quantified by HPLC-ultraviolet detection (UV) according to ISO 14502-2 (278 nm) as the sum of individual catechins. The pH of the extract at room temperature was recorded prior to synthesis. These analytical characterizations enable conversion of extract volume to phenolic equivalents for reaction stoichiometry.

2.2. Green synthesis of bimetallic Ag/Au nanoparticles

Three distinct bimetallic Ag/Au nanoparticle formulations were synthesized, corresponding to Ag:Au molar ratios of 4:1 (Sample C), 1:1 (Sample D), and 1:4 (Sample E). For each formulation, AgNO₃ and HAuCl₄ were dissolved in deionized water to a total metal-ion concentration of 1 mM (100 mL). The solution was stabilized at 60°C under magnetic stirring (600 rpm). Camellia sinensis extract (25 mL) characterized for total phenolic content (TPC, mg gallic acid equivalent GAE/mL; Section 2.1.1) was added at a controlled rate (≈1 mL min^-1) to define a reproducible phenolic-to-metal input; extract pH was recorded prior to addition. The mixture was maintained at 60°C for 2 h in amber glassware to limit photochemical effects. A similar method has been used for synthesizing AgNPs (Sample A) and AuNPs (Sample B).

A visible color change from pale yellow to reddish-brown indicated successful nanoparticle formation. Upon completion of the reaction, the suspensions were centrifuged at 12,000 rpm for 20 min (Beijing LuXiangyi Centrifuge Instrument Co., Ltd.), and the resulting pellets were washed thrice with deionized water to remove unreacted phytochemicals and free ions. The purified nanoparticles were freeze-dried using a vacuum lyophilizer (SCIENTZ-18ND, Ningbo Scientz Biotechnology Co., Ltd.) and stored at –20°C in sealed glass vials. A step-wise schematic of extraction, synthesis, purification, and storage has been provided in Scheme 1.

Close

Scheme 1.

Green synthesis workflow for Ag/Au nanoparticles using Camellia sinensis extract.

Export to PPT

2.3. Lactobacillus culture and exposure conditions

Lactobacillus gasseri CGMCC 1.3724 was obtained from the China General Microbiological Culture Collection Center (CGMCC). The strain was cultivated in de Man, Rogosa, and Sharpe (MRS) broth (Hopebio, Qingdao, China) under anaerobic conditions at 37°C using a Whitley A35 anaerobic workstation (Microbiology International, China). Log-phase cultures were diluted to an optical density at 600 nm (OD₆₀₀) of 0.1 before experimental treatments.

Nanoparticle stock solutions (1 mg mL^-1) were prepared by resuspending lyophilized powders in sterile phosphate-buffered saline (PBS, pH 7.4), followed by sonication for 10 minutes to minimize aggregation. Working concentrations of 10, 50, and 100 µg mL^-1 were freshly prepared and filtered (0.22 µm) before use. Lactobacillus cultures were incubated with each nanoparticle formulation (A, B, C) for 12 and 24 h, respectively, with untreated cells serving as controls. Each treatment was conducted in triplicate.

2.4. Growth analysis and cell viability

Growth kinetics were monitored by measuring OD₆₀₀ at hourly intervals using a BioTek Synergy multimode plate reader. For viability assessment, a Live/Dead BacLight Bacterial Viability Kit (Thermo Fisher Scientific, China) was employed according to the manufacturer’s instructions. Fluorescence was observed using a Leica DMi8 inverted fluorescence microscope, and the percentage of viable cells was quantified using ImageJ software.

2.5. Metabolite profiling via GC-MS and HPLC-MS

Fecal-like metabolite extraction was simulated by centrifuging exposed cultures at 8,000 rpm for 10 minutes, followed by resuspension in cold methanol:water (1:1, v/v) solution. Metabolites were extracted via ultrasonication (30 min, 4°C) and filtered through 0.22 µm membranes prior to analysis.

Short-chain fatty acids (SCFAs), including acetate, propionate, and butyrate, were quantified using a Thermo TRACE 1310 GC coupled with a TSQ 8000 EVO MS (Thermo Fisher Scientific, China) equipped with a TG-WAXMS (polyethylene glycol) capillary column (TG-WAXMS) column (30 m × 0.25 mm × 0.25 µm). The oven temperature was programmed from 80°C (2 min) to 200°C at 10°C min^-1. The injector and detector temperatures were maintained at 250°C. Calibration curves were established using external standards.

For broader metabolomic profiling, a Dionex Ultimate 3000 UHPLC system coupled with a Q Exactive Plus Orbitrap MS (Thermo, China) was used in both positive and negative electrospray ionization modes. Chromatographic separation was achieved on a Hypersil GOLD C18 column (2.1 × 100 mm, 1.9 µm) at 40°C with a flow rate of 0.3 mL min^-1. Mobile phases consisted of 0.1% formic acid in water (A) and in acetonitrile (B). Raw liquid chromatography-mass spectrometry (LC-MS) data were processed in Compound Discoverer 3.2 (Thermo Scientific) for peak detection, alignment, and gap filling. Feature intensities were first scaled to the corresponding internal standards and then normalized using probabilistic quotient normalization (PQN) to mitigate sample-to-sample dilution effects, followed by log₁₀ transformation and Pareto scaling prior to statistical analysis. A pooled quality-control (QC) sample, prepared by combining equal aliquots of study extracts, was injected at the start and then at regular intervals (∼every 10 injections) to monitor analytical stability; when temporal drift was detected, features were corrected using QC-based robust locally estimated scatterplot smoothing (LOESS) signal correction (QC-RLSC) [27].

2.6. Microbial composition analysis via 16S rDNA sequencing

Total genomic DNA was extracted from exposed bacterial pellets using a TIANamp Bacteria DNA Kit (Tiangen Biotech, Beijing, China). The V3–V4 regions of 16S rRNA genes were amplified using the universal primers 341F and 806R and sequenced on an Illumina MiSeq platform (Guangzhou Gene Denovo Biotechnology Co., Ltd.).

Raw reads were quality-filtered using QIIME 1.9.1. Chimeric sequences were removed with USEARCH, and operational taxonomic units (OTUs) were clustered at 97% identity using Uparse. Taxonomic classification was performed against the SILVA 132 database. α- and β-diversity indices were calculated, and differential abundance analysis was conducted using the Linear discriminant analysis effect size (LEfSe) algorithm with an Linear discriminant analysis (LDA) cutoff score of 3.5.

2.7. Statistical analysis

Unless otherwise stated, all experiments were performed with n = 3 independent biological replicates per condition. Continuous outcomes are presented as mean ± standard deviation (SD); figures display error bars as SD, with individual data points overlaid when feasible. Normality (Shapiro–Wilk) and homogeneity of variance (Levene) were assessed. For growth curves (optical density (OD)₆₀₀ vs time), treatment × time effects were tested using two-way repeated-measures ANOVA with the Geisser–Greenhouse correction, followed by Tukey’s post hoc comparisons; where normality assumptions were violated, a Friedman test with Dunn’s post hoc test was used. For single-time-point endpoints (colony-forming units (CFU), short-chain fatty acids (SCFAs), lipopolysaccharide (LPS), malondialdehyde (MDA), reduced glutathione (GSH), superoxide dismutase (SOD)), one-way ANOVA with Tukey’s post hoc test was applied (or Kruskal–Wallis with Dunn’s post hoc test if non-normal). For 16S rDNA data, α-diversity differences (Shannon, Chao1) were evaluated by Wilcoxon rank-sum tests; β-diversity differences (Bray–Curtis) were assessed by PERMANOVA (999 permutations). Differential genus abundances used Wilcoxon tests with Benjamini–Hochberg false discovery rate (FDR) adjustment; q-values are reported.

3.1. Material characterization

The progression of nanoparticle formation via Camellia sinensis extract was monitored over a 5-h period using UV–visible spectroscopy. The reaction kinetics and optical properties of the nanoparticles were strongly influenced by the metal composition and ratio of silver to gold precursors. Figure 1 presents the time-resolved UV–Vis absorption spectra of monometallic AgNPs (A), AuNPs (B), and bimetallic Ag/Au nanoparticles with molar ratios of 4:1 (C), 1:1 (D), and 1:4 (E). In the case of AgNPs (Figure 1a), a characteristic surface plasmon resonance (SPR) peak emerged at ∼440 nm [28,29], which increased in intensity over the first 2 h and plateaued by the 3-h mark [30]. The delayed formation and the time-dependent sharpening of the SPR band reflect the slower reduction kinetics of Ag⁺ ions under plant-extract-mediated conditions. The reaction medium transitioned visually from colorless to pale yellow-brown, consistent with Ag⁰ formation [31]. In contrast, AuNP synthesis (Figure 1b) was markedly faster [32]. The SPR peak at ∼545 nm, characteristic of spherical AuNPs, appeared within the first 10 min post-extract addition and stabilized in intensity and position within 1 h [33]. The color transition from yellow to ruby red correlated with rapid Au³⁺ reduction, likely facilitated by the high antioxidant activity and polyphenolic content of the Camellia sinensis extract.

Close

Figure 1.

(a) UV-visible spectra of AgNPs, (b) AuNPs, and (c–e) bimetallic Ag/Au NPs synthesized with Camellia sinensis extracts as a function of reaction time. Samples correspond to Ag:Au molar ratios of (c) 4:1, (d) 1:1, and (e) 1:4. Spectra were recorded at time points from 5 min to 5 h after extract addition.

Export to PPT

The bimetallic Ag/Au systems exhibited distinct spectral behavior depending on the Ag: Au molar ratio. For the Ag/Au (4:1) mixture (Figure 1c), no clear Au SPR peak at 545 nm was observed. Instead, a broad absorbance profile resembling AgNP spectra dominated, with maximum absorbance near 440 nm. The absence of distinct gold-associated peaks suggests either slow Au³⁺ reduction or the formation of Ag-core/Au-shell architectures in which the gold shell suppresses or shifts its characteristic plasmonic resonance [34]. For the 1:1 and 1:4 Ag/Au formulations (Figure 1d and e), the absorbance curves demonstrated more defined single SPR bands centered near 525–530 nm, with increasing intensity and red shift over time [35]. Notably, these bands did not show the dual-peak signature expected from physical mixtures of AgNPs and AuNPs, indicating that alloyed or core–shell structures had formed rather than segregated nanoparticles [36]. The faster and stronger response observed in the Au-rich samples further supports the hypothesis that gold nucleation dominates under high Au³⁺ concentrations, while silver plays a more passive or shell-forming role [37]. Over the 5-h observation window, no significant blue or red shift was observed after the 3-h mark for any sample, indicating completion of the reduction process and stabilization of nanoparticle morphology. The observed kinetics and spectral profiles are consistent with previously reported green syntheses where Au³⁺ is preferentially and rapidly reduced compared to Ag⁺ under identical phytochemical conditions [38].

Taken together, the UV–Vis spectra suggest that particle formation was composition-dependent: Ag-rich samples favored delayed growth with Ag-dominated spectral signatures, while Au-rich systems showed rapid formation of gold-like or alloy-type particles. The presence of a single, stable SPR peak in the bimetallic systems supports the formation of either homogenous alloyed NPs or core–shell morphologies with optically dominant outer gold layers, aligning with prior mechanistic suggestions from plant-mediated synthesis studies.

To elucidate the functional groups responsible for the reduction and stabilization of synthesized nanoparticles, Fourier transform infrared spectroscopy (FTIR) spectroscopy was performed on dried AgNPs, AuNPs, and three bimetallic Ag/Au formulations (Ag:Au = 4:1, 1:1, and 1:4). The resulting spectra have been presented in Figure 2(a). Several prominent absorption bands were observed across all samples, indicating the involvement of phytochemicals, predominantly polyphenols, flavonoids, and polysaccharides, from the Camellia sinensis extract in mediating nanoparticle formation [39]. In the spectrum of AgNPs, major bands were detected at 3281 cm⁻¹ (O–H stretching), 2924 and 2850 cm⁻¹ (C–H stretching), 1631 cm⁻¹ (C=O stretching or aromatic ring vibrations), 1512 cm⁻¹ and 1441 cm⁻¹ (N–H bending and aromatic C=C stretching), 1235 cm⁻¹ (C–N stretching of proteins), 1062 cm⁻¹ (C–O–C vibration in polysaccharides), and 604 cm⁻¹ (C–H out-of-plane bending) [40]. These features indicate that hydroxyl, carbonyl, and amine-containing biomolecules participated in metal ion reduction and subsequent nanoparticle capping. The AuNPs spectrum (Figure 2b) showed similar peak positions, with slightly higher intensity in the O–H and C=O regions, including a broad peak around 3410 cm⁻¹ and sharp bands at 1636 and 1510 cm⁻¹ [41,42]. These findings suggest stronger or more extensive interactions between Au³⁺ ions and polyphenolic ligands compared to Ag⁺ ions. Additionally, the fingerprint region (<800 cm⁻¹) showed reduced complexity, which may reflect the higher uniformity and stability of gold nanoparticle capping layers.

Close

Figure 2.

(a) FTIR spectroscopy analysis and (b) XRD patterns of AgNPs, AuNPs, and bimetallic Ag/Au NPs synthesized with Camellia sinensis extracts.

Export to PPT

For the bimetallic nanoparticles, FTIR patterns varied subtly depending on composition: Sample C (Ag:Au = 4:1) displayed a spectral profile closely resembling AgNPs, with characteristic peaks at 3287, 1630, 1442, 1233, and 1061 cm⁻¹ [43]. The prominent features in the polysaccharide and protein regions (1200–1000 cm⁻¹) indicate the dominant interaction with silver species [44], possibly forming Ag-core-rich nanostructures [45]. Sample D (Ag:Au = 1:1) exhibited overlapping peaks from both AgNPs and AuNPs, with the most distinct transmittance bands at 3402, 2920, 2851, 1634, 1513, 1386, 1230, and 1064 cm⁻¹ [46]. The presence of shared functional groups across all nanoparticle types reinforces the hypothesis that Camellia sinensis phytochemicals reduce both metal ions simultaneously [47]. The slight shifts in peak positions may indicate coordination interactions with bimetallic interfaces or alloy surfaces [48,49]. Sample E (Ag:Au = 1:4) showed a dominant Au-like spectrum [50], with key features at 3409, 1635, 1509, and 1154 cm⁻¹ [51]. The intensities of bands related to protein and polysaccharide vibrations (e.g., 1060–1000 cm⁻¹ region) were comparatively reduced [52], suggesting lower organic coverage or a more compact capping structure typical of gold-dominated systems [53].

Overall, the FTIR spectra confirm that multiple classes of biomolecules—especially polyphenols (aromatic O–H and C=C), proteins (amide I/II bands), and sugars (C–O–C, C–H)—are involved in nanoparticle formation and stabilization. The conservation of specific spectral features across all samples suggests that reduction and capping processes follow a common mechanism irrespective of the metal composition, although differences in peak sharpness and position imply structural rearrangements in the organic shell depending on the Ag/Au ratio. These findings support the proposed green synthesis pathway in which Camellia sinensis extract serves as both reducing agent and stabilizer, enabling tunable nanoparticle production via metal ratio modulation.

To confirm the crystalline nature and phase structure of the synthesized nanoparticles, X-ray diffraction (XRD) analysis was conducted on freeze-dried AgNPs, AuNPs, and three bimetallic Ag/Au nanomaterials. The resulting diffraction patterns have been presented in Figure 2b. All samples exhibited characteristic peaks corresponding to the face-centered cubic (FCC) lattice structures of elemental silver and gold.

For AgNPs, four distinct Bragg reflections were observed at 2θ values of approximately 38.37°, 44.89°, 64.97°, and 77.73°, which can be indexed to the (111), (200), (220), and (311) planes of metallic silver Joint Committee on Powder Diffraction Standards, no need expand (JCPDS card no. 04-0783). The dominant intensity of the (111) peak suggests preferential orientation along this crystallographic plane, commonly associated with energetically favorable growth in Ag nanocrystals. The AuNPs pattern showed comparable reflections at 2θ = 38.52°, 44.91°, 65.05°, and 77.97°, indexed to the same planes of face-centered cubic (FCC) gold (JCPDS card no. 04-0784). The sharper and more intense diffraction peaks indicate a higher degree of crystallinity and relatively larger grain size in AuNPs compared to AgNPs [54], in agreement with visual and UV–vis observations [55].

For the bimetallic Ag/Au nanoparticles, Sample C displayed peak positions closer to metallic silver, with broadened reflections and slightly lower intensity, possibly indicating strain or defect incorporation during co-reduction [56]. Sample D showed merged and broadened peaks between those of pure Ag and Au, indicative of a homogeneous alloy structure with intermediate lattice constants. Sample E exhibited peak positions that closely matched AuNPs, with enhanced (111) plane intensity, suggesting gold-rich composition or potential formation of Au-shell-dominated structures.

No additional impurity peaks or secondary phases were detected, supporting the successful synthesis of pure metallic nanocrystals [57]. Using the Debye–scherrer equation, the estimated crystallite sizes were 18.4 nm (AgNPs), 22.7 nm (AuNPs), and 16.2, 14.1, and 13.3 nm for Samples C, D, and E, respectively. The decreasing size trend in bimetallic samples may be attributed to nucleation dynamics and alloying effects during bioreduction.

Scanning electron microscopy (SEM) images (Figure 3a-c) showed uniformly distributed nanoparticles with predominantly spherical morphology across all three formulations. Sample C particles appeared slightly aggregated with an average diameter of ∼25–30 nm, while Sample E displayed more monodisperse and discrete particles around 15–20 nm in size. Transmission electron microscopy (TEM) micrographs (Figure 3d-f) provided further insight into particle shape and internal structure. Samples C and E exhibited lattice fringes with d-spacing values of ∼0.235 nm, indicative of (111) planes of FCC Au–Ag alloy. No core-shell contrast was observed, corroborating alloy-type bimetallic structures.

Close

Figure 3.

SEM and TEM analysis of monometallic and bimetallic nanoparticles synthesized using Camellia sinensis extracts. (a–c) SEM images of Sample C (Ag:Au = 4:1), Sample D (1:1), and Sample E (1:4), showing surface morphology and size distribution. (d–f) Corresponding TEM images displaying spherical shape, lattice fringes, and crystalline structure. High-resolution TEM revealed an interplanar spacing of ∼0.235 nm, corresponding to the (111) planes of FCC Ag/Au alloy. No core–shell contrast was observed, indicating homogenous alloy formation.

Export to PPT

Energy-dispersive X-ray spectroscopy (EDS) elemental mapping (Figure 4) confirmed the homogeneous distribution of Ag and Au throughout the particles. For Sample C, Ag accounted for ∼78 wt% and Au for ∼22 wt% [58]; Sample D had nearly equal proportions (∼49 wt% Ag, ∼51 wt% Au); Sample E showed a dominant Au content (∼81 wt%) with a minor Ag component (∼19 wt%). No significant signals from other elements were detected, indicating effective purification [59]. These data validate the molar ratios used during synthesis and support successful alloy formation [60].

Close

Figure 4.

EDS elemental mapping of bimetallic Ag/Au nanoparticles synthesized with Camellia sinensis extracts. (a) Sample C (Ag:Au = 4:1), (b) Sample D (1:1), and (c) Sample E (1:4).

Export to PPT

To further validate the morphological and crystallographic integrity of the synthesized bimetallic Ag/Au nanoparticles, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and high-resolution transmission electron microscopy (HRTEM) were employed (Figure 5). These advanced imaging techniques offer atomic-scale resolution, enabling detailed investigation of lattice structure, elemental homogeneity, and crystallographic phase. Figure 5(a) shows a representative HAADF-STEM image of a single bimetallic nanoparticle, appearing as a bright, spherical entity with uniform contrast [61,62]. The even brightness throughout the particle suggests a homogenous composition, with no visible core–shell boundaries or phase separation. This supports the alloyed nature of the Ag/Au nanostructures, as previously inferred from UV–vis, EDS, and X-ray photoelectron spectroscopy (XPS) analyses [63,64]. The enlarged HRTEM image in Figure 5(b) highlights well-resolved lattice fringes across the nanoparticle surface [65,66]. Fast Fourier transform (FFT) of the HRTEM image (Figure 5(c)) reveals distinct diffraction spots corresponding to the (111) and (200) planes, characteristic of a FCC structure common to both elemental Ag and Au [67]. These reflections are aligned with those observed in XRD patterns, reinforcing the presence of highly crystalline domains [68,69]. Figure 5(d) provides a direct measurement of the lattice fringe spacing, with an interplanar distance of 2.3 Å. This value is intermediate between that of pure Ag (2.36 Å for (111)) and pure Au (2.31 Å for (111)), further confirming the formation of a substitutional Ag–Au alloy. No signs of dislocations, stacking faults, or twin boundaries were detected, indicating excellent structural order. Collectively, these electron microscopy results provide strong evidence for the uniform alloyed structure and high crystallinity of the biosynthesized Ag/Au nanoparticles. The structural integrity and atomic-level homogeneity are expected to contribute significantly to the physicochemical stability and consistent biological activity observed across assays.

Close

Figure 5.

(a) High-angle annular dark-field (HAADF) STEM image of a well-defined spherical nanoparticle. The red solid lines indicate the region selected for HRTEM analysis. (b) Magnified HRTEM image acquired from the region marked in panel (a). The red dashed square highlights the area used for fast Fourier transform (FFT) analysis and lattice fringe measurement. (c) Fast Fourier transform (FFT) pattern obtained from the selected region in panel (b), which can be indexed to the (111) and (200) crystallographic planes of a face-centered cubic (FCC) lattice. (d) High-resolution lattice image corresponding to the red dashed square in panel (b). The red solid lines denote the measured interplanar spacing of approximately 2.3 Å, consistent with the (111) planes of the FCC structure.

Export to PPT

XPS was utilized to investigate the surface elemental composition and oxidation states of the Ag/Au bimetallic nanoparticles synthesized using Camellia sinensis extract. The survey spectra confirmed the presence of Ag, Au, C, and O (Figure 6a), with no detectable signals for foreign contaminants, supporting the high purity of the nanoparticle surfaces. High-resolution spectra of the Au 4f and Ag 3d regions provided detailed insight into the valence states of the metallic components. The Au 4f region (Figure 6b) showed two strong, symmetric peaks centered at 84.0 eV (Au 4f₇/₂) and 87.7 eV (Au 4f₅/₂), which are characteristic of metallic gold (Au⁰) [70]. Deconvolution of these peaks revealed a dominant contribution from Au⁰, with minor shoulders attributed to trace amounts of Au¹⁺ (∼85.1 and ∼88.3 eV), likely arising from surface-bound Au–O or Au–S interactions with capping ligands [71,72]. However, no signals associated with Au³⁺ (∼86–89 eV range) were detected, suggesting complete reduction of HAuCl₄ during green synthesis. Similarly, the Ag 3d spectrum (Figure 6c) exhibited two peaks located at 368.2 eV (Ag 3d₅/₂) and 374.2 eV (Ag 3d₃/₂), assigned to metallic Ag⁰ [73,74]. The absence of satellite features or shifted peaks indicates minimal surface oxidation [75,76]. Deconvolution confirmed that over 95% of the detected silver existed in the zero-valent state. Minor components detected between 367.5–368.0 eV may correspond to Ag–O or Ag–S bonds formed with phytochemicals, but their relative intensities were negligible.

Close

Figure 6.

(a) XPS analysis of AgNPs, AuNPs, and bimetallic Ag/Au NPs synthesized with Camellia sinensis extracts. (b) Core level of Au_4f and (c) Ag_3d of Ag/Au NPs.

Export to PPT

Table 1 summarizes the key physicochemical parameters of the three nanoparticle formulations. Sample C exhibited the largest average hydrodynamic diameter (∼32.8 nm) and the highest negative ζ-potential (–28.4 mV), indicative of good colloidal stability. Sample E had the smallest particle size (∼18.5 nm) and moderate ζ-potential (–21.7 mV). The specific surface area, determined by BET analysis, increased from 27.4 m² g^-1 in Sample C to 39.2 m² g^-1 in Sample E, consistent with the decreasing particle size trend.

3.2. Antimicrobial effects on Lactobacillus strains

To evaluate the biological effects of the green-synthesized Ag/Au NPs on beneficial gut microbiota, Lactobacillus gasseri was selected as a representative model strain due to its well-established role in gastrointestinal homeostasis. The temporal growth dynamics of L. gasseri were monitored over 24 h following exposure to each NP formulation at a concentration of 50 µg mL^-1. Figure 7 illustrates the OD₆₀₀-based growth profiles for all experimental and control groups.

Close

Figure 7.

Growth curves of Lactobacillus gasseri exposed to Ag/Au nanoparticles (50 µg/mL). Points show mean OD₆₀₀ ± SD (n = 3 independent cultures) at 1-h intervals. Statistics: two-way repeated-measures ANOVA (treatment × time) with Geisser–Greenhouse correction; Tukey post hoc comparisons across treatments at each timepoint; significance markers indicate p < 0.05 vs control unless otherwise noted. Inset: area-under-the-curve (AUC) ± SD over 0–24 h.

Export to PPT

The untreated control group demonstrated typical sigmoidal growth kinetics [77], entering exponential phase at 4 h [78] and reaching stationary phase by 14 h [79,80]. In contrast, Sample C (Ag-dominant) significantly inhibited bacterial growth, with OD₆₀₀ values plateauing below 0.25, indicating an early onset of growth arrest [81]. Sample D caused a moderate delay in the log phase, with final OD₆₀₀ reaching ∼0.45 [82], while Sample C (Au-rich) displayed only a slight deviation from the control, with OD₆₀₀ approaching ∼0.62 [83].

These data suggest that the antimicrobial potency of Ag/Au NPs is directly proportional to silver content [84], consistent with known bactericidal mechanisms of Ag⁰, including membrane disruption [85,86], oxidative stress induction [87], and metabolic interference. The relative biocompatibility of Sample E implies a more inert surface chemistry associated with gold-rich compositions [88,89].

To quantify the impact of nanoparticle exposure on culturability, treated L. gasseri suspensions were plated on MRS agar and incubated for 24 h. Colony-forming units (CFUs) were counted and have been summarized in Table 2. Sample C resulted in a 92% reduction in CFUs compared to the control, while Samples D and E yielded reductions of 61% and 23%, respectively. The trend parallels OD₆₀₀ and Confocal laser scanning microscopy (CLSM) findings, providing robust multi-modal evidence for silver-driven antimicrobial activity.

3.3. Effects on metabolic activity

Beyond viability and culturability, it was essential to assess whether sub-lethal concentrations of Ag/Au NPs influenced the metabolic output of L. gasseri. This was particularly important given the role of bacterial metabolites, such as SCFAs, organic acids, and amino acids, in host signaling, pH modulation, and microbial cross-feeding. Metabolomic profiling was performed using high-resolution HPLC-MS after 24-h NP exposure (50 µg mL^-1). Prior to statistical analysis, feature tables were normalized by internal-standard scaling and probabilistic quotient normalization (PQN), then log10-transformed and Pareto-scaled. Pooled QC samples were used to evaluate analytical stability across the run; when minor signal drift was detected, QC-based robust LOESS signal correction (QC-RLSC) was applied to affected features.

Figure 8(a) presents a PCA score plot generated from normalized metabolite intensities [90]. The control group clustered tightly, reflecting metabolic homogeneity [91]. Sample C-treated cells formed a distinct cluster far from the control, indicating profound metabolic perturbation [92]. Sample D showed partial overlap with both control and Sample C, while Sample E remained closely aligned with the control group [93]. The spatial separation of groups in PCA space confirms that Ag/Au NPs, especially those with higher Ag content, significantly modulate bacterial metabolic activity [94].

Close

Figure 8.

Metabolomic analysis of Lactobacls &aseri exposed to Ag/Au NPs (50 ug ml-1), (a) PCA score plot showing sample clusterins; (b) Volcano plot highlighting significant metabolic alterations in Sample C; (c) Heatmap ofton 50 metabolites across treatments.

Export to PPT

A heatmap of the top 50 metabolites (Figure 8b) further illustrated differences in metabolomic fingerprints [95]. Sample C exhibited marked downregulation of key fermentation byproducts, including lactic acid, acetic acid, and succinate. Sample D showed moderate reductions in amino acid-derived metabolites such as tryptophan and histidine [96,97], while Sample E’s profile was largely preserved, with only minor fluctuations [98]. Hierarchical clustering based on Euclidean distance further segregated Sample C from other groups, reinforcing its metabolic divergence [99]. This suggests that Ag-induced stress suppresses central carbon metabolism and biosynthetic pathways in L. gasseri.

Volcano plot analysis (Figure 8c) identified statistically significant (p < 0.05, fold change ≥ 1.5) metabolites altered by NP exposure. In Sample C, 34 metabolites were downregulated and 11 upregulated, including biomarkers linked to oxidative stress (e.g., oxidized glutathione) and cell wall degradation (e.g., N-acetylglucosamine) [100]. Sample D exhibited 18 differential metabolites [101], while Sample E showed only 5, none of which were SCFA-related.

3.4. Gut bacterial composition alterations

To further investigate whether Ag/Au nanoparticles exert indirect effects on microbial ecology beyond a single species model, we extended the analysis to mixed-culture systems simulating simplified gut microbial communities. 16S rDNA sequencing was used to determine community-level changes in bacterial composition after 24-h exposure to the three NP formulations. This analysis allowed us to assess whether sublethal stress imposed by the nanoparticles could selectively influence microbial diversity, relative abundance, and taxonomic balance.

Alpha diversity was evaluated using Shannon and Chao1 indices (Figure 9a). The untreated control group exhibited a high level of species richness and evenness (Shannon index = 3.45 ± 0.07; Chao1 = 128 ± 5), consistent with a stable mixed-culture community derived from Lactobacillus, Bifidobacterium, Turicibacter, and Enterococcus strains.

Close

Figure 9.

Microbial composition after 24-h exposure to Ag/Au NPs. (a) Relative genus-level abundance (stacked bars; per-sample values shown). (b) α-diversity indices (Shannon, Chao1), mean ± SD (n = 3). Statistics: Wilcoxon rank-sum tests vs control with Benjamini–Hochberg FDR adjustment; q-values reported next to brackets.

Export to PPT

Sample C exposure led to a significant reduction in both diversity indices (Shannon = 2.11 ± 0.09; Chao1 = 72 ± 6), suggesting substantial taxonomic contraction, likely driven by silver-induced cytotoxicity. Sample D caused a moderate decline (Shannon = 2.98; Chao1 = 104), while Sample E showed no statistically significant changes (Shannon = 3.39; Chao1 = 124).

The relative abundance of bacteria at the genus level is presented in Figure 9(b). The control group exhibited dominance of Lactobacillus (42.3%), followed by Bifidobacterium (21.7%), Turicibacter (15.1%), and Enterococcus (12.8%). In Sample C-treated cultures, Lactobacillus abundance decreased to 11.4%, while Turicibacter and Enterococcus populations were nearly eradicated. In contrast, opportunistic taxa such as Streptococcus and Escherichia emerged, comprising over 25% of total reads—indicative of competitive niche takeover following commensal depletion. Sample D showed partial depletion of Lactobacillus (27.6%) and a slight increase in Escherichia (6.4%), suggesting transitional dysbiosis. Notably, Bifidobacterium abundance was relatively preserved across all conditions. Sample E maintained microbial profiles closely resembling the control, affirming its compatibility with core commensal taxa. These observations suggest that the composition of Ag/Au NPs not only determines their direct antibacterial potency, but also their ecological influence on microbial balance, a critical factor for potential oral or dietary applications.

To identify taxonomic biomarkers altered by nanoparticle exposure, linear discriminant analysis effect size (LEfSe) was employed with an LDA cutoff of 3.5. The resulting LDA scores have been illustrated in Figure 10. Sample C was enriched with stress-tolerant genera such as Streptococcus and Clostridium sensu stricto, while depletions were noted for Turicibacter, Enterococcus, and Lactobacillus [102,103]. Sample D yielded similar, though less dramatic, shifts [104]. No significant biomarkers were detected in Sample E-treated cultures, further corroborating its minimal ecological impact. LEfSe thus highlights the potential of Ag-rich nanoparticles to selectively alter the gut microbial composition in favor of opportunistic or pro-inflammatory taxa, a finding of concern for long-term exposure or ingestion-related applications.

Close

Figure 10.

LEfSe analysis identifying taxonomic biomarkers significantly enriched or depleted by nanoparticle exposure. For pairwise differential genus abundance, Wilcoxon rank-sum tests were performed with Benjamini–Hochberg FDR correction.

Export to PPT

Table 3 lists the genera showing statistically significant changes in relative abundance (p < 0.05, Wilcoxon rank-sum test) across treatments. The largest shifts were observed in Sample C, affecting both dominant and minor taxa. Of particular interest was the reduction in Turicibacter and Lactobacillus, known producers of short-chain fatty acids and contributors to mucosal barrier function.

3.5. Short-chain fatty acids and lipopolysaccharide levels

SCFAs, including acetate, propionate, and butyrate, are key microbial metabolites that regulate gut barrier function, mucosal immunity, and energy homeostasis [105,106]. Given the observed taxonomic and metabolic disruptions induced by Ag/Au nanoparticles, particularly in silver-rich formulations, we quantified SCFA concentrations in bacterial culture supernatants using gas chromatography–mass spectrometry (GC-MS). Concurrently, extracellular levels of LPS, a potent pro-inflammatory endotoxin released by Gram-negative bacteria, were measured to assess inflammatory potential [107,108].

Figure 11(a) illustrates the SCFA profiles of the untreated control and NP-treated groups after 24 h. The control group showed robust production of acetate (5.71 ± 0.34 mM), propionate (1.63 ± 0.12 mM), and butyrate (0.94 ± 0.09 mM), consistent with healthy metabolic activity of Lactobacillus, Turicibacter, and Bifidobacterium species.

Close

Figure 11.

(a) SCFA concentrations and (b) extracellular LPS after 24-h exposure to Ag/Au NPs (50 µg/mL). Bars show mean ± SD (n = 3). Statistics: one-way ANOVA with Tukey’s post hoc test across treatments (or Kruskal–Wallis with Dunn’s post hoc test if non-normal); exact p or q values annotated.

Export to PPT

Sample C exhibited a sharp decline in all SCFA species, with acetate, propionate, and butyrate reduced by 67.2%, 58.9%, and 73.4%, respectively, compared to the control. These reductions likely reflect both the depletion of SCFA-producing taxa and direct inhibition of fermentation pathways by silver stress. Sample D showed intermediate decreases (30–40% range), while Sample E maintained SCFA levels statistically similar to the control (p > 0.05), indicating preservation of fermentation capability.

Table 4 provides the absolute concentrations and percent changes for each SCFA across treatments. Notably, the acetate:propionate ratio remained stable in samples D and E but was altered in Sample C, suggesting broader dysregulation of metabolic fluxes.

To simulate downstream inflammatory consequences of microbial disruption, LPS concentrations were measured in filtered supernatants using a chromogenic Limulus Amebocyte Lysate assay (LAL) assay. As shown in Figure 11(b), LPS levels were significantly elevated in Sample C (1.43 ± 0.18 EU mL^-1), representing a 2.8-fold increase relative to the control (0.51 ± 0.07 EU mL^-1). This increase is likely due to lysis of Gram-negative bacteria (e.g., Escherichia) and subsequent endotoxin release. Sample D also showed a moderate elevation in LPS (0.92 ± 0.13 EU mL^-1), while Sample E did not differ significantly from baseline. These data suggest that Ag/Au NPs, particularly Ag-rich types, may promote a low-grade inflammatory state by reshaping microbial populations and increasing LPS bioavailability.

The inverse relationship between SCFA levels and LPS accumulation further supports the hypothesis that Ag-driven microbiota perturbation reduces beneficial anti-inflammatory metabolites while elevating pro-inflammatory components. This imbalance could have downstream consequences if applied in vivo, particularly in contexts of intestinal inflammation or dysbiosis.

3.6. Oxidative stress and inflammatory biomarkers

Given the observed nanoparticle-induced disruptions in Lactobacillus viability, metabolite output, and microbial ecology, we further examined cellular oxidative stress and inflammatory signaling as potential mechanistic pathways underlying these biological responses. Specifically, intracellular malondialdehyde (MDA), reduced glutathione (GSH), and superoxide dismutase (SOD) activity were measured as canonical markers of oxidative damage and defense.

Figure 12 summarizes the levels of oxidative stress biomarkers in L. gasseri cells after 24-h exposure to Ag/Au NPs (50 µg mL^-1). The untreated control group showed baseline MDA levels of 1.22 ± 0.13 nmol mg^-1 protein, GSH at 14.7 ± 0.8 µmol mg^-1 protein, and SOD activity at 105.3 ± 5.4 U mg^-1 protein. Sample C-treated cells exhibited significant oxidative damage, with MDA levels rising to 3.58 ± 0.21 nmol/mg (2.9-fold increase), accompanied by GSH depletion (7.9 ± 0.5 µmol mg^-1) and SOD suppression (63.4 ± 4.2 U mg^-1). These findings indicate elevated lipid peroxidation and compromised antioxidant defenses, likely due to Ag-mediated reactive oxygen species (ROS) generation. Sample D showed moderate stress, with MDA at 2.04 ± 0.17 nmol mg^-1, GSH at 11.1 ± 0.6 µmol mg^-1, and SOD at 85.7 ± 3.7 U mg^-1. Sample E closely mirrored the control values, confirming the inert oxidative profile of gold-dominant nanoparticles. These results suggest that the antimicrobial and metabolic effects of Ag/Au NPs may be partially mediated by redox imbalance, especially in Ag-rich formulations that promote electron transfer and ROS propagation.

Close

Figure 12.

Intracellular oxidative stress biomarkers in Lactobacillus gasseri after 24-h exposure to Ag/Au NPs (50 µg mL^-1). MDA (malondialdehyde) levels indicate lipid peroxidation; GSH (reduced glutathione) and SOD (superoxide dismutase) reflect antioxidant capacity. Sample C induced significant oxidative damage.

Export to PPT

3.7. Hypothetical mechanism of interaction between Ag/Au nanoparticles and lactobacillus

To integrate the multi-layered findings of this study, from physicochemical characterization to microbial viability, metabolic function, community shifts, and biochemical stress markers, we propose a mechanistic model describing how bimetallic Ag/Au nanoparticles interact with Lactobacillus at different levels (Figure 13).

Close

Figure 13.

Schematic diagram of the proposed mechanism by which Ag/Au nanoparticles interact with Lactobacillus gasseri.

Export to PPT

Upon exposure, Ag-rich nanoparticles (Sample C and, to a lesser extent, Sample D) exhibit higher surface reactivity due to silver’s well-known affinity for thiol groups and its capacity to undergo oxidative dissolution, releasing Ag⁺ ions into the surrounding medium [109,110]. These ions interact with bacterial membrane proteins and lipids, compromising membrane integrity, as confirmed by confocal imaging and reduced CFU counts [111]. Additionally, direct physical contact between the nanoparticle surface and the bacterial envelope may induce structural damage via localized charge transfer or membrane depolarization [112]. This primary membrane disruption not only inhibits cell division but also facilitates the leakage of intracellular contents and the influx of extracellular ROS and LPS, both of which exacerbate physiological stress [113,114].

Direct intracellular ROS quantification was not performed in this study; thus, our inference of redox imbalance is based on indirect yet established endpoints, elevated malondialdehyde (MDA), reduced glutathione (GSH), and suppressed superoxide dismutase (SOD), reported in Section 3.6. These markers indicate lipid peroxidation and impaired antioxidant capacity but do not by themselves identify the specific ROS involved or their intracellular localization. In future work, we will complement these endpoints with bacterial ROS assays such as CM-H2DCFDA/DCFH-DA or CellROX Green for whole-cell ROS, dihydroethidium for superoxide, and HPF for •OH, analyzed by flow cytometry or plate fluorimetry with appropriate spectral controls to mitigate nanoparticle quenching/scattering artifacts; we will also include ROS-scavenger controls (e.g., N-acetylcysteine, thiourea, catalase) and consider EPR spin trapping for speciation. This clarification strengthens the mechanistic interpretation while acknowledging that our current conclusions about oxidative stress derive from indirect measures rather than direct intracellular ROS detection.

The accumulation of intracellular ROS, either generated on the nanoparticle surface or via Fenton-like reactions with Ag⁺, triggers lipid peroxidation (MDA elevation) [115,116], depletes key redox buffers (GSH) [117,118], and inhibits enzymatic defenses (SOD) [119]. This oxidative imbalance impacts central metabolic pathways, reducing the production of essential bacterial metabolites, such as SCFAs and aromatic amino acids, as observed in metabolomics profiling. These metabolic shifts likely result from both enzyme inactivation [120,121] and transcriptional downregulation [122] in response to ROS accumulation. As SCFA production declines, so too does the bacterial community’s ability to regulate pH [123,124], inhibit pathogens [125,126], and support mucosal immunity [105,127].

At the community level, Ag-rich NPs selectively suppress SCFA-producing and anti-inflammatory genera (Lactobacillus, Turicibacter, Enterococcus), while opportunistic taxa such as Streptococcus and Escherichia gain a competitive edge [128]. This compositional shift is supported by LEfSe and 16S rDNA sequencing results [129] and results in an altered functional output of the microbiota [130,131]. The release of LPS from Gram-negative bacteria, particularly under stress, serves as a surrogate for immunogenic potential [132] and likely mimics an inflammatory signal cascade in vivo [133]. When coupled with elevated oxidative stress and loss of beneficial metabolic outputs, this sets the stage for a potentially pro-inflammatory gut microenvironment [134].

In contrast, Sample E (Au-rich) demonstrated minimal effects across all biological assays. This likely stems from the inert nature of Au⁰ surfaces [135,136], limited ion release [137], and weaker interactions with bacterial membrane structures [138]. Additionally, the phytochemical capping derived from Camellia sinensis extract may enhance biocompatibility by providing antioxidant moieties on the nanoparticle surface, further reducing ROS generation and shielding bacterial cells. This highlights the potential of using gold-dominant formulations for applications requiring microbial safety, particularly in food, probiotic packaging, or gastrointestinal diagnostic interfaces.