Silver nitrate (AgNO₃), ethylenediaminetetraacetic acid (EDTA), potassium bromide (KBr), and ethanol were purchased from Sigma-Aldrich (USA). Peptone, yeast extract, and sodium chloride were collected from Unichem (China). Trypan blue dye was purchased from Alfa Aesar (United Kingdom). Agar powder was purchased from Titan Biotech Ltd. (India). Trichloroacetic acid (TCA) was purchased from Merck (Germany) and thiobarbituric acid (TBA) was bought from JT Baker (USA). Biochemical assay kits for serum AST, ALT and γ-GT were purchased from Vitro Scient (Egypt). On the other hand, serum creatinine assay kit was bought from Crescent Diagnostics (The Kingdom of Saudi Arabia). Both pathogenic (i.e., S. typhi, S. aureus, V. cholerae, enteropathogenic Escherichia coli (EPEC), and E. faecalis), and nonpathogenic (i.e., E. coli DH5α, and B. subtilis RBW) bacterial strains were obtained from the Department of Biotechnology and Genetic Engineering, Jahangirnagar University, Savar, Dhaka 1342, Bangladesh. Twenty-five Wistar male rats (weighing ∼60 g) were obtained from the Department of Biochemistry and Molecular Biology, Jahangirnagar University, Dhaka, Bangladesh.

Both aqueous and ethanolic extracts of D. malabarica fruit were used to prepare biogenic Ag nanoparticles (bAgNPs). The fruit extracts were prepared according to our previously published protocol (Shubhra et al., 2019; Shubhra et al., 2021). A freshly prepared AgNO₃ (10 mM) solution was added separately to aqueous (Aq) and ethanolic (Et) fruit extracts at a ratio of 9:1 (ml). The mixtures of AgNO₃ solution and fruit extracts were incubated in the dark at 30°C for 24 h upon constant stirring. Here, the plant extracts act as reducing agents, and the transformation of colorless AgNO₃ solution into brown color confirms the reduction of Ag⁺ to Ag⁰ (Ahmed et al., 2016; Niloy et al., 2020). The unreacted plant extracts as well as AgNO₃ were removed through centrifugation at 16,000xg for 60 min. The supernatant was discarded, and the bAgNPs generated by the aqueous as well as ethanolic extracts were washed twice with distilled water before redispersing in distilled water.

UV-visible spectrophotometer (Specord^® 205, Analytik Jena, Germany) was used to characterize the as-synthesized biogenic AgNPs. FTIR spectroscopy (IRPrestige-21, SHIMADZU, Japan) and powder X-ray diffractometer (GNR X-Ray Explorer, Italy) equipped with Cu K_α (0.154 nm) at 25 mA were used to characterize the chemical and crystal property of bAgNPs, respectively. The XRD spectrum was obtained by scanning the diffraction angle (2θ) region at 30 kV. The diffraction angle varied in the range of 10–50° and the scanning rate was 5°/s. The morphology and the composition of biogenic nanoparticles were characterized using TEM having a field emission gun (HF-2200; Hitachi, Tokyo, Japan) coupled with energy-dispersive X-ray spectroscopy (EDAX Genesis; AMETEK, Pennsylvania, United States) operating at 200 kV. The hydrodynamic size and zeta potential of bAgNPs were measured using a zeta size analyzer (Nano-ZS90; Spectris PLC, Egham, U.K.).

The antibacterial activity of as-synthesized biogenic AgNPs was investigated against five pathogenic (i.e., S. typhi, S. aureus, V. cholerae, EPEC, and E. faecalis) and two non-pathogenic (i.e., E. coli DH5α, and B. subtilis RBW) bacterial strains. The chosen strains include both Gram-positive and negative bacterial strains. The antibacterial activity of both aqueous and ethanolic extract mediated biogenic AgNPs was determined through MIC and disc diffusion assay, and trypan dye exclusion assay (Polash et al., 2017b; Hossain et al., 2019).

The peroxidation potential of bacterial lipid membrane by Aq@AgNPs and Et@AgNPs was investigated according to our previously established protocol (Ranjan Sarker et al., 2019). Briefly, TCA solution (10%) was added to bAgNPs-treated bacterial suspension and the insoluble cellular components were separated through centrifugation to collect the supernatant. The supernatant was centrifuged again to precipitate any remaining proteins and cellular debris. Malondialdehyde (MDA), a polyunsaturated fatty acid, presents in the supernatant reacts with the freshly prepared TBA (0.67%) solution and produce MDA-TBA adduct when incubated in a hot water bath for 10 min. The adduct was then cooled down to room temperature before measuring the absorbance at 532 nm by a UV-visible spectrophotometer.

The compatibility of human and rat erythrocytes with bAgNPs (i.e., Aq@AgNPs and Et@AgNPs) was investigated according to our previously established protocol (Ranjan Sarker et al., 2019; Sarker et al., 2022). Briefly, human and rat red blood cells (RBCs) were isolated from serum through centrifugation at 500 g for 10 min. The serum was then removed before resuspending RBCs in 5 ml phosphate-buffered saline (PBS) and centrifuged again. The cell suspension was washed twice with 150 mM NaCl solution at 3,000 g for 3 min. Finally, 0.1 ml of RBCs solution was mixed with 0.4 ml of various amounts of bAgNPs (i.e., 20, 40, 60, and 80 μg) and incubated for 30 min at 37°C and 150 rpm before centrifuging at 1,377 g for 5 min. The supernatant was taken out to measure its optical density at 570 nm. Herein, erythrocytes incubated with PBS and water were considered as negative and positive controls, respectively. The hemolysis rate was calculated using the following formula: %   o f   h e m o l y s i s =   a b s o r b a n c e   o f   s a m p l e − a b s o r b a n c e   o f   n e g a t i v e   c o n t r o l a b s o r b a n c e o f p o s i t i v e   c o n t r o l − a b s r o b a n c e o f n e g a t i v e   c o n t r o l   × 100

Wistar male rats were used to investigate the toxicity of bAgNPs in vivo. Twenty-five rats were divided into five groups (5 per group) and were maintained in a hygienic environment followed in our previous study (Hossain et al., 2019). Two doses of Aq@AgNPs and Et@AgNPs (i.e., 3 and 6 mg/kg of rat body weight) were delivered intravenously via the tail vein. After 7 days, about 3 ml of blood was collected and the serum was separated through centrifugation at 538 g for 10 min at room temperature.

To determine whether biogenic Aq@AgNPs and Et@AgNPs have any toxic effects on rat tissues, the concentration of several biomarkers was measured. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma-glutamyltransferase (γ-GT) were measured as liver function biomarker. On the other hand, the level of serum creatinine was measured as kidney function biomarker. The concentration of all the biomarkers was determined according to the instructions provided with the reagent kit. All the animal experiments performed in the study were approved by the Biosafety, Biosecurity & Ethical Committee of Jahangirnagar University [BBEC, JU/M 2019(4)1].

The colorless AgNO₃ solution turned brown after the addition of D. malabarica (aqueous and ethanolic) fruit extract confirming the reduction of Ag⁺ to Ag⁰. The change of color is due to the presence of secondary metabolites especially phenols, flavonoids, and andrographolides in D. malabarica fruit extract. These secondary metabolites act as reducing agents. They also bind to the surface of the biogenic nanoparticles as capping agents. The UV-visible spectra confirm the generation of biogenic AgNPs since the λ_max of Aq@AgNPs and Et@AgNPs are 428 and 416 nm, respectively (Figure 1A). The data are in the same line with the previously published literature that showed the surface plasmon resonance (SPR) value of AgNPs is 420 nm (Ramalingam et al., 2016; Hossain et al., 2019). The SPR is very specific for metals as it exhibits strong scattering and absorption properties of AgNPs by causing collective oscillation when electrons on the metal surface are excited by light at a specific wavelength (Li et al., 2010).

The crystalline structure of the as-synthesized Aq@AgNPs and Et@AgNPs was investigated by powder XRD. The analysis of XRD spectra of as-synthesized bAgNPs showed diffraction peaks at 32.2°, 38.3°, and 44.4° (Figure 1B) that can be assigned to the planes (122), (111), and (200) facets of Ag crystal, respectively (Ashraf et al., 2016; Goudarzi et al., 2016). On the other hand, FTIR spectroscopy was carried out to confirm the presence of secondary metabolites in the fruit extracts as well as in the as-synthesized bAgNPs, since the secondary metabolites (e.g., alkaloids, flavonoids, and so on) are responsible for the reduction of Ag⁺ and stabilization of bAgNPs (Elamawi et al., 2018; Hossain et al., 2019). Absorption bands at 3,015, (2917 and 2847), 2360, 1489–1855, (961 and 912), and (718 and 666) cm⁻¹ were attributed to symmetric CH₂ stretching, C-H asymmetric stretching, C≡C stretching, carbonyl group, C-OH stretching, and C-H stretching of the aromatic compound in Aq@AgNPs and Et@AgNPs (Figure 1C). FTIR spectra of all the precursor compounds are available in Supplementary Figure S1.

The average hydrodynamic diameter of Et@AgNPs was larger than that of Aq@AgNPs (Figure 2A). The hydrodynamic diameters of Aq@AgNPs and Et@AgNPs were 287 ± 117 and 375 ± 69 nm, respectively. Moreover, the zeta potentials of Aq@AgNPs and Et@AgNPs were −9.8 ± 2.6 and −12.2 ± 1.9 mV, respectively (Figure 2B). The negative zeta potential of bAgNPs is due to the adsorption of bioactive molecules on the particle surface (Rao and Paria, 2015). On the other hand, the polarity of solvents used to extract phytoconstituents from D. malabarica fruits plays an important role, since the chemical structure of secondary metabolites differs from each other. Hence, the composition of aqueous fruit extract differs from that of ethanolic extracts. The different chemical composition of different solvent extracts influences the surface potential of as-synthesized bAgNPs.

The shape and morphology of biogenic Aq@AgNPs and Et@AgNPs were determined using a scanning transmission electron microscope (STEM) (Figure 3). STEM images confirmed that bAgNPs were surrounded by polymers and difficult to observe directly. In addition, TEM or High-angle annular dark-field (HAADF) images also exhibited that bAgNPs were coated with polymers and most of the NPs were spherical. EDS analysis confirmed that both NPs were composed of Ag, C, and O (Figure 3). A characteristic Ag peak was observed at ∼3 keV. Other peaks such as Cu and Si are due to artifacts of TEM grid and vacuum pump silicone oil analysis, respectively. Hence, it can be concluded that bAgNPs are coated with bioactive polymers present in the fruit extracts of D. malabarica.

The broth dilution method was used to estimate the MIC values of as-synthesized Aq@AgNPs as well as Et@AgNPs. In the case of Aq@AgNPs, the lowest MIC value was 0.25 μg/ml against S. typhi, E. faecalis, E. coli DH5α, and V. cholerae (Table 1). On the other hand, the lowest MIC value (i.e., 0.125 μg/ml) for Et@AgNPs was found against B. subtilis and S. aureus (Table 1). A clear zone on the agar media surrounding the sterile metrical filter paper disk confirmed the antibacterial/bactericidal activity of as-synthesized bAgNPs. The highest antibacterial activity of Aq@AgNPs and Et@AgNPs (amount: 80 μg) was observed against E. faecalis (ZOI: 17.77 mm) and B. subtilis (ZOI: 20 mm), respectively (Table 1). On the other hand, the lowest antibacterial activity of Aq@AgNPs (13.22 mm) and Et@AgNPs (15.89 mm) was observed against enteropathogenic E. coli or EPEC. Aqueous fruit extract, ethanolic fruit extract, and silver nitrate were used as control, and they showed lower antibacterial activity when compared to that of the as-synthesized bAgNPs against all the tested bacteria (Table 1).

The phytoconstituents of D. malabarica fruit extracts contain a variety of biologically active compounds including long chain hydrocarbons, hydroxyls, as well as carboxyl groups containing bioactive molecules that can act as the hydrophilic moieties. Therefore, noncovalent (i.e., hydrophobic) interactions between the bAgNPs and bacteria were responsible for the bacterial death. Increased uptake of the negatively charged Aq@AgNPs and Et@AgNPs is associated with their higher hydrophobicity when compared to that of the plant extracts and silver nitrate (Li et al., 2014; Hossain et al., 2019). The interaction between the bAgNPs and bacteria could also be due to molecular clustering effect (Arakha et al., 2015). Finally, the release of silver ions and/or their intracellular deposition is responsible for the antibacterial mechanism of bAgNPs (Kim et al., 2016; Hoseinnejad et al., 2018). The Ag⁺ ions released from the as-synthesized bAgNPs can interact with sulfur as well as phosphorus containing proteins present in the plasma membrane as well as in the bacterial cell wall via electrostatic interactions and result in the formation of pores in the membrane (Hindi et al., 2009). The resultant membrane pores allow the outflow of bacterial intracellular contents and bring about bacterial death. Bharadwaj et al. (2021) reported the concentration-dependent antibacterial activity of bio-synthesized AgNPs where the diameter of the clear zone of inhibition (ZOI) was about 12–14 mm at 1,000 μg/ml concentration (Bharadwaj et al., 2021). Herein, we investigated the antibacterial activity of both Aq@AgNPs and Et@AgNPs using much lower concentration (amount: 80 µg). Our observed antibacterial activity in the presence of lower amount of nanoparticles indicates the promising antimicrobial potential of D. malabarica fruit extract derived biogenic AgNPs.

Trypan blue dye exclusion assay was performed to confirm the damage of bacterial cell wall due to interaction with bAgNPs. Hydrophobic interactions between the nanoparticles and bacterial cell wall were responsible for the membrane damage. As a result, blue dye deposited inside the bacterial cytosol and appeared blue under the phase contrast microscope (Supplementary Figures S2–S8). However, intact cell wall of live bacteria inhibits the dye from getting inside.

To investigate the oxidative potential of bacterial membrane fatty acids by bAgNPs, we performed lipid peroxidation (LPO) assay. The MDA-TBA adduct was formed after the treatment of bacteria with Aq@AgNPs and Et@AgNPs due to their strong hydrophobic interactions with bacterial cell wall. The highest amount of MDA-TBA adduct was formed when Et@AgNPs was used to treat Gram-positive B. subtilis (Figure 4). On the other hand, the highest amount of MDA-TBA adduct was formed when E. faecalis were treated with Aq@AgNPs (Figure 4). Different amount of MDA-TBA adduct was formed when different bacterial strains were treated with bAgNPs. This is because of the differences in the membrane composition of different bacterial strains that influence their interaction with the nanoparticles. Thus, lipid peroxides are formed due to the oxidation of bacterial cell membrane fatty acids by the as-synthesized bAgNPs.

As transition metals (e.g., Ag) put oxidative stress on fatty acids, the reactive oxygen species (ROS) are generated after the treatment of bacterial cell membrane with bAgNPs (Li et al., 2012). The generated ROS oxidizes fatty acids in the bacterial membrane and produces lipid peroxides. Furthermore, ROS-mediated oxidative stress inhibits the electron transport chain and alters bacterial metabolic responses (Wang et al., 2017). Thus, it induces apoptosis to the corresponding bacterial cells by stimulating the expression of apoptosis genes (Wang et al., 2017).

The compatibility of biogenic Aq@AgNPs and Et@AgNPs to human and rat RBCs was investigated in a dose-dependent manner (i.e., 20–80 μg). Both the nanoparticles showed excellent hemocompatibility to human as well as rat RBCs (Figure 5). The percentage of hemolysis in the presence of 80 μg Aq@AgNPs was 4.1 and 3.9% in the case of human and rat RBCs, respectively. On the other hand, the percentage of hemolysis in the presence of 80 μg Et@AgNPs was 4.2 and 2.3% in the case of human and rat RBCs, respectively. Biogenic Et@AgNPs showed greater compatibility to both human and rats RBCs than that of Aq@AgNPs. This could be due to greater negative surface potential (i.e., −12.2 ± 1.9 mV) of Et@AgNPs when compared that of Aq@AgNPs. The negative surface potential brings about stronger electrostatic repulsion with the negatively charged RBCs. In both cases, the percentage (%) of hemolysis was less than that of the therapeutically acceptable hemolytic value (i.e., 5%) (Nayak et al., 2016; Hossain et al., 2019; Sarker et al., 2022). Furthermore, the variation in hemolysis percentage is due to the differences in membrane composition of human and rat RBCs (da SilveiraCavalcante et al., 2015).

Neither Aq@AgNPs nor Et@AgNPs showed any significant toxic effect on the liver (Figure 6) and kidney tissues (Figure 7) at both doses (3 and 6 mg/kg) of the biogenic nanoparticles. The colloidal AgNPs are usually deposited to the liver and kidneys and interfere with their functions (Kim et al., 2007; Ahmad and Zhou, 2017; Hossain et al., 2019). To determine the toxic effects of bAgNPs, liver function was investigated by measuring the level of serum ALT, AST, and γ-GT enzymes after the intravenous delivery of bAgNPs through tail vein. Any decrease in ATP level indicates damage of the liver tissues that causes the release of hepatic enzymes (i.e., ALT and AST) (Ramaiah, 2007; Hossain et al., 2019). On the other hand, γ-GT is a hepatobiliary enzyme whose level rises within hours of cholestasis (Ramaiah, 2007). Meanwhile, the toxic effect of bAgNPs on kidneys was also investigated by determining the level of serum creatinine. Low level of serum creatinine indicates the kidneys’ ability to filter waste products from the blood and excrete them in the urine (Kaur et al., 2012).

Our data suggest that bAgNPs (up to 6 mg/kg) are therapeutically safe for the liver, since no statistically significant elevation of the level of serum biomarkers was observed. Furthermore, intravenous administration of different doses (up to 6 mg/kg) of bAgNPs for 7 days also did not show any significant elevation in the level of serum creatinine. Hence, both Aq@AgNPs and Et@AgNPs did not show any toxic effect to the liver and kidneys.