TEM micrographs of citrate Au and Ag NPs were obtained with NP solutions after centrifugation and re-suspension in DI water, and TEM micrographs of poly-allylamine hydrochloride (PAH) Au NPs were acquired with NPs in solution as they were made. In the case of Au NPs prepared in citrate, spherical NPs with 20−30 nm diameter were observed (Figure 1A), gold nanoparticles were well dispersed after rounds of centrifugation and re-suspension. PAH stabilized Au NPs were spherical and approximately 22 nm in diameter (Figure 1B), and the presence of polyelectrolyte forced the formation of monodispersed gold nanoparticle ‘aggregates’. Ag NPs were produced in Triton X-100 (TX-100) solution and reached an average diameter of 30 nm (Figure 1C). All NPs showed consistent spherical morphology and narrow size distribution between 20 and 30 nm. This consistency helped minimize the shape and size differences that might result in different antibacterial effects. Figure 1D shows thin section of E. coli grown in LB liquid medium. Bacterial cells exhibited well-defined cellular contents and were normal in size with intact intracellular structures.

Nanoparticles in solution were added to fresh LB medium for four hours to examine the level of aggregation, as shown in Figure 2, and the final concentration of NPs were chosen to be 10 μg/ml. The level of aggregation of citrate Au NPs was higher in LB medium (Figure 2A) as compared to in DI water (Figure 1A). This might have been expected, since the medium is full of nutrients as well as free ions both of which could exchange with citrate on the nanoparticles surface and cause aggregation. The exchange process is less likely to occur when high molecular weight molecules, such as polyelectrolyte (PAH), were used for surface modifications. The result is the level of aggregation stayed the same after adding LB (Figure 2B). Ag NPs display aggregation after suspending them in LB medium (Figure 2C).

NPs were added to LB medium to reach predesigned concentrations. E. coli was exposed to 0.1, 1, 5, 10 μg/ml citrate or PAH Au NPs and 1, 10 μg/ml Ag NPs. At selected times, 10 μl of medium was sampled, diluted and then cultured on LB ager plates. OD₆₀₀ was also monitored over time. Considering the OD₆₀₀ value of NPs in LB medium as background, it was expected that we could estimate bacterial numbers by subtracting background from each measurement. It was found that there was no OD₆₀₀ additive relationship between NPs and bacteria (data not shown). After NPs were added to the bacterial culture, bacterial cells attracted NPs. Thus, the OD₆₀₀ reading that would normally be generated by this portion of NPs was absent. As a result, CFU is likely to be the only accurate method to determine bacterial numbers with little influence from the NPs. CFU data were converted to bacterial numbers per ml as shown in Figure 3. The presence of NPs reduced the bacterial numbers during the exponential phase. High concentrations of citrate Au NPs have less of an inhibitory effect on bacterial growth as compared to low concentrations (Figure 3A). At 10 μg/ml, citrate Au NPs appeared to form aggregates which contain multiple mono-dispersed nanoparticles, and there is no opportunity for NPs inside ‘aggregates’ to interact with cells (Figure 2A). This also left fewer non-aggregated NPs which can be easily absorbed by bacteria. Hence, bacterial growth was only slightly inhibited at high citrate Au NPs concentrations. In contrast, E. coli growth was significantly inhibited by citrated Au NPs at 0.1, 1, 5 μg/ml. This observation could be explained by the fact that smaller aggregates exist at lower NPs concentrations. In such, more NPs can interact with bacterial cells and cause further damage. Figure 3B presents a dose-dependent relationship between PAH Au NPs and bacteria inhibition. It suggests that bacterial growth was less inhibited when 0.1 and 1 μg/ml PAH Au NPs were present, so there are similar growth curves compared to the negative controls. E. coli was largely inhibited by 5 and 10 μg/ml of PAH Au NPs. The effects of PAH toxicity on bacterial killing was evaluated by determining the minimal inhibitory concentration (MIC) for PAH. The initial concentration of PAH for making PAH Au NPs was 1000 mg/L, and most of PAH was removed after NPs formation by centrifugation and re-suspension in water. As shown in Figure 3D, the only concentration of PAH that display bacterial inhibitory effects is 2000 mg/L and lower concentrations showed no inhibitory effect. Hence, the antibacterial effect of PAH Au NPs is not due to concentrations of PAH that could remain after processing. At 260 min, hygromycin treated E. coli started to replicate while 5 and 10 μg/ml PAH Au NP treated groups still displayed no growth. This suggests that there are stronger antibacterial effects at these concentrations as compared to 20 μg/ml of hygromicin. Strong antibacterial activities were observed at all Ag NP concentrations (Figure 3C).

Nanopaticle treated E. coli was thin sectioned for TEM imaging in order to study the mechanisms of the antibacterial interactions. The following concentrations were selected because they display strong antibacterial effects: 2 μg/ml citrate Au NPs, 5 μg/ml PAH Au NPs and 20 μg/ml Ag NPs. Samples were withdrawn and fixed at 3, 6, and 9 hours. These points were picked since the interaction is more obvious. Seen from TEM micrographs, NPs are rarely found within E. coli cells after 3 hours of incubation with citrate Au NPs (Figure 4A, arrow indicates Au NPs aggregates). Most of the NPs are either free in the medium or aggregated between bacteria. Bacterial cell walls started to retain some NPs at 6 hours (arrows in Figure 4B). Figure 4C demonstrates an increased number of NPs inside bacterial cells after 9 hours (arrows in Figure 4C). The NP complexes inside bacterial cells in Figure 4D and 4E were measured to be 100 nm in diameter, which is equivalent to about 50 individual NPs and their morphology is quite different from more dispersed ‘aggregates’ in Figure 2A. It is easy to distinguish a single NP within the 100 nm complex at 6 hours (inset in Figure 4D). However, given more time, the outlines of individual NP became blurry within these complexes, which suggests that single NPs physically merged with each other and ‘aggregates’ turned into larger particles (9 h; inset in Figure 4E). PAH Au NPs showed different antibacterial activity when they were applied in E. coli culture, cell lysis was observed at 3 h, 6 h, and 9 h (Figure 4F−4I). Au NPs were trapped within what appeared to be released cytoplasm and no free NPs were found in the medium (arrows in Figure 4F and 4G). Release of the cytoplasm left hollow and disfigured cell walls (arrows in Figure 4H and 4I). TEM micrographs of Ag NPs treated E. coli at different time points are shown in Figure 4J−4N. Arrows in Figure 4J indicate Ag aggregates formed in medium, there are no aggregated Ag NPs complexes inside of cells. As seen in Figure 4M and 4N, NPs maintain their original size and spherical shape. A large portion of the Ag NPs was trapped within the cell walls (Figure 4L−4N).

FE-SEM was used to examine the interactions of NPs with bacteria in more detail and the results are shown in Figure 5. E. coli was fixed after 30 min of incubation with citrate Au NPs. In secondary electron mode, Au NPs were found aggregated (Figure 5A), which is in accordance with the TEM result in Figure 3A. Besides aggregated Au NPs, some Au NPs were found dispersed among cells, as shown through back-scattered imaging (Figure 5B).

The relationship between viable cell numbers and fluorescence was shown in Figure 6A. The level of fluorescence was correlated with bacterial numbers, as confirmed by CFU (R² = 0.994). All NPs were found to have no emission at 554 nm excitation. Hence, there is no interference of NPs with fluorescence emission. After 5 days of BCG/NPs co-culture, fluorescence was measured and the results are shown in Figure 6B−6D. BCG fluorescence was inhibited by citrate Au NPs at all concentrations tested (Figure 6B). The inhibitory effect of citrate Au NPs was more obvious at 0.1 and 1 μg/ml which is in accordance with E. coli inhibition. Low fluorescence levels were found when PAH Au NPs were at high concentrations (Figure 6C). This dose-dependent relationship for PAH Au NPs was observed for both E. coli and BCG. When BCG was treated with 1, 5, 10 μg/ml of Ag NPs, it displayed similar fluorescence to that of the hygromycin treated group ( 6D).

Citrate stabilized gold (Au) NPs were prepared following previously reported methods 17. Citrate ions produce negative charges on the surface of nanoparticles. By controlling citrate and hydrogen tetrachloroaurate tri-hydrate (HAuCl₄) concentrations, 20−30 nm diameter Au NPs were synthesized. Poly-allylamine hydrochloride (PAH) stabilized Au NPs with 22 nm mean size were synthesized in an aqueous solution of HAuCl₄ and PAH and in the presence of gold seed particles 15. A positive charge was demonstrated on Au NPs’ surface because of the presence of PAH. Silver (Ag) NPs were produced by photoirradiating AgNO₃ in Triton X-100 (TX-100) solution for 60 min, the average size was around 30 nm 18. Citrate Au NPs and Ag NPs solutions were centrifuged at 8000 rpm for 15 min to remove extra citrate and TX-100. The precipitated NPs were suspended in DI water. This process was repeated 3−4 times. PAH Au NPs were used immediately after they were made, since precipitated NPs were unable to dissolve in water.

pBlueScriptKSII + plasmid (Agilent Technologies, Santa Clara, CA) was transformed into Escherichia coliDH5α (Invitrogen, Carlsbad, CA). E. coli was grown in liquid Luria-Bertani (LB) medium at 37°C and 250 rpm. Bacillus Calmette-Guérin (Pasteur) cultures expressing tdTomato fluorescent protein were grown in M-OADC-TW broth with Middlebrook (M) 7 H9 broth (Difco) supplemented with 10% (v/v) oleic acid albumin dextrose complex, 0.5% (v/v) glycerol, and 0.05% (v/v) Tween 80. Medium was supplemented with 25 μg/ml kanamycin. NPs in solution were added into bacteria cultures to reach specific concentrations.

NPs samples for TEM imaging were prepared by slow evaporation of freshly made NPs in solution on a carbon-coated copper grid at room temperature. Bacterial samples were fixed in 2% (v/v) glutaraldehyde for 0.5 h at 37°C by adding fixative into culture medium. The following process was performed with cold microwave technology in the BioWave. The microwave was set for a 6 minute cycle (2 min power on, 2 min power off, 2 min power on) at 200 W for primary fixation. Alternating vacuum cycles of 30 seconds were used during 6 minutes of fixation. The microwave temperature was set at 20°C for all steps. Bacteria were spun down and washed 3 times with 0.1 M HEPES for 1 min at 200 W with consistent vacuum. Subsequently, 1% (v/v) osmium tetroxide in HEPES buffer was used for overnight post fixation. Samples were microwaved for 6 min at 100 W at the same power and vacuum cycles as used in primary fixation. Methyl alcohol was used for dehydration in 10% (v/v) steps from 10% to 100%. Microwave was set at 100 W with 1 min per step. Quetol 651 epoxy resin was used for specimen embedding. Thin sections were picked up on 200 mesh copper grids and imaged with a JEOL 1200EX TEM with 100 keV acceleration voltage 19. FE-SEM sample preparation followed the same fixation and dehydration procedure described above. Images were taken on an FEI Quanta 600 FE-SEM at an acceleration voltage of 10 keV. ImageJ software was used for measuring nanoparticle size distribution.

E. coli cells were grown in LB liquid medium at 37°C for 12 hours before they were diluted in fresh LB liquid medium to reach OD₆₀₀ = 0.003 (optical density). Gradient concentrations of NPs were then added to the culture medium. Bacteria/NP mixed cultures were put into a 37°C incubator. At different time points of 80, 120, 160, 200, and 260 min the medium was withdrawn from each sample. Dilutions were then made and cultured on LB agar plates. Plates were incubated overnight at 37°C and CFU was determined. NP-free E. coli was used as a negative control and 20 μg/ml hygromycin was used as a positive control for bactericidal activity. The toxicity of PAH was tested by determining its minimum inhibitory concentration (MIC). Overnight cultured E. coli suspensions were adjusted in fresh LB liquid medium to reach an OD₆₀₀ = 0.1 and diluted by a factor of 1:100. PAH was dissolved in LB media. After mixing 50 μL E. coli dilution and 50 μL PAH solution, 10 μL of the sample was immediately drawn out and a series of dilutions made. Samples were plated on LB agar plates, and colonies were counted the next day 20.

BCG::tdTomato with OD₆₀₀ = 0.05 was cultured in 96 well plates with different NPs at various concentrations. Five days after mixing, fluorescence was measured using a Perkin Elmer Envision spectrophotometer with an excitation wavelength of 554 nm and emission wavelength of 581 nm. Wells with NP-free BCG::tdTomato were used as negative controls, and the wells with 20 μg/ml hygromycin in NP-free 7 H9 medium were used as positive controls for comparison. The t test (student version) was used to determine the significance of results.