Grantee Research Project Results

The magnitude and mechanisms of quantum dot (QD) toxicity differ for CdSe vs. CdTe QDs and across bacterial strains. In a study completed during this project (Priester, et al., 2009), planktonic Pseudomonas aeruginosa PG201 bacteria were cultured with similar total cadmium concentrations as either fully dissolved cadmium acetate (Cd(CH₃COO)₂) or ligand capped CdSe QDs, and cellular morphology, growth parameters, intracellular reactive oxygen species (ROS), along with the metal and metalloid fates were measured. QDs dissolved partially in growth media, but dissolution was less in biotic cultures compared to sterile controls. Dose-dependent growth effects were similar for low concentrations of either cadmium salts or QDs, but effects differed above a concentration threshold of 50 mg/L (total cadmium basis) where: 1) the growth of QD-treated cells was more impaired, 2) the membranes of QD-grown cells were damaged, and 3) QD-grown cells contained QD-sized CdSe cytoplasmic inclusions in addition to Se₀ and dissolved cadmium. For most concentrations, intracellular ROS were higher for QD- versus cadmium salts-grown bacteria. Taken together, QDs were more toxic to this opportunistic pathogen than cadmium ions, and were affected by cells through QD extracellular stabilization, intracellular enrichment, and cell-associated decay. This report clearly shows a “nanoparticle effect” to the bacteria studied under these conditions, and attributes the mode of action mainly to ROS leading to membrane damage, uptake of QDs, and further intracellular destruction from additional ROS formation in the cytoplasm. In a comprehensive study of CdTe ( Dumas, et al., 2009) QD toxicity to gram-positive and gram-negative bacteria, including oxidative stress-sensitive (OxyR) mutants, under both light and dark conditions, P. aeruginosa was more resistant to toxic effects than Escherchia coli, owing to the protective effects of thick extracellular polymers typical of P. aeruginosa. Growth curves were relatively complex for the gram positive strains, potentially due to sporulation, but Bacillus subtilis appeared to be the most resistant of all four strains. Oxidative stress-sensitive mutants of E. coli were far more sensitive than the wild-type, with stress attributable to the following factors: a) small amounts of Cd ions released early during growth, 2) irradiation which enhanced membrane damage, 3) ROS production, enhanced with irradiation and, 4) which was determined to be, specifically, superoxide. This report clearly shows the importance of strain specificity, light versus dark conditions, and concludes with reiterating the value of quantifying superoxide directly via a reduction-based assay that circumvents confounding issues with other ROS-indicating stains requiring oxidation, and thus which can be oxidized directly by nanoparticles.

 

Charge transfer characteristics, and thus effects on cells, of quantum dots varies with QD core chemistry and conjugation. As above, both CdSe and CdTe quantum dots are toxic to bacteria through particle-related initiation of ROS formation, with membrane damage and uptake. The variations in toxicity of CdSe versus CdTe quantum dots to bacteria may be related to QD charge transfer characteristics. Generally, these semiconductor nanoparticles can generate ROS when illuminated; organic conjugates have the potential to change ROS production and also QD fluorescence. In a detailed study (Cooper, et al., 2009) of the lifetime fluorescence of CdSe and CdTe QDs and the effects of dopamine (a neurotransmitter used as a QD-conjugated electron donor in this case), illumination, and conditions of either oxygenation or deoxygenation, on the generation of different forms of ROS from QDs were characterized. Timeresolved spectroscopy was used to investigate the photoluminescence intensities and lifetimes of CdSe/ZnS and CdTe QDs as a function of blue light illumination. Conjugates of the particles to the electron donor dopamine also were investigated, and the effect of the antioxidant beta mercaptoethanol was explored. Both types of QDs showed signs of direct electron transfer to the conjugate, but enhancement was much more pronounced in CdSe/ZnS QDs. A model of the two different types of enhancement was proposed. This study was followed by another (Cooper, et al., 2010), where oxidative and reductive ROS production were probed abiotically. Quantum dots included: as-synthesized in toluene, water-solubilized, unconjugated QDs, QDs conjugated to dopamine, and dopamine alone. Results of indirect fluorescent ROS assays, both in solution and inside mammalian cells, were compared with those of spin-trap electron paramagnetic resonance spectroscopy (EPR). QDs in toluene did not produce substantial ROS, while solublilized mercaptopropionic acid (MPA)-capped QDs did produce ROS but not singlet oxygen; QDs conjugated to dopamine produced singlet oxygen. Several different fluorescence assays were compared for their performance in reporting ROS and singlet oxygen specifically.The effect of the particles on the metabolism of mammalian cells was shown to be dependent upon light exposure and proportional to the amount of ROS generated, most notably singlet oxygen. A novel study of direct interfacial electron transfer from bacteria to CdTe QDs (Dumas, et al., 2010) was undertaken. It had been recently found that C60 nanoparticles could cause direct oxidative damage to bacterial proteins and membranes, including causing a loss of cell membrane potential (depolarization). However, this did not correlate with toxicity. In this study, a similar analysis was performed, using fluorescent CdTe quantum dots, adapting tools used in prior studies 3, 4 to make use of particle fluorescence. The findings included that two Gram positive strains showed direct electron transfer to CdTe, resulting in changes in CdTe fluorescence lifetimes. These two strains also showed changes in membrane potential upon nanoparticle binding. Two Gram negative strains did not show these effects; nevertheless, they are over 10-fold more sensitive to CdTe than the Gram positives. Subtoxic levels of Cd2+ were released from the particles upon irradiation of the particles, but there was significant production of hydroxyl radicals, suggesting that the latter is a major source of toxicity. These results help establish mechanisms of toxicity and also provide caveats for use of certain reporter dyes with fluorescent nanoparticles which will be of use to anyone performing the assays. The findings also suggested future avenues of inquiry into electron transfer processes between nanomaterials and bacteria, which may be a mechanism of growth inhibition and toxicity overall that is common across many nanoparticles and bacteria strains.

 

Nanoparticles enter bacteria and biofilms, affording the use of exotic metal nanoparticles for improved imaging in research. While it has been known for some time that CdSe QDs can enter bacteria and impart their fluorescence intracellularly, and thus provide for improved and specific fluorescence imaging (reference JL Nadeau in Kloepfer, et al., Applied and Environmental Microbiology 2005 and 2007), a major goal in both assessing toxicity relationships to uptake and for improving the use of nanoparticles in imaging is to use combined fluorescence and high resolution electron microscopy approaches, and to also use nanoparticles with oligonucleotides or other probes to resolve organisms in complex environmental matrices. Two studies in this research were towards addressing these interests. In one study (Erhardt, et al., 2009), a nanoparticulate gold particle-based approach was used to probe bacteria using an oligonucleotide tethered to the particle and to enhance visualization in environmental scanning electron microscopy (ESEM). A new nanogold in situ hybridization method that increases the concentration of nanogold probes bound to rRNA targets within the cell was developed and applied to a range of specimens (Bacteria, Archaea, planktonic and surfaceassociated cells). The approach made individual hybridization events directly visible with secondary electron SEM imaging. While not developed or applied with QDs, the method with nanogold may be adaptable for use with QDs. Few simple labeling methods exist for simultaneous fluorescence and electron microscopy of bacteria and biofilms. In another study (Clarke, et al., 2010), we synthesized, characterized, and applied fluorescent nanoparticle quantum dot conjugates to target microbial species, including difficult-to-label Gram negative strains. The QD conjugates imparted contrast for both ESEM and fluorescence microscopy, permitting observation of living and fixed bacteria and biofilms. The probes were applied for studying biofilms extracted from perennial cold springs in the Canadian High Arctic, which is a particularly challenging system. In the native biofilms, sulfur-metabolizing bacteria live in close association with unusual sulfur mineral formations. Following simple labeling protocols with the QD conjugates, it was possible to image the organisms in fully-hydrated samples and to visualize their relationship to the sulfur minerals using both ESEM and fluorescence microscopy. Scanning transmission electron microscopy was used to observe precipitated sulfur around individual cells and within the biofilm lattice. The possible mechanisms of biofilm and mineral structure formation were revealed through the use of these novel approaches. The new QD conjugates and techniques are highly transferable to many other microbiological applications, especially those involving Gram-negative bacteria, and could be useful for correlated fluorescence and electron microscopy.

 

CdSe Quantum Dots are trophically transferred, and biomagnified, from Pseudomonas aeruginosa bacterial prey into its predator, Tetrahymena thermophila. (manuscript submitted for publication) The ecological impacts of engineered nanomaterials are mostly unknown. The potential for trophic transfer of bare cadmium selenide quantum dots (QDs) was investigated by growing the ciliate Tetrahymena thermophila exclusively on Pseudomonas aeruginosa bacteria that were either untreated or had bioaccumulated cadmium as either QDs or dissolved Cd. We discovered that Tetrahymena entered rapid growth with all treatments. QDs, but not dissolved Cd, inhibited protozoan digestion of bacteria in food vacuoles. After one doubling, protozoa in QD treatments grew slowly, while those in dissolved Cd treatments ceased growing. Cadmium concentrations in Tetrahymena were biomagnified by 33- and 23-fold in QD and dissolved Cd treatments, respectively, relative to their prey. A significant fraction of QDs remained intact within the protozoan cells. Because the poisoned protozoan cells did not lyse, QDs could potentially remain bioavailable to higher trophic levels.

 

TiO₂ nanoparticles disagglomerate in the presence of dense cultures of Pseudomonas aeruginosa (manuscript submitted for publication). Upon environmental release, nanoparticles could inhibit bacterial processes, as evidenced by laboratory studies. Less is known regarding bacterial alteration of nanoparticles, including whether bacteria affect physical agglomeration states controlling nanoparticle settling and bioavailability. Here, the effects of an environmental strain of Pseudomonas aeruginosa on TiO₂ nanoparticle agglomerates formed in aqueous media were studied. Environmental scanning electron microscopy and cryogenic scanning electron microscopy visually demonstrated bacterial dispersion of large agglomerates formed in cell culture medium and in marsh water. For experiments in cell culture medium, quantitative image analysis verified that the degree of conversion of large agglomerates into small nanoparticle-cell combinations was similar for either 12 hour growth or shortterm cell-contact experiments. Dispersion in cell growth medium was further characterized by size fractionation: for agglomerated TiO₂ suspensions in the absence of cells, 81% by mass was retained on a 5 μm pore-sized filter, compared to only 24% retained for biotic treatments. Filtrate cell and agglomerate sizes were characterized by dynamic light scattering, revealing that the average bacterial cell size increased from 1.4 μm to 1.9 μm because of nano-TiO₂ biosorption. High magnification scanning electron micrographs showed that P. aeruginosa dispersed TiO₂ agglomerates by preferential biosorption of nanoparticles onto cell surfaces. These results suggest a novel role for bacteria in the environmental transport of engineered nanoparticles, i.e. growth-independent, bacterially-mediated size and mass alterations of TiO₂ nanoparticle agglomerates.

 

TiO₂ Nanoparticle inhibition of Pseudomonas aeruginosa growth correlates with nanoparticle surface area as related to nanoparticle size (unpublished findings; work in progress). Nanoscale TiO₂ toxicity to P. aeruginosa was inferred from bacterial growth characteristics (growth rate, and yield) when bacteria were grown planktonically with distinctly sized, laboratory synthesized (flame spray pyrolysis, all ca. 80% anatase + 20% rutile) TiO₂ nanoparticles (7.4 nm, 16.0 nm, and 30.0 nm) at two concentrations (0.1 and 0.5 mg/mL) in the dark. During growth studies, rapid association of initially-agglomerated nanoparticles at cell surfaces occurred within 10 min, thus cells were exposed to particles at the outer bacterial membrane. Greater dose-dependent toxicity (here, defined as a decrease in growth rate or cell yield) was observed for the smallest (7.4 nm) nanoparticles. Mid-sized (16.0 nm) nanoparticles exerted intermediate dose-dependent toxicity, and the largest (30.0 nm) nanoparticles had no effect on growth characteristics, compared to TiO₂-free controls. Alone, this suggests greater toxicity of the smallest nanoparticls, particularly at the high (0.5 mg/mL). However, as particle size is inversely related to the surface per volume ratio, we considered that what appears to be a size-dependent toxicity phenomenon could, in fact, be a result of increased available TiO₂ surface area as particle size decreases. Assuming nanoparticle sphericity (as observed in TEM images), surface area dosage (cm2/mL) was calculated for each particle size at each concentration. When cell yield (maximum iDNA, ng/mL) was plotted against surface area dosage, a highly correlated (R2 = 0.89) but insignificant (P = 0.24) inverse trend was observed. However, a highly correlated (R2 = 0.84) and significant (P = 0.038) inverse linear relationship was observed between bacterial growth rate and surface area dosage. This suggests that surface area may be a reasonable dose metric. However, this research is ongoing with further methodological development including predispersing nanoparticles prior to the initiation of the growth experiment, quantifying nanoparticle agglomeration state and association with bacteria, and relating effects to exposures actualized at cells.

 

Growth inhibition of environmentally-relevant bacteria with a range of industrial metal oxide nanoparticles occurs at similar threshold dosages across particle and bacterial types (manuscript in preparation). Across several gram positive and gram negative bacterial strains, and across seven industrial metal oxide nanomaterials (with additional support), all of varying core chemistries, a synthetic comparison reveals that the mass concentration dosage administered in laboratory culture that results in exponential growth rate reduction in comparison to no-nanoparticle controls is very similar. This suggests that there are possibly similar effects mechanisms of nanoscale industrial metal oxide particles to bacteria. Further data analysis and synthetic research is underway to understand this observation.