Separation of Bacteria, Protozoa and Carbon Nanotubes by Density Gradient Centrifugation

Abstract

Sustainable production and use of carbon nanotube (CNT)-enabled materials require efficient assessment of CNT environmental hazards, including the potential for CNT bioaccumulation and biomagnification in environmental receptors. Microbes, as abundant organisms responsible for nutrient cycling in soil and water, are important ecological receptors for studying the effects of CNTs. Quantification of CNT association with microbial cells requires efficient separation of CNT-associated cells from individually dispersed CNTs and CNT agglomerates. Here, we designed, optimized, and demonstrated procedures for separating bacteria (Pseudomonas aeruginosa) from unbound multiwall carbon nanotubes (MWCNTs) and MWCNT agglomerates using sucrose density gradient centrifugation. We demonstrate separation of protozoa (Tetrahymena thermophila) from MWCNTs, bacterial agglomerates, and protozoan fecal pellets by centrifugation in an iodixanol solution. The presence of MWCNTs in the density gradients after centrifugation was determined by quantification of ¹⁴C-labeled MWCNTs; the recovery of microbes from the density gradient media was confirmed by optical microscopy. Protozoan intracellular contents of MWCNTs and of bacteria were also unaffected by the designed separation process. The optimized methods contribute to improved efficiency and accuracy in quantifying MWCNT association with bacteria and MWCNT accumulation in protozoan cells, thus supporting improved assessment of CNT bioaccumulation.

Keywords: Pseudomonas aeruginosa; Stokes’ law; Tetrahymena thermophila; bioaccumulation; bioconcentration; carbon-14; iodixanol; sucrose.

Figure 1

Representative phase contrast images of P. aeruginosa incubated with 5 mg/L MWCNTs for 1 h: (a) pelleted bacteria after differential centrifugation, white arrow points to MWCNT agglomerate; and (b) bacteria after density gradient centrifugation in sucrose.

Scheme 1

Schematic illustration of observed separation results for bacteria and protozoa after either bacterial exposure to MWCNTs, or protozoan exposure to bacterial prey with their cell-associated MWCNTs: (a) separation of MWCNT-associated bacteria (shown in red) from MWCNTs; and (b) separation of MWCNTs, bacteria and fecal pellets from protozoa (shown in yellow) that were grown with MWCNT-encrusted bacteria; free MWCNTs are included as a component of the system to account for possible “shedding” of MWCNTs from bacteria before or after the uptake into protozoan food vacuoles. Blue lines denote liquid levels in the tubes and interfaces of different liquids. FA—fixed angle rotor, SW—swinging bucket rotor (Table S2).

Figure 2

Natural log-transformed T. thermophila cell counts per milliliter over 22 h growth with control or MWCNT-treated P. aeruginosa as a food source. Data points are the average of three replicates, and error bars indicate standard deviation values. Letters designate growth phases: L—lag phase, E—exponential phase, LE—late exponential phase, and S—stationary phase.

Figure 3

Nomarski images of T. thermophila grown with MWCNT-encrusted bacteria (a,b) and control bacteria (c,d), taken before (a,c) and after (b,d) separation steps (differential centrifugation and density gradient centrifugations in iodixanol) at different growth phases. The round shapes inside protozoan cells are food vacuoles filled with bacteria. White arrows indicate bacteria provided as food source and black arrows show protozoan fecal pellets, i.e., secreted food vacuoles containing digested or partly digested bacteria.

Figure 4

Representative phase contrast images of the pelleted fraction after centrifugation of T. thermophila suspensions through 10% iodixanol. The two images serve as replicates and are both shown since bacterial agglomerates (left image) and free bacterial cells (right image) were each observed. White arrows indicate bacterial agglomerates or cells that pelleted with the protozoan cells.

Figure 5

Protozoan cell recovery percentages after density gradient centrifugation steps in iodixanol when sampled at four different protozoan growth phases. Protozoa were grown with control (untreated, white bars) or MWCNT-treated bacteria (blue bars). Data bars represent the mean values of three measurements and error bars are standard deviations. The values of “Number of protozoan cells before centrifugation” denote the total protozoan cell number in the concentrated samples after differential centrifugation but before density gradient centrifugation in iodixanol. Asterisks indicate significant difference from the recovery percentage at lag phase, both in control and MWCNT-amended exposure, using Tukey’s multiple comparisons test across all eight conditions, p ≤ 0.05 (**) and p ≤ 0.1 (*). There was no significant difference between the percentages of recovered protozoan cells grown with control bacteria (white bars) versus protozoan cells grown with MWCNT-encrusted bacteria (blue bars) within each growth phase.

Figure 6

(a) Comparison of food vacuole numbers in protozoan cells before and after differential and density gradient centrifugations; and (b) correlation between the food vacuole numbers and MWCNT mass per protozoan cell when sampled at four different growth phases (indicated by the letters: L—lag phase, E—exponential phase, LE—late exponential phase and S—stationary phase). Numbers at each data point in panel (a) indicate mean food vacuole numbers before and after centrifugation (x; y), which were not statistically different based on the two-sample t-test (p ≤ 0.05) for each of the growth phases. Food vacuole numbers per cell are the mean values of 5 to 11 individual cells. MWCNT masses are based on the mean values of three LSC measurements of ¹⁴C associated with radiolabeled MWCNTs. Error bars indicate standard deviations.