In order to compare agglomeration properties of CNP, various suspension media vehicles were selected based on the types used in previous publications. CNP were suspended at 5 mg/ml and 10 μl samples were analyzed by light microscopy at 400× magnification to examine relative CNP agglomeration states. Figures 1 thru 7 are organized whereby each figure represents a suspension media. The progression from Figure 1 to Figure 7 represents the best CNP dispersal media to the worst CNP dispersal media. CNP types from different sources are placed side-by-side for reference (e.g., A compared to B, C compared to D, and E compared to F). Figures A, C and E are CNP from SES Research and B, D and F are CNP from alternative sources (described below).

Descriptive data (median size area and maximum size area) on all dispersed CNP can be found in Table 1. These data were obtained by ImagePro software as described in Methods. Vehicles that produced mass agglomeration (e.g., SWNT and MWNT in 1% tween 80 and DMSO), were omitted from the table and analyses. All images that could be analyzed for particle area produced similar exponential histograms with a large number of smaller agglomerates and a small number of large agglomerates (with some being very large relative to the median agglomerate area). A larger median area indicates the presence of more frequent large agglomerates, whereas the maximum area is the area for the largest single agglomerate analyzed indicating the most extreme agglomerate state for a particular CNP in a particular media. Taken together, the results in Table 1 indicate that the median areas for all suspended agglomerates are relatively consistent regardless of particle type or vehicle. In contrast, the area of the largest observed agglomerate is extremely variable depending on particle type, particle source and vehicle used.

CNP suspended in 100% FCS are shown in Figure 1. C60CS dispersed well with only a few visible large agglomerates (Figures 1A and 1B). SWNT (Figures 1C and 1D) dispersed with uniform small CNP agglomerates. The only visible difference between the SWNT agglomeration states is that the SWNT from SES agglomerated in clumps, whereas the SWNT from CNI agglomerated in fibre-like stands. The MWNT appeared to agglomerate in uniform small to medium sized clumps regardless of source (Figures 1E and 1F).

CNP suspended in 7.5% BSA/PBS are shown in Figure 2. There was a significant difference in how the C60CS dispersed in this media with the SES C60CS forming large and small agglomerates (Figure 2A), with the C60CS from BuckyUSA forming only few visible small agglomerates (Figure 2B). The SWNT appeared to disperse in a similar manner to what was described in Figures 1C and 1D, with the formation of uniform small agglomerates (Figure 2C and 2D). The MWNT appeared to disperse better in this media compared to FCS with the formation of small visible agglomerates (Figures 2E and 2F). However, larger agglomerates appeared in the MWNT from NanoLab (Figure 2F).

CNP suspended in RPMI media with 10% FCS are shown in Figure 3. With regard to C60CS, this vehicle was very similar to the BSA/PBS with large agglomerates only appearing in the SES C60CS sample (Figure 3A), although larger agglomerates were also present in the BuckyUSA C60CS sample (Figure 3B). This is possibly the best dispersal media for SWNT and MWNT from SES as only small agglomerates were visible (Figure 3C and 3E). In contrast, SWNT from CNI became a swirling mass agglomerate (Figure 3D), and MWNT from NanoLab formed very large agglomerates (Figure 3F). This figure illustrates how CNP from different sources can be completely different with regard for formation of agglomerates in a particular vehicle.

CNP suspended in 100% delipidated FCS are shown in Figure 4. This vehicle appears to be a good dispersing media for all CNP with the exception of C60CS from SES where very large agglomerates formed (Figure 4A), and MWNT from NanoLab with the formation of some larger agglomerates (Figure 4F). The other CNP only created small agglomerates in this vehicle.

Figure 5 illustrates the CNP suspended in 1% tween 80 in PBS. This represents the first of the generally poor dispersal vehicles. Large agglomerates appeared with C60 from SES (Figure 5A). In contrast, this vehicle dispersed well for C60CS from BuckyUSA (Figure 5B). The SWNT from SES formed a cloud of agglomerated CNP with some more solid agglomerates visible (Figure 5C). The SWNT from CNI formed a swirling massive agglomeration of CNP (Figure 5D). MWNT from SES dispersed relatively well in the tween 80 vehicle (Figure 5E), whereas the MWNT from NanoLab agglomerated similarly to the SWNT samples described above (Figure 5F).

The results shown in Figure 6 represent another poor dispersal vehicle, PBS alone. The PBS vehicle was ineffective in dispersing both SWNT and MWNT regardless of source, as massive CNP clumps were apparent (Figure 6C–F). In contrast, the C60CS from both sources appeared to be relatively dispersed (Figures 6A and 6B).

The worst dispersal vehicle (100% DMSO) is shown in Figure 7. All CNP tested formed large agglomerates in this vehicle. The C60CS were characterized by large solid clumps (Figure 7A and 7B), and the other CNP formed large loose clusters of agglomerated nanoparticles.

Figure 8 represents the control particle, which is a crude fullerene carbon ash dispersed in all 7 of the vehicles tested. All of the vehicles produced a similar pattern of dispersal with very small to medium sized agglomerates formed. The number of agglomerates was larger in these samples due to the density of the particle compared to CNP. This figure illustrates that the differences seen in the CNP suspended in various media above were due to structural/physical properties in the specific CNP particular to the nano scale. Crude carbon particles not on the nano scale do not react to the differences in the vehicle makeup.

In order to determine the effect of the CNP vehicle on lung dispersion during particle instillation, BALB/c mice were given 250 μg instillations in either 100% FCS or PBS (representing the two extreme agglomerate states of the SWNT). After 24 hours the lungs showed significant differences in SWNT deposition as shown in Figure 9. Figure 9A (PBS) and Figure 9B (100% FCS) represent the respective vehicle controls at 100× magnification. The comparative SWNT dispersion patterns can be found in Figure 9C and Figure 9D, respectively. The blackened areas represent the accumulation of SWNT which is more generally dispersed in Figure 9D which used the 100% FCS vehicle. This observation was more pronounced at higher magnifications (200×), as illustrated in Figures 9E and 9F. Lung histology prepared 7 days post-CNP instillation showed only trace amounts of SWNT in the lungs using the PBS vehicle (Figure 9G). In contrast, the SWNT instilled with 100% FCS had persisted in the lung tissue and produced areas of increased cellularity indicative of sustained inflammation (Figure 9H).

Several CNP were used for these studies. The particles used and their respective sources are listed as follows: C60CS from SES Research (Houston, TX), C60CS from BuckyUSA (Houston, TX), SWNT from SES Research (Houston TX), SWNT from Carbon Nanotechnologies Inc. (CNI, Houston, TX), MWNT from SES Research (Houston, TX), MWNT from Nanolab (Newton, MA), and crude fullerene (elemental carbon not on the nanoscale) was used as a control particle (generous donation from Maria Morandi). Detailed CNP characterization can be found in Table 2.

Various suspension media were used in these studies in order to demonstrate the hydrophobic characteristics of CNP, and the potential need for a lipid interaction, causing better particle dispersion. Seven media were used in total and they were as follows: cell culture media RPMI (HEPES buffered w/L-glutamine, MediaTech) with 10% fetal calf serum (FCS), FCS alone (Hyclone), phosphate buffered saline (PBS, Diamedix), dimethyl sulfoxide (DMSO, stored under argon in amber vials, Sigma, St. Louis, MO), 1%Tween80 (Sigma, St. Louis, MO) in PBS, delipidated FCS (Biomeda Corporation), and 7.5% bovine serum albumin (BSA)/PBS solution (Sigma, St. Louis, MO). All particle suspensions were freshly prepared before instillation by suspension in sterile media and dispersed by sonication (Sonicator Ultrasonic Processor – Misonix, Farmingdale, NY) for one minute and vortexed for 10 sec before sonication and prior to mounting onto microscope slide.

The various stock CNP suspensions were made at a concentration of 5 mg/ml and 10 μl was pipetted onto a pre-cleaned and charged microscope slide, and cover-slipped. They were photographed at 400× under white light using Nuance Imaging software. In order to determine particle size and count for frequency distribution analysis, ImagePro software was employed to determine particle frequency and area. The threshold sensitivity of the imaging was set to include all visible particles. The area for each particle was calculated in number of pixels. This pixel count was then converted to square microns. One square micron was equal to 12544 pixels. Particle area was calculated as number of pixels per particle divided by 12544 equalling square microns of particle area.

Balb/c mice (Jackson Laboratories, Bar Harbor, ME) were used for all of the in vivo studies. All mice were used at 12 weeks of age. Animals were housed in microisolators on a 12:12 h light-dark cycle. The mice were maintained on an OVA-free diet and given deionized water ad libitum. Euthanasia was performed by intraperitoneal injection of a lethal dose of sodium pentobarbital. All animal procedures were approved by the University of Montana Institutional Animal Care and Use Committee.

All CNP exposures were administered intratracheally at a dose of 250 μg per mouse. Mice were anesthetized using 100 mg/kg Ketamine and 5 mg/kg Xylazine via IP injection. Mice were checked after a few minutes to verify that the anesthetic agents were effective. The mouse was then placed on its back and feet secured with tape. Hair was shaved from the upper thorax area and ethanol was applied. An incision, approximately 1 cm in length was made using a sterile blade in upper thorax area. Salivary glands were pushed aside to expose the trachea. Particulates were administered by injection into the trachea using a 23G sterile needle. The incision was closed with VetBond adhesive. The mice were monitored until they fully recovered from the procedure and then returned to the animal care facility. The mice were left for one or seven days before removal of the lungs for histological analysis.

Mice were euthanized by intraperitoneal injection of a lethal dose of sodium pentobarbital. The whole lungs were removed are fixed in 4% paraformaldehyde overnight. They were rinsed in a phosphate saline buffer (PBS) and processed for paraffin embedding. Sections were cut at 5 μm and mounted onto charged microscope slides. They were stained with hematoxylin and eosin (basic H&E stain) and analysed under white light at various magnifications.

Nonparametric descriptive statistics including median area, and maximum area were used to describe the particle distributions. When particle distributions could be determined the distribution had an exponential curve with the frequency of smaller particle areas much greater than the frequency of larger particle areas resulting in the exponential fall off in frequency over increasing area.