Electron microscopy of antibody-conjugated, lutetium-177 lanthanide gold-coated nanoparticles: Proof of concept of targeted loci—A potential theranostic agent

Nanoparticles (NPs) are being investigated as a potentially very efficient medium for targeted drug delivery with direct relevance to cancer treatment.

This study was undertaken to verify and prove the accuracy of the targeted NPs in our prior study¹ and to document the pharmacokinetics in vivo by demonstrating the cellular loci of the injected NPs. In that report, the Single Photon Emission Computed Tomography (SPECT) scans revealed the NP uptake in the lung, liver, spleen, and small bowel. The electron microscopic (EM) images in this study demonstrate the unequivocal targeting of the NPs toward the thrombomodulin antigen loci in the lungs. The progression of the NPs from the lungs to liver to biliary tract to small bowel portrays the pharmacodynamics of the NPs in vivo, paving the way for their use as cancer theranostics. This study demonstrates the elegant juncture of biology, physics, and engineering.

In our prior report, 300 µCi/190 µl of [¹⁷⁷Lu]Lu_0.5Gd_0.5(PO₄)@Au@PEG₈₀₀@Ab anti-thrombomodulin nanoparticles (NPs) were injected intravenously into two groups of mice: group 1 (control group of three untreated mice, no clodronate) and group 2 (experimental group of three mice that received intraperitoneal administration of 100 µl clodronate liposomes 72 h prior to NP injection). Clodronate liposomes (bisphosphonate) were injected to deplete the circulating macrophages and to prevent most of the macrophage phagocytosis of the circulating NPs.¹ Consequently, the circulating NPs successfully attached to their targeted antigens. The use of clodronate in group 2 resulted in a 47% increase in injected conjugated NP dose accumulation in the lungs compared to group 1.¹ In our previous work, the effect of the polyethylene glycol (PEG) linker length on the biodistribution of the radiolabeled Lu_0.5Gd_0.5(PO₄)@Au@PEG nanoparticles was investigated by changing the length of the PEG linker from 800 to 5000 Da. The [¹⁷⁷Lu]Lu_0.5Gd_0.5(PO₄)@Au@PEG800 nanoconjugate with the shortest PEG linker (800 Da) accumulated more rapidly and predominately in the lungs—85% of the injected dose (ID) early after the intravenous administration in clodronate-treated mice [Fig. 1(a)]. Over the course of 24 h, the [¹⁷⁷Lu]Lu_0.5Gd_0.5(PO₄)@Au@PEG800 nanoconjugates were filtered through the lungs and excreted through the liver and spleen (reticuloendothelial system). Despite the fact that both groups showed a dominant clearance pathway through the liver and spleen, a higher steady accumulation of PEG conjugated nanoconjugates was found in the liver of mice that were not treated with clodronate [Fig. 1(b)]. As the PEG linker was shortened, three phenomena were observed: maximum lung retention increased, spleen uptake decreased, and the observed results between the clodronate and untreated mice became more pronounced.

The SPECT scans described in our previous work demonstrated uptake in the lungs, liver, and spleen, with a decrease in the lungs and an increase in the liver and spleen ∼24 h after injection, as estimated with the Standard Uptake Value (SUV) analysis.¹ Uptake values in mice that were previously administered clodronate [Fig. 2(a)] show a much higher initial (40%) uptake by the lungs (target organ). Note that the concentration in the lungs remains high for over an hour and the concentration in the liver and spleen slowly increases over 24 h, with levels in the spleen approaching those initially seen in the lungs.

Figure 3 shows rapid lung uptake in both the control and clodronate-treated mice. There is also more concentration in the spleen in clodronate-treated mice and more concentration in the liver in the untreated mice. Note the early activity in the small bowel [marked with an asterisk (^*)] and liver. There is also a marked increase in activity in the spleen at 1440 and 1200 min.

The tissues obtained and examined for this report were acquired during the last study in accordance with the Canadian Council on Animal Care (CCAC), and the protocol was approved by the Animal Care Committee (ACC) of the University of British Columbia (A16-0150).¹ 

After the SPECT scans, the mice were killed by CO₂ asphyxiation under isofluorane anesthesia. Cardiac puncture was promptly performed to recover blood. The organs of interest (lungs, liver, spleen, and kidneys) were then harvested and placed in a modified Karnovsky fixative—94% pfa, 2.5% glutaraldehyde, and 0.02% picric acid in 0.1M sodium cacodylate buffer (pH 7.4)—for 24 h at 4 °C. After fixation, the tissues were submerged in 0.1M sodium cacodylate buffer solution (pH 7.4) and stored at 4 °C. The tissues were then washed and placed in a fresh buffer every 14 days for 66.5 days (ten half-lives of Lu-177). After ten half-lives of decay, no significant radiation remained, and the tissues were prepared for EM. The samples were washed one more time in the buffer and fixed in osmium tetroxide (OsO₄) [1% OsO₄-1.5%K-ferricyanide (aqueous)] for one hour. They were washed three more times with the buffer and one time with double distilled 18 MΩ water. The samples were stained with 1.5% uranyl acetate in water for one hour. Next, they were dehydrated through an ethanol series—50%, 70%, 85%, 95%, 100%, 100%, and 100% ethanol (EtOH)—sequentially for 15 min each. The samples were placed in 100% EtOH mixed 1:1 with acetonitrile for 10 min followed by acetonitrile (undiluted) for 15 min and acetonitrile 1:1 with inactive epoxy resin. They were then stored overnight in a sealed vial at room temperature. The next day, the samples were immersed in activated resin (with a catalyst), which was changed after 8 h prior to infiltration with an Epon analog resin (EMbed812-Electron Microscopy Sciences, methyl-5-norbornene-2-3-dicarboxylic anhydride [NMA], dodecenylsuccinic anhydride [DDSA], and benzyldimethylamine [BDMA]). Finally, the samples were embedded in fresh resin and polymerized in a 50 °C oven for 36 h.

The samples were cut using a DiATOME Ultra 45° diamond knife (Hatfield, PA), using Leica Ultracut T (Wetzlar, Germany) and MT-6000 microtomes (RMC-Boeckeler, Tucson, AR). The sections were placed on 200-mesh copper grids (EMS, Hatfield, PA). Four out of every five grids were placed in a 0.2% lead citrate solution for 3 min and then washed with 0.01N NaOH and then deionized water. A short staining time prevented the retained lead citrate particles from obscuring the NPs. Those grids were imaged using a JEOL JSM 1400 electron microscope (Akishima, Tokyo) at 5k, 20k, and 50k magnifications. The “fifth grid” was not stained with lead to determine whether the lead particles obscured the NPs. The difference between the lead-stained and unstained grids was negligible, that is, the lead particles did not obscure the NPs.

Figures 1–3 illustrate the localization of NPs in type I and type II pneumocytes. The ultrastructural localization of gold-coated nanoparticles (NPs) containing lutetium-177 (Lu-177) conjugated to the anti-thrombomodulin antibody (mAb-201b) following injection into mice was determined. Our findings correlate with prior work localizing NPs using SPECT scans. We previously described the methods to make the NPs and how the SPECT scans demonstrated their biodistribution in the body via gamma spectroscopy of lung, liver, spleen, and kidney tissues.¹ mAb-201b antibodies are attracted to the thrombomodulin, an integral membrane protein that is ubiquitous in the endothelium of the pulmonary vasculature. The circulating macrophages, which phagocytose NPs, were reduced by prior injection with clodronate (bisphosphonate) liposomes. Although some NPs were found in four organs, 85% of the injected dose was localized in type I and type II pneumocytes [see Figs. 4, 5(a), and 5(b)], but no other cell types in the lungs. Type II pneumocytes are identified by their lamellar bodies.⁹ Small concentrations were found in secondary lysosomes in hepatocytes, in splenic macrophages, and in an intravascular monocyte/macrophage in the kidney (see Fig. 6). There was no evidence of inflammation. There was no NP clumping or aggregation. There was no EM evidence of apoptosis or necrosis. No surfactant was visible. Presumably, it was washed away from the lining of the alveolar capillary epithelium by perfusion fixation.¹⁰ 

This study (1) represents proof of principle that these NPs can target specific antigens in murine cells in vivo, (2) validates the SPECT scans from our prior work, and (3) suggests that these NPs may be used as a theranostic agent for a variety of cancers, depending on the labeling antigen.

Figures 7(a)–7(c) illustrate the localization of NPs in the liver; NPs were found only in hepatocyte lysosomes. There was no EM evidence of apoptosis or necrosis in the liver The NPs were not observed in the Kupffer cells (differentiated macrophages) lining the sinusoids of the liver.

As shown in Figs. 8(a) and 8(b), NPs were observed in macrophage lysosomes of the spleen with no EM evidence of apoptosis or necrosis. The NP concentration per lysosome in the splenic macrophages is much higher than that in the liver hepatocytes, which is consistent with the results of the SPECT scans from our prior work that also demonstrated greater activity in the spleen than the liver.

No NPs were found in the renal parenchyma, and there was no EM evidence of apoptosis or necrosis. However, NPs were found in a circulating intravascular monocyte/macrophage. This phenomenon is consistent with the minimal proportion of the injected dose localized to the kidneys, as per the measurement of the radioactivity accumulated in the kidneys (as described in our prior report).¹ 

Thrombomodulin is an intramembrane protein and the target of the anti-thrombomodulin-labelled NPs.^2,3 Following NP attachment, EM of the lung revealed the ultrastructural loci of the NP to be intracellular—inside the pneumocytes. Upon contact with a physiological environment, the NPs’ surface charge attracts a corona of proteins from the serum.⁴ When the antibody binds to its ligand on the pneumocyte cell membrane, these proteins cause the cell to interact with the NPs as a protein and to be transported inside the cytosol.⁵ The ubiquitous protein coronas—visualized as the dense, dark, “fluffy” material surrounding the NPs—are present in the electron micrographs both in this report and in the publication by Kokkinopoulou et al.⁴ These NP + protein coronas are either lysed or simply exocytosed back into circulation and picked up by the liver and spleen. The NPs in every target organ are found exclusively in the endocytes or lysosomes.

The NPs, upon first pass in the circulation, are picked up by their immunological target—thrombomodulin. However, since the cells interact with the protein corona and not the nanoparticle itself,⁵ the NPs are taken up by the pneumocytes and stored as proteins. They are presumably exocytosed back into the circulation when the cell cannot “process” them. This presumed exocytosis is based on the SUV values in our prior report, which 1 h after injection began to decrease in the lungs and increase in the liver and spleen (splenic macrophages).¹ After being exocytosed from the pneumocytes into the circulation, the NPs likely attract a new corona⁶ and continue through the general circulation to the liver. When interacting with the liver, the NPs are sequestered in hepatocytes. Although foreign material usually interacts with Kupffer cells (specialized macrophages that line the hepatic sinusoids) first, these NPs, ∼70 nm¹, are smaller than the liver’s sinusoidal fenestrations (150–200 nm)⁷ and are not sequestered in the Kupffer cells. They pass through the space of Disse and enter the hepatocytes where they are metabolized and excreted through the bile canaliculi/bile and enter the proximal small bowel, as seen on the SPECT images in our prior report.¹ 

NPs are not excreted through the kidneys. The glomerular filtration threshold of the capillary walls—35 nm in diameter—limits passage of any particles with larger hydrodynamic diameters (∼70 nm-diameter NPs).⁸ This glomerular filtration threshold demonstrates the NP pathway and physiological life through the body, starting with their antibody target and ending with excretion through the bile/stool.

This study demonstrates that antibody labeled NPs—specifically, antibody-conjugated, Lu-177 lanthanide gold-coated NPs—successfully target their specific protein antigens. The NPs are removed from circulation by liver hepatocytes and splenic macrophages. This study also demonstrates the anti-thrombomodulin-labeled NP loci. We determined the ultrastructural localization of gold-coated nanoparticles (NPs) containing lutetium-177 (Lu-177) conjugated to the anti-thrombomodulin antibody (mAb-201b) following injection into mice. Our findings correlate with prior work localizing NPs using SPECT scans. We also outline the pharmacokinetics of these conjugated NPs throughout the body, from the lungs to liver to biliary tract and small bowel. We previously described the methods to make the NPs and how the SPECT scans demonstrated their biodistribution in the body via gamma spectroscopy of lung, liver, spleen, and kidney tissues. mAb-201b antibodies are directed against thrombomodulin, an integral membrane protein that is ubiquitous in the endothelium of the pulmonary vasculature. SPECT images demonstrated NPs in the lungs, liver, spleen, and proximal small bowel. The presence of circulating macrophages—which phagocytose NPs—was reduced with injections of clodronate (bisphosphonate) liposomes, prior to administration of NP injections. At 24 h after injection of the NPs and following the final SPECT imaging as described previously, the lungs, liver, spleen, and kidneys were harvested and processed for transmission EM. Although some NPs were found in all four organs, 95% of the injected dose was localized in type I and type II pneumocytes. Small concentrations were found in secondary lysosomes in hepatocytes, in splenic macrophages, and in an intravascular monocyte/macrophage in the kidney. Furthermore, there was no apoptosis or necrosis in any of the tissues, highlighting the relative safety of radionuclide NPs, whose primary interaction with non-targeted organs/tissues is in the filtration process.

This study represents proof of principle that these NPs can target specific antigens in murine cells in vivo. The results suggest that these NPs may be used as theranostic agents only in cell types known to have membranous thrombomodulin, as shown by their ultrastructural location inside the lungs (matching the previously published SPECT data).

Demonstration of successful targeting and removal of these NPs from the body is key to their further application in biological systems, notably as a potential theranostic agent for pulmonary or other metastatic diseases. More importantly, the results suggest that appropriately targeted antibody-labeled NPs will seek primary or metastatic cancers. The data presented and cited support the conclusions, but in the work that is currently underway on other tumors and radionuclide theranostics, biodistribution and biokinetics in other organs and the central nervous system are being pursued.

The authors would like to thank the technicians at the UBC Centre for Comparative Medicine (CCM, Vancouver, Canada), especially the assistance and insights of Jana Hodasova, MVDr.

All authors contributed equally to this work.

The data that support the findings of this study are available within the article.