Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Drug capture materials based on genomic DNA-functionalized magnetic nanoparticles

Drug capture materials based on genomic DNA-functionalized magnetic nanoparticles ARTICLE DOI: 10.1038/s41467-018-05305-2 OPEN Drug capture materials based on genomic DNA- functionalized magnetic nanoparticles 1 1 2 2 2 Carl M. Blumenfeld , Michael D. Schulz , Mariam S. Aboian , Mark W. Wilson , Terilynn Moore , 2 1 Steven W. Hetts & Robert H. Grubbs Chemotherapy agents are notorious for producing severe side-effects. One approach to mitigating this off-target damage is to deliver the chemotherapy directly to a tumor via transarterial infusion, or similar procedures, and then sequestering any chemotherapeutic in the veins draining the target organ before it enters the systemic circulation. Materials capable of such drug capture are yet to be fully realized. Here, we report the covalent attachment of genomic DNA to iron-oxide nanoparticles. With these magnetic materials, we captured three common chemotherapy agents—doxorubicin, cisplatin, and epirubicin—from biological solutions. We achieved 98% capture of doxorubicin from human serum in 10 min. We further demonstrate that DNA-coated particles can rescue cultured cardiac myoblasts from lethal levels of doxorubicin. Finally, the in vivo efficacy of these materials was demonstrated in a porcine model. The efficacy of these materials demonstrates the viability of genomic DNA- coated materials as substrates for drug capture applications. Arnold and Mabel Beckman Laboratories for Chemical Synthesis, Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA 91125, USA. Interventional Radiology Research Laboratory, Department of Radiology and Biomedical Imaging, University of California-San Francisco, San Francisco, CA 94143, USA. These authors contributed equally: Carl M. Blumenfeld, Michael D. Schulz. Correspondence and requests for materials should be addressed to R.H.G. (email: rhg@caltech.edu) NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 1 1234567890():,; ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 he systemic toxicity of chemotherapy is a widely recog- suppression. Consequently, a balance must be struck in order to nized problem in oncology. Off-target damage often per- maximize drug dose, leading to better tumor suppression, while Tsists indefinitely, adversely affects patient survival, and simultaneously avoiding catastrophic off-target toxicity. Although 1,2 restricts dose and treatment options . Direct administration of limiting a patient’s lifetime cumulative dose is the most effective chemotherapy agents to the tumor via transarterial che- way to avoid cardiotoxicity, this approach necessarily limits anti- moembolization (TACE), or similar procedures, followed by cancer efficacy . sequestration of any chemotherapeutic that enters systemic cir- The unwanted systemic toxicity of chemotherapy agents has culation would mitigate this damage if materials capable of such inspired a number of more targeted approaches. One such 3,4 drug capture were fully realized . approach is TACE, during which liver blood flow is occluded in Hepatocellular carcinoma (HCC) is the third leading cause of conjunction with administration of high dose chemotherapy 5 3,16 cancer-related deaths worldwide . Liver transplantation is the directly to the tumor . Both during TACE and after liver blood most definitive approach for treatment; however, less than 30% of flow is restored, however, up to 50% of residual chemother- 6 17 HCC patients are eligible . Direct delivery of drug to a tumor via apeutics enter systemic circulation and cause off-target toxicity . intraarterial chemotherapy (IAC) and its variant, TACE, is often Efforts have been made toward reducing non-targeted toxicity used as a bridge to transplantation, shrinking HCC or at least during TACE. In 2014 Patel and coworkers proposed che- controlling its growth through recurrent treatments until curative motherapy filtration devices (“ChemoFilters”) that employed transplant is possible. In cases where surgery is untenable, che- sulfonated ion-exchange resins with affinity for DOX. Such a motherapy is often the only recourse. Targeted therapy, however, device could be deployed via catheter in the hepatic vein, does not completely eliminate side-effects. “downstream” from the site of chemotherapy administration, Three of the most common drugs used to treat HCC are where it can intercept any residual chemotherapy agents before doxorubicin (DOX), epirubicin (EPI), and cisplatin (Fig. 1c), all they reach the heart and enter systemic circulation. They of which act on DNA . DOX and EPI function by intercalating demonstrated a 52% reduction in DOX concentration from between DNA base pairs, while cisplatin is a DNA crosslinker porcine serum over 10 min, and showed that such a device could 8,9 3 that functions by binding to guanine . A major problem for be successfully deployed during a simulated TACE procedure .In these anticancer compounds is toxicity in non-targeted tissues. 2016, the ChemoFilter approach inspired the development of DOX and EPI toxicity can result in cardiomyopathy and con- more elaborate block copolymer membranes for DOX capture, 8–10 gestive heart failure . Similarly, cisplatin elicits side-effects which achieved up to 90% removal of DOX in 31 min from 11,12 4 including extensive nephrotoxicity and neurotoxicity .To phosphate buffered saline (PBS) . reduce the likelihood of cardiac toxicity, cumulative dosage of A ChemoFilter device is intended to be placed downstream of DOX is generally limited by clinicians to 400–450 mg/m , though blood outflow from the tumor that is being treated with intra- lower cumulative dosages (300 mg/m ) are known to increase the arterial chemotherapeutics. In the case of HCC, which is located 13,14 risk of congestive heart failure . Still, a single standard dose of in the liver, the ChemoFilter would be placed into the suprahe- DOX (50–75 mg) can result in severe side-effects, yet higher patic inferior vena cava (IVC) immediately prior to administra- dosages of DOX are known to be associated with greater tumor tion of IAC into the hepatic artery and then removed shortly after IONP-Pt-DNA approach CI CI H Pt N NH HN NH 2 2 Si O SiOMe Si O H O K PtCl 2 4 Fe O Fe O 3 4 3 4 H N CI Fe O 3 4 Fe O 3 4 NH N H N 2 N CI Si O Si O O N(CH CH CI) SiOMe 2 2 3 Fe O 3 4 Fe O 3 4 IONP-HN3-DNA approach bc O OH O O OH O OH OH OH DOX OH Fe O 3 4 O O OH O NH O O OH O NH OH OH Cl NH OO OH Cisplatin Pt OH OH Cl NH O O OH O NH OH OO OH OH EPI >95% DOX OH DOX removal Serum Solution O O OH O NH OH Fig. 1 Experimental approach. a Two synthetic approaches for covalently attaching genomic DNA onto iron-oxide nanoparticle (IONP) surfaces. b Drug capture concept. c Three common chemotherapy agents used in this study 2 NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 ARTICLE 3,18 drug capture has been achieved . Here we report the devel- hydroxylated surface of Fe O was silylated with N-(2-ami- 3 4 opment of a drug-capture material based on genomic DNA and noethyl)-3-aminopropyltrimethoxysilane exposing a chelating iron oxide particles. In vitro studies confirm that these materials diamine functionality. This sample was treated with an excess can rescue cells from the toxic effects of DOX. A ChemoFilter of potassium tetrachloroplatinate to create an analog of cisplatin device is constructed from this material that rapidly removes by which DNA could be anchored to the surface. Cisplatin’s chemotherapy agents from relevant biological solutions, including cytotoxicity is thought to stem from its coordination with human serum and porcine blood, and in vivo studies confirm that nucleophilic N7-sites of purine bases, resulting in crosslinks . DOX can be removed from the bloodstream by an intraarterial We hoped to accomplish DNA crosslinking to the surface device constructed from these iron oxide/genomic DNA through this mechanism. The sample was then exposed to DNA materials. to produce IONP-Pt-DNA. The second approach was modeled on nitrogen mustard chemotherapy agents. IONP-HN3-DNA samples were prepared Results and discussion first by functionalizing Fe O with 4-aminobutyltriethoxysilane to Material design and synthesis. Inspired by the ChemoFilter 3 4 install free amines on the surface. This particle was then treated concept, we designed and synthesized DNA-functionalized with excess tris(2-chloroethyl)amine hydrochloride (HN3·HCl) to materials based on magnetite (Fe O ) nanoparticles, capable of 3 4 create a scaffold for DNA functionalization. HN3·HCl, the rapidly capturing chemotherapy agents. Central to our approach hydrochloride salt of the nitrogen mustard HN3, undergoes is the direct covalent attachment of genomic DNA. Functiona- aziridinium formation when deprotonated, and is attacked readily lizing surfaces with DNA has historically involved tagging either by the nucleophilic moieties of DNA . The functionalized the backbone or bases of synthetic DNA with an appropriate particle was exposed to DNA resulting in IONP-HN3-DNA. Both moiety, or attaching the DNA via a reactive end-group. These materials were characterized by scanning electron microscopy, approaches are highly useful and enable complete control of the electron dispersive scattering (EDS), elemental analysis, and DNA sequence used resulting in the development of numerous 19–26 infrared spectroscopy (see Supplementary Figs. 2& 3 and 11–24, interesting materials ; however, they are limited by the rela- and Supplementary Table 1). Microscopy images of the particles tively high cost of synthetic DNA. The synthesis of large amounts in solution revealed significant aggregation resulting in an average of such materials would be prohibitively expensive for most particle diameter of 4.2 μm with several larger (>10 μm) applications. aggregates. Elemental analysis indicated that these aggregates Functionalization with genomic DNA is an alternative were 18% DNA by mass in the case of IONP-HN3-DNA and approach that may be appropriate for certain applications; 14.7% DNA by mass in the case of IONP-Pt-DNA. however, this approach is relatively unexplored. Pierre and coworkers recently synthesized magnetic nanoparticles with surface-bound intercalating groups, and showed that such In vitro testing in simple solutions, serum, and blood. In order materials can bind to genomic DNA . To our knowledge, to evaluate the efficacy of our materials at scavenging che- however, no one has reported the covalent attachment of genomic motherapy agents from solution we studied DOX-binding in PBS DNA to a surface. Here, we report two methods of attaching and human serum at 37 °C to approximate the biological envir- genomic DNA to nanoparticles, both on multi-gram scale onment in which these materials would have to operate (Fig. 1b). (Fig. 1a). We show that the resulting materials are capable of We found that IONP-HN3-DNA was able to capture 93% of removing DNA-targeting chemotherapy agents from solution DOX, on average, from a 0.05 mg/mL solution of human serum both rapidly and in the presence of potential biological in 25 min, while IONP-Pt-DNA averaged 79% (Fig. 2a). In both intereferents (e.g., serum proteins and other blood components). cases, the kinetics were extremely rapid, with about 50% of DOX DNA-alkylating agents are a common motif in chemotherapy. capture occurring within 1 min in the case of IONP-Pt-DNA and By forming covalent crosslinks between DNA strands, these drugs over 65% DOX capture occurring within 1 min for IONP-HN3- prevent the DNA from being accurately duplicated, ultimately DNA. Based on these results, we carried out all further tests with leading to apoptosis. To attach genomic DNA to magnetic IONP-HN3-DNA. nanoparticles, we used an approach analogous to DNA-alkylat- Interestingly, both materials were highly effective, despite the ing/crosslinking drugs (Fig. 1a). The first approach was inspired known binding of DOX with serum albumin. It is known that 30,31 by cisplatin. To synthesize IONP-Pt-DNA samples, the DNA intercalation is the kinetically more favorable process , a b 0.05 IONP-HN3-DNA Control 0.9 IONP-Pt-DNA IONP-HN3-DNA 0.8 0.04 0.7 0.6 0.03 0.5 0.4 0.02 0.3 0.2 0.01 0.1 0 0 0 5 10 15 20 25 30 02468 10 12 Time (min) Time (min) Fig. 2 DOX capture in human serum and porcine blood. a Decrease in DOX concentration in human serum, determined by fluorescence, as a result of DOX capture by IONP-HN3-DNA and IONP-Pt-DNA; 100 ± 5 mg particle in 20 mL (0.05 mg/mL), 1 mg total DOX, 37 °C; error bars= 1 standard deviation (n = 3). b Decrease in DOX plasma concentration as a result of DOX capture by IONP-HN3-DNA from porcine whole blood; 100 ± 5 mg IONP-HN3-DNA in 20 mL (0.05 mg/mL), 1 mg total DOX, 37 °C; error bars= 1 standard deviation (n = 3) NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 3 DOX concentration (mg/mL) Normalized DOX concentration ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 and we believe that this kinetic advantage enabled our material to material added per mg DOX, resulting in ~90% DOX capture in capture DOX from serum solution, despite the thermodynamics 10 min. Further DOX capture appears less favorable after this being in favor of serum binding overall. We posit that over longer point. We believe this plateau is the result of competition with timescales, serum binding would be the dominant process; serum binding, which makes that portion of DOX unavailable for however, since TACE is a relatively short procedure (<1 h), we capture by our particles, as well as the typical kinetic effects of believe that kinetic factors will dominate in the performance of diminishing concentration. The absorption of DOX onto the any material or device. particles was further verified by performing confocal fluorescence Drug capture was also evaluated in porcine whole blood, by microscopy (Fig. 3c and d). This technique allowed visualization measuring DOX plasma concentration over time. We observed of the fluorescence of DOX bound to the surface of the particles. some DOX removal due to binding to the non-plasma blood Our approach is general for all DNA-targeting chemotherapy components, which we cannot deconvolute from capture by our agents. To demonstrate this fact, we performed further experi- materials. Nevertheless, there is rapid reduction of DOX ments on two additional common DNA-targeting chemother- concentration in the blood plasma within 1 min after exposure apeutics, cisplatin and EPI. We performed an initial cisplatin- to our material, reaching a 92% reduction in DOX plasma binding experiment in PBS solution with IONP-HN3-DNA and concentration over 10 min, in stark contrast to the control monitored the decrease of cisplatin concentration by inductively experiment (Fig. 2b). This experiment conclusively demonstrates coupled plasma-mass spectrometry (ICP-MS). Approximately that our materials are capable of capturing DOX from whole 20% of the cisplatin was captured from solution over 30 min, with blood. little improvement over longer time periods (see Supplementary To better understand the DOX-capture capacity of IONP- Fig. 4). We confirmed the presence of captured cisplatin on the HN3-DNA, we performed a series of experiments in which surface of the particles by x-ray photoelectron spectroscopy. We nanoparticle loading was systematically varied (Fig. 3a, b, further believe these relatively low levels of drug capture are due to data in the Supplementary Information). These experiments cisplatin not being in the aquo state, which happens intracellu- revealed a roughly linear trend in DOX-capture as a function of larly and is necessary in order to bind to DNA. Because the the amount of particle added, up to a plateau around 100 mg particles were not highly effective for cisplatin capture in PBS, we did not perform further experiments in more complex media such as serum or blood. 0.05 Along with DOX and cisplatin, EPI is among the most commonly used chemotherapeutic agents for treating HCC. We 0.04 evaluated the efficacy of our materials for capturing EPI using a set of experiments analogous to those we used with DOX (see 0.03 Supplementary Fig. 5). Our particles were highly effective at 0.02 sequestering EPI from serum, with 68% captured after 25 min. The sequestered amount would lead to a reduction in unwanted 0.01 side-effects if achieved in vivo. 0 2468 10 12 In vitro evaluation of biological efficacy. The ability of IONP- Time (min) HN3-DNA to detoxify DOX was tested in vitro in an H9C2 rat 10 mg 20 mg 30 mg 40 mg 50 mg 100 mg IONP-HN3-DNA heart myoblast cell culture assay (Fig. 4). These experiments demonstrated that DNA-coated particles could rescue cultured cardiac myoblasts from lethal levels of DOX more effectively than the ion exchange resin Dowex, which itself had been previously shown to reduce levels of DOX in vivo . DOX 10 mg 20 mg 40 mg 30 mg 100 mg Serum solution In vivo evaluation of drug capture. A device (Fig. 5a) consisting of IONP-HN3-DNA magnetically adhered to the surface of cylindrical rare-earth magnets strung along a PTFE-coated nitinol wire was evaluated using a closed loop flow model (see Sup- c d plementary Fig. 25) and subsequently tested in vivo using a porcine model. The device was inserted into the IVC and DOX was injected over 10 min at a rate of 2.5 mL/min into the left common iliac vein proximal to the device (Fig. 5b). As the drug flowed through the IVC, it made contact with the IONP-HN3- DNA adherent to the surface of the device and was captured. Blood aliquots were taken proximal to (upstream), adjacent to the midpoint of, and distal to (downstream from) the device using separate catheters. Peak DOX concentration was observed at 3 min, since the blood at the injection site had recirculated and live injection was still underway. At 3 Fig. 3 DOX capture with IONP-HN3-DNA particles. a DOX capture as a min (peak concentration), a 60% reduction in serum DOX function of the amount of IONP-HN3-DNA from a DOX serum solution concentration was observed half-way across the device, while (0.5 mg total DOX, 0.05 mg/mL); average of three experiments, data set a total reduction of 82% was observed at the end of the devi- with error bars in Supplementary Fig. 6. b Resulting serum solutions from ce (Fig. 5c, d). experiments summarized in a. c Brightfield image of IONP-HN3-DNA In conclusion, we have demonstrated two viable synthetic aggregates bound to DOX; scale bar = 50 μm. d Fluorescence from DOX pathways to genomic DNA-functionalized magnetic particles, bound to IONP-HN3-DNA; scale bar = 50 μm both on multi-gram scale. Moreover, these methodologies for 4 NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications Doxorubicin concentration (mg/mL) NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 ARTICLE a b DOX + Dowex DOX c DOX + IONP-DNA d Control DOX DOX + DOX + Dowex IONP-DNA Fig. 4 IONP-HN3-DNA rescue cultured cells from DOX toxicity (scale bar = 5 μm). a Rat heart myoblasts H9C2 cultured with DOX (0.05 mg/mL)— myoblasts are killed by adding DOX to culture (DOX, d). b Myoblasts treated with DOX (0.05 mg/mL) and 0.19 g Dowex ion-exchange resin beads— myoblasts are partially rescued ((DOX+ Dowex, d). c Myoblasts treated with DOX (0.05 mg/mL) and 0.19 g IONP-HN3-DNA—myoblasts are completely rescued (DOX+ IONP-DNA, d). d Cell counts for each experimental condition; error bars= 1 standard deviation (n = 3) a b 0.03 Pre 0.025 Mid 0.02 Post Peripheral 0.015 Pre-device Mid-device Post-device 0.01 0.005 0 5 10 15 20 25 30 Time (min) Fig. 5 In vivo results. a Device containing 25 magnets (1 cm × 0.5 cm) with IONP-HN3-DNA coating (above), and the same device after the in vivo experiment (below), demonstrating minimal loss of particles after removal of the device. b Fluoroscopy images during in vivo porcine experiment demonstrating the inferior vena cava with opacified right renal veins (scale bar = 5 mm). The device was placed within the inferior vena cava. The sampling catheters were placed immediately proximal to the device, prior to the renal vein, and distal to the device. c DOX concentration measurements from pre- device, mid-device, post-device, and peripheral locations. d Plasma solutions from the experiment described in c: Left 3, pre-filter samples taken at 1, 3, and 10 min; Center 3, mid-filter samples taken at 1, 3, and 10 min; Right 3, post-filter samples taken at 1, 3, and 10 min NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 5 Concentration (mg/mL) Cell number ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 DNA surface functionalization are not limited to magnetic metal introduced through the left internal jugular vein. Prior to the start of the experi- ments, patency of the venous system was demonstrated using contrast injection oxides, but may also be exploited for other substrates. The (Omnipaque). DOX was injected over 10 min at a rate of 2.5 mL/min into the left synthesized materials captured three commonly used chemother- common iliac vein proximal to the magnetic device. The pre-device DOX con- apy agents from relevant biological solutions (PBS, human serum, centration was measured by sampling with a 5 Fr catheter downstream of the DOX or porcine whole blood), at therapeutically relevant concentra- infusion. Blood aliquots were taken proximal to, adjacent to the midpoint of, and distal to the device using separate catheters. To clear the sampling catheters, 2 mL tions and timescales. A proof of concept device was developed, of blood was drawn immediately prior to taking the aliquot (3 mL). The blood which demonstrated efficient capture of DOX in vivo. Similar samples were placed on ice until they were centrifuged to isolate the plasma devices could be readily developed that would potentially reduce fraction for analysis. A control experiment was also performed using the same the off-target toxicity and damaging side-effects associated with procedures but with no device inserted. the use of DOX, cisplatin, and EPI during TACE or similar procedures. Ultimately, we believe our approach is general for all Particle synthesis. IONP-Pt: 3.31 g Fe O was dried in vacuo at 120 °C. Upon DNA-targeting chemotherapy drugs, and while further 3 4 cooling, the sealed material was introduced into an inert atmosphere nitrogen development is needed, we hope that this work will provide a glovebox. To the Fe O was added 23 mL anhydrous toluene along with 4 mL N- 3 4 foundation for future work on DNA-based materials and drug (2-aminoethyl)-3-aminopropyltrimethoxysilane. The reaction was mechanically capture approaches both for oncologic and non-oncologic stirred on the bench at 110 °C for 2 h and subsequently dried in vacuo at 110 °C for applications. 20 h. The reaction mixture along with 1.0 g K PtCl , was stirred at 70 °C for 21 min 2 4 and then washed three times with water. Following this, the mixture was diluted to a total volume of 450 mL with 18.2 MΩ water was treated with 1.3 g KCl and an Methods additional 10 mL water. Instrumentation. Fluorescence measurements were made using a 96-well plate on IONP-Pt-DNA: IONP-Pt materials along with 5.1 g deoxyribonucleic acid from a Molecular Devices FlexStation 3 Multi-mode microplate reader. Scanning elec- herring sperm were mechanically stirred in 450 mL 18.2 MΩ water at 37 °C for tron micrographs (SEM), as well as EDS measurements were made on a Zeiss 20 h. To ensure covalent attachment as opposed to being physically adsorbed, the 1550VP field emission SEM equipped with an Oxford EDS module. ICP-MS was particles were isolated from the reaction mixture, washed three times under carried out on an HP 4500 ICP-MS equipped with a Cetac ASX-500 autosampler. vigorous mechanical stirring with 18.2 MΩ water (400 mL) in order to remove Infrared measurements were made on a Nicolet iS50 Fourier transform infrared unbound DNA, frozen, and lyophilized to afford 3.08 g IONP-Pt-DNA (79% yield spectrometer equipped with a DuraScope ATR unit. C, H, N analyses were carried based on elemental analysis of DNA content). out using a PerkinElmer 2400 Series II CHN Elemental Analyzer. Fluorescence IONP-NH2: 4.2 g Of Fe O was dried in vacuo at 120 °C. The Fe O was 3 4 3 4 microscopy was performed on an inverted laser scanning confocal Zeiss LSM 710 allowed to cool to room temperature under vacuum. To the Fe O was added to 3 4 microscope equipped with an argon laser and photomultiplier tube detector, and 25 mL toluene (freshly dried over magnesium sulfate) and 3.2 mL 4- particle size was determined by image analysis of at least 100 particles measured on aminobutyltriethoxysilane. The reaction was sealed and stirred mechanically for 2 h their widest dimension. at 120 °C. The reaction was removed from heat and the particles were isolated from the toluene solution. The reaction mixture was washed once with toluene and subsequently dried in vacuo at 120 °C for 1 h and 45 min. 4.02 g of IONP-HN3 was General procedures. Unless otherwise stated reactions were carried out on the isolated. bench. Fe O (40 nm APS, 99%) was purchased from Nanostructured & Amor- 3 4 IONP-HN3-DNA: 3.4750 g IONP-HN3 was added to a vial along with 1.02 g phous Materials, Inc. Silane reagents were purchased from Gelest, Inc. Genomic HN3·HCl and dimethylformamide (30 mL). The reaction was stirred mechanically DNA (isolated from Herring sperm), human serum (OptiClear), H9C2 rat heart for 1 h at room temperature at which point, the particles were isolated from the myoblasts, and cisplatin were purchased from Sigma Aldrich. DOX was purchased dimethylformamide. The particles were then washed three times with from LC Labs and EPI was purchased from Biotang Inc. Potassium tetra- dimethylformamide. The isolated particle as well as 3.35 g deoxyribonucleic acid chloroplatinate was purchased from Pressure Chemicals. All reagents not otherwise from herring sperm were transferred into a flask along with 400 mL 18.2 MΩ water. mentioned were purchased from Sigma Aldrich, and were used without further The reaction was mechanically stirred at 38 °C for 17 h and 45 min. To ensure purification. covalent attachment and to remove any unbound DNA, the particles were then washed thoroughly under vigorous mechanical stirring three times with 18.2 MΩ Device construction. Twenty-five cylindrical rare-earth magnets (N52 grade, water (400 mL) and magnetic separation. The particles were then frozen in liquid nitrogen and lyophilized to afford 3.79 g of IONP-HN3-DNA (89% yield as 5 mm OD × 1 mm ID × 5 mm L, magnetized through the diameter) were strung along the length of a PTFE-coated nitinol wire (Terumo Glidewire). IONP-HN3- calculated above). DNA (1.0 g) was suspended in water and subsequently magnetically adhered to the surface of this device. Representative binding studies. DOX: To a scintillation vial was added 19 mL human serum. Drug was injected at a concentration of 1 mg/mL from a con- Flow model experiments. A closed-circuit flow model was used to measure DOX 3,32 centrated stock, to bring the total concentration to ∼0.05 mg/mL. An initial time clearance in a setting that simulates suprahepatic IVC conditions . In this model, point is taken before drug capture. DNA particles (100 ± 5 mg) were added to the the porcine blood is circulated through the polyvinyl chloride tubing at a rate serum mixture, which is constantly, mechanically stirred. 20 s before a time point is of ∼150 mL/min. The tubing size matches the average human hepatic vein mea- 18 taken, a strong, rare earth, magnet is used to isolate the particles at which point a suring 1.2 cm as described previously . Testing was performed with 200 mL 100 μL aliquot is taken and placed in a 96-well microplate. The solutions are then porcine blood and samples were obtained from the tubing downstream from measured by way of fluorescence on a microplate reader. the device. Cisplatin: Phosphate-buffered saline solution (19 mL) was added to a scintillation vial. Cisplatin solution (1 mL, 1 mg/mL solution) was then injected, In vitro experiments. H9C2 rat heart myoblasts (procured from Sigma Aldrich) followed by 117 ± 5 mg of IONP-HN3-DNA, and the mixture was mechanically were cultured in well plates (four replicates per condition) for 48 h after which stirred over the course of an hour. At predetermined time points the magnetic cells were imaged with a light microscope and counted. In the control experiment, materials were temporarily isolated using an external magnet so that 100 μL cells were cultured in RPMI medium with 10% fetal bovine serum. In the DOX aliquots could be taken, which were diluted 200× in 2% nitric acid solution and experiment, cells were cultured with 0.05 mg/mL DOX added to the medium. In subsequently analyzed by ICP-MS to determine the concentration of platinum the cell rescue experiments, 0.19 g of either Dowex ion-exchange resin or IONP- remaining in solution. HN3-DNA was added to the culture media prior to introduction of DOX (final EPI: Human serum (19 mL) was added to a scintillation vial. EPI solution in DOX concentration: 0.05 mg/mL). water (1 mL, 1 mg/mL solution) was then added. The particles (100 ± 5 mg IONP- HN3-DNA) were then added and the solution was mechanically stirred over the course of 25 min. At predetermined time points, the magnetic materials were In vivo porcine experiments. In vivo device testing was performed in farm swine temporarily isolated using an external magnet and 100 μL aliquots were taken, (n = 1, 45–50 kg), which was under humane care. Experimentation was under which were subsequently diluted 100× in water and analyzed by fluorescence compliance with UCSF IACUC protocols. The animal was monitored with blood on a microplate reader in order to characterize the amount of EPI remaining in pressure, pulse oximetry, heart rate, and electrocardiogram while under general solution. anesthesia with isoflurane. Using fluoroscopic guidance, an 18Fr sheath was placed into the left external iliac vein for introduction of the device. A pre-device sampling catheter was introduced through the right external iliac vein with the tip termi- nating in the left common iliac vein near the bifurcation. An additional catheter Data availability. All data supporting the findings of this study are available within was introduced via the right internal jugular vein with the tip distal to the device in the article and its Supplementary Information. All other data are available from the the IVC (post-device). The mid-device catheter and peripheral catheters were corresponding author upon reasonable request. 6 NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 ARTICLE Received: 12 May 2017 Accepted: 20 June 2018 28. Siddik, Z. H. Cisplatin: mode of cytotoxic action and molecular basis of resistance. Oncogene 22, 7265–7279 (2003). 29. Polavarapu, A., Stillabower, J. A., Stubblefield, S. G. W., Taylor, W. M. & Baik, M. H. The mechanism of guanine alkylation by nitrogen mustards: a computational study. J. Org. Chem. 77, 5914–5921 (2012). 30. Agudelo, D. et al. Probing the binding sites of antibiotic drugs doxorubicin and N-(trifluoroacetyl) doxorubicin with human and bovine serum albumins. References PLoS ONE 7,1–13 (2012). 1. Cleeland, C. S. et al. Reducing the toxicity of cancer therapy: recognizing 31. Agudelo, D., Bourassa, P., Bérubé, G. & Tajmir-Riahi, H. A. Intercalation of needs, taking action. Nat. Rev. Clin. Oncol. 9,1–8 (2012). antitumor drug doxorubicin and its analogue by DNA duplex: structural 2. Tatonetti, N. P., Ye, P. P., Daneshjou, R. & Altman, R. B. Data-driven features and biological implications. Int. J. Biol. Macromol. 66, 144–150 prediction of drug effects and interactions. Sci. Transl. Med. 4, 125ra31 (2012). (2014). 3. Patel, A. S. et al. Development and validation of endovascular chemotherapy 32. Patel, A. et al. O-016 development and validation of an endovascular filter device for removing high-dose doxorubicin: preclinical study. J. Med. chemotherapy filter device for removing high-dose doxorubicin from the Device. 8, 0410081–0410088 (2014). blood: <em>In Vivo</em> Porcine Study. J. Neurointerv. Surg. 6,A9LP–A9 4. Chen, X. C. et al. Block copolymer membranes for efficient capture of a L9 (2014). chemotherapy drug. ACS Macro Lett. 5, 936–941 (2016). https://doi.org/ 10.1021/acsmacrolett.6b00459 5. Altekruse, S. F., McGlynn, K. A. & Reichman, M. E. Hepatocellular carcinoma Acknowledgements incidence, mortality, and survival trends in the United States from 1975 to The authors gratefully acknowledge the financial support from the NIH (R01CA194533, 2005. J. Clin. Oncol. 27, 1485–1491 (2009). Hetts; 5T32EB001631-13, Aboian). Confocal imaging was performed in the Biological 6. Belghiti, J. & Kianmanesh, R. Surgical treatment of hepatocellular carcinoma. Imaging Facility, with the support of the Caltech Beckman Institute and the Arnold and HPB (Oxford). 7,42–49 (2005). Mabel Beckman Foundation. The authors additionally acknowledge support from the 7. Marelli, L. et al. Transarterial therapy for hepatocellular carcinoma: Which Beckman Institute of the California Institute of Technology to the Molecular Materials technique is more effective? A systematic review of cohort and randomized Research Center. The authors wish to thank Dr. Andres Collazo for assistance with studies. Cardiovasc. Interv. Radiol. 30,6–25 (2007). imaging, Mr. Daryl Yee for assistance with X-ray photoelectron spectroscopy, and 8. Cheung-Ong, K., Giaever, G. & Nislow, C. DNA-damaging agents in cancer Wesley Kuo for performing flow model experiments and for assisting with fluoroscopy chemotherapy: serendipity and chemical biology. Chem. Biol. 20, 648–659 (2013). image acquisition during in vivo porcine experiments. We would like to thank UCSF 9. Minotti, G., Menna, P., Salvatorelli, E., Cairo, G. & Gianni, L. Anthracyclines: Interventional Radiology laboratory members for assisting with in vivo porcine experi- molecular advances and pharmacologic developments in antitumor activity ments including Carol Stillson, Dr. Maythem Saeed, Anqi Liang, Joshua Fisher, Jay Yu, and cardiotoxicity. Pharmacol. Rev. 56, 185–229 (2004). and Dr. Caroline Jordan. We would also like to thank Sravani Kondapavulur for expert 10. Ryberg, M. et al. Epirubicin cardiotoxicity: an analysis of 469 patients with advice on the assembly of the magnetic device. metastatic breast cancer. J. Clin. Oncol. 16, 3502–3508 (1998). 11. Miller, R. P., Tadagavadi, R. K., Ramesh, G. & Reeves, W. B. Mechanisms of cisplatin nephrotoxicity. Toxins (Basel) 2, 2490–2518 (2010). 12. Wang, D. & Lippard, S. J. Cellular processing of platinum anticancer drugs. Author contributions Nat. Rev. Drug Deliv. 4, 307–320 (2005). C.M.B. and M.D.S. designed materials and devices, synthesized materials, and assisted 13. Rahman, A. M., Yusuf, S. W. & Ewer, M. S. Anthracycline-induced with in vitro experiments. M.S.A performed cell culture experiments, and performed cardiotoxicity and the cardiac-sparing effect of liposomal formulation. Int. J. in vitro and in vivo experiments with the magnetic device. T.M., M.W.W., and S.W.H. Nanomed. 2, 567–583 (2007). supervised and performed in vivo experiments. R.H.G. and S.W.H. supervised the pro- 14. Buzdar, A. U., Marcus, C., Smith, T. L. & Blumenschein, G. R. Early and ject. All authors discussed the results. delayed clinical cardiotoxicity of doxorubicin. Cancer 55, 2761–2765 (1985). 15. Yeh, E. T. H. & Bickford, C. L. Cardiovascular complications of cancer therapy. J. Am. Coll. Cardiol. 53, 2231–2247 (2009). Additional information 16. Tsochatzis, E. A., Germani, G. & Burroughs, A. K. Transarterial Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- chemoembolization, transarterial chemotherapy, and intra-arterial 018-05305-2. chemotherapy for hepatocellular carcinoma treatment. Semin. Oncol. 37, 89–93 (2010). Competing interests: S.W.H. and M.W.W. have a royalty agreement through the 17. Hwu, W. J. et al. A clinical-pharmacological evaluation of percutaneous University of California and currently licensed to Penumbra, Inc., should a medical isolated hepatic infusion of doxorubicin in patients with unresectable liver device resulting from the underlying endovascular filtration technology be developed and tumors. Oncol. Res. 11, 529–537 (1999). marketed. A patent application (WO2018048829A1) has been filed by the California 18. Aboian, M. S. et al. In vitro clearance of doxorubicin with a DNA-based Institute of Technology that covers the materials described in this paper. The remaining filtration device designed for intravascular use with intra-arterial authors declare no competing interests. chemotherapy. Biomed. Microdevices 18, 98 (2016). 19. Hurst, S. J. et al. Synthetically programmable DNA binding domains in Reprints and permission information is available online at http://npg.nature.com/ aggregates of DNA-functionalized gold nanoparticles. Small 5, 2156–2161 reprintsandpermissions/ (2009). 20. Cohen, G., Deutsch, J., Fineberg, J. & Levine, A. Covalent attachment of Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in hybridizable oligonucleotides to glass supports. Nucleic Acids Res. 25, 911–912 published maps and institutional affiliations. (1997). 21. Beier, M. & Hoheisel, J. D. Versatile derivatisation of solid support media for covalent bonding on DNA-microchips. Nucleic Acids Res. 27, 1970–1977 (1999). 22. Kumar, A., Larsson, O., Parodi, D. & Liang, Z. Silanized nucleic acids: a Open Access This article is licensed under a Creative Commons general platform for DNA immobilization. Nucleic Acids Res. 28, E71 (2000). Attribution 4.0 International License, which permits use, sharing, 23. Yao, G. et al. Clicking DNA to gold nanoparticles: poly-adenine-mediated adaptation, distribution and reproduction in any medium or format, as long as you give formation of monovalent DNA-gold nanoparticle conjugates with nearly appropriate credit to the original author(s) and the source, provide a link to the Creative quantitative yield. NPG Asia Mater. 7, e159 (2015). Commons license, and indicate if changes were made. The images or other third party 24. Mirkin, C. A., Letsinger, R. L., Mucic, R. C. & Storhoff, J. J. A DNA-based material in this article are included in the article’s Creative Commons license, unless method for rationally assembling nanoparticles into macroscopic materials. indicated otherwise in a credit line to the material. If material is not included in the Nature 382, 607–609 (1996). article’s Creative Commons license and your intended use is not permitted by statutory 25. Macfarlane, R. J. et al. Nanoparticle superlattice engineering with DNA. regulation or exceeds the permitted use, you will need to obtain permission directly from Science 334, 204–208 (2011). 26. Lipshutz, R. J., Fodor, S. P., Gingeras, T. R. & Lockhart, D. J. High density the copyright holder. To view a copy of this license, visit http://creativecommons.org/ synthetic oligonucleotide arrays. Nat. Genet. 21,20–24 (1999). licenses/by/4.0/. 27. Smolensky, E. D., Peterson, K. L., Weitz, E. A., Lewandowski, C. & Pierre, C. Magnetoluminescent light switches—dual modality in DNA detection eric. © The Author(s) 2018 J. Am. Chem. Soc. 135, 8966–8972 (2013). NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 7 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nature Communications Springer Journals

Drug capture materials based on genomic DNA-functionalized magnetic nanoparticles

Loading next page...
 
/lp/springer-journals/drug-capture-materials-based-on-genomic-dna-functionalized-magnetic-sRZJonrAAy

References (32)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
eISSN
2041-1723
DOI
10.1038/s41467-018-05305-2
Publisher site
See Article on Publisher Site

Abstract

ARTICLE DOI: 10.1038/s41467-018-05305-2 OPEN Drug capture materials based on genomic DNA- functionalized magnetic nanoparticles 1 1 2 2 2 Carl M. Blumenfeld , Michael D. Schulz , Mariam S. Aboian , Mark W. Wilson , Terilynn Moore , 2 1 Steven W. Hetts & Robert H. Grubbs Chemotherapy agents are notorious for producing severe side-effects. One approach to mitigating this off-target damage is to deliver the chemotherapy directly to a tumor via transarterial infusion, or similar procedures, and then sequestering any chemotherapeutic in the veins draining the target organ before it enters the systemic circulation. Materials capable of such drug capture are yet to be fully realized. Here, we report the covalent attachment of genomic DNA to iron-oxide nanoparticles. With these magnetic materials, we captured three common chemotherapy agents—doxorubicin, cisplatin, and epirubicin—from biological solutions. We achieved 98% capture of doxorubicin from human serum in 10 min. We further demonstrate that DNA-coated particles can rescue cultured cardiac myoblasts from lethal levels of doxorubicin. Finally, the in vivo efficacy of these materials was demonstrated in a porcine model. The efficacy of these materials demonstrates the viability of genomic DNA- coated materials as substrates for drug capture applications. Arnold and Mabel Beckman Laboratories for Chemical Synthesis, Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA 91125, USA. Interventional Radiology Research Laboratory, Department of Radiology and Biomedical Imaging, University of California-San Francisco, San Francisco, CA 94143, USA. These authors contributed equally: Carl M. Blumenfeld, Michael D. Schulz. Correspondence and requests for materials should be addressed to R.H.G. (email: rhg@caltech.edu) NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 1 1234567890():,; ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 he systemic toxicity of chemotherapy is a widely recog- suppression. Consequently, a balance must be struck in order to nized problem in oncology. Off-target damage often per- maximize drug dose, leading to better tumor suppression, while Tsists indefinitely, adversely affects patient survival, and simultaneously avoiding catastrophic off-target toxicity. Although 1,2 restricts dose and treatment options . Direct administration of limiting a patient’s lifetime cumulative dose is the most effective chemotherapy agents to the tumor via transarterial che- way to avoid cardiotoxicity, this approach necessarily limits anti- moembolization (TACE), or similar procedures, followed by cancer efficacy . sequestration of any chemotherapeutic that enters systemic cir- The unwanted systemic toxicity of chemotherapy agents has culation would mitigate this damage if materials capable of such inspired a number of more targeted approaches. One such 3,4 drug capture were fully realized . approach is TACE, during which liver blood flow is occluded in Hepatocellular carcinoma (HCC) is the third leading cause of conjunction with administration of high dose chemotherapy 5 3,16 cancer-related deaths worldwide . Liver transplantation is the directly to the tumor . Both during TACE and after liver blood most definitive approach for treatment; however, less than 30% of flow is restored, however, up to 50% of residual chemother- 6 17 HCC patients are eligible . Direct delivery of drug to a tumor via apeutics enter systemic circulation and cause off-target toxicity . intraarterial chemotherapy (IAC) and its variant, TACE, is often Efforts have been made toward reducing non-targeted toxicity used as a bridge to transplantation, shrinking HCC or at least during TACE. In 2014 Patel and coworkers proposed che- controlling its growth through recurrent treatments until curative motherapy filtration devices (“ChemoFilters”) that employed transplant is possible. In cases where surgery is untenable, che- sulfonated ion-exchange resins with affinity for DOX. Such a motherapy is often the only recourse. Targeted therapy, however, device could be deployed via catheter in the hepatic vein, does not completely eliminate side-effects. “downstream” from the site of chemotherapy administration, Three of the most common drugs used to treat HCC are where it can intercept any residual chemotherapy agents before doxorubicin (DOX), epirubicin (EPI), and cisplatin (Fig. 1c), all they reach the heart and enter systemic circulation. They of which act on DNA . DOX and EPI function by intercalating demonstrated a 52% reduction in DOX concentration from between DNA base pairs, while cisplatin is a DNA crosslinker porcine serum over 10 min, and showed that such a device could 8,9 3 that functions by binding to guanine . A major problem for be successfully deployed during a simulated TACE procedure .In these anticancer compounds is toxicity in non-targeted tissues. 2016, the ChemoFilter approach inspired the development of DOX and EPI toxicity can result in cardiomyopathy and con- more elaborate block copolymer membranes for DOX capture, 8–10 gestive heart failure . Similarly, cisplatin elicits side-effects which achieved up to 90% removal of DOX in 31 min from 11,12 4 including extensive nephrotoxicity and neurotoxicity .To phosphate buffered saline (PBS) . reduce the likelihood of cardiac toxicity, cumulative dosage of A ChemoFilter device is intended to be placed downstream of DOX is generally limited by clinicians to 400–450 mg/m , though blood outflow from the tumor that is being treated with intra- lower cumulative dosages (300 mg/m ) are known to increase the arterial chemotherapeutics. In the case of HCC, which is located 13,14 risk of congestive heart failure . Still, a single standard dose of in the liver, the ChemoFilter would be placed into the suprahe- DOX (50–75 mg) can result in severe side-effects, yet higher patic inferior vena cava (IVC) immediately prior to administra- dosages of DOX are known to be associated with greater tumor tion of IAC into the hepatic artery and then removed shortly after IONP-Pt-DNA approach CI CI H Pt N NH HN NH 2 2 Si O SiOMe Si O H O K PtCl 2 4 Fe O Fe O 3 4 3 4 H N CI Fe O 3 4 Fe O 3 4 NH N H N 2 N CI Si O Si O O N(CH CH CI) SiOMe 2 2 3 Fe O 3 4 Fe O 3 4 IONP-HN3-DNA approach bc O OH O O OH O OH OH OH DOX OH Fe O 3 4 O O OH O NH O O OH O NH OH OH Cl NH OO OH Cisplatin Pt OH OH Cl NH O O OH O NH OH OO OH OH EPI >95% DOX OH DOX removal Serum Solution O O OH O NH OH Fig. 1 Experimental approach. a Two synthetic approaches for covalently attaching genomic DNA onto iron-oxide nanoparticle (IONP) surfaces. b Drug capture concept. c Three common chemotherapy agents used in this study 2 NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 ARTICLE 3,18 drug capture has been achieved . Here we report the devel- hydroxylated surface of Fe O was silylated with N-(2-ami- 3 4 opment of a drug-capture material based on genomic DNA and noethyl)-3-aminopropyltrimethoxysilane exposing a chelating iron oxide particles. In vitro studies confirm that these materials diamine functionality. This sample was treated with an excess can rescue cells from the toxic effects of DOX. A ChemoFilter of potassium tetrachloroplatinate to create an analog of cisplatin device is constructed from this material that rapidly removes by which DNA could be anchored to the surface. Cisplatin’s chemotherapy agents from relevant biological solutions, including cytotoxicity is thought to stem from its coordination with human serum and porcine blood, and in vivo studies confirm that nucleophilic N7-sites of purine bases, resulting in crosslinks . DOX can be removed from the bloodstream by an intraarterial We hoped to accomplish DNA crosslinking to the surface device constructed from these iron oxide/genomic DNA through this mechanism. The sample was then exposed to DNA materials. to produce IONP-Pt-DNA. The second approach was modeled on nitrogen mustard chemotherapy agents. IONP-HN3-DNA samples were prepared Results and discussion first by functionalizing Fe O with 4-aminobutyltriethoxysilane to Material design and synthesis. Inspired by the ChemoFilter 3 4 install free amines on the surface. This particle was then treated concept, we designed and synthesized DNA-functionalized with excess tris(2-chloroethyl)amine hydrochloride (HN3·HCl) to materials based on magnetite (Fe O ) nanoparticles, capable of 3 4 create a scaffold for DNA functionalization. HN3·HCl, the rapidly capturing chemotherapy agents. Central to our approach hydrochloride salt of the nitrogen mustard HN3, undergoes is the direct covalent attachment of genomic DNA. Functiona- aziridinium formation when deprotonated, and is attacked readily lizing surfaces with DNA has historically involved tagging either by the nucleophilic moieties of DNA . The functionalized the backbone or bases of synthetic DNA with an appropriate particle was exposed to DNA resulting in IONP-HN3-DNA. Both moiety, or attaching the DNA via a reactive end-group. These materials were characterized by scanning electron microscopy, approaches are highly useful and enable complete control of the electron dispersive scattering (EDS), elemental analysis, and DNA sequence used resulting in the development of numerous 19–26 infrared spectroscopy (see Supplementary Figs. 2& 3 and 11–24, interesting materials ; however, they are limited by the rela- and Supplementary Table 1). Microscopy images of the particles tively high cost of synthetic DNA. The synthesis of large amounts in solution revealed significant aggregation resulting in an average of such materials would be prohibitively expensive for most particle diameter of 4.2 μm with several larger (>10 μm) applications. aggregates. Elemental analysis indicated that these aggregates Functionalization with genomic DNA is an alternative were 18% DNA by mass in the case of IONP-HN3-DNA and approach that may be appropriate for certain applications; 14.7% DNA by mass in the case of IONP-Pt-DNA. however, this approach is relatively unexplored. Pierre and coworkers recently synthesized magnetic nanoparticles with surface-bound intercalating groups, and showed that such In vitro testing in simple solutions, serum, and blood. In order materials can bind to genomic DNA . To our knowledge, to evaluate the efficacy of our materials at scavenging che- however, no one has reported the covalent attachment of genomic motherapy agents from solution we studied DOX-binding in PBS DNA to a surface. Here, we report two methods of attaching and human serum at 37 °C to approximate the biological envir- genomic DNA to nanoparticles, both on multi-gram scale onment in which these materials would have to operate (Fig. 1b). (Fig. 1a). We show that the resulting materials are capable of We found that IONP-HN3-DNA was able to capture 93% of removing DNA-targeting chemotherapy agents from solution DOX, on average, from a 0.05 mg/mL solution of human serum both rapidly and in the presence of potential biological in 25 min, while IONP-Pt-DNA averaged 79% (Fig. 2a). In both intereferents (e.g., serum proteins and other blood components). cases, the kinetics were extremely rapid, with about 50% of DOX DNA-alkylating agents are a common motif in chemotherapy. capture occurring within 1 min in the case of IONP-Pt-DNA and By forming covalent crosslinks between DNA strands, these drugs over 65% DOX capture occurring within 1 min for IONP-HN3- prevent the DNA from being accurately duplicated, ultimately DNA. Based on these results, we carried out all further tests with leading to apoptosis. To attach genomic DNA to magnetic IONP-HN3-DNA. nanoparticles, we used an approach analogous to DNA-alkylat- Interestingly, both materials were highly effective, despite the ing/crosslinking drugs (Fig. 1a). The first approach was inspired known binding of DOX with serum albumin. It is known that 30,31 by cisplatin. To synthesize IONP-Pt-DNA samples, the DNA intercalation is the kinetically more favorable process , a b 0.05 IONP-HN3-DNA Control 0.9 IONP-Pt-DNA IONP-HN3-DNA 0.8 0.04 0.7 0.6 0.03 0.5 0.4 0.02 0.3 0.2 0.01 0.1 0 0 0 5 10 15 20 25 30 02468 10 12 Time (min) Time (min) Fig. 2 DOX capture in human serum and porcine blood. a Decrease in DOX concentration in human serum, determined by fluorescence, as a result of DOX capture by IONP-HN3-DNA and IONP-Pt-DNA; 100 ± 5 mg particle in 20 mL (0.05 mg/mL), 1 mg total DOX, 37 °C; error bars= 1 standard deviation (n = 3). b Decrease in DOX plasma concentration as a result of DOX capture by IONP-HN3-DNA from porcine whole blood; 100 ± 5 mg IONP-HN3-DNA in 20 mL (0.05 mg/mL), 1 mg total DOX, 37 °C; error bars= 1 standard deviation (n = 3) NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 3 DOX concentration (mg/mL) Normalized DOX concentration ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 and we believe that this kinetic advantage enabled our material to material added per mg DOX, resulting in ~90% DOX capture in capture DOX from serum solution, despite the thermodynamics 10 min. Further DOX capture appears less favorable after this being in favor of serum binding overall. We posit that over longer point. We believe this plateau is the result of competition with timescales, serum binding would be the dominant process; serum binding, which makes that portion of DOX unavailable for however, since TACE is a relatively short procedure (<1 h), we capture by our particles, as well as the typical kinetic effects of believe that kinetic factors will dominate in the performance of diminishing concentration. The absorption of DOX onto the any material or device. particles was further verified by performing confocal fluorescence Drug capture was also evaluated in porcine whole blood, by microscopy (Fig. 3c and d). This technique allowed visualization measuring DOX plasma concentration over time. We observed of the fluorescence of DOX bound to the surface of the particles. some DOX removal due to binding to the non-plasma blood Our approach is general for all DNA-targeting chemotherapy components, which we cannot deconvolute from capture by our agents. To demonstrate this fact, we performed further experi- materials. Nevertheless, there is rapid reduction of DOX ments on two additional common DNA-targeting chemother- concentration in the blood plasma within 1 min after exposure apeutics, cisplatin and EPI. We performed an initial cisplatin- to our material, reaching a 92% reduction in DOX plasma binding experiment in PBS solution with IONP-HN3-DNA and concentration over 10 min, in stark contrast to the control monitored the decrease of cisplatin concentration by inductively experiment (Fig. 2b). This experiment conclusively demonstrates coupled plasma-mass spectrometry (ICP-MS). Approximately that our materials are capable of capturing DOX from whole 20% of the cisplatin was captured from solution over 30 min, with blood. little improvement over longer time periods (see Supplementary To better understand the DOX-capture capacity of IONP- Fig. 4). We confirmed the presence of captured cisplatin on the HN3-DNA, we performed a series of experiments in which surface of the particles by x-ray photoelectron spectroscopy. We nanoparticle loading was systematically varied (Fig. 3a, b, further believe these relatively low levels of drug capture are due to data in the Supplementary Information). These experiments cisplatin not being in the aquo state, which happens intracellu- revealed a roughly linear trend in DOX-capture as a function of larly and is necessary in order to bind to DNA. Because the the amount of particle added, up to a plateau around 100 mg particles were not highly effective for cisplatin capture in PBS, we did not perform further experiments in more complex media such as serum or blood. 0.05 Along with DOX and cisplatin, EPI is among the most commonly used chemotherapeutic agents for treating HCC. We 0.04 evaluated the efficacy of our materials for capturing EPI using a set of experiments analogous to those we used with DOX (see 0.03 Supplementary Fig. 5). Our particles were highly effective at 0.02 sequestering EPI from serum, with 68% captured after 25 min. The sequestered amount would lead to a reduction in unwanted 0.01 side-effects if achieved in vivo. 0 2468 10 12 In vitro evaluation of biological efficacy. The ability of IONP- Time (min) HN3-DNA to detoxify DOX was tested in vitro in an H9C2 rat 10 mg 20 mg 30 mg 40 mg 50 mg 100 mg IONP-HN3-DNA heart myoblast cell culture assay (Fig. 4). These experiments demonstrated that DNA-coated particles could rescue cultured cardiac myoblasts from lethal levels of DOX more effectively than the ion exchange resin Dowex, which itself had been previously shown to reduce levels of DOX in vivo . DOX 10 mg 20 mg 40 mg 30 mg 100 mg Serum solution In vivo evaluation of drug capture. A device (Fig. 5a) consisting of IONP-HN3-DNA magnetically adhered to the surface of cylindrical rare-earth magnets strung along a PTFE-coated nitinol wire was evaluated using a closed loop flow model (see Sup- c d plementary Fig. 25) and subsequently tested in vivo using a porcine model. The device was inserted into the IVC and DOX was injected over 10 min at a rate of 2.5 mL/min into the left common iliac vein proximal to the device (Fig. 5b). As the drug flowed through the IVC, it made contact with the IONP-HN3- DNA adherent to the surface of the device and was captured. Blood aliquots were taken proximal to (upstream), adjacent to the midpoint of, and distal to (downstream from) the device using separate catheters. Peak DOX concentration was observed at 3 min, since the blood at the injection site had recirculated and live injection was still underway. At 3 Fig. 3 DOX capture with IONP-HN3-DNA particles. a DOX capture as a min (peak concentration), a 60% reduction in serum DOX function of the amount of IONP-HN3-DNA from a DOX serum solution concentration was observed half-way across the device, while (0.5 mg total DOX, 0.05 mg/mL); average of three experiments, data set a total reduction of 82% was observed at the end of the devi- with error bars in Supplementary Fig. 6. b Resulting serum solutions from ce (Fig. 5c, d). experiments summarized in a. c Brightfield image of IONP-HN3-DNA In conclusion, we have demonstrated two viable synthetic aggregates bound to DOX; scale bar = 50 μm. d Fluorescence from DOX pathways to genomic DNA-functionalized magnetic particles, bound to IONP-HN3-DNA; scale bar = 50 μm both on multi-gram scale. Moreover, these methodologies for 4 NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications Doxorubicin concentration (mg/mL) NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 ARTICLE a b DOX + Dowex DOX c DOX + IONP-DNA d Control DOX DOX + DOX + Dowex IONP-DNA Fig. 4 IONP-HN3-DNA rescue cultured cells from DOX toxicity (scale bar = 5 μm). a Rat heart myoblasts H9C2 cultured with DOX (0.05 mg/mL)— myoblasts are killed by adding DOX to culture (DOX, d). b Myoblasts treated with DOX (0.05 mg/mL) and 0.19 g Dowex ion-exchange resin beads— myoblasts are partially rescued ((DOX+ Dowex, d). c Myoblasts treated with DOX (0.05 mg/mL) and 0.19 g IONP-HN3-DNA—myoblasts are completely rescued (DOX+ IONP-DNA, d). d Cell counts for each experimental condition; error bars= 1 standard deviation (n = 3) a b 0.03 Pre 0.025 Mid 0.02 Post Peripheral 0.015 Pre-device Mid-device Post-device 0.01 0.005 0 5 10 15 20 25 30 Time (min) Fig. 5 In vivo results. a Device containing 25 magnets (1 cm × 0.5 cm) with IONP-HN3-DNA coating (above), and the same device after the in vivo experiment (below), demonstrating minimal loss of particles after removal of the device. b Fluoroscopy images during in vivo porcine experiment demonstrating the inferior vena cava with opacified right renal veins (scale bar = 5 mm). The device was placed within the inferior vena cava. The sampling catheters were placed immediately proximal to the device, prior to the renal vein, and distal to the device. c DOX concentration measurements from pre- device, mid-device, post-device, and peripheral locations. d Plasma solutions from the experiment described in c: Left 3, pre-filter samples taken at 1, 3, and 10 min; Center 3, mid-filter samples taken at 1, 3, and 10 min; Right 3, post-filter samples taken at 1, 3, and 10 min NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 5 Concentration (mg/mL) Cell number ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 DNA surface functionalization are not limited to magnetic metal introduced through the left internal jugular vein. Prior to the start of the experi- ments, patency of the venous system was demonstrated using contrast injection oxides, but may also be exploited for other substrates. The (Omnipaque). DOX was injected over 10 min at a rate of 2.5 mL/min into the left synthesized materials captured three commonly used chemother- common iliac vein proximal to the magnetic device. The pre-device DOX con- apy agents from relevant biological solutions (PBS, human serum, centration was measured by sampling with a 5 Fr catheter downstream of the DOX or porcine whole blood), at therapeutically relevant concentra- infusion. Blood aliquots were taken proximal to, adjacent to the midpoint of, and distal to the device using separate catheters. To clear the sampling catheters, 2 mL tions and timescales. A proof of concept device was developed, of blood was drawn immediately prior to taking the aliquot (3 mL). The blood which demonstrated efficient capture of DOX in vivo. Similar samples were placed on ice until they were centrifuged to isolate the plasma devices could be readily developed that would potentially reduce fraction for analysis. A control experiment was also performed using the same the off-target toxicity and damaging side-effects associated with procedures but with no device inserted. the use of DOX, cisplatin, and EPI during TACE or similar procedures. Ultimately, we believe our approach is general for all Particle synthesis. IONP-Pt: 3.31 g Fe O was dried in vacuo at 120 °C. Upon DNA-targeting chemotherapy drugs, and while further 3 4 cooling, the sealed material was introduced into an inert atmosphere nitrogen development is needed, we hope that this work will provide a glovebox. To the Fe O was added 23 mL anhydrous toluene along with 4 mL N- 3 4 foundation for future work on DNA-based materials and drug (2-aminoethyl)-3-aminopropyltrimethoxysilane. The reaction was mechanically capture approaches both for oncologic and non-oncologic stirred on the bench at 110 °C for 2 h and subsequently dried in vacuo at 110 °C for applications. 20 h. The reaction mixture along with 1.0 g K PtCl , was stirred at 70 °C for 21 min 2 4 and then washed three times with water. Following this, the mixture was diluted to a total volume of 450 mL with 18.2 MΩ water was treated with 1.3 g KCl and an Methods additional 10 mL water. Instrumentation. Fluorescence measurements were made using a 96-well plate on IONP-Pt-DNA: IONP-Pt materials along with 5.1 g deoxyribonucleic acid from a Molecular Devices FlexStation 3 Multi-mode microplate reader. Scanning elec- herring sperm were mechanically stirred in 450 mL 18.2 MΩ water at 37 °C for tron micrographs (SEM), as well as EDS measurements were made on a Zeiss 20 h. To ensure covalent attachment as opposed to being physically adsorbed, the 1550VP field emission SEM equipped with an Oxford EDS module. ICP-MS was particles were isolated from the reaction mixture, washed three times under carried out on an HP 4500 ICP-MS equipped with a Cetac ASX-500 autosampler. vigorous mechanical stirring with 18.2 MΩ water (400 mL) in order to remove Infrared measurements were made on a Nicolet iS50 Fourier transform infrared unbound DNA, frozen, and lyophilized to afford 3.08 g IONP-Pt-DNA (79% yield spectrometer equipped with a DuraScope ATR unit. C, H, N analyses were carried based on elemental analysis of DNA content). out using a PerkinElmer 2400 Series II CHN Elemental Analyzer. Fluorescence IONP-NH2: 4.2 g Of Fe O was dried in vacuo at 120 °C. The Fe O was 3 4 3 4 microscopy was performed on an inverted laser scanning confocal Zeiss LSM 710 allowed to cool to room temperature under vacuum. To the Fe O was added to 3 4 microscope equipped with an argon laser and photomultiplier tube detector, and 25 mL toluene (freshly dried over magnesium sulfate) and 3.2 mL 4- particle size was determined by image analysis of at least 100 particles measured on aminobutyltriethoxysilane. The reaction was sealed and stirred mechanically for 2 h their widest dimension. at 120 °C. The reaction was removed from heat and the particles were isolated from the toluene solution. The reaction mixture was washed once with toluene and subsequently dried in vacuo at 120 °C for 1 h and 45 min. 4.02 g of IONP-HN3 was General procedures. Unless otherwise stated reactions were carried out on the isolated. bench. Fe O (40 nm APS, 99%) was purchased from Nanostructured & Amor- 3 4 IONP-HN3-DNA: 3.4750 g IONP-HN3 was added to a vial along with 1.02 g phous Materials, Inc. Silane reagents were purchased from Gelest, Inc. Genomic HN3·HCl and dimethylformamide (30 mL). The reaction was stirred mechanically DNA (isolated from Herring sperm), human serum (OptiClear), H9C2 rat heart for 1 h at room temperature at which point, the particles were isolated from the myoblasts, and cisplatin were purchased from Sigma Aldrich. DOX was purchased dimethylformamide. The particles were then washed three times with from LC Labs and EPI was purchased from Biotang Inc. Potassium tetra- dimethylformamide. The isolated particle as well as 3.35 g deoxyribonucleic acid chloroplatinate was purchased from Pressure Chemicals. All reagents not otherwise from herring sperm were transferred into a flask along with 400 mL 18.2 MΩ water. mentioned were purchased from Sigma Aldrich, and were used without further The reaction was mechanically stirred at 38 °C for 17 h and 45 min. To ensure purification. covalent attachment and to remove any unbound DNA, the particles were then washed thoroughly under vigorous mechanical stirring three times with 18.2 MΩ Device construction. Twenty-five cylindrical rare-earth magnets (N52 grade, water (400 mL) and magnetic separation. The particles were then frozen in liquid nitrogen and lyophilized to afford 3.79 g of IONP-HN3-DNA (89% yield as 5 mm OD × 1 mm ID × 5 mm L, magnetized through the diameter) were strung along the length of a PTFE-coated nitinol wire (Terumo Glidewire). IONP-HN3- calculated above). DNA (1.0 g) was suspended in water and subsequently magnetically adhered to the surface of this device. Representative binding studies. DOX: To a scintillation vial was added 19 mL human serum. Drug was injected at a concentration of 1 mg/mL from a con- Flow model experiments. A closed-circuit flow model was used to measure DOX 3,32 centrated stock, to bring the total concentration to ∼0.05 mg/mL. An initial time clearance in a setting that simulates suprahepatic IVC conditions . In this model, point is taken before drug capture. DNA particles (100 ± 5 mg) were added to the the porcine blood is circulated through the polyvinyl chloride tubing at a rate serum mixture, which is constantly, mechanically stirred. 20 s before a time point is of ∼150 mL/min. The tubing size matches the average human hepatic vein mea- 18 taken, a strong, rare earth, magnet is used to isolate the particles at which point a suring 1.2 cm as described previously . Testing was performed with 200 mL 100 μL aliquot is taken and placed in a 96-well microplate. The solutions are then porcine blood and samples were obtained from the tubing downstream from measured by way of fluorescence on a microplate reader. the device. Cisplatin: Phosphate-buffered saline solution (19 mL) was added to a scintillation vial. Cisplatin solution (1 mL, 1 mg/mL solution) was then injected, In vitro experiments. H9C2 rat heart myoblasts (procured from Sigma Aldrich) followed by 117 ± 5 mg of IONP-HN3-DNA, and the mixture was mechanically were cultured in well plates (four replicates per condition) for 48 h after which stirred over the course of an hour. At predetermined time points the magnetic cells were imaged with a light microscope and counted. In the control experiment, materials were temporarily isolated using an external magnet so that 100 μL cells were cultured in RPMI medium with 10% fetal bovine serum. In the DOX aliquots could be taken, which were diluted 200× in 2% nitric acid solution and experiment, cells were cultured with 0.05 mg/mL DOX added to the medium. In subsequently analyzed by ICP-MS to determine the concentration of platinum the cell rescue experiments, 0.19 g of either Dowex ion-exchange resin or IONP- remaining in solution. HN3-DNA was added to the culture media prior to introduction of DOX (final EPI: Human serum (19 mL) was added to a scintillation vial. EPI solution in DOX concentration: 0.05 mg/mL). water (1 mL, 1 mg/mL solution) was then added. The particles (100 ± 5 mg IONP- HN3-DNA) were then added and the solution was mechanically stirred over the course of 25 min. At predetermined time points, the magnetic materials were In vivo porcine experiments. In vivo device testing was performed in farm swine temporarily isolated using an external magnet and 100 μL aliquots were taken, (n = 1, 45–50 kg), which was under humane care. Experimentation was under which were subsequently diluted 100× in water and analyzed by fluorescence compliance with UCSF IACUC protocols. The animal was monitored with blood on a microplate reader in order to characterize the amount of EPI remaining in pressure, pulse oximetry, heart rate, and electrocardiogram while under general solution. anesthesia with isoflurane. Using fluoroscopic guidance, an 18Fr sheath was placed into the left external iliac vein for introduction of the device. A pre-device sampling catheter was introduced through the right external iliac vein with the tip termi- nating in the left common iliac vein near the bifurcation. An additional catheter Data availability. All data supporting the findings of this study are available within was introduced via the right internal jugular vein with the tip distal to the device in the article and its Supplementary Information. All other data are available from the the IVC (post-device). The mid-device catheter and peripheral catheters were corresponding author upon reasonable request. 6 NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05305-2 ARTICLE Received: 12 May 2017 Accepted: 20 June 2018 28. Siddik, Z. H. Cisplatin: mode of cytotoxic action and molecular basis of resistance. Oncogene 22, 7265–7279 (2003). 29. Polavarapu, A., Stillabower, J. A., Stubblefield, S. G. W., Taylor, W. M. & Baik, M. H. The mechanism of guanine alkylation by nitrogen mustards: a computational study. J. Org. Chem. 77, 5914–5921 (2012). 30. Agudelo, D. et al. Probing the binding sites of antibiotic drugs doxorubicin and N-(trifluoroacetyl) doxorubicin with human and bovine serum albumins. References PLoS ONE 7,1–13 (2012). 1. Cleeland, C. S. et al. Reducing the toxicity of cancer therapy: recognizing 31. Agudelo, D., Bourassa, P., Bérubé, G. & Tajmir-Riahi, H. A. Intercalation of needs, taking action. Nat. Rev. Clin. Oncol. 9,1–8 (2012). antitumor drug doxorubicin and its analogue by DNA duplex: structural 2. Tatonetti, N. P., Ye, P. P., Daneshjou, R. & Altman, R. B. Data-driven features and biological implications. Int. J. Biol. Macromol. 66, 144–150 prediction of drug effects and interactions. Sci. Transl. Med. 4, 125ra31 (2012). (2014). 3. Patel, A. S. et al. Development and validation of endovascular chemotherapy 32. Patel, A. et al. O-016 development and validation of an endovascular filter device for removing high-dose doxorubicin: preclinical study. J. Med. chemotherapy filter device for removing high-dose doxorubicin from the Device. 8, 0410081–0410088 (2014). blood: <em>In Vivo</em> Porcine Study. J. Neurointerv. Surg. 6,A9LP–A9 4. Chen, X. C. et al. Block copolymer membranes for efficient capture of a L9 (2014). chemotherapy drug. ACS Macro Lett. 5, 936–941 (2016). https://doi.org/ 10.1021/acsmacrolett.6b00459 5. Altekruse, S. F., McGlynn, K. A. & Reichman, M. E. Hepatocellular carcinoma Acknowledgements incidence, mortality, and survival trends in the United States from 1975 to The authors gratefully acknowledge the financial support from the NIH (R01CA194533, 2005. J. Clin. Oncol. 27, 1485–1491 (2009). Hetts; 5T32EB001631-13, Aboian). Confocal imaging was performed in the Biological 6. Belghiti, J. & Kianmanesh, R. Surgical treatment of hepatocellular carcinoma. Imaging Facility, with the support of the Caltech Beckman Institute and the Arnold and HPB (Oxford). 7,42–49 (2005). Mabel Beckman Foundation. The authors additionally acknowledge support from the 7. Marelli, L. et al. Transarterial therapy for hepatocellular carcinoma: Which Beckman Institute of the California Institute of Technology to the Molecular Materials technique is more effective? A systematic review of cohort and randomized Research Center. The authors wish to thank Dr. Andres Collazo for assistance with studies. Cardiovasc. Interv. Radiol. 30,6–25 (2007). imaging, Mr. Daryl Yee for assistance with X-ray photoelectron spectroscopy, and 8. Cheung-Ong, K., Giaever, G. & Nislow, C. DNA-damaging agents in cancer Wesley Kuo for performing flow model experiments and for assisting with fluoroscopy chemotherapy: serendipity and chemical biology. Chem. Biol. 20, 648–659 (2013). image acquisition during in vivo porcine experiments. We would like to thank UCSF 9. Minotti, G., Menna, P., Salvatorelli, E., Cairo, G. & Gianni, L. Anthracyclines: Interventional Radiology laboratory members for assisting with in vivo porcine experi- molecular advances and pharmacologic developments in antitumor activity ments including Carol Stillson, Dr. Maythem Saeed, Anqi Liang, Joshua Fisher, Jay Yu, and cardiotoxicity. Pharmacol. Rev. 56, 185–229 (2004). and Dr. Caroline Jordan. We would also like to thank Sravani Kondapavulur for expert 10. Ryberg, M. et al. Epirubicin cardiotoxicity: an analysis of 469 patients with advice on the assembly of the magnetic device. metastatic breast cancer. J. Clin. Oncol. 16, 3502–3508 (1998). 11. Miller, R. P., Tadagavadi, R. K., Ramesh, G. & Reeves, W. B. Mechanisms of cisplatin nephrotoxicity. Toxins (Basel) 2, 2490–2518 (2010). 12. Wang, D. & Lippard, S. J. Cellular processing of platinum anticancer drugs. Author contributions Nat. Rev. Drug Deliv. 4, 307–320 (2005). C.M.B. and M.D.S. designed materials and devices, synthesized materials, and assisted 13. Rahman, A. M., Yusuf, S. W. & Ewer, M. S. Anthracycline-induced with in vitro experiments. M.S.A performed cell culture experiments, and performed cardiotoxicity and the cardiac-sparing effect of liposomal formulation. Int. J. in vitro and in vivo experiments with the magnetic device. T.M., M.W.W., and S.W.H. Nanomed. 2, 567–583 (2007). supervised and performed in vivo experiments. R.H.G. and S.W.H. supervised the pro- 14. Buzdar, A. U., Marcus, C., Smith, T. L. & Blumenschein, G. R. Early and ject. All authors discussed the results. delayed clinical cardiotoxicity of doxorubicin. Cancer 55, 2761–2765 (1985). 15. Yeh, E. T. H. & Bickford, C. L. Cardiovascular complications of cancer therapy. J. Am. Coll. Cardiol. 53, 2231–2247 (2009). Additional information 16. Tsochatzis, E. A., Germani, G. & Burroughs, A. K. Transarterial Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- chemoembolization, transarterial chemotherapy, and intra-arterial 018-05305-2. chemotherapy for hepatocellular carcinoma treatment. Semin. Oncol. 37, 89–93 (2010). Competing interests: S.W.H. and M.W.W. have a royalty agreement through the 17. Hwu, W. J. et al. A clinical-pharmacological evaluation of percutaneous University of California and currently licensed to Penumbra, Inc., should a medical isolated hepatic infusion of doxorubicin in patients with unresectable liver device resulting from the underlying endovascular filtration technology be developed and tumors. Oncol. Res. 11, 529–537 (1999). marketed. A patent application (WO2018048829A1) has been filed by the California 18. Aboian, M. S. et al. In vitro clearance of doxorubicin with a DNA-based Institute of Technology that covers the materials described in this paper. The remaining filtration device designed for intravascular use with intra-arterial authors declare no competing interests. chemotherapy. Biomed. Microdevices 18, 98 (2016). 19. Hurst, S. J. et al. Synthetically programmable DNA binding domains in Reprints and permission information is available online at http://npg.nature.com/ aggregates of DNA-functionalized gold nanoparticles. Small 5, 2156–2161 reprintsandpermissions/ (2009). 20. Cohen, G., Deutsch, J., Fineberg, J. & Levine, A. Covalent attachment of Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in hybridizable oligonucleotides to glass supports. Nucleic Acids Res. 25, 911–912 published maps and institutional affiliations. (1997). 21. Beier, M. & Hoheisel, J. D. Versatile derivatisation of solid support media for covalent bonding on DNA-microchips. Nucleic Acids Res. 27, 1970–1977 (1999). 22. Kumar, A., Larsson, O., Parodi, D. & Liang, Z. Silanized nucleic acids: a Open Access This article is licensed under a Creative Commons general platform for DNA immobilization. Nucleic Acids Res. 28, E71 (2000). Attribution 4.0 International License, which permits use, sharing, 23. Yao, G. et al. Clicking DNA to gold nanoparticles: poly-adenine-mediated adaptation, distribution and reproduction in any medium or format, as long as you give formation of monovalent DNA-gold nanoparticle conjugates with nearly appropriate credit to the original author(s) and the source, provide a link to the Creative quantitative yield. NPG Asia Mater. 7, e159 (2015). Commons license, and indicate if changes were made. The images or other third party 24. Mirkin, C. A., Letsinger, R. L., Mucic, R. C. & Storhoff, J. J. A DNA-based material in this article are included in the article’s Creative Commons license, unless method for rationally assembling nanoparticles into macroscopic materials. indicated otherwise in a credit line to the material. If material is not included in the Nature 382, 607–609 (1996). article’s Creative Commons license and your intended use is not permitted by statutory 25. Macfarlane, R. J. et al. Nanoparticle superlattice engineering with DNA. regulation or exceeds the permitted use, you will need to obtain permission directly from Science 334, 204–208 (2011). 26. Lipshutz, R. J., Fodor, S. P., Gingeras, T. R. & Lockhart, D. J. High density the copyright holder. To view a copy of this license, visit http://creativecommons.org/ synthetic oligonucleotide arrays. Nat. Genet. 21,20–24 (1999). licenses/by/4.0/. 27. Smolensky, E. D., Peterson, K. L., Weitz, E. A., Lewandowski, C. & Pierre, C. Magnetoluminescent light switches—dual modality in DNA detection eric. © The Author(s) 2018 J. Am. Chem. Soc. 135, 8966–8972 (2013). NATURE COMMUNICATIONS | (2018) 9:2870 | DOI: 10.1038/s41467-018-05305-2 | www.nature.com/naturecommunications 7

Journal

Nature CommunicationsSpringer Journals

Published: Jul 20, 2018

There are no references for this article.