Nanobot synthesis

Nanobots were prepared as previously described³³. In brief, MSNPs were synthesized using a modified Stöber method⁴¹, reacting triethanolamine (35 g), ultrapure water (20 ml) and hexadecyltrimethylammonium bromide (CTAB; 570 mg) at 95 °C for 30 min while stirring. Tetraethyl orthosilicate (1.5 ml) was subsequently added dropwise; the mixture was left to react for 2 h at 95 °C and the resulting MSNPs collected by centrifugation and washed in ethanol (three times, 2,500g, 5 min). To remove the CTAB template, MSNPs were placed under reflux in acidic methanol (1.8 ml HCl, 30 ml methanol) for 24 h. Then, MSNPs were collected by centrifugation and washed three times in ethanol (2,500g, 5 min) before incorporating amine modification by adding APTES (6 µl per mg of MSNP) to MSNPs (1 mg ml⁻¹) in a 70% ethanolic solution at 70 °C, stirring vigorously for 1 h. MSNPs-NH₂ were collected and washed three times in ethanol and three times in water by centrifugation (three times, 1,150g, 5 min). MSNPs-NH₂ were resuspended in PBS at a concentration of 1 mg ml⁻¹ and total volume of 900 µl, and activated with glutaraldehyde (100 µl) for 2.5 h at room temperature. The activated MSNPs-NH₂ were collected and washed in PBS three times by centrifugation (1,150g, 5 min), resuspended in a solution of urease (3 mg ml⁻¹) and heterobifunctional PEG (1 μg PEG per mg of 5 kDa HS-MSNPs-NH₂) in PBS, and reacted for 24 h at room temperature. The resulting nanobots were then collected and washed three times in PBS by centrifugation (1,150g, 5 min) before resuspending them in a dispersion of AuNPs, prepared as previously described⁵¹, leaving them react for 10 min, and thoroughly washing by centrifugation (three times, 1,150g, 5 min).

Hydrodynamic size distribution and surface charge of the MSNPs, MSNPs-NH₂, nanobots and AuNP-decorated nanobots were determined using a Wyatt Mobius dynamic light scattering system and a Malvern Zetasizer, respectively. In all cases, concentration was 20 µg ml⁻¹ and acquisition time 5 s, using three runs per experiment. Three measurements per particle type were performed.

Synthesis of FITC MSNPs

A mixture of FITC (2 mg), ethanol (5 ml) and APTES (400 µl) was prepared and stirred for 30 min. Then, the previously described protocol for MSNP synthesis was followed, except that we added tetraethyl orthosilicate (1.25 ml) dropwise in combination with the FITC–APTES mixture (250 µl). The functionalization steps to obtain FITC-labelled nanobots were as aforementioned.

Synthesis of AuNPs

AuNPs were synthesized using a reported method³³. In brief, all materials were cleaned using freshly prepared aqua regia, thoroughly rinsed with water, and air-dried. Subsequently, a 1 mM AuCl₄ solution was heated to its boiling point while stirring in a round-bottom flask integrated into a reflux system. Following this, 10 ml of sodium citrate solution (30.8 mM) was added, and the solution was boiled for 20 min, resulting in a red colour. The solution was then allowed to cool to room temperature while stirring for 1 h. The resulting AuNPs were stored in the dark and characterization was conducted using transmission electron microscopy.

Enzymatic activity

Enzymatic activity of nanobots, ¹⁸F-nanobots and ¹³¹I-nanobots was measured using phenol red. To do so, 2 µl of nanobots (1 mg ml⁻¹) were added to a 96-well plate and mixed with 200 µl of different urea solutions (0, 50, 100, 200 mM) in 1.1 mM phenol red. Absorbance at 560 nm was measured over time at 37 °C.

Nanobot motion dynamics through optical microscopy

Optical videos of nanobots were acquired using a Leica Thunder microscope, coupled with a Hamamatsu high-speed CCD camera and a ×1.25 objective. For this, the nanobots were centrifuged and resuspended in 50 µl of PBS (final concentration of 20 mg ml⁻¹). Then, a Petri dish was filled with 3 ml of either PBS or a 300 mM solution of urea (in PBS) and observed under the microscope. A 5 µl drop with nanobots (20 mg ml⁻¹) was then added to the liquid-filled Petri dish and videos were recorded at 25 frames per second. Video pixel intensity distributions in ROIs were analysed at 15 s intervals using ImageJ software.

Radiolabelling of nanobots with [¹⁸F]F-PyTFP

Synthesis of [¹⁸F]F-PyTFP

[¹⁸F]F-PyTFP was synthesized in a Neptis xSeed module (Optimized Radiochemical Applications), following a previously reported method³³.

Synthesis of ¹⁸F-labelled nanobots

Nanobots were labelled with [¹⁸F]F-PyTFP, on the basis of a previously established procedure with minor modifications³³. In brief, 200 µl of nanobot solution (1 mg ml⁻¹) was centrifuged (10 min, 13,853g), resuspended in 10 µl of PBS (1 mM, pH 8), and incubated with 4 µl of [¹⁸F]F-PyTFP in acetonitrile (about 37 MBq) for 35 min at room temperature. After incubation, the reaction mixture was diluted with water (200 µl) and purified by centrifugation (5 min, 13,853g). The resulting pellet was then rinsed three times with water before being measured in a dose calibrator (CPCRC-25R, Capintec). Radiochemical yield was calculated as the ratio between the amount of radioactivity present in the nanobots after washing and the initial amount of radioactivity. Radiochemical purity after purification was ≥99%, as determined by radio thin-layer chromatography (radio-TLC) using iTLC-SG chromatography paper (Agilent Technologies) and dichloromethane and methanol (2:1) as the stationary and mobile phases, respectively. TLC plates were analysed using a TLC reader (MiniGITA, Raytest).

Stability of ¹⁸F-nanobots

The stability of ¹⁸F-labelled nanobots was determined using the following media: (1) 300 mM urea, (2) water, and (3) urine from tumour-bearing animals. ¹⁸F-labelled nanobots (10 µl) were incubated with the corresponding solution (100 µl) for 1 h at room temperature. Then, nanobots and supernatant were separated by centrifugation and collected, and radioactivity measured in a dose calibrator (CPCRC-25R).

Radiolabelling of nanobots with ¹³¹I

The radioiodination of urease nanobots was performed by incubating nanobots with injectable [¹³¹I]NaI solution (925 MBq ml⁻¹; GE HealthCare). In brief, 400 µl of urease nanobot solution (1 mg ml⁻¹) was centrifuged (13,853g, 5 min), resuspended in 100 µl of PBS (10 mM, pH 7.4) and incubated with 25 µl or 185 µl of injectable [¹³¹I]NaI (about 42.55 or 277.5 MBq, respectively) for 30 min, depending on the desired final activity. After incubation, the reaction mixture was purified by centrifugation (13,853g, 5 min). The resulting precipitate was washed three times with water (100 µl). Radioactivity in the supernatant and precipitate was determined using a dose calibrator (CPCRC-25R), and both fractions were analysed by radio-TLC, as for ¹⁸F-nanobots.

Animal model development

Mice were maintained and handled in accordance with European Council Directive 2010/63/UE and internal guidelines. All experimental procedures were approved by the CIC biomaGUNE ethics committee and local authorities (Diputación Foral de Guipuzcoa, PRO-AE-SS-276). Image analysis (both PET and MRI) was blinded towards group distribution of the animals.

The orthotopic murine model of bladder cancer was generated by intravesical administration of MB49 cells (murine carcinoma bladder cell line) to C57BL/6JRj female mice (8 weeks old, Janvier). For experiments aimed at determining tumour accumulation (four groups; details below), six animals were inoculated per group, as determined using precision analysis, with the following assumptions: required precision, 20%; expected s.d., ±20%; confidence, 95%; animal loss, 20%. For therapeutic efficacy experiments (six groups; details below), ten animals were included per group, as calculated using a one-tailed Student t-test, difference between two independent means, with the following assumptions: null hypothesis, treatment does not affect tumour growth; α, 0.05; 1 − β, 0.95; s.d., ±50%; expected differences between groups, 50%; animal loss, 20%. As the experiment was conducted in two batches for operational reasons, one control group was included in both batches (Table 2), and then all animals were pooled. For tumour establishment, mice were anaesthetized by inhalation of 3% isoflurane in pure O₂ and maintained by 1.0–1.5% isoflurane in 100% O₂. Then, the bladder was emptied, and chemical lesions induced on the urothelium by intravesically instilling 50 µl of poly-l-lysine (Sigma-Aldrich) through a 24-gauge catheter for 15 min. Subsequently, the bladder was emptied again and MB49 cells (10⁵ cells) in high-glucose DMEM (100 µl) were instilled for 1 h before removing the catheter and emptying the bladder via abdominal massage. Throughout the experiments, mice were monitored and weighed for health and welfare monitoring. A human endpoint was applied if weight loss exceeded 20% or on the basis of clinical symptoms, under the criteria of the veterinarian in charge.

Tumour size tracking

MRI studies were conducted 7 and 14 days after tumour induction, using a 7 T Bruker BioSpec USR 70/30 scanner (Bruker BioSpin) equipped with a BGA-12S gradient insert of 440 mT m⁻¹ and a 112/086 QSN resonator (T12053V3) for radiofrequency¹⁴ transmission, and a rat brain surface coil (T11205V3) for RF reception (both operating at 300 MHz). Animals were anaesthetized with isoflurane (4% for induction and 1.5% for maintenance in a 50% O₂/50% N₂ mixture) and placed on an MR-compatible cradle. Body temperature and respiration rate were continuously monitored using an MR-compatible monitoring device (model 1030 SA, Small Animal Instruments), interfaced to a small-rodent air heater system to maintain body temperature. After acquiring reference images, a spin-echo-based diffusion-weighted imaging sequence was used to image tumours, using the following parameters: echo time (T_E) = 22.3 ms, repetition time (T_R) = 2,500 ms, n = 2 averages, one A0 image (basal image with b = 0 s mm⁻²) and one DW image acquired using diffusion gradients in the (1, 0, 0) direction with a gradient duration δ = 4.5 ms and a gradient separation Δ = 10.6 ms, giving b = 650 s mm⁻², a 16 × 16 mm² field of view, image matrix size of 160 × 160 points, 20 consecutive slices of 0.5 mm thickness (no gap, acquired in interleaved mode) and a bandwidth of 192.9 Hz per pixel. To visualize tumours, images were postprocessed with ImageJ software, dividing images acquired with a diffusion gradient (b = 650 s mm⁻²) by those acquired without (b = 0 s mm⁻²), and applying a 3D Gaussian filter (σ_x = σ_y = σ_z = 0.7) to the result. Tumours were manually delineated to determine their volume.

In vivo biodistribution

On day 15 after tumour induction, mice were randomized into four groups to obtain homogeneous average tumour volume distributions among groups. PET-CT scans (MOLECUBES β and X-CUBE scanners) were acquired 3 h after intravesically administering 100 µl of ¹⁸F-BSA (groups 1 and 2) or ¹⁸F-urease (groups 3 and 4) nanobots at a concentration of 200 µg ml⁻¹, using either water (groups 1 and 3) or 300 mM urea in water (groups 2 and 4) as vehicle (Table 1). For image acquisition, animals were induced with anaesthesia (5% isoflurane in pure oxygen) and placed in a supine position before massaging the abdominal region for bladder evacuation. Immediately afterwards, the corresponding ¹⁸F-labelled nanobots (¹⁸F-BSA/¹⁸F-urease in water/urea) were instilled in the bladder through a 24-gauge catheter and incubated for 1 h, before removing the catheter, emptying the bladder and leaving the mice to recover from anaesthesia. At t = 3 h after administration, animals were re-anaesthetized and 10 min static whole-body PET images acquired, followed by CT scans. PET images were reconstructed using the 3D ordered subset expectation maximization reconstruction algorithm with random, scatter and attenuation corrections. PET-CT images of the same mouse were co-registered and analysed using the PMOD image processing tool. Plots of concentration of radioactivity versus time were obtained by creating a volume of interest on the upper bladder region using a 3D contour tool and measuring activity (decay corrected) in kilobecquerels per organ. Results were corrected by applying a calibration factor and then normalized by MRI-derived tumour volume.

Ex vivo studies

Histopathologic analyses

After completing all imaging, selected bladders (n = 3 per group) from tumour-bearing and healthy animals were removed in aseptic conditions and immediately fixed in 4% formaldehyde. Then, bladders were embedded in paraffin before taking 2–3 µm sections for haematoxylin–eosin staining. Representative images were obtained from all conditions for histopathologic examination.

ICP-MS analysis

Measurements were performed on a Thermo iCAP Q ICP-MS (Thermo Fisher Scientific) coupled with an ASX-560 autosampler (CETAC Tech). After completing all imaging, animals were killed, and selected bladders (n = 2 per group; four groups) collected and digested in 1 ml of HNO₃:HCl (4:1 mixture). The dispersion was boiled until organs were completely dissolved. Then, the solution was cooled to room temperature and analysed using ICP-MS to determine the concentration of Au in each sample, transforming the results into percentages of injected dose per gram of tissue (%ID g⁻¹).

Immunohistochemistry and confocal microscopy imaging

For immunohistochemistry analyses, tumour-bearing animals received FITC-labelled nanobots in water or 300 mM urea (n = 4 per group), as described above, for PET-CT studies. Additionally, tumour-bearing animals without nanobots served as a control group (n = 2). In all cases, bladders were collected, frozen and cut into 10 µm sections that were immediately fixed in 10% formaldehyde for 15 min, washed with 10 mM PBS and then incubated in 50 mM NH₄Cl in PBS for 5 min before rinsing again with PBS. Permeabilization was performed with methanol:acetone (1:1) for 5 min at room temperature and 0.1% Triton in PBS for 5 min. After PBS washing, samples were saturated with a solution of 5% BSA–0.5% Tween in PBS for 15 min at room temperature and incubated for 1 h at room temperature with mouse anti-FITC (1:100, Abcam) in 5% BSA–0.5% Tween. Sections were washed three times with 10 mM PBS for 5 min and incubated for 30 min at room temperature with secondary antibody Alex Fluor 647 donkey anti-mouse IgG (Molecular Probes, Life Technologies, 1:1,000) in 5% BSA–0.5% Tween in PBS, washed again in PBS (3 × 5 min) and mounted with a ProLong antifade kit with 4,6-diamidino-2-phenylindole (DAPI; Molecular Probes, Life Technologies). Images were acquired with a Leica STELLARIS 5 confocal microscope (UPV/EHU Scientific Park) with identical settings for all sections: ×10 magnification with tile imaging and stitching (typically 4 × 5 field of view). Laser line and detection windows were 405 nm and 440–503 nm for DAPI, 489 nm and 494–602 nm for FITC white laser and 653 nm and 660–836 nm for Alexa647 white laser.

Optical clearing

After perfusion with 4% paraformaldehyde and PBS, bladder samples were removed and further fixed in 4% paraformaldehyde overnight at 4 °C, then embedded in a 5 ml syringe with 0.8% low-melting-point agarose to form a cylindrical block and enable easy mounting in the quartz cuvette. The entire block was progressively dehydrated using methanol:H₂O at 4 °C (30%:70% for 1 h, 50%:50% for 1 h, 70%:30% for 1 h, 100%:0% for 1 h, then 100% methanol overnight and again for 4 h) and finally immersed in benzyl alcohol–benzyl benzoate (BABB) as refractive index matching solution for imaging. For in vitro comparisons of green FITC nanobots with commercial red particles, we used DiagNano (Creative Diagnostics) red fluorescent silica nanoparticles, 1 µm diameter, resistant to BABB clearing.

Autofluorescence and polarized sLS imaging

Light-sheet imaging was performed on MacroSPIM, a custom system for cleared whole-organ imaging developed at IRB Barcelona^44,45. In brief, samples are embedded in an agarose block, cleared together with the sample and imaged inside a quartz cuvette. Autofluorescence imaging used lasers at 488, 561 or 638 nm delivering illumination through a 50 mm achromatic doublet cylindrical lens (ACY254-050-A, Thorlabs). To reduce stripe artefacts, the light sheet is pivoted with a resonant scanner SC-10 (EOPC) along a 4f telescope with G322288322 100 mm achromatic doublet lenses (QI Optic Photonics). Tissue autofluorescence is collected through band- or long-pass fluorescence filters and recorded with an ORCA Flash v2 camera (Hamamatsu Photonics). Imaging was performed at ×9.6 with a ×8 zoom, ×2 lens and ×0.6 tube lens. The light sheet was flattened across the field of view, yielding 5–6 µm of axial resolution. 3D imaging was done in steps of 2.5 µm. Whole-bladder imaging was performed in 2 × 3 or 3 × 4 XY tiles, depending on organ size.

sLS imaging was achieved by removing the fluorescence filter or using any filter transmitting the laser. Light-sheet pivoting reduced laser speckle noise, resulting in temporal averaging of laser coherence as shown earlier⁵². The orientation of linear light-sheet polarization in illumination was controlled by rotating a half-wave plate (AHWP05M-600, Thorlabs) before the pivot scanner. The detected signal was selected in polarization using a rotating linear polarizer (LPVISC100, Thorlabs) before the filter wheel in detection, introducing >50% intensity loss in fluorescence detection. While sLS signal distribution in general changes with the polarizer’s orientation, the tissue autofluorescence signal remains unaffected by the polarizer’s rotation. sLS yields a spatial resolution of 2.4 ± 0.3 µm in BABB, which is comparable to the resolution in fluorescence light-sheet imaging (confirmed by fitting a Gaussian function to the XY image response of a single particle, Supplementary Fig. 8l–m) and close to the theoretical resolution in air (1.53 µm with numerical aperture (NA) = 0.2 at maximum macro zoom ×8).

Image processing and 3D analysis

Image processing, segmentation and analysis of light-sheet datasets was done with ImageJ/Fiji, while Figs. 3 and 4 were generated with Imaris Viewer 9.9 (https://imaris.oxinst.com/imaris-viewer) and Supplementary Video 3 was generated with Imaris 9 (https://imaris.oxinst.com/) (Bitplane, Oxford Instruments). Tiled light-sheet datasets were stitched with MosaicExplorerJ⁵³. Bladder tissue 3D segmentation was performed using custom ImageJ/Fiji macros for semi-automated 3D annotation of large volumes in virtual mode. In brief, a first script, ‘Macro1’, loads 3D image stacks, enables user annotation of ROIs in several planes and automatically interpolates the ROIs to generate and export 3D masks. ROIs were drawn every 15 planes (every 37.5 µm) to facilitate good segmentation continuity while keeping annotations to a reasonable minimum. A second script, ‘Macro2’, performs the mathematical or Boolean operations, plane by plane without loading the entire stacks into memory, either between 3D masks or between a 3D mask and the original data, saving the result as a new stack. All masks were generated by annotating autofluorescence images.

Both tumour and healthy tissue surface layers (Fig. 3) were delineated using Fiji’s wand and lasso tools on the bladder cavity in a mask. Calling this first iteration BC1, subsequent runs of Macro1 then automatically dilate this 3D contour by a defined pixel amount to yield new mask iterations, BC2, BC3 and so on, with increasing dilations. The first layer containing both tumour and healthy tissue, mask L1, is obtained by subtracting mask BC1 from BC2 and so forth, yielding L2 and L3 as concentric layers. The tumour volume closest to the cavity was obtained by annotating the tumour with wand and lasso tools to create a mask T1, while the healthy urothelium 3D layer was detected separately into mask U1. Subtracting U1 from L1 yields the surface layer of the tumour, and so forth: L2 − U1, L3 − U1. Conversely, the first layer of the urothelium is obtained by subtracting T1 from L1. All layers in Fig. 3 were defined to have 33 µm thickness.

The same suite of macros and procedures (ImageJ wand tool, digital erosion of 500 µm and so on) were used to delineate and segment the inner part of the bladder tissue and then estimate the bladder’s internal tissue volume (Fig. 4, see above for details). Histograms of the scattered signal intensity were created in Fiji by combining the scattered signal and mask.

RNT using ¹³¹I-nanobots

Between days 8 and 15 after tumour implantation, animals were divided into six groups (groups 1–6), trying to achieve similar average tumour volumes across groups (Table 2). For the experiments, animals were induced with anaesthesia (5% isoflurane in pure O₂) and positioned supine before emptying the bladder by massaging the abdominal region. Immediately afterwards, 100 µl of the appropriate treatment at a concentration of 400 µg ml⁻¹ (Table 2) was instilled into the bladder using a 24-gauge catheter. Treatment and vehicle (water or urea) remained in the bladder for 1 h before removing the catheter. The bladder was emptied again by abdominal massage and mice recovered from anaesthesia in their cages, replacing animal cage sawdust 24 h after treatment to remove radioactive contamination.

Therapeutic efficacy determined by MRI

Two MRI studies were performed on each mouse: (1) between days 7 and 14 after tumour inoculation to randomize animals among groups and measure initial (pretreatment) tumour volumes; (2) between days 16 and 21 after tumour inoculation (post-treatment) to evaluate therapeutic efficacy. MRI was conducted using 7 T Bruker BioSpec and 11.7 T Bruker BioSpec scanners (both with ParaVision 7 software), depending upon availability. This did not affect the results since the external field is not critical for anatomical imaging¹⁴. Imaging experiments were conducted using the same imaging parameters and processing as explained above (Tumour size tracking). In the case of the 11.7 T scanner the set-up consisted of a mouse heart surface coil for the reception and a volumetric coil for transmission. Tumour volumes in each slice were determined from manually drawn volumes of interest covering the tumour area.

Statistical analysis

In PET imaging studies, percentages of injected dose (% ID) and injected dose per tumour volume (% ID cm⁻³) were compared using one-way ANOVA. Differences between groups were determined using Tukey’s multiple comparisons test. NTV in RNT section was obtained from a t-test of unpaired values. Data distribution was assumed to be normal, but this was not formally tested. Statistical analyses were performed with GraphPad Prism v.8.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

• SEO Powered Content & PR Distribution. Get Amplified Today.
• PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
• PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
• PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
• PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
• Source: https://www.nature.com/articles/s41565-023-01577-y