Urease-powered nanobots for radionuclide bladder cancer therapy

Abstract

Bladder cancer treatment via intravesical drug administration achieves reasonable survival rates but suffers from low therapeutic efficacy. To address the latter, self-propelled nanoparticles or nanobots have been proposed, taking advantage of their enhanced diffusion and mixing capabilities in urine when compared with conventional drugs or passive nanoparticles. However, the translational capabilities of nanobots in treating bladder cancer are underexplored. Here, we tested radiolabelled mesoporous silica-based urease-powered nanobots in an orthotopic mouse model of bladder cancer. In vivo and ex vivo results demonstrated enhanced nanobot accumulation at the tumour site, with an eightfold increase revealed by positron emission tomography in vivo. Label-free optical contrast based on polarization-dependent scattered light-sheet microscopy of cleared bladders confirmed tumour penetration by nanobots ex vivo. Treating tumour-bearing mice with intravesically administered radio-iodinated nanobots for radionuclide therapy resulted in a tumour size reduction of about 90%, positioning nanobots as efficient delivery nanosystems for bladder cancer therapy.

Fig. 1. Fabrication, radiolabelling, characterization and motion dynamics of urease-powered nanobots.

a, Schematic representation of the nanobot fabrication process and radiolabelling. Ur, urease. b, Left: nanobot characterization by dynamic light scattering (n = 3, technical replicates). Data are presented as mean values and error bars represent the s.e.m. Centre: zeta potential (n = 3, technical replicates) Data are presented as mean values and error bars represent the s.e.m. Right: transmission electron microscopy image. c, Snapshots depicting the nanobot motion dynamics in the absence and presence (300 mM) of urea as fuel, and corresponding pixel intensity histograms for the ROI marked by a circle. Panel a created with BioRender.com. Source data

Fig. 2. In vivo studies of nanobot accumulation in a bladder cancer orthotopic murine model.

a, Left: tumour volumes (determined by MRI) on days 7 and 14 after cell implantation for the different study groups (n = 6 per group, biological replicates). Results are expressed in a box plot (centre line at the median; upper and lower bounds at 75th and 25th percentiles, respectively; one dot per animal) with whiskers at minimum and maximum values. Right: 2D DW-MRI images of the bladder (hypointense circular region) of two representative mice at t = 7 and 14 days after inoculation of MB49 cells. Scale bars, 2 mm. 3D renders of whole bladders (transparent) and tumours (purple) are presented next to each MRI image. b, Haematoxylin–eosin stains showing bladders containing tumours (delineated by red dotted lines) for one representative animal per group. Scale bars, 1 mm. c, 2D DW-MRI images of the bladder (hypointense circular region) of one representative animal per study group. Tumours appear hyperintense (tumours delineated by a red dotted line and bladders in yellow). Scale bars 2 mm. d, Coronal PET 2D images overlaid on CT images of one representative animal per study group. The dotted cyan cross shows the bladder position and the radioactive intensity has been colour-coded (given as the percentage of injected dose per millilitre, %ID cm⁻³). Scale bars, 5 mm. e, 3D renders of the whole bladder (transparent) and radioactive signal accumulation (colour-coded) of the PET-CT images shown in d. f, Box (centre line at the median; upper and lower bounds at 75th and 25th percentiles, respectively; one dot per animal) and whisker (minimum and maximum values) plot of radioactivity accumulation normalized by tumour volume (determined by MRI) in animals from all groups (n = 6 for BSA-NPs in water and nanobots in urea, n = 5 for the other groups, biological replicates), shown as percentages of the injected dose per cubic centimetre of tumour (%ID cm⁻³). Statistical analysis was performed via one-way analysis of variance (ANOVA). g, Correlation of tumour accumulation obtained by PET and ICP-MS results for all groups (n = 2 per group, biological replicates). Source data

Fig. 3. Nanobots penetrate the bladder tumour.

a, Tiled acquisition layout in XY and XZ showing the centred light-sheet waist for each XZ column; sLS principle where laser light is scattered by particles (for example, nanobots) and passes a filter for detection versus autofluorescence (AF) where the laser is blocked. b, Selected plane in bladder centre with tumour; light-sheet excitation from right. Tumour (Tu) detectable with a dashed yellow line delineating its surface exposed to the bladder inner cavity. Healthy tissue defined with a dark lamina propria (Lp) and a bright layer of urothelium (Ut), surrounded by detrusor muscles (Dm). Delineating tumour against inner tissues is more challenging; the tumour’s limits could lie anywhere within the orange dashed area (Supplementary Video 3). n = 1. ‘Glow’ colour scale: sLS showing nanobots (inside bladder) and agarose crystals (outside). c, sLS signal only, showing nanobots inside bladder, agarose crystals outside and scattered signal in the periphery (muscle). The dashed line shows external tissue boundaries; the dotted line shows the same region after 500 µm digital erosion. d, Area inside dotted line in c. e, Maximum-intensity projection (MIP) of sLS signal, where all external scatterers (agarose, muscle) contribute to signal. f, MIP of scattered signal from volume shown in d, containing only nanobots. g, Integrated sLS signal intensity normalized by layer volume inside four masks shown in h,j,k (Supplementary Methods). TL, tumour layer; HL, healthy layer. a.u., arbitrary units. h, Optical sections at different depths with annotations of tissue layers for quantification (j). i, 3D surface render of bladder, same orientation as e,f. j, Left: tumour and healthy layers of tissues along bladder cavity, detected and annotated (masks) in three dimensions. Cyan, urothelium; HL, healthy tissue; red, tumour surface at 0–33 µm depth; orange, 34–67 µm depth; yellow, 68–100 µm depth. Right: sLS signal coloured with masks. k, Left: 3D renders of masks from h,j. Right: 3D MIP of sLS signal inside masks. Scale bars: b–d,i,k, 400 µm; f,h,j, 500 µm. Source data

Fig. 4. Comparing nanobot retention in bladder tissues with tumours.

a–j, Each tissue undergoes identical processing and visualization (n = 1). Samples are imaged in autofluorescence and with V-LS (vertically-polarized light-sheet) and V-CAM (vertically-polarized detection) sLS. Left to right: single slice in mid-plane with autofluorescence (cyan) and processed mask (magenta) of internal tissue after removing internal lumen and digitally eroding an external layer of 500 µm from the outer edge of the bladder; summed intensity of sLS signal inside the mask; single slice of sLS signal inside the mask; MIP of sLS signal inside the mask and two insets showing the same sLS signal with different look-up tables and intensity scales (j). Rows represent different animals and conditions. a,b, Tumour in bladder with nanobots in urea (same sample as in Fig. 3 and Supplementary Fig. 5a–f). c,d, Tumour in bladder with nanobots in water, that is, ‘without fuel’. e,f, Tumour in bladder without nanobots (same sample as Supplementary Fig. 5h–m). g,h, Healthy bladder without tumour and with nanobots in urea. i,j, Healthy bladder without tumour and without nanobots (same sample as in Supplementary Fig. 4o–x). k,l, Voxel intensity histograms (3D) of the masked sLS volumes from a,c,e,g,i (that is, summed intensity inside masks) showing the distribution of the scattered signal on linear (k) and logarithmic (l) (16-bit) axes. While background distributions in l vary for different mask volumes, pink and green curves (nanobots in urea) correlate with MIP images in a,b,g,h, with a long high-intensity tail indicating a prominent signal; in other conditions (without nanobots or with nanobots in water) intensity drops off quickly with no counts above intensities of about 7,000. In g, the cavity inside the bladder was too thin and collapsed to be segmented, hence the mask includes the full volume (including the lumen), although residual cavity background does not contribute to the high intensities. Look-up table scales (‘Red Hot’ and ‘Inverted Red Hot’) and intensity limits shown in i,j apply to all panels. Scale bars, a,c,e,g,i, 1 mm; b,d,f,h,j insets, 200 µm. Source data

Fig. 5. RNT studies using ¹³¹I-nanobots in a bladder cancer orthotopic murine model.

a, Schematic representation and timeline of the RNT studies. b, Changes in body weight over time, showing mean and s.e.m. (n = 9 per group, except n = 13 for nanobots in urea and n = 7 for high-dose (HD) ¹³¹I-nanobots in water and urea, biological replicates). LD, low-dose. c, DW-MRI 2D slices through the bladder of tumour-bearing mice before and after treatment with radionuclides; low-dose (LD) and high-dose (HD) denote 1.85 MBq and 18.5 MBq doses of ¹³¹I, respectively; yellow dotted lines show the bladders. Scale bars, 2 mm. d, NTV obtained by MRI before and after treatment. Tumour volumes were normalized by the means of the pretreatment values of each group (n = 9 per group, except n = 13 for nanobots in urea and n = 7 for high-dose ¹³¹I-nanobots in water and urea, biological replicates). Data are presented as mean values and error bars represent the s.e.m. Statistical significances are based on a two-tailed unpaired t-test. Inset: tumour volume changes with respect to pretreatment (see e for legend). e, Post-treatment tumour volumes normalized to control condition (non-treated group). Panel a created with BioRender.com. Source data