Cell culture

Brown adipocytes were differentiated from an immortalized brown preadipocyte mouse cell line kindly provided by B. Spiegelman (Harvard University) as previously described¹⁹. Preadipocytes were maintained and expanded at low confluence (<50%) in a growth medium (Dulbecco’s modified Eagle’s medium, high glucose (Gibco), supplemented with 20% foetal bovine serum (Sigma-Aldrich), 20 mM HEPES (Sigma-Aldrich) and 100 U ml⁻¹ penicillin–streptomycin (Gibco)). For differentiation, preadipocytes were washed twice with phosphate-buffered solution (PBS) (Gibco, pH 7.4, 1×), dissociated with TrypLE Express, and seeded at a density of approximately 23,000 cells cm⁻² in the growth medium. The medium was changed after 24 h to a differentiation medium (Dulbecco’s modified Eagle’s medium, high glucose, supplemented with 10% foetal bovine serum, 20 nM insulin, 1 nM triiodothyronine and 100 U ml⁻¹ penicillin–streptomycin) and the cells were cultured for 2 days. Cell differentiation was then induced by the addition of an induction medium (differentiation media supplemented with 0.125 mM indomethacin, 0.5 μM dexamethasone and 0.5 mM isobutyl methylxanthine) for 2 days, after which the medium was changed to the differentiation medium for two additional days. Differentiated adipocytes were washed once with a starvation medium (Dulbecco’s modified Eagle’s medium, low glucose (Gibco), supplemented with glucose to a final concentration of 8 mM, 0.5% bovine serum albumin (Sigma-Aldrich) and 100 U ml⁻¹ penicillin–streptomycin) and incubated for 2 h in the starvation medium before treatment with insulin NanoRods or NanoRods diluted in the starvation medium for the times indicated in the main text. Following treatment, the cells were washed once in PBS and harvested for the isolation of protein for immunoblots or RNA for gene expression. For IR analysis by DNA-PAINT, preadipocytes were differentiated as described above with the following modifications. After treatment with the induction medium, the cells were dissociated with TrypLE Express and seeded in the differentiation medium in μ-Slide 18 Well Glass Bottom wells (ibidi). After 2 days, the cells were incubated with the starvation medium for 2 h followed by treatment with 10 nM insulin in the starvation medium for 10 min. In control experiments, the addition of insulin was omitted. Before treatment, differentiated adipocytes were visually analysed to assess the cell surface density and to evaluate the differentiation of cells, and then randomly allocated to the treatment and control groups.

The cells were maintained, expanded and differentiated in a humidified atmosphere containing 5% CO₂ at 37 °C.

DNA-PAINT of IR

Differentiated adipocytes were fixed for 12 min at room temperature (RT) with pre-warmed 4% paraformaldehyde in PBS, washed three times with PBS, blocked for 90 min at RT with a blocking solution (3.0% foetal bovine serum/0.1% Triton X-100 in PBS) and incubated with rabbit anti-IR β antibody (Cell Signaling (4B8); 1:300 dilution in the blocking solution) for 2 days at 4 °C. The cells were then washed three times with PBS and incubated with the nanobody FluoTag-XM-QC anti-rabbit IgG (Massive Photonics) diluted 1:200 in a blocking buffer (Massive Photonics) for 1 h at RT. Following three washes with PBS, the cells were incubated with 80 nm gold nanoparticles (Sigma-Aldrich; 1:5 dilution in the blocking buffer) for 10 min. The cells were washed once with PBS and incubated with 5 nM Cy3b-labelled strands (Massive Photonics) diluted in an image buffer (Massive Photonics).

Imaging was carried out on a Nikon ECLIPSE Ti-E microscope with a Perfect Focus system (Nikon Instruments), applying an objective-type total internal reflection fluorescence configuration using an iLAS2 circular total internal reflection fluorescence module (Gataca Systems) with an oil-immersion 1.49-numerical-aperture CFI Plan Apo total internal reflection fluorescence ×100 objective (Nikon Instruments) equipped with ×1.5 auxiliary Optovar magnification corresponding to a final pixel size of 87 nm. The laser used was an OBIS 561 nm LS 150 mW (Coherent) with custom iLas input beam expansion optics (Cairn) optimized for reduced field super-resolution imaging. The fluorescent light beam was passed first through a filter cube (89901, Chroma Technology) containing an excitation quadband filter, a quadband dichroic filter and a quadband emission filter (ZET405/488/561/640x, ZET405/488/561/640bs and ZET405/488/561/640m, Chroma Technology). Fluorescence light was then spectrally filtered with an emission filter (ET595/50m, Chroma Technology) and imaged on an iXon Ultra 888 electron-multiplying charge-coupled device camera (Andor). Micro-Manager software v. 1.4 was used to acquire 12,000 frames with 10 MHz readout frame, 130 ms exposure and no electron multiplication gain. A total of ten cells from three independent experiments were imaged in each condition (Supplementary Note and Supplementary Fig. 1).

INS-DNA production and purification

Insulin (Merck, 1 mg ml⁻¹) was reacted with dibenzocyclooctyne-sulfo-N-hydroxysuccinimidyl ester (DBCO-sulfo-NHS, Click Chemistry Tools; 690 µM) in 100 mM Na₂CO₃ buffer (pH 11.5) for 20 min at RT. The reaction was then quenched for 5 min through the addition of Tris base (Sigma-Aldrich, 100 mM). The solution was washed three times with 400 µl of 100 mM Na₂CO₃ using Amicon Ultra 0.5 ml centrifugal filter units with a 3 kDa cut-off membrane (Merck). At each washing step, the columns were spun for 10 min at 14,000×g. After the final washing step, insulin–DBCO (690 µM) was mixed with azide-modified DNA (Biomers, 35 µM; Supplementary Table 3) in 100 mM Na₂CO₃ and left to react for 3 h at RT. The reaction was quenched by adding NaN₃ (Sigma-Aldrich, 6.9 mM). The samples were run on a native polyacrylamide gel (6% 19:1 polyacrylamide in 1× TAE, 20 min, 200 V, TAE running buffer) and stained with SYBR Gold (Thermo Fisher) according to the manufacturer’s instructions, for confirming INS-DNA formation. Imaging was performed using an ImageQuant LAS 4000 gel imager. The insulin–DBCO-sulfo-NHS conjugation protocol was optimized to promote the binding of the ssDNA oligo to lysine-29 of the B chain (B29 lysine) compared with the amine groups at the N terminus of insulin A and B chains. The pKa value of the B29 amine is higher than those of the amines of the two N termini (11.2 versus 8.6 and 6.8). At high pH, the NHS group of the crosslinker is predicted to preferentially react with the most basic amine group, which is the B29 lysine amine³². Therefore, the pH reaction conditions were optimized to promote a single INS-DNA product.

INS-DNA reaction mixes were purified by a reverse-phase high-performance liquid chromatography C₁₈ column (Agilent Poroshell 120 EC-C₁₈) on an Amersham Pharmacia Biotech ÄKTA Ettan LC. Buffer A (50 mM triethylamine acetate) and buffer B (90% acetonitrile and 10% buffer A) were used in a gradient profile, in which the percent of buffer B was increased from 30% to 50% over 20 min. Fractions were collected and spun on a vacuum concentrator (Thermo Scientific SpeedVac Savant DNA 120) for 30 min at high heat to remove the volatile components of the high-performance liquid chromatography buffers. Selected peaks were buffer exchanged into PBS using Amicon Ultra 0.5 ml centrifugal filter units with a 3 kDa cut-off membrane (Merck), by spinning three times for 10 min at 14,000×g and washing each time with 400 µl PBS. Samples of purified fractions were run on a native polyacrylamide gel and stained with SYBR Gold (Thermo Fisher) to visualize the purified INS-DNA. The final purity of conjugates was analysed for every preparation via three methods: comparison of silver staining band intensities (Pierce Silver Stain Kit) on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (Invitrogen, 4–12% Bolt gel) against insulin standards (Merck), comparison of SYBR Gold band intensity on native polyacrylamide gel electrophoresis against DNA standards (Integrated DNA Technologies) and through the Qubit ssDNA Assay Kit (Qubit 4 Fluorimeter, Invitrogen). Final concentrations of INS-DNA were calculated based on Qubit ssDNA measurements. Purified INS-DNA was frozen and stored at −20 °C until further use.

Production of NanoRods and insulin NanoRods

Origami structures were prepared by mixing scaffold plasmid DNA (p7560, Tilibit, 10 nM) with the appropriate staple strands (Integrated DNA Technologies, 100 nM) (Supplementary Tables 4–9) in a folding buffer (5.0 mM Tris at pH 8.5 (Sigma-Aldrich), 1.0 mM EDTA (Panreac AppliChem) and 12.5 mM MgCl₂ (Sigma-Aldrich)). The mix was then placed in a thermocycler (MJ Research PTC-225 Gradient Thermal Cycler) and annealed by heating to 80.0 °C for 5 min, cooling to 60.0 °C at 1.0 °C per min over 20 min and then slowly cooling to 20.0 °C at 0.5 °C per min. Excess staples were removed using Amicon Ultra 0.5 ml centrifugal filter units with a 100 kDa cut-off membrane (Merck) by spinning five times for 2 min at 14,000×g and washing each time with 400 µl of the folding buffer. The concentration of the purified structure was determined by measuring the DNA absorbance at 260 nm (Thermo Scientific NanoDrop 2000). Purified INS-DNA was then added in 3× stoichiometric excess to available extended strand-binding sites on the NanoRod structure and annealed in a thermocycler by heating to 37.0 °C for 1 h, cooling to 22.0 °C at 0.1 °C per min, incubating at 22.0 °C for 14 h and cooling to 4.0 °C at 0.1 °C per min. Unbound INS-DNA was removed using Amicon Ultra 0.5 ml centrifugal filter units with a 100 kDa cut-off membrane (Merck), spinning five times for 2 min at 14,000×g and washing each time with 400 µl of PBS + 10 mM MgCl₂. Insulin NanoRods were stored at 4 °C until further use.

Agarose gel electrophoresis

NanoRod structures were analysed by running samples on agarose gels on ice for 4 h at 70 V. Gels were composed of 2% agarose (Thermo Scientific TopVision Agarose) in 0.5× TBE buffer (Panreac AppliChem) plus 10 mM MgCl₂ (Sigma-Aldrich) and 1× SYBR Safe DNA stain (Invitrogen). Imaging was performed using an ImageQuant LAS 4000 gel imager.

Dynamic light scattering

NanoRod and insulin NanoRod samples were prepared in PBS + 10 mM MgCl₂, syringe filtered using 0.1 µm membranes (Merck) and analysed on a Zetasizer Ultra instrument (Malvern Panalytical). Three measurements were taken at 25 °C in a low-volume cell (ZSU1002) and then averaged.

oxDNA simulation of the NanoRod

NR, NR-1, NR-2, NR-4, NR-7 and NR-15 structures were analysed using oxDNA coarse-grained modelling (https://oxdna.org/). NanoRod structures with dsDNA strands extending from the sites of insulin incorporation were created using vHelix and converted to the oxDNA format using the tacoxDNA web site (http://tacoxdna.sissa.it/). Structures were submitted to the oxDNA.org web server for simulation at 37 °C, with 1 as the salt concentration, 1 × 10⁸ time steps with a dt value of 0.0001, and a preliminary relaxation step with the default parameters. Simulations were visualized and videos were made using the oxView tool (https://oxdna.org/).

Negative-staining TEM

Purified NanoRods (10 nM) were dispensed on a glow-discharged carbon-supported copper TEM grid (TEM-CF200CU50, Thermo Fisher Scientific) and incubated for 60 s before removing the solution. The grids were then stained for 10 s with 5 µl of 2% w/v uranyl formate, which was subsequently removed. The staining procedure was repeated seven times and the TEM grids were air dried for 30 min before imaging. Imaging was performed on a Talos 120C G2 (120 kV, Ceta-D detector) at ×92,000 for near-field views. Raw images were processed using ImageJ software (v1.53).

AFM

NanoRod structures were imaged on a disc of mica fastened with epoxy adhesive to the centre of a microscope slide and enclosed by a plastic ring attached to the slide using Reprorubber. Nanostructures were diluted to 1 nM in TE-Mg buffer (5 mM Tris base, 1 mM EDTA, 10 mM MgCl₂, pH 8.0) and 10 µl was pipetted onto freshly cleaved mica. After 30 s, 4 µl of 5 mM NiSO₄ was added and incubated for a further 4.5 min. The surface was then rinsed with 1.0 ml of 0.1 µm-filtered TE-Mg buffer after which 1.5 ml of filtered TE-Mg buffer was added to the mica disc for imaging. Imaging was performed in liquid using a JPK Instruments NanoWizard 3 Ultra atomic force microscope with a Bruker AC40 cantilever in the a.c. mode.

DNA-PAINT of insulin NanoRod nanostructures

The µ-Slide 18 Well Glass Bottom wells (ibidi) were cleaned with isopropanol and dried with N₂. The wells were incubated with 1 mg ml⁻¹ Biotin bovine serum albumin (Thermo Fisher) in buffer A (10 mM Tris-HCl, 100 mM NaCl and 0.05% (v/v) Tween 20 at pH 8.0) for 5 min at RT, washed three times with buffer A and incubated with 0.5 mg ml⁻¹ neutravidin (Thermo Fisher) in buffer A for 5 min at RT. Wells were then washed three times with buffer A followed by three washes with buffer B (5 mM Tris-HCl, 10 mM MgCl₂, 1 mM EDTA and 0.05% (v/v) Tween 20 at pH 8.0). Then, 500 pM NanoRods incorporating four biotin-labelled DNA strands (Extended Data Fig. 2a), with INS-DNA containing a 9-nucleotide PAINT docking sequence (DS1; Supplementary Table 3), was added to each well for 5 min. The wells were washed three times with buffer B. Then, 1 nM of Atto-647N imager strand (IS1; Supplementary Table 10) in buffer B supplemented with oxygen scavengers (protocatechuic acid (Sigma), protocatechuate 3,4-dioxygenase (Sigma) and Trolox (Sigma)) was added to each well. For dual-exchange PAINT, three docking sequences (DS2; Supplementary Table 10) were added on both ends of the NanoRods. Samples were prepared as previously described, with wells being washed ten times with buffer B between each imaging acquisition. Each imaging acquisition was done with a different Atto-647N imager strand (IS1 and IS2; Supplementary Table 10). Micro-Manager software was used to acquire 9,000 frames with 10 MHz readout frame, 200 ms exposure and no electron multiplication gain (Supplementary Note and Supplementary Figs. 2 and 3).

SPR

A Biacore T200 instrument (Cytiva) was used to perform the SPR experiments and data were acquired using Biacore T200 system control software v. 2.01. Biotinylated ECD-IR (Nordic BioSite) was immobilized on a streptavidin Sensor Chip SA (Cytiva). HBS-P+ buffer (Cytiva) was used as the running buffer. After the immobilization of ECD-IR, a stabilization time of 15 min was introduced to reach a stable baseline. NR-1, NR-2, NR-4, NR-7, NR-15, NR-8 and NR-8dsDNA were injected at 11.4 nM concentration of insulin in the running buffer. Insulin and INS-DNA were injected at 55 nM, the minimum concentration to obtain a binding curve that could be analysed. NR was injected as a negative control at a concentration equal to the highest concentration of NanoRod used within insulin NanoRod injections (NR-1 = 11.4 nM). Alternatively, NR-2, NR-4, NR-7 and NR-15 were injected at 5.7 nM concentration of nanostructure (Extended Data Fig. 4e), and NR-7^K-PEG was also injected a 50 nM concentration of insulin (Extended Data Fig. 9c). The injection of each sample was performed using an association phase of 180 s and a dissociation phase of 300 s. The dissociation equilibrium constant (K_D), association rate constant (k_on) and dissociation rate constant (k_off) were determined using the BIAevaluation 3.0 software. The t_1/2 values, which define the residence time, were determined using the formula ln2/k_off. To compare the binding of NR-7 structures between IR and IGF1R, ECD-IR and ECD-IGF1R proteins (Nordic BioSite) were immobilized on two different flow cells of a CM5 sensor chips via amine coupling reactions, according to the manufacturer’s instructions. The binding of insulin and INS–DNA was tested by injecting different concentrations of insulin (6.2, 18.5, 55.6, 166.7 and 500.0 nM) in the running buffer (HBS-P+) in the single-cycle kinetic mode, using an association phase of 140 s and a dissociation phase of 300 s. Binding of NR-7 was tested by injecting a single concentration of structure (11.4 nM of insulin) using an association phase of 180 s and a dissociation phase of 300 s.

Coomassie gel

Here 0.5 µg of recombinant ECD-IR (Nordic BioSite) was resuspended in Laemmli sample buffer (Bio-Rad). For reducing conditions, 2-mercaptoethanol was added to a final concentration of 2.5%. The samples were denaturated at 80 °C for 10 min, resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and stained with GelCode Blue safe protein stain (Thermo Fisher Scientific).

Gel shift assay

NanoRods (NR and NR-7) at 20 nM were incubated with 300 nM recombinant extracellular domain of human IR (Nordic BioSite) or with PBS for 30 min at 4 °C. The samples were then run on a 2% agarose gel and stained with SYBR Safe.

Immunoblotting

The cells were washed with PBS, lysed in radioimmunoprecipitation assay buffer (Sigma-Aldrich) supplemented with protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific) and incubated on ice with shaking for 30 min. The lysate was cleared by centrifugation (20,000×g for 20 min at 4 °C) and the protein lysates were quantified using the Bradford protein assay (Bio-Rad). Protein lysates were resuspended in the Laemmli sample buffer (Bio-Rad) containing 2.5% of 2-mercaptoethanol, denaturated at 80 °C for 10 min, resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes. Membranes were incubated 1 h in a blocking solution (Tris-buffered saline with 0.1% Tween 20 (TBST) and 5.0% non-fat dry milk), followed by overnight incubation at 4 °C with primary antibodies against phospho-IR beta/IGF1R beta (CST 3024, 1:1,000), phospho-AKT-S473 (CST 4058, 1:1,000) or GAPDH (Invitrogen PA1-987, 1:5,000). After three washes with TBST, the membranes were incubated with horseradish-peroxidase-conjugated secondary antibodies (Invitrogen, 31460, 1:5,000) for 1 h at RT. Detection of horseradish peroxidase was performed by chemiluminescent substrate Immobilon Forte on a ChemiDoc Imaging System (Bio-Rad). Band densiometry was performed using the ImageJ software.

Flow cytometry

Differentiated, serum-starved adipocytes were prepared as described above. Adipocytes were subsequently washed twice with Hanks’ balanced salt solution (HBSS) and dissociated in collagenase D solution (1.5 U ml⁻¹ collagenase D (Roche) and 10 mM CaCl₂ in HBSS) for 20 min at 37 °C. The cells were resuspended in HBSS, filtered through a 35 µm HBSS-equilibrated cell strainer (BD Biosciences), pelleted at 300×g for 5 min and resuspended in a staining buffer (1× PBS and 1% bovine serum albumin). The dead cells were labelled with LIVE/DEAD Fixable Yellow Dead Cell Stain Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions, and subsequently, the cell suspension was pelleted at 300×g for 5 min and resuspended in the staining buffer. Here ~100,000 cells were incubated with 10 nM ATTO-647-labelled NanoRod structures in a final volume of 100 µl for 10 min at 37 °C. The cells incubated without the NanoRod structures were used as the untreated control. The cells were then washed twice with the staining buffer by centrifugation at 300×g for 5 min. Flow cytometry measurements were performed on a BD FACSCANTO II with BD FACSDIVA software v. 9.0 (BD Biosciences). Live adipocyte cells were initially identified by gating cells on FSC-A versus AmCyan-A, followed by gating on FSC-A versus FSC-H to detect singlets. The acquired data were analysed using FlowJo 10.7.1 software (BD Biosciences). The geometric mean of the fluorescence intensity after normalization to the untreated control was used to define the degree of cell labelling by the NanoRods.

RNA-seq library preparation and sequencing

The total RNA was isolated from differentiated adipocytes using Quick-RNA Microprep Plus Kit (Zymo Research), and 500 ng of purified RNA was used for mRNA-seq library preparation using TruSeq RNA Library Prep Kit v2 according to the manufacturer’s low-sample protocol. Library quantification was performed using the QuantiFluor dsDNA system (Promega) according to the manufacturer’s multiwell plate protocol on a Varioskan LUX multimode microplate reader (Thermo Fisher). The library size and quality were assessed using Bioanalyser 2100 and High Sensitivity DNA Kit (Agilent). Libraries were denatured and diluted using NextSeq standard normalization protocol (Illumina), and sequencing was performed using single-end reads (1 × 75 bp) with NextSeq 500/550 High Output Kit v. 2.5 (75 cycles) on a NextSeq 550 platform (Illumina).

RNA-seq quantification, DEG analysis and GSEA

Sequencing reads were mapped against a reference transcriptome of Mus musculus protein-coding transcript sequences (release M29, GRCm39; https://www.gencodegenes.org/mouse/) and quantified using Salmon 1.7.0 (ref. ³³). Count tables were generated using the tximport package³⁴ and lists of DEG were obtained using the DESeq2 package (v. 1.34.0)³⁵, where only genes with adjusted P values equal to or below 0.001 and a log₂fold change cut-off at ±0.58 were considered for further analysis. Heat maps and UpSet plots were generated using ComplexHeatmap (v. 2.10.0)³⁶. GSEA for biological processes with both Gene Ontology terms and KEGG pathways was performed using a ranked list of genes as input to clusterProfiler (v. 4.2.2)³⁷ and a significance of false-discovery-rate-adjusted P values below 0.10 and 0.05, respectively.

Zebrafish microinjections and free glucose quantification

NanoRod and insulin NanoRods were mixed with oligolysine-PEG (K₁₀-PEG_5K, Alamanda Polymers) at a 1:1 ratio between the amines of lysines in K₁₀-PEG_5K and the phosphates in DNA³⁰, and incubated at RT for 30 min before a microinjection of 2 nl of the sample into each zebrafish larva. The samples for injection were prepared at a final concentration of 100 nM of structures for the coated NanoRod samples and at a final concentration of 100 nM insulin for the coated insulin NanoRod samples (corresponding to 100.0 and 14.3 nM of NanoRods for NR-1^K-PEG and NR-7^K-PEG, respectively). Injection of 1 nl of human insulin at 100 nM concentration in zebrafish larvae has been shown to induce a decrease in free glucose levels and transcriptional changes consistent with insulin signalling³⁸. We, therefore, injected 2 nl of 100 nM insulin concentration in our assays to evaluate the insulin-mediated effects on free glucose levels. Since the total blood volume for a 2 dpf zebrafish is 60–89 nl (ref. ³⁹), the estimated concentration of the injected insulin in our assays would be around 2–3 nM. In these assays, we were also limited in the amount of sample injected, with higher injection levels (3 and 4 nl) resulting in poor larvae survival.

The maintenance and crossing of zebrafish (D. rerio) lines were conducted in compliance with Swedish legislation on animal welfare regulations approved by Stockholms djurförsöksetiska nämnd. Since for β-cell ablation and free glucose assay experiments only animals younger than 5 days were used, no ethical permit was required according to 2010/63/EU. Zebrafish transgenic lines used have been previously described, namely, Tg(ins:CFP–NTR)^s892 (ref. ²⁷) and Tg(ins:Kaede)^s949 (ref. ²⁸).

Ablation of β-cells was performed in two-day-old Tg(ins:CFP–NTR) and Tg(ins:CFP–NTR);Tg(ins:Kaede) embryos by treatment with 10 mM MTZ (Sigma-Aldrich) diluted in 1% DMSO (VWR) in an egg water solution (E3) supplemented with 0.2 mM 1‐phenyl‐2‐thiourea (PTU, Acros Organics) for 24 h. Following β-cell ablation, three-day-old Tg(ins:CFP–NTR) larvae (72 hpf) were anaesthetized in 0.01% tricaine and injected with 2 nl of 1× PBS, unmodified insulin or coated NanoRod/insulin NanoRods into the common cardinal vein (duct of Cuvier)⁴⁰. Phenol red (Sigma-Aldrich) to a final concentration of 0.1% was added to the PBS, insulin or coated insulin NanoRod samples to aid in the visualization of the microinjection process and the determination of successfully injected larvae. Zebrafish larvae were randomly assigned to the treatment groups. Free glucose levels were measured as described elsewhere⁴¹ using a fluorescence-based enzymatic kit (BioVision). Groups of three to six injected larvae were used per condition/replicate.

Confocal imaging

Tg(ins:CFP–NTR);Tg(ins:Kaede)-ablated larvae were collected 24 h after ablation treatment, anaesthetized and injected following the previously stated protocol and fixed in 4% paraformaldehyde solution before analysing the β-cell numbers by confocal imaging. The confocal images were acquired with a Leica TCS SP8 microscope and the LAS X software (v. 3.5.5.19976). The primary pancreatic islets of the β-cell-ablated Tg(ins:CFP–NTR);Tg(ins:Kaede) larvae were scanned with a ×40 water-immersion objective and the z stacks were analysed using Fiji software (v1.53). All the displayed images were acquired from the same experiment and their contrast values were adjusted for visualization purposes. The quantification of β-cells was performed on original unmodified images.

Statistical analysis

No statistical methods were used to predetermine the sample sizes, but the sample sizes were similar to those reported in previous publications^28,38,42. Cell culture samples and animals were randomly assigned to the control and treatment groups. Data collection and analysis were not performed blind to the conditions of the experiments. Individual data points are plotted for most graphs. Sample size (n) of the number of experimental biological repeats and the statistical methods used are indicated in the corresponding figure legends. Datasets were tested for Gaussian distribution followed by the appropriate statistical test. Statistical analysis and graphical representation of the data were processed with GraphPad Prism 9.4.0. Two-tailed Mann–Whitney test was performed to compare the cluster properties between the control and insulin-treated cells. For western blot quantifications, one-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparisons test was carried out. For the quantification of β-cells, the Kruskal–Wallis test followed by Dunn’s multiple comparisons test was carried out. The analysis of free glucose values was performed using one-way ANOVA with Tukey’s multiple comparisons test.

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/s41565-023-01507-y