All key resources are summarized in Table 1.
Ethics approval
All animal experiments were performed according to all relevant ethical regulations and were approved by the Division of Veterinary Resources and the Institutional Animal Care and Use Committee of the University of Miami IAUC protocol 19-068.
Human islets and samples from male and female cadaveric donors were acquired through commercial vendors and/or tissues banks with no information linking them to the donors. According to regulation 45 CFR 46.of the U.S. Department of health and human services (https://www.hhs.gov) defining a human subject as “a living individual about whom an investigator (whether professional or student) conducting research” with emphasis given on the living, any research involving cadavers, autopsy material, or biospecimens from now deceased individuals does not meet the regulatory definition of “human subject research” and since no protected health information were used in the study, no IRB oversight nor approval was required for this study.
HT-“Cluster-cell” SELEX and Toggle SELEX
The random DNA library (Supplementary Table 3), previously described by the Sullenger group91, was amplified by PCR using recombinant Taq (Invitrogen) with the Sul5′ and the Sul3′-short primers (Supplementary Table 3) with the following cycling condition: 95 °C 5′, 3× (94 °C 30″, 52 °C 20″, 72 °C 25″), 20× (94 °C 30″, 54 °C 20″, 72 °C 25″), 72 °C 5′. The amplicons were purified using the PCR purification kit (Qiagen) and transcribed in vitro by Durascribe T7 RNA synthesis kit (Lucigen, USA). The resulting 2′Fluoro-RNA aptamers were purified using the RNeasy kit (Qiagen). In the initial selection round, 200 pmoles of RNA were suspended in 450 μl PBS (pH 7.4, Life technologies, MO, USA), heated at 65 °C for 5′ and then cooled at room temperature (25 °C—RT) for 10′. In each cycle of selection, the RNA aptamer library (200 picomoles in cycles 1–2 and 100 picomoles in subsequent cycles) derived from the initial random library or the previous selection cycle (cycles 2–8) was first depleted of non-specific aptamers by a 15′ incubation at RT in 1 ml of PBS supplemented with yeast RNA (10 µg/ml, Ambion) on a rotator with dissociated exocrine tissue (~106 cells) previously depleted from islets by hand. The suspension was spun down (10′ @300 × g 4 °C), and the supernatant passed through a 0.2 µm PES filter. The filtered supernatant containing the unbound RNA aptamers was then used to resuspend handpicked islets (~200 IEQ in cycles 1–4 and ~100 IEQ in cycles 5–8) and incubated for 10′–20′ in rotation at RT. The preparation was then washed 3–6 times with 1 ml of PBS and yeast RNA (10 µg/ml) and centrifuged (10′, @300 × g, 4 °C) and islet pellet lysed with Trizol (Thermo fisher Scientific). Recovered RNA aptamers were then cleaned with RNAeasy and reverse transcribed using Sul3′-short primer and the SuperScript® III Reverse Transcriptase. In the HT “cluster-cell” SELEX, we performed eight selection cycles using islet and acinar tissues from 4 different cadaveric donors. Stringency was gradually increased by: (a) limiting the quantity of RNA aptamer used for the selection to 100 picomoles starting from cycle 3, (b) reducing the number of islets to 100 IEQ in cycle 5–8, (c) reducing the incubation time with the islets to 10′ from cycle 6, and by increasing the number of washes after the positive selection as follow: 3 in cycle 1–3, 4 in cycles 5 and 6, and 6 in cycle 7 and 8. Additionally, to minimize PCR artifacts, PCR cycles were reduced to 15 cycles from cycles 4–8. A similar strategy was followed with the toggle SELEX using islets and acinar tissues from BALB/c mice in the first eight cycles (Cycles M1–M8) and islets and acinar tissue from one cadaveric donor in the last two cycles (Cycles M8H1 and M8H2).
Preparation of libraries for HT-sequencing
cDNA from cycles 1–8 in the HT “cluster-cell” and cycles M8, M8H1, and M8H2 from the toggle cell SELEX were sampled for HT-sequencing by tagging aptamers’ constant region with two sequential PCRs. The first PCR reactions were performed in 100 µl of water containing 1X PCR buffer, MgCl2 solution (1.5 mM), dNTPs (200 µM each), DNA template (5 ng/µl), recombinant Taq polymerase (5 U, Invitrogen), and the PFA and PRA primers (Supplementary Table 3) corresponding to each cycle described above. The reactions were performed in the GS482 thermocycler (G-STORM) using the following program: 95 °C 5′, 5× (95 °C 1′, 56 °C 30″, 72 °C 30″), 72 °C 10′. PCR was purified via gel extraction using the QIAquick Gel Extraction Kit (QIAGEN) following manufacturer instructions. The second PCR was performed using the same condition described above but using the UFB and PRB primers (Supplementary Table 3) and the following cycles: 95 °C 5′, 6× (95 °C 30″, 65 °C 30″, 72 °C 30″, 72 °C 10′. Products were purified by gel extraction; quality was evaluated via bio-analyzer (Agilent). Library quantitation and pooling took place at the Hussman Institute for Human Genomics-Center for Genome Technology using the KAPA Library Quantification Kit for Illumina platforms (part# KK4854). 10-13 pM of pooled samples were loaded on the Illumina cBot for cluster generation according to the manufacturer’s recommendations. Sequencing was performed on an Illumina HiSeq 2000/2500 (HCS 2.0.12.0) using the reagents provided in the Illumina TruSeq PE Cluster Kit v3 and the TruSeq SBS Kit-HS (200 cycles) kit. Data processing was done using HiSeq’s Real-Time Analysis (RTA) from Casava software. Base-calling files were transformed into zipped FASTQ files containing raw reads with base qualities. These raw read files were then filtered by Illumina’s internal filter resulting in 2 FASTQ files (1 per read) containing all pass-filter reads. FASTQ files were used as input for frequency evaluation and for Clustal-Ω and APTANI analyses25,26. HT-sequencing raw and processed data from SELEX experiments are available in GEO as GSE197262.
Bioinformatic analysis of aptamer libraries
FastQ files were processed to identify the different aptamers. Briefly, the aptamers’ variable region (40–44 nucleotides) was determined using the surrounding constant regions in 5′ and 3′. Clustering (Clustal-Ω v 1.1.1.2013.05.31) was performed on cycle 8 of human “cluster-Cell” SELEX on the 7158 sequences whose frequency was higher than 10−6, resulting in 208 families of which only 31 (Supplementary Table 1) had a cumulative frequency (i.e., the sum of all individual sequence frequency within the family) higher than 10−4 and accounting for 66.7% of the sequences in the library. The most frequent aptamer in each of these 31 families was selected and compared by Clustal Ω with those from the toggle SELEX.
Thirty-nine aptamers from toggle-SELEX (Supplementary Table 2) were selected by: (i) selecting those that were undetected in the M8 library and had a frequency higher than 10−4 in MH2, (ii) filtering those whose frequency was higher than 5 × 10−6 in the M8 library, higher than 106 in MH2, and expand more than ten times after the human selections, and (iii) choosing the one with a frequency higher than 104 in M8 and at least doubling in frequency after the human selection.
Clustal Ω analyzed the resulting 80 aptamers from the two selection strategies, and 15 of those were chosen considering the relative similarity in the cluster analysis and convergent selection using the two different selection strategies as representative for each cluster branch for empirical validation.
Library complexity among cycles was calculated as 100× (number of unique sequences)/(total reads). PCA analysis was performed using Jalview with the “DNA” default parameters92 on the top 5000 sequences aligned with Clustal Ω.
Tissue staining
Cy3 aptamers
Aptamers were labeled with cy3 using the Silencer siRNA Labeling Kit—cy3 (Ambion). Fresh frozen tissues and tissue arrays (AMSBIO) were fixed in 10% neutral-buffered formalin (BDH) for 15′ at RT, incubated with dextran sulfate sodium/PBS (1:2 m/V—Pharmacia Biotech) for 30′ and washed with PBS. Then, tissues were stained with cy3-labeled aptamer (10 µg/ml) in PBS for 30′. In the cold target inhibition experiments, the putative protein was added at the indicated molar ratio before staining to cy3-labeled aptamers (60 nM). The aptamer-stained slides were masked with blocking buffer (2% BSA and 10% FBS in PBS) and counterstained with anti-glucagon (Cell Signaling- 1:300 dilution) and anti-insulin (DAKO 1:300) primary antibodies overnight at 4 °C. The slides were washed three times with PBS for 10′ at RT, stained with AlexaFluor-647 anti-guinea pig and AlexaFluor-488 anti-rabbit secondary antibodies for 90′ at RT, washed three times (10′ at RT in PBS), and counterstained with DAPI (10′ at RT followed by two washes with PBS). Sections were imaged using a fluorescence microscope (Zeiss Apotome or Keyence BZ-X800), the Hi-throughput VS120 Olympus slide scanner, or the Leica SP5 confocal microscope.
Biotinylated aptamers
Frozen tissues were fixed for 20′ in neutral-buffered formalin solution, washed three times with PBS, treated with Image-IT (Invitrogen) for 30′ to remove background fluorescence, washed with PBS, and blocked with superblock (Invitrogen)-BSA3%-Tween 20 0.05% for 1 h at RT. Specimen were rinsed in PBS, and endogenous biotins were blocked using the endogenous biotin blocking kit (Invitrogen). Tissues were dextran sulfate sodium/PBS (1:2 m/V—Pharmacia Biotech) for 30′ at RT, rinsed with PBS, and incubated with biotinylated aptamers (0.5 µM) in the presence of yeast RNA (Ambion) 1 mg/ml for 30′ at RT. Tissues were washed four times with PBS (5′ RT) and counterstained with guinea-pig anti-human insulin (4C O/N). Tissues were washed four times with PBS Tween 20 (10′ wash at RT), rinsed with PBS, and stained for 1 h at RT with AF594-goat anti-guinea-pig antibody, AF488 mouse anti-human glucagon, and AF647-streptavidin. Sections were washed three times with PBS-tween 20 (0.5%), once in PBS, mounted with Fluorogel + DAPI with the coverslip, and imaged on Keyence fluorescence microscopy.
Cell profiler
Tiff files from whole slide scan or individual pictures were converted in grayscale and optimized with ImageJ using CLAHE93 to reduce eventual illumination artifacts using for DAPI block size 200 and maximum slope 3.5, and for insulin, glucagon, and aptamers channel block size of 20 and maximum slope of 1.1. Images were then sliced into 600 × 600 pixel images using ImageSlicer (https://www.coolutils.com/TotalImageSlicer), each identifiable as metadata for column and row. Sliced images were loaded in the cell-profiler software. Nuclei were identified as the primary object using the blue (DAPI) channel by setting the diameter of the nuclei between 2 and 10 pixels, using the three classes Otsu Adaptive threshold method with a correction factor of 1, and the lower and upper bounds on threshold 0.1–1.0. Clumped objects were distinguished by shape, the size of the smoothing filter and the minimum allowed distance between local maxima were automatically calculated. Secondary objects (i.e., cells) were identified using the autofluorescence and fluorescence of the merged image from the channels acquired using the nuclei propagation method with three classes Otzu Adaptive threshold method, 0.9 as threshold correction factor (0.0–1.0 range) and 0.02 as regularization factor. The cytoplasm as the tertiary object was identified as the area included in the cells (secondary object) but not in the nuclei (primary object). For each cell, the integrated intensity means of the channels in the nuclei, cells, or cytoplasm were exported as a “cpout” file. Cpout files were analyzed using FCS Express 7 PLUS. Examples of segmentation and analysis are provided in Supplementary Fig. 5.
Cell culture
Human islets and acinar tissues were isolated by the human islets GMP isolation core at the Diabetes Research Institute maintained in PIM (R) media (PRODO) and used between 24 and 48 h after isolation. Only islets with more than 90% viability and purity have been used for selection. Mouse islets and exocrine tissues were purified by the small animal core at the Diabetes Research Institute from 5 to 6 pancreas of BALB/c mice. Islets and acinar tissues were maintained in PIM (R) media (5.8 mM glucose) and used within 48 h from isolation. A549 cells, a human lung cancer cell line isolated from a male patient94, were acquired by ATCC and maintained in RPMI 1640 media (10% FBS, 1% Pen/Strep—Invitrogen Thermo Fisher) in 5% CO2 at 37 °C.
In vitro islet assays
Islets (500 IEQ) were cultured in 24-well plates in 1 ml of PIM (R) media (5.8 mM glucose) with RNA aptamers (2.5 nMoles) or scrambled aptamer for 24 h at 37C 5%CO2. Islets were then washed and assessed for viability function. Islets viability was assessed following the recommendation of the IDT consortium using the Fluorescein Diacetate/ Propidium Iodide (FDA)/(PI) Viability Assay. 20–30 fields with ×4 magnification each containing 5–20 islets were acquired using a “revolve microscope” and images were analyzed with ImageJ to determine “red” and “green” area and calculate the viability index as follow: Viability index = (green area)/(green area + red area) × 100.
Islet functionality was evaluated by perifusion (dynamic GSIS) experiments as previously described95. Briefly, we used a PERI4-02 machine (Biorep Technologies, Miami, FL, USA) that allows parallel perifusion for up to six independent channels. For each experiment, 100 human IEQ were handpicked and loaded in Perspex microcolumns between two layers of acrylamide-based microbead slurry (Bio-Gel P-4, Bio-Rad Laboratories, Hercules, CA, USA). Perifusion buffer containing 125 mM NaCl, 5.9 mM KCl, 1.28 mM CaCl2, 1.2 mM MgCl2, 25 mM HEPES, and 0.1% BSA at 37 °C with selected glucose or KCl (25 mM) concentrations was circulated through the columns at a rate of 100 μL/min. After 60 min of washing with low glucose (3 mM) solution for stabilization, islets were stimulated with the following sequence: 8 min of low (3 mM) glucose, 20 min of high (11 mM) glucose, 15 min of low glucose, 10 min of KCl (25 mM), and 10 min of low glucose (3 mM). Samples (100 μL) were collected every minute from the outflow tubing of the columns in an automatic fraction collector designed for a multi-well plate format. Islets and the perifusion solutions were kept at 37 °C in a built-in temperature-controlled chamber while the perifusate in the collecting plate was kept at <4 °C to preserve the integrity of the analytes. Insulin concentrations were determined by ELISA (Mercodia Inc., Winston Salem, NC, USA). Data were normalized on the DNA contents (evaluated using the dsDNA Picogreen kit, Thermo fisher) of the islets recovered after perifusion.
Challenge of islets with inflammatory cytokines
Human islets (500 IEQ) in 24-well plates 2 ml of PIM(R) media were transfected with the aptamer chimeras 24 h after isolation. Forty-eight hours later, each well was split, and half of the wells were treated with TNF-α, IL-1β, and IFN-γ (1000 U/ml each) (Peprotech, Rocky Hill, NJ), and half were left untreated. Twenty-four hours after the challenge, islets were dissociated and analyzed by flow cytometry.
Flow cytometry
Islet cultures were spun down at 250 × g for 6 min. The supernatant was decanted, and islets incubated with 400 µl of 4 °C trypsin for 5–10 min′. The reaction was quenched with 20% FBS containing RPMI. Islets were passed five times through a 5/8 26G needle, and cells spun at 500 × g for 6.5 min and washed with PBS. Cells were stained with the fixable Live Dead dye (Live/dead yellow, ThermoFisher) for 20 min at RT, washed, and incubated for 30 min at 4 °C with cy3-labeled RNA aptamers (38 pmoles/250 IEQ). The cell suspension was then washed once in PBS, spun at 550 × g for 6.5 min, permeabilized and fixed with the 1 Perm/fix solution (BD Bioscience) for 20 min at 4 °C, washed twice with 1× perm/wash buffer, and stained with anti-insulin and anti-glucagon antibodies for 30 min at 4 °C. MIN6 cells were detached with trypsin and incubated for 30′ at 4 C with either aptamer-chimera hybridized to Cy5 guide RNA or biotinylated aptamer complexed with streptavidin-AF647 in HBBS. Samples were washed once again with perm/wash buffer and once with PBS, resuspended in 300 µl of PBS, and analyzed on an LSR2 flow cytometer equipped with 405, 488, 532, and 635 nm lasers (BD Bioscience) or the Cytoflex 18. Data were acquired with FACS DIVA v 8.0.1 or Cytexpert v2.3 software and analyzed using FCS v6 and v7 express plus software (Denovo-Software).
Aptamer-streptavidin conjugation
Aptamers biotinylated at the 3′end were synthesized by oligofactory and added (at a 4 to 1 molar ratio) to the streptavidin conjugated with AlexaFluor-647 (Biolegend) or with AlexaFluor-750 (Thermo fisher) gradually (1/6 of the volume added every 5 min). The mixture was incubated ON at RT in rotation in the dark, concentrated with an Amicon Ultra-4 centrifugation filter (50 kDa, Millipore), washed twice with 1 ml PBS, and brought to the 62.5 μM. The aptamer-streptavidin complexes were heated to 65 °C for 10′ and then allowed to cool to room temperature for at least 10 min before. Effective conjugation was evaluated by EMSA on a 2% agarose gel.
Selection of saRNAxiap
Bioinformatics screening of the XIAP promoter was screened with the algorithm developed by Wang et al.50. Briefly, the 1100 bp region between nucleotide 123858524-123859624 of Chromosome X upstream of XIAP transcriptional starting site (−1200 to −100 bp) was retrieved from the ensemble GCA_000001405.28 from the Ensembl Genome Browser and fed to the excel macro developed by Wang et al.50. The candidate sequences with Wang’s score ≥4 were screened with Blat96 using default parameters and Blast (with filtered E value <0.5 and >90% identity) to remove those sequences with more than one hit in the genome. This resulted in seventy-five 19-nucleotide long sequences that were synthesized (Sigma-Aldrich) as double-stranded saRNA. All sequences were synthesized with 3′dTdT overhang for empirical validation on A594 cells. In particular, 104 A549 cells cultured O/N at 37 °C in complete media in flat bottom 96 well plate (104 cells/well) and transfected with saRNA (100 nM) using lipofectamine 3000 (Thermo Fisher) following the manufacturer’s instruction. Media was changed 30 min and 2 days after transfection. Ninety-six hours after transfection, RNA was isolated with TRIzol, and XIAP expression was quantified by qRT-PCR (Applied Biosystems) and normalized to 18 S expression. The effect size was calculated from triplicates with the following formula: effect size = (2−ΔCTexp − 2−ΔCTctrl)/SDexp.
Aptamer-si/saRNA chimera transfection and fluorescent probes
Aptamer chimeras were generated as previously described28. Briefly, aptamer-passenger strand conjugates were produced by PCR and T7RNA polymerase using the Sul5′ and the Sul3′ primers elongated in 5′ with the relevant passenger sequence (Supplementary Table 3). The guide strand RNA sequences were then annealed at an equimolar ratio in a thermocycler using the following conditions: 70 °C for 10 min, cooling to 25 °C at 0.1 °C/s. Annealing was confirmed by gel shift electrophoresis on a 3% agarose gel. Non-dissociated islets (250 IEQ) were transfected by adding the relevant or scrambled si/saRNA-aptamer chimera (150picomoles/100 IEQ for siRNA and 200 picomoles/100 IEQ for saRNA) to human islets in PIM (R) media. Clusterin siRNA was previously validated97, whereas TMED6 siRNA was selected from the first 200 nt of the TMED6 mRNA (NM_144676.4) using the Invivogen siRNA wizard and validated in vitro. When used as an immunofluorescence probe, aptamer-chimera was annealed to the complementary RNA strand labeled in 5′ with Cy5.
qRT-PCR
TMED6 and clusterin expression was evaluated by qRT-PCR using 200 ng of human islets total RNA extracted with Trizol. Data were normalized on 18S expression (Applied biosystem) and expressed as follows: expression = 2−(CTexp-CT18S) × 1000).
Protein arrays
HuProt 2.0 arrays (ArrayIt) were prepared for binding as recommended by the manufacturer. The protein surface of the array was deactivated with 3 ml of “Chem block” buffer (Arrayit) for 1 h at RT. After deactivation, the array was blocked with blocking buffer (Arrayit) for 1 h and then hybridized for 90 min at 4 °C with the cy3-labeled aptamer (40 picomoles) in 100 µl of blocking buffer supplemented with BSA (2%, Sigma) and yeast RNA (0.1 mg/ml, Ambion). After hybridization, the arrays were washed (5 times × 5 min) in PBS, dried using the Microarray High-Speed Centrifuge (MHC—ArrayIt), and images were acquired with the Genepix 4000B microarray reader and acquired with GenePix Pro Microarray Analysis Software (Molecular Devices).
Protein array analysis
The raw intensity signals were extracted from the gpr files using the PAA package (v. 1.0)98 in R 3.4.4; before preprocessing, each microarray was visually inspected for any artifact using the plotArray function. Then, the median fluorescence intensities at 635 nm wavelength were inputted as raw intensity foreground signals; specifically, intensity signals of replicated spots were mean-centered, and background corrected using default parameters, e.g., the saddle variant of the normexpr method in the backgroundCorrect function, and were finally normalized using the quantile method of the normalizeArrays function. Not annotated spots, e.g., without both gene Symbol and Refseq annotations (n.1,975), were excluded from following analysis. Statistically significant differences in aptamer binding to the proteins have been defined on the logged (log2) binding values using the t test method in the diffAnalysis function and considered significant if the p value ≤0.05 and the absolute fold change ≥2. Raw data are available in GEO as GSE162273.
Gene expression analysis
Gene expression data was recovered from the GEO datasets GSE2109 and GSE15543, both for the whole pancreas tissues (GSM53046, GSM325790, GSM325838, GSM277701, GSM277726, GSM277736, GSM231922, GSM203675, GSM203703, GSM203761, GSM179781, GSM179869, GSM152744, GSM137958, GSM117645, GSM117647, GSM89045) and for the islets cells (GSM388749, GSM388750, GSM388753, GSM388754, GSM388759, GSM388760, GSM388766, GSM388767) using the GEOquery package (v. 2.46.15)99 in R (v 3.4.4). The raw intensity signals were extracted from CEL files and normalized using the justRMA function of the affy package (v. 1.56)100. Fluorescence intensities were background-adjusted and normalized using the quantile normalization; afterward, log2 expression values were calculated using median polish summarization and custom chip definition files for Human U133 plus2 array based on Entrez genes (HGU133Plus2_Hs_ENTREZG version 20.0.0; 19,363 unique genes) from the Brain Array webpage101. Statistically significant differences in gene expression were determined using the moderated t test in the limma package (v. 3.34.9)102; a gene was defined significant if with absolute fold change ≥two and adjusted p value ≤0.05 after multiple testing corrections with the Benjamini and Hochberg method, e.g., FDR.
Aptamer-mediated immunoprecipitation and mass spectrometry
Human islets (1000 IEQ) were dissociated for 5 min with 0.025% Trypsin EDTA and washed three times with PBS-1%BSA (2 ml, 514 g 6 min). The cell suspensions were then incubated with biotinylated aptamers (100 pmoles) in binding buffer (5 mM Mg2+ PBS with masking oligo 1–3 at 5 µM each—Supplementary Table 3) for 30 min at 4 °C. Cells were washed three times with PBS-BSA1%. To separate membranes from intracellular components, cells were incubated in a mild hypotonic lysis buffer containing 10 mM Tris–HCl, 50 mM NaCl, 500 µM MgCl, 1 µM DTT, protease inhibitor cocktail, masking oligo (5 µM), and yeast RNA (4 mg/ml) for 2 min on ice. Immediately after incubation, cells were gently homogenized in a Dounce homogenizer, ten times on ice, mixed with streptavidin-conjugated magnetic beads (Dynabeads, Thermo Fisher) in binding buffer for 1 h at RT in rotation. Beads were magnetically recovered, washed with 1 ml of PBS, resuspended in 30 µl of Laemmli buffer, and captured proteins run on a 4–20% gradient SDS-PAGE. Gels were stained using Brillant Coomassie Blue (Pierce). Bands from the m12-3773 immune precipitate and corresponding area (i.e., same weight) from the scrambled controls were cut out and sequenced by microcapillary LS/MS/M and analyzed by Mascot software at the University of Kentucky Mass Spectrometry Facility (Supplementary Table 5).
Mice
All animal experiments were approved by the Division of Veterinary Resources and the Institutional Animal Care & Use Committee of the University of Miami. Eight to ten weeks old male NOD.Cg-Prkdcscid Il2rgtm1Sug/JicTac (NOG) (Taconic), female Balb/C, and C57Bl/6J mice (Jackson) were purchased, allowed free access to food and water, and were maintained on a 12-h light/dark cycle at room temperature (range 20–23 °C) and controlled humidity (range 30–70%) in individually ventilated cages at the pathogen-free animal facilities at the University of Miami on a chlorophyll free diet. Mice were allowed to acclimate for at least 1 week before starting any experiments. NOG mice were transplanted with variable numbers (150–500 IEQ) of human islets in the EFP103 or under the kidney capsule. Transplants were performed by the small animal core at the Diabetes Research Institute. Unless otherwise stated, mice were treated with the aptamer at least 21 days after transplant. BALB/c mice were transplanted with B6 and Balb/c islets (250 IEQ each) dorsally in the flank using a thrombin-plasma biological scaffold as described104.
IVIS analysis
Isofluorane-anesthetized, islet-transplanted mice were analyzed by the In Vivo Imaging System (Xenogen IVIS Spectrum Perkin Elmer) 4 h after i.v. injection of 12.5 pmol/g of aptamers conjugated with either streptavidin-Alexa Fluor-750 or streptavidin-Alexa Fluor-647. Fluorescence was quantified with the Living Image v4.3 software (Perkin Elmer). The signal to background ratio was calculated by evaluating the radiance in the region of interest (i.e., EFP or liver area) over an ROI of the same size drawn on the lungs region after subtracting the autofluorescence signal from non-injected mice.
Marginal mass islet transplantation experiments
Diabetes was induced in 10–16 weeks old immunodeficient NOG female mice (Taconic, Rensselaer, NY) with 5 low doses of streptozotocin (50 µg/g, i.p., q.d.). Mice were kept euglycemic with a subcutaneous insulin pellet until human islets were available (~5–20 days). Forty-eight hours before the scheduled transplantation, the insulin pellet was removed. Blood glucose was monitored, and only hyperglycemic mice were used for the experiments. On the day of the transplantation, mice were anesthetized and transplanted by the DRI small animal core under the kidney capsule as previously described105 with 500 IEQ of human islets treated with aptamer chimeras or left untreated. To evaluate the quality of the preparation, 1–2 mice were transplanted with 1200 IEQ from each preparation as positive controls. Blood glucose was monitored by venipuncture three times a week.
Statistical analysis
Sigmaplot 12.5 (Systat Software) was used for data analysis. Statistical tests (one-way ANOVA followed by Holm–Sidak test for multiple pairwise comparisons or student T test) were applied as indicated in the figure legends in a two-sided, unpaired fashion after normality was evaluated by the Shapiro–Wilk test. In vitro analyses and in vivo experiments were repeated two to five times to ensure reproducible conclusions; the exact number of repetitions is stated in each figure legend. Log-rank test was used for survival analysis followed by all pairwise multiple comparison procedures (Holm–Sidak method). Data from multiple experiments were cumulated unless otherwise indicated in the figure legends. No experimental data point was excluded from the analyses. The sample size was chosen by power analysis using effect size determined by pilot experiments.
Reporting summary
Further information on experimental design is available in the Nature Research Reporting Summary linked to this paper.
