Research ArticleBIOCHEMISTRY

Sub–100-nm metafluorophores with digitally tunable optical properties self-assembled from DNA

See allHide authors and affiliations

Science Advances  21 Jun 2017:
Vol. 3, no. 6, e1602128
DOI: 10.1126/sciadv.1602128
  • Fig. 1 DNA-based metafluorophores.

    (A) Labeling pattern for DNA origami–based metafluorophores. Cylinders represent DNA double helices. Selected strands are extended with 21–nucleotide (nt) handles on the 3′ end, which bind complementary fluorescently labeled antihandles. Labeling patterns are represented as pictograms, where each colored dot represents a dye-labeled handle. (B to D) Fluorescence intensities increase linearly with the number of dyes attached to a metafluorophore (here, up to 132 dyes per structure). Insets show diffraction-limited fluorescence images of metafluorophores and the corresponding labeling pattern. Image sizes, 1.2 × 1.2 μm2. (E to G) Metafluorophores allow dense labeling (~5-nm dye-to-dye distance) without self-quenching. Pictograms illustrate dense and sparse labeling patterns for 14 dyes. Corresponding intensity distributions of the two patterns overlap for each color, showing no significant change in intensity. a.u., arbitrary units.

  • Fig. 2 Multicolor metafluorophores.

    (A to C) “Randomly” labeled metafluorophores may result in significant decrease in fluorescence intensity (B and C) due to FRET, when labeled with spectrally distinct dyes. Metafluorophores with only 44 dyes of the same color serve as references (gray distributions). If Atto 647N, Cy3, and Atto 488 are all present on the same structure (44 dyes each), the intensity distributions (colored) for Cy3 (B) and Atto 488 (C) are significantly shifted to lower values. However, this fluorophore arrangement does not provide an acceptor for Atto 647N fluorescence; thus, its intensity distribution is not altered (A). Pictograms illustrate labeling patterns. (D to F) Column-like metafluorophore labeling pattern prevents energy transfer (FRET). Metafluorophores labeled with 44 dyes of one species (gray) show identical intensity distributions as structures labeled with all three species (colored). Pictograms illustrate labeling patterns.

  • Fig. 3 Metafluorophores for intensity barcoding.

    (A) Intensity distributions for Atto 488 (red, 6 dyes per structure; purple, 14 dyes; blue, 27 dyes; green, 44 dyes). Nonoverlapping intensity distributions can be achieved by the precise control over the number of dyes per metafluorophore. (B) Fluorescence image of 124 distinct metafluorophores deposited on a glass surface. Scale bar, 5 μm. (C) Matrix of representative fluorescence images of 124 distinct metafluorophores. (D) Metafluorophore-based intensity barcodes (124) in one sample. A total of 5139 barcodes were recorded, and all 124 barcode types were detected. (E) Subset of 25 of 124 barcodes. A total of 2155 barcodes were recorded, where 86.5% were qualified barcodes and 87.4% thereof were expected barcodes. (F) Subset of 12 of 64 barcodes. All barcodes have all three fluorophore species, making their detection more robust. A total of 521 barcodes were recorded, where 92.5% were qualified and 95.4% thereof were expected barcodes. (G) Subset of 5 of 20 barcodes. A total of 664 barcodes were recorded, where 100% were qualified and 99.6% thereof were expected barcodes.

  • Fig. 4 Quantitative nucleic acid detection.

    (A and B) Schematic of the hybridization reaction. A metafluorophore is programmed to hybridize to part (green) of a specific nucleic acid target. A biotinylated capture strand binds to a second region (red) and thus immobilizes the complex on a streptavidin-coated surface, yielding fluorescence images comparable to Fig. 3B. Each positively identified metafluorophore indicates a single copy of a nucleic acid target. (C) The number of detected targets is directly proportional to their concentration in the sample of interest. Targets were added with defined concentrations (blue bars) and subsequently identified in the expected ratios (green bars). Five different field of views have been recorded, enabling the calculation of error bars. The lowest target concentration [target 3 (t3) and target 4 (t4)] was 1.5 pM. Target 1 (t1) and target 2 (t2) were not added, and corresponding identifications are due to nonspecific binding of metafluorophores to the surface and potential false-positive identifications.

  • Fig. 5 Metafluorophores with different photostability.

    (A) Time-lapsed fluorescence micrographs of a sample composed of two spectrally indistinct metafluorophore species: one containing 44 Atto 647N dyes (more photostable) and one containing 44 Alexa 647 dyes (less photostable). Images were acquired at t1 = 0 s, t2 = 20 s, and t3 = 40 s, with an integration time of 10 s, whereas the sample was constantly illuminated during acquisition. The time-lapsed micrographs show two species where one bleaches faster than the other. The two species can be visually identified by superimposing the images taken at t1 (false color blue) and t3 (false color yellow). The metafluorophore containing more photostable dyes (that is, Atto 647N) appears green (blue + yellow), whereas the one with the less photostable dyes (that is, Alexa 647) appears blue. The fluorescence decay constant can be used as a parameter to quantitatively describe the photostability. The decay constant is obtained by fitting a single exponential decay to the intensity versus time trace. Scale bars, 5 μm. (B) Intensity versus decay constant histograms for three different metafluorophore samples containing Atto 647N dyes (green), Alexa 647 dyes (blue), and both dyes (red), respectively (note that only one species was present in each sample). (C) 1D histogram of the decay constants shows three distinguishable decay constant distributions (schematics in the legend show the dye arrangement on the metafluorophores).

  • Fig. 6 Triggered assembly of metafluorophores.

    (A) Schematic of triggered assembly of triangular metafluorophores constructed from 10 metastable Cy3-labeled DNA hairpin strands. A so-called capture strand (labeled with Alexa 647) is attached to a glass surface via biotin-streptavidin coupling. A long “trigger strand” can hybridize to the capture strand. The trigger strand consists of four concatenated domains, “1-A,” where the subdomain “1” is 20 nt long and the subdomain “A” is 12 nt long. Hairpin strands coexist metastably in the absence of the trigger and only assemble into the desired structure upon exposure to the trigger. More specifically, the introduction of a repetitive single-stranded trigger initiates the assembly of kinetically trapped fluorescent hairpin monomers, which produce a second row of binding sites. These binding sites further enable the assembly of successive rows of monomers, with each row containing one fewer monomer than the previous. After assembly of 10 hairpins (labeled with Cy3) to a single trigger strand, no further trigger sequences are displayed, and assembly is terminated, yielding a triangular-shaped metafluorophore of fixed dimensions. (B) Fluorescence images of triangles assembled in situ on a glass surface. The capture strands are labeled with Alexa 647 (red), whereas the hairpins are labeled with Cy3 (green). DNA origami with 10 Cy3 and 44 Atto 488 (blue) dyes were added to the sample as intensity references. DNA origami can be identified at the positions where Atto 488 and Cy3 signals colocalize. In the schematic below the overlay fluorescence image, the blue spot indicates the Atto 488–labeled origami marker; green spots indicate the expected overlay of Alexa 647–labeled capture strand and the triangle composed of Cy3-labeled hairpin monomers. Gray crosses indicate nonspecific binding of hairpins to the surface. Scale bars, 1 μm. (C) Triangular metafluorophores (green) and reference DNA origami (blue) intensity distributions are overlapping, thus indicating the formation of the triangles as expected.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/6/e1602128/DC1

    fig. S1. caDNAno DNA origami design.

    fig. S2. Schematic DNA origami staple layouts of single-color metafluorophores (6 to 132).

    fig. S3. Linear dependence of intensity on the number of dyes per DNA origami (calibrated).

    fig. S4. Intensity distributions for 6 to 132 dyes.

    fig. S5. Excitation power variation.

    fig. S6. Integration time variation.

    fig. S7. Refocusing performance.

    fig. S8. Photostability.

    fig. S9. Schematic DNA origami staple layouts of self-quenching study.

    fig. S10. FRET investigation dye patterning (random and column-wise).

    fig. S11. Intensity barcode dye patterns.

    fig. S12. Intensity distributions for 124 barcodes in one sample.

    fig. S13. Exemplary fluorescent image of nucleic acid detection.

    fig. S14. DNA detection calibration.

    fig. S15. Triggered assembly formation gel assay.

    table S1. DNA origami staple sequences.

    table S2. M13mp18 scaffold sequence.

    table S3. Fluorescently labeled DNA sequences.

    table S4. Intensity barcode subset (25 of 124).

    table S5. Intensity barcode subset (12 of 64).

    table S6. Intensity barcode subset (5 of 20).

    table S7. DNA detection sequences and corresponding barcodes.

    table S8. Triggered assembly sequences.

    protocol S1. DNA origami self-assembly.

    protocol S2. Microscopy sample preparation.

    protocol S3. Triggered assembly on surface.

    protocol S4. Triggered assembly in solution and gel assay.

    Materials

    Optical setup

    Software section

  • Supplementary Materials

    This PDF file includes:

    • fig. S1. caDNAno DNA origami design.
    • fig. S2. Schematic DNA origami staple layouts of single-color metafluorophores (6 to 132).
    • fig. S3. Linear dependence of intensity on the number of dyes per DNA origami (calibrated).
    • fig. S4. Intensity distributions for 6 to 132 dyes.
    • fig. S5. Excitation power variation.
    • fig. S6. Integration time variation.
    • fig. S7. Refocusing performance.
    • fig. S8. Photostability.
    • fig. S9. Schematic DNA origami staple layouts of self-quenching study.
    • fig. S10. FRET investigation dye patterning (random and column-wise).
    • fig. S11. Intensity barcode dye patterns.
    • fig. S12. Intensity distributions for 124 barcodes in one sample.
    • fig. S13. Exemplary fluorescent image of nucleic acid detection.
    • fig. S14. DNA detection calibration.
    • fig. S15. Triggered assembly formation gel assay.
    • table S1. DNA origami staple sequences.
    • table S2. M13mp18 scaffold sequence.
    • table S3. Fluorescently labeled DNA sequences.
    • table S4. Intensity barcode subset (25 of 124).
    • table S5. Intensity barcode subset (12 of 64).
    • table S6. Intensity barcode subset (5 of 20).
    • table S7. DNA detection sequences and corresponding barcodes.
    • table S8. Triggered assembly sequences.
    • protocol S1. DNA origami self-assembly.
    • protocol S2. Microscopy sample preparation.
    • protocol S3. Triggered assembly on surface.
    • protocol S4. Triggered assembly in solution and gel assay.
    • Materials
    • Optical setup
    • Software section

    Download PDF

    Files in this Data Supplement:

Navigate This Article