SUPPLEMENTARY INFORMATION · 2015-11-04 · Supplementary Information High-throughput optical sensing of nucleic acids in a nanopore array Shuo Huang, Mercedes Romero-Ruiz, Oliver

Supplementary Information

High-throughput optical sensing of nucleic acids in a nanopore array

Shuo Huang*, Mercedes Romero-Ruiz*, Oliver Castell, Hagan Bayley & Mark Wallace**

*These authors contributed equally

**Correspondence to: [email protected]

This PDF file includes:

Supplementary Materials

Supplementary Methods

Supplementary Text

Supplementary Table 1 to 7

Supplementary Figures 1 to 17

Captions to Supplementary Videos 1 to 7

Other Supplementary Materials for this manuscript include:

Supplementary Videos 1 to 7

SUPPLEMENTARY INFORMATIONDOI: 10.1038/NNANO.2015.189

NATURE NANOTECHNOLOGY | www.nature.com/naturenanotechnology

© 2015 Macmillan Publishers Limited. All rights reserved

http://dx.doi.org/10.1038/nnano.2015.189

1

Table of Contents

Supplementary Materials ....................................................................................................................... 2

Supplementary Methods ........................................................................................................................ 3

Supplementary Text ............................................................................................................................... 5

Supplementary Table 1 | Nucleic acid abbreviations and sequences. ................................................... 6

Supplementary Table 2 | Statistics of the residual fluorescence of αHL using oSCR. .......................... 6

Supplementary Table 3 | Statistics of the residual fluorescence of MspA using oSCR. ........................ 6

Supplementary Table 4 | Statistics of the residual current of αHL in a BLM. ......................................... 6

Supplementary Table 5 | Statistics of the residual current of MspA in a BLM. ...................................... 6

Supplementary Table 6 | Statistics of the mean duration time for a full miRNA unzipping cycle. .......... 7

Supplementary Table 7 | The rate constants for different probe/miRNA combinations. ........................ 7

Supplementary Figure 1 | The DIB measurement device. ..................................................................... 8

Supplementary Figure 2 | Electrical measurementof αHL pores in aDIB. ............................................. 9

Supplementary Figure 3 |The Current-Voltage (I-V) curve of an αHL pore in a BLM. ......................... 10

Supplementary Figure 4 | Fluorescence-Voltage (F-V) curve of an αHL pore determined by oSCR. . 11

Supplementary Figure 5 | Parallel recording using oSCR in a DIB. ..................................................... 12

Supplementary Figure 6 | The minimum spatial resolution of oSCR. .................................................. 13

Supplementary Figure 7 | BLM measurements of αHL WT with three tethered oligonucleotides. ...... 14

Supplementary Figure 8 | BLM measurements of MspA pores with individual DNA homopolymers. .. 15

Supplementary Figure 9 | Fitting miRNA unzipping traces. ................................................................. 16

Supplementary Figure 10 | miRNA unzipping of various probe/miRNA pairs. ..................................... 17

Supplementary Figure 11 | Pore clogging by strongly hybridized probe/miRNA pairs. ....................... 18

Supplementary Figure 12 | HHBa measurement chamber. ................................................................. 19

Supplementary Figure 13 | Hydrogel chip fabrication procedures. ...................................................... 20

Supplementary Figure 14 | Electrodes for the HHBa chip. .................................................................. 21

Supplementary Figure 15 | Programmed chip loading with a spotting robot. ...................................... 22

Supplementary Figure 16 | Optical single channel recording in a multiplexed HHBa chip. ................. 23

Supplementary Figure 17 | Parallel oSCR recording in full frame. ...................................................... 24

Supplementary Video 1 | Real-time oSCR of nanopore blockades with DNA. .................................... 25

Supplementary Video 2 | High-throughput nanopore tracking and recording. ..................................... 25

Supplementary Video 3 | Parallel DNA detection using MspA. ............................................................ 25

Supplementary Video 4 | miRNA unzipping in parallel. ....................................................................... 25

Supplementary Video 5 | Hydrogel chip loading. ................................................................................. 25

Supplementary Video 6 | Hydrogel bilayer array formation. ................................................................ 25

Supplementary Video 7 | Multi-sample recording in the hydrogel array. ............................................. 25

Supplementary References ................................................................................................................. 26


2

Supplementary Materials The lipid used in this paper is 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC) (Avanti Polar Lipids). The oil is a 1:1 (v:v) mixture of hexadecane (Sigma-Aldrich) and silicone oil AR20 (Sigma-Aldrich).To dissolve the lipid in the oil, the DPhPC powder is first dissolved in pentane (Sigma-Aldrich) in a 7 mL glass vial. It is then air dried with nitrogen gas to form a thin film of lipid on the inner wall of the vial. The lipid film is treated in a desiccator for more than 4 h to remove the residual pentane. Finally the lipid film is dissolved in the oil.

Ethylenediaminetetraacetic acid (EDTA) (Sigma-Aldrich), agarose for routine use (Sigma-Aldrich), low melting point agarose (Sigma-Aldrich), 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (Sigma-Aldrich), potassium chloride (Sigma-Aldrich), calcium chloride (Sigma-Aldrich), pentane (Sigma-Aldrich), SU-8 2035 photoresist (MicroChem), poly-methylmethacrylate (PMMA 495 a5) (MicroChem), poly-dimethylsiloxane (PDMS, Sylgard 184) (Dow Corning), fluo-8 (ABD Bioquest), chelex (BioRad chelex 100 Resin, Biotechnology Grade, 100-200 mesh), and streptavidin (New England Biolabs)are used as received without further purification.

HPLC-purified DNA (ATDbio and Sigma Aldrich) and RNA (IDT) samples are purchased and used without further purification.

The protein nanopores used in this paper are αHL WT and MspA NNNRRK or M2. The αHL WT heptamer protein is expressed in E. coli and purified based on the published protocols1. Mutant MspA NNNRRK or M22 are produced by expression in an E. coli in vitro transcription/translation system (E.coli T7 S30 Extract system for Circular DNA) and oligomerized into heptameric pores on rabbit red blood cell membranes3.

Potassium chloride buffer (1.5 M KCl, 10 mM HEPES, pH 7.0 [αHL] / pH 8.0 [MspA]) and the calcium chloride buffer (0.75 M CaCl2, 10 mM HEPES, pH 7.0 [αHL] / pH 8.0 [MspA]) are prepared and membrane-filtered (0.2 µm cellulose acetate, Nalgene) prior to use. KCl buffer is treated with Chelex 100 resin overnight before use.

DNA samples are dissolved in DNase/RNase free water prior to use. Streptavidin and biotinylated-ssDNA are mixed at a 1:1 molar ratio in potassium chloride buffer and incubated for 20 min prior to use.


3

Supplementary Methods

1 | miRNA/Probe Annealing

miRNA and DNA probe samples are dissolved in DNase/RNase free water prior to use. miRNA and probe are mixed at a 1:1 molar ratio in potassium chloride buffer. To facilitate hybridization, the mixture is heated to 95 ℃ for 5 min and cooled down to room temperature at a rate of -5 ℃/min in a thermal cycler (Veriti, Life Technologies).

2 | TIRF Microscopy

TIRF measurements are performed using a Nikon Eclipse Ti microscope equipped with an oil immersion objective (60x Plan Apo TIRF, Nikon). Fluorescence is excited at 473 nm using a DPSS laser (Shanghai Dream Laser Technologies) and imaged with an electron-multiplying CCD camera (Ixon3 897, Andor). The field of view is 150 µm by 150 µm.

3 | Fluorescence Trace Extraction and Normalization for αHL WT

Images are analyzed using a custom LabVIEW program. Fluorescence traces are extracted by summing pixel intensities within a 5 µm by 5 µm region around individual spots. The intensity is normalized to the background fluorescence around the pore.

4 | Fluorescence Trace Extraction and Normalization for MspA

For MspA experiments, a histogram-based background subtractor for ImageJ developed by the MOSAIC Group at ETH Zürich (mosaic.mpi-cbg.de/Downloads/BGS_manual.pdf) is used to normalize the original image. A custom IGOR Pro program (Wavemetrics, Lake Oswego, OR, USA) then performs 2D Gaussian fitting for each fluorescence spot. Fluorescence traces are then normalized to the open pore fluorescence at -100 mV and 0 mV.

5| BLM Measurements and Data Analysis

BLM measurements are performed similarly to the method published previously4. Briefly, a lipid (DPhPC) bilayer is formed across a Teflon (Goodfellow, 25 µm thick) aperture (100 ± 5 µm), which separates the cis (electrically grounded) and the tans compartments of the measurement apparatus (0.5 mL or 1mL volume on both sides). To mimic the optical recording measurement, the BLM is recorded under asymmetric buffer conditions (cis: KCl; trans: CaCl2). The ionic current through a single nanopore in the BLM is low-pass filtering at 1 kHz and recorded (Axopatch 200B, Molecular Devices) with a sampling rate of 5 kHz (Digidata 1440A digitizer, Molecular Devices). Streptavidin-tethered ssDNA (267 nM [αHL] or 1.6 µM [MspA]) is added to the cis chamber and the chamber is magnetically stirred. A repeating voltage protocol (100 mV, 0.9 sec; -100 mV [MspA]/-140 mV [αHL], 0.05 sec; 0 mV, 0.05 sec) is applied to accumulate sufficient events for statistical analysis. Data analysis is performed with pClamp software (Molecular Devices) and the event histograms are fit to a Gaussian function (Origin 8.5.1). Three independent experiments are performed for each condition


4

and the standard deviations between independent experiments are calculated (Supplementary Table 4, 5).

6| Photolithography

Photomasks are designed (AutoCAD) and printed on a transparent film (JDphoto). Micro-patterns of SU-8 pillars are fabricated according to the standard photolithography protocols (MicroChem):

1. Spin coating: 1 mL of SU-8 2035 photoresist is spin coated on the 6-inch silicon wafer at 500 rpm for 15 s followed by 2000 rpm for 35 s.

2. Pre-bake: The wafer is baked at 60 ℃ for 2.5 min and 95 ℃ for 7 min.

3. Exposure: The wafer, which is covered with the photomask, is exposed to UV (200 mJ/cm2) for 30 s.

4. Develop: The wafer is then sprayed and rinsed with the developer for 6 min.

5. Wash: The developed wafer is cleaned with isopropanol and air dried in a nitrogen stream.

6. Hard Bake: The wafer is baked at 150 ℃ for 10 min to finalize the lithography process.

The thickness of the fabricated pillar structures is approximately 40 µm according to the calibration profiled (MicroChem). The micro-patterned wafer can be re-used.

7| Soft Lithography

Soft lithography is performed according to published protocols5. Briefly, a PDMS base and curing agent are mixed in a 10:1 volume ratio. The mixture is poured over the micro-pattered silicon wafer pillar array in a petri-dish and degassed for 1 h. The PDMS mixture is then heated to 80 ℃ for 4 h. The cast PDMS elastomer is then peeled from the wafer. This PDMS mould could be re-used (Supplementary Figure 13).

8| Comparison of Electrical and Optical Recording

To compare signal to noise between electrical and optical measurements the mean current (or fluorescence) level for each blocking event was calculated. The signal to noise was calculated as the difference between two adjacent levels (X3 and X5) divided by the standard error of the mean. Values are given in Table S6 and this data is shown in Fig. 2C.


5

Supplementary Text

1 | Pore Densities Estimation (theoretical limit)

Assuming that the pores are assembled into a hexagonal array, each hexagonal unit cell contains 3 nanopores. If the pore to pore distance is d, then the area per pore is:

𝐴𝑟𝑒𝑎/𝑝𝑜𝑟𝑒 =32𝑑! (1)

According to the full width at half maximum (FWHM) of each fluorescence spot (Supplementary Figure 6), the minimum pore to pore separation for independent optical recording is 3 µm. The estimated pore density is:

1.3×10! 𝑝𝑜𝑟𝑒𝑠 ∙𝑚𝑚!!

Based on the published results6, 7 of nanopore sequencing rate (40 nucleotides/s), a sequencing density of:

5.2×10! 𝑛𝑢𝑐𝑙𝑒𝑜𝑡𝑖𝑑𝑒𝑠 ∙𝑚𝑚!! ∙ 𝑠!!

could be achieved with oSCR. In principle, for a human genome (~ 3×10!nucleotides), sequencing, admittedly without additional coverage, could be completed in ~15 min (900 s) within a 1 mm2 sized array.

2| Unzipping Kinetics Modelling

The miRNA unzipping kinetics is modelled similarly as reported before8, 9. As a single step, first-order reaction, the rate constant from state A (hybridized state) to state B (unzipped state) is defined to be 𝑘, which reflects the hybridization strength between the miRNA and the probe. The rate constant 𝑘 is a function of the temperature (T) and the applied potential (V):

𝑘(𝑇,𝑉) = 𝑘!exp (−𝐸! − 𝑞!""𝑉

𝑘!𝑇) (2)

We assume that the unzipping process is driven by a constant electrical force. And the effective charge (q!"") remains constant during the whole unzipping process. Under the applied potential, the effective activation energy (E!) is lowered by q!""V .


6

Supplementary Table 1 | Nucleic acid abbreviations and sequences. “X” in the sequence represents an abasic position.

Abbreviations Nucleic Acid Sequence

X5 3’-Biotin-TEG-CCCCCCCCCCCCCCCCCCCCCCCCXXXXXCCCCCCCCCCC-5’

X3 3’-Biotin-TEG-CCCCCCCCCCCCCCCCCCCCCCCCCXXXCCCCCCCCCCCC-5’

C40 3’-Biotin-TEG-CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC-5’

C65 5’-Biotin-CCCCCCCCCCCC-CCC-C35-CCTGTCTCCCTGCCG-3’

T65 5’-Biotin-TTTTTTTTTTTT-TTT-T35-CCTGTCTCCCTGCCG-3’

A65 5’-Biotin-AAAAAAAAAAAA-AAA-A35-CCTGTCTCCCTGCCG-3’

G3 in A65 5’-Biotin-AAAAAAAAAAAA-GGG-A35-CCTGTCTCCCTGCCG-3’

Plet7a 3’-C30-ACTCCATCATCCAACATATCAA-C30-5’

Plet7i 3’-C30-ACTCCATCATCAAACACGACAA-C30-5’

Let7a 3’-UUGAUAUGUUGGAUGAUGGAGU-5’

Let7i 3’-UUGUCGUGUUUGAUGAUGGAGU-5’

Supplementary Table 2 | Statistics of the residual fluorescence of αHL using oSCR. The fluorescence amplitude is normalized according to the two reference levels (unblocked fluorescence intensity at 0 mV and -50 mV).

Mean Fluorescence Intensity / a.u. Standard Deviation / a.u. Measurements C40 0.39 0.08 132 X3 0.57 0.07 140 X5 0.69 0.07 121

Supplementary Table 3 | Statistics of the residual fluorescence of MspA using oSCR. The fluorescence amplitude is normalized according to the two reference levels (unblocked fluorescence intensity at 0 mV and -100 mV).

Mean Fluorescence Intensity / a.u. Standard Deviation / a.u. Measurements C65 0.34 0.04 49 T65 0.52 0.04 53 A65 0.71 0.07 43 G3 in A65 0.88 0.08 38

Supplementary Table 4 | Statistics of the residual current of αHL in a BLM. The mean residual current and standard deviation are calculated from the Gaussian fitting of a current histogram (Supplementary Figure 7).

Mean Residual Current / pA Standard Deviation / pA C40 8.02 0.13 X3 9.49 0.11 X5 10.46 0.12

Supplementary Table 5 | Statistics of the residual current of MspA in a BLM. The mean residual current and standard deviation are calculated from the Gaussian fitting of a current histogram (Supplementary Figure 8).

Mean Residual Current / pA Standard Deviation / pA C65 12.66 0.13 T65 14.59 0.84 A65 18.33 1.10 G3 in A65 22.12 0.76


7

Supplementary Table 6 | Statistics of the mean duration time for a full miRNA unzipping cycle. TII, TIII and TIV stands for the mean duration time for different periods in a full miRNA unzipping cycle. σ TII , σ TIII and σ TIV stands for the corresponding standard deviation of the distribution.

TII / ms σ TII / ms TIII / ms σ TIII / ms TIV / ms σ TIV / ms Counts Plet7i/Let7i 393.41 769.20 957.40 1287.51 125.63 284.87 541 Plet7i/Let7a 39.13 134.21 306.14 506.16 49.25 73.55 350 Plet7a/Let7a 315.98 695.61 521.73 781.33 84.2 230.37 378 Plet7a/Let7i 43.68 109.59 476.17 660.77 90.86 273.04 561 Let7i 0 0 832.15 1263.18 128.37 177.75 87 Let7a 0 0 707.22 786.37 47.67 113.23 59

Supplementary Table 7 | The rate constants for different probe/miRNA combinations. The rate constant is calculated from the exponential fitting of the histogram for TII (Figure 4d). Fully complementary hybridization between probe and miRNA generates lower rate constant values, while the unmatched counterparts generate higher rate constant values.

Plet7a / Let 7a Plet7i / Let 7a Plet7i / Let 7i Plet7a / Let 7i Rate Constant / s-1 13.66 68.80 7.06 47.45


8

Supplementary Figure 1 | The DIB measurement device. The device is manufactured as reported before10 with a CNC milling machine (Modela MDX-40, Roland). a-c, Three-view projection of the device used in the experiment. Scale unit: mm. The agarose (0.75% agarose, low melting point) coated coverslip sticks to the bottom of the plastic plate when the device is filled with molten agarose (2.5% agarose, low melting point) (b). d, The measurement chamber as shown in schematic form in (a-c). Inlet (yellow arrow) and outlet (red arrow) holes for molten agarose are shown. The extra hole on the outlet side helps air bubbles to escape during the filling. Each of the 16 holes (central part of the device) accommodates a single DIB.


9

Supplementary Figure 2 | Electrical measurementof αHL pores in aDIB. Nanopores and streptavidin-tethered ssDNA (C40, 267 nM) are placed in the droplet. Multiple nanopores insert in the DIB and the total ionic current is recorded using a patch clamp amplifier. a, A representative electrical trace of DNA blockages. Streptavidin-tethered ssDNA in the droplet blocks each pore sequentially after a positive potential is applied. b, Histogram of the electrical trace in (a). Separations between the adjacent peaks (82.9±6.5 pA, N=12) are comparable to the typical DNA blockade amplitude in a BLM measurement, which implies that the potential drop in the agarose is negligible compared to that in the lipid membrane.


10

Supplementary Figure 3 |The Current-Voltage (I-V) curve of an αHL pore in a BLM. The I-V curve of a single nanopore is measured in a BLM. I-V measurements for all four buffer combinations in the cis and trans compartments are recorded. a, The schematic diagram of I-V curve measurement with BLM. The cis side is defined as the compartment to which nanopores are added and is electrically grounded. To compare the I-V characteristics of the same nanopore with different buffer combinations, the buffer in either the cis or the trans side is exchanged without breaking the bilayer. b, I-V curves with all four different combinations of buffer 1 (0.66 M CaCl2, 8.8 mM HEPES, pH 7.0) and buffer 2 (1.32 M KCl, 8.8 mM HEPES, pH 7.0): 1) Cis: buffer 1/Trans: buffer 1; 2) Cis: buffer 1/Trans: buffer 2; 3) Cis: buffer 2/Trans: buffer 2; 4) Cis: buffer 1/Trans: buffer 2. At positive potentials, the conductance of a single pore is lower (see arrow) when the calcium chloride buffer is on the trans side.


11

Supplementary Figure 4 | Fluorescence-Voltage (F-V) curve of an αHL pore determined by oSCR a, The fluorescence intensity (marked with the red dots) gradually drops as the potential is ramped from +100 mV to -100 mV. b, The applied voltage potential protocol. Selected image frames on the right show the change of the spot brightness at different potentials.


12

Supplementary Figure 5 | Parallel recording using oSCR in a DIB. Multiple nanopores can be recorded simultaneously. a-f, Selected image frames of activity from three pores. Cartoon schematics on the top of each frame demonstrate the independent activities of three pores blocked by streptavidin-tethered C40. Scale bar: 10 µm. g, Independent fluorescence traces from the 3 pores in the same field of view.


13

Supplementary Figure 6 | The minimum spatial resolution of oSCR. In principle, fluorescence spots which are separated by more than their FWHM should be resolved (Supplementary Text 1). To test the minimum spatial resolution of oSCR, the agarose coating below the DIB is over hydrated to increase the mobility of the nanopores in the DIB. The drift of the pores (Supplementary Video 2) generates various pore to pore distance with time. The tracking resolution is thus estimated by analyzing pore pairs at small separations. a, Image frame containing many nanopores. The tracking is marked by colour-coded circles on the spots. Scale bar: 15 µm. b, Example of the minimum pore to pore distance at which adjacent pores are distinguished. The pore-to-pore separation in this case is 3.2 µm.


14

Supplementary Figure 7 | BLM measurements of αHL WT with three tethered oligonucleotides. Three types of DNA (streptavidin-tethered C40, X3 and X5) are mixed, each at the same concentration (89 nM). A ssDNA with abasic nucleotides may be captured at a higher rate due to its reduced molecular mass. Mean residual currents and standard deviations are determined after Gaussian fitting of the histograms (Supplementary Table 4). Different types of steptavidin-ssDNA are added to the measurement chamber sequentially and the corresponding peak in the histogram appears accordingly. In the histogram, the peaks from left to right correspond to event counts of C40, X3 and X5.


15

Supplementary Figure 8 | BLM measurements of MspA pores with individual DNA homopolymers. Independent BLM measurements are performed with a single nanopore. For each experiment, a streptavidin-tethered homopolymer (400 nM) is added to the cis compartment. Asymmetric electrolyte conditions (Supplementary Methods 5) are used to mimic oSCR experiments (Figure 3). a-d, Event histograms based on 500 voltage cycles each for C65(5’-Biotin-CCCCCCCCCCCC-CCC-C35-CCTGTCTCCCTGCCG-3’) (a), T65 (5’-Biotin-TTTTTTTTTTTT-TTT-T35-CCTGTCTCCCTGCCG-3’) (b), A65(5’-Biotin-AAAAAAAAAAAA-AAA-A35-CCTGTCTCCCTGCCG-3’) (c) and G3 in A65 (5’-Biotin-AAAAAAAAAAAA-GGG-A35-CCTGTCTCCCTGCCG-3’) (d). Mean residual currents and standard deviations are determined after Gaussian fitting of the histograms (Supplementary Table 5).


16

Supplementary Figure 9 | Fitting miRNA unzipping traces. The fluorescence trace is recorded in a DIB at +160 mV. a, A representative miRNA unzipping cycle., as described in figure 3 in the main text. b, The 1st order derivative of the trace in a. The peak centers determine the accurate transition times (dashed lines) between states. c, Gaussian fitting of peaks in the 1st order derivative trace. d, Automated event selection. A typical miRNA unzipping event has three characteristic current levels. By analyzing the amplitudes and the transition times in the fluorescence trace, miRNA unzipping events can be picked up automatically with a custom LabVIEW program. Periodically, -50 mV is applied across the DIB to clear the pore in the event that it is clogged. Incomplete events (0-1 s in d) and events at unclogging potentials (17.5-18.5 s in d) are not included in the statistics.


17

Supplementary Figure 10 | miRNA unzipping of various probe/miRNA pairs. The measurements are done in DIBs over 15 s. All four combinations of matched/mismatched probe/miRNA are tested. In general, long TII events (Figure 4a) (>1 s) are only detected when the miRNA is complementary to the probe. a, Plet7a / Let7a. b, Plet7a / Let7i. c,Plet7i / Let7i. d, Plet7i / Let7a.


18

Supplementary Figure 11 | Pore clogging by strongly hybridized probe/miRNA pairs. At a low applied potential (+100 mV), strongly hybridized probe/miRNA may clog the αHL nanopore. Clogging is relieved by switching the potential to a negative value (-50 mV). a, Diagram describing αHL clogging by Plet7i/Let7i at +100 mV. b, Diagram showing re-opening at -50 mV. c, A representative fluorescence trace showing long unzipping times at low applied potential (+100 mV) for Plet7i/Let7i. Under these conditions, TII is often more than 1 s and reversal of the potential may be needed to unclog the pore. d, The corresponding voltage protocol for the trace in c.


19

Supplementary Figure 12 | HHBa measurement chamber. The device is manufactured with a CNC milling machine (Modela MDX-40, Roland). a-c, Three-view projection of the HHBa chamber. Scale unit: mm. The agarose coated (0.2% agarose, low melting point) coverslip sticks to the bottom of the device when molten agarose (2% agarose, low melting point) is introduced into the setup (b). d, Photograph of the HHBa measurement chamber. Inlet (yellow arrow) and outlet (red arrow) holes on the device are designed to allow molten agarose to fill into the channel. The extra hole on the outlet side helps air bubbles to escape during the filling.


20

Supplementary Figure 13 | Hydrogel chip fabrication procedures. The hydrogel chip is fabricated according to the following procedure. Steps a-c are performed according to a standard soft lithography protocol5. Steps d-f are similar to the fabrication of an agarose stamp11 and the PDMS mould can be reused for this process. a, Formation of SU-8 pillars with the standard photolithography protocol (MicroChem). b, Casting a PDMS mould with SU-8 pillar array (a). c, Peeling off the PDMS mould from the SU-8 pillar array. d, Filling the micro-cavities of the PDMS mould with molten agarose (3% (v/v) agarose for routine use, 1.32 M KCl, 8.8 mM HEPES, 0.4 mM EDTA, pH 7.0). e, Desiccator degassing. Air bubbles in the hydrogel chip must be removed. A piece of coverslip is immediately placed on the back of the hydrogel. f, Gelling of the agarose at low temperature (4 ℃, 30 min). g, Peeling off the cast hydrogel chip from the PDMS mould. h, Spin coating of the hydrogel chip(4000x rpm, 60 s) with PMMA 495/A5 (0.2% in anisole). Immediately after the spin coating, a PMMA film will form in the gaps between the pillars. i, Immersing the chip in hexadecane to avoid de-hydration. Fluo-8 dye (1 µg/µL) is allowed to diffuse into the hydrogel. It takes ~4 h to achieve a homogeneous distribution of the dye. The chip immersed in hexadecane can be stored at room temperature for > 48 h. j, Comparing the size of the chip with a 5 cent Euro coin. Scale bar: 8 mm. The chip can be manipulated with tweezers. Image inset: Bright field microscopic image of the hydrogel pillars on the chip. Scale bar: 140 µm.


21

Supplementary Figure 14 | Electrodes for the HHBa chip. An electrode ring (Ag/AgCl) is designed both to hold the chip and secure an electrical connection. a, Top-view schematic of the electrode. b, Side-view schematic of the electrode with a chip. The chip is placed onto the electrode with tweezers. The electrode is coated with a layer of agarose to make the surface more hydrophilic so that the electrode and the chip spontaneously form a tight electrical connection upon physical contact. c, Photograph of an electrode. Scale bar: 5 mm. d, Diagram showing HHBa formation in3 mM lipid in oil. HHBa formation is more clearly visible in a video than in a static image (Supplementary Video 5).


22

Supplementary Figure 15 | Programmed chip loading with a spotting robot. Biological samples can be selectively spotted onto the surfaces of the pillars with a sharp capillary tip on a spotting robot (Patchstar, Scientifica), which loads samples (αHL or DNA from source droplets) and spots them onto the pillars of the chip by physical contact. The loading/spotting is controlled by a LabVIEW program which enables automated and accurate positioning (as low as 100 nm/step).The automated spotting process is also demonstrated in Supplementary Video 4. a-d, Monitoring the sequential spotting. When the spotting tip is in physical contact with the pillar (a, d), the capillary apex is clearly visible as a focused spot. Scale bar: 210 µm. e-h, Cartoons describing the spotting process seen in a-d. Biological samples (red) are spotted onto the pillars by passive diffusion while the tip and the pillars are in contact. The spotting efficiency depends on various conditions: the initial sample concentration in the capillary; the spotting duration; the diameter of the capillary tip. The arrows indicate the direction of movement of the capillary in the following step. i, Multi-sample spotting. Different biological samples can be spotted onto the same chip from various source droplets.


23

Supplementary Figure 16 | Optical single channel recording in a multiplexed HHBa chip. a, A cast hydrogel chip, scale bar: 4 mm (Supplementary Figure 13).The image inset shows an array of micro-pillar patterns on the chip surface, scale bar: 140 µm. b, HHBa measurement procedure. Biological samples (in this case αHL) are spotted onto the pillars (Supplementary Figure 15) and then the chip is flipped over to form HHBa with the substrate agarose in lipid/oil. c, Bilayer array detachment. After HHBa formation (Supplementary Video 5), the chip is lifted gradually to separate the bilayers. Clear boundary lines between intact and separated bilayers can be seen (white arrows). Scale bar: 140 µm. d, The intensity change of the fluorescence spots in the HHBa with alternating applied potential. αHL is spotted onto specific pillars (red arrow). The inserted pores show strong fluorescence at +50 mV and weak fluorescence at -50 mV. Each spot represents a single inserted nanopore. Scale bar: 35 µm. The fluorescence image is background normalized (The background profile is generated by averaging 100 fluorescence intensity images followed with Gaussian Blurring by ImageJ) to optimize image contrast. e, Fluorescence traces of the areas in d (indicated by red and purple arrows) in response to a voltage protocol. The nanopores show alternating fluorescence in response to the alternating applied potential. The fluorescence of the unspotted HHB (purple arrow) remains constant.


24

Supplementary Figure 17 | Parallel oSCR recording in full frame. a, Parallel recording of >100 pores in full frame (~150 µm x 150 µm). This gives ~ 5x103 pores/mm2 measurement density in full frame. In smaller area with higher pore density (red dashed square, 50 µm x 50 µm), the nanopore measurement density achieves > 104 pores/ mm2. Scale bar: 25 µm. b, 24 traces recorded by simultaneously oSCR showing miRNA (Plet7a) translocation events.


25

Supplementary Video 1 | Real-time oSCR of nanopore blockades with DNA. Raw fluorescence image recordings played back at real-time speed. The extracted trace (Supplementary Methods 5) and the colour coded Z-axis intensity profile are also shown. The nanopore in this video is αHL WT. The DNAs in this video are streptavidin-tethered C40 and X5, which are labelled as C and X near the Z-axis intensity profile.

Supplementary Video 2 | High-throughput nanopore tracking and recording. The substrate agarose coating is over hydrated on purpose to increase the bilayer fluidity and pore mobility. The nanopore in this video is αHL WT. At +100 mV, the fluorescence spots get bright transiently and the αHL pores are then quickly blocked by streptavidin-tethered ssDNA (C40).

Supplementary Video 3 | Parallel DNA detection using MspA. At +100 mV, streptavidin-tethered homopolymers in the droplet are captured by MspA. The voltage protocol applied for each oligonucleotide capture cycle consisted of three steps: 1) 0mV 2) ssDNA capture at +100 mV and 3) release at -100 mV. Scale bar: 30 µm.

Supplementary Video 4 | miRNA unzipping in parallel. At +160 mV, probe/miRNA samples in the droplet are captured by the αHL WT pores and are forced to unzip. The fluorescence spots blink due to repetitive miRNA unzipping and translocation. The probe/miRNA sample being tested in this video is Plet7i/let7i.

Supplementary Video 5 | Hydrogel chip loading. A demonstration of automated chip loading/spotting with a robot. The loading/spotting is automated by a LabVIEW program which guarantees accuracy and consistency.

Supplementary Video 6 | Hydrogel bilayer array formation. Upon physical contact with the substrate agarose, the hydrogel chip gradually forms HHBa by annealing to the agarose coating on the coverslip in 3 mM lipid/oil.

Supplementary Video 7 | Multi-sample recording in the hydrogel array. Single-molecule nanopore activities can be monitored within four bilayers simultaneously in HHBa with TIRF microscopy. Two types of biological sample are spotted on different pillars (top left: αHL+DNA, top right: αHL only). Many pore blockades (sudden fluorescence intensity drops) can be seen in the top left bilayer. While in the top right bilayer, such events are rare. The DNA used in this video is streptavidin-tethered C40.


26

Supplementary References 1. Castell OK, Berridge J, Wallace MI. Quantification of Membrane Protein Inhibition by Optical

Ion Flux in a Droplet Interface Bilayer Array. Angewandte Chemie-International Edition 2012, 51(13): 3134-3138.

2. Butler TZ, Pavlenok M, Derrington IM, Niederweis M, Gundlach JH. Single-molecule DNA

detection with an engineered MspA protein nanopore. Proceedings of the National Academy of Sciences of the United States of America 2008, 105(52): 20647-20652.

3. Cheley S, Malghani MS, Song LZ, Hobaugh M, Gouaux JE, Yang J, et al. Spontaneous

oligomerization of a staphylococcal alpha-hemolysin conformationally constrained by removal of residues that form the transmembrane beta-barrel. Protein Eng 1997, 10(12): 1433-1443.

4. Stoddart D, Heron AJ, Mikhailova E, Maglia G, Bayley H. Single-nucleotide discrimination in

immobilized DNA oligonucleotides with a biological nanopore. Proceedings of the National Academy of Sciences of the United States of America 2009, 106(19): 7702-7707.

5. Qin D, Xia YN, Whitesides GM. Soft lithography for micro- and nanoscale patterning. Nature

Protocols 2010, 5(3): 491-502. 6. Manrao EA, Derrington IM, Laszlo AH, Langford KW, Hopper MK, Gillgren N, et al. Reading

DNA at single-nucleotide resolution with a mutant MspA nanopore and phi29 DNA polymerase. Nature Biotechnology 2012, 30(4): 349-353.

7. Cherf GM, Lieberman KR, Rashid H, Lam CE, Karplus K, Akeson M. Automated forward and

reverse ratcheting of DNA in a nanopore at 5-angstrom precision. Nature Biotechnology 2012, 30(4): 344-348.

8. Jin Q, Fleming AM, Burrows CJ, White HS. Unzipping Kinetics of Duplex DNA Containing

Oxidized Lesions in an alpha-Hemolysin Nanopore. Journal of the American Chemical Society 2012, 134(26): 11006-11011.

9. Sauer-Budge AF, Nyamwanda JA, Lubensky DK, Branton D. Unzipping kinetics of double-

stranded DNA in a nanopore. Physical Review Letters 2003, 90(23): 238101. 10. Leptihn S, Castell OK, Cronin B, Lee EH, Gross LCM, Marshall DP, et al. Constructing droplet

interface bilayers from the contact of aqueous droplets in oil. Nature Protocols 2013, 8(6): 1048-1057.

11. Mayer M, Yang J, Gitlin I, Gracias DH, Whitesides GM. Micropatterned agarose gels for

stamping arrays of proteins and gradients of proteins. Proteomics 2004, 4(8): 2366-2376.


Documents

SUPPLEMENTARY INFORMATION · 2015-11-04 · Supplementary Information High-throughput optical sensing of nucleic acids in a nanopore array Shuo Huang*, Mercedes Romero-Ruiz*, Oliver

SUPPLEMENTARY INFORMATION · 2015-11-04 · Supplementary Information High-throughput optical sensing of nucleic acids in a nanopore array Shuo Huang, Mercedes Romero-Ruiz, Oliver