Research Progress

Activities

1.     Sensing materials synthesis and nanoparticle release for Adenosine Detection

1.1   Synthesis of sensing materials

Streptavidin-modified SNPs (~490 nm) and magnetic MPs (~5 µm) were obtained from Spherotech (Lake Forest, IL). The adenosine binding aptamer sequence and complementary sequence (CS) were selected based on a previously published study (Sexton et al. 2007). These oligonucleotide sequences were custom ordered with a biotin modification at the 5’ end from Integrated DNA technologies (Coralville, IA). For the modification of SNPs, 1 mg of particles were reacted with 3 nmole aptamer in 200 µL binding buffer solution (10mM Tris, 1mM EDTA, 2M NaCl) at 4°C overnight. Similarly, 10 mg of MPs were reacted overnight with 8 nmole of CS in 200 µL binding buffer solution at 4°C. After the reaction, the particles were pelleted and washed in  phosphate-buffered saline (PBS) five times to remove the excess DNAs. The amount of DNAs that remained in the supernatant was measured via Nanodrop 2000. The amount of DNAs which remained in the supernatant were subtracted from the total amount of DNAs used for the reaction to estimate the amount of DNAs conjugated to the particles. The conjugation was verified by assessing the zeta potential of the particles before and after DNA modification using a Zetasizer Nano ZS (Malvern Panalytics).

1.2  Sample preparation and NP release

The aptamer-modified SNPs and CS-modified MPs were reacted overnight at either a 10:1 or 100:1 ratio in 100 µL binding buffer solution at room temperature on a rotator. After the reaction, the SNPs-MPs complexes were precipitated by magnetic attraction and the excess/unbound SNPs were collected from the solution.   The SNP-MP complexes were washed with PBS five times. For concentration dependent adenosine triggered release study, 2 x 106 microparticles were incubated in 50 µL of the release medium at room temperature on rotator. The release medium was prepared with different adenosine concentration (0.1 nM to 10 mM).  The adenosine treat-ment was applied for one hour. The whole release medium was collected by separating the microparticles. For time dependent release study, 1 µM and 1 nM adenosine con-centration were chosen for triggering nanoparticle release. The whole release medium was collected at the predeter-mined time points (0 to 60 minutes).

2. Demonstrate biomolecule detection with high sensitivity using a micropore resistive pulse sensor research

Details of the sdetection principle and the design of the micro resistive pulse sensor were described in the Year 2 report. They are illustrated in the following plot.



Figure 1. Mechanism of detecting small biomolecules via aptamer reconfiguration and nanoparticle counting. (a): Target biomolecules induce the release of nanoparticles from microparticles. The released nanoparticles are counted by a resistive pulse sensor. (b) Target-induced molecular reconfiguration of aptamer and separation of aptamer-CS duplex (CS: complementary sequence).

2.1 Detection principle, fabrication and test setup

The RPS was fabricated by using standard soft lithography method. An SU8 mold was fabricated using photography that are complementary to the following components: (1) a 2 µm x 2 µm x 10 μm (width x height x length) micropore (sensing channel) to detect the released nanoparticles, (2) a filtering structure consisting an array of 5 µm x 2 µm x 10 µm microchannels to prevent the debris and remaining MPs blocking the sensing channel, (3) two electrode holes to place a pair of Ag/AgCl electrodes to measure the voltage pulses, and (4) inlet/outlet reservoirs and connection channels. These microfluidic components were fabricated by pouring polydimethylsiloxane (PDMS) onto the SU8 mold, followed by degassing and curing the PDMS for 2 hours at 70 ⁰C. Then the inlet/outlet reservoir and electrode holes were punched by 1.5 mm and 1 mm in diameter biopsy punches, respectively. Next, the PDMS slab was treated by air plasma (200 mTorr, 50W, 50s), followed by being bonded onto a glass substrate to complete the fabrication. The dimensions of the micropore measured by the profilometer (Dektak 150, Veeco Instrument, NY, USA) were 2.28 ± 0.15 μm (width), 2.17 ± 0.22 μm (depth), and 10.87 ± 1.13 μm (length). Finally the Ag/AgCl electrodes (1-mm diameter) were inserted into the electrode holes to complete the RPS.

For each test, the SNP suspension was loaded into the inlet reservoir and driven by a syringe pump (KDS Legato 185, KD Scientific) at a flow rate of 36 μL/hr. A DC voltage (1V) was applied across the Ag/AgCl electrodes. The voltage pulse signals by the SNPs were amplified with an instrumentation amplifier (AD620BN, Analog Devices Inc, USA) and recorded with an NI – DAQ board (PCI-6133, National Instrument USA) at a sampling rate of 500 kHz. The recorded signals were denoised using a custom MATLAB program to analyze voltage pulses.

2.2       Adenosine Detection with the Micropore Resistive Pulse Sensor.

Adenosine was used as the target small biomolecules to prove the biomolecule detection method. Before mixing adenosine samples with nanoparticles-microparticles complex (NPs-MPs), we first measured the background/residual NP concentration of the NPs-MPs solution. The residual nanoparticles were separated from the solution by centrifugation, which were counted by the resistive pulse sensor. This background concentration was subtracted in the subsequent experiments with different adenosine concentrations. For concentration-dependent adenosine triggered release studies, 2 x 106 MPs were incubated in 50 µL of the release medium at room temperature on a rotator. The release medium was prepared with different adenosine concentration (0.1 nM to 10 mM). The adenosine treatment was applied for one hour. The SNP release medium was collected by separating the MPs via magnetic attraction. For time-dependent release studies, 1 µM and 1 nM adenosine concentrations were chosen for triggering SNP release. The SNP release medium was collected at the predetermined time points (0 to 60 minutes). Likewise, the SNP release medium was collected by separating the MPs via magnetic attraction.

 

3.          Study the feasibility of signal multiplexing based on the geometry modulation

To achieve fast counting of NPs (and hence high throughput detection of biomolecules), we designed a 4-channel device that are multiplexed based on geometry modulation. Sensing channels with different geometries were used to generate signals with a unique code. Code division multiple access), was used to enable encoding and decoding the signals in multiple channels from a combined voltage output. We chose the code sequences (1010110, 1101100, 0111111, 1001011) for each channel geometry design, as shown in Fig. 2a, b. These codes are orthogonal to each other, as there is a low cross-correlation between the codes. While the correlation of the code with itself, the auto-relation is one (maximum), shown in Fig. 2c.


Figure 2. Design and analysis of geometry modulation for code-multiplexed resistive pulse sensor. (a) Geometry design for each channel based on specific gold code sequences. (b) The chosen gold code sequence in our design. (c) Periodic auto-correlation and cross-correlation of the Gold sequences designed to encode electrical signals in out sensor.


We designed 4-channel device with a different geometry to generate a specific electrical signal based on the specific codes. For decoding the combined electrical signal, we need to determine (1) the signal duration that the particle passing through the channel, (2) the specific code that the auto-correlation is maximized. To demultiplex the combined signal to individual signals, first, we used 4 different code signals to correlate the recorded signal to obtain the auto-relation peak representing the strongest signal correlation. The desired signal with the maximum auto-correlation was subtracted from the original signal. The process was iterated until the residual signal did not contain any clear signal of channel. Correspondingly, the process was terminated when the normalized cross-correlation coefficient was less than 0.4 (weak correlation).

 

 

Findings

4.     Developing and characterizating sensing materials for nanoparticle release


Fig.1: Formation and characterization of microparticle-nanoparticle (MP-NP) complexes. (a) Number of DNA sequence bound per particle. (b) Zeta potential of unconjugated and DNA conjugated particles. (c-e) Examination of NP to MP ratio in hybridization reaction. (c) Representative images of NP-MP complexes. Scale bar: (d) Flow cytometry analysis of NP-MP complexes. Black: 0:1, Blue: 10:1, Orange: 100:1. (e) Average number of NPs bound to one MP.  

 

DNAs were first attached to the SNPs and MPs using biotin-streptavidin interactions. Due to the larger surface area of MPs, two orders of magnitude more DNAs were attached to a single MP that to a SNP (Figure 1a). In addition, the zeta potential of the particles before and after the surface modification was assessed. As shown in Figure 1b, due to the negative charge of DNA, the zeta potentials of MPs and NPs became more negative after DNA modification. A higher shift in the zeta potential of SNPs was observed. This could be explained by the presence of a larger number of SNPs per w/v % compared to MPs as the same w/v % of MPs and SNPs was used for measurement. Figure 1a and 1b confirmed the DNAs attachment on the surface of the particles.  After the formation of SNP-MP complexes, an inverted fluorescent microscope (Olympus IX73) was used to verify the SNPs attachment to the MPs at various SNP:MP ratios, i.e. 0:1 (No NPs), 10:1 and 100:1 as shown in Figure 1c. A  fluorescent signal was detected after the FITC labeled SNPs attached to the CS-MPs. As more SNPs attached per MP, a higher fluorescent intensity per unit area was observed. Figure 1d showed a significant shift in the FITC signal intensity after the SNPs attachment. Moreover, the FITC signal increased approximately 10-fold when the SNP:MP ratio increased from 10:1 to 100:1 ratio, indicating more SNP attachment per MP (Figure 1d). These flow cytometry results confirmed the observations under the fluorescent microscope. The number of SNPs attached per MP were calculated indirectly from the excess SNPs remaining in the solution after the reaction. As shown in Figure 3e, the calculated number of attached SNPs per MP was 44.2 ± 8.7 when NP:MP ratio was 100:1 and 7.5 ± 0.5 for 10:1. The subsequent experiments were conducted at NP:MP ratio of 100:1.

 

5.     Demonstrate small biomolecule detection with ultra-high sensitivity.


Figure 2: Experimental results. (a) The released SNP concentration measured by the resistive pulse sensor vs adenosine concentration. (b) The released SNP concentration at different incubation time, ranging from 5 min to 60 min. Left: the concentration of adenosine was 1nM. Right: the concentration of adenosine was 1μM. Each error bar represents standard deviation of 3 separate measurements from 3 batches. (c) Comparison of nanoparticle release over time. (d) Enhancement efficiency of a higher target molecule concentrations with respect to incubation time. Error bars represent the standard deviation calculated from 3 separate measurements of 3 sample replicates.


Adenosine samples with concentrations ranging from 0.1 nM to 10 mM were mixed with the SNP-MP complexes. After incubating the SNPs-MPs with adenosine sample for one hour, the medium was centrifuged to pellet the MPs and collect the supernatant containing the released SNPs. Next, the collected samples were diluted to 300 μL (1:15 dilution) with DPBS to avoid clogging the micropore because of the high concentration of SNPs. After agitation, the SNP samples were observed under a microscope to ensure the SNPs were uniformly dispersed. The SNPs were then counted by the resistive pulse sensor. Each sample was measured for 1 hour at a flow rate of 36 μL/hr. The released SNP concentration was calculated from SNP counts.

The relationship between the released SNP concentration and adenosine concentration is shown in Figure 2a. This result shows the released SNP concentration increased as the adenosine concentration increased. A log scale was used for both the horizontal and vertical axes. At adenosine concentrations from 0.1 nM to 101 nM, the released SNP concentration increased faster than at higher concentrations ( from 101 nM to 107 nM). This can be explained as follows. As each adenosine molecule collides with a SNP-MP complex it disrupts the DNA duplex and initiates the release of the bound nanoparticles into the solution. The interaction of the aptamer and its target is dynamic. The target molecule, adenosine, is likely to dissociate from its aptamer and be released back into the solution where it can bind to another aptamer and trigger more SNP release. This results in an amplified release. However, at a higher adenosine concentration, there are more adenosine competing to interact with aptamers.  Thus, the amplification effect is reduced..  

In comparison to a standard ELISA for Adenosine a1 Receptor ELISA Kit (detection range from 0.313 ng/mL – 20 ng/mL), our method achieved a lower LoD and wider detection range. Compared to other detection methods using aptamer modified NPs, which typically can measure a biomolecule concentration up to 150 nM, our method has a larger detection range.

In addition, to explore the NP release process, we also measured the released NP concentration at different mixing/incubation times. Two adenosine concentrations, 1nM and 1 μM were chosen. Adenosine samples were mixed with NPs-MPs solution and were incubated at 5min, 10min, 20 min, 40 min and 60min. The released NPs were collected at each incubation time and counted by the resistive pulse sensor . The results are shown in Figure 2b. The concentration of released SNPs increased with increasing incubation time (Figure 2c). For each adenosine concentration, the concentration of released SNP increased dramatically from 5 min to 40 min. After 40 minutes, the SNP release slowed down, especially for the lower adenosine concentration (e.g. 1nM). The released SNPs ratio (for 1 uM:1 nM adenosine concentration)  peaks at nearly 12:1 during the initial stages of incubation (first 5 minutes), and stabilizes at approximately 3:1 after 20 minutes of incubation (Figure 2d). This study shows that a longer incubation time reduces the variability in SNP. 

 

6.     Study of feasibility of signal multiplexing based on the geometry modulation

20 microns microparticles (Sigma-Aldrich, 74491) were used to test the 4-channel device. The code sequences for the 4 channels are 1010110, 1101100, 0111111, and 1001011. We used three different concentrations for microparticles counting, which are 240/μL, 480/μL and 720/μL. Using the signal multiplexing, results are shown in Fig. 3. The measured particle concentrations were 248.4 ± 11.7, 478.9 ± 45.2 and 702 ± 33.9. The measured particles sizes were 19.36 ± 0.68, 20.09 ± 0.39, and 20.25 ± 0.43. Both the measured concentrations and diameters are in good agreement with the actual particle concentration and sizes (20 ± 0.3 μm).


Figure 3: Comparison of measured and actual MP concentrations and diameters. (a) Concentration comparison. (b) Diameter comparison.


Figure 4: Decoding measured electrical signal step by step. The measured signal was correlated with the standard signals to identify the right code signal that has the maximum auto-correlation peak.

For decoding part, we illustrated how this modulation method can demultiplexed the combined electrical signals in Fig. 4. First, we used 4 unique code signals to determine the autocorrelation peak, according to the peak signal interference. Then, the desired signal which has the maximum auto-correlation was subtracted from the original signal (Fig. 4, 1st row). The remaining signal would continue to correlate the different code standard signal until the correlation coefficient was less than 0.4 (weak correlation), which means the residual did not contain any desired code signal (Fig. 4, 3rd row).  Signals from individual channels (with a signature code) can be obtained from the subtracted signals.