Resistive pulse sensing method for biomarker detection
Activities
We used the resistive pulse sensor to detect a real biomarker, human ferritin. The detection of human ferritin were performed in both phosphate buffer saline (PBS) and 10% fetal bovine serum (FBS) to prove that this method can be applied to a real biomarker detection environment in complex media such as blood. We also demonstrated that the detection range is adjustable by using different microparticle concentrations.
2.1 Detection Concept and Device Design
Fig. 1: A microscopy picture of microfabricated sensing channel for biomarker-microparticle aggregates detection.
The resistive pulse sensor consists of a micromachined channel (nominal dimensions: 30 μm (L) × 10 μm (W) × 10 μm (H)), one inlet reservoir, one outlet reservoirs and additional two reservoirs to accommodate a pair of Ag/AgCl electrodes. Ag/AgCl electrodes were used to reduce electrical double layer effect. Sample solution was loaded in the inlet reservoir and driven by a syringe pump. Biomarkers and polyclonal antibody-functionalized microparticles formed large-size biomarker-microparticle aggregates, which were detected by the sensing channels. The passage of each aggregate through the sensing channel caused a resistive pulse; the magnitude of the resistive pulse is proportional to volume of a biomarker aggregate. The measured volume fraction of aggregates indicates the biomarker concentration. This approach dramatically amplified signals of the resistive sensor, eliminating the need to modify the microchannel, and fabricate nanoscale sensing channels.
2.2 Device Fabrication
The resistive pulse sensor was fabricated using the standard soft lithography method. A two-layer structure was applied to the fabrication of the device. The thickness of the sensing channel is 10 µm and the thickness of the reservoirs is 40 µm. This two-layer structure can significantly reduce the resistance of the inlet and outlet channels of the sensor, and increase the sensitivity. To fabricate the device, first, SU8-5 negative photoresist was spin-coated onto a silicon wafer with a thickness of 10 µm. Photolithography was used to define the mold pattern for the microchannel. Then the second layer of SU8-2025 negative photoresist was spin-coated on to the silicon wafer with a thickness of 40 µm to define the pattern of the reservoirs. Once the desired pattern was obtained on the silicon wafer, polydimethylsiloxane (PDMS) was poured over the mold and cured to transfer the pattern onto the PDMS. The PDMS layer was then peeled off from the mold. Finally, the device was completed by bonding the PDMS layer and a glass slide by treating the two surfaces in air plasma. Fig. 1 shows the picture and the sensing channel of a fabricated device.
2.3 Sensor Testing
The measurement setup is shown in Fig. 2. A differential amplifier circuit was used to test the resistive pulse sensor which is modeled as RC in the circuit. R1 is a variable resistor which was set equal to RC. R2 = R3 = 500 kΩ. A signal generator (Agilent 33220A) was used to apply a DC voltage Vin = 2.4 V. The differential gain of amplifier AD620 was set to be 50 by using appropriate RG value (1 kΩ). The sensor response was recorded at a sampling rate of 1 MHz using a NI USB-6251 data acquisition board and a LabView interface. For each test, 15 μl of sample solution was loaded in to the device using a syringe pump (KDS Legato 270) at flow rate of 60 μl/hr.
Fig. 2: Measurement setup
Findings
With the demonstration of the feasibility of measuring biomarker concentrations via immunoaggregation in section 1, next we used human ferritin as a model biomarker to prove that microparticle-based immunoaggregation assay could be applied to real human biomarkers. In this study, the goat anti-human ferritin antibody was used as the capture probe on MPs. The biomarker corresponding to antibodies conjugated on the microparticle surface cause aggregation of Ab-MPs. We then used a microfluidic resistive pulse sensor to measure the sizes and count the number of aggregates. The volume fraction of aggregates can be calculated, which indicates the biomarker concentration.
Figure 3. Volume fraction of aggregates at different concentrations of human ferritin from 1.04 to 280.00 ng/ml in PBS buffer. The concentration of Ab-MP was 53.40 µg/ml for all tests.
As shown in Fig. 3, the volume fraction of aggregates increased with the increase of ferritin concentration in the range of 1.04 ng/ml to 62.40 ng/ml. The max volume fraction (35.4%) occurred at 62.40 ng/ml. Above 62.40 ng/ml, the volume fraction was reduced with the increase of ferritin concentration. This is because the high concentration of ferritin would saturate anti-ferritin Abs on surfaces of Ab-MP; hence the number of unreacted anti-ferritin Abs on Ab-MPs was too low to cause the aggregation.
Figure 4. Volume fraction of aggregates at different concentrations of human ferritin from 0.1 to 416.00 ng/ml in 10% FBS. Three sets of Ab-MP concentrations were used: 13.35 µg/ml, 53.40 µg/ml, and 213.40 µg/ml, for black (circle), red (triangle) and blue (square) lines respectively.
Next, we proved the detection of ferritin in 10% FBS to mimic the real biomarker detection environment in complex media. As shown in Fig. 4, the volume fraction of aggregates increased as increasing the ferritin concentration even in complex media (10% FBS). More importantly, tests showed that the detection range was adjustable by using different set of Ab-MP concentrations. It is important to point out here that volume fractions of nonspecific aggregates for the negative control were 0.045 ± 0.008, 0.054 ± 0.011 and 0.049 ± 0.002 for Ab-MP concentrations of 13.35 µg/ml, 53.4 µg/ml and 213.40 µg/ml, respectively, which are close to that in PBS at the concentration of 53.4 µg/ml. Results indicate that Ab-MPs are stable in the complex medium and nonspecific binding of non-target protein has negligible effect on the aggregation of Ab-MPs. With an Ab-MP concentration of 53.40 µg/ml (red curve), the max volume fraction also occurs at 62.40 ng/ml (see Fig. 3), and the aggregate volume fraction (32.4%) was comparable to that in PBS (35.4%). The result further confirms the immunoaggregation based biomarker detection method is insensitive to other biomolecules present in the complex media, implying that this approach can be applied to biomarker detection in complex media, i.e. human blood. Additionally, the detection range was shifted to the lower concentration range (0.1 ng/ml to 10.4 ng/ml) by using a lower concentration of Ab-MP (13.35 µg/ml). Similarly, a higher detection concentration range (0.1 ng/ml to 208 ng/ml) was achieved with a higher concentration of Ab-MP (213.40 µg/ml). Hence the detection range of the biomarker can be accurately tuned by changing the microparticle concentration. This biomarker detection method can be readily adapted for the detection and quantification of any biomacromolecules as long as there are high affinity capture probes available.