High Accuracy Protein Identification: Fusion of Solid-State Nanopore Sensing and Machine Learning

Proteins are arguably one of the most important class of biomarkers for health diagnostic purposes. Label-free solid-state nanopore sensing is a versatile technique for sensing and analyzing biomolecules such as proteins at single-molecule level. While molecular-level information on size, shape, and charge of proteins can be assessed by nanopores, the identification of proteins with comparable sizes remains a challenge. Here, solid-state nanopore sensing is combined with machine learning to address this challenge. The translocations of four similarly sized proteins is assessed using amplifiers with bandwidths (BWs) of 100 kHz and 10 MHz, the highest bandwidth reported for protein sensing, using nanopores fabricated in <10 nm thick silicon nitride membranes. F-values of up to 65.9% and 83.2% (without clustering of the protein signals) are achieved with 100 kHz and 10 MHz BW measurements, respectively, for identification of the four proteins. The accuracy of protein identification is further enhanced by classifying the signals into different clusters based on signal attributes, with F-value and specificity of up to 88.7% and 96.4%, respectively, for combinations of four proteins. The combined use of high bandwidth instruments, advanced clustering and machine learning methods allows label-free identification of proteins with high accuracy.

Enhanced Signal to Noise Ratio Enables High Bandwidth Nanopore Recordings and Molecular Weight Profiling of Proteins

Fast protein translocations often lead to bandwidth-limited amplitude-attenuated event signatures. In this study, we developed a protein- and electrolyte chemistry-centric pathway to construct a readily executable decision tree for the detection of non-attenuated protein translocations using conventional electronics. Each optimization encompasses increasing capture rate (C_R), signal-to-noise ratio (SNR), and minimizing irreversible analyte clogging to collect >10⁴ events/pipette spanning a host of electric fields. This was demonstrated using 11 proteins ranging from ∼12 kDa to ∼720 kDa. Moreover, both symmetric and asymmetric electrolyte conditions (cis and trans chamber electrolyte concentration ratios _<^> 1) were explored. As a result, asymmetric electrolyte conditions were favorable on the extreme ends of the size spectrum (i.e., larger, and smaller proteins) and while the remainder of proteins were best sensed under symmetric electrolyte conditions. Under these optimal conditions, only ≲10% of events were attenuated at 500 mV (≲ 5% for most proteins at 500 mV with only ≲1-5% of the population faster than ∼7 μs, which is the theoretical attenuation threshold for 100 kHz bandwidth). Finally, applied voltage (V_app), peak current drop (ΔI_p), electrolyte conductivity (K), and open-pore conductance (G₀) were used to generate a linear relationship to evaluate the molecular weight of the protein (M_w) using plots of (dΔI_p)/(dV_app) vs M_w/(G₀/K).

Nanopores in Graphene and Other 2D Materials: A Decade's Journey toward Sequencing

Nanopore techniques offer a low-cost, label-free, and high-throughput platform that could be used in single-molecule biosensing and in particular DNA sequencing. Since 2010, graphene and other two-dimensional (2D) materials have attracted considerable attention as membranes for producing nanopore devices, owing to their subnanometer thickness that can in theory provide the highest possible spatial resolution of detection. Moreover, 2D materials can be electrically conductive, which potentially enables alternative measurement schemes relying on the transverse current across the membrane material itself and thereby extends the technical capability of traditional ionic current-based nanopore devices. In this review, we discuss key advances in experimental and computational research into DNA sensing with nanopores built from 2D materials, focusing on both the ionic current and transverse current measurement schemes. Challenges associated with the development of 2D material nanopores toward DNA sequencing are further analyzed, concentrating on lowering the noise levels, slowing down DNA translocation, and inhibiting DNA fluctuations inside the pores. Finally, we overview future directions of research that may expedite the emergence of proof-of-concept DNA sequencing with 2D material nanopores.

Pressure-Biased Nanopores for Excluded Volume Metrology, Lipid Biomechanics, and Cell-Adhesion Rupturing

Nanopore sensing has been widely used in applications ranging from DNA sequencing to disease diagnosis. To improve these capabilities, pressure-biased nanopores have been explored in the past to-primarily-increase the residence time of the analyte inside the pore. Here, we studied the effect of pressure on the ability to accurately quantify the excluded volume which depends on the current drop magnitude produced by a single entity. Using the calibration standard, the inverse current drop (1/ΔI) decreases linearly with increasing pressure, while the dwell drop reduces exponentially. We therefore had to derive a pressure-corrected excluded volume equation to accurately assess the volume of translocating species under applied pressure. Moreover, a method to probe deformation in nanoliposomes and a single cell is developed as a result. We show that the soft nanoliposomes and even cells deform significantly under applied pressure which can be probed in terms of the shape factor which was introduced in the excluded volume equation. The proposed work has practical applications in mechanobiology, namely, assessing the stiffness and mechanical rigidity of liposomal drug carriers. Pressure-biased pores also enabled multiple observations of cell-cell aggregates as well as their subsequent rupture, potentially allowing for the study of microbial symbioses or pathogen recognition by the human immune system.

Fast Fabrication of Solid-State Nanopores for DNA Molecule Analysis

Solid-state nanopores have been developed as a prominent tool for single molecule analysis in versatile applications. Although controlled dielectric breakdown (CDB) is the most accessible method for a single nanopore fabrication, it is still necessary to improve the fabrication efficiency and avoid the generation of multiple nanopores. In this work, we treated the SiNx membranes in the air–plasma before the CDB process, which shortened the time-to-pore-formation by orders of magnitude. λ-DNA translocation experiments validated the functionality of the pore and substantiated the presence of only a single pore on the membrane. Our fabricated pore could also be successfully used to detect short single-stranded DNA (ssDNA) fragments. Using to ionic current signals, ssDNA fragments with different lengths could be clearly distinguished. These results will provide a valuable reference for the nanopore fabrication and DNA analysis.

Modulation of electrophoresis, electroosmosis and diffusion for electrical transport of proteins through a solid-state nanopore

Nanopore probing of molecular level transport of proteins is strongly influenced by electrolyte type, concentration, and solution pH. As a result, electrolyte chemistry and applied voltage are critical for protein transport and impact, for example, capture rate (C_R), transport mechanism (i.e., electrophoresis, electroosmosis or diffusion), and 3D conformation (e.g., chaotropic vs. kosmotropic effects). In this study, we explored these using 0.5–4 M LiCl and KCl electrolytes with holo-human serum transferrin (hSTf) protein as the model protein in both low (±50 mV) and high (±400 mV) electric field regimes. Unlike in KCl, where events were purely electrophoretic, the transport in LiCl transitioned from electrophoretic to electroosmotic with decreasing salt concentration while intermediate concentrations (i.e., 2 M and 2.5 M) were influenced by diffusion. Segregating diffusion-limited capture rate (R_diff) into electrophoretic (R_diff,EP) and electroosmotic (R_diff,EO) components provided an approach to calculate the zeta-potential of hSTf (ζ_hSTf) with the aid of C_R and zeta potential of the nanopore surface (ζ_pore) with (ζ_pore–ζ_hSTf) governing the transport mechanism. Scrutinization of the conventional excluded volume model revealed its shortcomings in capturing surface contributions and a new model was then developed to fit the translocation characteristics of proteins.

Figure shows hSTf protein translocating through a solid-state nanopore under an applied electric field and the resulting current traces. The transport mechanism is determined by the interplay of electrophoretic and electroosmotic force.

Microfluidic Assessment of Drug Effects on Physical Properties of Androgen Sensitive and Non-Sensitive Prostate Cancer Cells

The identification and treatment of androgen-independent prostate cancer are both challenging and significant. In this work, high-throughput deformability cytometry was employed to assess the effects of two anti-cancer drugs, docetaxel and enzalutamide, on androgen-sensitive prostate cancer cells (LNCaP) and androgen-independent prostate cancer cells (PC-3), respectively. The quantified results show that PC-3 and LNCaP present not only different intrinsic physical properties but also different physical responses to the same anti-cancer drug. PC-3 cells possess greater stiffness and a smaller size than LNCaP cells. As the docetaxel concentration increases, PC-3 cells present an increase in stiffness and size, but LNCaP cells only present an increase in stiffness. As the enzalutamide concentration increases, PC-3 cells present no physical changes but LNCaP cells present changes in both cell size and deformation. These results demonstrated that cellular physical properties quantified by the deformability cytometry are effective indicators for identifying the androgen-independent prostate cancer cells from androgen-sensitive prostate cancer cells and evaluating drug effects on these two types of prostate cancer.

Rapid Discrimination of Extracellular Vesicles by Shape Distribution Analysis

A rapid and simple cancer detection method independent of cancer type is an important technology for cancer diagnosis. Although the expression profiles of biological molecules contained in cancer cell-derived extracellular vesicles (EVs) are considered candidates for discrimination indexes to identify any cancerous cells in the body, it takes a certain amount of time to examine these expression profiles. Here, we report the shape distributions of EVs suspended in a solution and the potential of these distributions as a discrimination index to discriminate cancer cells. Distribution analysis is achieved by low-aspect-ratio nanopore devices that enable us to rapidly analyze EV shapes individually in solution, and the present results reveal a dependence of EV shape distribution on the type of cells (cultured liver, breast, and colorectal cancer cells and cultured normal breast cells) secreting EVs. The findings in this study provide realizability and experimental basis for a simple method to discriminate several types of cancerous cells based on rapid analyses of EV shape distributions.

Assessment of 1/f noise associated with nanopores fabricated through chemically tuned controlled dielectric breakdown

Recently, we developed a fabrication method-chemically-tuned controlled dielectric breakdown (CT-CDB)-that produces nanopores (through thin silicon nitride membranes) surpassing legacy drawbacks associated with solid-state nanopores (SSNs). However, the noise characteristics of CT-CDB nanopores are largely unexplored. In this work, we investigated the 1/f noise of CT-CDB nanopores of varying solution pH, electrolyte type, electrolyte concentration, applied voltage, and pore diameter. Our findings indicate that the bulk Hooge parameter (α_b ) is about an order of magnitude greater than SSNs fabricated by transmission electron microscopy (TEM) while the surface Hooge parameter (α_s ) is ∼3 order magnitude greater. Theα_s of CT-CDB nanopores was ∼5 orders of magnitude greater than theirα_b , which suggests that the surface contribution plays a dominant role in 1/f noise. Experiments with DNA exhibited increasing capture rates with pH up to pH ∼8 followed by a drop at pH ∼9 perhaps due to the onset of electroosmotic force acting against the electrophoretic force. The1/f noise was also measured for several electrolytes and LiCl was found to outperform NaCl, KCl, RbCl, and CsCl. The 1/f noise was found to increase with the increasing electrolyte concentration and pore diameter. Taken together, the findings of this work suggest the pH approximate 7-8 range to be optimal for DNA sensing with CT-CDB nanopores.

Ion Current Rectification in Extra-Long Nanofunnels

Nanofluidic systems offer new functionalities for the development of high sensitivity biosensors, but many of the interesting electrokinetic phenomena taking place inside or in the proximity of nanostructures are still not fully characterized. Here, to better understand the accumulation phenomena observed in fluidic systems with asymmetric nanostructures, we study the distribution of the ion concentration inside a long (more than 90 µm) micrometric funnel terminating with a nanochannel. We show numerical simulations, based on the finite element method, and analyze how the ion distribution changes depending on the average concentration of the working solutions. We also report on the effect of surface charge on the ion distribution inside a long funnel and analyze how the phenomena of ion current rectification depend on the applied voltage and on the working solution concentration. Our results can be used in the design and implementation of high-performance concentrators, which, if combined with high sensitivity detectors, could drive the development of a new class of miniaturized biosensors characterized by an improved sensitivity.

Mechanical characterization of vesicles and cells: A review

Vesicles perform many essential functions in all living organisms. They respond like a transducer to mechanical stress in converting the applied force into mechanical and biological responses. At the same time, both biochemical and biophysical signals influence the vesicular response in bearing mechanical loads. In recent years, liposomes, artificial lipid vesicles, have gained substantial attention from the pharmaceutical industry as a prospective drug carrier which can also serve as an artificial cell-mimetic system. The ability of these vesicles to enter through pores of even smaller size makes them ideal candidates for therapeutic agents to reach the infected sites effectively. Engineering of vesicles with desired mechanical properties that can encapsulate drugs and release as required is the prime challenge in this field. This requirement has led to the modifications of the composition of the bilayer membrane by adding cholesterol, sphingomyelin, etc. In this article, we review the manufacturing and characterization techniques of various artificial/synthetic vesicles. We particularly focus on the electric field-driven characterization techniques to determine different properties of vesicle and its membranes, such as bending rigidity, viscosity, capacitance, conductance, etc., which are indicators of their content and mobility. Similarities and differences between artificial vesicles, natural vesicles, and cells are highlighted throughout the manuscript since most of these artificial vesicles are intended for cell mimetic functions.

Analysis Method of the Ion Current-Time Waveform Obtained from Low Aspect Ratio Solid-state Nanopores

Low aspect ratio nanopores are expected to be applied to the detection of viruses and bacteria because of their high spatial resolution. Multiphysics simulations have revealed that the ion current-time waveform obtained from low aspect ratio nanopores contains information on not only the volume of viruses and bacteria, but also the structure, surface charge, and flow dynamics. Analysis using machine learning extracts information about these analytes from the ion current-time waveform. The combination of low aspect ratio nanopores, multiphysics simulation, and machine learning has made it possible to distinguish different types of viruses and bacteria with high accuracy.

Translocation of soft phytoglycogen nanoparticles through solid-state nanochannels

Phytoglycogen nanoparticles are soft, naturally-derived nanomaterials with a highly uniform size near 35 nm. Their interior is composed of a highly-branched polysaccharide core that contains more than 200% of its dry mass in water. In this work, we measure the translocation of phytoglycogen particles by observing blockade events they create when occluding solid-state nanochannels with diameters between 60 and 100 nm. The translocation signals are interpreted using Poisson-Nernst-Planck calculations with a "hardness parameter" that describes the extent to which solvent can penetrate through the interior of the particles. Theory and experiment were found to be in quantitative agreement, allowing us to extract physical characteristics of the particles on a per particle basis.

Characterization of Flagellar Filaments and Flagellin through Optical Microscopy and Label-Free Nanopore Responsiveness

In this study, we investigated the translocation characteristics of flagellar filaments (Salmonella typhimurium) and flagellin subunits through silicon nitride nanopores in tandem with optical microscopy analysis. Even though untagged flagella are dark to the optical method, the label-free nature of the nanopore sensor allows it to characterize both tagged (Cy3) and pristine forms of flagella (including real-time developments). Flagella were depolymerized to flagellin subunits at ∼65 °C (most commonly reported temperature), ∼70 °C, ∼75 °C, and ∼80 °C to investigate the effect of temperature (T_depol) on depolymerization. The change in conductance (ΔG) profiles corresponding to T_depol ∼65 °C and ∼70 °C were bracketed within the flagellin monomer profile whereas those of ∼75 °C and ∼80 °C extended beyond this profile, suggesting a change to the native protein state. The molecular radius calculated from the excluded electrolyte volume of flagellin through nanopore-based ΔG characteristics for each T_depol of ∼65 °C, ∼70 °C, ∼75 °C, and ∼80 °C yielded ∼4.2 ± 0.2 nm, ∼4.3 ± 0.3 nm, ∼4.1 ± 0.2 nm, and ∼4.7 ± 0.5 nm, respectively. This, along with ΔG (plateaued values) and translocation time profiles, points to the possibility of flagellin misfolding at ∼80 °C.

Electrophoretic transport and dynamic deformation of bio-vesicles

Study of the deformation dynamics of cells and other sub-micron vesicles, such as virus and neurotransmitter vesicles are necessary to understand their functional properties. This mechanical characterization can be done by submerging the vesicle in a fluid medium and deforming it with a controlled electric field, which is known as electrodeformation. Electrodeformation of biological and artificial lipid vesicles is directly influenced by the vesicle and surrounding media properties and geometric factors. The problem is compounded when the vesicle is naturally charged, which creates electrophoretic forcing on the vesicle membrane. We studied the electrodeformation and transport of charged vesicles immersed in a fluid media under the influence of a DC electric field. The electric field and fluid-solid interactions are modeled using a hybrid immersed interface-immersed boundary technique. Model results are verified with experimental observations for electric field driven translocation of a virus through a nanopore sensor. Our modeling results show interesting changes in deformation behavior with changing electrical properties of the vesicle and the surrounding media. Vesicle movement due to electrophoresis can also be characterized by the change in local conductivity, which can serve as a potential sensing mechanism for electrodeformation experiments in solid-state nanopore setups.

Molecular-Level Profiling of Human Serum Transferrin Protein through Assessment of Nanopore-Based Electrical and Chemical Responsiveness

In this study, we investigated the voltage and pH responsiveness of human serum transferrin (hSTf) protein using silicon nitride (Si _xN _y) nanopores. The Fe(III)-rich holo form of hSTf was dominant when pH > pI, while the Fe(III)-free apo form was dominant when pH < pI. The translocations of hSTf were purely in an electrophoretic sense, thus depended on its pI and the solution pH. With increasing voltage, voltage driven protein unfolding became prominent which was indicated by the trends associated with change in conductance, due to hSTf translocation, and in the excluded electrolyte volume. Additionally, analysis of the translocation events of the pure apo form of hSTf showed a clear difference in the event population compared to that of the holo form. The results obtained demonstrate the successful application of nanopore devices to distinguish between the holo and apo forms of hSTf in a mixture and to analyze its folding and unfolding phenomenon over a range of pH and applied voltages.

Mechanical characterization of HIV-1 with a solid-state nanopore sensor

Enveloped viruses fuse with cells to transfer their genetic materials and infect the host cell. Fusion requires deformation of both viral and cellular membranes. Since the rigidity of viral membrane is a key factor in their infectivity, studying the rigidity of viral particles is of great significance in understating viral infection. In this paper, a nanopore is used as a single molecule sensor to characterize the deformation of pseudo-type human immunodeficiency virus type 1 at sub-micron scale. Non-infective immature viruses were found to be more rigid than infective mature viruses. In addition, the effects of cholesterol and membrane proteins on the mechanical properties of mature viruses were investigated by chemically modifying the membranes. Furthermore, the deformability of single virus particles was analyzed through a recapturing technique, where the same virus was analyzed twice. The findings demonstrate the ability of nanopore resistive pulse sensing to characterize the deformation of a single virus as opposed to average ensemble measurements.

A low voltage nanopipette dielectrophoretic device for rapid entrapment of nanoparticles and exosomes extracted from plasma of healthy donors

An insulator-based dielectrophoresis (iDEP) is a label-free method that has been extensively utilized for manipulation of nanoparticles, cells, and biomolecules. Here, we present a new iDEP approach that can rapidly trap nanoparticles at the close proximity of a glass nanopipette’s tip by applying 10 V/cm direct current (DC) across the pipette’s length. The trapping mechanism was systemically studied using both numerical modeling and experimental observations. The results showed that the particle trapping was determined to be controlled by three dominant electrokinetic forces including dielectrophoretic, electrophoretic and electroosmotic force. Furthermore, the effect of the ionic strength, the pipette’s geometry, and the applied electric field on the entrapment efficiency was investigated. To show the application of our device in biomedical sciences, we demonstrated the successful entrapment of fluorescently tagged liposomes and unlabeled plasma-driven exosomes from the PBS solution. Also, to illustrate the selective entrapment capability of our device, 100 nm liposomes were extracted from the PBS solution containing 500 nm polystyrene particles at the tip of the pipette as the voltage polarity was reversed.

Collision and Coalescence of Single Attoliter Oil Droplets on a Pipet Nanopore

We describe the use of a quartz pipet nanopore to study the collision and coalescence of individual emulsion oil droplets and their subsequent nanopore translocation. Collision and coalescence of single toluene droplets at a nanopore orifice are driven primarily by electroosmosis and electrophoresis and lead to the fast growth of a trapped oil droplet. This results in a stepwise current response due to the coalesced oil droplet increasing its volume and its ability to partially block the nanopore's ionic current, allowing us to use the resistive-pulse method to resolve single droplet collisions. Further growth of the trapped oil droplet leads to a complete blockage of the nanopore and a nearly 100% current decay. The trapped oil droplet shows enormous mechanical stability at lower voltages and stays in its trapped status for hundreds of seconds. An increased voltage can be used to drive the trapped droplet into the pipet pore within several milliseconds. Simultaneous fluorescence imaging and amperometry were performed to examine droplet collision, coalescence, and translocation, further confirming the proposed mechanism of droplet-nanopore interaction. Moreover, we demonstrate the unique ability to perform fast voltammetric measurements on a nanopore-supported attoliter oil droplet and study its voltage-driven ion transfer processes.

Multiple consecutive recapture of rigid nanoparticles using a solid-state nanopore sensor

Solid-state nanopore sensors have been used to measure the size of a nanoparticle by applying a resistive pulse sensing technique. Previously, the size distribution of the population pool could be investigated utilizing data from a single translocation, however, the accuracy of the distribution is limited due to the lack of repeated data. In this study, we characterized polystyrene nanobeads utilizing single particle recapture techniques, which provide a better statistical estimate of the size distribution than that of single sampling techniques. The pulses and translocation times of two different sized nanobeads (80 nm and 125 nm in diameter) were acquired repeatedly as nanobeads were recaptured multiple times using an automated system controlled by custom-built scripts. The drift-diffusion equation was solved to find good estimates for the configuration parameters of the recapture system. The results of the experiment indicated enhancement of measurement precision and accuracy as nanobeads were recaptured multiple times. Reciprocity of the recapture and capacitive effects in solid state nanopores are discussed. Our findings suggest that solid-state nanopores and an automated recapture system can also be applied to soft nanoparticles, such as liposomes, exosomes, or viruses, to analyze their mechanical properties in single-particle resolution.

Hydrophilic and size-controlled graphene nanopores for protein detection

This paper describes a general approach for transferring clean single-layer graphene onto silicon nitride nanopore devices and the use of the electron beam of a transmission electron microscope (TEM) to drill size-controlled nanopores in freely suspended graphene. Besides nanopore drilling, we also used the TEM to heal and completely close the unwanted secondary holes formed by electron beam damage during the drilling process. We demonstrate electron beam assisted shrinking of irregularly shaped 40-60 nm pores down to 2 nm, exhibiting an exquisite control of graphene nanopore diameter. Our fabrication workflow also rendered graphene nanopores hydrophilic, allowing easy wetting and use of the pores for studying protein translocation and protein-protein interaction with a high signal to noise ratio.

Nanoparticle mechanics: deformation detection via nanopore resistive pulse sensing

Solid-state nanopores have been widely used in the past for single-particle analysis of nanoparticles, liposomes, exosomes and viruses. The shape of soft particles, particularly liposomes with a bilayer membrane, can greatly differ inside the nanopore compared to bulk solution as the electric field inside the nanopores can cause liposome electrodeformation. Such deformations can compromise size measurement and characterization of particles, but are often neglected in nanopore resistive pulse sensing. In this paper, we investigated the deformation of various liposomes inside nanopores. We observed a significant difference in resistive pulse characteristics between soft liposomes and rigid polystyrene nanoparticles especially at higher applied voltages. We used theoretical simulations to demonstrate that the difference can be explained by shape deformation of liposomes as they translocate through the nanopores. Comparing our results with the findings from electrodeformation experiments, we demonstrated that the rigidity of liposomes can be qualitatively compared using resistive pulse characteristics. This application of nanopores can provide new opportunities to study the mechanics at the nanoscale, to investigate properties of great value in fundamental biophysics and cellular mechanobiology, such as virus deformability and fusogenicity, and in applied sciences for designing novel drug/gene delivery systems.