WO2022266398A1 - Method, system and apparatus for single molecule measurements of binding kinetic and enzyme activities using molecular electronic sensors - Google Patents

Abstract

The disclosed embodiments relate to molecular electronic sensors formed on a CMOS chip to detect various biomolecular interactions. The disclosed embodiments provide a platform to detect measurement of binding kinetics and enzyme activities. In one embodiment, the disclosure relates to an apparatus to detect small molecule interaction. The apparatus includes a first nanoelectrode and a second nanoelectrode separated by a gap; a molecular wire in electrical communication with the first and the second nanoelectrodes; a probe molecule connectable to the molecular wire, the probe molecule further comprising means to interact with a target molecule; a detector to detect one or more pulses indicating interaction kinetics between the probe molecule and the target molecule; and a sensor to receive the one or more detected pulses from the detector and identify at least one of the plurality of pules as a reaction indication type.

Description

METHOD, SYSTEM AND APPARATUS FOR SINGLE MOLECULE MEASUREMENTS OF BINDING KINETIC AND ENZYME ACTIVITIES USING MOLECULAR ELECTRONIC SENSORS

PRIORITY

[0001] The present disclosure claims priority to the U.S. Provisional Application No. 63/211,532 Filed June 16, 2021; the specification of which is incorporate herein in its entirety.

BACKGROUND

[0002] Rapid, quantitative, specific, and sensitive measurements of target analytes are the goal of many methods used in molecular biology and biotechnology. The bulk methods typically use a binding molecule to recognize the target molecule, combined with an indirect optical reporter mechanism, such as a fluorescent dye emission detected by a fluorimeter or microscope. Such classical methods detect an average of many molecules binding at once. In contrast, single molecule biosensors will dynamically bind and release analyte molecules repeatedly at statistically well-defined rates, such as the kinetic rates of binding and releasing (kon and k_0ff). The approaches used for such single molecule analysis fall into categories based on detection method. Most are fluorescence-based single molecule biosensors but other, specialized physical techniques have been used including electrochemical sensors plasmonic sensors, surface-enhanced Raman spectroscopy, and methods coupled to nanopore detection.

[0003] Single molecule optical methods suffer from fundamental limitations in signal and resolution and require complex labeling procedures to add fluorescent reporter molecules to targets of interest. The primary limitation for single molecule fluorescence methods is the rate of photon production from single dye molecules, which in turn is limited by rapid photo- bleaching of the dyes. This fundamentally limits the Signal to Noise Ratio (SNR) and limits the ability to make short integration-time measurements. This also limits the ability to make long time measurements, as many dyes undergo photolysis in less than about 100 seconds under intense illumination, thus making observation of rare events impractical as well. While microscopy can yield images of multiple flours at once, they often occupy random (and moving) locations in the field of view, requiring complex identification and tracking to achieve spatial resolution. More fundamentally, diffraction limits the ultimate spatial resolution or density of multiple distinct optical reporters, although advanced super resolution imaging methods can exceed these density limits in constrained situations.

[0004] Moving away from photon-based detection to all-electronic electron detection, both removes these fundamental constraints on SNR, scaling and bandwidth, and moreover is maximally compatible with implementation on CMOS chip devices. The single molecule sensing modality presented here is entirely non-optical and all-electronic, based only on measurements of electrons, and more specifically, based on using single molecules as electrical elements in a circuit. The scientific advances on the electrical nature of molecules and bioelectronic inspirations led to the proposal in the early 1970’s that single molecules could be engineered for use as circuit elements, to perform useful circuit functions such as a rectifier or switch. However, due to limitations of nanofabrication technology, it was not until the late 1990’s that the first single molecule circuits was demonstrated experimentally. Interest in the field of molecular electronics expanded dramatically after that point. The accompanying editorial noted that integrating molecules into chips would be the critical advance needed for this new field to have broad impact. [0005] Molecular electronics was originally envisioned as a way to make the ultimate miniaturized components such as switches or transistors. However, in the intervening decades since the first molecular circuits were demonstrated, silicon transistor technology, embodied in CMOS chips, has continued to make exponential progress in accord with Moore’s Law, and thus the potential need for miniaturizing molecular logic elements has instead been filled by the continuing advances of silicon-chip fabrication. It is therefore proposed here that the greatest value of molecular circuit elements is not for use as miniaturized logic, but instead for use as a miniaturized and universal form of sensor element. Molecules are the ultimate transducers of many forms of biophysical interaction, and are ideal for this role, whereas silicon or other solid state semiconductor materials are fundamentally deficient for this highly diverse set of sensor applications.

[0006] The scientific study of single molecule electronic sensing was initially based on carbon nanotube sensor devices. Research on Carbon nanotubes devices began in the early 1990’s, and their potential as sensors for single molecule interactions became apparent (20), initially in the context of sensing gas molecules and then chemical reactions. In 2011, Nuckolls, Shepard and colleagues described a single molecule sensor based on single-wall carbon nanotubes as a molecular wire that could monitor single-molecule DNA-DNA binding (hybridization) processes, through a specific functionalization of the nanotube with a single DNA oligomer “probe” molecule . Their sensor was a point-functionalized carbon nanotube with such a probe, acting as a field-effect device that could give time-resolved measurements at up to 10 kHz bandwidth. These sensors demonstrated measurements of single-molecule binding kinetics of DNA oligonucleotides in PBS buffer. A pure field effect mechanism of action was demonstrated by showing that the signal magnitude was reduced to zero as the solution Debye length was decreased through an increase in salt concentration. Collins and Weiss have also shown carbon nanotubes can be used to monitor single enzyme molecules activity for an enzyme attached to the nanotube, including Polymerase enzymes, and have also rigorously demonstrated a dominant field effect mechanism of action in this setting. They have also shown that such measurements can be performed at MHz bandwidth, to fully resolve the temporal kinetics of enzyme activity. The scientific study of single molecule electronic sensing was primarily based on carbon nanotube sensor devices. Research on Carbon nanotubes devices began in the early 1990’s, and their potential as sensors for single molecule interactions became apparent, initially in the context of sensing gas molecules and then chemical reactions. In 2011, Nuckolls, Shepard and colleagues described a single molecule sensor based on single-wall carbon nanotubes as a molecular wire that could monitor single-molecule DNA-DNA binding (hybridization) processes, through a specific functionalization of the nanotube with a single DNA oligomer “probe” molecule . Their sensor was a point-functionalized carbon nanotube with such a probe, acting as a field-effect device that could give time-resolved measurements at up to 10 kHz bandwidth. These sensors demonstrated measurements of single-molecule binding kinetics of DNA oligonucleotides in PBS buffer. A pure field effect mechanism of action was demonstrated by showing that the signal magnitude was reduced to zero as the solution Debye length was decreased through an increase in salt concentration . Collins and Weiss have also shown carbon nanotubes can be used to monitor single enzyme molecules activity for an enzyme attached to the nanotube, including Polymerase enzymes, and have also rigorously demonstrated a dominant field effect mechanism of action in this setting. They have also shown that such measurements can be performed at MHz bandwidth, to fully resolve the temporal kinetics of enzyme activity. [0007] At present there is no way to mass manufacture precision carbon nanotube devices having precise functionalizations, nor is there a manufacturing process to integrate nanotubes into CMOS chip devices. Despite the efforts of the IBM Watson Research Center and decades of attention, there is still no established path to a manufacturable commercial chip system. At present there is no way to mass manufacture precision carbon nanotube devices having precise functionalizations, nor is there a manufacturing process to integrate nanotubes into CMOS chip devices. Although carbon nanotubes are extremely sensitive and have allowed pioneering work in the science of single-molecule electronic sensing, despite decades of attention. There is still no established path to a manufacturable commercial chip system, despite, for example, an extensive industry effort at IBM Watson Research Center to integrate carbon nanotube integration into semiconductor chips.

[0008] It is therefore ideal to have molecular wire sensors that can be precision engineered for site specific conjugation with probe molecules, and for conjugation into the nanoelectrodes on a CMOS chip, and which are also readily available through existing processes. This generally limits the candidates to peptides, proteins or DNA as molecular wires, as these are in fact the only conducting polymers for which there are well-developed and highly sophisticated commercial precision synthesis capabilities, including extensive means of precision functionalization for diverse conjugation needs.

[0009] Double-stranded DNA helices (dsNDA) and protein alpha helices have both been studied as molecular wires. Detailed tunneling probe methods have also recently been used to study conduction through larger proteins. While not nearly as conductive as carbon nanotubes, these biopolymers have the great advantage for present purposes of allowing complete precision engineering using existing commercialized manufacturing methods. The sensors described in this work in particular use as a molecular wire an alpha-helical peptide, 25nm in length, with a specific conjugation site engineered into the side chain of an amino acid near the middle of the peptide, for attachment of probe molecules, and metal specific conjugation groups (1) engineered on to the ends for selective binding and self-assembly to the metal nanoelectrodes.

SUMMARY

[00010] Assays based on detecting molecular interactions are widely used to quantify analytes, detect biomarkers, and measure specific binding or enzyme activity in the presence of other biological molecules. This provides the foundation for many important biotechnology applications in society and industry, such as diagnostics, DNA sequencing, drug discovery and enzyme evolution. The disclosure generally describes a new and universal platform for measuring such interactions at the single-molecule scale, in real-time, and label-free, using a highly scalable, customizable sensor array CMOS chip device. This provides the unique ability to perform highly sensitive and highly multiplexed real-time monitoring of molecular interactions. These and other embodiments provide an ideal means - both practically manufacturable in the near term and with a durable long-term scaling potential - to bring the full power of modern CMOS chip technology to the broad and important field of biosensing.

[00011] The disclosure realizes a scientific vision to have single molecules integrated into electronic chips circuits as the ultimate form of miniaturized electronics. The discloed embodiments provide a biosensor chip platform with an extremely rapid future scaling roadmap, which can translate into powerful, compact and low-cost systems and tests. An observation from the discloed embodiments is that molecules can be ideal elements to act as sensors, rather than as logic circuit elements that have been well-served by silicon semiconductors. As such, they provide a universal sensor capability across all kinds of biosensing.

[00012] The molecular electronics sensor CMOS chips disclosed can be functionalized with specific biomolecules, including single-strand DNA and RNA probes, antibodies, antigens, aptamers, or enzymes, and then used to measure the interaction kinetics as these probe molecules interact with biomolecules in solution. The disclosure demonstrates detection of a range of molecular interactions from DNA hybridization to protein/DNA and protein/protein binding, including antibody/antigen binding and aptamer/target binding, as well as activities of enzymes important to biotechnology such as DNA polymerase, as used for sequencing, and CRSPR Casl2a, as used for diagnostics.

[00013] The resulting molecular electronics sensor chip platform provides the basis of simple new biosensor instruments for a broad range of applications, such as drug-target interaction characterization, diagnostic testing, DNA sequencing, or proteomic analysis, or environmental monitoring, with the potential for unprecedented speed, sensitivity, and multiplexing capabilities, using low-cost, mass-produced chips and compact, low-cost instruments ideal for highly decentralized, portable or mobile deployment.

[00014] Measuring molecular interactions is fundamental to many applications in biotechnology, such as drug discovery, which relies on measurement of drug-target interactions, molecular diagnostics, which rely on detecting antibody and DNA binding, and

DNA sequencing, which relies on monitoring DNA polymerase enzymes interacting with their substrates. For technological advancement of such applications, it is desirable to measure these interactions on a CMOS chip device. This allows the sensor to inherit the low- cost mass manufacturing, speed and miniaturization that are the hallmark of modern CMOS- chip based devices such as portable computers and cell phones. It also can take advantage of the durable roadmap for future improvements provided by 50 years of Moore’s Law scaling of CMOS chips, and the corresponding chip foundry infrastructure and supply chains.

[00015] The full potential of this vision of moving biosensing on-chip can be realized if it is achieved using an ideally matched sensor concept. It is important to understand that the approach presented here is guided by three essential design criteria that should be considered for evaluating any proposed on-chip biosensor solution: manufacturability, scalability, and universality.

[00016] Manufacturability: Foremost, it is essential that the approach be fully compatible with existing CMOS foundry fabrication processes, so that it can truly leverage this mass-manufacturing base and economy of scale. The chip industry has invested trillions of dollars in infrastructure, and the ideal is to fully leverage the existing and future foundries, and related supply chains. For practical purposes of making real world impact, it is critical that the chip and any other sensor-related elements are mass-manufacturable using existing industrial processes. In contrast, if new manufacturing processes must be made, it can take decades to achieve commercially attractive manufacturing capacity.

[00017] Scalability: It is desirable to solve the “More-than-Moore” problem, which otherwise limits the scalability of chip-based approaches: devices based on chips but also adding non-CMOS features — such as sensors — typically cannot keep up with the pace of scaling of Moore’s Law, and thus ultimately limit access to the benefits of tremendous cost and performance scaling that CMOS logic chips have enjoyed. The ideal on-chip solution should in no way impede miniaturization of the circuits, either for advancing among existing generations of CMOS foundries, which extend from two-decade old but widely used 180nm

“nodes” (the node is loosely related to minimum feature size, but directly related to achieving

Moore’s Law of denser circuitry), down to the state-of-the-art 5nm nodes, as well as moving to future foundries which are projected to extend towards lnm nodes. The ideal way to solve this is to have the sensor element be “fully scaled” at the onset, i.e., to already be at the nanometer scale, so that it does not ever require any further miniaturization as the supporting CMOS measurement circuitry is miniaturized. This also ensures durability of the technology, since there is a long and unimpeded roadmap for technology improvement merely by advancing the sensor chip manufacturing to existing and future finer node foundries.

[00018] Universality: Third, it is important to have a universal sensor concept, so that one chip platform can serve the great diversity of biosensing applications. This means the sensor should be applicable to the diverse range of sensing targets that include small molecules, antigens, proteins, antibodies, aptamers, DNA oligos, enzymes and their substrates. This also means that the sensor should support performance specifications up to the ultimate limits of such measurements, which would be single-molecule, real-time and label-free detection.

[00019] It is proposed that molecular electronics, with the particular form of such sensors presented here, provides an ideal solution to all these challenges, and thus provides an ideal way to put biosensing on chip. First, the molecular electronics sensor proposed is a single molecule in a circuit, and thus solves the “Moore-than-More” scaling problem: the critical sensor element is already shrunk to the smallest dimensions of the nanometer scale, and thus shrinking the circuits does not require any changes at all to the critical sensor element.

[00020] Second, the concept presented here, as illustrated in Figure 1A, is a sensor concept, which in principle applies to any form of molecular interaction, providing single molecule, real-time and label-free detection. Finally, the sensor chip proposed here is fully manufacturable for commercialization: the chip, including the nano-electrodes, is fully compatible with fabrication using photolithography processes in standard CMOS foundries, and the precision engineered molecular wire that is the foundation for the sensor element is also readily manufacturable by existing industry methods of synthetic chemistry and protein engineering.

BRIEF DESCRIPTION OF THE DRAWINGS

[00021] The detailed description is provided with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The use of the same reference numbers in different figures indicates similar or identical items.

[00022] Figure 1. Illustrates a Molecular Electronic Sensor and Chip according to one embodment of the disclosrue. (A). Senor Concept: Given a pair of molecules that undergo an interaction, one of the pair is selected as a probe molecule and conjugated to a precision molecular wire (here a synthetic alpha-helical protein) that spans a nano-scale gap between metal nanoelectrodes. These connect it to a driving voltage source and current meter circuit to provide real-time readout of current vs. time, for the current passing through the molecular wire/probe complex. When the target molecule binds to the probe, the resistance of the complex changes, resulting in an observed change in current. The resulting current trace has on/off pulses that provide a direct representation of the molecular interactions. (B). CMOS Chip device: A large scale array of sensors are fabricated on the surface of a CMOS chip. Shown is an annotated image of the CMOS chip device used in these studies. This chip has 16,000 sensors and the circuitry needed to digitize and transfer sensor readings off-chip, at a rate of 1000 frames per second. This chip is implemented inl80nm CMOS. [00023] Figure 2. CMOS Molecular Electronics Sensor Pixel. The sensor pixel concept is fabricated in CMOS as shown: A. Cross-Sectional illustration of the CMOS layer stack- up, consisting of the Front-End-Of-Line (FEOL) transistor layers, and the Back-End-Of-Line (BEOL) layers providing metal interconnects (lateral wires) and vias (vertical wires). This shows how nano-electrodes fabricated at an upper planar layer are connected down into the transistor layer using a series of vias and interconnects. B. Illustration of how the idealized molecular sensor is configured in the exposed nano-electrodes. C. Electron micrograph of actual chip nanoelectrodes, fabricated by CMOS-foundry compatible photo-lithography processes, to have an approximately 20nm gap open for the molecular wire. D. The pixel measurement circuit is shown schematically. This is located within the FEOL layers. As indicated, the measurement circuit integrates current flowing through one electrode (referred to as the “drain” side) onto a capacitor, for readout by switching into off-pixel array column circuits. Source and drain side inputs in turn are connected to the nanoelectrode and sensor molecular wire using the vias and interconnects as shown.

[00024] Figure 3. A Diversity of Molecular Electronic Sensors: The variety of sensors used in the work reported here are indicated, shown to scale with idealized molecular structures: (Upper left) DNA - DNA hybridization binding sensor; (Mid left) Aptamer - Protein Binding sensor (SARS-CoV-2 Spike Protein); (Lower left) CRSPR Cas Enzyme activity sensor (Casl2a); (Top right) Small molecule binding sensor (nucleotide binding in polymerase pocket); (Mid right) Antigen-Antibody binding sensor (Fluorescein antigen); (Lower right) DNA Polymerase activity/sequencing sensor (Phi29 Polymerase). All molecular graphics are drawn to the scale.

[00025] Figure 4. The DNA binding sensor. (A) A 17-mer oligonucleotide (5’- TACGTGCAGGTGACAGG-3’) is conjugated to the bridge using conventional click chemistry at its 5’ end. (B) Example current vs. time trace, showing 6 seconds of data sampled at 1kHz, taken with the sensor exposed to a 20 nM concentration of the target complementary 14-mer strand (5’-CCTGTCACCTGCAC) in Buffer A. Each pulse of current above the baseline represents a single DNA binding event. The durations of the events and time between events occur stochastically, with exponential distributions as summarized in Figure 15 - 18. (C) The distribution of measured current values in the trace suggests that the probe DNA is bound to target approximately 22% of the time.

[00026] Figure 5. Response of DNA binding sensor to Target Concentration. Binding of the 17-mer DNA probe on the sensor to the 14-mer target, for different target concentrations: Traces A-C showl -second-long traces of sensor current for exposure to 10 nM, 100 nM, and 1000 nM solutions of the target 14-mer. The width of the peaks (Dwell time) remains constant (~25 ms), but the time between peaks decreases from 45 to 4 ms as the concentration is increased, reflecting more frequent concentration-driven interactions. The Fraction of time bound was estimated from the current measurement histograms (insets at right). D shows the resulting fits to the measurements of dwell time and fraction bound vs. concentration. As expected for DNA binding, dwell time remains constant with concentration, but fraction of time bound shows a classical saturation curve, from which is computed a binding affinity, Kd, of 39nM. (N.B. the temperature for the runs shown in figs. 4-5 is not controlled or measured.)

[00027] Figure 6. Single-molecule thermal melting curves derived from the DNA binding sensor. For these experiments, a 45-mer ssDNA probe is attached to the bridge. Two different complementary target oligos that are closely matched in properties but having different melting points were designed by taking one of length 15 (5’- CCTCTGTGAAGGCCT) and the other to be an extension of this to length 20 (5’- CCTCTGTGAAGGCCTGATCG). These were added to the chip at a concentration of 20 nM, performed in series. For each solution, a thermal controlled mounted under the chip was used to set and sweep the temperature, in 2 °C steps, from 41°C to 55 °C. Standard DNA melting curves were fit using the method of (59), which were then used to derived the empirical melting points, T_m, shown. The results shown agree with the classical predictions for the difference in T_m between the oligos, but here measured entirely in a single molecule context.

[00028] Figure 7. Specificity of the DNA binding sensor for mismatched DNA. To finely illustrate the specificity of the DNA binding sensor, and to illustrate the rich information in the detailed single molecule kinetics, the sensor was used to probe mismatched targets. For these experiments, the sensor was a peptide bridge with a 45-mer oligonucleotide (5’-CGATCAGGCCTTCACAG AGGAAGTATCCTGTCGTTT AGCATACCC). Targets (all 20-mers having 0, 1, 2, or 3 mismatches as indicated) were added sequentially, and signal traces were used to measure the fraction of time bound and dwell time. The result is a significant downward trend in both fraction bound (shown) and dwell time (not shown), as the number of mismatched nucleotides is increased. These measurements were done using -4000 binding events each. The change in sensor response to even a single mismatch in the target 20-mer is readily detectable from either fraction bound or dwell time.

[00029] Figure 8. Protein and small molecule binding kinetics. As a model system for showing protein binding and small molecule binding, this sensor is configured to observe (A) DNA polymerase binding a primer/template, and (B) a nucleotide binding into the polymerase pocket, as shown in visualizations above. For this model system, a 17-mer ssDNA template is conjugated to the peptide bridge at its 5’ end (with 3’ end blocked to prevent polymerase binding that site). A complementary 14-mer primer strand is then bound to this, on the distal end of the 17-mer, to create a primer site on the sensor with the 3’ -OH available for polymerase binding. (A) shows the summary kinetics (dwell time, fraction of time bound) for Klenow DNA polymerase binding to the primer site, as polymerase concentration is titrated from 0.008 to 3.8 mM, in a background of 100 nM 14-mer primer to suppress primer dissociation. The inferred binding affinity of the polymerase, Kd, is 530nM. (B) A nucleotide titration is performed to observe the binding in the polymerase pocket, in a non-catalytic buffer so as to observe the binding kinetics. In this case, a 45mer template was put on the bridge, and a 31-mer primer was used that binds at the distal end, and such the first template base (dA) is complementary to the nucleotide being tested (dT). The nucleotide was added in concentrations of 2.5, 5, and 15 pM along with the 100 nM polymerase and primer in the presence of a buffer that has 10 mM Sr²⁺ (without Mg²⁺), in which nucleotide incorporation cannot occur. In this non-catalytic buffer, the dNTP will simply repeatedly bind and dissociate from the polymerase pocket, and the resulting summary binding kinetics are shown. This also serves to illustrate the detection of small molecule binding.

[00030] Figures 9A and 9B. Aptamer sensors. Aptamer Sensors were constructed, here targeting the SARS-CoV-2 S Protein with a 57-mer DNA aptamer (illustrated above) and targeting the SARS-CoV-2 N-Protein with a 97-mer DNA aptamer, both taken from the literature. Shown (lower) is the dose-response titration curves for both the S-aptamer and N- aptamer sensors, for a range of applied target protein concentrations. The binding affinities, Kd, derived from these curves (6.4nM, 39nM) are similar to those reported for standard bulk aptamer binding assays in solution.

[00031] Figures 10A and 10B. Antibody-antigen sensors. As a model system, a fluorescein - Anti -fluorescein antigen-antibody pair was used, with the fluorescein antigen presented on the sensor (Figure 10A) by tethering it to the bridge using a ssDNA oligo as a linker for synthetic convenience. A 45-mer oligonucleotide was used, with the 3’ (distal) end of the DNA capped with a fluorescein during synthesis. A commercial anti-fluorescein antibody (Fab) was added in TKS buffer on the chip. The summary kinetics are shown (Figure 10B) for dwell time and fraction of time bound, as the concentration of antibody is titrated over the range shown. The inferred binding affinity, Kd, was 1.3 mM. As a control, it was also observed that all binding signals were extinguished when 4 mM free fluorescein was added to saturate the antibody, verifying the specificity of the binding signal.

[00032] Figure 11. A CRSPR Cas sensor. To assemble a programmed Cas enzyme as a probe on the bridge, first a guide RNA targeting a dsDNA target for a CRSPR Casl2a enzyme was conjugated to the bridge, and these were assembled on chip. The casl2a was then allowed to bind onto the guide RNAs, thereby programming it for the target dsDNA, and also effectively tethering it to the bridge as a probe. For these experiments, the gRNA is a 40-mer, attached to the bridge peptide using the click chemistry at its 13^th nucleotide, which is the base that extends furthest outside the enzyme in the pseudo-knot loop. The kinetics are summarized in the titration curve, showing fraction of time bound saturating as the dsDNA target varies in concentration, in the presence of a concentration of 20 nM free Casl2a enzyme. Thus, this configuration acts directly as a sensor for the dsDNA target, without assessing post-target nonspecific ssDNAse activity. This later nonspecific activity is also observable on the sensor (not shown). The observed binding affinity for the target, Kd, is ~3 pM. These experimental buffer was 20 mM Tris HCl pH 8.0, 20 mM KC1, 10 mM SrCh, 4 mM DTT.

[00033] Figures 12A-12C. DNA Polymerase Activity Sensor. As visualized above, a phi29 DNA polymerase is conjugated to the bridge using the SpyTag-SpyCatcher conjugation method, where the SpyTag peptide is conjugated directly to the bridge, and the SpyCatcher protein is fused to the N terminus of the polymerase. Figure 12A shows a 25 second current trace, recorded after adding a primed template (3’-15G25A, see below), and dCTP and dATP nucleotides to the polymerase enzyme sensor chip. The blue curve is measured data, the dashed black line is an HMM state classification. A series of discrete pulses are observed, representing putative nucleotide incorporation events The plot has wide spaced, narrow pulses in its left half, and densely spaced, broader pulses to the right suggesting putative C incorporation vs A incorporation segments of the trace. In contrast, in control phases, with either no template DNA or template DNA but no matching nucleotides, only very narrow and small pulses (peak height ~5 pA) are observed. Figure 12B shows the results of a detailed analysis of the pulse features, further supporting the hypothesis that the dATP and dCTP incorporation pulses are distinguishable as indicated in Figure 12A. An HMM analysis was used to segment out the pulses, and each pulse was annotated with seven extracted features, including height and width and wait time for the next pulse. A Principal Component Analysis was performed on these 7 feature metrics, and the pulses are plotted projected into the first two principal component space. As shown, there are two well separated clusters, the respective red and green pulses from trace A fall into these two clusters. Thus, the pulses are statistically distinguishable, at a -90% level. The loading of the 7 primary feature metrics onto the first component PCI are shown in Figuer 12C, where it can be seen that time between pulses is the major factor, while pulse width and pulse height (mean or max) are also important factors. The quantitative results suggest dCTP incorporations occur at a slower rate (0.68 sec/base) than the dATP incorporations (0.29 sec/base), while the dCTP incorporation events are faster, at a pulse width of 0.12 seconds, versus dATP events at a pulse width of 0.22 seconds. (See Table 3, Fig. 2, for details). The DNA template sequence was 5’-

TTTTTTTTTTTTTTTTTTTTTTTTTGGGGGGGGGGGGGGG TC AGTC ACGTCTAGATGC AGTC AG-3 ’ with primer sequence 5'-CTGACT GCATCTAGACGTGACTGA, so the expectation is incorporation of 15 dC’s followed by 25 dA’s.

[00034] Figure 13A and 13B. Using the DNA binding sensor to detect pathogen sequences under assay conditions. Results illustrate a mock assay, in which the DNA binding probe sensor chip is used to detect target DNA in the context of a mock viral detection assay, with confounding complex background. Figure 13A. The titration curves show fraction of time sensors were bound to target DNA, for various targets, either synthetic 24-mer in buffer, or a PCR product from a contrived saliva sample. Salmon Sperm DNA was added as a mock background DNA at a level expected for contaminating genomic DNA in a raw saliva sample, 2 pg/ml. Such DNA binding sensor measurements were also performed in crude saliva extract (not shown), with comparable performance. Figure 13B. A signal trace at the lowest concentration tested, 100 pM, showing 5.2% fraction of time bound. Since the background baseline is quiet compared with the binding events, this does not represent the lower limit of detection, and substantially lower concentrations would be detectable with longer observations than the 3 seconds shown, and/or using additional such pixels to the same effect. The absolute number of target DNA molecules introduced to the chip was approximately 60 million (0.1 femtomole). By using means of longer time observation, multiple pixels, and voltage driven concentration of target DNA near the sensors, this could be reduced towards the single molecule limit, potentially allowing direct PCR-free detection of nucleic acid targets.

[00035] Figure 14A shows Pixel Noise Power Spectrum for Sensor Chip Pixel. (Upper) Representative signal trace from a pixel on a dry chip, showing level of electronic noise and baseline leakage current that would be present prior to beginning a wet experiment of the type reported herein (chip previsouly cleaned for use, but no molecules in the nano gap, signal measured in air, at IV DC applied voltage). Current y-axis is in ADC output counts, each count corresponds to ~0.4pA of current. As shown, the signal fluctuates rapidly (comparable to the 1kHz sampling rate) by 1 to 2 ADC counts (~ 0.4 to 0.8 pA) over the 30 seconds of observation. (Lower). Correspodng power spectrum computed from the trace (with DC offet removed) show the absence of noise tones. The majority of the trace above frequencies of lHz reflects pure thermal noise, as ideally expected.

[00036] Figure 14B shows a series of Pixel ADC linearity measurments. This shows standard ADC linearity metrics, verifying that the amplifier and ADC are highly linear over the dynamic range of current measurment. Measures are: (Upper) Uniformity of the ADC code histogram, as the input calibration current is swept linearly from 0 to 150 pA, showing unform ADC code utilization. (Middle) Differetial Nonlinearity of the ADC, showing less than 1 Least Significat Bit (LSB) of variation. (Bottom) Integral Non-linearity (INL) of the ADC, also genrally less than 1 LSB.

[00037] Figure 14C shows sensor chip deployment instrument used for running experiments in The disclosure. (Left) The 16k sensor CMOS chip is mounted in small form- factor custom-designed instrument. The chip is wirebonded to a PCB interposer board, which mounts in a flow-cell socket in the center of the motherboard. (Right) System control software (seen on screen) runs on an adjacent laptop computer (not shown), and data are transferred off-instrument to computer for analysis.

[00038] Figure 14D a schematicl of an exemplary CMOS Sensor Pixel Basic

Architecture Detail: The major organizational elements of the pixel are as follows, corresponding to the pixel level circuit schematic illustration shown: (1) Each sensor element has a so-called “drain” and “source” electrode, which are spanned by the molecular bridge, and which are distinguished functionally in the circuit in that the actual current measured is the current flowing into the drain: as shown, a switched capacitor transimpedance amplifier is tied to the drain, which integrates the current onto a capacitor while maintaining a constant voltage differential across the source-drain electrode pair. (2) The transimpedance amplifier has a typical gain of 5mV/pA (or 5 Giga-ohm) and dynamic range of lOOpA with a root mean square (rms) noise floor of 0.4pA. A high dynamic range option converts the range to 400pA. (3) When the row select switch is on, the stored voltage on the capacitor is transferred to the column ADC for digitization.

[00039] Figure 14E is an Annotated die image for the 16k CMOS sensor array chip. The 16k pixel array is organized as 4 banks of 4k pixels. Readout uses 128 column-pitch- matched ADCs on top and bottom of the device, to read out a full row of 256 pixels, with the Row Selection Control at left. Serializers on top and bottom output the bit data at high speed, supporting the 1000 fps readout.

[00040] Figure 14F illustrates an exemplary CMOS Sensor Array Chip Architecture: The major architectural elements of the sensor array chip are as follows, corresponding to the chip-level circuit schematic illustration shown: (1) The chip sensor array organized as 64 rows x 256 columns of pixels (pixel inset shown, and detailed in Figure 14E) (2) Each column is digitized by an ADC operating at 64K samples/second. (3) 16 ADCs are daisy chained and serialized into a single CMOS 1.8V output at (64Kxl6) 1M samples/second implying 11Mbps (bit depth of 11 bits/sample) (4) This in turn can be considered one unit cell of the chip; this unit cell is repeated 16 times to create the 64x256 array. This implies there are 16 parallel output lanes that carry data to the off-chip FPGA, which handles primary off-chip data transfer. (5) There are 256 ADCs in the chip, and they are time interleaved between the 64 rows. (6) The row decoder block is central to the chip and generates the reset and row select control signals for all the rows and facilitates the time interleaved digitization of the 64 rows by 256 different ADCs.

[00041] Figure 15A-15E show DNA Binding Signal Trace analysis using a Hidden Markov Model (HMM). Example showing how an HMM can be fit to a current-time trace to quantify information about single-molecule interaction kinetics. Figuer 15A is a raw data trace from the DNA binding probe sensor of Figure 2 (1kHz raw sampling rate). The 2-state HMM model fit to the data is shown in Figure 15B (black dashed line) and enlarged in Figure 15C to show the to (waiting time) and xi (dwell time) parameters for each binding event. Histograms of the observed To and xi times are shown in Fiures 15D and 15E, affirming that both waiting and dwell fit well to an exponential distribution (darker green line), as expected for a first-order stochastic process obeying Boltzmann kinetics (also see Figure 16).

[00042] Figure 16 shows DNA Binding Signal Trace analysis using a Hidden Markov Model (HMM). In this example, the analysis was of an entire phase (450 seconds) of the DNA probe binding its target at one concentration. Here, the bridge had a 45mer attached at its 5’ end (5’-CGATCAGGCCTTCACAGAGGAAGTATCCTGTCGTTTAGCATACCC). The target oligo binding was an 18mer (5’-TCTGTGAAGGCCTGATCG) at 10 nM concentration. Running the HMM on the entire trace extracts 2049 binding events. The resulting summary statistics include (see Table SI) this count of states, an estimate the fraction of time bound, 0.0957, the average waiting time of 0.101 seconds and the average dwell time of 0.0107 s. These in turn allow estimation of the kinetic rates, k₀ff = 1/0.0107 = 93 sec ^_1 and k₀n = 1/ (0.101*1X10⁷) = 9.9 x 10⁷ M ¹ sec ¹ (this makes use of the fact that in an ideal exponential distribution the mean of all the x_n states is equal to the reciprocal of the decay constant). More sophisticated analyses may be needed if the distribution diverges far from the ideal case. The temperature for this experiment was not controlled but can be estimated to be effectively >50°C, based on the oligo melting point and bound fraction.

[00043] Figure 17 represents Table 1 which is the statistics derived from the data represented in Figure 16.

[00044] Figure 18 shows the Exponential Distribution of Binding Parameters for DNA Binding Probe. Data is from the Binding of the 14-mer DNA target to the 17-mer DNA attached to the sensor bridge, as in Figure 14. The dwell and waiting times for a total of 44,500 binding events like those shown in Figure 14 were plotted as distributions to assess fit to an exponential distribution. Both the distributions conform closely to the expected exponential as shown on this semi-log plot, spanning nearly 3 decades, indicating that the binding process in consistent with a single molecule binding reaction with a constant reaction rate, and that the sensor is not substantially distorting the measurements. The temperature of the chip in this experiment was not controlled or recorded but is effectively over 50° C based on the dwell times. From these fits, the mean dwell time is 8.2 ms, and the kinetic on/off rates are k_0n = 3.1 X10⁹ M V¹ and k_0ff = 122 ± 5 sec ¹.

[00045] Figure 19 shows DNA Binding Probe Response to Titration of Salt Levels. The DNA binding probe and target are the same as in Figure 14, the 17-mer probe and 14- mer target, with target concentration of 20 nM. The salt concentration of KC1 is here titrated from 2 mM to 2000 mM. From resulting traces, kinetic parameters were estimated dwell time (defined as pulse width) and pulse height (a measure of signal strength, defined as the difference between state 0 and state 1 mean currents). As shown, dwell times did not vary significantly with salt, while signal strength in terms of pulse height dropped by approximately 70% in going from 2 mM and 2000 mM, which corresponds to dropping the solution Debye length (electric field decay length) from nearly 10 nm at low salt, down to below 0.3 nm at high salt (solution Debye Length is shown on the upper axes). Results suggest that a substantial component of the signal is due to a field effect arising from charges more than 0.3nm from the bridge, which is screened out as the Debye length decreases.

[00046] Figure 20 represents Table 2 which is varius control targets in the Covid-19 moch assay (see Figre 13).

[00047] Figures 21A-21F illustrates ribbon models of molecular electronics sensors used in the disclosed sensors which are intended to provide more visual detail for complex probe molecules.

[00048] Figure 20 illustrates DNA Hybridization sensor in which an ssDNA Oligo probe is conjugated to the attachment site of the 25nm alpha-helical peptide.

[00049] Figure 21B-1 Shows DNA Aptamer Probe. A ssDNA Aptamer probe is conjugated to the attachment site of the 25nm alpha-helical peptide. Shown is the Aptamer and Target for the SARS-CoV-2 S-Protein.

[00050] Figure 21B-2 Shows DNA Aptamer Probe. A ssDNA Aptamer probe is conjugated to the attachment site of the 25nm alpha-helical peptide. Shown is the 94-mer Aptamer and Target for the SARS-CoV-2 N-Protein (N protein dimerization domain shown, PDB 6yun).

[00051] Figure 21C illustrates an Antigen-Antibody Probe. The antigen, here a fluorescein molecule, is tethered to attachment site of the 25nm alpha-helical peptide, using a DNA oligo tether. Shown for perspective is a model Antibody target. [00052] Figure 21D shows (Upper) CRSPR Casl2a Probe for dsDNA target. A guide RNA for a Casl2a enzyme is tethered to the attachment site of the 25nm alpha-helical peptide (Lower), at a site on the RNA that is the most exposed base of the pseudo-knot loop. The Casl2a enzyme is then in turn tethered to the bridge through its docking to the guide RNA. Shown is the resulting casl2a-guide RNA complex interacting with its specific target dsDNA.

[00053] Figure 21E shows DNA Polymerase probe for detecting nucleotide incorporation activity. A Phi29 DNA polymerase is conjugated to the attachment site of the 25nm alpha-helical peptide, using a SpyTag SpyCatcher conjugation scheme. The spy tag peptide is conjugated to the attachment site on the bridge, and the SpyCatcher motif is fused to the N-terminus of the Phi29 polymerase. Shown is the resulting complex further bound to a primed template DNA and interacting with an incoming nucleotide.

[00054] Figure 21F shows DNA Fig. S10.D Small Molecule Binding Probe:

Nucleotide Binding in the active pocket of a Klenow DNA Polymerase: a ssDNA template oligo (gold) is conjugated to the attachment site of the 25nm alpha-helical peptide, and a primer oligo is hybridized to this (brown). A Klenow polymerase is recruited to the primer site. Using a non-catalytic buffer (Sr++ replacing Mg++), this complex interacts with an incoming matched nucleotide (dCTP shown), and in the non-catalytic state, this small molecule binding transiently in the pocket. Thus, this probe complex acts to detect the small molecule binding on a target dNTP.

[00055] Figure 22 shows TABLE 3 which is the Pulse Metrics and Statistics for the

Phi29 DNA Polymerase Sensor Pulse Analysis shown in Fig. 6. The features of a current pulse segmented from the current-vs-time trace are indicated, which are: min within the pulse, max within the pulse, mean of the pulse, standard deviation within the pulse, time width of the pulse, extrema within the pulse, and waiting time to the next pulse. These 7 features of each pulse were the inputs to a Principal Component Analysis (PCA), for which the top two components (PCI, PC2) are shown in Fig. 6. The table below shows the means and deviations of metrics for the two major pulse clusters shown in Fig. 6.

DETAILED DESCRIPTION

[00056] Manufacturable Molecular Electronics Sensor CMOS Chip - The disclosed chip embodiment is a CMOS integrated circuit chip that supports an array of nano-electrode- based molecular electronics sensors, as illustrated and described in Figures 1 and 2. These sensors are based on a molecular wire bridge, spanning typically a 20-nanometer gap between the tips of nanoelectrodes, which themselves can connect into a current meter circuit for real-time measurement of the current vs. time flowing through the molecular element. For a specific sensor application, the bridge molecule can be conjugated in a site-specific manner to a biomolecule of interest, which acts as a probe for its molecular interactions with target molecules. This results in observed current pulses that represent these dynamic interactions. Since these sensors are isolated single molecules, they reveal binding events discretely and with high sensitivity, without the need for labels. The result is a direct electrical measurement of molecular interactions with corresponding kinetics. The sensor chip operates with millisecond time resolution.

[00057] Each sensor circuit (or “pixel”) is a dedicated current meter, may be implemented as a CMOS transimpedance amplifier that amplifies pico-Amp (pA) scale currents to milli-Volt (mV) scale voltages, specifically with a 5 Giga-Ohm gain, so that each pico-Amp of current through the molecule is amplified to 5mV of output voltage. The amplifier circuit in effect measures the current flowing through the biomolecule, and the chip transfers the measured currents from the sensor array in digital format off chip for subsequent analysis, at a rate of 1000 frames per second, and with 10 bits of measurement precision. The dynamic range of measurement is from 0 to 400 pA, and the amplifier is designed to be highly linear over this range, with sub-picoamp noise (Figures 1A, B), to preserve the detailed shapes of the sensor pulses for subsequent analysis. In one application, the sensor circuits are arrayed on chip with a 20-micron pitch, for a total array of 16,384 (16k) sensors in an area that is just 1.2 mm by 4.8 mm. The present CMOS chip is implemented in a 180nm CMOS node.

[00058] As shown in Figure 2, the nano-electrodes are fabricated using CMOS foundry-compatible photolithography processes and tools, for full integration of manufacturing into a commercial CMOS chip foundry. Further details of the chip architecture and performance parameters are provided in Figures 14 C-F.

[00059] The disclosed chip architecture and manufacturing methods readily extend to larger chips that can provide over 1 million pixels in 180nm CMOS. The 1 -million-pixel scale supports future ultra-high-throughput applications, such as sequencing a whole human genome on one chip, in less than 1 hour. Subsequent scaling to finer CMOS nodes (the 180nm node was introduced commercially in 1999, and while still heavily used, current commercial state-of-the-art is the 5nm node, representing a 1000-fold increase in transistor density), as well as advances in the pixel circuit design, can support future scaling to deep sub-micron sensor pixels, and thus potentially a 10,000-fold increase in future sensor density on chip, with the implied opportunities for comparable chip cost or size reductions. There is no need to alter the nanoelectrode and molecular sensor structure to gain this size reduction - it only requires the industry standard methods of shrinking circuits that have provided 50 years of Moore’s Law, and which moreover are already provided by the existing foundries and chip design tools.

[00060] For detailed investigation of the chip and sensor performance, the disclosure presents the well-studied model system of a single-stranded DNA oligo as a probe for hybridization to its complementary strand. This is a useful reference system, since oligo binding has been extensively studied, both empirically and theoretically, given it is the basis for classical ensemble hybridization assays such as the so-called Southern blots and DNA microarrays, as well as the basis for the primer binding process that drives PCR, and is also important in fundamental biology and other diverse applications of DNA in biotechnology. Oligo binding has also been extensively studied experimentally at the single molecule level using carbon nanotube sensors.

[00061] To illustrate the breadth of different types of molecular interactions that can be deployed on the platform, the disclosure demonstrates sensors that detect binding of protein molecules, small molecules, aptamers and antibodies, and well as probes to monitor the activity of enzymes that are important for both applications and fundamental biology, DNA polymerase and CRSPR Cas enzymes. All these sensors have a common format, of a probe molecule precisely conjugated to the molecular wire - which here is a synthetic peptide with a 25 nm long alpha-helical conformation - and with the probe having an associated specific target molecule.

[00062] Figure 3 visually summarizes the molecular electronic sensors described in The disclosure. Figures 21A-21F also show more detailed ribbon models of all these molecular electronic sensor constructs.

[00063] Results [00064] The experiments reported here require attaching bridge molecules to metal electrodes on chips with nanoelectrode spacings in the desired range of 15 nm to 20 nm. The molecular bridge used is a proprietary 25 nm alpha-helical peptide, with specific metal binding peptide sequences at both ends. The end groups allow it to self-assemble onto the electrodes under proper conditions. When the bridge molecule is delivered to the nanoelectrodes by passive diffusion, binding spanning the gaps can take 24 hours (using 20 nM peptide concentration and dilute buffer (2 mM Tris pH 7.5)). However, an active bridging protocol uses electrical forces to attract the bridge to the electrodes, to radically accelerate and enhance this assembly process, allowing assembly to be completed in seconds. Dielectrophoresis is a highly efficient process for using electrical forces to attract molecules or nanoparticles to the gap between micro- or nano-electrodes. The dielectrophoretic trapping protocol relies on the application of an AC voltage (here 100 kHz, 1.6 V peak-to- peak). These voltages are simultaneously applied to all nanoelectrodes on the chip, by switching in special on-chip AC driving circuits. Dielectrophoretic trapping readily shortens bridging time to 10 seconds (10,000-fold), while simultaneously working at 1000-fold lower input concentrations of bridge molecules. The effective concentration increase of the bridges near the gaps is thus on the order of at least one million-fold. To assess effective bridging, the DC current on sensor the after bridging is compared with the value prior, and a sufficient jump in current indicative of successful bridging. A population of sensors showing substantial bridging current increases is thereby observed, typically over 10% of all available pixels on the chip, indicating the presence of the 25 nm peptide bridge spanning the electrode gap.

[00065] DNA-DNA hybridization binding and Sensor Validation - The chip and sensor performance can be validated using the well-studied model system of a single-stranded DNA oligo as a probe for hybridization to its complementary strand. This can be a useful reference system, since oligo binding has been extensively studied, both empirically and theoretically, including at the single molecule level using carbon nanotube sensors. The probe molecule attached to the molecular wire bridge for the work can be a single-stranded 17-mer DNA oligonucleotide with a specific sequence as seen in Figure 4.

[00066] The specific binding target for this probe is the complementary sequence DNA oligomer. This DNA probe was conjugated the bridge molecule in a precision site-specific manner to the central amino acid on the bridge, using conventional alkyne/azide copper-less click chemistry and purified by HPLC for application on chip. Once the bridge molecules (with probe) are attached to the electrodes of the chip, baseline current is measured, which is typically steady within a range of a few pA when a voltage of 700-1000 mV is applied. While continuing to monitor sensor currents, the “target” oligonucleotide is added at a particular concentration.

[00067] As shown in Figure 4, the sensor on chip responds to the presence of the target in solution with current pulses of that can be interpreted as individual DNA binding events with the DNA strand attached to the sensor, in a dynamic equilibrium between bound and un-bound states.

[00068] Control experiments (Figure 5) revealed that this binding changes with the target concentration and (Figure 6) temperature as would be expected for DNA-DNA binding. It is important to note that while various summary statistics described below provide the classical kinetic rate parameters, the complete time trace contains extremely rich information about the molecular interaction, such as possibly indications of multiple conformations and partial interactions. [00069] Analysis of single-molecule binding data using Hidden Markov Models

(HMM) - In some embodiments, HMM methodology is used to quantify the primary signal traces. HMM have previously been successfully applied to timeseries data from single molecule biophysics experiments. In some embodiments, the HMM assigns the “hidden” bound and unbound states of the sensor to segments of the observed signal that have statistically different current levels, with lower currents for the unbound state and higher currents for the bound state. Figures 15 - 18 show a detailed HMM analysis of the data shown in Figure 4. In this case, the unbound state is identified as a low-current range of ~30 pA and the bound state is identified as the high current range of 50 pA — 70 pA. The fundamental time parameters extracted from the HMM segmented signal trace are the individual waiting times between binding events, TO, and the individual dwell times or time spent bound, Tl

(Figure 15). As shown in the empirical distribution (Figure 18), the shorter duration states are more numerous, obeying an exponential dirtribution as expected for 2-state first-order chemical reaction kinetics. While these times are ideally exponentially distributed, note the complete empirical distributions contains richer information about the more complex nature of the binding interactions. From these fundamental parameters, it is possible to compute standard binding kinetic parameters of interest: the off rate, k_0ff, the on rate, k_on. In addition, the total fraction of time spent bound (denoted “Fraction Bound” in all figures) is a convenient summary statistic that is readily related to the concentration of the target in solution. The expectation is that the fraction of time bound is proprtional to concentration at low concentration, and saturates (ideally at 1, but in practice at some full-occupation value) at high concentration. These characteristic saturation curves are evident in the data shown. These curves also allow calculation of the binding affinity of the interaction, Kd, which at the single-molecule level is defined here as the target concentration at which the single probe molecule spends equal time bound and unbound. Note K_d can be visualized and calculated as the inflection point in curves where fraction of time bound is titrated against target concentration, as shown throughout The disclosure. The fraction of time spent in the bound and unbound states can also be conveniently visualized using vertical histograms of all the measured current values in a signal of segment, as shown to the right of the traces in Figures 4B and 5. The details of these histograms may also contain of richer information about the interactions, such as indications of additional bound states conformations.

[00070] Single Molecule Thermodynamics: Melting Curves - Another application of this same type of assay is to determine the melting temperature (T_m) of the DNA duplex, which is defined here at the single-molecule level as the temperature at which the probe DNA molecule spends equal amounts of time in the bound and unbound states. This is directly observable in the single molecule binding traces. As shown in Fig. 6, measuring of fraction bound at eight temperatures allows fitting of the data to a classical DNA hybridization melting curve. Here measured temperature values shown are temperatures taken on the chip die itself, using on-chip temperature sensors, as a Peltier-drive heating plate in direct contact with the die is used to set to different temperatures (and allowing time to equilibrate to each new temperature) in succession in one chip experiment. From these curves it is clear that a 20-mer target oligo melts at a higher temperature than 15-mer segment of this 20-mer oligo, as expected. Determination of T_m serves to validate that this is measurement of the DNA- DNA hybridization binding reaction as intended, and also has practical value in selecting the experimental operating temperature for using such a probe to make concentration measurements. As is done classically, this single-molecule melting curve be used to increase the signal-to-noise for detecting the target of interest, or to characterize details of targets that contain mismatches. [00071] Mismatch Sensitivity - The single-molecule binding probe signal trace contains rich information about the binding reaction and is also highly sensitive to the specific binding target. This can be illustrated in fine detail for DNA oligo binding by looking at the impact of single-base mismatches in the target oligo sequence.

[00072] As shown in Figures 7 and S5, several variants of the exact match 20-mer target sequence were made, introducing from 0 to 3 mismatched bases. Results from applying these variant targets to the probe, and measuring bound fraction, are shown. With each additional mismatch, the bound fraction decreases. These differences are readily measured and can be further magnified by performing the measurements at a temperature nearer to the T_m of the matched target, or by performing a temperature melting curve as in Figure 6.

[00073] The binding dwell time is also reduced by the presence of mismatches, and can provide a comparable — and independent — indicator (data not shown). This sensitivity to mismatches can have applications for DNA binding assays in which sequence variants relative to a reference sequence probe might be of interest, such as in detecting novel strains of a viral genome, or detecting diverse somatic mutations in a cancer genome, or in detecting SNP genotype variants. The probe of Figure 7 has the following sequence:

[00074]

[00075] Sensor Signal Generation Mechanism - Carbon nanotube biophysics experiments with a DNA probe as well as enzyme probes show that the dominant signal generation mechanism in the single-walled carbon nanotube is a field-effect, where electrical fields emanating from the target-probe molecule complex gate the flow of current in the nanotube wire in an action-at-a-di stance manner. This is demonstrated by increasing salt concentration to screen out fields in the solution (reduce solution Debye length), and showing the signal is lost for sources more distant than the screening (Debye) length. For the present molecular electronic sensor, similar experiments were run to examine the effect of increasing salt concentration in the binding buffer on both signal strength and dwell time, with salt concentration (KC1) increasing from 2 mM to 2000 mM (Debye length decreasing from 6.8 nm to 0.22 nm). As shown in Figure 19, the signal strengths (recorded as pulse heights (in pA)) decrease by nearly 70% with salt concentration, suggesting a major portion of the signal is due to a field effect, but not as much as seen on the carbon nanotube devices. This suggests that there is another component of the signal not coming from a gating field-effect on the bridge, which persists in the presence of very strong screening of fields (0.22 nm Debye length). This result suggests that while the dominant part of the current modulations observed are due to the field effect demonstrated for the carbon nanotube sensors, there may be other minor forms of conduction modulation also involved. For example, these may relate to direct changes in electron conduction paths through the molecular complex, or to changes in the ion clouds in the solution around the molecules, or to electrochemical state changes during conduction, such as due to redox charge transfer reactions. Using this DNA-DNA binding model system, further investigation is underway to fully characterize the detailed conduction modulation mechanisms and the controlling system parameters. One benefit is that the sensor allows for measurements of binding even in high salt conditions, much higher than is possible for the carbon nanotube sensors. [00076] Protein and Small Molecule Binding - Figures 8A and 8B show two binding processes fundamentally related to DNA polymerase (indicated at the molecular level in the top illustration): the binding of the protein to a 3’ -OH primer site, and the nucleotide substrate binding in the active pocket of the enzyme. These particular sensor modalities are useful for the studies of DNA polymerases, but they also serve to generally illustrate detection of protein binding (here polymerase docking to a priming site) and small molecule binding (here a dNTP interacting with the polymerase binding pocket). This could be considered as a model for a small molecule drug interacting with a binding pocket on a protein target, and indeed nucleotide derivatives are used as drugs in certain contexts, such as blocking DNA replication in chemotherapy or antiviral therapy. To construct the present sensor, a DNA template oligo is first tethered to bridge, as above in the DNA binding studies. A complementary primer oligo is then bound to this, to form a primer site, which can act as a probe for binding a polymerase.

[00077] Figure 8 A shows the resulting titration curve for Klenow DNA polymerase binding to the primer/template complex on the bridge. This produces a typical saturation curve, and a resulting binding constant Kd. The final process, shown in Figure 8B, is the binding of a nucleotide substrate to the binary complex of primer-polymerase. This is performed in a non-catalytic buffer (lacking Mg²⁺), such that the Klenow DNA polymerase used here cannot incorporate the nucleotide, which instead maintains a dynamic equilibrium entering and existing the binding pocket and producing a corresponding binding signal trace.

[00078] Aptamer - Target Binding - The single stranded DNA oligomer probe on the bridge can also be a DNA aptamer, which is a DNA oligomer with a sequence that is empirically selected to bind a specific molecule of interest, such as a protein. These DNA (or RNA) aptamer probes can be attached to the sensor peptide in exactly the same manner as the oligonucleotides used for binding probes. This provides a general class of aptamer binding probes with a great diversity of possible targets, and in particular these are well suited for rapidly developing binding probes for protein targets, for use in protein detection and identification, and proteome characterization and profiling.

[00079] One particularly timely application of such protein binding aptamers is in testing for infectious disease pathogens. In particular, The COVID-19 pandemic has highlighted the need for cost-effective, highly specific, rapid and distributed testing for viruses. Aptamers against the SARS-CoV-2 N protein and the S protein have been described in the literature, which give the selective and high affinity binding necessary for a diagnostic test. Antigen tests have been developed using these aptamers which show high accuracy in detecting SARS-CoV-2 infection. To test the use of aptamers on the present sensor chip platform, both the N and S protein aptamers were attached to the bridge to study their binding kinetics against SARS-CoV-2 proteins. After attachment of these aptamer-functionalized bridges to respective sensor chips, the corresponding N or S protein were titrated in solution on the chip and the fraction bound time was calculated as previously described. The plot of the fraction bound against the protein concentration is shown in Figure 9.

[00080] From these plots the binding kinetics for the N-protein were k_0ff = 40 sec ¹ and kon = 6.2 X 10⁸ M ¹ sec ¹ with a kinetically calculated Kd of 64 nM. Similarly, for S-protein koff = 35 sec ¹ and k_0n = 3.3 X 10⁹ M ¹ sec ¹ with a kinetically calculated Kd of 6.1 nM.

[00081] The detection of SARS-CoV-2 spike protein by single molecule binding to aptamers on the molecular electronics sensor illustrates the possibility for a rapid diagnostic test for viral infection, ideally suited for low-cost testing deployed on highly compact point - of-care or home-use systems. Further, the massively parallel and scalable nature of these sensor arrays allows for many targets to be screened in a single chip run. This enables a highly multiplexed testing where targets from any number of viruses or viral variants can be screened for in a single test. In contrast, multiplexing of more than a few targets is not possible in the standard lateral flow or qPCR methods presently used for viral testing.

[00082] Antibody - Antigen Binding - To demonstrate the application to antibody- antigen binding, one option is to place the antigen on the bridge and observe the binding to the antibody present in solution. A convenient model system for this is to use a fluorescein dye molecule as the antigen, and a commercially available anti-fluorescein antibody, as illustrated in Figures 10.

[00083] The fluorescein is conveniently mounted on the bridge using the same DNA oligo probe attachment chemistry as described above, with a synthetic DNA oligonucleotide synthesized to have the fluorescein dye at its 3’ (distal) end, thereby presenting the antigen. A commercially available anti-fluorescein antibody was titrated on this chip and binding activity was readily observed. The results are shown in Figure 10 with the apparent Kd being about 1.3 mM. In general antigens or antibodies can be attached to the bridge to observe antibody-antigen binding. A DNA oligo tether such as used here can provide one convenient means of conjugating antigen molecules to the bridge in particular cases, but many well- known methods of conjugating can be used, including the conjugation methods that may already be present on existing antigen or antibody libraries.

[00084] CRSPR Cas Enzyme Activity - The CRISPR-Cas enzymes originally used for gene editing have recently been proposed at tools for sensitive DNA detection for diagnostics and other applications. In general, these CRSPR Cas enzymes bind to a short guide RNA strand that serves to program them to a sequence-specific form, and then subsequently bind to and cleave the specific target DNA strand. The ability to monitor the single-molecule kinetics of these fascinating enzymes could in general be useful to understanding their multiple complex activities, and also may help in enzyme evolution studies to provide high throughput screening for useful mutant phenotypes. In addition, such enzymes have potential uses for diagnostics, based on monitoring for indications that the programmed enzyme has bound its specific target. The originally discovered Cas9 enzyme has been widely adopted for gene editing functions, but more recently discovered Cas enzyme families such as Casl2, Casl3 and Casl4 have been proposed for diagnostics, since they undergo more dramatic and readily detectible transformations after encountering their target, and therefore simplify the optical reporter methods.

[00085] The disclosed single molecule sensor is capable of observing the primary DNA target capture, and thus any of these enzymes could potentially be used diagnostically and in a highly multiplex target fashion, on these sensor array chips. Shown in Figure 11 and 21 are results from binding experiments using the CRSPR Casl2a enzyme, which is commonly used as the basis for such diagnostics, programmed by a guide RNA designed to detect a 20 base DNA sequence taken from the S gene of the SARS-CoV-2 virus.

[00086] For this purpose, as illustrated in the Figure 11 and in Figure 21D in more detail, the guide RNA was conjugated to the peptide bridge, and several trial configurations were studied, wherein specific sites in the RNA were conjugated to the bridge. Resulting guide RNA bridges were assembled onto the chip, which was then first used to observe titration of the Casl2a enzyme binding to the guide RNA, over a protein concentration range of 0-1 mM. Binding was observed with a Kd of 40 nM when the RNA was attached to the base of the 13^th nucleotide of the guide RNA (which corresponds to the outermost exposed point in the pseudo-knot loop, Fig. 21D) but was not observed when the attachment was at the 3’ terminus of the guide RNA. It is likely that this configuration results in steric hinderance to Cas binding, based on the known Casl2a-guide RNA structure (Fig. 11, 21D). When the chip with gRNA was then incubated with 20 nM Cast 2a to bind and program the enzyme, and then titrated with 0.1-64 pM target dsDNA, an apparent dissociation constant for the target dsDNA of 3 pM was observed, similar to the results reported elsewhere for similar enzymes. By tethering guide RNAs to the bridges, this provides a means to deploy multiplex targeted CRSPR Cas enzymes on chip, and monitoring their primary detection activity in real-time and in parallel. This could provide for highly multiplex CRSPR diagnostics, deployable on a highly advantageous chip-based format. This could also provide for massively parallel mutant enzyme phenotype screening.

[00087] DNA Polymerase Enzyme Incorporation Activity - In this case, a single DNA polymerase molecule is conjugated to the molecular bridge, using a site-specific conjugation method. The resulting sensor chip device is provided with a primed DNA template, and the required dNTPs, for the polymerase to extend the primer on the template. This sensor configuration, illustrated in Fig. 6 (and in greater detail in Fig. 2 IE) allows single-molecule observation of the polymerase activity as it incorporates nucleotides in real-time. Exemplary results are summarized in Figure 12 and Table 3 which shows a 25-second portion of the experiment that has about 40 pulses with dwell times longer than ~10 ms along with about four times as many brief ones. Analysis of this signal trace suggests that each major pulse represents a nucleotide binding and incorporation event, consistent with the known behavior of DNA polymerase and carbon nanotube observations of polymerase activity.

[00088] Details of the pulse shape actually seem to discriminate the dATP binding and incorporation events from the dCTP events. For the DNA template here, there should be 15 dC incorporations followed by 25 dA incorporations. Visual inspection suggests the pulse height, pulse width and inter-pulse spacing are important factors in discriminating putative A pulses from the putative C pulses of the trace. Shown in Figures 12B and 12C is a more extensive pulse discrimination analysis, where 7 features of each pulse are automatically extracted and put into a Principal Component Analysis (PCA). When the pulses from the trace are displayed in the two dominant principal component space (Figure 12B), they fall into two nearly disjoint clusters based on these shape features that also correspond to the pulses from the putative C (left) segment (coded Red) and A (right) segment (coded Green) of the signal trace. Thus, the pulses are highly distinct in their features, consistent with the hypothesis that A and C produce distinctive pulses.

[00089] The loading chart at right (Figure 12C) show the relative contribution of the shape features to the first principal component, and make it clear the time between pulses as well as amplitude and width are dominant distinguishing factors. This example illustrates the potential for constructing a sequencing sensor, based on monitoring the detailed single molecule kinetics of a DNA polymerase as it copies a template.

[00090] Sensing in Complex Backgrounds: A Model Viral Detection Assay - The basic detection of molecular interactions demonstrated above can be used to develop many applications. One such example is viral detection, which has been highly relevant to the COVID-19 pandemic. However, for use in practical diagnostic tests, it is important that a sensor be able to reject complex backgrounds of off-target molecules, and it is even more ideal that the sensor could actually function in crude samples such as saliva. Here it is demonstrated that these molecular electronics sensors, even though they have the extreme sensitivity to detect single target molecule binding, can reject highly complex backgrounds and perform in crude saliva samples.

[00091] The CDC has issued a recommendation for a qPCR test for the Covid-19 infection, based on testing for the presence of two sequences in the N-gene. The sensor chip was prepared with a DNA oligo probe that would target one strand of the PCR product. PCR products were made using the CDC forward and reverse primers for SAR-CoV-2, with a synthetic plasmid for the N gene serving as the positive control target template, which was spiked into contrived samples. This resulting PCR product, unpurified, was applied to the chip in various ways to assess the ability of the molecular electronics DNA hybridization sensor to reject complex backgrounds and still detect its specific target, as well as to work with the complex mixtures produced by PCR, all of which arise in actual diagnostic tests. For the oligo probe on the bridge, the present study uses a 21-mer (CCGCATTACGTTTGGTGGACC) taken from the CDC qPCR TaqMan probe sequence (2019-nCoV_N 1 -P; F AM- ACC CCG CAT TAC GTT TGG TGG ACC-BHQ1R77T Various target sequences to test against this probe are shown in Table S2, and these were synthesized for study in The disclosure.

[00092] Each of these targets (as listed in Table 2 rows a-g ) were tested for binding on chip and found to conform to positive and negative controls as expected. Only the last example (double-stranded PCR product) displayed weaker binding than the others, suggesting it has partially re-natured (thereby lowering the concentration of target single strands) prior to testing on chip.

[00093] The targets were tested for binding on chip at concentrations of 10 pM, 100 pM, 1 nM, and 10 nM, in buffer A. The PCR products registered similar sensor responses to pure oligo target samples, even though they represent a much less pure sample owing to PCR off-target byproducts and reagents. In order to further assess the impact of complex background materials on this highly sensitive sensor, buffer A was mixed with heat- inactivated human saliva (from 10% by volume to 50% by volume), and in another approach it was mixed with a high concentration of highly complex background DNA: salmon sperm DNA added at a concentration typical for DNA contamination of saliva (2 pg/ml). Neither one of these crude sample challenges had substantial impact on the sensor readout, showing that the sensor is highly specific for its target and robust against complex and even crude saliva samples.

[00094] In terms of the overall sensitivity of the sensor, note that at 100 pM. target concentration, the signal consisted of pulses of about 6 pA magnitude occurring about 8 times per second and lasting 5% of the total time. This corresponds to a sample of about 1 fmol of oligonucleotide applied to the chip. This is, however, not close to the fundamental limit of on-chip detection, since these strong binding pulse events of single molecules are ultimately counted to obtain the information about fraction bound and dwell time. Therefore, at lower concentrations, if sensors are observed for longer time, or multiple sensors for the same target are observed in parallel, it is possible to observe a statistically meaningful number of binding events with much lower target concentration, providing a powerful means to reduce the limits of detection in such a test. If means of electronically concentrating target near the sensor are also applied, such as the dielectrophoretic trapping force used to attract the bridges to the electrodes, it may be possible to approach the limits of single molecule detection in the primary sample, without PCR amplification.

[00095] Discussion & Analysis - As a fundamental new sensor platform fulfilling a 50- year-old technology vision, the molecular electronics sensor chip presented here has a number of novel features, as well as broad potential for future applications. In particular, as a sensor platform, it is unique in its combination of universality, scalability, and single molecule sensitivity, while the CMOS chip format also provides for manufacturable realization of sensitive, multiplex, rapid, low-cost tests, on compact instruments. These combined features enable near-term technology disruption in applications from drug discovery to diagnostics to DNA sequencing, and moreover, also provide each of these with a long-term, faster-than-Moore’s Law scaling path to ever lower costs and greater speeds, leading to highly durable technology solutions. These points are briefly discussed here:

[00096] Sensitivity - The molecular electronics sensor has intrinsic single-molecule detection sensitivity. This is a relatively unique capability in biosensing, where methods that have single-molecule detection capability often rely on biological signal amplification (such as in PCR or ELISA assays) to increase the signal to the point where it can be detected by a detector that does not have single molecule sensitivity. Having the platform based on a true single-molecule sensor therefore provides the potential for the ultimate limits of sensitive detection, with or possibly without combination with biological signal amplification for various assays.

[00097] It is known that such highly sensitive electronic amplifier systems can experience “random telegraph noise” (RTN) resulting from, for example, single trapped charges changing state. Indeed, the signal pulse trains produced by the molecular electronics sensor (e.g., Figures 4, 5) generally look similar to RTN. Such spurious RTN noise modes do indeed appear sporadically on the present type of sensors, and therefore it is important that experiments be designed with controls to distinguish these spurious noise modes from properly detected signals. One useful method is to titrate the concentration of the target and verify that the sensor pixel in question responds properly to this titration. For example, Figures 5 shows the output for different target concentrations for the DNA hybridization sensor. The fraction of time the probe DNA is bound responds properly, as concentration increases from lOnM (about 20% fraction bound), to 100 nM (50% fraction bound), tol mM (over 70% fraction bound). This can thereby rule out pixels subject to RTN that could otherwise confound measurements. In addition, the dwell times for RTN pulses will not match the expected dwell time or temperature distribution for the target of interest, and this can further be used to reject RTN artifacts. Such calibrations and quality controls should ideally be built into assay protocols.

[00098] Electronic Target Amplification - One particularly attractive feature of the present platform is the ability to electrically concentrate the target near the sensor, using electrophoretic or dielectrophoretic methods. As noted, this is used in the present work as the means of rapidly attracting the bridge molecule to assembly on the nanoelectrodes. In practice this attraction provides a million-fold rate enhancement for sensor assembly, in terms of concentration times reaction time (i.e., a lower concentration input is assembled in less time, as compared to passive diffusion transport). If this same electronic concentration ability can be applied to the concentrate targets near the sensor, it may radically increase the ability to detect low concentrations of target. This may in some cases eliminate the need for PCR in the case of nucleic acid detection. Moreover, for detections of proteins or other non- nucleic acid targets, this provides a powerful new electronic amplification technique for targets that cannot undergo physical amplification. Since the primary sensor already has single molecule sensitivity, this can enable other detection modalities to approach the single molecule sensitivity traditionally associated with PCR for detecting DNA.

[00099] Specificity - The molecular electronic sensors can provide highly specific detection, as long as the primary molecular interaction is specific. This is perhaps surprising, given the extreme sensitivity of the sensors. This specificity is best demonstrated with the DNA hybridization binding sensors, where the DNA binding reaction is in effect sequence- programmable in the strength of the interaction, and finely tuned off-target interactions can be studied by introducing mismatched bases in the target, as well as very large numbers of off-target interactions with complex background DNA. The sensors were challenged with such complex interactions to demonstrate the limits of specificity. [000100] As shown in Figure 7, a matching oligonucleotide target produces a signal trace readily distinguishable from one possessing even a single mis-matching nucleotide. The difference in affinity is readily apparent by observing fraction bound and also dwell time. This also has implications for practical assays, since such binding probes can therefore be used to discern the presence of a single-nucleotide polymorphisms for SNP genotyping, or to detect isolated mutations such as are relevant for viral strain detection or cancer mutation analysis. At the other extreme of specificity, the sensor demonstrated the ability of the probe to reject highly complex off-target backgrounds, such as genomic (salmon sperm) DNA and crude saliva. This also has implications for practical assays, enabling them to be robust and have simple, rapid sample preparation.

[000101] Multiplex Measurements - The universality of the sensor format was demonstrated, including enzymes (DNA polymerase and Casl2a nuclease), aptamers (for SARS-CoV-2 S- and N- proteins) and an antigen (fluorescein) with antibody target, as well as proteins and small molecules. The scalable pixel array chip therefore brings powerful and practically unlimited multiplexing capabilities to all these types of detection. Since each pixel is an independent sensor, even the present chip provides a capacity to multiplex from 1 up to 16,000 probes, and future scalability can readily take the upper limit to millions and far beyond (see below). As a practical matter, for single molecule measurements it may be desirable to have 10 — 100 replicates of each probe, so tens to hundreds of different probes could still be so represented on even the present chip. For example, it was shown that a DNA binding probe could detect a specific gene from SARS-CoV-2; through such multiplexing, similar DNA binding probes could target many different viruses or viral strains on one chip, applied to a single sample. [000102] Several methods are available to efficiently assemble such multiplex probe sets on the chip sensor array. The most desirable approach is to use voltage-directed assembly of probes into sensor pixels, such that a single probe-bridge solution is exposed to the chip, and the bridge molecules therein are attracted only to pixels that are supplied with a driving voltage.

[000103] This process can be done serially for each distinct probe type. Since it is fast for each trapping phase, requiring just seconds, tens, hundreds or even thousands of different probes could practically be assembled into the array using electrical “programming” of probe locations. An alternate approach that may be advatnageus for much higher levels of probe multiplexing is to pool the many different probe-bridges, assemble them at random into pixels, and then apply test samples to decode which sensor is present at which pixel sites. With well-known combinatorical decoding methods, N different probes can have their locations mapped in Log2(N) pooled decoding reactions (by pooling appropriate subsets of approximately N/2 probes in each decoding pool), so this process can be quite efficient for very large N. (For example, 1000 distinct probes can have their locations decoded with just 10 such reactions, and 1 million with 20). Finally, direct mechanical partitioning of the chip is also possible, using a gasket or other barriers, and different probe solutions applied to each zone. All these methods can be used singly or in combination to achieve efficient multiplex probe deposition, across all scales of multiplexing.

[000104] Finally, utilizing the full pixel capacity of such chips for unlimited massive multiplexing is readily applicable to DNA sequencing or Proteomic identification, where the anonymous single molecule targets are captured at each sensor site (DNA target or protein target, respectively) and there undergo many interrogation reactions (polymerase processing, or antibody/aptamer response profiling, respectively) to fully identify or characterize each anonymous DNA or protein target. Theses extremely high throughput applications can directly benefit from the multiplexing provided by future chips with 10 million, 100 million, or even 1 billion or more sensors.

[000105] Rapid Detection - The single molecule sensors demonstrated here expose the dynamic nature of single-molecule binding interactions. Because just a few seconds of data can survey enough binding events to gather quantitative statistics, this can enable extremely rapid measurements and rapid testing. In general, near the chemical equilibrium point of the interaction (e.g., melting point, T_m, for DNA-DNA binding), the rate of binding nearly equals the rate of un-binding, and numerous events can be observed in a short time, such as the span of seconds shown in Figure 3. Controlling key reaction variables such as temperature and target concentration can be used to adjust the interaction kinetics into such a regime, where just seconds of observation can identify and quantify targets of interest.

[000106] Low-Cost Tests - In many use cases, such as for diagnostics, the CMOS sensor chip would be a single-use disposable. Because of the economy of scale of CMOS chip manufacturing, CMOS chip are extremely low cost when produced at high volume, and therefore support low-cost testing. For example, circa 2021, in the 180nm CMOS node at medium-to-high volume, finished commercial CMOS 200mm wafers from a foundry cost in the range of $1000-$ 1400 per 200mm wafer, or approximately 4 cents per square millimeter. Finer nodes, such as 65nm, 22nm, and 7nm are only several fold more expensive, while enabling orders of magnitude higher sensor densities. As shown in Figure 2, each square millimeter can contain many thousands - and potentially many millions - of sensors (see below). Thus, molecular electronics chips enable extremely low-cost diagnostic tests, and perhaps even whole human genomes sequencing, to be performed on pennies worth of

CMOS. [000107] There have been recent proposals for penny-scale diagnostics relying on low- cost materials such as PDMS, paper, and ink-jet printed nanoparticles yet the extreme economics of CMOS manufacturing allow millions of highly sophisticated sensor circuits to be fabricated on pennies-worth of finished CMOS. The potential of CMOS to provide extremely low-cost diagnostics should not be ignored, when paired with universal, maximally scalable electronic sensors. In addition to low-cost production, it is important to note that production capacity is unparalleled: the global foundry capacity is estimated to be the equivalent of several hundred million such wafers per year, and the industry currently delivers over 1 trillion chips per year. Even for the most extreme imaginable high volume testing scenarios — such as a future pandemic where such chips are used to test the global population on a near-daily basis — the CMOS chip industry, uniquely, has the required manufacturing capacity. Even for diagnostic concepts based on simple, low-cost materials, in the absence of a manufacturing base it can take decades to reach these production scales

[000108] Highly Compact Instruments - The instrument needed to run CMOS chip- based assays can be compact, comparable in size to a portable computer or cell phone or USB sticks, so that diagnostic tests could be run at the site of use, such as medical point of care, or transportation hubs or other public sites, or in homes. For reference, the instrument used for the experimental work reported here is shown in FigureMC, which is smaller than a laptop and yet was not at all optimized for small size — the actual electronics required is less than that needed in a modern smart phone. This potential for ultra-small form factors could support novel environmental sensing methods, such as drone-deployed pathogen sensors actively surveying air or wastewater. Compact, low-cost electronic devices, no more complex than a digital thermometer, could also be suitable for home screening assays of general interest like health and wellness biomarkers panels, early indicators of disease, and or at-home diagnostics.

[000109] Scaling Limits - The disclosed sensor embodiments solve the More-than- Moore scaling problem: the physical extent of the sensors here is defined by the bridge, which is 25nm long (and which was chosen to be large enough to accommodate all biomolecules of interest as probes). This is already substantially smaller than the Minimum Metal Pitch (MMP— the closest possible metal electrode spacing, at the first contact layer used to make contact transistors) on all existing CMOS nodes (MPP is ~36nm on the state- of-the-art (circa 2021) 5nm CMOS nodes), as well as for all projected nodes spanning the next 10 years (3nm, 2nm, 1.4nm) (82, 83), and thus the size of the molecular element in no way limits the ability to shrink the CMOS pixel circuits (from the current 20 micron pitch reported here, rendered in a 180nm CMOS node) on current or foreseeable foundries. Indeed, allowing for reasonable engineering feasibility in existing foundries , future molecular electronic sensor could have a limiting pitch approaching 100 nm, which allows ample room for the nanoelectrodes contacting vias, and a number of transistors for circuit implementation. This corresponds to a density of 100 million sensors per square millimeter of CMOS. Thus chips the size of a square millimeter, costing pennies, could provide fantastic sensing bandwidth, sufficient to read whole human genomes (i.e., reading 100 Giga-bases) in well-under an hour (e.g., based on base-reading rates of polymerases, such as shown in Figure 12). Conversely, a large chip, such as having 10 mm x 10 mm sensor array die area, would still cost only a few dollars to fabricate at high volume, but provide for up to 10 billion sensors, which would have the capacity to sequence whole human genomes in seconds, or to support even higher bandwidth future applications such as the readout for Exabyte-scale DNA data storage, or whole proteome digital analysis (see below). [000110] Applications - These general features of the platform provide support many applications, with the potential for disruptive capabilities and dramatic performance improvements:

[000111] Molecular Electronic Microarravs - Of particular note, the hybridization sensor here, deployed in a massively multiplex fashion with many different hybridization oligos represented on one sensor array chip, is the molecular electronics equivalent of a classical DNA microarray, and could be used for many of the same applications. However, recast in this framework, it becomes a rapid readout, real-time, label-free detection array, that is deployed in an all-electronic chip-based fashion compatible with field deployment on low cost, compact devices. This next generation microarray thus confers many benefits. If the probes are taken to be aptamers, this can further provide for diverse targeting far beyond DNA targets, such as proteins. This illustrates the general principle that classical binding assays, when their molecules are recast as molecular electronic sensors on this chip platform, inherit many major performance advantages, as well as a long roadmap of further performance improvements.

[000112] Drug-Target Interaction Characterization - The ability to provide label-free, time-resolved detection of small molecule-protein and antibody-antigen interactions enables drug discovery applications, especially such as characterizing very weak binding interactions that may represent the earliest stages of drug candidate selection for poorly druggable targets. In addition, since the chips are inherently sensitive to single molecules, assays may operate with minimal demand for input materials, which may make them ideal for testing of novel or rare compounds. The potential for massive multiplexing on chip could here translate into high throughput screening of drug candidates, or for molecular evolution programs that rely on screening many mutant protein phenotypes, such as for antibody engineering, developing new CRSPR Cas genome editing enzymes, or directed evolution of proteins.

[000113] Diagnostic Testing - The basic sensor types demonstrated here provide a unifying foundation for transferring content from all existing molecular diagnostics platforms. For example, DNA hybridization is the basis of many forms of nucleic acid detection, such as in qPCR or DNA microarrays, as used in nucleic acid tests for viruses and infectious disease pathogens. Antigen-Antibody binding (or Aptamer binding) is the basis for immunoassays, such as commonly used in lateral flow detection of various antigens, such as pregnancy tests, or detection of protein biomarkers, or precision screening of panels of molecular allergens as used in the diagnosis of allergy and autoimmune disorders. CRSPR Cas enzymes interacting with their targets have recently been proposed as the basis for new types of diagnostics. The disclosed platform offers the potential to unify all these disparate diagnostics onto a common chip platform, and provide the benefits of massive multiplexing, and a compact, simple, all-electronic deployment format ideal for the future of Point-of-Care testing. And of course, DNA sequencing is itself a diagnostic testing modality of special importance, as a cornerstone of Precision Medicine, which can also potentially be unified onto this platform through DNA polymerase sensors (see below).

[000114] Proteome Analysis - The ability to detect a protein interacting with an aptamer or antibody target in solution provides a fundamental basis for identifying and quantifying proteins. Specific aptamers or antibodies on chip can provide for highly multiplexable identification of various known protein targets in a sample. Conversely, for whole-proteome analysis, the anonymous protein molecules in a sample can be mounted as single molecule sensors on a sensor array, and this “proteome representation chip” can be used to interrogate panels of hundreds of aptamers or antibodies, so as to provide an interaction “fingerprints” for each sensor that can be used to characterize and thereby identify and digitally count for quantification the proteins in the sample under analysis. In order to have several orders of magnitude dynamic range of counting, over the -20,000 canonical proteins of the human proteome, this requires tens of millions to billions of sensors on a chip, thereby benefiting from the upper scaling limits of the platform.

[000115] DNA Sequencing and Digital DNA Data Reading - The sensor with a DNA polymerase probe can monitor the activity of the polymerase as it copies a template, with resolution of the individual nucleotide addition events, and discrimination between bases. This ability to monitor polymerase activity enables the “sequencing by synthesis” methods, first introduced by Sanger with chain-termination sequencing. Such methods have dramatically increased in throughput and decreased in cost since their introduction in the late 1970’s through the introduction of next-generation massively parallel sequencers, and have even progressed to the first single-molecule sequencing platforms and the first CMOS chips sequencing devices. For the present molecular electronics chip platform, millions of such polymerase sensors, running in parallel on this scalable CMOS chip device, would offer the potential to generate a whole Human Genome sequence in less than an hour, on chips that need only cost several dollars to mass-produce, and which run on low-cost, compact instruments. Furthermore, access to Moore’s Law scaling ensures there can be a long roadmap of continuous improvements in cost and speed well beyond introduction, providing the potential for a highly durable sequencing technology. Such platform durability is rare but is in fact a general hallmark of disruptive technologies that also have a long roadmap for improvement, such as the jet engine has achieved for powering commercial aviation, and CMOS chips have achieved for powering computers. Such platforms have an “end game” quality, as once realized, they are very difficult to ever economically displace. In the present case, the molecular electronic CMOS sensor scaling limits noted above suggest the long term potential for chips costing tens of pennies, with the capacity of sequencing human genomes in tens of seconds.

[000116] Among the many potential applications of this new sensor chip platform, the disruptive whole genome sequencing described here is perhaps the grandest challenges to pursue. Not only can this provide a foundation for the future use of the genome in precision medicine, but the long-term scalability of the platform can also enable the ultra-high- throughput data reading needed for the future of molecular information storage, where zettabyte-scale digital data may one day be stored in DNA.

[000117] Future Developments - As the scaling potential of the platform approaches billion sensor chips, the capacities to read vast amounts of molecular interaction data efficiently and economically move from the realms of science fiction to engineering feasibility, making whole genome sequencing, whole proteome profiling, DNA data storage, as well as diverse diagnostic assays ever more affordable and accessible. Through chips such as these, the 50-year-old vision of molecular electronics as the ultimate means of chip miniaturization may be realized — but driven by their value for advancing the power of sensors, rather than processors.

[000118] Exemplary Materials and Methods

[000119] CMOS Sensor Array Chips. The proprietary CMOS sensor array chips used in this study were designed and fabricated using a 180nm CMOS node foundry. The chips present a 16k sensor pixel array. Pixels are post-processed at a foundry to have the tips of Ruthenium nano-electrodes exposed on the solution-facing surface of the chip, with such electrodes fabricated using either photolithography or E-beam lithography methods. The 16k electrodes were fabricated to have various nano-gap sizes in different ranges: 10-12 nm, 14- 16 nm, 17-20nm and 20-30 nm. Gaps of 14-20 nm were used for present experiments, and other sizes were not analyzed for present experiments. The chips were mounted in custom- built instruments to supply support to chip operations and sensor pixel data collection. The data are collected from the 16k sensor array at a frame rate of 1000Hz, and current measurements have 10 bits of resolution.

[000120] Alpha Helix molecular wire bridge preparation. The peptide is a helix-forming sequence 242 amino acids in length, including an N-terminal FLAG sequence and metal binding motifs at each end. In the alpha-helical conformation the length is ~25nm. A single cysteine residue is present in the middle position as the attachment point for probes using alkyne/azide click chemistry. To attach a DNA to the peptide, it was first modified using a thiol-reactive (103) 3-arylpropiolonitrile (APN)-PEG4- bicyclo [6.1.0] nonyne (BCN) (Conju-Probe, San Diego, CA) yielding a reactive bicyclo nonyne alkyne on the peptide. Typically, 100 pL of peptide solution (3 to 4 mg/mL in PBS) was first mixed with freshly prepared DTT or TCEP (2 mM final) and left at room temperature for an hour. Then the APN-BCN reagent dissolved in DMSO (1 M stock), is added to a final concentration of 0.01 M and mixed thoroughly by pipetting. The reaction is left at 4°C for a minimum of 48 hours. The excess APN-BCN is removed by size-exclusion chromatography. The purified peptide- BCN is stored at -20°C until needed. Further reactions are done using DNA or RNA oligos with azide placed at a specific site to obtain the bridges used in this study.

[000121] The reaction of BCN with azide was performed in PBS with a molar excess of the oligo-azide to purified peptide-BCN. The final reaction product was further chromatographically purified to more than 95%. The oligos were blocked at the free 3’ end with a fluorescent dye (FAM or Cy3 to help detection of peptide on SDS-PAGE). A gel shift on SDS-PAGE confirmed the bridge conjugation to oligos.

[000122] Sequences of oligos used in this study as DNA probes:

17mer- TACGTGCAGGTGACAGG/FAM/

45mer- CGATCAGGCCTTCACAGAGGAAGTATCCTCGTTTAGCATACCC/FAM/

[000123] Analysis of single-molecule binding data - Waiting times, To, and bound times, Tl, are extracted from signal traces using HMM analysis (see Supplmentary Materials for further details). Kinetic rate parameters are computed as: off rate, , on rate,

k_on = (here “bar” denotes the mean). The total fraction of time bound is then given by f_b =

). The fraction of time bound is modelled as obeying a classical Michaelis-Menten kinetic formula, depending on target concentration,^], as f_b = F ( —

^K )/(l + ( — )), where K_d is the empirical binding affinity, d ^Kd which has units of concentration, and F is an empirical constant. Thus, Kd is defined at the single-molecule level of interest here as the target concentration at which the single probe molecule spends equal time bound and unbound.

[000124] DNA oligo binding experiments - All oligo binding experiments performed in a buffer 50 mM Tris HC1 pH 7.5, 4 mM DTT, 10 mM KC1 and 10 mM SrC12 (Buffer A). Primer P-3 binds with its 3’ terminus 3 nucleotides away from the bridge; the sequence is:

5’-CCTGTCACCTGCAC, complementary to the 17mer. [000125] DNA melting temperature experiments and T_m estimation - All temperature melt experiments were using the 45mer probe-peptide bridges. The two oligos used in this analysis are 2P-0: CCTCTGTGAAGGCCTGATCG and 2P-5: CCTCTGTGAAGGCCT.

[000126] The temperature changes were controlled by the software interface that communicates with a Peltier device sitting attached to the chips. The temperature ramps were recorded as ignore and resume phases while every two-degree step were recorded continuously for four minutes of data collection stabilized at the temperature desired.

[000127] DNA Match-mismatch binding experiments - For assessing the binding kinetics for match and mismatch oligos following oligos designed against the 45mer probe bridge.

Exact Match 5’- CCTCTGTGAAGGCCTGATCG, 1 Mismatch 5’- CCTCTCTGAAGGCCTGATCG, 2 Mismatch 5’- CCTCTGTGAACCCCTGATCG, 3 Mismatch 5’- CCAGAGTGAAGGCCTGATCG. Targets (all 20-mers) were added separately, and binding kinetics monitored to tabulate fraction bound and other parameters.

[000128] Aptamer-Protein binding experiments - DNA Aptamers were conjugated to the peptide bridge the same click chemistry of azide to BCN described earlier. The SARS-CoV- 2 Nucleocapsid protein (N-protein) was target using a 94-nucleotide DNA aptamer:

/5AzideN//iCy3/TTTTTTGCAATGGTACGGTACTTCCGG

AT GC GG A A AC T GGC T A AT T GGT G AGGC T GGGGC GGT C GT GC AGC A A A AGT GC AC GC TACTTTGCTAA (63).

[000129] The SARS-CoV-2 Spike Protein (S-protein) was targeted using a 57-mer DNA aptamer: /5AzideN//iCy3/TTTTTTCAGCACCGACCTTGTGCTTTGGGAGTGCTGGTCCAAGGGCGT TAATGGACA (64). S

[000130] ARS-CoV-2 Spike Protein (S-protein) (RBD, His Tag)

(CatalogNumber:40592-V08B) was purchased from Sino Biological. The protein was analyzed using an SDS-PAGE to confirm the identity of the protein by molecular weight and purity. The control experiments used the Influenza A H1N1 (A/California/07/2009) Nucleoprotein / NP Protein (His Tag) Sino Biological (CatalogNumber:40205-V08B).

[000131] DNA polymerase binding experiments - E. coli DNA polymerase I, Klenow Fragment (3’ to 5’ exo ) (New England Biolabs®, Cat No: M0212M, 75.7 mM) was used for this study. For these experiments, the peptide bridge had the standard 45-mer oligo attached at its 5’ end and was annealed to a complementary 35-mer such that its free 3’ end was positioned 10 bases from the bridge attachment point, providing a binding site for DNA polymerase. These experiments were run using E. coli DNA pol. I Large Fragment exo- (New England Biolabs®) in the absence of nucleoside triphosphates without a thermostat in Mg²⁺- free buffer with SrCh.

[000132] Nucleotide binding experiments - As described above, the 45mer bridge was pre-complexed with a primer, and polymerase in Buffer A with a nucleotide added successively at concentrations of 0, 2.5, 5 and 15 pM, and binding events recorded. The nucleotide used here is a modified dT nucleotide with an anionic peptide tag.

[000133] Effect of salt on oligo binding - A 17mer bridge is used in these experiments with a binding oligo P-3 5’-CCTGTCACCTGCAC. The binding oligo concentration was lOOnM and the potassium chloride concentration was increased over range of 2mM to 2000mM in 10-fold increments. The assay was performed in Buffer A, except for the concentration of KC1 which is titrated as described. Debye length calculations use the standard formula valid for aqueous solutions of low ionic strength,

0.304 nm

[000134] l_ΰ

Ji

[000135] where I is the Molar ionic strength, as in (23).

[000136] Binding of Casl2a endonuclease enzyme - To use the Casl2a enzyme system, a guide RNA was designed to direct activity to the coding region of the SARS CoV-2 S gene. The chosen sequence: (rUrA rArUrU rUrCrU rArCrU rC/iAzideN/rU rGrUrA rGrArU rGrArG rUrCrC rArArC rCrArA rCrArG rArArU rCrU) was synthesized and attached to the sensor bridge peptide using click chemistry as described. The Casl2a enzyme was purchased from Integrated DNA Technologies® and used as obtained (Casl2a [Cpfl] V3, #1081068).

[000137] The target DNA strands (50-mer; 5'-AACTTCTAACTTT AGAGTCCAACCAACAGAATCTATTGTTAGATATCCTA and 5’- TAGGATATCT AACAATAGATTCTGTTGGTTGGACTCTAAAGTTAGAAGTT) were mixed at equimolar concentrations, heated to 95°C and slowly cooled to 4°C.

[000138] Antibody binding to Fluorescein on 3’ end of 45-mer probe - An FAB fragment (Anti fluorescein POD Fab fragment Ref No 1146346910 Roche®) at various dilutions was used. The fluorescein amidite modification capping the 3’ free end of 45-mer bridge is the target of the antibody binding.

[000139] COVID-19 Mock Assay - The detection target for a COVID-19 nucleic acid assay a segment of the N-gene. The following oligos are used in positive control experiments:

[000140] The probe oligo conjugated to the bridge is NlBr21 =

/5AzideN//iCy3/CCGCATTACGTTTGGTGGACC. A synthetic complementary oligo target used as positive a control is Nl-full-24 = 3'-TGGGGCGTAATGCAAACCACCTGG-5'. The CDC assay for SARS-CoV-2 detection kit (2019-nCov CDC EUA Kit, Cat No. 10006770) is used for generating positive control samples. PCR amplification using the control plasmid and primers from the kit are used for generating the N-gene PCR products. For on-chip assays the PCR DNA product was either digested with Bacteriophage Lambda Exonuclease (to remove the one phosphorylated + strand) or heat denatured and quick-chilled before using on the chips. For assessing the impact of complex background genomic DNA that may be present in real samples, separate runs were also performed wherein Salmon sperm DNA is added at a concentration of 2 pg/ml. Experiments were also performed by spiking in saliva samples from healthy individuals, and with 10% up to 50% saliva by volume going on the chip, so as to mimic possible sample contamination conditions. Results are not shown, but sensor readout was not substantially different from what is shown in Figures 13A, 13B with salmon sperm background.

[000141] DNA Polymerase Activity - The polymerase used was an engineered fusion protein, with a Phi29 DNA polymerase fused at the N terminus with the SpyCatcher protein. This was conjugated to the SpyTag-bridge conjugate post bridging on to the chip, to assemble the final polymerase activity sensor. The DNA template sequence used was:

5 , TTTTTTTTTTTTTTTTTTTTTTTTT GGGGGGGGGGGGGGG TCAGTCACGTCTA GATGCAGTCAG and the primer sequence was 5'-CTGACT GCATCTAGACGTGACTGA.

[000142] Nucleotides dCTP and dATP were supplied for the synthesis at a concentration of 200 mM.

[000143] Attachment of bacteriophage Phi29 DNA Polymerase to the peptide bridge - An exo Phi29 DNA polymerase-SpyCatcher fusion was expressed in E. Coli and purified using a combination of three chromatographic steps. The peptide bridge was expressed in E. Coli and purified using anti-FLAG resin and reacted with APN-BCN to produce the bridge-BCN conjugate as described above. The N-terminus of the spy Tag peptide was modified with azide for click reaction with BCN. (Add Spy ref. here) To form the final polymerase-bridge complex the Phi29-Spycatcher and bridge-BCN were mixed with slight excess of Phi29-Spycatcher (1 bridge: 1.5 polymerase) and incubated. The crude reaction was then purified on SEC.

[000144] Visualizations of Molecular Electronics Sensors - The sensors as shown in Figure 3, Figure 21, and throughout the paper, were visualized as follows: All solid and ribbon models and atoms/molecules were visualized using the ChimeraX program

(h tips :// www . rhvi ucsf . edu/c hi m erax/). version 1.2 (2021-05-24).

[000145] The bridge molecule 25nm alpha-helix is the exact amino acid sequence of the molecular wire peptide folded and visualized within ChimeraX. All other molecules used are scaled to be in proportion relative to the 25nm length of the bridge. The small-molecule conjugation between the bridge and probe is indicated schematically by a short, black zig zag line, as is the conjugation between the fluorescein and tether oligo in Figure 3 middle right. The respective other molecules shown are PDB structures (and their location in Figure3) are as follows: DNA oligo probe/target (upper left) custom sequences input and visualized as DNA helices; Klenow polymerase from 1KFD (upper right); SARS-CoV-2- Spike Protein from 6VSB, and the S aptamer is the exact Aptamer DNA sequence, folded into a 3D secondary structure using RNAComposer (http vVrnacomposer.es. put poznan.pl), with structure visualized in ChimeraX (middle left); Antibody from 1IGT (middle right); Casl2a from 5B43 (lower left); Phi29 Polymerase from 2PYL, combined with SpyTag- SpyCatcher conjugation complex from 4MLS (Figure 3 lower right).

Claims

What is Claimed is:

1. An apparatus to detect small molecule interaction, comprising: a first nanoelectrode and a second nanoelectrode separated by a gap; a molecular wire in electrical communication with the first and the second nanoelectrodes; a probe molecule connectable to the molecular wire, the probe molecule further comprising means to interact with a target molecule; a detector to detect one or more pulses indicating interaction kinetics between the probe molecule and the target molecule; and a sensor to receive the one or more detected pulses from the detector and identify at least one of the plurality of pules as a reaction indication type;

Wherein the interaction between the probe molecule and the target molecule is detected by the detector in real time and the interaction is substantially label-free; and wherein the sensor is configured to identify the reaction indication type by applying a Hidden Markov Model (HMM) to a timeseries of pulses.

2. The apparatus of claim 1, wherein at least one of the probe molecule or the target molecule is selected from the group consisting of functionalized with specific biomolecules, single-strand DNA and RNA probes, antibodies, antigens, aptamers and enzymes.

3. The apparatus of claim 1, wherein the apparatus is used to detect a change in charge to indicate probe-target interaction selected from the group consisting of drug-target interaction, molecular diagnostics, DNA sequencing, aptamer testing, proteomic analysis, environmental monitoring.

4. The apparatus of claim 1, wherein the first nanoelectrode the second nanoelectrode are separated by a gap of about 15-20 nanometers to form an electrode pair and wherein at least 16,000 electrode pairs are arranged to form an array.

5. The apparatus of claim 1, wherein the probe further comprises a first ribbon and the target further comprises a second ribbon.