FIELD OF THE INVENTION

The present invention is related to the field of the detection of biomarkers, blood analysis systems for disease and health monitoring, and microfluidic devices for blood sample analysis.

BACKGROUND TO INVENTION

In modern healthcare, in vitro diagnostics go far beyond simply telling a doctor whether a patient has a certain disease or not. Today, diagnostic tests are an integral part of the clinical decisionmaking process along the entire continuum of a patient's health related topics, enabling physicians to make full use of in vitro diagnostics along the healthcare value chain. In vitro diagnostics have been influencing over 60% of clinical decisions. Diagnostic testing empowers doctors to make the right decisions for their patients at the right time for the effective management and prognosis of disease. The development of personal diagnostics will allow people to have improved control over their health and to do so in the convenience and privacy of their own-homes.

Management and prognosis of disease requires the measurement of biomarkers. Quantitative detection of low abundance biomarkers is of great interest for health monitoring and biological applications because protein biomarkers can be present at low concentrations of 10-15 to 10-12 molar (M) in biofluids. Furthermore, detection of biomarkers in human body fluids such as whole blood is very desirable. New diagnostic platforms provide for changes in how disease diagnosis and health monitoring can be conducted. Microfluidic devices have the potential to enable multiplexed biomarker analysis of ultra-low volume (~5 μΙ) whole blood samples with high- sensitivity and broad dynamic range and to do so in a point-of-care (POC), home-based, or resource-limited setting.

Microfluidics has been developed to precisely manipulate nanoliter (nl) or picoliter (pi) volumes, enable large-scale integration and improve bioassay performance. Numerous microfluidic systems have been developed for miniaturization of immunoassays over the years. Although the microfluidic format has enhanced the performance of immunoassays, no integrated microfluidic assay exists that can provide high-sensitivity, high-dynamic range multiplexed analysis of ultralow blood volume samples. Also, the background art does not describe any portable and low-cost device control and readout instrumentation.

Research and clinical laboratories have extensively used Enzyme-Linked Immunosorbent Assays (ELISA) for protein detection. Although the standard ELISA format is robust, it suffers from several limitations. The ELISA format requires large volumes of reagents and samples, as well as long incubation times. Multiplexed alternatives to ELISAs have lately been developed including bead-based methods, PCR and biosensors. Hirsch et al. presented a simple immunoassay using gold nanoshells that is able to detect analyte in serum or whole blood without requiring sample preparation (Hirsch et al., 2003). In this method, an analyte is added to a solution of disperse antibody conjugated nanoparticles. Upon addition of the analyte, the absorption spectrum of the nanoparticle solution diminishes, such that the presence of the analyte is detectable. However, this assay is limited due to the need for large reagent and sample volumes and its non-portable nature. The detection limit of the method is over the range of 88-0.8 ng/mL within 10-30 mins in saline, serum and 20% whole blood.

Fan et al. achieved rapid multiplexed protein biomarker detection from 10 μΙ whole blood samples diluted to 90 μΙ with buffer using an integrated blood barcode chip device (Fan et al., 2008). The device of Fan comprises a microfluidic device having a control layer, a flow layer and a glass slide patterned with a plurality of channels wherein each channel is coated with a distinct DNA-directed immobilized antibody. The antibody binds a specific plasma protein, which is then recognized by a biotin-labeled detection antibody and streptavidin-Cy5 florescence probe. This detection method corresponds to a variant of a standard ELISA assay. The method of Fan produces an analogue read-out signal that can be used to detect changes in the concentration of plasma proteins in a patient's blood.

Wang et al. describes an updated version of the device of Fan in which adsorbent filter paper provides power to drive all the steps of the ELISA-like immunoassay (Wang et al., 2010). This device is able to quantitate picomolar concentrations of more than ten cancer biomarkers and cytokines concurrently in cancer patient blood. However, the time required to prepare the device prior to sample introduction is 45 minutes, followed by 10 minutes for sample introduction. The authors do not provide the time required for post sample processing. Furthermore, the device readout requires an expensive research grade DNA microarray scanner, thereby restricting the use of the device to POC applications.

Inci et al. discloses a nanoplasmonic electrical field-enhanced resonating device (NE ² RD), that achieves a sensitivity of 400 fg/ml (-23 fM) for detecting IFN-γ in human serum with a dynamic range of 8 orders of magnitude (Inci et al., 2015). The device comprises antibodies that are immobilized to a polystyrene surface coated with gold nanoparticles. The antibodies capture specific plasma proteins. Spectral measurements are then performed to monitor capture events, which change the spectral colour of the surfaces in terms of both wavelength and extinction intensity parameters, providing an analogue readout. The NE ² RD doesn't support multiplexed analysis of biomarkers and requires an exceedingly large sample volume of 100 μΙ per analyzed biomarker. The cost of the instrumentation required for the device readout is not discussed by the authors. MITOMI is a micro-mechanical method recently developed to allow the quantitative analysis of molecular interactions (Maerkl and Quake, 2007). MITOMI consists of a freestanding "button" membrane, which can be actuated by pneumatic or hydraulic pressure, similarly to standard micro-mechanical valves generated by multilayer soft-lithography. When the button membrane is actuated it physically contacts a circular area on the glass surface of the microfluidic device. When the button membrane is in contact with the glass surface it protects the surface from solute and solvent. MITOMI can be used to mechanically trap surface bound molecules between a surface and the button membrane, preventing dissociation of these molecules and thus allowing the measurement of transient molecular interactions. A MITOMI analogue high-throughput microfluidic platform for the quantification for antigen has been described (Volpetti et al., 2015). In this device, primary antibodies are isolated in individual detection regions by MITOMI to eliminate cross-talk between individual assay units of the device. The primary antibody binds a specific analyte, which is then recognized by a fluorescently labelled secondary antibody. The primary antibody-analyte-secondary antibody complex is trapped by the button membrane. The analogue detection signal produced by binding of the fluorescently labelled secondary antibody is quantitated using a DNA microarray scanner. The platform has a limit of detection (LOD) of 4-30 pM, depending on the antigen. Wang et al. discloses a microfluidic device that uses microwells to enhance the sensitivity of a chemi-fluorescence detection assay for antibodies in 5% BSA PBS (Wang et al., 2013). The microfluidic device comprises a pneumatic control layer, a flow (fluidic) layer and a surface, wherein the flow comprises a series of inter-connected assay chambers. The assay chambers comprise a movable element that may be reversibly defected such that it is brought into contact with the surface, at relatively low pressure (30-80 kPa, equivalent to 4.3-1 1 .6 psi), to create an array of femtoliter detection volumes interconnected by a thin layer of liquid (Wang et al. 2013, Figure 1 ). The femtoliter structure of the moveable element enhances the sensitivity of fluorescent detection in a chemi-fluorescence detection assay. The femtoliter wells are not suitable for immunoassays, and produce an analogue readout ("our readout method is fundamentally different from a digital immunoassay", Wang et al. 2013, page 4193, paragraph 2), as fluorophores captured by the moveable element are able to diffuse across the interconnected wells. The device of Wang requires 10 μΙ of sample (Wang et al. 2013, Table 1 ). Digital immunoassays have emerged as a robust technology for sub-picomolar detection of proteins. Some of these assays use micrometer-sized wells or droplets to isolate individual target molecules for single-molecule detection.

Rissen et al. describes a singulated bead ELISA in which microscopic beads coated in antibodies capture target analyte from a sample (Rissin et al., 201 1 ). The capture beads are then sealed into an array of femtoliter-sized isolated microwells together with a droplet of enzyme substrate. Beads associated with the bound enzyme produce a locally high concentration of florescent product in the well. Positive wells are counted to provide a digital readout. This digital ELISA is capable of detecting proteins in 100 μΙ serum samples at concentrations as low as 0.4 fM.

Although digital immunoassays, such as that described by Rissen et al., have great potential to drastically improve assay sensitivity, one limitation of digital measurements is that the number of unique digital units limits the dynamic range. A second limitation of digital assays is that they have not yet been implemented in an integrated miniaturized assay platform. The method of Rissen, for example, requires 100 μΙ of sample (Rissin et al., 201 1 ). Previous attempts at incorporating a digital assay into a microfluidic device have been unsuccessful (Wang et al., 2013). Therefore, despite advancements in the field of biomarker analysis, substantial improvements are desired and necessary for detection of fM concentrations of an analyte in ultra-low volume clinical samples.

SUMMARY OF INVENTION

The present invention relate to a microfluidic device comprising a control layer, a flow layer and a surface, wherein:

a) the flow layer and surface define a plurality of digital assay units, wherein the digital assay units comprise a digital moveable element, wherein the digital moveable element:

i. may be reversibly deflected,

ii. is positioned such that reversible deflection of the element brings it into contact with the surface, and

iii. is patterned such that contact of the element with the surface forms an array of discrete enclosed wells;

a) the pressure in the control layer controls deflection of the moveable element and b) the digital units are connected by a fluid flow path.

In another aspect the present invention relates to a method for digitally representing the concentration of an analyte in a sample, comprising

a) trapping the analyte directly from the analyte sample in an array of discrete enclosed wells, wherein the wells are formed from contact between a movable element and a surface, wherein

i. the analyte is immobilised within the wells

ii. the analyte is randomly distributed within the array,

b) targeting an enzyme to the analyte containing wells of the array

c) trapping the enzyme in the analyte containing wells in the presence of its substrate, wherein enzyme catalysis of the substrate produces a detectable signal, d) determining the rate of catalysis within each well and

e) summing the number of wells having a rate of catalysis above a predetermined background level so as to provide a digital readout.

According to one aspect of the present invention, a multiplexed microfluidic system is provided that combines digital and analog detection based on mechanically induced trapping of molecular interactions (MITOMI) to create a portable, integrated, digital-analog hybrid microfluidic device capable of detecting biomarkers with high-sensitivity and dynamic range.

Although the microfluidic format has enhanced the performance of immunoassays, currently no integrated microfluidic assay exists that simultaneously fulfills the following performance criteria: i. high-sensitivity (sub-picomolar),

ii. high-dynamic range (5-6 orders of magnitude or better),

iii. multiplexed biomarker analysis (12<),

iv. compatible with ultra-low volume (3-5 μΙ) whole blood samples.

According to one aspect of the present invention, a multiplexed microfluidic system is provided that combines digital and analog detection based on mechanically induced trapping of molecular interactions (MITOMI) to create a portable, integrated, digital-analog hybrid microfluidic device capable of detecting biomarkers with high-sensitivity and dynamic range. Although the microfluidic format has enhanced the performance of immunoassays, currently no integrated microfluidic assay exists that simultaneously fulfils the following performance criteria:

i. high-sensitivity (sub-picomolar)

ii. high-dynamic range (5-6 orders of magnitude or better)

iii. multiplexed biomarker analysis (12<)

iv. compatible with ultra-low volume (3-5 μΙ) whole blood samples.

The proposed integrated microfluidic platform allows the quantitation of multiple protein biomarkers from a 5 μΙ whole blood sample obtainable by a pinprick, without the need for any pre-processing of the sample on or off chip. The proposed microfluidic device can also be combined with a low-cost, portable do-it-yourself (DIY) microfluidic control system and a cheap fluorescence universal serial bus (USB) microscope to enable diagnostic testing at the POC, at home, or in resource limited field-environments. The device has been extensively tested by performing single enzyme measurements, and digital immunoassays in serum, achieving single molecule detection in both cases and a sensitivity as low as -10 fM (330 fg/ml). Also, the experimental results have also demonstrated the applicability of the device to clinically relevant tests by detecting anti-Ebola antibodies in whole blood and IgG/lgE antibodies in human serum for allergy testing. According to an aspect of the present invention, the microfluidic digital-analog hybrid immunoassay device consists of a flow and a control fluidic layer fabricated by multilayer soft lithography. The PDMS device measures 20 mm in length, 14 mm in width, ~4 mm in height and is bonded to a 25 mm x 75 mm glass slide. The device includes 16 independent assay units, schematically shown as a circuit. The assay chambers contain a deflectable button membrane (analog MITOMI) in one half of the assay units and a deflectable button membrane patterned with femtoliter wells (digital MITOMI) in the other half. Analog detection is achieved using the MITOMI mechanism: increasing analyte concentration results in increasing fluorescence signal density under the MITOMI button, showing a schematic cross-sectional view of the MITOMI detection area. In digital-MITOMI, increasing analyte concentrations give rise to increasing numbers of positive wells containing one, two, or more molecules. A sandwich immunoassay is performed with either fluorophore-labeled (analog MITOMI) or enzyme- conjugated (digital MITOMI) antibodies.

ABBREVIATIONS

~ approximately

3D three-dimensional

FDS microFluidic Diagnostic System

μΙ microliter

μηι micrometer

DIY do-it-yourself

ELISA enzyme-linked immunosorbent assays

G galactosidase

FDG fluorescein di-β-ϋ- galactopyranoside

fg femtogram

fM femtomolar

ig immunoglobulin

IFN interferon

kPa kilopascals

LOD limit of detection

M molar

MCS microfluidic control system

MITOMI mechanically induced trapping of molecular interactions

mm millimeters NE ² RD nanoplasmonic electrical field-enhanced resonating device

ng nanogram

nl nanoliter

PCR polymerase chain reaction

PDMS poly-di-methyl siloxane

Pl picoliter

POA point-of-care

psi pound per square inch

USB universal serial bus

W watts

wH watt hours

DETAILED DESCRIPTION

Brief description of the Figures

Figure 1 The detection device

A: Schematic of the detection device. B: Schematic of an assay unit having an analogue and digital MITOMI. C: Schematic of the MITOMI detection areas.

Figure 2 The MITOMI detection areas

Phase contrast microscope images of an analog (A) and digital (B) assay unit.

Figure 3 MITOMI detection assays

The device may be used to perform a sandwich immunoassay may be performed with either fluorophore-labelled (analog MITOMI, A) or enzyme-conjugated (digital MITOMI, B) antibodies

Figure 4 Determination of analyte concentration by analogue and digital MITOMIs

A: Schematic representing the relationship between increasing analyte concentrations signal densities in analog measurements and positive femtowells in a digital measurement. B: Relationship between fluorescent intensity and biomarker concentration as measured by analogue or digital detection.

Figure 5 Visualization of analog and digital detection

Phase contrast images (A, B) and fluorescence images (C, D) of an analog MITOMI button (A, C) and a digital MITOMI button having 400 femtocells (B, D). Scale bar = 150 μηι. Figure 6 Microfabricated diagnostic devices

A: Design schematic of an analog MITOMI microfluidic device is represented showing flow (blue) and control (grey) layers. The device has four rows; each row contains four unit cells for a total of 16 analog assay units. Each assay unit (inset) contains: a MITOMI button valve 1 , a sandwich valve 2, two neck valves 3, 4 for the top chamber / ^' and two 5, 6 for the bottom one // ^' , and a peristaltic pump 8. B: A digital MITOMI device. C: A digital-analog hybrid MITOMI device, with 8 digital and 8 analog units.

Figure 7 Integrated peristaltic pump

A: Schematic of the serpentine-shape (S-shape) delay peristaltic pump and magnified view of a single column. B: Schematic illustrations of the difference between a conventional peristaltic pump, in which three (3) independent valves are activated in sequence, and the S-shape delay peristaltic pump, in which a single channel is able to perform the mixing. C: A Cy3 labeled solution mixing measurement at a frequency of ~ 8.3 Hz (topen = tclose = 60 ms) and P = 25 psi. D: schematic illustration of different flow modes. Figure 8 MicroFluidic Diagnostic System

A FDS including a microfluidic control system (MCS), a fluorescence USB microscope, a portable computing device, for example but not limited to a netbook, portable computer, tablet computer, laptop computer, and the microfluidic device.

Figure 9 Capillary sampling of whole blood

Sample volumes for analysis can be as small as 5 μΙ of whole blood obtainable by a pinprick

Figure 10 Processing of a blood sample

The blood sample is aspirated and transferred to a control device of the μ FDS with a micro- hematocrit tube coated with heparin.

Figure 11 3D rendering, block diagram and overview of the graphical user interface of the MSC

A: 3D rendering of the MSC showing how the different components were arranged. B: Schematic representation of the system including of an Arduino Mega 2560 coupled with an electronic PCB (custom shield), a touchscreen, a battery, a pneumatic system (manifold, pressure regulators, pumps), and an imaging/analysis module (USB microscope + netbook). C: Several possible control modes are available to the user: (1 ) manual control of the compressor, vacuum pump, peristaltic pump and individual valves; (2) manual control of pressure; (3) preprogrammed routines for GFP, IgG and IgE detection.

Figure 12 Circuit diagram

A: ULN2803 transistor array. One digital pin of the Arduino is connected to each IN pin of the ULN. Applying +5 V to the IN, allows current to pass from the corresponding OUT pin to the ground and thus through the load that is connected to the OUT pin - in this case, a solenoid valve. Three ULN2803's are used to drive the 24 valves of the MSC. B: The pressure regulators. A low-pass filter converts the PWM signal from the Arduino ("pr1_in") to an analog voltage that can be read by the regulator on pin 2. Pin 3 is the output voltage indicating the current pressure, and goes to the Arduino pin labeled "pr1_out". C: The transistors for the pumps. Applying +5 V to the control + pin allows current to pass through the pump and from the drain to the source. The diode wired in parallel with the pump protects the circuit from voltage spikes occurring when the pump is switched off.

Figure 13 Schematic of digital MITOMI

An enzyme/substrate mixture (βG + FDG) was introduced on chip after surface derivatization

Figure 14 Substrate turnover rates of molecules for different enzyme concentrations

Histograms of the substrate turnover rates of βG molecules for different enzyme concentrations, with a bin size for all histograms being 50 s ^-1 inserts show representative microscope images of individual digital-MITOMI buttons with a scale bar of 50 μηι.

Figure 15 Activity traces and observed turnover rates of enzymes for the assay

A: Activity traces of enzymes for the βG assay. Turnover rates are determined by linear regression fitting (solid red lines) to the fluorescence trajectories after background subtraction. B: Histograms of observed turnover rates for individual femtowells are shown. The numbers of wells are plotted versus the increase in observed rate. For 0.5, 5, and 25 pM the peaks were fitted with a sum of Gaussians. The values displayed above each peak are the centers of these Gaussians. The inserts show the probability distribution. The black bars represent the experimentally obtained probability distributions calculated from the Gaussian fits. Red data points represent the values obtained from the Poisson distribution with the

parameter λ equal to the expected number of enzymes per chamber for 0.5, 5, and 25 pM of βG. Blue data points are the values obtained from the Poisson distribution for 12.5 pM of βG. The bin size for all histograms is 50 s ^~ . Figure 16 Correlation of the on-chip signal with the known solution concentration of βG.

The sum of turnover rates are plotted against βG concentration, and the dashed red line is a linear fit to the data (R2 = 0.99). Figure 17 GFP sandwich assays

A: Schematic representation of the GFP sandwich assays performed on the microfluidic device and detected with digital-MITOMI, analog amplification, and analog-MITOMI. B, C: Detection of GFP in 2% BSA PBS buffer solution over concentrations ranging from 10 fM to 5 pM are presented in a chart. D: the Digital-MITOMI detection of GFP in human serum over concentrations ranging from 12 fM to 50 fM. Intensity of fluorescein signal is plotted as function of GFP concentration. E, F: detection of GFP in 2% BSA PBS buffer solution with concentrations ranging from 500 fM to 5 nM is shown., analog amplification (E) and analog- MITOMI detection (F) of GFP. Data were obtained with a fluorescence USB microscope. LODs were determined from the signal of a negative control plus 3 standard deviations. G: representative fluorescence images of one assay unit taken with the USB microscope for different concentrations of GFP in 2% BSA PBS using the analog amplification method, the scale bar indicated being 350 μηι.

Figure 18 Multiplexing capability of the system

Details of spotting (A) and immunoassay (B). (C, D): A concentration of 5 pM (75 pg/ml) anti- Bundibugyo GP IgG, 1 pM (15 pg/ml) anti-Reston GP IgG spiked into human serum, 3 independent chips. D: Measurements of 1 pM anti-Reston Ebola GP IgG antibody. E: A negative control was measured on three (3) independent devices. C-E: The sum of the signal in positive femtowells, and examples of a digital-MITOMI button for each sample are shown. All measurements were performed in human serum, and the scale bar is 700 μm. Figure 19 Anti-Ebola virus detection in a whole blood sample

A: Schematic representation of the assay for anti-Ebola virus detection in a whole blood sample. (B-D): Data for different concentrations of anti-Zaire Ebola GP IgG antibody (B) anti- Bundibugyo Ebola GP IgG antibody (C) anti-Reston Ebola GP IgG antibody (D) spiked in whole blood. Zaire, Bundibugyo and Reston glycoproteins are spotted in quadruplicate to yield two repetitions for the analog and the analog amplification detection mechanisms. Goat anti-rabbit IgG PE conjugated antibody and goat anti-rabbit IgG β-Galactosidase conjugated antibody were spotted in the top and bottom part of the chip, respectively, to allow the use of both detection methods on the same device. Data represent means (n=2). Error bars correspond to the standard deviation of the means.

Figure 20 Representative fluorescent images of 10 nM anti-Zaire Ebola GP IgG antibody spiked in whole blood

A: The analog-MITOMI signal imaged using a fluorescence USB microscope with a 570 nm emission filter. B: The amplified analog signal detected with a fluorescence USB microscope using a 510 nm emission filter, and the scale bar used is 700 μηι.

Figure 21 Peanut allergy

A: The device was programmed with two repeats of three different allergens: Arachis hypogaea allergen 2 (Ara h2), Felis domesticus allergen 1 (Fel d1 ), Dermatophagoides pteronyssinus allergen 2 (Der p2), and a negative control. B: The clinical serum sample (6 μΙ) was diluted 1 :2 in an incubation buffer and flowed in the chip following the protocol described below and IgE and IgG were detected. C: Quantification of IgE levels for a sample positive to peanuts allergy and the corresponding fluorescent images. D: Quantification of IgG levels for a sample positive to peanuts allergy and the corresponding fluorescent images. Bars are means (n=2), with individual data points shown as dots (·).

Figure 22 Charts of activity traces and observed turnover rates of enzymes for the digital-MITOMI immunoassay.

A: enzyme activity traces for GFP detection in 2% BSA. Turnover rates are determined by linear regression fitting (solid red lines) to the fluorescence trajectories after background subtraction. B: histograms of observed turnover rates of individual femtowells. C: enzyme activity traces for GFP detection in human serum. D: histograms of observed turnover rates. The bin size for all histograms is 50 s ^-1 .

Figure 23 GFP quantitation.

A, B: detection of GFP in 2% BSA PBS buffer solution over concentrations ranging from 10 fM to 5 pM. A: results from the digital-MITOMI GFP immunoassay. Plotted is the sum of slopes of all femtowells above a background threshold. B: the results from the analog-MITOMI GFP immunoassay. Plotted is the intensity of Cy3 signal as function of GFP concentration. C: the digital-MITOMI detection of GFP in human serum over concentrations ranging from 12 fM to 50 fM. Intensity of fluorescein signal is plotted as function of GFP concentration. Data were obtained with a research grade fluorescent microscope. D, E: detection of GFP in 2% BSA PBS buffer solution with concentrations ranging from 500 fM to 5 nM. D: analog amplification detection of GFP. E: analog-MITOMI detection of GFP. Data were obtained with a USB fluorescent microscope. LODs (dashed red lines) were defined as the signal of a negative control sample plus 3 standard deviations.

Figure 24 Multiplexed digital Ebola diagnostics

A: Enzyme activity traces for detection of anti-Bundibugyo, anti-Reston and control. Turnover rates are determined by linear regression fitting (solid red lines) to the fluorescence trajectories after background subtraction. B: histograms of observed rates of enzymes. Dashed rectangles indicate the position where the positive reaction took place along the channel. The bin size for all histograms is 50 s ^-1 . Definitions

And/or is herein defined as "either or both", for example "the digital and/or analogue assay units" is defined as "the digital assay unit, the analogue assay unit or both the digital and analogue assay units".

An analogue readout or representation of analyte concentration is one in which increasing analyte concentration result in increasing signal density.

Anti-analyte antibody is herein understood to mean an antibody capable of binding the relevant analyte, i.e. the analyte to be detected.

An assay unit is a discrete unit in a microfluidic device suitable for performing an assay. Individual or independent assay units may be in fluid flow connection or isolated from other assay units in the device. An analogue assay unit produces an analogue readout, whilst a digital assay unit produces a digital readout.

A background level is herein described as the signal observed or detected when a sample known to contain no analyte is analysed by the same method as a test sample, or the known detection limit of the assay, as is commonly understood in the art. The background level may also be termed the background threshold.

A control layer of a microfluidic device controls the fluid flow through the device. A digital readout or representation of analyte concentration is one in which increasing analyte concentration gives rise to increasing numbers of positive wells containing one, two, or more molecules.

A flow layer of a microfluidic device is a layer through which liquid can flow.

A fluid flow path is an area through which a liquid may pass when there is no obstruction blocking the path. A fluid flow path may be blocked by an obstruction such as a closed valve.

A functionalised surface is one that has been modified to provide a function, for example the ability to non-specifically or specifically bind proteins.

Mechanically trapping a molecule means to force a surface bound molecule against the surface such that it does not move, whilst excluding solutes that are not bound to the surface (Maerkl and Quake, 2007).

A moveable element is a structure or "button" membrane that can be actuated (Maerkl and Quake, 2007; Volpetti et al., 2015). A digital moveable element forms part of a digital assay unit. An analogue moveable element forms part of an analogue assay unit.

Pressure is described herein in psi units, wherein 1 psi is equivalent to -6.89 kPa.

Sample processing is herein defined as ex vivo manipulation of the sample on or off chip, including purification and cell separation or sorting. The addition of heparin to a blood sample does not constitute sample processing.

A USB immunofluorescence microscope is a low-powered digital microscope that can connect to a computer via a USB port.

A well is a cavity capable of holding liquid. A discrete enclosed well is one that can be individually distinguished and structurally defined by its capacity to independently hold liquid and solutes without any leakage of the liquid or solute into neighbouring wells or the surrounding environment. A solute or immobilised molecule trapped in a discrete enclosed well cannot escape from the well.

Description

The present invention relates to a microfluidic device comprising a control layer, a flow layer and a surface, wherein:

a) the flow layer and surface define a plurality of digital assay units, wherein the digital assay units comprise a digital moveable element, wherein the digital moveable element:

i. may be reversibly deflected ii. is positioned such that reversible deflection of the element brings it into contact with the surface, and

iii. is patterned such that contact of the element with the surface forms an array of discrete enclosed wells

a) the control layer controls deflection of the moveable element and

b) the digital units are connected by a fluid flow path.

In one embodiment the invention also comprises

a) the flow layer and surface define a plurality of digital assay units, wherein the digital assay units comprise a digital moveable element, wherein the digital moveable element:

i. may be reversibly deflected ii. is positioned such that reversible deflection of the element brings it into contact with the surface, and

iii. is not patterned such that contact of the element with the surface forms an array of discrete enclosed wells

b) the control layer controls deflection of the analogue moveable element and c) the analogue units are connected by a fluid flow path.

In another embodiment, the digital assay units comprise one or more reaction chambers, a reagent chamber, and one or more sandwich valves, wherein:

a) the moveable element is positioned within the reaction chamber, the reaction chamber and each of the reagent chambers within an assay unit are connected by a assay unit fluid flow path,

the one or more sandwich valves may be in an open or closed conformation, wherein i. the closed conformation blocks the assay unit fluid flow path and isolates the reagent chamber from the reaction chamber

ii. the open conformation allows flow along the assay unit fluid flow path.

In another embodiment, the analogue assay units comprise one or more reaction chambers, a reagent chamber, and one or more sandwich valves, wherein:

a) the moveable element is positioned within the reaction chamber,

b) the reaction chamber and each of the reagent chambers within an assay unit are connected by a assay unit fluid flow path,

b) the one or more sandwich valves may be in an open or closed conformation, wherein i. the closed conformation blocks the assay unit fluid flow path and isolates the reagent chamber from the reaction chamber and

ii. the open conformation allows flow along the assay unit fluid flow path.

In one embodiment the conformation of the sandwich valves of the analogue assay units is determined by the pressure in the control layer. In one embodiment the conformation of the sandwich valves of the digital assay units is determined by the pressure in the control layer.

In one embodiment the discrete, enclosed wells are femtoliter wells. In another embodiment the flow and control fluidic layers are fabricated by multilayer soft lithography. The flow and control layers may be fabricated from poly-di-methyl-siloxane (PDMS). In another embodiment the device measures approximately 20 mm in length, 14mm in width and 4mm in height. The surface may comprise a functionalised microscope slide. The slide may be made from glass and may be functionalised with epoxysilane. In yet another embodiment the device of the invention may comprise 16 independent assay units. The changes in fluorescence intensity within the digital assay units of the invention may be detected with an immunofluorescence microscope. The changes in fluorescence intensity within the digital assay units of the invention may be detected with an immunofluorescence microscope, preferably a USB immunofluorescence microscope. The changes in fluorescence intensity within the analogue assay units of the invention may be detected with an immunofluorescence microscope. The changes in fluorescence intensity within the analogue assay units of the invention may be detected with an immunofluorescence microscope, preferably a USB immunofluorescence microscope.

In another embodiment the surface of the device of the invention may be spotted with antibodies and the flow layer aligned over the surface such that the antibody spots are positioned within the reagent chambers of the digital assay units and not within the reaction chamber of the digital assay units. These antibodies may be anti-viral antibodies, preferably anti-Ebola antibodies, more preferably anti-Ebola virus glycoprotein (GP) IgG antibodies.

In another embodiment the of the device of the invention may be spotted with antibodies and the flow layer aligned over the surface such that the antibody spots are positioned within the reagent chambers of the analogue assay units and not within the reaction chamber of the analogue assay units. These antibodies may be anti-viral antibodies, preferably anti-Ebola antibodies, more preferably anti-virus glycoprotein (GP) IgG antibodies.

The device of the invention may also comprise neck valves, wherein the neck valves may be in an open or closed conformation, wherein

a) the closed conformation blocks the fluid flow path between two adjacent analogue assay units and isolates the two adjacent analogue assay units from each other and b) the open conformation allows flow along the fluid flow path between two adjacent analogue assay units. The device of the invention may also comprise neck valves, wherein the neck valves may be in an open or closed conformation, wherein

a) the closed conformation blocks the fluid flow path between two adjacent digital assay units and isolates the two digital assay units from each other and

b) the open conformation allows flow along the fluid flow path between two digital assay units.

The conformational state of the neck valves of the invention may be determined by the pressure in the control layer. In another embodiment the digital assay unit fluid flow path is powered by a peristaltic pump, wherein the peristaltic pump comprises a plurality of intersections between the control layer and the flow layer such that pressure changes in the control layer alter the conformation of the flow layer at a plurality of positions along the digital assay unit fluid flow path. In one embodiment the plurality of intersections produces a serpentine arrangement of the control layer relative to the digital and/or analogue assay unit, such that the peristaltic pump operates as a delay pump.

In another embodiment the analogue assay unit fluid flow path is powered by a peristaltic pump, wherein the peristaltic pump comprises a plurality of intersections between the control layer and the flow layer such that pressure changes in the control layer alter the conformation of the flow layer at a plurality of positions along the analogue assay unit fluid flow path. In one embodiment the plurality of intersections produces a serpentine arrangement of the control layer relative to the digital and/or analogue assay unit, such that the peristaltic pump operates as a delay pump. In another embodiment, the reaction chamber of the invention comprises a detection area comprising 2 reagent chambers, wherein each reagent chambers is connected to the reaction chamber by a channel that allows flow through the reaction chambers and reagent chambers by actuating the peristaltic pump. In one embodiment the digital moveable element of the device is reversibly deflected by a pressure of over 12 psi in the control layer, preferably over 15 psi, most preferably 25 psi.

In another embodiment the discrete enclosed wells of the invention have a diameter of approximately 5μηι, a height of approximately δμηι and/or a volume of approximately 100 fl_.

In another aspect the invention relates to a micro-fluidic diagnostic system ^FDS) comprising: a) the device of the invention;

b) a microfluidic control system (MCS);

c) a fluorescence USB microscope and

d) a portable computing device,

wherein the MCS controls the flow and pressure in the control and flow layers of the device. The portable computing device is preferably a netbook, portable computer, tablet computer or laptop computer. The MCS may also function as the portable computing device. The MCS may comprise:

a) a printed circuit board;

b) a microcontroller, preferably a Arduino™ microcontroller;

c) a touch sensitive display and

d) a battery pack,

wherein the MCS may be controlled by the touch sensitive display.

The battery pack preferably has a capacity of 100 Wh or less, preferably wherein the MCS has an average power consumption of 25 W or less.

The invention also relates to a digital assay unit forming part of the flow layer of a microfluidic device, the assay unit comprising a digital moveable element and a surface, wherein the digital moveable element:

a) may be reversibly deflected,

b) is positioned such that reversible deflection of the element brings it into contact with the surface, and

c) is patterned such that contact of the element with the surface forms an array of discrete, enclosed wells

In another aspect the invention relates to a method for counting enzymes, in particular a method for digitally representing the concentration of an enzyme in a test solution, comprising:

a) trapping enzymes directly from the solution in an array of discrete enclosed wells, in the presence of the enzyme substrate, wherein

the enzymes are randomly distributed within the array, enzyme catalysis of the substrate produces a detectable signal and the wells are formed from contact between a movable element and a surface;

b) determining the rate of catalysis within each well and

c) summing the wells having a rate of catalysis above a background level so as to provide a digital readout.

This method may be performed using a first test solution and a second control solution having a known enzyme concentration, and comparing the digital readout for the first solution and second solution. In another aspect the invention relates to a method of determining the concentration of an enzyme in a solution, comprising:

trapping enzymes directly from A in an array of discrete, enclosed wells, in the presence of the enzyme substrate, wherein

i. the enzymes are randomly distributed within the array, ii. enzyme catalysis of the substrate produces a detectable signal and

iii. the wells are formed from contact between a movable element and a surface;

determining the rate of catalysis within each well and summing the wells having a rate of catalysis above a background level so as to provide a digital readout for A, and

trapping enzymes directly from a solution B of known enzyme concentration in an array of discrete enclosed wells, in the presence of the enzyme substrate, wherein

i. the enzymes are randomly distributed within the array;

ii. enzyme catalysis of the substrate produces a detectable signal and

iii. the wells are formed from contact between a movable element and a surface,

determining the rate of catalysis within each well and summing the wells having a rate of catalysis above a background level so as to provide a digital readout for B, and

STEP 3

Comparing the digital readout for STEP 1 and STEP 2.

In another aspect, the invention relates to a method for digitally representing the concentration of an analyte in a sample, comprising a) trapping the analyte directly from the analyte sample in an array of discrete enclosed wells, wherein

i. the analyte is immobilised within the wells, preferably by anti- analyte antibodies

ii. the analyte is randomly distributed within the array, and iii. the wells are formed from contact between a movable element and a surface

b) targeting an enzyme to the analyte containing wells of the array

c) trapping the enzyme in the analyte containing wells in the presence of its substrate, wherein enzyme catalysis of the substrate produces a detectable signal, d) determining the rate of catalysis within each well,

e) summing the wells having a rate of catalysis above a background level so as to provide a digital readout.

In another aspect the invention relate to a method for representing the concentration of an analyte in a sample, comprising:

• STEP 1

a) A method for digitally representing the concentration of an analyte in a sample, comprising

b) trapping the analyte directly from the analyte sample in an array of discrete enclosed wells, wherein

a. the analyte is immobilised within the wells, preferably by anti-analyte antibodies,

b. the analyte is randomly distributed within the array, and

c. the wells are formed from contact between a movable element and a surface

c) targeting an enzyme to the analyte containing wells of the array d) trapping the enzyme in the analyte containing wells in the presence of its substrate, wherein enzyme catalysis of the substrate produces a detectable signal,

e) determining the rate of catalysis within each well,

f) summing the wells having a rate of catalysis above a background level so as to provide a digital readout, and • STEP 2

An analogue method for representing the concentration of the analyte in the solution, the analogue method comprising:

a) mechanically trapping the analyte directly from the analyte sample b) targeting a detectable molecule to the trapped analyte

c) mechanically trapping the targeted detectable molecule

d) detecting the trapped detectable molecule so as to provide an analogue readout, wherein, the same sample is used in STEP 1 and STEP 2, preferably wherein STEP 1 and STEP 2 are performed synchronously.

In one embodiment of the invention the volume of the sample is less than 10ΟμΙ, preferably less than 50μΙ, more preferably less than 10μΙ, most preferably 5μΙ.

In one embodiment the detectable molecule of the method of the invention comprises a fluorophore and/or an anti-analyte antibody. The detectable signal may be a change in florescence intensity and a rate of catalysis within each well determined by imaging each well at 3 4, 5, 6 or 7 minute intervals, preferably for a total period of approximately 15, 20, 25, 30 or 45 minutes.

The enzyme of the invention may be β-Galactosidase, preferably wherein the enzyme substrate is Fluorescein di-3-D- galactopyranoside.

These methods for representing the concentration of an analyte in a sample may comprise performing the method of the invention using a first test sample and a second control sample having a known analyte concentration and comparing the digital readout for the first sample and second sample.

In another aspect the invention relates to a method of determining the concentration of an analyte in a solution (A), given a solution of a known concertation of the same analyte (B), comprising a) the detection method of the invention, wherein A is the analyte in a sample, b) the same method as a), wherein B is the analyte in a sample and

c) comparing the digital readout of a) and b). In one embodiment, the analyte of the invention may be an anti-viral antibody, more preferably an anti-Ebola antibody, more preferably anti-Ebola virus glycoprotein (GP) IgG antibody. The sample may comprise whole blood, urine, saliva, amniotic fluid, bile and/or cerebrospinal fluid. The sample is preferably not diluted. In one embodiment the sample has not been processed, for example by purification or cell separation.

In another embodiment, the invention relates to a method for diagnosing Ebola infection in a patient comprising

a) preforming the method of detection of the invention, wherein the analyte is anti-virus antibody, preferably anti-Ebola antibody, more preferably anti-Ebola virus glycoprotein (GP) IgG antibody and

b) preforming the method of detection of the invention, wherein the analyte is an inflammatory cytokine,

wherein the sample is obtained from the patient, preferably wherein the sample is whole blood. In another aspect, the invention relates to the device of the device, system or assay unit of the invention for use in any of the methods of the invention.

The proposed integrated microfluidic platform allows the quantitation of multiple protein biomarkers from a 5 μΙ whole blood sample obtainable by a pinprick. The proposed microfluidic device can also be combined with a low-cost, portable do-it-yourself (DIY) microfluidic control system and a cheap fluorescence universal serial bus (USB) microscope to enable diagnostic testing at the POC, at home, or in resource limited field-environments. The device has been extensively tested by performing single enzyme measurements and digital immunoassays in serum, achieving single molecule detection in both cases and sensitivity as low as -10 fM (330 fg/ml). The experimental results have also demonstrated the applicability of the device to clinically relevant tests by detecting anti-Ebola antibodies in whole blood and IgG/lgE antibodies in human serum for allergy testing. Modern diagnostics, such as the system, device and method described herein, reduces the costs of modern healthcare by diminishing subsequent health problems, reducing hospitalization and avoiding unnecessary treatment. An important part of sustainable healthcare depends on the development and commercialization of next generation diagnostic platforms, and the system, device, and method described herein provides for a low-cost, portable, fast, and performant diagnostic system that has a high economic potential because it enables personalized home-based diagnostics.

According to another aspect of the present invention, the microfluidic digital-analog hybrid immunoassay device consists of a flow and a control fluidic layer fabricated by multilayer soft lithography. The PDMS device measures 20 mm in length, 14 mm in width, ~4 mm in height and is bonded to a 25 mm x 75 mm glass slide. As shown in Figure 1A, the device includes 16 independent assay units, schematically shown as a circuit. The assay chambers contain a deflectable button membrane (analog MITOMI) in one half of the assay units and a deflectable button membrane patterned with femtoliter wells (digital MITOMI) in the other half, as shown in Figure 1 B. Analog detection is achieved using the MITOMI mechanism: increasing analyte concentration results in increasing fluorescence signal density under the MITOMI button, see upper section of Figure 1 C and Figure 2A, showing a schematic cross-sectional view of the MITOMI detection area. In digital-MITOMI, increasing analyte concentrations give rise to increasing numbers of positive wells containing one, two, or more molecules, see lower section of Figure 1 C and Figure 2B. As shown in Figures 2A and 2B, a sandwich immunoassay is performed with either fluorophore-labeled (analog MITOMI) or enzyme-conjugated (digital MITOMI) antibodies. As schematically shown in Figure 3 and 4, increasing analyte concentrations will lead to increasing signal densities in analog measurements, and to increasing numbers of positive femtowells in a digital measurement. Figure 5A and 5C show phase contrast images, and Figures 5B and 5D shows fluorescence images of the analog and digital MITOMI buttons with a scale. The scale bar shown in Figures 4, 5A, 5B is 150 μηι. The digital assay unit includes an exemplary 400 femtocells.

Each assay unit comprises 2 spotting chambers i, ii, a MITOMI button 1 , and an S-shaped peristaltic pump 7. The spotting chambers i, ii are pre-programmed with assay reagents by spotting these on a glass microarray and aligning the array to the microfluidic chip. Variants of the device include fully analog cells, as shown in Figure 6A or fully digital cells, as shown in Figure 6B. The inventors have previously shown that highly-multiplexed biomarker detection is possible using analog MITOMI (Volpetti et al. , 2015), and that such devices can be preprogrammed with reagents and stored at elevated temperatures for at least 2-3 weeks (Volpetti et al., 2015). In order to reduce assay time, we incorporated active mixing using a simple serpentine-shaped or S-shaped peristaltic pump 7. The integrated peristaltic pump 7 permits complete mixing of the spotted reagents in 1 minute, in contrast to the 2.5 hours required when solely relying on passive diffusion, as shown in Figure 7. Another aspect of the present invention is the provision of a micro-Fluidic Diagnostic System μFDS) and the operation thereof (Figure 8). Sample volumes for analysis can be as small as 5 μΙ of whole blood obtainable by a pinprick, as exemplarily shown in Figure 9. The blood sample is aspirated and transferred to a control device of the μFDS with a micro-hematocrit tube coated with heparin, as exemplarily shown in Figure 10. To load the sample into the control device of the μFDS, vacuum based loading can be used, by applying vacuum to the outlet of the control device and drawing the blood sample into and through the control device. The blood sample could be loaded directly onto the control device and on-chip biomarker quantitation could be performed without requiring any sample-pretreatment or removal of haematocytes. Prior to this approach, microfluidic approaches generally relied on either off-chip haematocyte removal or integrated on-chip separation approaches, which complicated chip design and assay implementation. According to an aspect of the present invention, it has been shown that cell- separation is not necessary, and as a consequence, simplifying the assay and enabling multiplexed biomarker detection from an ultra-low volume whole blood sample. To enable the use of the above described biomarker detection and diagnostic testing at the POC, in a home-based setting, for field tests, or in a resource-limited environment, a portableμFDS has been designed and tested. The μFDS includes a microfluidic control system (MCS), a fluorescence USB microscope, a portable computing device, for example but not limited to a netbook, portable computer, tablet computer, laptop computer, and the microfluidic device. A three-dimensional (3D) rendering, block diagram and user interface of the system are schematically depicted in Figure 11. Moreover, in another variant, the μFDS can be integrated such that one or more processors of the MCS take over all the functions of the portable computing device. In the variant shown, the MCS includes a custom designed printed circuit board shown as a circuit diagram in Figure 12 coupled to a low-cost Arduino™ microcontroller, a touch screen, and a battery pack with a capacity of 98 Wh. The MCS controls 24 solenoid valves, a compressor, a vacuum pump and two pressure regulators. The pneumatic subsystem generates two air pressures to drive the microfluidic valves at 15 psi and load reagents at 3 psi. Using commercially available electronic components, the MCS can be built at costs of about $1 ,666, with the solenoid valves of about $700, and the two pressure regulators about $343.84 contributing the majority of the cost. The MCS has an average power consumption of 25W allowing at least one complete test to be performed on a single battery charge. The MCS can be controlled directly via the touch sensitive display, or indirectly via the portable computing device, in the variant shown a netbook. The fluorescence USB microscope and netbook cost about $828 and about $134, respectively, bringing the hardware cost of the complete μFDS to about $2,628.

According to yet another aspect of the present invention, digital enzyme measurements are performed to count single enzyme molecules. It was first tested with experimental results whether it is possible to count single enzyme molecules with the digital- MITOMI buttons reproducing results, and the experimental results confirmed the initial proof-of-concept tests with β-Galactosidase (G) and Fluorescein di-β-D- galactopyranoside (FDG) in buffer. In order to provide for experimental results for the kinetics of substrate turnover by single βG molecules, an enzyme/substrate mixture (βG + FDG) as schematically shown in Figure 13 was introduced on chip after surface derivatization. Thereafter, the digital MITOMI buttons were closed and the fluorescent intensity in femtowells of 4 buttons was measured using a fluorescence microscope. The time elapsed between combining the enzyme-substrate solutions and acquiring the first image was 15 minutes and the buttons were imaged for 15 minutes with a frequency of 5 minutes. Wells that contained one or more active βG molecule displayed an increase in fluorescence intensity. Figure 14 shows histograms of the substrate turnover rates of βG molecules for different enzyme concentrations, with a bin size for all histograms being 50 s ^' inserts show representative microscope images of individual digital-MITOMI buttons. The histograms were consistent with Poisson statistics that are shown in Figure 15, as expected for a random distribution of molecules inside the wells. Summing the slopes of all femtowells above a background threshold permitted the correlation of the on-chip signal with the known solution concentration of βG as shown in Figure 16. The sum of turnover rates are plotted against βG concentration, and the dashed red line is a linear fit to the data (R2 = 0.99). Therefore, the results of the experimental tests have shown that solution phase single enzyme counting can be performed on an integrated digital-MITOMI assay. Figure 17A shows a schematic representation of the GFP sandwich assays performed on the microfluidic device and detected with digital-MITOMI, analog amplification, and analog-MITOMI is shown. The detection of GFP in 2% BSA PBS buffer solution over concentrations ranging from 10 fM to 5 pM. Next, in Figure 17B, the results of the digital-MITOMI GFP immunoassay are shown, and plotted is the sum of slopes of femtowells above background as a function of GFP concentration. For activity traces and corresponding histograms of GFP molecules see Figure 22. In Figure 17C, the results of analog-MITOMI GFP immunoassay are shown. Plotted is the intensity of Cy3 signal as function of GFP concentration. Next, in Figure 17D, the digital- MITOMI detection of GFP in human serum over concentrations ranging from 12 fM to 50 fM is shown. The intensity of the fluorescein signal is plotted as function of GFP concentration. These data were obtained with a research grade fluorescence microscope. In addition, in Figure 17E and F, detection of GFP in 2% BSA PBS buffer solution with concentrations ranging from 500 fM to 5 nM is shown. Figure 17E shows the results of analog amplification detection of GFP. Furthermore, in Figure 17F, the results of analog-MITOMI detection of GFP is shown. Data were obtained with a fluorescence USB microscope. LODs were determined from the signal of a negative control plus 3 standard deviations. Figure 17G shows representative fluorescence images of one assay unit taken with the USB microscope for different concentrations of GFP in 2% BSA PBS using the analog amplification method, the scale bar indicated being 350 μηι.

Having successfully conducted single enzyme counting, it has been determined whether it is possible to conduct digital immunoassays with the integrated digital-MITOMI device. A recent attempt at conducting digital immunoassays using an integrated microwell/membrane approach failed to give digital results and suggested that it was not possible to establish a sufficient seal between the membrane and the glass surface to allow digital measurements (Wang et al., 2013). A coupled analog-digital immunoassay has been developed and detected his-tagged GFP using a surface immobilized primary goat anti-GFP antibody, an Alexa Fluor 555 labeled mouse anti-Penta-His antibody, which in turn is detected using a βG conjugated goat anti- mouse IgG antibody.

Unexpectedly and surprisingly, successful digital immunoassays could be conducted using a surface based immunoassay and integrated digital-MITOMI. As discussed above, detection of as little as -10 fM (330 fg/ml) of GFP in a buffer solution using digital-MITOMI (LOD: 2.6 fM (86 fg/ml)) could be done, which is an improvement of two to three (2-3) orders of magnitude compared to detection of GFP with analog- MITOMI (LOD: 1 .7 pM (55.59 pg/ml)). Therefore, it has been shown that digital-MITOMI leads to a significant increase in sensitivity for microfluidic immunoassays.

In addition, combining digital-MITOMI and analog-MITOMI detection on a single platform increases the dynamic range, in this case covering a range ~6 orders of magnitude between 10 fM (330 fg/ml) to at least 5 nM (164 ng/ml) without requiring any sample dilutions or other processing steps. To test whether digital-MITOMI is compatible with clinical samples, GFP in human serum has been detected and observed at a similar level of sensitivity of 12 fM (400 fg/ml).

Although a microfluidic assay that requires relatively expensive microscope/optics and digital camera for analysis could be applied in a POC setting, with current market prices, it is not feasible to use such methods for personal, home-based diagnostics or in resource-limited environments, where hardware size, weight, and cost become important factors. Based on these cost limitations, experimental results were made using a relatively cheap, commercially available fluorescence USB microscope. The USB microscope was able to detect GFP via analog-MITOMI with a LOD of 562 pM (18.37 ng/ml).

In order to improve the sensitivity of the assay, a standard ELISA was performed, using R>G and FDG turnover for signal amplification. Combining on-chip signal amplification with a cheap USB microscope permitted detection of GFP concentrations as low as 5 pM (163.5 pg/ml) with an LOD of 1 .9 pM (62.13 pg/ml). This performance is comparable to that previously achieved by analog-MITOMI using a research grade optical microscope (Volpetti et al., 2015).

According to still another aspect of the present invention, multiplexed digital detection of anti- EBOLA antibodies in human serum can be provided. After digital detection of GFP in human serum, experimental tests were made to see whether digital detection in human serum was also possible for a clinically relevant biomarker. Ebola virus infection is diagnosed by detecting the presence of anti-Ebola virus glycoprotein (GP) IgG antibodies. Infectious hemorrhagic fevers caused by the Ebola virus results in mortality rates of up to 90%, and no effective vaccines or therapeutics are currently available. The highly infectious and lethal nature of this virus requires the development of novel diagnostic methods in order to monitor and control outbreaks. Although anti-Ebola IgG levels begin to rise 8-10 days after disease onset, detection of IgG antibodies represents a viable diagnostic test for symptomatic and asymptomatic individuals. An optimal Ebola diagnostic test has a multiplexed detection of virus antigen, IgM, IgG, and inflammatory cytokine levels, which could be readily achieved by the multiplexed microfluidic system. Due to the difficulty of obtaining relevant molecules and the biosafety hazard of working with actual Ebola patient samples, the experimental tests were limited to recombinant IgG antibodies specific to recombinant virus glycoproteins (GP).

To demonstrate the multiplexing capability of the system, experimental tests were made with a single device with Ebola GPs from 3 virus species: Bundibugyo, Reston, and Zaire, plus a BSA negative control. Details of the spotting and the immunoassay are shown in Figure 18A and Figure 18B, and more details on the methods and materials involved are discussed below. A concentration of 5 pM (75 pg/ml) anti-Bundibugyo GP IgG, 1 pM (15 pg/ml) anti-Reston GP IgG spiked into human serum, and a negative control sample were measured on 3 independent chips. For each test, the femtowells of 4 digital-MITOMI buttons of the corresponding capture agents (Bundibugyo GP, Reston GP, Zaire GP and BSA) were analyzed. When the human serum sample was spiked with anti-Bundibugyo IgG, a large number of femtowells programmed with Bundibugyo GP were positive. Very minor crosstalk could be observed with Reston and Zaire GP. Summing the fluorescence signal of positive wells for the 3 GPs and the negative control results in a clear signal for Bundibugyo, low signals for Reston and Zaire and baseline levels for the negative control. A similar result was obtained when testing the presence of anti- Reston GP IgG, which was more specific. The negative control sample showed no appreciable signal for any of the viral species tested. Therefore, based on these experimental results, digital- MITOMI provides a sensitive approach for successful rapid multiplexed diagnostic testing of biomarkers in human serum samples.

Furthermore, multiplexed detection of anti-Ebola IgG in mouse whole blood was been tested. To test whether the device could be used to detect Ebola specific IgG levels in a resource- limited, field-based setting, similar experimental tests as described above were performed, but the above described μFDS was used for device readout and biomarkers spiked into whole blood samples. The average quantitative load of specific IgG molecules during a viral infection ranges from 6.6 nM - 6.6 μΜ (1 - 1000 μg/ml), which defines the concentration range that the device needs to detect. Results of the multiplexed Ebola diagnostics in whole blood using the μFDS are shown in Figure 19 and 20. Figure 19A shows a schematic representation of the assay for anti-Ebola virus detection in a whole blood sample. Data for different concentrations are shown in Figure 19B for anti-Zaire Ebola GP IgG antibody, Figure 19C for anti-Bundibugyo Ebola GP IgG antibody and in Figure 19D for anti-Reston Ebola GP IgG antibody spiked in whole blood. Zaire, Bundibugyo and Reston glycoproteins are spotted in quadruplicate to yield two repetitions for the analog and the analog amplification detection mechanisms. Goat anti-rabbit IgG PE conjugated antibody and goat anti-rabbit IgG β-Galactosidase conjugated antibody were spotted in the top and bottom part of the chip, respectively, to allow the use of both detection methods on the same device. Data represent means (n=2). Error bars correspond to the standard deviation of the means. Figure 20 shows representative fluorescent images of 10 nM anti-Zaire Ebola GP IgG antibody spiked in whole blood. In Figure 20A, the analog-MITOMI signal was imaged using a fluorescence USB microscope with a 570 nm emission filter. In Figure 20B, the amplified analog signal was detected with a fluorescence USB microscope using a 510 nm emission filter, and the scale bar used is 700 μηι.

A series of experimental tests were performed, 1 experiment per concentration for a total 10 independent experiments over an IgG concentration range of 0.1 - 100 nM for the three Ebola species, Bundibugyo, Reston, and Zaire. Using analog-MITOMI the device was able to detect IgG antibodies specific to all three species at levels of 100 nM, with the exception of anti-Zaire IgG, which could be detected to a concentration of 1 nM. Employing on-chip enzymatic signal amplification improved the detection limit for all species to 10 nM and 0.1 nM for Zaire. The experimental results show that a dynamic range of three (3) orders of magnitude can be achieved in the range of 100 pM (0.15 μg/ml) to 100 nM (150 μg/ml) in unprocessed whole blood using a USB microscope which effectively covers the clinically relevant range.

According to another aspect of the present invention, multiplexed serological detection of IgE and IgG for allergy detection may be successfully tested with the device. Based on the background art, in vitro testing of specific IgE reactivaties for diagnosis of allergies is a routine serological analysis, while IgG detection, as a diagnostic indicator for allergies, remains controversial (Hantusch et al., 2005). Due to the widespread incidence rate of allergies in the developed world there is considerable interest in developing novel assays and devices that perform sensitive, fast and multiplexed allergy tests and do so in a POC or preferably a home- based setting. In serum the concentration of IgG is considerably higher than IgE, IgG thus readily outcompetes IgE for binding on the same immobilized allergen making IgE detection challenging. Commercial allergy tests are available based on lateral flow-based devices that allow the analysis of 10 IgE species, but require a large 1 10 μΙ whole blood, a dedicated instrument for readout, and the quantification is qualitative or semi quantitative (Eigenmann et al., 2009)

Therefore, according to another aspect of the invention, experimental results have been performed to demonstrate the ability of the device to perform detection of allergen-specific IgE and IgG, and a human serum sample was subjected to testing obtained from a patient diagnosed with peanut allergy. Generally, the device is capable of measuring up to 16 different allergens specific IgEs or up to 8 different allergens for parallel quantitation of IgE and IgG. In the experimental tests, the device was programmed with two repeats of three different allergens: Arachis hypogaea allergen 2 (Ara h2), Felis domesticus allergen 1 (Fel d1 ), Dermatophagoides pteronyssinus allergen 2 (Der p2), and a negative control, see schematic representation in Figure 21 A. The clinical serum sample (6 μΙ) was diluted 1 :2 in an incubation buffer and flowed in the chip following the protocol described below and IgE and IgG were detected as shown in Figure 21 B. In the top two rows of the device allergen bound IgE was measured, and the bottom two rows of the device were used to detect IgG. A clear signal was obtained for IgE and IgG bound to the recombinant allergen Ara h2, while the two other recombinant allergens Fel d1 and Der p2 showed no signal above the negative control. The signal obtained from IgG was considerably stronger than that of IgE, as expected given the vastly different concentrations of these molecules in serum. Figure 21 C shows the quantification of IgE levels for a sample positive to peanuts allergy and the corresponding fluorescent images. Figure 21 D shows the quantification of IgG levels for a sample positive to peanuts allergy and the corresponding fluorescent images. Bars are means (n=2), with individual data points shown as dots (·). Although the above tests were made with a research grade microscope, IgG levels were high enough to be observed directly using the low-cost fluorescence USB microscope indicating that a home-based, rapid allergy test is possible. As shown above, the diagnostic test is applicable in a POC setting to provide rapid and multiplexed analysis of a patient's allergic profile.

Accordingly, based on the above discussed experimental results, it has been shown that the system, device, and method according to the embodiments of the present invention are able to meet a number of challenging performance criteria including: i) high sensitivity, ii) high dynamic range, iii) multiplexed biomarker analysis, iv) compatibility with ultra-low volume (~5 μΙ) whole blood samples, and v) portable/low-cost control and readout instrumentation, providing for a system, device and method that can meet these criteria results in a high-performance diagnostic platform that can be applied in clinical or POC settings. Decreases in the costs of control and data readout systems allow such an assay to be performed in application fields were hardware costs are an important factor, and lead to wide-spread, democratized, personalized diagnostic assays, for example in a home-based setting. Additionally, portability and low-cost is important if diagnostic assays are to be performed in resource-limited environments. With the embodiments of the present invention, a digital-analog microfluidic diagnostic system, device, and method has been provided that unexpectedly and surprisingly improves the limit of detection by roughly 3 orders of magnitude to -10 fM, compared to previous MITOMI based immunoassays. By combining digital-MITOMI with standard analog- MITOMI on the same device, dynamic range of 6 orders of magnitude have been achieved. These levels of sensitivity and dynamic range can be achieved for a multiplexed analysis of a single, ultra-low volume (~5 μΙ) serum or whole blood sample. Microfluidic large-scale integration and MITOMI are also scalable based on the proposed design, and very high-throughput multiplexed immunoassays can be used to achieve up to 4,096 parallel immunoassays on a single device. The system, device, and method of the present invention can be configured to conduct 16 simultaneous and independent immunoassays, and drastically reduces assay time by incorporating active mixing. Therefore, the system, device, and method can achieve the above criteria of i) high sensitivity, ii) high dynamic range, iii) multiplexed analysis and iv) low sample volume compatibility. These performance characteristics, and the fact that essentially any protein biomarker for which an antibody pair exists, or can be developed, is measurable on the present system, device and method, and make it a viable and generically applicable POC diagnostic platform.

To further improve upon the fifth criterion v) of portability and low-cost hardware instrumentation, a DIY microfluidic control system has been developed, which can be assembled by a layperson from commercially available components. Programming the MCS is easily achieved via an Arduino board allowing completely automated control of the microfluidic device. The current cost of the MCS is about $1 ,666, which could be further reduced to about $1 ,322 by eliminating the expensive pressure regulators and touch display. For device readout, experimental tests have shown that a low-cost USB fluorescence microscope provides sufficient resolution and sensitivity to quantitate microfluidic diagnostic tests based on fluorescence readout, particularly when combined with on-chip signal amplification. The entire μFDS consisting of the MCS ($1 ,322), the USB fluorescence microscope ($828), and a netbook ($134) costs about $2,284. The complete μFDS system or device can be run on battery power alone and is completely self-sustained and portable. It has also recently been shown that MITOMI devices can be pre-programmed with antibody reagents and consequently stored at elevated temperatures of 40°C for at least 2 weeks without loss of function (Volpetti et al. , 2015), making the microfluidic device itself a self-contained reagent reservoir with a good shelf- life. Long-term storage of enzyme-based reagents is possible and could be easily implemented. Experimental results have shown similar sensitivities with the μFDS as with an expensive research grade fluorescence microscope and it has also been shown that the system could detect IgG antibody levels down to 100 pM in a 5 μΙ whole blood sample that can be obtained by a simple pinprick.

With the features of the system, device and method of the embodiments of the present invention, it is possible to further improve upon a couple of performance characteristics. With the experimental tests performed as discussed above, assay time usually requires a total of about 2 hours, which includes all necessary pre-assay surface derivatization steps, about 45 minutes, antibody immobilization steps / sample introduction, about 60 minutes, and readout of the device, about 10 minutes. However, it is possible to generate the required surface chemistry prior to device use, eliminating the first 45 minutes from the routine. Therefore, it is possible to introduce the sample within 25 minutes of beginning the test and results would be available within one hour of introducing the sample. Further improvements to the process flow and to reduce the times required for the various reagent and wash steps leads to the provision of test results within 30 minutes of introducing the sample on the device.

Based on the above discussed experimental results, a low-cost cheap fluorescence USB microscope can be used with improved magnification and/or optical resolution to permit the visualization of individual femtowells of the above described digital-MITOMI readout mechanism (Ghosh et al. , 201 1 ). It has been shown by the inventors that the fluorescent signal within the femtowells is sufficiently high to be detected with the fluorescence USB microscope used, but the resolution of the used USB microscope was not high enough to resolve individual femtowells. In a variant, a digital-MITOMI device could be used that contains larger wells or pitch individual femtowells far enough apart so that they can be resolved. Moreover, in another variant, the optical elements could be incorporated in the microfluidic device or could be placed directly on a CCD chip for detection.

Next, according still a further aspect of the present invention, different materials, test samples, manufacturing processes, operation, and other aspect the different embodiments are described. It is noted that the description given below is meant to be merely exemplary, and other variants thereof are also within the scope of the invention. For example, the different reagents, antibodies, proteins, and clinical samples include but are not limited to Chlorotrimethylsilane (92360), 2-propanol (IPA) (19516), β-Galactosidase from Escherichia coli (βΘ) (G3153), phosphate buffered saline (PBS) (P5244), β-mercaptoethanol (P6250), bovine serum albumin (BSA) (A3912), and Tween 20 (P1379) that were purchased from Sigma-Aldrich. Fluorescein di-3-D-galactopyranoside (FDG) (F-1 179) and goat anti-rabbit IgG PE conjugate (P- 2771 MP) were bought from Life Technologies. Biotinylated human serum albumin (HSA) (ab8033), goat anti-GFP antibody biotin conjugate (ab6658), goat anti-human IgG Fc (SureLight PE) (ab131612), goat anti-rabbit IgG βG conjugate (ab136774), and rabbit anti-goat IgG βG conjugate (ab136712) were obtained from Abeam. Rabbit anti-Bundibugyo Ebola virus glycoprotein IgG (BVGP41-A), rabbit anti-Reston Ebola virus glycoprotein peptide IgG (RVGP31 -A), recombinant Bundibugyo-Ebola virus glycoprotein (BVGP45-R-10) and recombinant Reston-Ebola virus glycoprotein (RVGP35-R-10) were acquired from Alpha Diagnostics. EZ-Link NHS-PEG4-Biotin and labeling kits (PI-21455), biotinylated BSA (29130), and neutravidin (31000) were purchased from Thermo Fisher Scientific. Mouse anti- Penta His biotin conjugate (34440) and Mouse anti-Penta His Alexa Fluor 555 conjugate (35310) were purchased from Qiagen. Rabbit anti-Ebola virus GP (1501003) and recombinant Ebola virus glycoprotein (0501-015) were purchased from IBT Bioservices. Goat anti-mouse IgG βG conjugate (401607) was bought from Merck Millipore and Enhanced Green Fluorescent Protein (EGFP) (4999100) from BioVision. Biotinylated natural Ara h2 (BI-AH2-1 ), recombinant Fel d1 (LTR-FD1 D-1 ), and recombinant Der p2 (RE-DP2A-1 ) were purchased from INDOOR biotechnologies. Micro hematocrit capillary tubes (7493 1 1 ) were purchased from VWR International. Human clinical samples were purchased from PlasmaLab International.

According to another aspect of the invention, the microfluidic device consists of a flow and a control layer. Molds for each layer were fabricated using standard lithography techniques on 4" silicon wafers. After exposure to 02 plasma the control layer mold was patterned with GM1050 SU-8 photoresist to a height of 30 μηι. The mold for the fluidic layer was fabricated by a three- step lithography process. Briefly, a micropost array, initially patterned to a height of 5 μηι using GM1040 SU8, was patterned on top of a ~ 10 μηι high SU8 layer. Then, -15 μηι high channel features were patterned with AZ9260 photoresist. Devices were cast in polydimethylsiloxane (PDMS) using multilayer soft lithography. PDMS was prepared at a 20: 1 ratio and spin-coated on the flow layer mold at 1900 rpm. PDMS at a 5: 1 ratio was cast on the control layer mold to a thickness of ~4 mm. Both layers were baked at 80°C for 30 minutes. The control layer was peeled off from its mold and manually aligned to the flow layer mold, followed by a baking step at 80°C for 90 minutes. The femtowells for digital-MITOMI were 5 μηι wide and 5 μηι tall, giving a volume of 100 fl_, with a pitch of 10 μηι.

Regarding the spotting and device alignment, microscope glass slides were coated with epoxy- silane, and the glass slides were not sonicated in toluene for 20 minutes. To generate microarrays, samples were pipetted into a 384-well plate and spotted on the epoxy-coated glass slide with a ~ 5 nl delivery volume spotting pin, for example the 946MP8XB from Arrayit, using a microarray robot QArray2 from Genetix, at 60% humidity. The microfluidic device was aligned on top of the spotted glass slide and bonded overnight at 40°C.

Moreover, with respect to the microfluidic diagnostic device, an exemplary device was made of an electronic printed circuit board (PCB) coupled with an Arduino Mega 2560 R3, two pressure controllers 990-005103-015 from Parker, an air pump TM40a from TOPSFLO, a vacuum pump E163- 11 -120, from Parker, a touch screen uLCD-43DT, 4D Systems, a pneumatic control subsystem S14-2936, Pneumadyne, and a PDMS device.

Moreover, with respect to the device function and operation, control lines on the device were primed with Phosphate Buffered Saline (PBS) at 5 psi. When the control lines were fully primed the pressure was increased to 25 psi. Flow lines were operated at 3 psi. For the initial surface derivatization steps the neck valves remained closed to avoid liquid from entering the chambers containing the spotted reagents. First, the surface area was derivatized by flowing a solution of either biotinylated BSA resuspended to 2 mg/ml in H20 or biotinylated HSA resuspended to 1 mg/ml in H20 for 10, followed by a 5 minutes 0.005% Tween 20 in PBS. Next a 500 g/ml neutravidin solution in PBS was flown for 10 minutes, followed by a 5 minutes 0.005% Tween 20 in PBS wash. The buttons were closed and all remaining accessible surface area was passivated with the same biotinylated solution as above for 10 minutes, followed by a 5 minutes 0.005% Tween 20 in PBS wash. To perform re-suspension and mixing of the spotted reagents and the assay chamber, first, valve 2 was closed, in order to isolate each unit, followed by opening valves 3 and 4, to allow mixing between the top spotting chamber and the assay chamber. Then the MITOMI button was opened and the pump activated. Once the mixing was completed, button and valves 3, 4, and 2 were closed sequentially. The same procedure was followed for mixing the bottom chamber, using valves 5 and 6 instead of 3 and 4 (Figure 7).

With respect to the processes for digital enzyme measurements, after surface derivatization various concentrations ofβG ranging from 0.05 pM (0.02 ng/ml) to 50 pM (23.26 ng/ml) and 50 μΜ of FDG were mixed and then flowed through the device. FDG was resuspended in a 100 mM phosphate buffer, including 1 mM MgCI2 and 50 mM β-mercaptoethanol. As a final step, the MITOMI buttons were actuated at 25 psi. Bright field and fluorescent images were taken 15 minutes after the closure of the buttons with an exposure time of 2000 ms. To establish a calibration curve with known βG , the observed rates of all wells above a threshold were summed defined by the negative control and plotted the sum of rates versus βG concentration.

Concerning the processes for microfluidic digital and analog immunoassay, more details of the experimental tests are provided. For the GFP immunoassay 15 μ I of each reagent were loaded into tygon tubing, connected to the device and flowed in sequence at 3 psi. After surface functionalization, a 66.7 nM (10 μg/ml) biotinylated goat anti-GFP solution in 2% BSA was flowed for 10 minutes and immobilized in the button region coated with neutravidin. To generate the standard curves, various concentrations of GFP ranging from 5 fM (0.17 pg/ml) to 5 nM (166.67 pg/ml) in 2% BSA, and 12 fM (0.4 pg/ml) to 50 fM (1.67 pg/ml) in human serum were flowed through the device for 10 minutes. Next, a 13 nM (2 μg/ml) mouse anti- Penta His Alexa Fluor 555 conjugate solution in 2% BSA was flowed for 10 minutes. After, a 100 pM (0.015 μg/ml) goat anti-mouse IgG G conjugate solution in 2% BSA was flowed for 8 minutes. The channels were washed with 0.005% Tween PBS flowed immediately after each reagent for 5 minutes. After the last washing step 50 μΜ FDG was flowed for 4 minutes and buttons closed with 25 psi pressure. Bright field and fluorescent images were taken 15 minutes after the closure of the buttons with an exposure time of 2000 ms.

Next, more details are provided with respect to the experimental tests of the multiplexed digital detection of anti-EBOLA antibodies in human serum, according to another aspect of the embodiments of the present invention. The recombinant virus antigens, along with the detection antibodies, were printed in a 4 x 4 format on epoxy slides. Briefly, three (3) different Ebola virus GPs and a BSA negative control in all upper spotting chambers of the unit have been spotted. Goat anti-rabbit IgG βG conjugated antibodies were printed in the lower spotting chamber of the unit cells. After surface functionalization, 15 μΙ of each reagent were loaded into individual tygon tubings, connected to the device and flowed sequentially at 3 psi for 10 minutes. First, a 7 nM (0.45 μg/ml) biotinylated mouse anti-Penta His solution in 2% BSA was flowed and immobilized in the button region coated with neutravidin. Then, 1.9 μΜ (136.8 μg/ml) of recombinant Reston GP, 1.9 μΜ (187.7 μg/ml) of recombinant Bundibugyo GP, and 1 μΜ (73 μg/ml) of recombinant Zaire GP were pumped for 10 minutes. 5 pM (75 pg/ml) of rabbit anti- Bundibugyo and 1 pM (15 pg/ml) of rabbit anti-Reston in human serum were flowed for 8 minutes, one species per chip. Next, a 100 pM (15 ng/ml) goat anti-rabbit βG antibody 2% BSA was pumped for 10 minutes. The channels were washed with 0.005% Tween 20 in PBS flowed after each reagent for 5 minutes. After the last wash step 50 μΜ FDG was flowed for 4 minutes and buttons were closed with 25 psi pressure. Bright field and fluorescent images were taken 15 minutes after the closure of the buttons with an exposure time of 2000 ms.

Moreover, with respect to the experimental tests regarding the multiplexed detection of anti- Ebola IgG in whole blood, in the multiplexed Ebola immunoassay, 3 different Ebola virus GPs and a BSA negative control in the upper spotting chambers were spotted, element A, chamber i, of the unit cells. Secondary antibodies were printed in the bottom spotting chambers, element A, chamber ii of the units. PE conjugated antibodies were spotted in the top half of the chip, to allow analog detection while βG conjugated antibodies where spotted in the bottom half of the chip to allow analog-amp detection. After surface functionalization, 15 μ I of each reagent were loaded into tygon tubing, connected to the device and flowed sequentially at 3 psi for 10 minutes. First, a 7 nM (0.45 μg/ml) biotinylated mouse anti-Penta-His solution in 2% BSA were immobilized in the button region coated with neutravidin. Then, 1.9 μΜ (136.8 μg/ml) of recombinant Reston GP, 1.9 μΜ (187.7 μg/ml) of recombinant Bundibugyo GP, and 1 μΜ (73 μg/ml) of recombinant Zaire GP were pumped for 10 minutes. To generate the standard curves, different concentrations, ranging from 0.1 nM (0.15 μg/ml) to 100 nM (150 μg/ml), of anti- Bundibugyo, anti-Reston, and anti-Zaire were spiked in mouse whole blood. A final volume of 3.5 μΙ was then flown through the device from a hematocrit tube that in turn was plugged into the device. The mouse whole blood sample from the hematocrit tube was loaded into the device by vacuum suction (8 minutes at -2.5 psi) as opposed to pressure driven flow. Next, 30 nM (1 1.7 μ9/ηιΙ) goat anti-rabbit IgG PE conjugated antibody and 2 nM (0.3 μg/ml) of goat anti- rabbit βG antibody in 2% BSA were pumped for 10 minutes in the upper and lower half of the device, respectively. The channels were washed with 0.005% Tween 20 in PBS immediately after each reagent for 5 minutes. After the last washing step 50 μΜ FDG was flowed and buttons immediately closed with 25 psi pressure. Bright field and fluorescent images were taken with the USB fluorescence microscope 15 minutes after the closure of the buttons with the following exposure settings: Lowest, Luma 240.

With respect to the experimental tests regarding multiplexed serological detection of IgE and IgG, three allergens, natural Ara h2, recombinant Fel d1 , recombinant Der p2, and a BSA negative control were spotted in order to perform an allergy diagnostic test. The human clinical serum sample was diluted 1 :2 (6.7 μΙ in 13.4 μΙ) with incubation buffer (0.05 M Tris/HCI pH 7.6, 0.15 M NaCI, Tween 20 0.02% v/v, BSA 1 % w/v) (49). After surface functionalization, 15 μΙ of each reagent were loaded into tygon tubing, connected to the device, and flowed sequentially at 3 psi for 10 minutes. First, 30.5 μΜ (0.55 mg/ml) of Ara h2, 41.15 μΜ (0.75 mg/ml) of Fel d1 , and 42.8 μΜ (0.6 mg/ml) of Der p2 were circulated through each unit cell by the S-shaped peristaltic pump for 20 minutes and immobilized in the button region coated via neutravidin. Then the clinical sample was flowed for 5 minutes. Next, 1 :100 rabbit anti-human IgE in 2% BSA and 30 nM (11.7 g/ml) of anti-rabbit IgG PE conjugated antibody in 2% BSA were flowed on the top half of the device while 6 nM (2.4 μg/ml) of goat anti-human IgG PE conjugated antibody in 2% BSA was flowed on the lower half for 10 minutes. Channels were washed with 0.005% Tween 20 in PBS after each step for 5 minutes. After the last washing step the buttons were closed with 25 psi pressure. Bright field and fluorescent images were taken using the Nikon microscope with an exposure time of 2000 ms for IgE and 300 ms for IgG.

Furthermore, additional details with respect to the experimental tests and the image acquisition, quantification and data analysis are provided. Pumping and mixing sequences were recorded with a Hamamatsu ORCA-Flash4.0 camera C1 1440. Mixing time was determined by analyzing fluorescent intensity changes over time in a portion of the microfluidic channel during pumping. The sandwich valve was closed and a Cy3 labeled solution was pumped in the assay chamber containing buffer. Fluorescent images were obtained Nikon ECLIPSE Ti microscope equipped with LEDFIuorescent ExcitationSystem, Cy3 and FITC filter sets, and a Hamamatsu ORCA- Flash4.0 camera, C1 1440. Images were taken with a 40x objective lens from Nikon, SPIan Fluor, ELWD 40x/0.60, °°/0.2, WD 3.6-2.8. Alternatively, fluorescent images were taken with a Cy3, AM4113T-YFBW) or FITC (AM41 13T-GFBW) Dino-Lite USB fluorescence microscopes. The resulting TIFF-images were analyzed with a microarray image analysis software from GenePix Pro v6.0, Molecular Devices, ImageJ, Fiji or Matlab from Mathworks. Averages, SDs and linear fits were calculated with Microsoft Excel. Vertical scatterplots, histograms and statistical analysis were created and performed in Prism v5.0 from Graphpad.

Figure 6 shows microfabricated diagnostic devices according to an aspect of the embodiments of the present invention. In section A of Figure 6 a design schematic of the analog MITOMI microfluidic device is represented showing flow (blue) and control (grey) layers. The device has four rows; each row contains four unit cells for a total of 16 analog assay units. Each assay unit (inset) contains: a MITOMI button valve 1 , a sandwich valve 2, two neck valves 3, 4 for the top chamber i and two 5, 6 for the bottom one ii, and a peristaltic pump 8. Section B shows the digital MITOMI device, and section C shows a digital-analog hybrid MITOMI device, with 8 digital and 8 analog units.

A multiplexed microfluidic diagnostic system, device, and method are provided with unprecedented and unexpected performance characteristics. The system, device, and method can be controlled and interrogated with low-cost off-the-shelf hardware, making it a viable diagnostic test in POC settings, and enables the introduction of diagnostics tests into the privacy and convenience of one's own home for personalized diagnostics, or deployment in resource- limited environments. The system is applicable to any protein biomarker or molecule that can be detected via an immunoassay based affinity reagent, which encompasses an extremely broad range of possible tests that can be done by the system. Accordingly, the system meets the requirements for new assays to meet the increasing demand for on-site diagnostic testing, companion diagnostics, generic health monitoring, as well as applications in biodefense and monitoring of endemic, epidemic, and pandemic disease outbreaks. The present system, device and method is furthermore compatible to be used with other bodily fluids other than blood, including, but not limited to, urine, saliva, amniotic fluid, bile and cerebrospinal fluid. EXAMPLES

Example 1. Design of the device

1.1. Structure of the device

An example microfluidic digital-analog hybrid immunoassay device is described. The device consists of a flow and a control fluidic layer fabricated by multilayer soft lithography. The PDMS device preferably measures 20 mm in length, 14 mm in width, ~4 mm in height and is bonded to a 25 mm x 75 mm glass slide. As shown in Figure 1A, the device includes 16 independent assay units, schematically shown as a circuit. The assay chambers contain a deflectable button membrane or moveable element (analog MITOMI) in one half of the assay units and a deflectable button membrane or moveable element patterned with femtoliter wells (digital MITOMI) in the other half, as shown in Figure 1 B and IC. Figure 2 depicts phase contrast microscope images of an analog (Figure 2A) and digital (Figure 2B) assay unit according to an aspect of the embodiments.

1.2. Detection assays

Analog detection is achieved using moveable elements: increasing analyte concentration results in increasing fluorescence signal density under the moveable element (button membrane). In digital detection (digital MITOMI), increasing analyte concentrations give rise to increasing numbers of positive wells containing one, two, or more molecules. In particular, the device may be used to perform a sandwich immunoassay with either fluorophore-labelled (analog MITOMI) or enzyme-conjugated (digital MITOMI) antibodies (Figure 3). Increasing analyte concentrations leads to increasing signal densities in analog measurements, and to increasing numbers of positive femtowells in a digital measurement (Figure 4), as can be detected using phase contrast or fluorescent microscopy (Figure 5).

1.3. Assay unit

Each assay unit comprises two spotting chambers, a MITOMI button and an S-shaped peristaltic pump. The spotting chambers are pre-programmed with assay reagents spotted on a glass microarray. Variants of the device include fully analog cells (Figure 6A) or fully digital cells, as shown in (Figure 6B). The inventors have previously shown that highly-multiplexed biomarker detection is possible using analog MITOMI and that these devices can be pre- programmed with reagents and stored at elevated temperatures for at least 2-3 weeks (Volpetti et al., 2015). 1.4. S-shaped peristaltic pump

In order to reduce assay time, the inventors incorporated active mixing into the assay unit using a serpentine-shaped or S-shaped peristaltic pump. The integrated peristaltic pump permits complete mixing of the spotted reagents in 1 minute, as opposed to 2.5 hours required when solely relying on passive diffusion (Figure 7).

1.5. μFDS

Another aspect of the present invention is the provision of a micro-Fluidic Diagnostic System μFDS) and the operation thereof. To enable the use of the above described biomarker detection and diagnostic testing at the point-of-care (POC), in a home-based setting, for field tests, or in a resource-limited environment, a portable microFluidic Diagnostic System μFDS) has been designed and tested. The μFDS includes a microfluidic control system (MCS), a fluorescence USB microscope, a portable computing device, for example but not limited to a netbook, portable computer, tablet computer, laptop computer and the microfluidic device (Figure 8).

Sample volumes for analysis can be as small as 5 μΙ of whole blood obtainable by a pinprick (Figure 9). The blood sample is aspirated and transferred to a control device of the μFDS with a micro-haematocrit tube coated with heparin (Figure 10). To load the sample into the control device of the μFDS , vacuum based loading can be used, by applying vacuum to the outlet of the control device and drawing the blood sample into and through the control device. The blood sample can be loaded directly onto the control device and on-chip biomarker quantitation can be performed without requiring any sample pre-treatment or removal of haematocytes. Prior to this approach, microfluidic approaches generally relied on either off-chip haematocyte removal or integrated on-chip separation approaches, which complicated chip design and assay implementation. According to an aspect of the present invention, it has been shown that cell- separation is not necessary, simplifying the assay and enabling multiplexed biomarker detection from an ultra-low volume whole blood sample.

A 3D rendering, block diagram and user interface of the μFDS are depicted in Figure 11. In another variant, the μFDS can be integrated such that one or more processors of the MCS take over all the functions of the portable computing device. In the variant shown, the MCS includes a custom designed printed circuit board shown as a circuit. 1.6. MCS

In one embodiment the MCS comprises a custom designed printed circuit board (Figure 12) coupled to a low-cost Arduino™ microcontroller, a touch screen, and a battery pack with a capacity of 98 Wh. The MCS controls 24 solenoid valves, a compressor, a vacuum pump and two pressure regulators. The pneumatic subsystem generates two air pressures to drive the microfluidic valves at 15 psi and load reagents at 3 psi.

1.7. Cost of the FDS

Using commercially available electronic components, the MCS can be built at costs of about $1 ,666, with the solenoid valves of about $700, and the two pressure regulators about $343.84 contributing the majority of the cost. The MCS has an average power consumption of 25W allowing at least one complete test to be performed on a single battery charge. The MCS can be controlled directly via the touch sensitive display, or indirectly via the portable computing device, in the variant shown a netbook. The fluorescence USB microscope and netbook cost about $828 and about $134, respectively, bringing the hardware cost of the complete \FDS to about $2,628.

Example 2. Detection assays

In one embodiment, the device of the present invention can perform digital enzyme measurements to count single enzyme molecules. As a proof of concept, the device was used to detect single β-Galactosidase (G) molecules. 2.1. Digital detection of single enzyme molecules

In order to assess the kinetics of substrate turnover by enzyme molecules, an enzyme/substrate mixture (βΘ + Fluorescein di-β-Ο- galactopyranoside (FDG)) in buffer was introduced on chip after surface derivatization (Figure 13). The digital MITOMI buttons were then closed and the fluorescent intensity in femtowells of 4 buttons was measured using a fluorescence microscope.

The time elapsed between combining the enzyme-substrate solutions and acquiring the first image was 15 minutes and the buttons were imaged for 15 minutes with a frequency of five minutes. Wells that contained one or more active βG molecule displayed an increase in fluorescence intensity.

Figure 14 shows histograms of the substrate turnover rates of βG molecules for different enzyme concentrations, with a bin size for all histograms being 50 s ^~ inserts show representative microscope images of individual digital-MITOMI buttons with a scale bar of 50 μηι. The histograms were consistent with a Poisson distribution (Figure 15), as expected for a random distribution of molecules inside the wells. The on-chip signal correlated with the known solution concentration of βG (Figure 16), demonstrating that solution phase single enzyme counting can be performed on an integrated digital-MITOMI assay.

2.2. Digital immunoassay

A recent attempt at conducting digital immunoassays using an integrated microwell/membrane approach failed to give digital results and strongly suggested that it is not possible to establish a sufficient seal between the membrane and the glass surface to allow digital measurements (Wang et al., 2013).

However, having successfully conducted single enzyme counting, the inventors next sought to determine whether it was also possible to conduct digital immunoassays with the integrated digital-MITOMI platform. Digital MITOMI, analog MITOMI and analogue amplification methods were used to detect soluble GFP using immobilized anti-GFP (Figure 17A). Surprisingly, GFP could be detected by digital MITOMI at a concentration ranging from 10 fM to 5 pM (Figure 17B), with a far lower LOD than analog MITOMI (Figure 17C).

Digital MITOMI could also be used to detect GFP in human serum (Figure 17D), over concentrations ranging from 12 fM to 50 fM, at a sensitivity of 12 fM (400 fg/ml). The detection limit of digital MITOMI was higher than that of analogue MITOMI (Figure 17E-G).

2.3. Coupled analog-digital immunoassay

Having successfully conducted single enzyme counting and digital MITOMI using the device, the inventors next sought to determine whether it was possible to also conduct a coupled analog-digital immunoassay. For this purpose, the inventors developed a coupled analog-digital immunoassay comprising detecting his-tagged GFP using a surface immobilized primary goat anti-GFP antibody (an Alexa Fluor 555 labeled mouse anti-Penta-His antibody, analog MITOMI detection), which in turn is detected using a βG conjugated goat anti-mouse IgG antibody (digital and analog amplification detection). Combining digital-MITOMI and analog-MITOMI detection on a single platform increased the dynamic range of the assay. The assay covered between 10 fM (330 fg/ml) to at least 5 nM (164 ng/ml) without requiring any sample dilutions or other processing steps, a range of ~6 orders of magnitude.

2.4. Clinical samples

A relatively cheap, commercially available fluorescence USB microscope was able to detect GFP via analog-MITOMI, with a LOD of 562 pM (18.37 ng/ml). In order to improve sensitivity, a standard ELISA was performed, using βG and FDG turnover for signal amplification. Combining this with on-chip signal amplification with the cheap USB microscope permitted detection of GFP concentrations as low as 5 pM (163.5 pg/ml) with an LOD of 1.9 pM (62.13 pg/ml). This performance is comparable to that previously achieved by analog-MITOMI and a research grade optical microscope (Volpetti et al., 2015).

Example 3. Digital detection of analyte

The inventors assessed whether digital detection in human serum was possible for a clinically relevant biomarker, in particular, anti-Ebola antibodies. Ebola virus infection is diagnosed by detecting the presence of anti-Ebola virus glycoprotein (GP) IgG antibodies. Infectious hemorrhagic fevers caused by the Ebola virus results in mortality rates of up to 90%, and no effective vaccines or therapeutics are currently available. The highly infectious and lethal nature of this virus requires the development of novel diagnostic methods in order to monitor and control outbreaks.

Although anti-Ebola IgG levels begin to rise 8-10 days after disease onset, detection of IgG antibodies does represent a viable diagnostic test for symptomatic and asymptomatic individuals. An optimal Ebola diagnostic test has a multiplexed detection of virus antigen, IgM, IgG, and inflammatory cytokine levels, which could be readily achieved by the multiplexed microfluidic system. Due to the difficulty of obtaining relevant molecules and the biosafety hazard of working with actual Ebola patient samples, the experimental tests were limited to recombinant IgG antibodies specific to recombinant virus glycoproteins (GP). 3.1. Detection in human serum

A single device was used to detect Ebola GPs in human serum from three virus species: Bundibugyo, Reston, and Zaire, plus a BSA negative control (Figure 18A, B). A concentration of 5 pM (75 pg/ml) anti-Bundibugyo GP IgG, 1 pM (15 pg/ml) anti-Reston GP IgG was spiked into human serum, and a negative control sample were measured on 3 independent chips. For each test, the femtowells of 4 digital-MITOMI buttons of the corresponding capture agents (Bundibugyo GP, Reston GP, Zaire GP, and BSA) were analyzed. When the human serum sample was spiked with anti-Bundibugyo IgG, a large number of femtowells programmed with Bundibugyo GP were positive (Figure 18C). Very minor crosstalk could be observed with Reston and Zaire GP. Summing the fluorescence signal of positive wells for the 3 GPs and the negative control results in a clear signal for Bundibugyo, low signals for Reston and Zaire, and baseline levels for the negative control. A similar result was obtained for anti-Reston GP IgG (Figure 18D). The negative control sample showed no appreciable signal for any of the viral species tested (Figure 18E). Digital-MITOMI therefore provides a sensitive approach for successful rapid multiplexed diagnostic testing of biomarkers in human serum samples.

3.2. Detection of biomarkers in whole blood

3.2.1. Ebola antibodies

Multiplexed detection of anti-Ebola IgG in mouse whole blood was assessed, to determine whether the device could be used to detect Ebola specific IgG levels in a resource- limited, field- based setting, the device was used to detect biomarkers spiked into whole blood samples (Figure 19A). The average quantitative load of specific IgG molecules during a viral infection ranges from 6.6 nM - 6.6 μΜ (1 - 1000 μg/ml), which defines the concentration range that the device needs to detect.

Using analog-MITOMI the device was able to detect IgG antibodies specific to all three virus species at levels of 100 nM, with the exception of anti-Zaire IgG, which could be detected to a concentration of 1 nM (Figure 19B-C). Employing on-chip enzymatic signal amplification improved the detection limit for all species to 10 nM and 0.1 nM for Zaire. Micrographs obtained with the fluorescence USB microscope for 10 nM anti-Zaire Ebola IgG for the analog-MITOMI (Figure 20A) and the analog amplification based method (Figure 20B). These data demonstrate that a dynamic range of 3 orders of magnitude can be achieved in the range of 100 pM (0.15 μg/ml) to 100 nM (150 μg/ml) in unprocessed whole blood using a USB microscope which effectively covers the clinically relevant range. 3.2.2. Allergy antibodies

Multiplexed serological detection of IgE and IgG for allergy detection was performed. In vitro testing of specific IgE reactivaties for diagnosis of allergies is a routine serological analysis, while IgG detection, as a diagnostic indicator for allergies, remains controversial (Hantusch et al., 2005). Due to the widespread incidence rate of allergies in the developed world there is considerable interest in developing novel assays and devices that perform sensitive, fast and multiplexed allergy tests and do so in a POC or preferably a home-based setting. In serum the concentration of IgG is considerably higher than IgE, IgG thus readily outcompetes IgE for binding on the same immobilized allergen making IgE detection challenging. Commercial allergy tests are available based on lateral flow-based devices that allow the analysis of 10 IgE species, but require a large 1 10 μΙ whole blood, a dedicated instrument for readout, and the quantification is qualitative or semi-quantitative (Eigenmann et al., 2009).

Human serum sample was obtained from a patient diagnosed with peanut allergy. The device was capable of measuring up to sixteen different allergens specific IgEs or up to 8 different allergens for parallel quantitation of IgE and IgG. The device was programmed with two repeats of three different allergens: Arachis hypogaea allergen 2 (Ara h2), Felis domesticus allergen 1 (Fel d1 ), Dermatophagoides pteronyssinus allergen 2 (Der p2), and a negative control (Figure 21 A). The clinical serum sample (6 μΙ) was diluted 1 :2 in an incubation buffer and flowed in the chip following the protocol described below and IgE and IgG were detected (Figure 21 B). In the top two rows of the device allergen bound IgE was measured, and the bottom two rows of the device were used to detect IgG. A clear signal was obtained for IgE and IgG bound to the recombinant allergen Ara h2, while the two other recombinant allergens Fel d1 and Der p2 showed no signal above the negative control. The signal obtained from IgG was considerably stronger than that of IgE, as expected given the vastly different concentrations of these molecules in serum. These assays were performed using a research grade microscope. However, the IgG levels were high enough to be observed directly using the low-cost fluorescence USB microscope indicating that a home-based, rapid allergy test is possible. The diagnostic test is therefore applicable in a POC setting to provide rapid and multiplexed analysis of a patient's allergic profile. Materials and Methods

Reagents, antibodies, proteins and clinical samples

Chlorotrimethylsilane (92360), 2-propanol (IPA) (19516), β-Galactosidase from Escherichia coli (βG) (G3153), phosphate buffered saline (PBS) (P5244), β-mercaptoethanol (P6250), bovine serum albumin (BSA) (A3912), and Tween 20 (P1379) that were purchased from Sigma-Aldrich. Fluorescein di-β-D-galactopyranoside (FDG) (F-1 179) and goat anti-rabbit IgG PE conjugate (P- 2771 MP) were bought from Life Technologies. Biotinylated human serum albumin (HSA) (ab8033), goat anti-GFP antibody biotin conjugate (ab6658), goat anti-human IgG Fc (SureLight■ PE) (ab131612), goat anti-rabbit IgG βG conjugate (ab136774), and rabbit anti- goat IgG βG conjugate (ab136712) were obtained from Abeam. Rabbit anti-Bundibugyo Ebola virus glycoprotein IgG (BVGP41-A), rabbit anti-Reston Ebola virus glycoprotein peptide IgG (RVGP31 -A), recombinant Bundibugyo-Ebola virus glycoprotein (BVGP45-R-10) and recombinant Reston-Ebola virus glycoprotein (RVGP35-R-10) were acquired from Alpha Diagnostics. EZ-Link NHS-PEG4-Biotin and labeling kits (PI-21455), biotinylated BSA (29130), and neutravidin (31000) were purchased from Thermo Fisher Scientific. Mouse anti- Penta His biotin conjugate (34440) and Mouse anti-Penta His Alexa Fluor 555 conjugate (35310) were purchased from Qiagen. Rabbit anti-Ebola virus GP (1501003) and recombinant Ebola virus glycoprotein (0501-015) were purchased from IBT Bioservices. Goat anti-mouse IgG βG conjugate (401607) was bought from Merck Millipore and Enhanced Green Fluorescent Protein (EGFP) (4999100) from BioVision. Biotinylated natural Ara h2 (BI-AH2-1 ), recombinant Fel d1 (LTR-FD1 D-1 ), and recombinant Der p2 (RE-DP2A-1 ) were purchased from INDOOR biotechnologies. Micro hematocrit capillary tubes (7493 1 1 ) were purchased from VWR International. Human clinical samples were purchased from PlasmaLab International.

Microfabrication

Molds for each layer were fabricated using standard lithography techniques on 4" silicon wafers. After exposure to 02 plasma the control layer mold was patterned with GM1050 SU-8 photoresist to a height of 30 μηι. The mold for the fluidic layer was fabricated by a three-step lithography process. Briefly, a micropost array, initially patterned to a height of 5 μηι using GM1040 SU8, was patterned on top of a ~ 10 μηι high SU8 layer. Then, -15 μηι high channel features were patterned with AZ9260 photoresist. Devices were cast in polydimethylsiloxane (PDMS) using multilayer soft lithography. PDMS was prepared at a 20: 1 ratio and spin-coated on the flow layer mold at 1900 rpm. PDMS at a 5:1 ratio was cast on the control layer mold to a thickness of ~4 mm. Both layers were baked at 80°C for 30 minutes. The control layer was peeled off from its mold and manually aligned to the flow layer mold, followed by a baking step at 80°C for 90 minutes. The femtowells for digital-MITOMI were 5 μηι wide and 5 μηι tall, giving a volume of 100 fl, with a pitch of 10 μηι.

Spotting and device alignment

Regarding the spotting and device alignment, microscope glass slides were coated with epoxy- silane, and the glass slides were not sonicated in toluene for 20 minutes. To generate microarrays, samples were pipetted into a 384-well plate and spotted on the epoxy-coated glass slide with a ~ 5 nl delivery volume spotting pin, for example the 946MP8XB from Arrayit, using a microarray robot QArray2 from Genetix, at 60% humidity. The microfluidic device was aligned on top of the spotted glass slide and bonded overnight at 40°C.

MDS

The system was made of an electronic printed circuit board (PCB) coupled with an Arduino™ Mega 2560 R3, two pressure controllers 990-005103-015 from Parker, an air pump TM40a from TOPSFLO, a vacuum pump E163- 1 1-120, from Parker, a touch screen uLCD-43DT, 4D Systems, a pneumatic control subsystem S14-2936, Pneumadyne, and a PDMS device.

Device function and operation

Control lines on the device were primed with Phosphate Buffered Saline (PBS) at 5 psi. When the control lines were fully primed the pressure was increased to 25 psi. Flow lines were operated at 3 psi. For the initial surface derivatization steps the neck valves remained closed to avoid liquid from entering the chambers containing the spotted reagents. First, the surface area was derivatized by flowing a solution of either biotinylated BSA resuspended to 2 mg/ml in H20 or biotinylated HSA resuspended to 1 mg/ml in H20 for 10, followed by a 5 minutes 0.005% Tween 20 in PBS. Next a 500 μg/ml neutravidin solution in PBS was flown for 10 minutes, followed by a 5 minutes 0.005% Tween 20 in PBS wash. The buttons were closed and all remaining accessible surface area was passivated with the same biotinylated solution as above for 10 minutes, followed by a 5 minutes 0.005% Tween 20 in PBS wash.

To perform re-suspension and mixing of the spotted reagents and the assay chamber, first, valve 2 was closed, in order to isolate each unit, followed by opening valves 3 and 4, to allow mixing between the top spotting chamber and the assay chamber. Then the MITOMI button was opened and the pump activated. Once the mixing was completed, button and valves 3, 4, and 2 were closed sequentially. The same procedure was followed for mixing the bottom chamber, using valves 5 and 6 instead of 3 and 4 (Figure 7).

Digital enzyme measurements

after surface derivatization various concentrations of β raGnging from 0.05 pM (0.02 ng/ml) to 50 pM (23.26 ng/ml) and 50 μΜ of FDG were mixed and then flowed through the device. FDG was resuspended in a 100 mM phosphate buffer, including 1 mM MgCI2 and 50 mM β- mercaptoethanol. As a final step, the MITOMI buttons were actuated at 25 psi. Bright field and fluorescent images were taken 15 minutes after the closure of the buttons with an exposure time of 2000 ms. To establish a calibration curve with known β,G the observed rates of all wells above a threshold were summed defined by the negative control and plotted the sum of rates versus βG concentration.

Microfluidic digital and analog immunoassay

Concerning the processes for microfluidic digital and analog immunoassay, more details of the experimental tests are provided. For the GFP immunoassay 15 μ I of each reagent were loaded into tygon tubing, connected to the device and flowed in sequence at 3 psi. After surface functionalization, a 66.7 nM (10 μg/ml) biotinylated goat anti-GFP solution in 2% BSA was flowed for 10 minutes and immobilized in the button region coated with neutravidin. To generate the standard curves, various concentrations of GFP ranging from 5 fM (0.17 pg/ml) to 5 nM (166.67 pg/ml) in 2% BSA, and 12 fM (0.4 pg/ml) to 50 fM (1.67 pg/ml) in human serum were flowed through the device for 10 minutes. Next, a 13 nM (2 μg/ml) mouse anti- Penta His Alexa Fluor 555 conjugate solution in 2% BSA was flowed for 10 minutes. After, a 100 pM (0.015 μg/ml) goat anti-mouse IgG G conjugate solution in 2% BSA was flowed for 8 minutes. The channels were washed with 0.005% Tween PBS flowed immediately after each reagent for 5 minutes. After the last washing step 50 μΜ FDG was flowed for 4 minutes and buttons closed with 25 psi pressure. Bright field and fluorescent images were taken 15 minutes after the closure of the buttons with an exposure time of 2000 ms.

Multiplexed digital detection of anti-Ebola antibodies in human serum

The recombinant virus antigens, along with the detection antibodies, were printed in a 4 x 4 format on epoxy slides. Briefly, three (3) different Ebola virus GPs and a BSA negative control in all upper spotting chambers of the unit have been spotted. Goat anti-rabbit IgG G conjugated antibodies were printed in the lower spotting chamber of the unit cells. After surface functionalization, 15 μ I of each reagent were loaded into individual tygon tubings, connected to the device and flowed sequentially at 3 psi for 10 minutes. First, a 7 nM (0.45 g/ml) biotinylated mouse anti-Penta His solution in 2% BSA was flowed and immobilized in the button region coated with neutravidin. Then, 1.9 μΜ (136.8 μg/ml) of recombinant Reston GP, 1.9 μΜ (187.7 μg/ml) of recombinant Bundibugyo GP, and 1 μΜ (73 μg/ml) of recombinant Zaire GP were pumped for 10 minutes. 5 pM (75 pg/ml) of rabbit anti-Bundibugyo and 1 pM (15 pg/ml) of rabbit anti-Reston in human serum were flowed for 8 minutes, one species per chip. Next, a 100 pM (15 ng/ml) goat anti-rabbit βG antibody 2% BSA was pumped for 10 minutes. The channels were washed with 0.005% Tween 20 in PBS flowed after each reagent for 5 minutes. After the last wash step 50 μΜ FDG was flowed for 4 minutes and buttons were closed with 25 psi pressure. Bright field and fluorescent images were taken 15 minutes after the closure of the buttons with an exposure time of 2000 ms.

3 different Ebola virus GPs and a BSA negative control in the upper spotting chambers were spotted, element A, chamber i, of the unit cells. Secondary antibodies were printed in the bottom spotting chambers, element A, chamber ii of the units. PE conjugated antibodies were spotted in the top half of the chip, to allow analog detection while βG conjugated antibodies where spotted in the bottom half of the chip to allow analog-amp detection. After surface functionalization, 15 μΙ of each reagent were loaded into tygon tubing, connected to the device and flowed sequentially at 3 psi for 10 minutes. First, a 7 nM (0.45 μg/ml) biotinylated mouse anti-Penta- His solution in 2% BSA were immobilized in the button region coated with neutravidin. Then, 1.9 μΜ (136.8 μg/ml) of recombinant Reston GP, 1.9 μΜ (187.7 μg/ml) of recombinant Bundibugyo GP, and 1 μΜ (73 μg/ml) of recombinant Zaire GP were pumped for 10 minutes. To generate the standard curves, different concentrations, ranging from 0.1 nM (0.15 μg/ml) to 100 nM (150 μg/ml), of anti-Bundibugyo, anti-Reston, and anti-Zaire were spiked in mouse whole blood. A final volume of 3.5 μΙ was then flown through the device from a hematocrit tube that in turn was plugged into the device. The mouse whole blood sample from the hematocrit tube was loaded into the device by vacuum suction (8 minutes at -2.5 psi) as opposed to pressure driven flow. Next, 30 nM (1 1.7 μg/ml) goat anti-rabbit IgG PE conjugated antibody and 2 nM (0.3 μg/ml) of goat anti-rabbit βG antibody in 2% BSA were pumped for 10 minutes in the upper and lower half of the device, respectively. The channels were washed with 0.005% Tween 20 in PBS immediately after each reagent for 5 minutes. After the last washing step 50 μΜ FDG was flowed and buttons immediately closed with 25 psi pressure. Bright field and fluorescent images were taken with the USB fluorescence microscope 15 minutes after the closure of the buttons with the following exposure settings: Lowest, Luma 240. Serological detection of IgE and IgG

Three allergens, natural Ara h2, recombinant Fel d1 , recombinant Der p2, and a BSA negative control were spotted in order to perform an allergy diagnostic test. The human clinical serum sample was diluted 1 :2 (6.7 μΙ in 13.4 μΙ) with incubation buffer (0.05 M Tris/HCI pH 7.6, 0.15 M NaCI, Tween 20 0.02% v/v, BSA 1 % w/v) (49). After surface functionalization, 15 μΙ of each reagent were loaded into tygon tubing, connected to the device, and flowed sequentially at 3 psi for 10 minutes. First, 30.5 μΜ (0.55 mg/ml) of Ara h2, 41.15 μΜ (0.75 mg/ml) of Fel d1 , and 42.8 μΜ (0.6 mg/ml) of Der p2 were circulated through each unit cell by the S-shaped peristaltic pump for 20 minutes and immobilized in the button region coated via neutravidin. Then the clinical sample was flowed for 5 minutes. Next, 1 :100 rabbit anti-human IgE in 2% BSA and 30 nM (1 1.7 μg/ml) of anti-rabbit IgG PE conjugated antibody in 2% BSA were flowed on the top half of the device while 6 nM (2.4 μg/ml) of goat anti-human IgG PE conjugated antibody in 2% BSA was flowed on the lower half for 10 minutes. Channels were washed with 0.005% Tween 20 in PBS after each step for 5 minutes. After the last washing step the buttons were closed with 25 psi pressure. Bright field and fluorescent images were taken using the Nikon microscope with an exposure time of 2000 ms for IgE and 300 ms for IgG.

Image acquisition, quantification and data analysis

Pumping and mixing sequences were recorded with a Hamamatsu ORCA-Flash4.0 camera C1 1440. Mixing time was determined by analyzing fluorescent intensity changes over time in a portion of the microfluidic channel during pumping. The sandwich valve was closed and a Cy3 labeled solution was pumped in the assay chamber containing buffer. Fluorescent images were obtained Nikon ECLIPSE Ti microscope equipped with LEDFIuorescent ExcitationSystem, Cy3 and FITC filter sets, and a Hamamatsu ORCA- Flash4.0 camera, C1 1440. Images were taken with a 40x objective lens from Nikon, SPIan Fluor, ELWD 40x/0.60, °°/0.2, WD 3.6-2.8. Alternatively, fluorescent images were taken with a Cy3, AM4113T-YFBW) or FITC (AM41 13T- GFBW) Dino-Lite USB fluorescence microscopes. The resulting TIFF-images were analyzed with a microarray image analysis software from GenePix Pro v6.0, Molecular Devices, ImageJ, Fiji or Matlab from Mathworks. Averages, SDs and linear fits were calculated with Microsoft Excel. Vertical scatterplots, histograms and statistical analysis were created and performed in Prism v5.0 from Graphpad. The liquid handling sequence for the assay described above are shown in the tables below.

REFERNCES

Eigenmann, P., Kuenzli, M., D'Apuzzo, V., Kehrt, R., Joerg, W., Reinhardt, M., Rudengren, M., Borres, M.P., and Lauener, R.P. (2009). The ImmunoCAP Rapid Wheeze/Rhinitis Child test is useful in the initial allergy diagnosis of children with respiratory symptoms. Pediatr. Allergy Immunol. 20, 772-779.

Fan, R., Vermesh, O., Srivastava, A., Yen, B.K.H., Qin, L, Ahmad, H., Kwong, G.A., Liu, C.-C, Gould, J., Hood, L, et al. (2008). Integrated barcode chips for rapid, multiplexed analysis of proteins in microliter quantities of blood. Nat. Biotechnol. 26, 1373-1378.

Ghosh, K.K., Burns, L.D., Cocker, E.D., Nimmerjahn, A., Ziv, Y., Gamal, A.E., and Schnitzer, M.J. (201 1 ). Miniaturized integration of a fluorescence microscope. Nat. Methods 8, 871-878.

Hantusch, B., Scholl, I., Harwanegg, C, Krieger, S., Becker, W.-M., Spitzauer, S., Boltz- Nitulescu, G., and Jensen-Jarolim, E. (2005). Affinity determinations of purified IgE and IgG antibodies against the major pollen allergens Phi p 5a and Bet v 1 a: discrepancy between IgE and IgG binding strength. Immunol. Lett. 97, 81-89.

Hirsch, L.R., Jackson, J.B., Lee, A., Halas, N.J., and West, J.L. (2003). A whole blood immunoassay using gold nanoshells. Anal. Chem. 75, 2377-2381.

Inci, F., Filippini, C, Baday, M., Ozen, M.O., Calamak, S., Durmus, N.G., Wang, S., Hanhauser, E., Hobbs, K.S., Juillard, F., et al. (2015). Multitarget, quantitative nanoplasmonic electrical field- enhanced resonating device (NE2RD) for diagnostics. Proc. Natl. Acad. Sci. U. S. A. 112, E4354-4363.

Maerkl, S.J., and Quake, S.R. (2007). A systems approach to measuring the binding energy landscapes of transcription factors. Science 315, 233-237. Rissin, D.M., Fournier, D.R., Piech, T., Kan, C.W., Campbell, T.G., Song, L, Chang, L, Rivnak, A.J., Patel, P.P., Provuncher, G.K., et al. (2011 ). Simultaneous Detection of Single Molecules and Singulated Ensembles of Molecules Enables Immunoassays with Broad Dynamic Range. Anal. Chem. 83, 2279-2285.

Volpetti, F., Garcia-Cordero, J., and Maerkl, S.J. (2015). A Microfluidic Platform for High- Throughput Multiplexed Protein Quantitation. PLoS ONE 10.

Wang, J., Ahmad, H., Ma, C, Shi, Q., Vermesh, O., Vermesh, U., and Heath, J. (2010). A self- powered, one-step chip for rapid, quantitative and multiplexed detection of proteins from pinpricks of whole blood. Lab. Chip 10, 3157-3162.

Wang, T., Zhang, M., Dreher, D.D., and Zeng, Y. (2013). Ultrasensitive microfluidic solid-phase ELISA using an actuatable microwell-patterned PDMS chip. Lab. Chip 13, 4190^1197.