Disposable cartridge platform for rapid detection of viral hemorrhagic fever viruses

Steven M. Scherr a, David S. Freedman b, Krystle N. Agans cd, Alexandru Rosca b, Erik Carter e, Melody Kuroda f, Helen E. Fawcett a, Chad E. Mire cd, Thomas W. Geisbert cd, M. Selim Ünlü ghi and John H. Connor *eh
aDepartment of Mechanical Engineering, Boston University, Boston, MA 02215, USA
bNexgen Arrays, LLC, Boston, MA 02215, USA
cGalveston National Laboratory, University of Texas Medical Branch, Galveston, TX 77555, USA
dDepartment of Microbiology and Immunology, University of Texas Medical Branch, Galveston, TX 77555, USA
eDepartment of Microbiology and National Emerging Infectious Disease Laboratories, Boston University, Boston, MA 02118, USA. E-mail: jhconnor@bu.edu
fBD, Research Triangle Park, NC 27709, USA
gDepartment of Electrical Engineering, Boston University, Boston, MA 02215, USA
hDepartment of Biomedical Engineering, Boston University, Boston, MA 02215, USA
iPhysics Department, Boston University, Boston, MA 02215, USA

Received 13th December 2016 , Accepted 2nd February 2017

First published on 6th February 2017


Light microscopy is a straightforward and highly portable imaging approach that is used for the detection of parasites, fungi, and bacteria. The detection of individual virus particles has historically not been possible through this approach. Thus, characterization of virus particles is typically performed using high-energy approaches such as electron microscopy. These approaches require purification of virions away from its normal milieu, significant levels of expertise, and only count a small number of particles at a time. To correct these deficiencies we created a platform that allows label-free, point-of-need virus imaging and counting. We adapted a multiplex-capable, interferometric imaging technique to a closed-system that allows real-time particle detection in complex mixtures. To maximize virus particle binding we constructed a disposable device with a constant flow rate of ∼3 μl min−1. Biosafety was achieved by having a sealable sample addition port. Using this platform we were able to readily identify virus binding in a 20 minute experiment. Sensitivity was comparable to laboratory-based assays such as ELISA and plaque assay, and showed equal or better sensitivity against paper-based assays designed for point-of-need use. Our results demonstrate a platform that can be used for rapid multiplexed detection and visualization of whole virus particles. We envision this technology as a sample-to-answer platform for detection and visualization of viruses without the need for prior labeling. This would enable both research investigation of virus particle behavior and morphology and have the potential to be used in a diagnostic context, where direct imaging from samples such as blood and urine would be valuable.


Introduction

Bacteria, fungi, and parasites are all found in a size range where they interact strongly with visible light. This has made light microscopy an important tool for the study of infectious disease. It is standard equipment for microbiology labs and useful for the identification of bacteria, fungi, and parasites. In clinical diagnosis situations, these microbes often have distinguishing size, shape, and staining characteristics. In contrast, virus particles are too small to be seen using standard light microscopy. Imaging of individual virus particles is typically performed using electron microscopy or atomic force microscopy. These techniques require particle purification, significant levels of expertise, and are low throughput.1,2 Viruses can be detected in a standard light microscope only after labeling, which involves purification or genetic modification.

The lack of an easy approach for the identification and counting of virus particles has limited the ability of researchers and health care providers. Assessment of individual particle size and shape in clinical and laboratory samples and virus stocks would benefit from a light microscopy technique that allowed high throughput measurements. Additionally, quantification of vaccine preparation monodispersity and uniformity would be more readily assessed with the development of techniques that allow the detection and characterization of large numbers of individual virus particles. The ability to use light microscopy for virus detection also has implications for improved point of care diagnostic approaches. Light microscopy is globally established as an effective approach for diagnosing diseases such as malaria at the point of need. The ability to carry out viral identification on a similar platform could advance the diagnostic capability.

Previously we have shown that a light microscopy approach that uses interference reflectance (called SP-IRIS, or single particle interferometric reflectance imaging sensor) is capable of visualizing and detecting individual virions in serum.4 This approach does not rely on virus labeling, instead utilizing the mass of the virus particle itself to generate a signal. The initial promise of this approach was tempered somewhat by the fact that early iterations involved open assay systems or active pumping approaches that are not desirable when handling potentially infectious samples.

Here we describe the development of an enclosed virus detection approach that uses a disposable polymer-paperfluidic cartridge. This allows time-resolved counting and analysis in complex solutions using a benchtop-sized light-microscope. Cartridge based virus detection showed good sensitivity in a short (20 min) timeframe. These advances result in a high performance platform and a disposable test which utilizes capillary-based passive fluid handling. This eliminates the need for pressure controllers or complex incubation machinery. Additionally, the disposable cartridge contains the sample to reduce exposure risk and requires minimal sample preparation and no cold chain storage.

Results

Disposable polymer-paperfluidic test cartridge

Paper based lateral flow immunoassays are commonly employed to produce inexpensive and disposable diagnostic tests. However, these often lack sensitivity and are difficult to implement robust fluid control or multiplexed testing.5,6 The ability of paper to passively transport fluids via capillary flow is an attribute we chose to take advantage of. By incorporating paper-based fluid handling we were able to reduce the need for external equipment. To achieve a steady controlled flow rate of 1–10 μL min−1 for at least 20 minutes a single-layer absorbent pad with a 270° fan shape was placed in the channel after the sensor7 (ESI). Using this shape, shown in Fig. 1A, the stem width and length can be experimentally varied to control the flow rate. Once the fluid front reaches the circular region, the increasing perimeter of the advancing flow front increases the driving force (due to surface tension) at approximately the same rate as the fluidic resistance increases, maintaining a quasi-steady flow rate.
image file: c6lc01528j-f1.tif
Fig. 1 Absorbent pad flow rate measurement. A) Shows the shape of the absorbing pad which is able to maintain a steady flow rate. The red area indicates the perimeter of the wetted region that expands radially as the pad absorbs fluid. The blue line marks the outer perimeter of the absorbent pad. B) Shows the flow rate as measured by taking the ratio of the wetted area to the full perimeter, multiplied by the total volume absorbed by the pad.

A 100 μL sample was placed in the reservoir and the fluid was pushed into the channel by screwing a luer cap onto the reservoir, thereby displacing air. Once the fluid contacted the absorbent pad, the material continued to wick the sample into the pad at a quasi-steady rate. This continued to draw fresh sample over the sensor without the need for external fluid actuation giving a repeatable incubation of the sensor. In order to measure the flow rate, images were taken every 30 seconds. The wetted region (shown in red in Fig. 1A) was calculated as a percent of the total pad area (shown in blue in Fig. 1A). The percent of wetted area was then multiplied by the total volume absorbed during the experiment to determine the flow rate at each time interval. Analysis of the fluid front moving across the pad showed that the designed pad shape did in fact maintain a quasi-steady flow rate of ∼3 μL min−1 for approximately 20 minutes, as shown in Fig. 1. This satisfied the requirements of maintaining a flow rate in the range of 1–10 μL min−1 for approximately 20 minutes, while using a sample volume of 100 μL or less.

The final cartridge design incorporates this 2-D paper based capillary driven flow into the multilayer polymer laminate cartridge as shown in Fig. 2B. This design was produced to be compatible with a reel-to-reel lamination manufacturing process that is highly scalable and promises a low cost device ($1–10) at high volumes. The final design has a sample reservoir, 3-dimensional fluidic network exposing the surface of the sensor to the sample, an optical imaging window, and capillary driven flow which also acts as waste reservoir, containing the sample completely within the cartridge (ESI). This represents a significant improvement over the previous design shown in Fig. 2A.3 The previous version of the cartridge was connected to tubing and flow was controlled via a syringe pump. This would have been a significant burden on the user and would require additional infrastructure and cost to operate the system.


image file: c6lc01528j-f2.tif
Fig. 2 Two generations of SP-IRIS cartridges. A) Shows the previous generation cartridge which relied on a syringe pump. B) Shows an image of the disposable passive cartridge with integrated paperfluidic flow control and imaging window.

Detection sensitivity using an Ebola GP-pseudotyped model virus

Having designed a cartridge with optimized capillary driven flow it was necessary to determine the performance and sensitivity of the system. Previously we have shown that the syringe pump driven cartridge was capable of sensitive detection of individual virus particles directly from serum.3 To determine the system performance, we conducted a dilution experiment using a recombinant vesicular stomatitis virus (VSV) expressing the glycoprotein of the Zaire (Mayinga variant) strain of Ebola virus (rVSV-EBOV) spiked into 100% fetal bovine serum (FBS). Serial dilutions were prepared from the stock sample of rVSV-EBOV at 1[thin space (1/6-em)]:[thin space (1/6-em)]5, 1[thin space (1/6-em)]:[thin space (1/6-em)]50, 1[thin space (1/6-em)]:[thin space (1/6-em)]500, and 1[thin space (1/6-em)]:[thin space (1/6-em)]5000 fold dilutions in FBS. The disposable cartridges were assembled with the sensor chip containing 4 spots each of anti-Ebola antibody (13F6), anti-Marburg antibody (74-1), anti-Lassa antibody (8.9F), and a negative control antibody (8G5) for wild type vesicular stomatitis virus glycoprotein (VSV-G) which should not be present in the sample. A 100 μL sample was pipetted into the sample reservoir of the disposable cartridge. The cartridge was then activated by screwing the cap on and immediately inserted into the automated reader (Methods section). The test was allowed to run for a total of 20 minutes with each spot being scanned once every 2 minutes. The three best quality spots of the four antibody spots for each condition were scanned during the experiment as a method of quality control for the spotting process.8 This gave a total of 12 individual spots being monitored during the experiment. This format represents a multiplex test for viral hemorrhagic fever viruses.

Fig. 3A shows the results of the highest concentration sample tested. Each data point represents the average number of viruses detected on the three spots scanned of each condition. As expected in this test, the anti-Ebola antibody (orange circles) showed rapid accumulation and detection of viruses, whereas the other antibodies showed no significant capture or detection of viruses. This suggested very little cross-reactivity and good specificity even at relatively high viral titers. Fig. 3B shows just the anti-Ebola antibody for the 5 different samples tested, ranging from a 5-fold dilution of the stock sample to a 5000-fold dilution, as well as a blank sample that contained no virus. The highest concentration shown in blue circles on Fig. 3B corresponds to the orange circles on Fig. 3A. Each sample dilution showed a decreasing signal in the form of fewer viruses enumerated during the test. All four concentrations tested in this experiment showed a positive detection signal, with the lowest concentration being just above the limit of quantitation, defined here as the mean of the blank sample plus three times the standard deviation of the blank sample.9


image file: c6lc01528j-f3.tif
Fig. 3 Sensitivity experiment. A) Shows the number of viruses detected on each antibody at the highest concentration tested, a 5-fold dilution of the stock virus sample in 100% FBS. B) Shows the number of viruses detected on just the anti-Ebola antibody for all 4 dilutions tested as well as a blank sample containing no virus.

Comparison of detection techniques

In order to further validate the test, the samples were also tested using ELISA, plaque assay, and PCR. Fig. 4A shows the total number of viruses detected after 20 minutes using the disposable cartridge for all four concentrations tested as well as the blank sample. The solid circle data points indicate the number of viruses detected on the anti-Ebola antibody, and the square data points represent the number of viruses detected on the anti-Lassa antibody. This further demonstrates the specificity of this test in a multiplexed setting. All red data points were considered positive detection for the presence of the pseudotyped virus. The dashed line indicates the limit of quantitation, which is less than the number of viruses detected at the lowest concentration, yet greater than those detected for the anti-Lassa antibody. Table 1 shows the average number of viruses detected on both the anti-Ebola and anti-Lassa antibody for each sample tested. The anti-Lassa antibody shows between zero and four viruses detected at various concentrations, which correlated well with the blank sample of the anti-Ebola antibody that shows a false detection of three viruses. False detection events can be caused by surface morphology or non-specific binding of viruses or other biological particles to the sensor.10
image file: c6lc01528j-f4.tif
Fig. 4 Comparison of techniques. A) Shows the SP-IRIS disposable cartridge results with positive detection of virus in red and negative detection in blue. The dotted line indicated the limit of detection. B) Shows the results obtained from ELISA. C) Shows the plaque assay results D) shows the q-PCR results.
Table 1 Comparison of techniques. Table 1 shows the results of the different techniques for virus detection shown in Fig. 4. Numbers in red represent a positive test results and numbers in blue represent a negative test result. All tests except for the blank sample should be positive
image file: c6lc01528j-u1.tif


Fig. 4B shows the results of the in-house ELISA using the same anti-Ebola antibody as the disposable cartridge. This test used the same samples as the disposable cartridge test with additional dilutions created from each sample. The red data points represent the positive detection of the presence of the virus and the blue data points would be considered negative for the presence of virus. The dotted line again represents the limit of quantitation. The ELISA test showed positive detection on only the two highest concentrations tested, as shown in Table 1. The two lower concentrations were not significantly greater than the limit of quantitation shown and were considered to be negative. Fig. 4C contains the quantification of the of the plaque assay run on the four samples containing virus. These results show a linear trend ranging from approximately 107–104 PFU mL−1 for the samples tested with values shown in Table 1. Finally the last plot in Fig. 4D shows the results of quantitative PCR run on these samples. PCR again shows a fairly linear trend with the number of cycles increasing for each dilution with values shown in the last column of Table 1.

Values in Table 1 shown in red were considered to be positive detection of virus, whereas values in blue were considered negative results. The disposable cartridge, plaque assay, and PCR all showed correct identification of samples containing virus and the blank sample. However, the ELISA showed a lower detection sensitivity producing false negative results for the two lowest concentration samples tested. Additionally the disposable cartridge showed no false positives with the additional antibodies supporting multiplex capability.

Comparison to rapid diagnostic test

The only blood based rapid diagnostic test currently available for Ebola is the Corgenix ReEbov rapid antigen test, which is a lateral flow assay. This singleplex test for Ebola detects the presence of the VP40 matrix protein and therefore requires a sample additive to rupture the viral capsid prior to testing. The SP-IRIS test captures intact virions through specific interactions with the surface glycoprotein (GP) on the viral envelope. Both the VP40 protein and GP confer the presence of Ebola specifically; however, VP40 is the most abundant of the viral proteins in Ebola Zaire strain.11 In order to compare the SP-IRIS disposable cartridge to the ReEbov rapid antigen test, we used gamma irradiated Ebola (Zaire strain-variant Mayinga) in cell culture media supplied by Rocky Mountain Laboratories. This sample was then diluted from the stock concentration in 100% FBS and tested at four different concentrations ranging from 10-fold to 10[thin space (1/6-em)]000-fold dilution. Both the SP-IRIS disposable cartridge and the ReEbov rapid antigen test were allowed to run for 20 minutes and were tested on the same sample. Fig. 5A shows the number of viruses counted on both the anti-Ebola antibody as well as the anti-Lassa antibody in the SP-IRIS cartridge. The red bars indicated positive detection of virions and correspond to the anti-Ebola antibody spots. The blue bars indicate negative for the presence of virions and correspond to the anti-Lassa antibody spots. The dotted line indicates the cut off for a positive signal. Fig. 5B shows the results of the ReEbov rapid antigen test. The top red line is the control line and a positive test results in a secondary red line appearing below. The two highest concentration samples tested showed a clear positive signal. The third dilution tested was inconclusive, and the most dilute sample was negative. The inconclusive result was caused by difficulty visualizing the red test line. Both positive and negative control tests were run as well (data not shown). The first two columns in the table in Fig. 5 show the number of viruses detected on the anti-Ebola antibody and the anti-Lassa antibody respectively. The SP-IRIS disposable cartridge showed a positive test results (shown in red) for all four concentrations tested, whereas the anti-Lassa antibody showed negative results indicating good specificity. The right most column in the table in Fig. 5 show the results portrayed of the ReEbov test. A red plus sign indicates yes, and a blue minus sign indicates no, with a question mark used for inconclusive results. These results suggest the SP-IRIS disposable cartridge is one to two orders of magnitude more sensitive than the ReEbov test strip under these test conditions.
image file: c6lc01528j-f5.tif
Fig. 5 Comparison to Corgenix ReEbov antigen rapid test with samples of gamma irradiated Ebola virus spiked into FBS. A) Shows the results of the SP-IRIS disposable cartridge with the number of viruses bound to the anti-Ebola antibody (red) and anti-Lassa antibodies (blue). B) Shows the results of the ReEbov rapid antigen test using the same samples tested in the SP-IRIS cartridge. The results of the SP-IRIS disposable cartridge as well as the Corgenix rapid antigen test are shown in table form in which red represents a positive test results and blue represents a negative test result.

Bona fide Ebola virus testing

To further validate the functionality of the platform beyond the model virus, testing was conducted using Ebola Virus, Zaire strain-variant Mayinga. This was performed at the biosafety level 4 facility of University of Texas Medical Branch in Galveston, Texas. To carry out these tests, stock virus solution in cell media was diluted using 100% FBS. Final dilutions were tested at expected plaque assay values ranging from 1 × 104–1 × 107 PFU mL−1. The same protocol was used from the previous experiments, using 100 μL of sample added to the cartridge reservoir and a 20 minute experimental time. Fig. 6A shows the output of the analysis with the highest concentration tested at 1 × 107 PFU mL−1. Data was acquired approximately every 2 minutes for 3 spots each of multiple different antibodies. Results of anti-Ebola and anti-Lassa antibodies are shown in Fig. 6A. The anti-Ebola antibody shows rapid accumulation of hundreds of viruses, while the anti-Lassa antibody shows no apparent virus binding. This shows that Ebola virus particles are specifically captured by the corresponding antibody. Fig. 6B shows the results of the dilution experiment using concentrations of 1 × 104–1 × 107 PFU mL−1 as well as a blank sample. The lowest concentration that showed clearly detectable virus binding greater than the blank sample in this experiment was 1 × 105 PFU mL−1 in just under 20 minutes. This represents a 10 fold improvement in sensitivity compared to the ReEbov antigen rapid test and the Oraquick Ebola rapid antigen test, which reported an analytical sensitivity of 1 × 106 PFU mL−1 and 1.64 × 106 PFU mL−1 respectively.12,13 This shows the model Ebola virus (rVSV-EBOV) used works as a test platform for assay development, and, more significantly the cartridge is capable of rapid detection of bona fide Ebola virus in a complex solution.
image file: c6lc01528j-f6.tif
Fig. 6 Sensitivity of bona fide Ebola. A) Shows the results from the highest concentration tested, at a viral titer of 1 × 107 PFU mL−1. The number of viruses counted on the anti-Ebola antibody and the anti-Lassa antibody are plotted. B) Shows the results of the dilution experiment ranging from 1 × 107 PFU mL−1 to 1 × 104 PFU ml−1 as well as a blank sample.

An additional advantage of SP-IRIS is that virus morphology can also be identified. Fig. 7 highlights the increased information provided. Fig. 7A shows one EBOV capture spot. The perimeter of the spot is denoted with a red circle. The spot shows hundreds of virus particles of various shapes and sizes in a single image. Fig. 7B shows an inset of the antibody spot highlighting the different populations of virus particles captured. This image dramatically increases confidence that EBOV is being captured and counted. There are many filamentous particles clearly visible. These particles vary in length (consistent with the potential for EBOV to be polyploid) and range from small spherical particles to filaments of several microns.14 Moreover, filamentous virus particles captured on the surface tend to align with the direction of flow, making their size and shape easier to determine. Fig. 7C shows a close up of a 5 μm filamentous particle exhibiting the stereotypical EBOV morphology. Fig. 7D shows an example of a 1 μm filamentous Ebola particle that appears to be a more abundant among the population shown. Fig. 7E shows a close up of what appears to be a spherical particle. The additional information gleaned from the images of the virus increase confidence that the particles captured are in fact Ebola virions. Thus, the enumeration of virus particles and characterization of morphology and polydispersity of a large number of particles is provided in a single approach.


image file: c6lc01528j-f7.tif
Fig. 7 SP-IRIS images of captured EBOV. A) Shows viruses immobilized following a 20 minute assay. The red circle drawn outlines the perimeter of the antibody spot, particles are visible in false-color based on normalized pixel intensity. B) Shows a zoomed area of interest. This image shows the polydispersity of the population of viruses captured from the sample, and shows clear evidence of filamentous particles. C) Shows a close up of a 5 μm long filamentous virus particle exhibiting the quintessential shepherd's crook shape of an Ebola virion. D) Shows a shorter 1 μm long filamentous particle. E) An example of the smaller spherical particles also seen.

Discussion

The real-time optical detection and spatial localization of individual viruses and nanoparticles has many applications beyond that of diagnostics. Real-time, sensitive, specific, label-free detection of pathogens and nanoparticles can be used for bio-aerosol detection, environmental contamination monitoring, and basic virology research.15,16 Rapid and accurate quantification of multiple viruses is not only important for clinical diagnosis but for virology research as well. The ability to simultaneously identify and distinguish between multiple viruses in a single sample is necessary to treat patients appropriately. For example, many diseases present with similar symptoms, and the ability to differentiate between pathogens is necessary to ensure that proper care is given.

We have demonstrated a platform capable of sensitive and specific label-free detection of viruses using a disposable cartridge and reader. To exhibit the capability of this technique, samples were tested and compared to results from ELISA, PCR, plaque assay, and the ReEbov rapid antigen test. The SP-IRIS disposable cartridge platform shows promising performance compared to standard ELISA, PCR, and plaque assay using an Ebola glycoprotein-pseudotyped VSV model virus (rVSV-EBOV), outperforming the ELISA test in a shorter time. The SP-IRIS disposable cartridge was also compared to the results of the ReEbov rapid antigen test using gamma irradiated Ebola virus in 100% FBS. These results showed superior detection limit performance, with a detection limit 10–100 times more sensitive than the rapid test, while requiring no sample preparation or cold chain storage. The SP-IRIS disposable cartridge was shown to be capable of rapidly detecting bona fide Ebola virions in complex solution, supporting that the model virus (rVSV-EBOV) was an appropriate test platform for assay development. Furthermore, individual particle size and shape can be assessed on a large number of viruses providing an additional dimension of information that is not provided by any of the standard laboratory or rapid tests. The SP-IRIS disposable cartridge represents a new platform for rapid, semi-quantitative detection and characterization of viruses with promising potential for rapid diagnostics as well as virology research.

Rapid diagnosis and intervention at the point-of-care is an invaluable tool in the detection and containment of disease. However, timely test results continue to be one of the main obstacles in the treatment of patients. This platform allows semi-quantitative results in a time comparable to lateral flow assays, as opposed to the traditional yes/no answer obtained from other rapid tests. This system exemplifies the compromise of a disposable cartridge with a reader that allows high performance without the infrastructure or user expertise required for a central laboratory test. Additionally, since the sample and waste are contained within the cartridge the reader does not need to be cleaned and the chance of accidental exposure is reduced. The disposable polymer cartridge protects the sensor during shipment and the microarray stabilizer allows the test to be stored at room temperature removing the need for a cold chain. The SP-IRIS disposable cartridge promises to be a high-performance, inexpensive, and easy-to-use rapid test platform.

Methods

NVDX10 reader

The NVDX-10 reader, is a prototype reader designed and built by Nexgen Arrays, LLC specifically for automating data acquisition in a BSL-4 environment. The reader has dimensions of 16′′ × 13′′ × 7′′ (40.6 × 33 × 18 cm) and weighs approximately 30 pounds. It requires 24 VDC power (average 12 Watts) and does not require regular maintenance. The reader illuminates the sensor chip with 535 nm wavelength light from an LED light source. The imaging system uses a 40 × 0.75 N.A objective and a CMOS camera to record the images to the computer. Custom acquisition software determines the antibody spot locations on the chip and sequentially scans each spot. The microarrays used were scanned every two minutes and then repeatedly scanned over the length of each experiment for 20 minutes. SP-IRIS images are analyzed by custom software written by Nexgen Arrays, LLC. Diffraction limited particles are automatically detected in the image and the contrast is then calculated based on peak intensity of the particle relative to the local particle-free background.

Sensor preparation

The sensors substrates were purchased from Silicon Valley Microelectronics Inc. The sensors consist of polished silicon with a thin film of thermally grown oxide on top. The oxide thickness was precisely controlled to optimize detection of nanoscale particles in a liquid environment using a given wavelength of light. The surface of the sensor was then coated in an anti-fouling polymer containing NHS groups which facilitate immobilization of antibodies on the surface. The anti-Ebola antibody (13F6) was provided by Larry Zeitlin of Mapp Biopharmaceutical, San Diego. The mouse monoclonal anti-Marburg antibody (74-1) was provided by Professor Ayato Takada from Hokkaido University and the human monoclonal anti-Lassa antibody (8.9F) was provided by Professor James Robinson of Tulane University. The antibody microarray was printed on the sensor using a Scienion S3 SciFlex Arrayer. The sensors then remained at room temperature for 12–16 hours to allow immobilization on to the sensor surface. Following immobilization the sensors were then washed with 1× PBST (0.1% Tween-20) followed by filtered de-ionized water and dried with compressed nitrogen. The sensors were stored at 4 °C until they were assembled into the cartridge.

Disposable cartridge preparation and use

The SP-IRIS disposable cartridge was fabricated using layers of polymer, pressure sensor adhesive (PSA), and absorbent pad prior to experimentation. To install the sensor in the cartridge a release liner was removed from a silicone PSA layer in the chip cut out and the sensor is firmly adhered to the PSA to make a liquid tight seal. StabilGuard Choice Microarray Stabilizer (Surmodics, Inc.) was flushed through the channel to coat the channel and sensors surface. Excess stabilizer is removed and the cartridge was dried under vacuum. To run the cartridge, 100 μL of sample was pipetted into the luer reservoir. A vented luer cap with an adhesive sealing tab was then screwed onto the luer reservoir to displace the sample through channels, contacting the absorbent pad. The tab on the luer cap is then removed, venting the cartridge. The cartridge is then placed in the NVDX10 reader for data acquisition. The absorbent pad continues to wick the sample through the cartridge and over the sensor for the remainder of the experiment.

Virus creation, preparation, and use

Recombinant vesicular stomatitis viruses (VSV) expressing the native Ebola glycoprotein were created by inserting the protein cDNA into an independent transcription start/stop sequence between the M and L genes in the VSV genome where the VSV glycoprotein sequence had been previously removed. All virus stocks were in cell media solutions and stored at −80 °C prior to use. Samples were diluted in 100% FBS for all experiments. Experiments were conducted using a single stock solution, serially diluted and aliquoted for different measurements. Ebola virus samples in cell media were inactivated using gamma irradiation and provided by Hideki Ebihara at Rocky Mountain Laboratories. Bona fide Zaire ebolavirus (EBOV) used in this study was the Mayinga variant from isolate Ebola virus H.sapiens-tc/COD/1976/Yambuku-Mayinga (passage 3; Vero 76 (ATCC, CRL-1587)).

Plaque assay

Plaque assays were performed by seeding 6-well tissue culture plates with Vero E6 cells to be 90% confluent on day of assay. Ten-fold dilutions of each virus sample were prepared in serum-free DMEM and kept on ice. Vero cells were washed three times with PBS and then inoculated with 100 μL of each stepwise dilution and incubated at 37 °C for 1.5 hours. Plates were rocked back and forth every 15 minutes to ensure even dispersal of virus over cells. Following incubation, a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 mixture of 2× DMEM at 37 °C and 2% agar in PBS at 55 °C was prepared and wells were overlaid with 4 mL of the mixture. Plates were incubated at 37 °C with 5% CO2 for ∼30 hours, then fixed with 4% formaldehyde and stained with 1% crystal violet.

ELISA

Wells of a 96-well microtiter plate were coated with 13F6 anti-Ebola GP antibody to a concentration of 10 μg mL−1 in carbonate/bicarbonate buffer. Wells were blocked with 1% bovine serum albumin (BSA) in PBS for one hour, then serial dilutions of rVSV-EBOV in FCS were added to wells in triplicate for one hour. Wells were washed three times with PBST and blotted on a paper towel to remove excess buffer. A horseradish peroxidase-conjugated anti-Ebov GP secondary antibody was added at a concentration of 50 ng mL−1 in 1% BSA in PBS for thirty minutes, followed by three washes in PBST. For this assay, a QuantaBlu Fluorogenic Peroxidase Substrate Kit (Thermo Fisher, 15169) was used. After blotting plate on paper towels to remove excess wash buffer, 100 μL of QuantaBlu WS was added to each well and incubated at room temperature. After 60 minutes, 100 μL of QuantaBlu stop solution was added to wells to stop the peroxidase activity. The relative fluorescence units (RFU) for each well were then read using a Tecan plate reader at an excitation of 325 nm and an emission maximum of 420 nm.

qPCR

Serial dilutions of rVSV-EBOV were processed to isolate viral RNA using a Zymo Viral RNA Extraction Kit (Zymo Research, R1035). RT-PCR was performed on the samples using a SuperScript III One-Step RT-PCR Platinum Taq HiFi Kit (Invitrogen, 12574-035) and the reaction was carried out and analyzed using a Biorad CFX96 Real-Time C1000 Thermo Cycler (Biorad, 1855096).

Acknowledgements

We thank Professor Bruce K. Gale and Christopher Lambert of the State of Utah Center of Excellence for Biomedical Microfluidics for their help with this work. This work was funded in part by the National Institutes of Health (R01AI1096159).

Notes and references

  1. C. S. Goldsmith and S. E. Miller, Clin. Microbiol. Rev., 2009, 22, 552–563 CrossRef PubMed.
  2. Y. G. Kuznetsov and A. McPherson, Microbiol. Mol. Biol. Rev., 2011, 75, 268–285 CrossRef CAS PubMed.
  3. S. M. Scherr, G. G. Daaboul, J. Trueb, D. Sevenler, H. Fawcett, J. H. Connor and M. S. Ünlu, ACS Nano, 2016, 10, 2827–2833 CrossRef CAS PubMed.
  4. G. G. Daaboul, C. A. Lopez, J. Chinnala, B. B. Goldberg, J. H. Connor and M. Selim Ünlü, ACS Nano, 2014, 8, 6047–6055 CrossRef CAS PubMed.
  5. S. Byrnes, G. Thiessen and E. Fu, Bioanalysis, 2013, 5, 2821–2836 CrossRef CAS PubMed.
  6. E. M. Fenton, M. R. Mascarenas, G. P. Lo and S. S. Sibbett, ACS Appl. Mater. Interfaces, 2009, 1, 124–129 CAS.
  7. S. Mendez, E. M. Fenton, G. R. Gallegos, D. N. Petsev, S. S. Sibbett, H. A. Stone, Y. Zhang and G. P. López, Langmuir, 2010, 26, 1380–1385 CrossRef CAS PubMed.
  8. E. Seymour, G. G. Daaboul, X. Zhang, S. M. Scherr, N. L. Ünlu, J. H. Connor and M. S. Ünlu, Anal. Chem., 2015, 87, 10505–10512 CrossRef CAS PubMed.
  9. D. A. Armbruster and T. Pry, Clin. Biochem. Rev., 2008, 29, 49–52 Search PubMed.
  10. G. G. Daaboul, A. Yurt, X. Zhang, G. M. Hwang, B. B. Goldberg and M. S. Ünlü, Nano Lett., 2010, 10, 4727–4731 CrossRef CAS PubMed.
  11. L. H. Elliot, M. P. Kiley and J. B. McCormick, Virology, 1985, 147, 169–176 CrossRef.
  12. R. W. Cross, M. L. Boisen, M. M. Millett, D. S. Nelson, D. Oottamasathien, J. N. Hartnett, A. B. Jones, J. S. Schieffelin, B. Kargbo, M. Gbetuwa, S. M. Gevao, R. B. Wilson, E. O. Saphire and K. R. Pitts, J. Infect. Dis., 2016, 214, 210–217 CrossRef PubMed.
  13. Orasure Technologies, Inc., OraQuick Ebola Rapid Antigen Test, 2016, pp. 1–10, http://www.orasure.com/docs/pdfs/products/ebola/Ebola_Oral_Fluid_PI-ENG.pdf Search PubMed.
  14. S. Welsch, L. Kolesnikova, V. Krähling, J. D. Riches, S. Becker and J. A. G. Briggs, PLoS pathog., 2010, 6, e1000875 Search PubMed.
  15. R. S. Salter, G. W. Durbin, E. Conklin, J. Rosen and J. Clancy, Appl. Environ. Microbiol., 2010, 76, 7803–7810 CrossRef CAS PubMed.
  16. T. M. Santiago-Rodríguez, C. Dávila, J. González, N. Bonilla, P. Marcos, M. Urdaneta, M. Cadete, S. Monteiro, R. Santos, J. S. Domingo and G. A. Toranzos, Water Res., 2010, 44, 4716–4725 CrossRef PubMed.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c6lc01528j

This journal is © The Royal Society of Chemistry 2017