key: cord-0747420-buu5k3hm authors: Colbert, Ashlee J.; Lee, Dong Hoon; Clayton, Katherine N.; Wereley, Steven T.; Linnes, Jacqueline C.; Kinzer-Ursem, Tamara L. title: PD-LAMP smartphone detection of SARS-CoV-2 on chip date: 2022-03-09 journal: Anal Chim Acta DOI: 10.1016/j.aca.2022.339702 sha: 893ee58e33e6c04bcc038489fd0d8713aac1c5da doc_id: 747420 cord_uid: buu5k3hm In 2019 the COVID-19 pandemic, caused by SARS-CoV-2, demonstrated the urgent need for rapid, reliable, and portable diagnostics. The COVID-19 pandemic was declared in January 2020 and surges of the outbreak continue to reoccur. It is clear that early identification of infected individuals, especially asymptomatic carriers, plays a huge role in preventing the spread of the disease. The current gold standard diagnostic for SARS-CoV-2 is quantitative reverse transcription polymerase chain reaction (qRT-PCR) test based on the detection of the viral RNA. While RT-PCR is reliable and sensitive, it requires expensive centralized equipment and is time consuming (∼2 h or more); limiting its applicability in low resource areas. The FDA issued Emergency Use Authorizations (EUAs) for several COVID-19 diagnostics with an emphasis on point-of care (PoC) testing. Numerous RT-PCR and serological tests were approved for use at the point of care. Abbott's ID NOW, and Cue Health's COVID-19 test are of particular interest, which use isothermal amplification methods for rapid detection in under 20 min. We look to expand on the range of current PoC testing platforms with a new rapid and portable isothermal nucleic acid detection device. We pair reverse transcription loop mediated isothermal amplification (RT-LAMP) with a particle imaging technique, particle diffusometry (PD), to successfully detect SARS-CoV-2 in only 35 min on a portable chip with integrated heating. A smartphone device is used to image the samples containing fluorescent beads post-RT-LAMP and correlates decreased diffusivity to positive samples. We detect as little as 30 virus particles per μL from a RT-LAMP reaction in a microfluidic chip using a portable heating unit. Further, we can perform RT-LAMP from a diluted unprocessed saliva sample without RNA extraction. Additionally, we lyophilize SARS-CoV-2-specific RT-LAMP reactions that target both the N gene and the ORF1ab gene in the microfluidic chip, eliminating the need for cold storage. Our assay meets specific target product profiles outlined by the World Health Organization: it is specific to SARS-CoV-2, does not require cold storage, is compatible with digital connectivity, and has a detection limit of less than 35 × 10(4) viral particles per mL in saliva. PD-LAMP is rapid, simple, and attractive for screening and use at the point of care. and surges of the outbreak continue to reoccur. It is clear that early identification of infected 23 individuals, especially asymptomatic carriers, plays a huge role in preventing the spread of the 24 disease. The current gold standard diagnostic for SARS-CoV-2 is quantitative reverse 25 transcription polymerase chain reaction (qRT-PCR) test based on the detection of the viral RNA. 26 While RT-PCR is reliable and sensitive, it requires expensive centralized equipment and is time 27 consuming (~2 hours or more); limiting its applicability in low resource areas. The FDA issued 28 Emergency Use Authorizations (EUAs) for several COVID-19 diagnostics with an emphasis on 29 point-of care (PoC) testing. Numerous RT-PCR and serological tests were approved for use at 30 the point of care. Abbott's ID NOW, and Cue Health's COVID-19 test are of particular interest, 31 which use isothermal amplification methods for rapid detection in under 20 minutes. We look to 32 necessary for a PoC device [34, 35] . These detection limits meet the requirements outlined by the 124 WHO (10 3 to 10 1 copies µL -1 ) for testing and have the advantage of minimal sample processing 125 [36] . 126 127 128 129 145 RT-LAMP was performed using a primer set developed in-house to target the nsp3 region of 146 ORF1ab (Table S1 ) and a primer set targeting the N2 region of the N gene developed by Butler 147 et al. (Table S2 ) [37] . The reaction mixture contained the following components (also listed in 148 Cutter (Silhouette America, Lindon, UT) with two passes using a 10-blade, force of 29, and a 180 speed of 10 (arbitrary units within the Silhouette Cameo software). The top imaging layer, made 181 from 50 μm COP, was hand-cut to cover the channel cutout on the 188 μm COP layer (Fig. S1) . 182 The first two layers of 188 μm were bonded together using a Carver 4386 hydraulic press 183 (Carver Inc., Wabash, IN) with 1.2 tons of pressure at 120°C for 2 minutes. After bonding the 184 two 188 μm layers together, the remaining 50 μm COP is bonded with the same pressure and 185 temperature settings in the hydraulic press for 1 minute. videos collected using an iPhone6-driven custom microscope device and processed using an in-236 house smartphone application, fully described in previous work [30, 39] . (Table S1 ) and tested it in an in-tube RT-LAMP reaction with 251 various concentrations of SARS-CoV-2 genomic RNA (gRNA). The in-house designed primers 252 could amplify the genomic RNA in less than 30 minutes without any negative amplification 253 occurring. We ran the assay for 30 minutes five times (Fig. S2) and found a LOD of 75 genomic 254 RNA (gRNA) copies per reaction (representative results shown in Fig. 2 ). 7.5 gRNA copies per 255 reaction did amplify 2 out of the 5 times. Therefore our limit of detection was determined to be 256 75 gRNA copies per reaction as this concentration amplified for every repeat (Fig. S2, N=5) . CoV-2 samples relative to controls in which no genetic template was added (no template 267 controls, NTC) was found for each repeat (N=3) (Fig. 3 and Fig. S3 ). Each chip was analyzed for 268 30 seconds, and the resulting diffusion coefficient measurements were recorded and analyzed 269 using a one-way ANOVA with post hoc Dunnett's against the NTC. The diffusion coefficients 270 were all significantly different from the NTC (p < 0.0001 for 5x10 1 , 5x10 2 minutes of amplification on chip, it was confirmed PD-LAMP was indeed specific to SARS-CoV-289 2 and did not amplify DENV3, MERS, or SARS (Fig. 4, N=3) . After the 35 minutes of reaction, 290 samples were removed from the chip and run on gel electrophoresis to visually confirm 291 amplification (Fig. 4B) . Diffusion coefficients of SARS-CoV-2 at various concentrations were 292 significantly different from each off-target virus (DENV3, MERS, SARS), using a Tukey's 293 multiple comparison test (* p < 0.05 for 5x10 1 , **** p < 0.0001 for 5x10 2 Thus, demonstrating that PD sensitively measured SARS-CoV-2 presence. Overall, diffusion 324 coefficients presented lower values in the presence of saliva, which was expected due to its higher 325 viscosity. Additionally, one of the three MERS samples amplified in 10% saliva. The MERS 326 amplification may have been due to a pipetting error, carryover contamination, or because the 327 diluted MERS template was contaminated. Due to this negative amplification event, MERS was 328 significantly different from the other negative control SARS (* p < 0.05). PD-LAMP was still 329 found to be specific for SARS-CoV-2 viral particles in 8% and 10% saliva (Fig. 5) . Detection of 330 SARS-CoV-2 via PD-LAMP was found to have a LOD of 50 particles per 20 µL reaction in both 331 8 and 10% saliva. 332 Reactions were unable to lyophilize with the full concentration of Betaine, an anti-freezing agent. 348 Therefore, the lyophilization protocol was optimized by reducing the amount of betaine in the 349 reaction to allow for proper freezing. The master mix containing two primer sets (S1 and S2 350 Tables) and 2.3x10 9 fluorescent beads mL -1 particles in a total reaction volume of 20 µL was 351 lyophilized in the chips. After at least one day of storage at room temperature the chips were 352 rehydrated with 20 µL of a mixture containing positive or negative template, 8% (v v -1 ) saliva and 353 water. MERS and SARS were again tested to evaluate cross-reactivity of the combined SARS-354 CoV-2 primer sets and the off-target viruses at a concentration of 1.6x10 4 virus particles per 355 reaction volume. SARS-CoV-2 was tested at concentrations from 5x10 3 to 5x10 1 virus particles 356 per reaction. PD-LAMP was successful in detecting positive samples from negative and off-target 357 samples (Fig. 6) . SARS-CoV-2 amplification resulted in lower diffusion coefficients for the 358 positive samples relative to NTC and off-target samples. Each of the SARS-CoV-2 positive 359 concentrations was found to be significantly different from each of the control samples using 360 Tukey's multiple comparison test (** p < 0.01 5x10 1 virus particles per reaction, *** p < 0.001 361 5x10 2 virus particles per reaction, and **** p < 0.0001 5x10 3 virus particles per reaction) (Fig. 6) . 362 Performing PD-LAMP from resuspending lyophilized reagents yielded a LOD of 30 particles µL -363 1 and was specific to SARS-CoV-2. Tracing and rapidly diagnosing infected patients are key to controlling disease spread. The 391 COVID-19 pandemic is a prime example of the importance of affordable and rapid diagnostics 392 [34, 41] . iNAAT-based methods offer an alternative to RT-PCR tests by eliminating the need for 393 expensive temperature controlling units [17, 42] . Furthermore, the ability to perform RT-LAMP 394 with minimal sample prep in a portable platform would provide a highly sensitive PoC device to 395 aid in rapid confirmation of cases and surveillance of the virus. RT-LAMP provides an alternative 396 to current serological tests that require strenuous sample prep. 397 398 Smartphone devices have advantages such as a GPS, camera, and advanced computational 399 processing power that can aid in medical diagnosis [43, 44] . The pairing of PoC testing with 400 automated sample processing is of great importance for scalability and rapid processing [21] . We assess the cross-reactivity of the reaction against other coronaviruses as well as with DENV3 408 inactivated virus, a virus with similar symptoms as SARS-CoV-2. We found SARS-CoV-2 PD-409 LAMP to be specific even when challenged with other coronaviruses and DENV3 (Fig. 4) . Next, 410 we performed RT-LAMP in various concentrations of saliva and determined the LOD in 8% and 411 10% saliva to be 50 virus particles per reaction (Fig. 5) . Diffusion coefficients measured by PD in 412 the presence of saliva were overall lower when compared to samples without saliva. This was 413 expected due to higher viscosity of saliva. Regardless, there still may be room for further 414 optimization of the assay to ensure no non-specific amplification occurs in the presence of saliva. 415 In the MERS samples, the average diffusion coefficient-value was significantly different than the 416 positive SARS-CoV-2 amplification, and not significantly different from other negative controls 417 (Fig. 5) , however in some repeats non-specific amplification and primer dimers may have occurred 418 in some cases, as seen in the increased standard deviation of MERS sample in Fig. 5B One group explored the creation of lyophilized LAMP buttons which were stable for at least 24 437 hours at room temperature but found a degradation in the intensity of the dyes needed for their 438 paper-based method [49] . Others have lyophilized the dyes in the cap of tubes separately from the 439 remaining reagents to combat the loss in dye intensity [50] . In the current work dyes are not a 440 necessary component of the assay, making the lyophilization process less complicated, however 441 optimization of lyophilization conditions was still necessary. We found that the lyophilization 442 process would not complete in concentrations of betaine greater than 200 mM due to its extremely 443 low freezing point. The concentration was lowered to 100 mM with no reduction in assay 444 sensitivity. We performed PD-LAMP on chip after at least 24 hours of room temperature storage 445 and saw a significant difference from the negative diffusion coefficients with a LOD of 30 virus 446 particles µL -1 and the technique was specific to SARS-CoV-2 (Fig. 6) . The performance of PD-447 LAMP reagents over a range of storage conditions remains to be studied. 448 449 Furthermore to truly make the device field ready, an analysis of more samples using the current 450 setup and microfluidic chips will be needed to create proper cut-off values for positive and negative 451 samples. Previously we have determined cut-offs for PD-LAMP of 150 blinded samples of V. 452 Cholerae in pond water [30] . Coefficients less than 7.0x10 -13 m 2 /s were considered positive and 453 coefficients greater than 7.2x10 -13 m 2 /s were negative with all coefficients in between yielding 454 results which were inconclusive and would need to be retested. However, the chips used for the 455 previous study were not the same design as the microfluidic chip described here and the differences 456 in saliva vs water samples are expected to affect background solution viscosity. While all negative 457 samples did maintain values greater than 7.0x10 -13 there was a wider range for the positive 458 samples. One limitation of the current chip design, is that it does not always create an airtight seal. 459 This allows for the sample to flow which can interfere with the true Brownian motion of particles 460 and in turn yield higher diffusion coefficients even in more complex matrices with higher 461 viscosities than water. After lyophilization, the contact angle of liquid in the chip changes also 462 inducing greater susceptibility to flow. This will need to be further explored to reduce flow or a 463 change in our in-house processing application made to account for flow. This would aid in 464 improving the robustness of the assay and readiness for use at the point-of-care. We would like to recognize Melinda Lake for providing feedback throughout the manuscript 505 writing process. We would also like to acknowledge Navaporn Sritong, Holly Witten, and Riley 506 Brown for helpful insights into primer optimization. 507 508 509 512 general-s-opening-remarks-at-the-media-briefing-on-covid Presymptomatic SARS-CoV-2 Infections and Transmission in a Skilled Nursing 523 Facility Estimating the asymptomatic 526 proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess 527 cruise ship World Health Organization, The selection and use of essential in vitro diagnostics World Health Organization, COVID-19 Weekly Epidemiological Update 22, World 533 Health Organization Clinical course and 537 risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a 538 retrospective cohort study World Health Organization, Recommendations for national SARS-CoV-2 testing 541 strategies and diagnostic capacities Interim guidance A comprehensive 544 review of detection methods for SARS-CoV-2, Microorganisms Saliva-547 based PCR tests for SARS-CoV-2 detection Types of assays for 550 SARS-COV-2 testing: A review Amplification Reaction by Turbidity Derived from Magnesium Pyrophosphate Formation, 554 Biochemical and Biophysical Research Communications Loop-mediated isothermal amplification (LAMP): a rapid, accurate, 557 and cost-effective diagnostic method for infectious diseases Loop-mediated isothermal amplification of DNA Rapid and 565 extraction-free detection of SARS-CoV-2 from saliva with colorimetric LAMP., 566 MedRxiv : The Preprint Server for Health Sciences Scaling 572 diagnostics in times of COVID-19: Colorimetric Loop-mediated Isothermal 1 Amplification (LAMP) assisted by a 3D-printed incubator for cost-effective and 2 scalable 574 detection of SARS-CoV-2 3 4 RT-LAMP for rapid diagnosis of coronavirus 577 SARS-CoV-2 Reverse Transcription Loop-Mediated Isothermal Amplification Assays Targeting Severe 581 Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Loop-mediated isothermal amplification of DNA A smartphone-587 based diagnostic platform for rapid detection of Zika, chikungunya, and dengue viruses Diagnostics for SARS-CoV-2 detection: A 590 comprehensive review of the FDA-EUA COVID-19 testing landscape ID NOW COVID-19 | Abbott Point of Care Testing Real-time COVID-19 testing | Cue Consistent Detection of Clinical Infectious Diseases : An Official Publication of the 601 Infectious Diseases Society of America Temporal profiles of viral load in posterior 607 oropharyngeal saliva samples and serum antibody responses during infection by SARS-608 CoV-2: an observational cohort study, The Lancet Infectious Diseases Rapid and 612 extraction-free detection of SARS-CoV-2 from saliva with colorimetric LAMP., 613 MedRxiv : The Preprint Server for Health Sciences Saliva is 622 more sensitive for SARS-CoV-2 detection in COVID-19 patients than nasopharyngeal 623 swabs Complications of COVID-19 Nasopharyngeal Swab Test Reverse Transcription Loop-Mediated Isothermal Amplification Assays Targeting Severe 630 Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Particle Diffusometry Platform for Sub-Attomolar Detection of Vibrio cholerae in 635 Environmental Water Towards the use of a smartphone imaging-based tool for 639 point-of-care detection of asymptomatic low-density malaria parasitaemia, Malaria 640 Microfabricated flow-through device 642 for DNA amplification -Towards in situ gene analysis A heater-integrated transparent 645 microchannel chip for continuous-flow PCR Novel coronavirus disease (COVID-19): Paving 647 the road for rapid detection and point-of-care diagnostics From lab on a chip to point of care 650 devices: The role of open source microcontrollers COVID-19 Shotgun Transcriptome and Isothermal Profiling of SARS-CoV-662 Infection Reveals Unique Host Responses, Viral Diversification, and Drug Interactions Brownian motion and correlation in particle image velocimetry Particle Diffusometry: An Optical Detection Method for Vibrio cholerae Presence in 669 Environmental Water Samples COVID-19 Development and evaluation of a rapid 680 CRISPR-based diagnostic for COVID-19 Smartphone-based clinical diagnostics: towards democratization of 684 evidence-based health care Medical diagnostics with mobile 687 devices: Comparison of intrinsic and extrinsic sensing Genetics and genomics of SARS-CoV-2: A 690 review of the literature with the special focus on genetic diversity and SARS-CoV-2 691 genome detection Mediated Isothermal Amplification Reagents for the Detection of Leptospira, MILITARY 695 MEDICINE Direct Blood Dry 697 LAMP: A Rapid, Stable, and Easy Diagnostic Tool for Human African Trypanosomiasis Comprehensive evaluation of molecular enhancers of 700 the isothermal exponential amplification reaction OPEN Paper-based loop-mediated isothermal amplification and lateral flow (LAMP-LF) assay 703 for identification of tissues of cattle origin 705 E4307BF2997888215B3DDFFFD36240212A7D774CB038705EED4ACE1154ACEEE15 Advanced Lyophilised Loop Mediated Isothermal Amplification (L-LAMP) based point 709 of care technique for the detection of dengue virus Highlights • Smartphone biosensor combines isothermal amplification and particle diffusometry • Limit of detection of 50 virus particles µL -1 from 10% saliva samples on chip • SARS-CoV-2 detected using assays lyophilized in microfluidic chip • Sample amplified with a portable miniaturized heater • Potential use in widespread screening out in the field for its portability Linnes are co-founders of OmniVis Inc., a spinout company of Purdue University to translate the smartphone PD-LAMP technology