key: cord-0926783-avmmbshl
authors: Damin, Francesco; Galbiati, Silvia; Gagliardi, Stella; Cereda, Cristina; Dragoni, Francesca; Fenizia, Claudio; Savasi, Valeria; Sola, Laura; Chiari, Marcella
title: CovidArray: A Microarray-Based Assay with High Sensitivity for the Detection of Sars-Cov-2 in Nasopharyngeal Swabs
date: 2021-04-03
journal: Sensors (Basel)
DOI: 10.3390/s21072490
sha: 2ebeac7769a475e1fb940bede5853a5d2be7fb19
doc_id: 926783
cord_uid: avmmbshl

A new coronavirus (SARS-CoV-2) caused the current coronavirus disease (Covid-19) epidemic. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) is used as the gold standard for clinical detection of SARS-CoV-2. Under ideal conditions, RT-qPCR Covid-19 assays have analytical sensitivity and specificity greater than 95%. However, when the sample panel is enlarged including asymptomatic individuals, the sensitivity decreases and false negatives are reported. Moreover, RT-qPCR requires up to 3–6 h with most of the time involved in RNA extraction from swab samples. We introduce CovidArray, a microarray-based assay, to detect SARS-CoV-2 markers N1 and N2 in the nasopharyngeal swabs. The method is based on solid-phase hybridization of fluorescently-labeled amplicons upon RNA extraction and reverse transcription. This approach combines the physical-optical properties of the silicon substrate with the surface chemistry used to coat the substrate to obtain a diagnostic tool of great sensitivity. Furthermore, we used an innovative approach, RNAGEM, to extract and purify viral RNA in less than 15 min. We correctly assigned 12 nasopharyngeal swabs, previously analyzed by RT-qPCR. Thanks to the CovidArray sensitivity we were able to identify a false-negative sample. CovidArray is the first DNA microarray-based assay to detect viral genes in the swabs. Its high sensitivity and the innovative viral RNA extraction by RNAGEM allows the reduction of both the amount of false-negative results and the total analysis time to about 2 h.

In December 2019, an unexplained acute respiratory disease, Covid-19, was first reported in Wuhan, China [1] . It was immediately determined that the disease's cause was a novel coronavirus named SARS-CoV-2 [2] . Since virus identification and sequencing in early January 2020 [3] , the primary approach for detecting viral RNA in respiratory specimens was the reverse transcription-quantitative polymerase chain reaction (RT-qPCR) [4, 5] . The RT-qPCR, initially used to confirm symptomatic patients' diagnosis, was increasingly used to screen asymptomatic contacts and subjects at risk. Several RT-qPCR assays have been developed and recommended by the World Health Organization, the United States RPP30, but, compared to the commercial RT-qPCR kits, our system has the additional potential to distinguish the SARS-CoV-2 variants or to differentiate it from other viral and bacterial respiratory tract infections simply by adding new primers and capture probes to the same array.

Another key improvement introduced by CovidArray is the faster analysis time. An important feature of a diagnostic assay during an outbreak is the overall execution time since a fast method would allow expanding the analytical throughput. The introduction of rapid antigen detection (RAD) tests based on immunoassays for qualitative determination of SARS-CoV-2 antigens allows one to obtain a result in about 15-30 min. However, due to their low sensitivity especially for asymptomatic subjects [25] , the actual negativity of the sample must in any case be confirmed by the nucleic acid test. Moreover, when they were compared with RT-qPCR, RADs were identified as positive samples containing large amounts of virus but even the most sensitive RAD failed in detecting viral antigens in samples containing small amounts of virus [26] . Thus, rapid tests have the potential to allow earlier detection of those samples testing positive, but the evidence actually is not solid enough to determine how useful they are in clinical practice [27] . The standard methodology for SARS-CoV-2 detection requires from 3 to 6 h to run a test [28] with most of the time involved in RNA extraction from swab samples, with CovidArray the time required is reduced to about 2 h. Many commercial RNA extraction kits such as, for example, the Roche MagNA Pure 96 or the QIAGEN QIAcube kits have been validated for viral RNA extraction purpose [24] . Although the RNA isolation kits are easy to use in automated instruments, it might be necessary to use alternative approaches to extraction to expand the analytical capability in the case of an epidemic. In an attempt to overcome the limits of the standard solid-phase extraction methods many other techniques based, for example, on electrokinetic concentration or isotachophoresis were tried over the years [29, 30] , but these methods are not still utilized on a large scale. In this work, we used an innovative single-tube approach to extract the viral RNA, by employing RNAGEM, a straightforward temperature-driven enzymatic method to extract RNA, commercially available by MicroGEM (MicroGEM UK Ltd., Southampton, UK). The main advantages of using this extraction methodology are: (i) minimal pipetting steps (manual or automated) leading to less contamination, virtually no loss of RNA and reduced amount of plastic consumables (pipette tips, tubes, etc. also in shortage during this pandemic), (ii) no need of using a harsh chemical which eliminates the washing steps, (iii) no need for further purification of the RNA for accurate RT-qPCR and qPCR analysis and (iv) single-tube workflow that provides purified RNA in 15 min, (v) extraction is conducted using a common laboratory thermocycler allowing to extract up to 96 samples simultaneously, (vi) the manual steps can be automated easily by using any liquid handler.

Moreover, here we demonstrated that combination of single-tube extraction by RNAGEM with highly sensitive multiplex microarray substrate with optimal properties allows to reduce the number of PCR cycles from 40 to 25 and to lead to an overall increase in accuracy and a reduction in analysis time.

Copoly(DMA-NAS-MAPS) (MCP-4) was obtained by Lucidant Polymers Inc., Sunnyvale, CA, USA. Ammonium sulfate ((NH4)2SO4), ethanolamine and 20× standard saline sodium citrate (SSC) solution (3 M sodium chloride, 0.3 M sodium citrate, pH 7.0), sodium dodecyl sulfate (SDS), were purchased from Sigma Aldrich (St. Louis, MO, USA). All the oligonucleotides were synthesized by Metabion International AG (Steinkirchen, Germany). Their sequences are reported in Supplemental Table S1 . Untreated silicon/silicon oxide chips with 100 nm thermal grown oxide (15 × 15 mm 2 ) were supplied by SVM, Silicon Valley Microelectronics Inc. (Santa Clara, CA, USA). Chips were pretreated using a HARRICK Plasma Cleaner, PDC-002 (Ithaca, NY, USA) connected to an oxygen line.

Spotting is performed using a SciFLEXARRAYER S12 (Scienion, Berlin, Germany). InnoScan 710 (Innopsys, Carbonne, France) was used to scan the hybridized chips. Data intensities were extracted with the Mapix software and the data analysis was performed for each experiment.

Five nasopharyngeal swabs have been collected at the Unit of Obstetrics and Gynecology, L. Sacco COVID19-hub Hospital, ASST Fatebenefratelli Sacco, and seven nasopharyngeal swabs have been collected at the IRCCS Mondino Foundation (Pavia). This study was developed on existing samples collected during standard diagnostic tests that were positive to SARS-CoV-2 RNA detection. Subjects participating in the study gave their informed consent (oral or written) for SARS-CoV-2 analysis. The protocol was approved by the local Medical Ethical and Institutional Review Board (Milan, area 1, #154082020). We obtained informed consent from the patients, according to CARE guidelines and in compliance with the Declaration of Helsinki principles.

Maxwell ® RSC Viral Total Nucleic Acid Purification Kit was used to extract RNA from 250 µL of the five nasopharyngeal swabs preservation media (from L. Sacco COVID19-hub Hospital) employing the Maxwell ® RSC Instrument (Promega, Fitchburg, WI, USA) while RNAs from 350 µL of the seven nasopharyngeal swabs UTM (IRCCS Mondino Foundation) have been isolated by Magnetic bead method using an automatic nucleic acid purification system (GenePure Pro BIOER) for qPCR testing.

Dual-labeled TaqMan probes with 5 -6-FAM fluorescent dye and 3 -BHQ-1 quencher for SARS-CoV-2 target sequences N1 was used for the detection of viral RNA. For internal reference control, a pair of primers and TaqMan probe for human Ribonuclease P gene (RPP30), labeled with 5 -HEX fluorescent dye and 3 -BHQ-1 quenchers were used. Primer For reaction mix, 3 µL of extracted RNA and 7 µL of Go-Script One-Step PCR mix (Promega, Madison, WI, USA) have been used for qPCR in CFX96 (BioRad, Richmond, CA) at Sacco Hospital, while 5 µL of extracted RNA, 5 µL of Reliance One-Step RT-qPCR Supermix (BioRad, Richmond, CA), 1 µL of RT enzyme and 4 µL of water have been used at Mondino Foundation, depending on the protocol used. Cycling conditions were 50 • C for 10 min, 95 • C for 3 min, followed by 40 cycles of amplification (95 • C for 10 s and 60 • C for 30 s). qPCR analysis has been considered valid in all samples in which RPP30 gene has been detected. Positive samples were determinate by Cycle threshold (Ct) of N1 and N2 gene minor of 40.

The viral RNA was extracted by mixing 89.5 µL of the inactivated universal transport medium (UTM) (70 • C for 1 h) containing a nasopharyngeal swab with 0.5 µL of RNAGEM enzyme (MicroGEM UK Ltd., Southampton) and 10 µL of 10× Blue Buffer. Subsequently, the extraction was conducted by incubating the reaction mix in a thermocycler at 75 • C for 10 min and 95 • C for 5 min. 

The N1, N2, and RPP30 sequences were amplified using the 5 -biotin forward and 5 -Cy3 labeled reverse primers reported in Table S1 .

The PCRs were performed in 20 µL of reactions containing 9 µL of cDNA previously diluted 1:20, 200 µM deoxynucleotide triphosphates, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 1 U of DNA polymerase (FastStart Taq, Roche) and 10 pmoles of each primer.

Cycling conditions were as follows: 95 • C for 5 min; 25 cycles at 95 • C for 30 s, 60 • C for 30 s, 72 • C for 30 s and finally 72 • C for 10 min.

In addition to the single amplification, we optimized a triplex PCR amplification in which the primers used for the N1, N2, and RPP30 amplification were mixed in the same PCR mixture.

The triplex PCR was performed in 20 µL of reactions containing 9 µL of diluted cDNA, 10 pmoles of each primer, and 4 µL of 5× HOT FIREPol Blend Master Mix Ready to Load (Solis BioDyne).

Cycling conditions were as follows: 95 • C for 12 min; 25 cycles at 95 • C for 20 s, 60 • C for 30 s, 72 • C for 30 s, and finally 72 • C for 7 min.

We selected three oligonucleotide sequences from the US CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-qPCR Diagnostic Panel corresponding to the two specific probes for the regions N1 and N2 of the virus nucleocapsid gene and to the human RPP30 gene. In addition, an oligonucleotide probe, not correlated to any viral or human sequences, was selected as negative hybridization control. Capture and control probes (reported in Table S1 ), amino-modified at the 5 end, were dissolved in the printing buffer (150 mM sodium phosphate pH 8.5, 0.01% Sucrose monolaurate) to a concentration of 10 µM and printed by a piezoelectric spotter, SciFLEX ARRAYER S12 (Scienion, Berlin, Germany) on silicon/silicon oxide slides coated with MCP-4 according to the protocol provided by the manufacturer. The spotting was performed at 20 • C in an atmosphere of 60% humidity.

After the spotting step the chips were incubated overnight, and all residual reactive groups of the coating polymer were blocked as previously described [31] .

The products of the PCR reactions were heated at 95 • C for 5 min to denature the DNA double-strand. The solution was quickly centrifuged and chilled on ice for 1 min then it was spread onto the microarray. A cover slip (large enough to cover the entire spotted surface) was carefully placed on the microarray to avoid any bubble capture. The slides were incubated in a sealed humid hybridization chamber at room temperature for 15 min. The hybridized silicon chips were then removed from the hybridization chamber and soaked briefly in 4× SSC buffer to remove the cover slips. Finally, the chips were washed at room temperature with 0.2× SSC for 1 min and 0.1× SSC for 1 min and then dried with a nitrogen flow.

The hybridized silicon chips were scanned with InnoScan 710 (Innopsys, Carbonne, France). A green laser (λ ex 532 nm) for the Cy3 dye was applied. The photomultiplier (PMT) tube gain and the laser power changed between different experiments. 16-bit TIFF images were analyzed at 5 µm resolution. Data intensities were extracted with the Mapix software and the data analysis was performed for each experiment.

We employed the QX100™ Droplet Digital™ PCR System (Bio-Rad Laboratories, Hercules, CA, USA). 9 µL of cDNA previously diluted 1:20 were mixed with primers and fluorophore-labeled commercial probes (2019-nCoV CDC EUA Authorized qPCR Probe Assay primer/probe mix, Integrated DNA Technologies) specific for the amplification of N1, N2 and RPP30 genes as previously reported. The volume of the PCR mix was 20 µL including 10 µL of ddPCR™ Supermix for Probes (No dUTP) and 1 µL of primers/probe. The droplet emulsion was thermally cycled on C1000 Touch Thermal Cycler (Bio-Rad) instrument. Cycling conditions were 95 • C for 5 min, followed by 40 cycles of amplification (94 • C for 30 s and 60 • C for 1 min), ending with 98 • C for 10 min, according to the manufacturer's protocol. The copies of the target gene were calculated automatically by the QuantaSoft™ software version 1.7.4 (Bio-Rad).

Plasmid controls contain the complete nucleocapsid gene from 2019-nCoV virus were provided by Integrated DNA Technologies and delivered at a concentration of 200,000 copies/µL in IDTE pH 8.0.

An oligonucleotide microarray targeting two regions (N1 and N2) of the Sars-CoV-2 nucleocapsid, and the human RPP30, was developed ( Figure 1A ).

We employed the QX100™ Droplet Digital™ PCR System (Bio-Rad Laboratories, 251

Hercules, CA, USA). 9 µL of cDNA previously diluted 1:20 were mixed with primers and 252 fluorophore-labeled commercial probes ( To interpret the results, the indications of the US CDC qPCR test were followed. In this assay, the fluorescence of capture spots indicates positivity. Briefly, a specimen is considered positive for SARS-CoV-2 if the two SARS-CoV-2 markers (N1, N2) produce a fluorescence signal that exceeds more than 3 times the standard deviation, the signal of the no-template control (NTC). On the contrary, a specimen is considered negative if the SARS-CoV-2 markers (N1, N2) show a signal non-discernible from that of the NTC. The RPP30 gene, in a positive sample, may or may not be positive. It is possible that some samples may fail to exhibit RPP30 fluorescence due to low cell numbers in the original clinical sample. A negative RPP30 signal does not preclude the presence of Sars-CoV-2 virus RNA in a clinical specimen. On the other hand, the absence of the RPP30 signal in a negative specimen makes the result invalid because the presence of an RPP30 signal in a sample negative for N1 and N2 confirms the correct extraction of RNA.

In this work, to greatly speed up the test, the viral RNA was extracted from the nasopharyngeal swabs (stored at −80 • C) using RNAGEM kit. RNA extraction by singlestep reagent isolation methods may leave behind contaminants that reduce amplification efficiency. However, the inhibition that, in standard media, would hamper the RT-PCR detection does not affect sensitivity of CovidArray assay thanks to the increased sensitivity of this microarray approach. To optimize the assay and further reduce the analysis time, the transcripts were amplified at a different number of cycles (data not shown). The optimization showed that the array is able to detect, after only 25 cycles, amplicons that are detectable at a higher number of cycles with standard RT-qPCR. A decrease of the number of cycles to 25 led to reduction of the analysis time (1h of reaction). Figure 1 shows a comparison between the workflows of the CovidArray platform and the standard RT-qPCR method. The limiting step of the standard workflow is the RNA extraction, which is overcome, in our workflow, by the use of the RNAGEM enzyme.

Firstly, it was demonstrated the correct assignment of samples previously assayed in the laboratory of Immunology at University of Milan (L. Sacco Hospital) with standard qPCR technology. In particular, five nasopharyngeal swabs (S1-S5) were analyzed, four of which positive for the target N1 and N2 by RT-qPCR and one negative for both (S3). The four positive samples presented different viral loads resulting in different threshold cycles (Ct) in qPCR. In particular, for N1, the sample S1 crossed the threshold line at 20.00 cycles due to its high viral load; also sample S2 has a high viral load (22.00 cycles) while S4 and S5 have lower viral load and are detected at 38 and 36 cycles respectively (Table 1) . Figure 2B shows the results of the microarray analysis of the swab samples and the control NTCs for the SARS-CoV-2 markers (N1 and N2) and for the RPP30 positive control. 

Amplicons of N1, N2, and RPP30 were separately incubated on microarray chips. 323

Three different silicon substrates were used to analyze one swab sample. As shown in 324 Figure 2B , the fluorescence signal appears only at the location where the immobilized cap-325 ture probe is complementary to the labeled PCR with no cross-hybridization and a good 326

reproducibility from spot to spot. The absence of fluorescence in the NTC subarrays is 327 essential as it allows discriminating low-signal samples from background signals. The 328 Amplicons of N1, N2, and RPP30 were separately incubated on microarray chips. Three different silicon substrates were used to analyze one swab sample. As shown in Figure 2B , the fluorescence signal appears only at the location where the immobilized capture probe is complementary to the labeled PCR with no cross-hybridization and a good reproducibility from spot to spot. The absence of fluorescence in the NTC subarrays is essential as it allows discriminating low-signal samples from background signals. The same samples were also assayed by ddPCR. Indeed, one of our samples (S3) from a patient negative according to the RT-qPCR assay, but with symptoms attributable to Covid-19, was found to be positive by CovidArray in agreement with ddPCR (N1 = 34.5 copies/µL; N2 = 7.6 copies/µL). CovidArray matches the sensitivity of ddPCR (Figure 3 ). 

Amplicons of N1, N2, and RPP30 were separately incubated on microarray chips. 323

Three different silicon substrates were used to analyze one swab sample. As shown in 324 Figure 2B , the fluorescence signal appears only at the location where the immobilized cap-325 ture probe is complementary to the labeled PCR with no cross-hybridization and a good 326

reproducibility from spot to spot. The absence of fluorescence in the NTC subarrays is 327 essential as it allows discriminating low-signal samples from background signals. The 328 same samples were also assayed by ddPCR. Indeed, one of our samples (S3) from a patient 329 negative according to the RT-qPCR assay, but with symptoms attributable to Covid- 19, 330 was found to be positive by CovidArray in agreement with ddPCR (N1 = 34.5 copies/µL; 331 N2 = 7.6 copies/µL). CovidArray matches the sensitivity of ddPCR (Figure 3 ). 332 333 Figure 3 . Plots of the relative fluorescence intensity (blue and green bars) of the images in Figure 2B , and the copies num- The CovidArray was further validated with seven nasopharyngeal swabs from IRCCS Mondino Foundation (Pavia), previously subjected to solid-phase extraction and RT-qPCR as reported in the experimental/materials and methods section. A 100% agreement between the two methods was found ( Table 2 ). The fluorescence images for samples B243, N053, and NTC are shown in Figure S1 . 

Since the new SARS CoV-2 emerged, researchers have struggled to develop highly sensitive molecular techniques to diagnose positive Covid-19 subjects effectively. Ac-cording to the WHO and the Centre for Disease Control and Prevention (CDC), the gold standard for the diagnosis is qPCR. However, many studies have highlighted the presence of false-negative results in RT-qPCR [32, 33] . Therefore, it is worthy to build up novel robust methodologies that ensure high sensitivity useful not only for diagnostic purposes but also for the follow-up of patients and for monitoring of the viral load. To evaluate our method's sensitivity, serial dilutions of linear DNA standard 2019-CoV Plasmid Control were tested using primer sets targeting N1 and N2 regions. The plasmid DNA was diluted to 50, 25, 5, 2.5, 0.5, 0.25, 0.05 copies/µL prior to undergo 25 cycles of PCR. To build the calibration curves for the two viral regions, the capture probes in 36 replicates (6 × 6 subarrays) were spotted on different coated silicon chips (one chip for each plasmid concentration). Samples with decreasing concentration of plasmid DNA were amplified, denatured, and finally hybridized for 15 min at room temperature.

The value of fluorescence intensity detected for each of the seven plasmid concentrations together with the background fluorescence of the control sample with no plasmid DNA was plotted versus the number of copies of the plasmid per µL in the starting solution. Figure 4A shows the calibration curves for the region N1 and N2, respectively. 

The LODs (lowest concentration of detectable plasmid DNA) extrapolated for each 376 marker are reported in Figure 4B . The determination of LOD is based on the equation: 3.3 377

σ/s where s is the slope of calibration curve and σ is the standard deviation of fluorescence 378 background in the control sample. The LODs found with this system are 1.16 copies/µL 379 for the N1 and 0.81 copies/µL for the N2, respectively. These LODs are comparable with 380 those declared by the various manufacturers of kits for qPCR with the difference that the 381 number of standard amplification cycles for those methods is 40 while, in our approach 382 only 25 cycles are sufficient to detect the target genes. 383

One of the peculiar features of microarray technology is its multiplexing capability. 385

Different genes can be revealed in a single hybridization assay by spotting onto the mi-386 croarray substrate different capture probes specific to the target. In this work, we ex-387 ploited the multiplexing capability of the CovidArray to detect the presence of the N1, N2 388 markers of SARS-CoV-2 and the RPP30 control gene on a single silicon chip in a single 389 hybridization assay. We performed a triplex PCR, amplifying simultaneously the genes 390 N1, N2, and RPP30 using the cDNA produced by reverse transcription of the RNA ex-391 The LODs (lowest concentration of detectable plasmid DNA) extrapolated for each marker are reported in Figure 4B . The determination of LOD is based on the equation: 3.3 σ/s where s is the slope of calibration curve and σ is the standard deviation of fluorescence background in the control sample. The LODs found with this system are 1.16 copies/µL for the N1 and 0.81 copies/µL for the N2, respectively. These LODs are comparable with those declared by the various manufacturers of kits for qPCR with the difference that the number of standard amplification cycles for those methods is 40 while, in our approach only 25 cycles are sufficient to detect the target genes.

One of the peculiar features of microarray technology is its multiplexing capability. Different genes can be revealed in a single hybridization assay by spotting onto the microarray substrate different capture probes specific to the target. In this work, we exploited the multiplexing capability of the CovidArray to detect the presence of the N1, N2 markers of SARS-CoV-2 and the RPP30 control gene on a single silicon chip in a single hybridization assay. We performed a triplex PCR, amplifying simultaneously the genes N1, N2, and RPP30 using the cDNA produced by reverse transcription of the RNA extracted from the same 5 nasopharyngeal swabs as reported in the "Materials and Method" section. The triplex PCR was hybridized with the probes spotted on the same substrate. The simultaneous appearance of fluorescence signals on the subarrays corresponding to the N1 and N2 regions confirmed the positivity of the sample. The fluorescence signal of the RPP30 subarray was also detectable. In Figure 5B five triplex PCR' hybridization results corresponding to the S1-S5 samples are shown. The NTC does not show significant fluorescence. In Figure 5C the histogram of fluorescence intensity indicates that the samples S1 and S2 have a higher fluorescence intensity confirming the higher viral load detected by the single-PCR CovidArray. Sample S3 which was considered negative by the RT-qPCR technique shows a weaker but detectable signal in agreement with ddPCR and single-PCR CovidArray. 

In summary, we describe a novel microarray platform, CovidArray, for the specific 411

and sensitive detection of SARS-CoV-2 in nasopharyngeal swabs. This approach combines 412 the physical-optical properties of the silicon substrate with the surface chemistry used to 413 bind to the substrate the oligonucleotide capture probes specific to the virus' genes to ob-414 tain a diagnostic tool of great sensitivity. In agreement with ddPCR, we correctly assigned 415 12 nasopharyngeal swabs of different origins. Thanks to the lower limit of detection of 416

CovidArray, we identified a false-negative sample. Another feature of our system, also 417 due to the high sensitivity of the CovidArray, is the decrease of the number of PCR cycles 418 required to detect the viral markers which, in turn, leads to a significant reduction of the 419 analysis time. A further contribution to speeding up the diagnostic assay is the use of an 420 alternative method for extracting the viral RNA from clinical samples. 421 

In summary, we describe a novel microarray platform, CovidArray, for the specific and sensitive detection of SARS-CoV-2 in nasopharyngeal swabs. This approach combines the physical-optical properties of the silicon substrate with the surface chemistry used to bind to the substrate the oligonucleotide capture probes specific to the virus' genes to obtain a diagnostic tool of great sensitivity. In agreement with ddPCR, we correctly assigned 12 nasopharyngeal swabs of different origins. Thanks to the lower limit of detection of CovidArray, we identified a false-negative sample. Another feature of our system, also due to the high sensitivity of the CovidArray, is the decrease of the number of PCR cycles required to detect the viral markers which, in turn, leads to a significant reduction of the analysis time. A further contribution to speeding up the diagnostic assay is the use of an alternative method for extracting the viral RNA from clinical samples.

In this work, we used an innovative approach, RNAGEM, commercially available by MicroGEM, to extract and purify viral RNA in less than 15 min. The total time required for the molecular test can thus range from about 3-6 h of a standard process to about 2 h with the CovidArray method. Moreover, RNAGEM provides an alternative to commercial RNA extraction kits that may undergo a shortage due to their massive use during the pandemic. Furthermore, in this work we have exploited the multiplexing capability of the microarray technology, to detect the presence of viral markers and the control sequence in a single assay.

The drawback of the current platform is the high degree of manual work required to perform the analysis. Our current throughput is still far from that of RT-qPCR. The analytical capability of our approach is only 16 samples at a time. However, there are no conceptual obstacles to integrating the assay into an automatic platform. The whole workflow, from viral RNA extraction to detection can be performed in a microtiter plate equipped with a microplate fluorescence reader.

Finally, it is worth noticing the versatility of this approach. In fact, CovidArray could potentially allow differentiating SARS-CoV-2 from other viral and bacterial respiratory tract infections by merely adding new primers and capture probes to the same array, becoming a promising diagnostic tool suitable for routine diagnosis of a wide range of respiratory diseases.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/s21072490/s1, Table S1 : Assay primer/probe sequences, Figure S1 : Cy3 fluorescence image of the CovidArray analysis of sample B243 and N053. 

A Novel Coronavirus from Patients with Pneumonia in China

The species Severe acute respiratory syndrome-related coronavirus: Classifying 2019-nCoV and naming it SARS-CoV-2

World Health Organization. Laboratory Testing for 2019 Novel Coronavirus (2019-nCoV) in Suspected Human Cases

General Office of the National Health and Health Commission Office of the NA of TCM. Diagnosis and Treatment of Pneumonitis with a New Type of Coronavirus Infection

New Novel Coronavirus Detection Products Approved by the National Medical Products Administration

United States Food and Drug Administration

False-negative results of initial RT-QPCR assays for COVID-19: A systematic review

Potential preanalytical and analytical vulnerabilities in the laboratory diagnosis of coronavirus disease 2019 (COVID-19)

False-Negative Results of Real-Time Reverse-Transcriptase Polymerase Chain Reaction for Severe Acute Respiratory Syndrome Coronavirus 2: Role of Deep-Learning-Based CT Diagnosis and Insights from Two Cases

COVID-19: Overcoming Supply Shortages for Diagnostic Testing

Large scale multiplex PCR improves pathogen detection by DNA microarrays

Multiple genotyping based on multiplex PCR and microarray

Technology for Simultaneous Detection and Species Identification of Seven Human Herpes Viruses

Analysis of KRAS, NRAS and BRAF mutational profile by combination of in-tube hybridization and universal tag-microarray in tumor tissue and plasma of colorectal cancer patients

Microarray approach combined with ddPCR: An useful pipeline for the detection and quantification of circulating tumor DNA mutations

COLD-PCR and microarray: Two independent highly sensitive approaches allowing the identification of fetal paternally inherited mutations in maternal plasma

Quantification of DNA and protein adsorption by optical phase shift

High sensitivity protein assays on microarray silicon slides

Characterization of a polymeric adsorbed coating for DNA microarray glass slides

A more accurate tool for SARS-CoV-2 detection in low viral load specimens

Sensitivity assessment of droplet digital PCR for SARS-CoV-2 detection

Centers for Disease Control and Prevention (CDC). CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-QPCR Diagnostic Panel

Evaluation of rapid antigen test for detection of SARS-CoV-2 virus

Comparison of Rapid Antigen Tests for COVID-19

COVID-19 Diagnostic Test Accuracy Group. Rapid, point-of-care antigen and molecular-based tests for diagnosis of SARS-CoV-2 infection

Sing, A. Rapid establishment of laboratory diagnostics for the novel coronavirus SARS-CoV-2 in Bavaria

Simultaneous RNA purification and size selection using on-chip isotachophoresis with an ionic spacer

One-Step Nucleic Acid Purification and Noise-Resistant Polymerase Chain Reaction by Electrokinetic Concentration for Ultralow-Abundance Nucleic Acid Detection

DNA microarray-based solid-phase PCR on copoly (DMA-NAS-MAPS) silicon coated slides: An example of relevant clinical application

Limits of detection of six approved RT-QPCR kits for the novel SARS-coronavirus-2 (SARS CoV-2)

Real-time RT-QPCR in COVID 19 detection: Issues affecting the results

We are grateful to Riccardo Vago (Università Vita-Salute S. Raffaele, Milan) and Greta Bergamaschi (SCITEC CNR, Milan) for their contribution without which this work would have been much more difficult. We also thank Giuseppina Sannino (MicroGEM UK Ltd., Southampton, UK) for her comments and for the supply of the RNAGEM Extraction Kit.

The authors declare no conflict of interest.Sensors 2021, 21, 2490