CRISPR/Cas9-Based Lateral Flow and Fluorescence Diagnostics

Clustered regularly interspaced short palindromic repeat (CRISPR/Cas) proteins can be designed to bind specified DNA and RNA sequences and hold great promise for the accurate detection of nucleic acids for diagnostics. We integrated commercially available reagents into a CRISPR/Cas9-based lateral flow assay that can detect severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) sequences with single-base specificity. This approach requires minimal equipment and represents a simplified platform for field-based deployment. We also developed a rapid, multiplex fluorescence CRISPR/Cas9 nuclease cleavage assay capable of detecting and differentiating SARS-CoV-2, influenza A and B, and respiratory syncytial virus in a single reaction. Our findings provide proof-of-principle for CRISPR/Cas9 point-of-care diagnosis as well as a scalable fluorescent platform for identifying respiratory viral pathogens with overlapping symptomology.


Introduction
The World Health Organization declared a pandemic in March of 2020 as a result of the global spread of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that causes coronavirus disease 2019 . Testing and contact tracing are cornerstones of prevention, mitigation, and control efforts. Patient demographics for those who succumb to COVID-19 are multifactorial; however, the ability to test at high capacity has been associated with reduced mortality [1]. The ability to test often correlates with better infrastructure, hospital capacity, healthcare quality, and effective public health systems. The 'gold standard' test for COVID-19 has been quantitative reverse transcription polymerase chain reaction (qRT-PCR) to detect viral nucleic acid and has been adopted by most public health agencies. Serological and antigen detection assays also comprise a minority of tests and are more streamlined with faster turnaround times. However, the nature of the immune response whereby IgM and IgG antibodies are produced can be days to weeks following exposure to the virus [2] and antigen testing is less sensitive, particularly for asymptomatic patient diagnoses [3,4]. Therefore, nucleic acid testing by qRT-PCR remains the most broadly applied procedure and requires sample collection, transport, RNA isolation, reverse transcription, amplification, sequence interrogation, and detection. Detection by qRT-PCR relies on a primer:probe set that specifically recognizes a target DNA sequence and generates a fluorescent signal during DNA amplification. This testing platform is based on specialized reagents, training, and infrastructure that is costly

Amplification
DNA template fragments were amplified with the indicated forward and reverse primers. Endpoint PCR was performed using 1 µM template and 0.2 µM final concentration of primers in a 100 µL reaction volume with Phusion High-Fidelity PCR Master Mix (Thermo Fisher, Waltham, MA) under the following conditions: 98 • C × 2 min and 34 cycles of 98 • C × 10 s, 62 • C × 10 s, and 72 • C × 15 s, with a final extension of 5 min.
Recombinase polymerase amplification (RPA): 1 µM of template was used with primers for recombinase polymerase amplification according to the manufacturer's instructions for the TwistDx Basic RPA Kit from TwistDx (Maidenhead, UK).

Soak DNA Oligonucleotides
The oligonucleotides were purchased from IDT (Coralville, IA, USA) and resuspended at 200 µM. Equal molar equivalents were mixed in Tris NaCl and denatured and renatured by heating for 5 min at 95 • C and cooling to room temperature at a rate of −0.1 • C/s.

Guide RNAs
All single guide RNAs (sgRNA) were purchased from Synthego (Menlo Park, CA, USA) and resuspended at 100 µM in Tris Ethylenediaminetetraacetic acid. Shown are the 20 bp sequences specific to the corresponding gene target. The remaining sgRNA architecture is the vendor supplied standard sequence for Streptococcus pyogenes Cas9 binding.

Biotinylated Cas9 and Lateral Flow
dCas9-3XFLAG™-Biotin Protein containing the D10A and H840A mutations from Milipore Sigma (Merck KGaA, Darmstadt, Germany) was resuspended in the included Reconstitution Solution to~1.7 mg/mL (8 pmol/µL), and 1 µL of 100 µM sgRNA was added to give~1.2 M excess of sgRNA. Cas9:sgRNA complexing was allowed to occur at room temperature for five minutes.
10 µL of unpurified PCR or RPA products was used with the Cas9:sgRNA complex in a 50 µL reaction at 37 • C. Where indicated, the double-stranded soak DNAs were used at the concentrations shown in the relevant figures. 20 µL of the Cas9 reaction was used for detection using the TwistDx Milenia HybriDetect1 lateral flow assay (Maidenhead, UK) under the manufacturer's recommendation.
For the assays in which Cas9 was included in the RPA, the reaction conditions and primer concentrations remained as above. Instead of using water to achieve the final 50 µL reaction volume, the indicated concentrations of soak ODNs were used to reach the final volume. RPA then proceeded at 37 • C with use of 20 µL of the post-reaction product for LFA.

CRISPR/Cas9 Fluorescence Assay
Equal volumes of amplification product and 100 µM disease-specific probe were mixed and denatured/renatured by heating for 5 min at 95 • C and cooling to room temperature at a rate of −0.1 • C /second. 20 µL of this product was used for single or multiplex fluorescence with 10 µg of Cas9 nuclease (Aldevron, Fargo, ND, USA) and 100 µM of sgRNA in 100 µL at 1X New England BioLabs buffer 3.1 (New England BioLabs, Ipswich, MA, USA).
Single fluorescence assays were performed in a black 96-well plate (ThermoFisher, Waltham, MA, USA) and signal was recorded using excitation: 485 nm, emission: 530 on a BioTek (Winooski, VT) plate reader. Multiplex fluorescence was performed in a 96-well skirted PCR plate (ThermoFisher, Waltham, MA, USA) and fluorescence was recorded in the FAM, VIC®, TAMRA (Carboxytetramethylrhodamine), and JUN™ channels every 30 s over one hour using the QuantStudio3 Real-Time PCR System (ThermoFisher, Waltham, MA, USA).

RT-PCR
Genomic RNA from severe acute respiratory syndrome-related coronavirus 2 (ATCC ® VR-1986D™) was acquired from the American Type Culture Collection (Baltimore, MD) and was deposited by the Centers for Disease Control and Prevention and obtained through BEI Resources, NIAID, NIH (Biodefense and Emerging Infections Research Resources Repository, National Institute of Allergy and Infectious Diseases, National Institutes of Health) Genomic RNA from SARS-Related Coronavirus 2, Isolate USA-WA1/2020, NR-52285.

Genome Analysis
SARS-CoV-2 genome sequence was obtained from the global initiative on sharing avian influenza data current to January 2021 (www.gisaid.org (accessed on 9 February 2021)).

Graphing and Statistics
Values were graphed using GraphPad Prism 9 (San Diego, CA, USA) and statistical evaluation was performed using one-way analysis of variance (ANOVA) and a post hoc Tukey's multiple comparisons test.

Images
Photography was performed with a Canon 5DIII with a Tamron 25-70 lens at the 70 mm setting. Figure images were produced with BioRender.com (Toronto, ON, Canada).

Rapid Nucleic Amplification and Lateral Flow Detection
Isothermal RPA amplification and detection via LFA are a simplified approach for nucleic acid analysis that avoids the need for specialized infrastructure ( Figure 1A) [6,25]. The LFA strips employed in our study require test material that is labeled with both FITC/FAM and biotin ( Figure 1B-E). FITC/FAM:biotin conjugated analytes are captured at the test band that contains a biotin ligand ( Figure 1C,E). FITC Au-NPs are in excess and a portion remain unbound and flow to the assay control band ( Figure 1D). This ensures that the LFA test strip is functional and suitable for interpretation. We designed forward and reverse PCR primers that were labeled with FITC and biotin respectively, to generate SARS-CoV-2 ORF8a gene amplicons that flanked a single nucleotide polymorphism at genomic nucleotide position 28144 that causes a L84S amino acid substitution (Figure 2A,B). ORF8a is a~100 amino acid protein with putative ER import signal sequences [29]. While the Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) assays detect the N and E genes respectively, we designed our assay for the ORF8a region of the SARS-CoV-2 genome because it does not overlap with other viral sequences and the L84S alteration was the first mutation observed in the USA (Figure 2A) [30]. This mutation is proximal to a protospacer-adjacent motif (PAM) of -NGG (N = any nucleotide and G = guanine) for the Streptococcus pyogenes Cas9, making it suited for Cas9 detection (Figure 2A). The use of dual-labeled (i.e., FAM and biotin) primers resulted in positive LFA test bands in either the presence or absence of an ORF8a template ( Figure 2C). We hypothesized that primer dimers that resulted in a single complex containing FITC and biotin were the cause of this false positive result and corrected it by titrating the primer concentration. It was observed that higher primer concentrations contributed to dimerization and high false positive rates and lower primer amounts diminished sensitivity ( Figure S1, in the Supplementary Materials). These results showed that direct amplification with labeled primers may be suboptimal for unambiguous nucleic acid detection of ORF8a by LFA.

Cas9 Allows for Specific SARS-Co-V2 ORF8a Sequence Detection
In order to avoid prohibitively high false positive test results, the labeling of the target amplicon was next approached using a FAM/FITC forward primer and an unlabeled reverse primer ( Figure 2D). We posited that biotinylation of this FITC-labeled PCR product could be accomplished using a biotinylated, nuclease inactivated ('dead') version of Cas9 (bdCas9) ( Figure 2D). FITC ORF8a amplicons generated by standard PCR were incubated with bdCas9 and an ORF8a-specific sgRNA (Figure 2A) or a mismatched control sgRNA (Supplementary Figure S2A). Under these conditions, we observed a readily observable test band using the COVID-19 sgRNA and a faint one using the control sgRNA with no homology to SARS-Co-V2 ( Figure 2E). Densitometry has been applied to ascribe values for semi-quantitative results as well as to differentiate the test bands of experimental and controls [31]. Because Cas9 is physically associated with DNA while scanning for target sites to cleave [32], we predicted that this interaction led to non-specific DNA labeling and a faint but visible test band in the control sgRNA ( Figure 2E). Therefore, to improve definitive interpretation, a competing 'soak' DNA that was rich in PAM-NGG sequences was designed in order to prevent indiscriminate Cas9:DNA interactions ( Figure 2F and Figure S2B). This allowed for specific detection using a COVID-19 sgRNA, while the control mismatched sgRNA did not show a test band, therefore false positive detection was avoided ( Figure 2G). Conversely, an irrelevant amplicon was not recognized by the COVID-19 sgRNA (Supplementary Figure S2C). These results confirmed the ability of bdCas9 to label a FITC amplicon for detection via LFA.
To merge the capabilities of RPA and our bdCas9-based detection method, we investigated the conditions for rapid SARS-Co-V2 sequence detection. First, RPA using a FITC-labeled forward and unlabeled reverse primer was performed in the presence of bdCas9. Even with a large excess of competitor soak DNA, a positive test band was observed on LFA strips with either COVID or control sgRNAs when a SARS-Co-V2 DNA template was present ( Figure 3A). The two reaction components were then separated, first by performing a 20 min RPA at room temperature and then using RPA products for bdCas9 detection. With a 20 min bdCas9 incubation, test bands were visible that became more defined at 40 and 60 min and with increasing amounts of soak DNA ( Figure 3B). Collectively, these data showed that SARS-CoV-2 DNA amplified by RPA can be detected with bdCas9 and LFA in the presence of a competitor/soak DNA.

Cas9-Nuclease-Based Diagnostics for Single and Multiplexed Targets
The nuclease properties of Cas9 also hold potential to serve as a diagnostic platform by cleaving a fluorescent probe in a sequence-specific manner ( Figure 4A). COVID-19specific probes labeled with a fluorescent marker and quencher were designed, that when hybridized with SARS-CoV-2 amplicons, were cleaved by Cas9 nuclease together with a COVID-19 sgRNA ( Figure 4B). The ability to multiplex Cas9 with multiple sgRNAs also allowed us to test whether we could achieve simultaneous detection of viruses with overlapping symptomology. First, we designed, built, and tested DNA probes with distinct fluorophores for SARS-CoV-2, influenza A and B, and RSV, respectively. These were tested and showed specificity of fluorescent signaling only for matched sgRNAs ( Figure 5A,B). Next, the four distinct viral detection components were all combined in a single reaction mixture and analyzed simultaneously under isothermal (37 • C) conditions in a single tube using real-time fluorometry with a standard quantitative PCR instrument in a 96-well format ( Figure 5C). Distinct fluorescence signals from cleavage of the pathogen-specific probe by the disease-specific sgRNA were observed. These results showed that all four viral pathogens could be detected in a multiplex fashion.

LFA and Fluorescence Assay Validation with SARS-CoV-2 Genomic RNA
Genomic RNA from the USA-WA1/2020 isolate was diluted and reverse transcribed followed by qRT PCR with the designated primer:probes employed by the Centers for Disease Control and Prevention for the N gene (Figures 2A and 6A, and Supplementary Figure S3). Using ORF8a primers, the same dilution series was performed and resolved by gel elec-trophoresis ( Figure 6B). This amplicon was then used for establishing the limit of detection (LOD) with the fluorescence and LFAs. Fluorescence signal above background was observed for each sample, indicating that the Cas9-based assay limit of detection (LOD) was similar to that of qRT-PCR ( Figure 6A,C). We further quantified ORF8a amplicons and performed a copy number LOD. Fluorescence above background was observed with 9 × 10 9 copies of target DNA ( Figure 6D). Under the parameters of 35 cycle exponential PCR amplification (2 35 ), this represents <5 copies of starting material. LFA optimization with soak DNA and irrelevant sgRNA was performed (Supplementary Figure S4). Under these conditions, test bands were observable at a sensitivity that was an order of magnitude below that of qRT-PCR or Cas9 fluorescence ( Figure 6E). Together, these results show that LFA and Cas9 fluorescence can be used for SARS-CoV-2 viral nucleic acid detection.

Cas9 Analysis of a SARS-CoV-2 Variant
Amino acid substitutions D614G and N501Y in the S gene and L84S in ORF8a have been suggested to result in increased viral load [33,34]. S gene D614G is highly prevalent [35] and N501Y has led to COVID surges [36]. Being able to distinguish SARS-CoV-2 strains may aid in whether certain strains are associated with differential clinical outcomes and/or could provide rapid information to public health departments. However, because the S gene between coronaviruses are highly homologous, we avoided targeting it with CRISPR/Cas9 to avoid false positives that may occur from a coronavirus other than SARS Co-V2. Instead, we targeted L84S caused by a SNP in the ORF8a gene that is unique to SARS-Co-V2 and delineated the relationships between S gene D614G, N501Y, and ORF8a L84S (Figure 2A and Figure S5). We tested the ability of Cas9 nuclease and bdCas9 to distinguish between cytosine and thymine at nucleotide position 28144 in ORF8a (Figure 2A). We interrogated the sequence of an ORF8a DNA amplicon with a perfectly matched fluorescent probe and sgRNAs with matched complementarity (thymine) or a one base pair mismatch (cytosine). Either wild-type or a high-fidelity version of Cas9 (SpyFi™) [37] were employed, and each showed differential fluorescence signal between matched and mismatched; however, it was not statistically significant ( Figure S6A,B, in the Supplementary Materials). We then assessed the ability of bdCas9 in the LFA assay to distinguish targets at the single nucleotide level. For this, we utilized two soak DNA candidates: the PAM-rich soak DNA or a more homologous competitor soak that differed from the target and sgRNA by a single nucleotide ( Figures S2B and S6C, in the Supplementary Materials). A test band was observed for both the perfectly matched and one base pair mismatched DNA target when the PAM-rich soak was used ( Figure S6D,E, in the Supplementary Materials). In contrast, when using soak DNA containing the single nucleotide mismatch from the target, only the sgRNA with perfect match to the target yielded a test band ( Figure S6D,E). These data showed that, using an appropriate soak DNA sequence, bdCas9 and LFA could resolve DNA targets at the single nucleotide level.

Discussion
The granting of emergency use authorization of SARS-Co-V2 vaccines represents a hopeful end to a pandemic that has infected more than 100 million people worldwide and claimed greater than 2 million lives from January 2020 to January 2021. It is predicted that widespread vaccine administration will not be available until the second or third quarter of 2021, making continued testing and mitigation efforts crucial to minimize more loss of life and continued global social, economic, and in-person schooling disruptions.
We set out to leverage the ability and specificity of Streptococcus pyogenes Cas9 to interrogate and identify SARS-Co-V2 sequences to develop testing platforms for both field-based and more specialized laboratory testing. The former requires simplified methodologies and rapid readouts and would be particularly helpful in rural areas that lack laboratory facilities able to perform molecular diagnostics. Rural COVID-19 case rates are increasing [38], rural residents are at elevated risk of COVID-19-related serious illness [39], medical care capacity in lowly populated areas can be quickly overwhelmed [40], and testing is challenging due to a lack of local facilities and funding [41]. To address these testing shortfalls, we employed Cas9 for targeting a portion of the SARS-CoV-2 ORF8a gene (Figure 2A). Presently, the CDC and WHO test for sequences in the N and E genes, respectively (Figure 2A). We chose ORF8a for our targeting strategies because there are seven coronaviruses that infect humans [42] and the ORF8a sequence is dissimilar between them, making it an ideal gene to target for SARS-CoV-2 detection without false positives occurring from these other viruses. Further, this site was ideal for designing a CRISPR/Cas9 sgRNA with an -NGG PAM (Figure 2A). The LFA test strips employed in our study require dual labeling of a candidate molecule for detection that is based on capture of the analyte by embedded Au-NPs coated with rabbit anti-FITC antibodies ( Figure 1B,C). If the test material lacks a biotin label, it flows past the test band that contains a biotin ligand and accumulates at the assay control band that is coated with anti-rabbit antibodies. The presence of this band, that we termed the assay control, ensures that the LFA reagents and test stick device are functioning properly for detection ( Figure 1D). If the reaction components are labeled with both FITC and biotin, a positive result is observed following Au-NP accumulation at the test band ( Figure 1E). Using recombinase polymerase amplification (RPA) that allows for isothermal amplification of nucleic acids, we observed that the use of PCR primers labeled with FITC and biotin resulted in false positive test bands as a result of primer dimerization ( Figure 2B,C and Supplementary Figure S1). In order to prevent this, we employed an amplification strategy using one FITC/FAM-labeled and one unlabeled primer ( Figure 2D). To achieve the requisite FITC and biotin conjugation for LFA detection, we employed a nuclease inactive, biotinylated Cas9 with SARS-CoV-2 (COVID-19) or control sgRNAs. Under these conditions, we avoided the false positives resulting from primer dimers; however, control sgRNA with no homology to ORF8a showed the presence of a test band ( Figure 2E). Cas9 can remain stably bound to DNA as it scans for sequences of homology required to initiate DNA cleavage [43]. In the absence of nucleolytic properties, such as with 'dead' Cas9, DNA is also scanned and the on/off rate is rapid and impacted by Cas9 concentration [44]. Therefore, the false positive test bands we observed may be due to high concentration of Cas9 that can associate and dissociate with the former, leading to test bands irrespective of the sgRNA. To remedy this, a bait or "soak" DNA sequence comprised of PAM-rich sequences was designed for inclusion in the assay in order to sequester non-specific binding events ( Figure 2F and Supplementary Figure S2). Under these new conditions, LFA test bands were observed using the COVID-19 but not the mismatched control sgRNA ( Figure 2G).
To further define the conditions for field-based testing, we explored the optimal settings for rapid amplification and detection via LFA. Previous studies using Cas enzymes have employed LAMP PCR and Cas in a 'one pot' approach for simultaneous amplification and detection [8,45]. In our system, this strategy yielded high levels of false positives even in the presence of large amounts of competitor soak DNA ( Figure 3A). In order to prevent this, isothermal RPA was performed followed by biotinylated Cas9 interrogation. Increased amounts of soak DNA and longer dbCas9:sgRNA:DNA incubation times from 20 to 60 min improved the resolution of detection ( Figure 3B). Centralized reference and public health diagnostic sites can still face backlogs, leading to increased turnaround time for results. Therefore, we sought to apply CRISPR/Cas9 toward developing a higher throughput methodology. The design was based on the nuclease properties of Cas9 to cleave a fluorescent probe hybridized to a target amplicon ( Figure 4A). Following a 20 min isothermal room temperature RPA, a fluorescently labeled SARS-CoV-2 probe was annealed to the reaction product and incubated with Cas9 peptide and an ORF8a sgRNA. This resulted in rapid fluorescence generation with a statistically significant difference between Cas9 with COVID-19 sgRNA and control and DNA:probe reactions observable in as little as 10 minutes ( Figure 4B). COVID-19 symptoms can mirror those of influenza, which is most prevalent during the winter months in the northern hemisphere, and it is possible for one infection to be confused for another or for co-infection with both agents to occur [46]. Therefore, the CDC has designed multiplex qRT-PCR assays capable of detecting multiple viral pathogens in the same sample [47]. We likewise assessed the multiplex capability of Cas9 to detect and distinguish SARS-CoV-2, influenza A and B, and further designed and added components for detecting RSV. Individual sgRNAs were first tested against specific pathogen DNA:probe hybrids to assess whether there was any cross-reactivity. Figure 5A shows that only the Cas9:sgRNA complex matched to the target DNA:probe generated a distinct fluorescent signal. These data and those of Figure 2C,E are complementary and support the mechanism of Cas that interacts with and scans DNA in a broad manner but cleaves DNA in a sequence-specific fashion [48]. As such, in the bdCas9 binding assay, a competitor soak DNA is required to inhibit this process in order to achieve specificity by LFA, while the nuclease-dependent fluorescent approach is capable of specificity without competitor soak DNA (Figures 2 and 5A). The detection kinetics for each target were distinguishable from mismatched sgRNA sample fluorescence signal ( Figure 5B). Because no promiscuous activity was observed between any sgRNA, the amplicon:probe hybrids were then pooled for single-well multiplex analysis. Rapid fluorescence was generated that was readily distinguishable from controls ( Figure 5C), showing the potential of CRISPR/Cas9 nucleasebased diagnostics. Importantly, this method can be performed on a quantitative PCR instrument, and while this represents specialized equipment, it is also standard equipment in many research laboratories and most public health facilities. The speed of detection, capability to multiplex, and the ability to perform reactions in a 96-well format (or greater) makes this a scalable platform for high-throughput analytics.
With proof-of-principle established for our methodology using synthetic fragments, we then validated both LFA and fluorescence detection using the USA-WA1/2020 coronaviral isolate. A dilution series was performed using the CDC qRT PCR assay with the N1 and N2 primer:probe set and the 1:10,000 dilution showed a Ct of 34.9 ± 0.78 for N1 and 36.7 ± 0.55 for N2 ( Figure 6A and Supplementary Figure S3). Our CRISPR/Cas9 fluorescence assay was also able to detect the 1:10,000 dilution above background ( Figure 6C). CDC guidelines are for positive tests to have a Ct < 40 [49] and other studies show reduced ability to isolate virus when Ct values exceed 35 [50]. Thus, our fluorescence assay is comparable to qRT-PCR with a LOD that correlates to a Ct of~35. In contrast, the LOD of LFA was an order of magnitude lower ( Figure 6E), which is in keeping with its detection by visualization vs fluorescence, making it less sensitive, and may result in an inability to detect patients with low viral titers and makes follow-on confirmation of rapid tests important. To facilitate streamlined confirmation of LFAs, part of our design strategy was to employ a reverse transcription step using oligo dT/random hexamer priming. This allows for the same sample to be tested by LFA, fluorescence, and/or qRT-PCR. This is differential to some 'one-pot' approaches, such as RT LAMP; however, using gene-specific priming in this manner, we observed unacceptably high false priming events during PCR (data not shown). Moreover, the use of oligo dT/hexamer priming allows for standard control gene analysis during confirmation testing using nucleic acid amplification tests in accordance with CDC and WHO guidelines. Additionally, our RT strategy will support whole-genome sequencing that will provide further knowledge on coronaviral strain distribution and prevalence [51]. Follow-on confirmation testing is critical for any field-based assay, particularly for our LFA that does not generate/evaluate a human control gene. Rather, the observed assay control band shows proper function of the LFA strips. Future improvements to our approach will include a human control gene that is differentially labeled such that viral target and control genes can be analyzed on the same LFA.
Coronaviruses have genetic proofreading systems [52]; however, mutations occur with potential to confer favorable properties, including increased infectivity or to diminish diagnostic capabilities [53]. S gene mutations such as D614G or N501Y and ORF8a L84S are suggested to have higher rates of infectivity [33][34][35]. The S genes in which the D614 and N501Y residues reside are homologous between SARS-CoV and SARS-CoV-2 [54]. In contrast, the ORF8a gene shows little homology between coronaviruses and the L846 polymorphism is in the seed sequence of the sgRNA using the CRISPR/Cas9 enzyme from Streptococcus pyogenes (Figure 2A). The seed sequence is the first~10 bp proximal to the PAM and dictates specificity to a higher degree than PAM distal sequences [55][56][57][58][59]. These properties make ORF8a a desirable target site for overall specificity and single nucleotide analysis. Using wild-type or a high-fidelity Cas9, we assessed the single nucleotide distinction capabilities of Cas9 nuclease for ORF8a L84 (thymine) or S84 (cytosine) in our fluorometric assay. These data showed a trend for each enzyme in generating higher fluorescence signals when perfect homology between the DNA:sgRNA was present; however, the differences were not statistically significant (Supplementary Figure S6A,B). In contrast, analysis with bdCas9 and LFA allowed for single-nucleotide resolution when a soak DNA was included that was mismatched from the target DNA sgRNA site by 1 bp (Figures S6C-E and S2B, in the Supplementary Materials). Being able to rapidly obtain information on viral polymorphisms could greatly enhance the ability to track the spread or infectivity of novel viral strains. The mutational rate of SARS-CoV-2 shows hotspots in the Orf1ab [60] gene, which may owe to its larger size, and diagnostics that require sequence specificity such as ours may be invalidated should a target site be mutated [53]. However, the plasticity of CRISPR/Cas9 targeting allows for the rapid development and deployment of new reagents to circumvent this. Further, our deliberate use of wholly commercial reagents, most of which are obtained lyophilized and therefore highly stable, supports the development and archival of assays and reagents for current and emergent biological threat events.
In this study, we showed proof-of-principle for Cas9 in detecting target sequences for analysis by LFA or fluorometry. Single nucleotide resolution by LFA can aid in strain identification and has broad applicability for rapidly assessing circulating viral pathogens or other targets for diagnostics, prognostics, drug metabolism, etc. [61]. Fluorescence-based analysis showed high specificity and sensitivity, and in a multiplex fashion, was able to identify four disparate respiratory viral pathogen sequences. LFA allows for field-based application, while the fluorescence assay is highly scalable, allowing for quicker turnaround times. As others have reported [62], the robust amplification obtained by RPA and LAMP can result in cross-contamination and special precautions are required, particularly in multi-step reaction conditions such as ours. In addition, faint bands on LFA can occur [63], particularly with increased exposure time mandating the inclusion of rigorous controls and readout standard operating procedures (i.e., evaluation in <5 min).

Conclusions
The approval of the SARS-CoV-2 vaccine is highly promising but the time between the first doses and herd immunity will be months. Using commercial reagents, we describe a Cas-9-based detection methodology for nucleic acid detection using lateral flow assays and fluorescence signal generation. Our approach adds to the armamentarium of testing methodologies that can be brought to bear to bridge the immunization-immunity gap.

Patents
Patent pending.  Figure S3. Quantitative reverse transcriptase PCR of coronaviral genomic RNA. RNA from the USA-WA1/2020 strain was diluted from 1:10 to 1:10,000, reverse transcribed, and analyzed using the CDCN gene primer:probe sets. Data are duplicates representative of three analyses and (A) is the N1 probe and (B) is the N2 probe. The y-axis shows Rn that is the reporter fluorescent signal normalized to ROX and the x-axis shows cycle number. Figure S4. Optimization of LFA for SARS-Co-V2 genomic RNA. DNA amplicons were incubated with Cas9 and a mismatched or COVID-19-specific sgRNA in the presence of various amounts of soak DNA. LFA products are shown with the blue arrow. Figure S5. Correlative genome analysis of SARS-Co-V2 variants. (A) The correlation between the D614G (nucleotide 1842) strain of SARS-Co-V2 and strain with L84S due to a single-nucleotide polymorphism at genome nucleotide coordinate 28114 was assessed for the Midwestern state of Minnesota (USA). A pie graph for 1610 patients shows the relationship between the amino acid aspartic acid (D) or glycine (G) at amino acid position 614 and cytosine (C) or thymine (T) nucleotide at 28114. (B) The relationship between the ORF8a L84 or S84 and N501Y is shown. The genomes were analyzed from 1816 patients in the United Kingdom. Figure