key: cord-102270-rfhtlodc
authors: Azhar, Mohd.; Phutela, Rhythm; Ansari, Asgar Hussain; Sinha, Dipanjali; Sharma, Namrata; Kumar, Manoj; Aich, Meghali; Sharma, Saumya; Rauthan, Riya; Singhal, Khushboo; Lad, Harsha; Patra, Pradeep Kumar; Makharia, Govind; Chandak, Giriraj Ratan; Chakraborty, Debojyoti; Maiti, Souvik
title: Rapid, field-deployable nucleobase detection and identification using FnCas9
date: 2020-04-21
journal: bioRxiv
DOI: 10.1101/2020.04.07.028167
sha: 
doc_id: 102270
cord_uid: rfhtlodc

Detection of pathogenic sequences or variants in DNA and RNA through a point-of-care diagnostic approach is valuable for rapid clinical prognosis. In recent times, CRISPR based detection of nucleic acids has provided an economical and quicker alternative to sequencing-based platforms which are often difficult to implement in the field. Here, we present FnCas9 Editor Linked Uniform Detection Assay (FELUDA) that employs a highly accurate enzymatic readout for detecting nucleotide sequences, identifying nucleobase identity and inferring zygosity with precision. We demonstrate that FELUDA output can be adapted to multiple signal detection platforms and can be quickly designed and deployed for versatile applications including rapid diagnosis during infectious disease outbreaks like COVID-19.

The rise of CRISPR Cas9 based approaches for biosensing nucleic acids has opened up a broad diagnostic portfolio for CRISPR products beyond their standard genome editing abilities 1, 2 . In recent times, CRISPR components have been successfully used for detecting a wide variety of nucleic acid targets such as those obtained from pathogenic microorganisms or disease-causing mutations from various biological specimens [3] [4] [5] [6] [7] [8] [9] [10] . At the heart of such a detection procedure lies the property of CRISPR proteins to accurately bind to target DNA or RNA, undergo conformational changes leading to cleavage of targets generating a reporter-based signal outcome [11] [12] [13] [14] [15] . To enable such a detection mechanism to be foolproof, sensitive and reproducible across a large variety of targets, the accuracy of DNA interrogation and subsequent enzyme activity is extremely critical, particularly when clinical decisions are to be made based on these results 1,2 .

Current technologies relying on using CRISPR components for nucleic acid detection can sense the identity of the target either through substrate cleavage mediated by an active CRISPR ribonucleoprotein (RNP) complex or by binding through a catalytically inactive RNP complex. Cleavage outcomes are then converted to a reporter-based readout with or without signal amplification. Among the CRISPR proteins that have been used so far, Cas9 and Cas12 or their inactive forms have been predominantly employed for detecting DNA sequences while Cas13 has been used for both DNA and RNA sequences. Each of these approaches has its own strengths and limitations that are related to sensitivity, specificity and read-out modes for an accurate diagnosis. The primary focus of these platforms is towards detection of low copy numbers of nucleic acids from body fluids of patients where signal amplification through collateral activity of fluorescent reporters has been proven to be advantageous. For genotyping individuals with high confidence, including careers of We have recently reported a Cas9 ortholog from Francisella novicida (FnCas9) showing very high mismatch sensitivity both under in vitro and in vivo conditions [16] [17] [18] . This is based on its negligible binding affinity to substrates that harbor mismatches, a property that is distinct from engineered Cas proteins showing similar high specificity 19 . We reasoned that FnCas9 mediated DNA interrogation and subsequent cleavage can both be adapted for accurately identifying any SNVs provided that the fundamental mechanism of discrimination is consistent across all sequences ( Figure   1A ). We name this approach FnCas9 Editor Linked Uniform Detection Assay (FELUDA) and demonstrate its utility in various pathological conditions including genetic disorders and infectious diseases including disease outbreaks like COVID-19 [20] [21] [22] .

To identify a SNV with high accuracy, we first sought to investigate if FnCas9 can be directed to cleave the wild type (WT) variant at a SNV by placing an additional mismatch in the sgRNA sequence specific to the SNV. To test this, we selected sickle cell anemia (SCA), a global autosomal recessive genetic disorder caused by one point mutation (GAG>GTG) [23] [24] . We identified an sgRNA that can recognize the main disease causing point mutation (GAG > GTG) on account of a PAM site in the vicinity ( Figure 1B ). We then fixed the position of this mutation with respect to PAM and changed every other base in the sgRNA sequence to identify which combination led to complete loss of cleavage of a wild type substrate in an in vitro cleavage (IVC) assay with FnCas9 ( Figure 1B, Supplementary Figure 1A ). We found that two mismatches at the 2 nd and 6 th positions away from the PAM completely abrogated the cleavage of the target ( Figure 1B ). Similarly other combinations (2 and 7, 2 and 8, 2 and 9, and 2 and 12, numbers referring to positions away from PAM) also abolished the cleavage, although to slightly lower levels ( Figure 1B) . Notably, on the SCA substrate, the 2 and 6 mismatch combination produced near complete cleavage suggesting that this combination may be favorable for discriminating between WT and SCA substrates ( Figure 1B) . Importantly, the same design principle can guide the allelic discrimination at every SNV that appears in these positions upstream of the trinucleotide NGG PAM in DNA.

We performed FELUDA with 4 synthetic sequences corresponding to SNVs reported for the Mendelian disorders like Glanzman's Thrombasthenia, Hemophilia A (Factor VIII deficiency), Glycogen Storage Disease Type I and X linked myotubular myopathy and observed an identical pattern of successful discrimination at the SNVs (Supplementary Figure 1B) . Taken together, these experiments suggest that FELUDA design can be universally used for detection of SNVs and and would not require extensive optimization or validation steps for new SNVs. To aid users for quick design and implementation of FELUDA for a target SNV, we have developed a webtool JATAYU (Junction for Analysis and Target Design for Your FELUDA assay) that incorporates the above features and generates primer sequences for amplicon and sgRNA synthesis (https:// jatayu.igib.res.in, Supplementary Figure 2 ). We next tested FELUDA in DNA from 6 SCA patients and a healthy control and found that in every case, the SC mutation containing substrates were cleaved to give a distinguishable signature while the WT substrate remained intact (Supplementary Figure 3A ). To establish that enzyme specificity for position-specific mismatches with respect to PAM site is the fundamental reason for this discrimination, we designed the mismatch combinations such that cleavage will occur only for the WT substrate and observed identical results thus confirming our hypothesis (Supplementary Figure   3B ). Notably, a simple agarose gel electrophoresis can be employed for this discrimination suggesting that FELUDA can be used in routine molecular biology labs to establish the presence of an SNV in a DNA sequence.

In recent times fluorescence based nucleic acid detection has been widely used for several CRISPRDx platforms, particularly where collateral cleavage of reporters has been employed. Although FELUDA results can be precisely determined by agarose or capillary electrophoresis, we envisioned fluorescence or chemiluminescence as alternate end-point readouts to expand the scope of devices that can suit FELUDA based detection. To implement this, we first investigated if FELUDA can be adapted to a non-cleavage, affinity-based method of detection which works with single nucleotide mismatch sensitivity.

To develop such a readout, we tested FELUDA with a catalytically dead FnCas9 (dFnCas9) tagged with a fluorophore (GFP) and investigated if its mismatch discrimination is regulated at the level of DNA binding (Supplementary Figure 1A) .

We performed microscale thermophoresis (MST) assays to measure the binding affinity of inactive FnCas9-GFP RNP complex with WT or SCA substrates. We observed that the SCA substrate showed moderately strong binding (K d = 187.2 ± 3.4 nM) whereas the WT substrate exhibited very weak binding (K d = 1037.4 nM ± 93.3 nM) consistent with the absence of cleavage on the IVC readout ( Figure 2A) .

We then developed a pipeline to adapt FELUDA for an affinity-based fluorescent read-out system, where the amplification step generates biotinylated products which can then be immobilized on magnetic streptavidin beads. Upon incubation with fluorescent components in FELUDA, the absence or presence of a mismatch guides the binding of FnCas9 molecules to the substrate leading to loss of fluorescence signal in the supernatant ( Figure 2B ). We tested FELUDA using WT or SCA substrates and observed distinct signatures that distinguished between the two alleles suggesting that FELUDA can be adapted for a fluorescence-based readout ( Figure 2B ).

Next, we investigated if FELUDA detection can be extended to a quick point of care diagnosis of SNVs using lateral flow strips for instrument-free visual detection.

Although such read-outs have been demonstrated with CRISPR effectors that have collateral activity, they rely on a secondary signal amplification step where fluorescent oligos are added to the reaction setup. We designed an assay using FAM labelled RNP complex and biotin labelled amplicons on commercially available paper strips. In such a reaction, RNP-bound substrate molecules in a solution can lead to aggregation of the complex on a distinct test line of the strip while anti-FAM antibody linked gold nanoparticles accumulate on a control line on the strip ( Figure   2C ). To enable FAM labelling of sgRNA, we first validated the successful activity of chimeric FnCas9 sgRNAs by altering the length of overlap between crRNA:tracrRNA and observed that 19nt overlap can efficiently cleave the target ( Figure 2D ). Since a single FAM labelled tracrRNA is compatible with multiple crRNAs, this design also reduces the time and cost of a FELUDA assay. Next, we performed this assay with the HBB target and were able to detect up to femtomolar levels of target DNA in a solution suggesting that even without signal amplification, FELUDA reaches sensitivity similar to that reported for CRISPRDx platforms employing collateral activity ( Figure 2E ). Finally, we tested FELUDA using WT and SCA samples and obtained clear distinguishing bands for either condition validating the feasibility of visual detection for FELUDA diagnostics with complete accuracy ( Figure 2F ).

Genotyping carrier individuals with heterozygosity though non-sequencing PCR based methods is often complicated and requires extensive optimization of primer concentration and assay conditions. Although sickle cell trait (SCT) individuals are generally non-symptomatic, carrier screening is vital to prevent the spread of SCA in successive generations and is widely employed in SCA control programs in various parts of the world 19 . We speculated that the high specificity of FELUDA can be extended to identifying carriers since FnCas9 would cleave 50% of the DNA copies carrying SCA mutation and thus show an intermediate cleavage pattern (Figure 3A .)

We also investigated the use of saliva instead of blood as a non-invasive source of genomic DNA that would allow genotyping children and aged subjects where drawing blood may not be feasible. We obtained a clear, distinguishable signature of DNA cleavage in an SCT subject that was intermediate between normal or SCA individuals suggesting that FELUDA can be successfully used for determining zygosity at a specific SNV ( Figure 3A ). Although this reinforced the inherent sensitivity of FELUDA for genotyping targets with 1bp mismatches, obtaining the three distinct readouts would necessitate significantly robust reaction components and high reproducibility across multiple subjects. We performed a blinded experiment using DNA obtained from 49 subjects with all three genotypes from a Tertiary care center. Remarkably, the FELUDA results perfectly matched with the sequencing data from the same samples performed in a different laboratory (CSIR Center for Cellular and Molecular Biology, Chandak Lab) and thus identified all three genotypes with 100% accuracy ( Figure 3B ).

Although several pathogenic SNVs are located close to a NGG PAM and can be accurately targeted by FELUDA, these form only a small subset of the total number of disease-causing SNVs catalogued in ClinVar database (Supplementary Figure   4 ) 25 . To detect the non-PAM proximal SNVs we designed an in-built PAM site in the amplification step of FELUDA. We tested this approach using 2 SNVs (A2142G and A2143G) present in Helicobacter pylori 23s rRNA gene and which do not have a NGG PAM in the vicinity. These SNVS confer variable clarithromycin resistance in patients with gastric ulcers and clinically pose a serious concern for physicians 26 . We validated that PAM-mer based amplification can be successfully used for FnCas9 based IVC by targeting synthetic DNA sequence containing the H. pylori 23s rRNA gene ( Figure 3C ). We then isolated bacteria DNA from patient gut biopsy samples and successfully distinguished the antibiotic resistance genotype from another closely matched synthetic wild type sequence ( Figure 3C ). Importantly, this procedure takes a few hours from obtaining sample to diagnosing the variant, a significant improvement over existing regimens which rely on antimicrobial susceptibility tests that can take several days to complete 27 .

Since our design incorporates the need for fixing the SNV at either 2 nd or 6 th position proximal to PAM, this limits the possibility of discriminating substrates using a single FELUDA reaction. To overcome this, we explored single mismatches in the sgRNA sequence that might lead to abrogation of cleavage or binding from the substrate particularly at positions 16-19 at PAM distal end, since these showed minimum cleavage from our previous study. Remarkably, we found that FnCas9 shows negligible cleavage at each of these positions (Supplementary Figure 5A ). In particular, mismatch at PAM distal 16 th base shows complete absence of cleavage and negligible binding affinity to mismatched substrate ( Figure 3D ). To confirm this strategy, we targeted the SNV rs713598 which is associated with either G or C or G/C (heterozygous) variants in different individuals 28 Figure 10B) . Its ease of design and implementation, as exemplified by its urgent deployment during the COVID-19 health crisis offers immense possibilities for rapid and wide-spread testing that has so far proven to be successful in spreading the progression of the disease in multiple countries.

The sequences for the Hbb (WT and SCA), EMX1 and VEGFA site 3 were PCR Similarly, a 500bp region flanking two SNVs (A2142G and A2143G) in Helicobacter pylori 23s rRNA gene were ordered as synthetic DNA cloned in pUC57 by EcoRV.

The cloned sequences were confirmed by Sanger sequencing.

All the plasmid constructs used were adapted from our previous study 16 

In vitro transcription for sgRNAs/crRNAs were done using MegaScript T7 Transcription kit (Thermo Fisher Scientific) using a T7 promoter containing template as substrates. IVT reactions were incubated overnight at 37℃ followed by DNase treatment as per kit instructions and then purified by NucAway spin column (Thermo Fisher Scientific) purification. IVT sgRNAs/crRNAs were stored at -20℃ until further use.

Human Genomic DNA was extracted from the blood using the Wizard Genomic DNA Purification kit (Promega) as per the instructions.

For genomic DNA extraction from saliva, 1ml of saliva was centrifuged at 13000 rpm followed by three washes with 1ml of 1X PBS. After washing, the pellet was lysed with 50µl of 0.2% Triton X100 at 95°C for 5 minutes. Then again centrifuged at 13000 rpm and supernatant was transferred into a fresh vial. A total volume of 1µl of the supernatant was used in PCR reaction or otherwise stored at -20 °C.

Genomic DNA was extracted from the biopsy samples (15) (16) (17) (18) (19) (20) 

RPA reaction was set up as recommended in the TwistAmp ® Basic kit. 3-5 ng of genomic DNA was used with normal or biotinylated primers and the reaction was performed at 39℃ for 20 minutes. Amplicons were then purified with Qiagen PCR clean up kit and visualized on agarose gel.

DNA samples were directly PCR amplified for the target region while RNA samples were converted to cDNA using Reverse Transcriptase kit (Qiagen) and PCR amplified. In vitro cleavage assay was performed as optimized in our previous study 16 

Chimeric gRNA (crRNA:TracrRNA) was prepared by mixing in-vitro transcribed crRNA and synthetic 3'-FAM labelled TracrRNA in a equimolar ratio using annealing buffer (100mM NaCl, 50mM Tris-HCl, pH 8.0 and 1mM MgCl 2 ), mix was heated at 95 C for 2-5 minutes and then allowed to cool at room temperature for 15-20 minutes. Chimeric gRNA-dead FnCas9 RNP complex (500nM) was prepared by mixing them in a buffer (20mM HEPES, pH7.5, 150mM KCl, 1mM DTT, 10% glycerol, 10mM MgCl 2 ) and incubated for 10 min at RT. RPA or PCR amplified biotinylated amplicons were then incubated with the RNP complexes for 15 minutes at 37°C. Dipstick buffer was added along with the Milenia Hybridetect paper strip (TwistDx) was added as per instructions, producing dark colored bands over the strip within 2-10 min.

FnCas9-sgRNA complex (500nM) was prepared by mixing them in a buffer (20mM HEPES, pH7.5, 150mM KCl, 1mM DTT, 10% glycerol, 10mM MgCl 2 ) and incubated for 10 min at RT. The reconstituted RNP complexes along with PCR amplified DNA amplicons were then used for IVC assays at different temperatures ranging from 10°C to 50°C for 30 minutes. The reaction was inhibited using 1µl of Proteinase K (Ambion). After removing residual RNA by RNase A (Purelink), the cleaved products were visualized on agarose gel.

FnCas9-sgRNA complex (500nM) was prepared by mixing them in a buffer (20mM HEPES, pH7.5, 150mM KCl, 1mM DTT, 10% glycerol, 10mM MgCl 2 ) and incubated for 10 min at RT. Linearized plasmid used here as a substrate was incubated with reconstituted RNP complexes at different time points starting from 0 h to 100 h with or without 10% sucrose in the reaction buffer respectively. Further, cleaved products were visualized on agarose gel.

For the binding experiment, dFnCas9-GFP protein was used along with PAGE purified respective IVT sgRNAs. Notably, IVT sgRNAs were purified by 12% Urea- with RNP complex at 37 0 C for 60 min in reaction buffer. NanoTemper standard treated capillaries were used for loading the sample. Measurements were performed at 25°C using 40% LED power in blue filter (465-490nm excitation wavelength; 500-550nm emission wavelength) and 40% MST power. All experiments were repeated at least three times for each measurement. All Data analyses were done using NanoTemper analysis software. The graphs were plotted using OriginPro 8.5

software.

The sequencing reaction was carried out using Big dye Terminator v3.1 cycle sequencing kit (ABI, 4337454) in 10μl volume (containing 0.5μl purified DNA, 0.8μl sequencing reaction mix, 2μl 5X dilution buffer and 0.6μl forward/ reverse primer) with the following cycling conditions -3 mins at 95°C, 40 cycles of (10 sec at 95°C, 10 sec at 55°C, 4 mins at 60°C) and 10 mins at 4°C. Subsequently, the PCR product was purified by mixing with 12μl of 125mM EDTA (pH 8.0) and incubating at RT for 5 mins. 50μl of absolute ethanol and 2μl of 3M NaOAc (pH 4.8) were then added, incubated at RT for 10 mins and centrifuged at 3800rpm for 30 mins, followed by invert spin at <300rpm to discard the supernatant. The pellet was washed twice with 100μl of 70% ethanol at 4000rpm for 15 mins and supernatant was discarded by invert spin. The pellet was air dried, dissolved in 12μl of Hi-Di formamide (Thermo fisher, 4311320), denatured at 95°C for 5 mins followed by snapchill, and linked to ABI 3130xl sequencer. Base calling was carried out using sequencing analysis software (v5.3.1) (ABI, US) and sequence was analyzed using Chromas v2.6.5 (Technelysium, Australia).

ClinVar dataset (version: 20180930) was used to extract disease variation spectrum that can be targeted by FELUDA 25, 36 

JATAYU (jatayu.igib.res.in) is a web tool which enables users to design sgRNAs and primer for the detection of variations. Users need to provide a valid genomic sequence with position and type of variation. JATAYU front-end has been created using Bootstrap 4 and jQuery. In the back-end, python-based Flask framework has been used with genome analysis tools BWA (Burrows-Wheeler aligner) and bedtools 41-42 . All oligos used in the study are listed in Table 1 .

We thank all members of Chakraborty and Maiti labs for helpful discussions and valuable insights. We are grateful to Mitali Mukerji, Rajesh Pandey and Mohd. Faruq 

The present study was approved by the Ethics Committee, Institute of Genomics and Step 2: Users need to provide mutation information such as Position and type of mutation.

Step 3: Confirmation of the mutation information.

Step 4: Design of sgRNA and primers for the given input sequence and mutation. Only one fluorescent RNP will bind to the target while the other RNPs will dissociate.

The process can be iteratively performed to identify each nucleobase in a sequence. 

Next-generation diagnostics with CRISPR. Science (80-. )

CRISPR/Cas Systems towards Next-Generation Biosensing

Low-Cost Detection of Zika Virus Using Programmable Biomolecular Components

CDetection: CRISPR-Cas12b-based DNA detection with subattomolar sensitivity and single-base specificity

FLASH: a next-generation CRISPR diagnostic for multiplexed detection of antimicrobial resistance sequences

SHERLOCK: nucleic acid detection with CRISPR nucleases

HOLMESv2: A CRISPR-Cas12b-Assisted Platform for Nucleic Acid Detection and DNA Methylation Quantitation

Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a, and Csm6. Science (80-. )

Pathogen detection in the CRISPR-Cas era

Highly Effective and Low-Cost MicroRNA Detection with CRISPR-Cas9

C2c2 is a single-component programmable RNAguided RNA-targeting CRISPR effector. Science (80-. )

CRISPR-Cas12a has both cis-and trans-cleavage activities on single-stranded DNA

CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity. Science (80-. )

Cas13b Is a Type VI-B CRISPR-Associated RNA-Guided RNase Differentially Regulated by Accessory Article Cas13b Is a Type VI-B CRISPR-Associated RNA-Guided RNase Differentially Regulated by Accessory Proteins Csx27 and Csx28

Cas13d Is a Compact RNA-Targeting Type VI CRISPR Effector Positively Modulated by a WYL-Domain-Article Cas13d Is a Compact RNA-Targeting Type VI CRISPR Effector Positively Modulated by a WYL-Domain-Containing Accessory Protein

Francisella novicida Cas9 interrogates genomic DNA with very high specificity and can be used for mammalian genome editing

Targeted activation of diverse CRISPR-Cas systems for mammalian genome editing via proximal CRISPR targeting

Structure and Engineering of Francisella novicida Cas9

targeting accuracy

World Report Developing antibody tests for SARS-CoV-2

SARS-CoV-2 Testing

The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak -an update on the status

When basic science reaches into rational therapeutic design: from historical to novel leads for the treatment of β globinopathies

Sickle Cell Disease

ClinVar: Public archive of relationships among sequence variation and human phenotype

Mutations in the 23S rRNA gene are associated with clarithromycin resistance in Helicobacter pylori isolates in Brazil

Molecular Patterns of Resistance Among Helicobacter pylori Strains in South-Western Poland

Association of the bitter taste receptor gene TAS2R38 (polymorphism RS713598) with sensory responsiveness, food preferences, biochemical parameters and body-composition markers. A cross-sectional study in Italy

Isothermal Amplification of Nucleic Acids

Clinical manifestations of dengue in relation to dengue serotype and genotype in Malaysia: A retrospective observational study

Serotype influences on dengue severity: A cross-sectional study on 485 confirmed dengue cases in Vitória

Recombinase polymerase amplification: Basics, applications and recent advances

Comment Scientific and ethical basis for socialdistancing interventions against COVID-19

Improved molecular diagnosis of COVID-19 by the novel, highly sensitive and specific COVID-19-RdRp/Hel real-time reverse transcription-polymerase chain reaction assay validated in vitro and with clinical specimens

Aerosol and Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1

ClinVar: improving access to variant interpretations and supporting evidence

A new coronavirus associated with human respiratory disease in China

Unique and Conserved Features of Genome and Proteome of SARS-coronavirus, an Early Split-off From the Coronavirus Group 2 Lineage

NCBI Viral Genomes Resource

SeqMap: mapping massive amount of oligonucleotides to the genome

Fast and accurate long-read alignment with Burrows-Wheeler transform

BEDTools: a flexible suite of utilities for comparing genomic features