key: cord-0286996-e2v4uin7
authors: Luo, Hao; Chen, Wentao; Mai, Zhida; Lin, Xiaomian; Yang, Jianjiang; Zeng, Lihong; Pan, Yuying; Xie, Qinghui; Xu, Qingqing; Li, Xiaoxiao; Liao, Yiwen; Feng, Zhanqin; Ou, Jiangli; Qin, Xiaolin; Zheng, Heping
title: Development and Application of Cas13a-based Diagnostic Assay for Neisseria Gonorrhoeae Detection and Identification of Azithromycin Resistance
date: 2021-05-21
journal: bioRxiv
DOI: 10.1101/2021.05.20.445076
sha: ae6c36819f1e04230250f0b788fabc0581cb01cc
doc_id: 286996
cord_uid: e2v4uin7

Gonorrhea caused by Neisseria gonorrhoeae has spread world-wide. Antimicrobial-resistant strains have emerged to an alarming level to most antibiotics including to the ceftriaxone-azithromycin combination, currently recommended as first-line dual therapy. Rapid testing for antimicrobial resistance will contribute to clinical decision-making for rational drug use and will slow this trend. Herein, we developed a Cas13a-based assay for N. gonorrhoeae detection (porA target) and azithromycin resistance identification (A2059G and C2611T point mutations). We evaluated the sensitivity and specificity of this method, and 10 copies per reaction can be achieved in porA detection and C2611T identification, with no cross-reactions. Comparison of the Cas13a-based assay (porA target) with Roche Cobas 4800 assay (n=23 urine samples) revealed 100% concordance. Isolated N. gonorrhoeae strains were used to validate the identification of A2059G and C2611T resistance mutations. All tested strains (8 A2059G strains, 8 C2611T strains, and 8 wild-type strains) were successfully distinguished by our assay and verified by testing MIC for azithromycin and sequencing the 23S rRNA gene. We adopted lateral flow for the SHERLOCK assay readout, which showed a visible difference between test group and NC group results. To further evaluate the capability of our assay, we tested 27 urethral swabs from patients with urethritis for N. gonorrhoeae detection and azithromycin-resistance identification. Of these, 62.96% (17/27) strains were detected with no mutant strains and confirmed by sequencing. In conclusion, the novel Cas13a-based assay for rapid and accurate N. gonorrhoeae detection combined with azithromycin drug resistance testing is a promising assay for application in clinical practice.

critical to cure and slow the spread of N. gonorrhoeae infections [4, 5] . However, due to 48 the use and abuse of antibiotics, antimicrobial resistance (AMR) of N. gonorrhoeae has 49 emerged to all first-line therapeutic drugs used to date [4] [5] [6] [7] . In particular, AMR to 50 azithromycin and ceftriaxone currently used as first-line dual therapy has been reported 51 as a cause of treatment failure in both the United Kingdom and Australia [8, 9] , and the 52 resistance has shown a gradual increasing trend according to the gonococcal surveillance 53 program data from Europe and the United States [10, 11] . There is, therefore, a need for 54 clinicians to rapidly acquire resistance data for antibiotics, which could help manage 55 10 Zhang et al. [19] . Briefly, the assay was carried out in a 25 L reaction volume consisting 174 of 40 mM Tris-HCl (pH7.5), 9 mM MgCl 2 , 1 mM rNTPs (New England Biolabs; 175 N0466L) , 50 U RNase inhibitor (New England Biolabs; M0314L), 37.5 U T7 RNA 176

Polymerase (New England Biolabs; M0251L), 225 nM crRNA, 45nM purified LwCas13a, 177 125 nM RNA reporter (5'-6FAM-UUUUU-BHQ1-3' as described by Gootenberg et 178 al.[36] ), and 1.25 L RPA reaction solution was added to the above mixture. The reaction 179 mixture was allowed to incubate at 37°C for 3 h in a 96-Well Half-Area Microplate 180 (Corning; CLS3694-100EA). Fluorescence emission (excitation 490 nm/detection 520 181 nm) was tested every 5 min. 182

The lateral flow dipstick (Milenia Biotec; MGHD 1) was used to acquire the results of 184 Cas13a collateral cleavage, which was based on the cleavage of the FITC-RNA-Biotin 185 reporter. It basically replaces the RNA reporter used in the system described above, with 186 a new RNA reporter (5'-FITC-UUUUUUUUUUUUUU-Biotin-3' described by Myhrvold 187 et al. [37] ), and then was subjected to the same process. Subsequently, a 20 μL volume of 188 Cas13a detection solution was added and the reaction mixture was incubated at 37℃ for 189 3h in 80 L dipstick buffer, with thorough mixing. A lateral flow dipstick was inserted 190 into the mixture to obtain the results. 191

The 23s rRNA gene of N. gonorrhoeae, which contains the A2059G and C2611T point 193 11 mutations, was amplified using the Q5 High-Fidelity DNA Polymerases (New England 194 Biolabs; M0492S). In a 25 μL reaction volume, comprised of 12.5 μL of 2x Master Mix, 195 1.25 μL of each 10 μM primer, 1 μL of input, and 8 μL ddH2O. PCR was performed as 196 follows: initial denaturation at 98℃ for 30 s, then 35 cycles of 98℃ for 10 s, 64℃ for 20 197 s, and 72℃ for 30 s, followed by 72℃ for 2 min. PCR products were verified by Sanger 198 sequencing (Sangon Biotech). The results of sequencing were blasted in SnapGene 4.1.9. 199 PCR primers are available in Table S2 . The SHERLOCK assay was performed as established by Zhang et al. [19] , and combined 209 RPA and Cas13a to create an isothermal detection system. The target sequence was 210 amplified by RPA, and the T7 promoter was appended to the front of the RPA product as 211 the template to be used to initiate subsequent RNA transcription. Synthetic crRNA guided 212 12 the Cas13a protein to recognize the specific target and enable its RNA cleavage and 213 collateral cleavage activities ( Fig. 1) [18] . For N. gonorrhoeae detection, we selected 214 porA as the target as was frequently used to identity N. gonorrhoeae in other methods [17, 215 38-40] . Based on the characteristics of its single-base resolution, we constructed two 216 crRNA sequences that could identify A2059G and C2611T separately. Although a single 217 synthetic mismatch of crRNA pairing to the target enabled the assay to identify the 218 A2059G mutation (Fig. S1A) , this design failed to achieve the identical results for the 219 detection of the C2611T mutation. Thus, we introduced one more synthetic mismatch of 220 crRNA for C2611T mutation testing ( Fig. S1B-D) . Both crRNA designs were 221 successfully utilized for 23s rRNA mutations detection. 222

To determine the sensitivity of the N. gonorrhoeae detection and azithromycin resistance 224 identification assay, we prepared serial dilutions of dsDNA template, ranging from 10 0 225 copy/μL to 10 5 copies/μL ( Table 1) . The RPA step included the addition of 1 μL input of 226 dsDNA template, which was then transferred to the mixture for Cas13a detection. 227

Detection of 10 copies/μL was achieved for porA and C2611T identification (Fig. S2A , 228 S2E). For A2059G identification, the detection limit was an order of magnitude lower, at 229 about 10 2 copies/μL (Fig. S2C) . We further evaluated the specificity of the assay using a 230 panel of urogenital tract pathogenic bacteria (n=12) and Neisseria meningitides ( Table 2) . 231 13 No cross-reactivity was observed for both N. gonorrhoeae detection and azithromycin 232 resistance identification (Fig. S2B, S2D, S2F ). All RPA primers and crRNA sequences 233 had been confirmed by BLAST before we tested its specificity. The SHERLOCK exhibits 234 robust capability for N. gonorrhoeae detection and azithromycin resistance identification. 235

Twenty-three clinical urine samples with low concentrations previously tested using the 237 standard procedure of the Roche Cobas 4800 NG/CT test were used to validate the 238 performance of SHERLOCK. DNA was extracted from urine samples after centrifugation 239 and a maximum volume of 6.6 μL of DNA was amplified by RPA, followed by Cas13a 240 detection. The method was repeated using 3 technical replicates and the fluorescence 241 signal of each sample was normalized against the negative controls. Using this method, a 242 total of 12 of 23 positive samples were detected, showing a 100% coincidence rate with 243 the Roche assay (Fig. 2) . The fluorescence signals of 3 samples (samples 3, 6, 10) were 244 weaker than other specimens, but still could be distinguished with the negative control. 245

A2059G mutant strains (n=8), C2611T mutant strains (n=8), and wild-type strains (n=8) 247 isolated from clinical specimens were used to validate Cas13a-based SNPs detection. We 248 measured the MICs of azithromycin in each strain and sequenced their 23S rRNA gene 249 ( Fig. 3, 4) . Strains containing either the A2059G or C2611T point mutation were more 250 14 likely to be a high-level azithromycin-resistant strain. Of the 8 A2059G isolated strains, 251 all 8 strains had MICs of >1 mg/L, and 5 A2059G strains had MICs ≥512 mg/L. Of the 8 252 C2611T isolated strains, 7 had MICs of >1 mg/L and 3 C2611T strains had MICs ≥512 253 mg/L. Compared with mutant strains, the wild-type strains possessed lower MICs, 254 corresponding to ≤1 mg/L in 7 strains and the MIC of the remaining strain was 4 mg/L, 255 which was above the average MICs of all mutant strains. We extracted DNA of all 256 mutant and wild-type strains to validate this Cas13a-based assay. Paired with A2059G 257 crRNA or C2611T crRNA, this assay successfully differentiated all 8 A2059G mutant 258 strains and 8 C2611T strains from 8 wild-type strains (Fig. 3, 4) . Sixteen strains 259 harboring A2059G and C2611T mutations were identified by exhibiting a higher 260 fluorescence intensity than wild-type strains by our assay (Fig. S3A, S3B) . Thus, we 261 successfully applied Cas13a based assay in azithromycin resistance identification. 262

We applied lateral flow to provide a more convenient readout tool. The FAM and BHQ1 264 markers in the RNA reporter were replaced by FITC and Biotin. Compared to the 265 fluorescent intensity detection, the lateral flow is inserted directly into the reaction liquid 266 instead of using a specific device or instrument for the readout of results. The lateral flow 267 contains a control band and a test band. Generally, a positive test will show only one test 268 band or two bands (test band and control band), due to its different cleavage efficiency 269 15 which will result in varying amounts of cleaved RNA reporter captured by the antibody 270 in the test band. We tested lateral flow for porA detection, A2059G identification, and 271 C2611T identification separately ( Fig. 5A-C) . Three positive groups generated visual 272 signals in the test bands, while all wild-type groups and the no-input group only showed a 273 single control band. 274

To confirm the efficacy of the assay in clinical specimens. We collected 27 urethral 277 swabs from patients with urethritis requiring differential diagnosis for potential 278 gonococcal infection and to determine whether azithromycin was still effective and this 279 information is important for the clinician. We extracted DNA from urethral swabs 280 directly and then tested all 27 samples with the SHERLOCK assay (Fig. 6) . Overall, 281 62.96% (17/27) samples showed porA positivity, and the fluorescence intensity of 17 282 samples was higher than that of the negative samples and the no-input group. We further 283 tested for azithromycin resistance with A2059G crRNA and C2611T crRNA, and no 284 mutant strain was discovered in the 17 porA positive samples. For 27 specimens, we 285 sequenced the 23S rRNA gene and the results showed a 100% coincidence rate with our 286 assay (Table S1 ). Sequencing data demonstrated that 17 porA positive samples were 287 wild-type strains and no signals were detected in 10 porA negative samples. Antimicrobial resistance is the largest concern in the treatment of gonorrhea, with its 291 decreasing susceptibility to antibiotics used in previous or current treatment approaches 292

[5]. Treatment failures have been reported to the current first-line drug combination of 293 azithromycin and ceftriaxone [8, 9] , and represent a significant treatment challenge to 294 clinicians. Rapid information regarding antimicrobial resistance would be beneficial to 295 rational drug use in the clinic and would slow this growing trend. Traditional 296 culture-based drug resistance methods have been widely used in the clinical laboratory 297 and are of great importance in antimicrobial resistance surveillance, although the testing 298 is time-consuming and is unable to meet clinical requirements rapidly [41] . Newly 299

NAATs-based assays may overcome some of the disadvantages of culture-based methods 300 and can be applied to identify antimicrobial resistance associated mutations 301 simultaneously. Current molecular methods to identify mutations are mainly based on 302 specially designed probes, HRM analysis, and mass spectrometry [13, 16, 17] . Compared 303 with culture-based antimicrobial resistance detection methods, these methods effectively 304 shorten assay times, but still present several limitations. If the alteration in gene has little 305 impact on Tm values or GC content, for example, C to G variations, the method would be 306 invalidated, moreover, short insertions or deletions may make the method unreliable [17, 307 17 42]. Furthermore, these assays are essentially PCR-based and require PCR amplification 308 instruments coupled with other detection equipment, which limits the convenience and 309 flexibility of the assay. Thus, the development of an assay with ultra-high resolution is 310 desired for distinguishing mutant from the wild-type. 311 312 CRISPR/Cas molecular diagnostics have been developed and applied for testing various 313 organisms, including SARS-CoV-2, HPV, Zika virus, Dengue virus, Ebola virus, and 314 plasmodium [20, 21, 37, 43, 44] . Benefiting from its high specificity, sensitivity, and 315 ability to identify SNP with the isothermal process, we have adopted a Cas13a-based 316 strategy for N. gonorrhoeae detection and antimicrobial resistance identification in this 317 study. The SHERLOCK contains two nucleic acid amplification steps: DNA 318 amplification by recombinase polymerase amplification and RNA amplification by T7 319 transcription. With double signal amplification cycles, this strategy allowed to detect low 320 levels of N. gonorrhoeae. Cross-reactivity is the major concern of currently developed 321 diagnostic methods [12] . Attempts have been made to introduce two targets to uniquely 322 identify a species, but this approach complicates the assay [13, [15] [16] [17] . The combination 323 of specially-designed RPA primers and crRNA makes the whole reaction extremely 324 specific. As expected, our assay exhibited high specificity in testing a panel of 325 non-gonococcal bacteria. In addition, our Cas13a-based assay showed an excellent 326 18 concordance rate with the Roche Cobas assay currently used for clinical urine samples. 327

With regard to SNPs detection, assay has been developed that exploits CRISPR/Cas13a 328 technology to recognize single point mutations [19, 36, 45] . Our Cas13a-based assay 329 achieved a sensitivity of 10 copies per reaction, which is more sensitive than previous 330 HRM-based assays [46] . The diagnostic capability of the Cas13a-based assay has also 331 been examined by testing clinical isolates harboring the SNP mutation. This isothermal 332 assay which relies on a reaction temperature of 37℃ over the entire process without 333 complex equipment has a great potential to be applied as a POCT device. 334 335 Azithromycin is a widely used macrolide antimicrobial agent and primarily acts on 336 domain Ⅴ of the 23S rRNA gene. Previous studies and our observation have revealed that 337 N. gonorrhoeae strains harboring A2059G and C2611T mutations in the 23s rRNA gene 338 are strongly associated with high-level azithromycin resistance [28, 32, 33, [47] [48] [49] . 339

Sixteen N. gonorrhoeae isolates containing 23s rRNA mutations were utilized to evaluate 340 the performance of our assay. The results showed that our Cas13a-based assay could 341 provide drug resistance information in real-time. We also tested a small number of 342 urethral swabs collected from the clinic in Guangzhou, China. In porA-positive(17/27) 343 urethral swabs from patients with urethritis, no 23s rRNA mutant was identified, which is 344 consistent with previous reports, indicating that the high-level azithromycin-resistant N. 345 19 gonorrhoeae has not wildly spread in Guangzhou, China [14, 16, 17] S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 Minimum inhibitory concentration(mg/L) S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 

Direct real-time PCR-based detection 420 of Neisseria gonorrhoeae 23S rRNA mutations associated with azithromycin 421 resistance

Mismatch Amplification Mutation 423

Assay-Based Real-Time PCR for Rapid Detection of Neisseria gonorrhoeae and 424

Antimicrobial Resistance Determinants in Clinical Specimens

Multiplex real-time PCR assays for the 427 prediction of cephalosporin, ciprofloxacin and azithromycin antimicrobial 428 susceptibility of positive Neisseria gonorrhoeae nucleic acid amplification test 429 samples

Multiplex High-Resolution Melting Assay for 431 Simultaneous Identification of Molecular Markers Associated with 432 Extended-Spectrum Cephalosporins and Azithromycin Resistance in Neisseria 433 gonorrhoeae

C2c2 is a single-component 435 programmable RNA-guided RNA-targeting CRISPR effector

Nucleic acid detection with 438 CRISPR-Cas13a/C2c2

Ultrasensitive CRISPR-based diagnostic for 440 field-applicable detection of Plasmodium species in symptomatic and 441 asymptomatic malaria

Deployable CRISPR-Cas13a 443 diagnostic tools to detect and report Ebola and Lassa virus cases in real-time

Update to CDC's Sexually transmitted diseases treatment guidelines, 2010: oral 448 cephalosporins no longer a recommended treatment for gonococcal infections

Emergence and dissemination 451 of three mild outbreaks of Neisseria gonorrhoeae with high-level resistance to 452 azithromycin in Barcelona

Characterization of two Neisseria 457 gonorrhoeae strains with high-level azithromycin resistance

First Case of High-Level Azithromycin-Resistant Neisseria gonorrhoeae in North 461

Emergence and Spread of Neisseria 463 gonorrhoeae Strains with High-Level Resistance to Azithromycin in Taiwan from 464

Neisseria gonorrhoeae in the USA in 2017: a genomic analysis of surveillance 467 data

The mosaic mtr locus as major genetic 469 determinant of azithromycin resistance of Neisseria gonorrhoeae

Mutation in 23S rRNA associated with 472 macrolide resistance in Neisseria gonorrhoeae

High-level azithromycin resistance occurs in 475

Neisseria gonorrhoeae as a result of a single point mutation in the 23S rRNA 476 genes

Molecular pathways to 478 high-level azithromycin resistance in Neisseria gonorrhoeae

Manual for the laboratory identification and antimicrobial 481 susceptibility testing of bacterial pathogens of public health concern in the 482 developing word

SHERLOCK: 486 nucleic acid detection with CRISPR nucleases

Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a, and 490 Csm6

Field-deployable viral diagnostics 492 using CRISPR-Cas13

Molecular Assay for Detection of 494

Genetic Markers Associated with Decreased Susceptibility to Cephalosporins in 495 Neisseria gonorrhoeae

Evaluation of six nucleic 497

 Table  647 35