View PDF Version

 

High-contrast imaging of cellular non-repetitive drug-resistant genes via in situ dead Cas12a-labeled PCR†

Ruijie Deng^a, Xinlei Zhang^a, Jijuan Cao^c, Xinmiao Liu^a, Yong Zhang^a, Feng Wang*^b and Xuhan Xia*^a
aCollege of Biomass Science and Engineering, Sichuan University, Chengdu 610065, China. E-mail: xxh23@scu.edu.cn
^bShenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, Guangdong, China. E-mail: f.wang@siat.ac.cn
^cKey Laboratory of Biotechnology and Bioresources Utilization of Ministry of Education, Dalian Minzu University, Dalian, Liaoning 116600, China

First published on 28th August 2024

Currently, fluorescence in situ hybridization (FISH) is widely applied for RNA imaging at single-cell resolution.^11,12 FISH can provide transcriptomic information and cellular heterogeneity by hybridizing mRNAs with multiple fluorophore-labeled nucleic acid probes.^13–15 However, it is less sensitive for labeling double-stranded genomic DNA.¹⁶ Attributed to the evolvement of nucleic acid amplification techniques, the detection sensitivity of DNA can be enhanced.¹⁷ Among them, polymerase chain reaction (PCR), as the classic gold standard method of nucleic acid amplification, can also be successfully applied to develop in situ PCR-based methods.^18–20 However, the drawback of combining FISH with in situ PCR lies in the difficulty of labeling double-stranded DNA amplicons. It can label amplicons only when subjected to high-temperature conditions or when utilizing base-modified DNA strands, like locked nucleic acids.²¹ Additionally, the inevitable non-specific amplification of PCR inside cells results in a reduced signal gain and a high background in FISH-based imaging. Hence, there is a demand to develop an efficient approach for tagging PCR amplicons.

In this work, we employed clustered regularly interspaced short palindromic repeat (CRISPR) arrays and their associated Cas proteins (CRISPR–Cas) for tracking and labeling cellular genes.^22–24 We developed a dead Cas12a (dCas12a)-labeled PCR (CasPCR) assay for high-contrast imaging of the drug-resistant gene oqxB of Salmonella enterica (S. enterica), which is associated with quinolone-resistance.^25,26 Quinolone is among the most commonly used antibiotics for treating bacterial infections in both animals and humans.^27,28 In the single-cell imaging strategy, the signal output is realized by binding the dCas12a fusion with a green fluorescent protein (GFP).^29,30 The involvement of dCas12a–GFP confers a dual-recognition process of the in situ PCR amplicons, thus avoiding the non-specific amplification and improving the signal-to-background ratio for detecting genes. The CasPCR assay allows the sensitive quantification of quinolone-resistant S. enterica strains and identification of quinolone-resistant strains collected from poultry farm isolated Salmonella spp.

The genomic DNA imaging was proceeded via in situ PCR and dCas12a–GFP recognition (Fig. 1). Although both dCas12a and dCas9 can target genomic DNA, Cas12a requires only a shorter CRISPR RNA (crRNA) (43 nucleotides) with a simpler structure compared to Cas9. Therefore, we chose dCas12a to label the PCR amplicon. First, the fusion protein dCas12a–GFP was expressed. We constructed a plasmid capable of expressing dCas12a–GFP by recombining plasmids that express dCas12a and GFP (Fig. 1A). For the gene imaging procedure, we designed a set of probes for PCR, including forward primer and reverse primer. In situ PCR was used to achieve signal amplification for the oqxB gene, which is associated with the resistance-nodulation-division efflux pump commonly found in Gram-negative bacteria.²⁶ After PCR, dCas12-GFP can recognize and light up the amplicon, enabling in situ imaging of quinolone-resistant strains. The inactivated dCas12 eliminates its cis-cleavage activity, facilitating target binding (Fig. 1B). The coupling of in situ PCR and dCas12a endowed a dual-recognition process of the target, thus improving specificity and avoiding the use of multiple fluorescent probes.

	Fig. 1 Working principle of the CasPCR assay for visualizing the oqxB gene inside bacterial cells. (A) The process of expressing dCas12a–GFP through plasmid recombination. (B) Single-cell imaging of quinolone-resistant S. enterica via in situ PCR and dCas12a–GFP recognition.

Electrophoresis analysis and cellular fluorescence imaging were employed to validate the working principle of the CasPCR assay for cellular oqxB gene detection. First, the oqxB-positive strain isolated from the poultry farm was selected to perform a minimum inhibitory concentration (MIC) test and growth curves to ciprofloxacin, a frequently used quinolone antibiotic (Fig. S3, ESI†). The results indicated that this oqxB-positive strain was inhibited from growing at a ciprofloxacin concentration of 16 μg mL⁻¹, confirming its resistance to quinolone. Then, we performed electrophoresis analysis to explore the capacity for amplification of the oqxB gene (Fig. 2A). The bands (lane 1 and lane 2) with quinolone-sensitive strain genes, forward primer, and reverse primer did not show amplicons at the target band. In contrast, the bands containing quinolone-resistant strain genes and primers showed an amplification product with 150 bp (lane 3 and lane 4), consistent with the designed amplicon length. In addition, we also identified the oqxB gene in quinolone-resistant S. enterica by Sanger sequencing and qPCR (Table S3 and Fig. S4, ESI†). After confirming that PCR could successfully amplify oqxB, cellular fluorescence imaging of quinolone-resistant and quinolone-sensitive strains was conducted using the CasPCR assay (Fig. 2B). For the quinolone-resistant strain, the bacterial cells were lighted-up and exhibited noticeable fluorescence. The outcome showed that the dCas12a–GFP assay successfully recognized the amplicons. In comparison, significant fluorescence decreases of the bacterial cells occurred when detecting the quinolone-sensitive strain. We quantified the gray values of each bacterial cell, with each dot corresponding to the mean fluorescence intensity of an individual cell (Fig. 2C). The mean fluorescence intensities for quinolone-resistant and quinolone-sensitive strains were 53.48 and 22.01, respectively. The fluorescence intensity in the quinolone-sensitive control group decreased by 58.9%, with a signal-to-background ratio of 2.43-fold. In addition, cellular fluorescence imaging was also performed for control groups lacking either crRNA or dCas12a–GFP (Fig. S5, ESI†). It has been observed that fluorescence was weak and the fluorescence intensity profile further confirmed that the dCas12a–GFP used in this method did not trigger non-specific binding within the cells. The results proved that the CasPCR assay can detect the oqxB gene inside cells, serving as an in situ analysis tool for quinolone-resistant strain analysis.

	Fig. 2 Demonstration of the CasPCR assay for quinolone-resistant strain detection. (A) Electrophoretic analysis of the PCR process for oqxB gene detection. (B) Example fluorescence images of quinolone-resistant and quinolone-sensitive strains measured by CasPCR assay detection. (C) The fluorescence intensity of quinolone-resistant and quinolone-sensitive strains in panel (B).

Considering that the signal output of the CasPCR assay lies in tagging amplicons via dCas12a–GFP and in situ PCR reaction, both procedures were optimized. First, we explored the design of dCas12a–GFP to the amplicons. dCas12a–GFP serves a key role in discriminating oqxB, and its performance is dependent on the locations of the protospacer adjacent motif (PAM, 5′-TTN-3′) recognized by crRNA. Five locations of PAM on the amplicon sequence were used to design crRNAs (Fig. 3A). Cellular fluorescence imaging was employed to select the optimal PAM locations. We observed that at PAM positions 1 to 3, quinolone-resistant and sensitive strains could be clearly distinguished in the imaging results (Fig. 3B–D). When PAM positions 4 and 5 were selected, the performance in detecting quinolone-resistant strains significantly decreased (Fig. 3E and F). The mean fluorescence intensities of quinolone-resistant bacteria were 42.6, 35.7, and 67.6, respectively. The signal-to-background ratio for distinguishing between quinolone-resistant and sensitive strains increased from 3.1-fold to 5.2-fold. Hence, the PAM position 3 was selected as the optimal dCas12a–GFP recognition site.

	Fig. 3 Design of dCas12a/crRNA. (A) Illustration of four crRNAs targeting different locations of the in situ PCR amplicon. (B) Example images and the cellular fluorescence intensity for crRNA-1. (C) Example images and the cellular fluorescence intensity for crRNA-2. (D) Example images and the cellular fluorescence intensity for crRNA-3. (E) Example images and the cellular fluorescence intensity for crRNA-4. (F) Example images and the cellular fluorescence intensity for crRNA-5.

Then, we designed 10 sets of primers along the oqxB gene, named P1 to P10, and performed in situ imaging tests. The imaging outcomes indicated that different PCR primer recognition sites significantly affect amplification efficiency (Fig. 4A). The results of cellular imaging showed that the quinolone-resistant bacteria could not be lighted up in some groups, indicating that the amplification failed. However, there was a significant fluorescence difference between quinolone-resistant and sensitive strains in the P3, P5, P6, and P7 groups. To obtain the optimal primers, we conducted cellular fluorescence intensity measurement (Fig. 4B) and calculated the mean fluorescence intensity of about 50 bacterial cells (Fig. 4C). When choosing P3 primers, the mean fluorescence intensity of the quinolone-resistant strain and quinolone-sensitive strain was 52.7 and 4.3, respectively, exhibiting a signal-to-background ratio up to 13.3-fold.

	Fig. 4 Investigation of PCR primers for cellular gene imaging. (A) Example images of quinolone-resistant and quinolone-sensitive S. enterica using 10 sets of designed primers. (B) The fluorescence intensity of each bacterial cell for imaging quinolone-resistant and quinolone-sensitive S. enterica using different primers. (C) The mean fluorescence intensity in (B).

Attributed to the CasPCR assay's ability of profiling the oqxB gene with single-cell resolution, we evaluated the potential to quantify quinolone-resistant S. enterica in the mixed strains. In the mixture, a series of different proportions of quinolone-resistant S. enterica were prepared, specifically, 0%, 20%, 40%, 60%, 80%, and 100%. We imaged the samples using the optimal primers and crRNAs and measured the mean fluorescence intensity of each bacterial cell. The detected proportions were 0%, 18.4%, 43.1%, 61.4%, 80.5%, and 100%, and were close to the concentrations of quinolone-resistant S. enterica added in the mixed strains (Fig. 5A). The result indicated that CasPCR assay can precisely profile heterogeneity information related to genomic genes.

	Fig. 5 Quantification performance of the CasPCR assay. (A) Example images and cellular fluorescence intensity of the quantification performance test of quinolone-resistant S. enterica in the mixed strains using the CasPCR assay. (B) Imaging of S. enterica isolated from a poultry farm. (C) PCR and electrophoresis analysis for oqxB gene of S. enterica isolated from a poultry farm. (D) The receiver operating characteristic (ROC) curve of in situ imaging and PCR results.

Finally, the CasPCR assay was explored for analyzing the quinolone-resistant strains of S. enterica isolated from poultry farms. We employed the assay to test 20 strains via in situ imaging and observed that 9 strains of them were lighted up, indicating their potential as quinolone-resistant S. enterica (Fig. 5B). To verify the reliability of these results, we extracted DNA from all strains and performed PCR and electrophoretic tests (Fig. 5C). The result further shows that the quinolone-resistant strains detected by cellular imaging contained the oqxB gene. The receiver operating characteristic (ROC) curve demonstrated the correlation between the CasPCR assay and PCR. Both the sensitivity and specificity values were 100%, indicating an agreement between these two methods. In addition, we performed a MIC test on these oqxB-positive strains (Fig. S6, ESI†), and found that the MIC of 9 strains ranged from 8 to 128 μg mL⁻¹, indicating that all the strains were quinolone-resistant. The successful identification of quinolone-resistant S. enterica isolated from poultry farms proves the feasibility of the assay to profile the correlation of cellular genotype and phenotype.

In summary, we developed an assay for in situ imaging genes via dCas12a-labeled PCR, termed CasPCR. The dual-recognition of PCR primers and Cas/crRNA confers a high specificity. By the use of the fusion of dCas12a and GFP, the CasPCR assay enables high-contrast imaging of cellular genes, allowing the precise profiling of the cellular heterogeneity of pathogenic bacteria. Collectively, we leveraged the optimized CasPCR assay to light up the oqxB gene, facilitating precise quantification of quinolone-resistant S. enterica in mixed bacteria. In addition, it enabled the detection of quinolone-resistant S. enterica isolated from poultry farms. This single-cell imaging tool will help us to explore cellular functions and heterogeneity in populations.

R. Deng: conceptualization, formal analysis, investigation, methodology, visualization, funding acquisition and writing – original draft; X. Zhang: data curation, investigation and validation; J. Cao: investigation and validation; X. Liu: project administration; Y. Zhang: methodology and project administration; F. Wang: formal analysis and validation; X. Xia: conceptualization, funding acquisition, project administration, resources, supervision, and writing – review & editing.

This work was financially supported by the National Key R&D Program of China (No. 2022YFF1103000), Natural Science Foundation of Sichuan Province (2024NSFSC1246), Open Fund of Key Laboratory of Biotechnology and Bioresources Utilization (Dalian Minzu University), and Ministry of Education (No. 2024003).

Data availability

Conflicts of interest

1. H. Gao, K. Zhang, X. Teng and J. Li, Trac-Tren. Anal. Chem., 2019, 121, 115700 CrossRef CAS .
2. S. Hocine, P. Raymond, D. Zenklusen, J. A. Chao and R. H. Singer, Nat. Methods, 2013, 10, 119–121 CrossRef CAS PubMed .
3. R. Deng, K. Zhang and J. Li, Acc. Chem. Res., 2017, 50, 1059–1068 CrossRef CAS PubMed .
4. X. Xia, F. Li, Y. Zhang, Y. Zhu, S. Song, G. He, J. Zhang, Q. He and R. Deng, Anal. Chem., 2023, 95, 10163–10171 CrossRef CAS PubMed .
5. X. Zhu, K. Zhang, X. Teng, X. Xia, R. Deng and J. Li, Chem, 2023, 9, 1094–1117 CAS .
6. N. Martínez, M. C. Martín, A. Herrero, M. Fernández, M. A. Alvarez and V. Ladero, Trends Food Sci. Tech., 2011, 22, 367–376 CrossRef .
7. J. Glökler, T. S. Lim, J. Ida and M. Frohme, Crit. Rev. Biochem. Mol., 2021, 56, 543–586 CrossRef PubMed .
8. Y. Zhao, F. Chen, Q. Li, L. Wang and C. Fan, Chem. Rev., 2015, 115, 12491–12545 CrossRef CAS PubMed .
9. K. Zhang, R. Deng, H. Gao, X. Teng and J. Li, Chem. Soc. Rev., 2020, 49, 1932–1954 RSC .
10. Z. Qing, J. Xu, J. Hu, J. Zheng, L. He, Z. Zou, S. Yang, W. Tan and R. Yang, Angew. Chem., Int. Ed., 2019, 58, 11574–11585 CrossRef CAS PubMed .
11. I. Dardani, B. L. Emert, Y. Goyal, C. L. Jiang, A. Kaur, J. Lee, S. H. Rouhanifard, G. M. Alicea, M. E. Fane, M. Xiao, M. Herlyn, A. T. Weeraratna and A. Raj, Nat. Methods, 2022, 19, 1403–1410 CrossRef CAS PubMed .
12. R. Amann and B. M. Fuchs, Nat. Rev. Microbiol., 2008, 6, 339–348 CrossRef CAS PubMed .
13. A. Safieddine, E. Coleno, F. Lionneton, A.-M. Traboulsi, S. Salloum, C.-H. Lecellier, T. Gostan, V. Georget, C. Hassen-Khodja, A. Imbert, F. Mueller, T. Walter, M. Peter and E. Bertrand, Nat. Protoc., 2023, 18, 157–187 CrossRef CAS PubMed .
14. A. Raj, P. van den Bogaard, S. A. Rifkin, A. van Oudenaarden and S. Tyagi, Nat. Methods, 2008, 5, 877–879 CrossRef CAS PubMed .
15. R. Rocha, J. M. Sousa, L. Cerqueira, M. J. Vieira, C. Almeida and N. F. Azevedo, Food Microbiol., 2019, 80, 1–8 CrossRef CAS PubMed .
16. H. Frickmann, A. E. Zautner, A. Moter, J. Kikhney, R. M. Hagen, H. Stender and S. Poppert, Crit. Rev. Microbiol., 2017, 43, 263–293 CrossRef CAS PubMed .
17. X. Xia, H. Yang, J. Cao, J. Zhang, Q. He and R. Deng, TrAC, Trends Anal. Chem., 2022, 153, 116641 CrossRef CAS .
18. M. Chen, X. Lan, L. Zhu, P. Ru, W. Xu and H. Liu, Foods, 2022, 11, 2675 CrossRef CAS PubMed .
19. X. Liu, H. Yang, J. Liu, K. Liu, L. Jin, Y. Zhang, M. R. Khan, K. Zhong, J. Cao, Q. He, X. Xia and R. Deng, Anal. Chem., 2024, 96, 2032–2040 CrossRef CAS PubMed .
20. O. Bagasra, Nat. Protoc., 2007, 2, 2782–2795 CrossRef CAS PubMed .
21. C. Almeida, J. M. Sousa, R. Rocha, L. Cerqueira, S. Fanning, N. F. Azevedo and M. J. Vieira, Appl. Environ. Microbiol., 2013, 79, 6293–6300 CrossRef CAS PubMed .
22. A. V. Anzalone, L. W. Koblan and D. R. Liu, Nat. Biotechnol., 2020, 38, 824–844 CrossRef CAS PubMed .
23. K. Zhang, R. Deng, X. Teng, Y. Li, Y. Sun, X. Ren and J. Li, J. Am. Chem. Soc., 2018, 140, 11293–11301 CrossRef CAS PubMed .
24. M. Wang, K. Chen, Q. Wu, R. Peng, R. Zhang and J. Li, Anal. Chem., 2020, 92, 2468–2475 CrossRef CAS PubMed .
25. X. Xia, B. Ma, W. Zhou, H. Yang, C. Lei, Y. Dong, J. Zhang, Q. He and R. Deng, Chem. Eng. J., 2023, 451, 138355 CrossRef CAS .
26. N. Bharatham, P. Bhowmik, M. Aoki, U. Okada, S. Sharma, E. Yamashita, A. P. Shanbhag, S. Rajagopal, T. Thomas, M. Sarma, R. Narjari, S. Nagaraj, V. Ramachandran, N. Katagihallimath, S. Datta and S. Murakami, Nat. Commun., 2021, 12, 5400 CrossRef CAS PubMed .
27. K. J. Aldred, R. J. Kerns and N. Osheroff, Biochemistry, 2014, 53, 1565–1574 CrossRef CAS PubMed .
28. A. Shaheen, A. Tariq, M. Iqbal, O. Mirza, A. Haque, T. Walz and M. Rahman, Antibiotics, 2021, 10, 1455 CrossRef CAS PubMed .
29. J.-D. Pédelacq, S. Cabantous, T. Tran, T. C. Terwilliger and G. S. Waldo, Nat. Biotechnol., 2006, 24, 79–88 CrossRef PubMed .
30. S. J. Remington, Protein Sci., 2011, 20, 1509–1519 CrossRef CAS PubMed .