The bacterium Klebsiella pneumoniae (Kp), belonging to the Enterobacteriaceae family (Tominaga, 2018), was a Gram-negative conditional pathogen that commonly colonized various sites such as the oral cavity, skin, respiratory tract, urinary tract, and intestines. Kp infection predominantly occurred in people with low immunity (Podschun and Ullmann, 1998; Li et al., 2014), was a main reason for hospital-acquired infection (Pitout et al., 2015), and could result in bacteremia, stroke-associated pneumonia, sepsis, urinary tract infections, liver abscesses, and so on (Wang et al., 1998; Chen et al., 2013; Yao et al., 2015). Moreover, the presence of a severe Kp infection often indicated a significant risk of mortality, with reported mortality rates ranging from 16% to 40% (Namikawa et al., 2019). With the use of antimicrobial drugs, Kp resistance was increasing and the emergence of multi-drug-resistant Kp, especially carbapenem-resistant Kp, had made infection with Kp a major global public health concern (Zhang et al., 2015). Therefore, the early detection of Kp was very important.

Current detection methods for Kp mainly include bacterial culture, biochemical identification, immunological assays, and molecular biology technologies. Among these methods, the culture method was considered the gold standard, but it required at least 24 hours of incubation time (Kloss et al., 2013). Some Enterobacteriaceae had a high degree of similarity, and identification and differentiation of Enterobacteriaceae based on biochemical characterization methods were prone to produce erroneous results (Zamkan and Mohamed, 2018). Matrix-assisted laser desorption ionization time-of-flight mass spectrometry was also commonly used for the identification of Kp (Huang et al., 2022) but still required bacterial culture and a long diagnostic time. Molecular diagnostic methods based on nucleic acid detection obviated the need for bacterial culture, thereby being suitable for early clinical diagnosis of Kp. The commonly used nucleic acid detection methods for Kp included polymerase chain reaction (PCR), isothermal amplification, etc. Compared to traditional detection methods based on bacterial culture, PCR significantly reduced the detection time. However, the dependence of PCR on sophisticated and expensive instruments restricted its application for point-of-care test and in resource-poor areas (Dai et al., 2019), and the detection time of PCR was still longer than that of isothermal amplification. The isothermal amplification method, characterized by its simple instrumentation and rapid reaction time, demonstrated enormous potential to replace PCR.

Recombinase polymerase amplification (RPA), a novel isothermal amplification technology, exhibited distinct characteristics including fast reaction speed, low reaction temperature (37–42°C), and low necessity for sophisticated instrumentation (Dai et al., 2019).Two important enzymes were used in RPA: recombinase and polymerase (Piepenburg et al., 2006). The recombinase could release DNA double strands during replication at low temperature, while polymerase could synthesize new DNA strands (Daher et al., 2016). The principle of RPA was shown in Figure 1A . Compared to other isothermal amplification method such as loop-mediated isothermal amplification (LAMP), the important advantages of RPA included its high tolerance to inhibitors in complex samples and easier primer design, whereas the LAMP principle was complex, required multiple pairs of primers, and had a high false positive rate (Dong et al., 2015). Therefore, RPA had attracted a great deal of attention from researchers.

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) was a revolutionary gene editing technology first discovered in the immune system of bacteria (Jansen et al., 2002). The CRISPR system demonstrated remarkable attributes of high specificity, sensitivity, and programmability, and was widely employed for nucleic acid detection. Cas12a was an important enzyme of the CRISPR system (Chen et al., 2018). In 2018, Chen et al. reported that CRISPR/Cas12a could precisely cis-cleave target sequences under CRISPR RNA (crRNA) guidance, non-specific trans-cleave ssDNA reporter to amplify the signal of RPA, and successfully used to detect papillomavirus in human patient samples (Chen et al., 2018). The SHERLOCK platform developed by Gootenberg et al. integrated RPA with CRISPR/Cas13a, and was successfully used to detect Zika and Dengue viruses as well as the genotype of human DNA (Gootenberg et al., 2017). Moreover, several studies had reported that pathogenic microorganisms including SARS-CoV-2 could be detected using RPA combined with CRISPR/Cas12a (Lin et al., 2022; Mao et al., 2022; Xiong et al., 2022; Cao et al., 2023). However, there were no reports about using RPA integrated with CRISPR/Cas12a to detect Kp.

In this study, three pairs of RPA primers were designed to target the rcsA of Kp. By optimizing the RPA and the CRISPR/Cas12a detection system, a platform for Kp detection using fluorescence detection (FD) and lateral flow test strips (LFTS) as the signal readout methods were successfully established, and was further compared with culture and qPCR methods. It was expected to address the clinical need for rapidly and accurately detecting Kp.

Kp was a frequently occurring pathogen that was closely associated with human health and hospital-acquired infection (Pitout et al., 2015) It had become a major global public health threat due to its high level of antimicrobial resistance (Wyres and Holt, 2016). Therefore, prevention and control of Kp infection was crucial. At present, the commonly used detection methods for Kp mainly included culture method, biochemical identification, nucleic acid detection based on PCR, etc (Dong et al., 2015). However, these methods were time-consuming, laborious, complex, and so on, which were not conducive to their application for point-of-care test and in resource-poor areas (Xu et al., 2022). In recent years, isothermal amplification technology such as LAMP had been gradually applied for the detection of Kp. As for LAMP, the reaction principle was complex, multiple pairs of primer increased the difficulty of primer design and complexity of the reaction system, and its false positive rate was relatively high (Sun et al., 2021). In contrast, benefiting from the application of recombinase, the principle of RPA was relatively simpler than LAMP. Additionally, RPA only had a pair of primers, which made the reaction system simpler than LAMP. Therefore, RPA was often considered to be a promising isothermal amplification technology. However, the non-specific amplification of basic RPA might produce false positive results, and the LOD of RPA was not low enough to satisfy the demand for ultra-low concentration of nucleic acid detection (Gong et al., 2022). Therefore, RPA signal amplification was necessary to improve its LOD and specificity. The CRISPR system had been extensively used in nucleic acid detection of pathogenic microorganisms. However, the LOD was also not good enough to detect ultra-low concentrations of nucleic acids when the CRISPR system was used alone for detection (Chen et al., 2022). Therefore, the integrated RPA and CRISPR system might compensate for their respective limitations during the detection of Kp.

In the present study, we developed an RPA-CRISPR/Cas12a platform to rapidly and accurately detect Kp and improved its performance through the optimization of each component. Eventually, the platform was able to detect Kp in 50 min at 37°C. The RPA-CRISPR/Cas12a platform had high specificity and did not cross react with other common clinical pathogens, which may be attributed to the specific primers of RPA and the high specificity recognition of RPA products by the crRNA of CRISPR/Cas12a system. The background of clinical sputum samples was complex, and the samples from patients with initial Kp infection always had low bacterial content, so the detection method must had an ultra-low LOD. In this study, the FD could detect DNA concentrations as low as 1 fg/μL, while the LOD of the LFTS was 10 fg/μL. The LOD of LFTS was poorer than that of FD, which was consistent with previous studies (Li Y et al., 2022). The basic RPA was only able to detect a DNA concentration of 1 pg/μL. The LOD of the RPA-CRISPR/Cas12a platform was significantly lower than that of the basic RPA, which could be attributed to that the Cas12a enzyme trans-cleaved the ssDNA reporter to exponentially amplify the RPA signal.

Herein, the RPA-CRISPR/Cas12a platform established in this study was further compared with the methods in other studies. The RPA-CRISPR/Cas12a platform was able to detect a lower concentration of Kp DNA than the RPA, RAA, LAMP, and LAMP-CRISPR/Cas12a method (Li N et al., 2022; Hou et al., 2022; Poirier et al., 2022; Qiu et al., 2022). In addition, the specificity of the RPA-CRISPR/Cas12a platform was the same as that of RPA (Hou et al., 2022), LAMP (Dong et al., 2015; Liao et al., 2020; Poirier et al., 2022), and LAMP-CRISPR/Cas12a (Qiu et al., 2022), and the sensitivity was the same as that of RPA (Li et al., 2022), RAA (Hou et al., 2022), and LAMP (Dong et al., 2015; Liao et al., 2020; Poirier et al., 2022). The RAA-LFS (Hou et al., 2022) was less specific than the RPA-CRISPR/Cas12a platform (84.6% vs 100%). This might be related to the further identification of the CRISPR/Cas12a, the selection of specific genes, and the stability of the test strips. Additionally, because the LOD of the LAMP-CRISPR/Cas12a was higher than that of the RPA-CRISPR/Cas12a platform, the LAMP-CRISPR/Cas12a (Qiu et al., 2022) was less sensitive than the RPA-CRISPR/Cas12a platform. As for the detection time, RPA, RAA, and LAMP did not combine with signal amplification step like CRISPR/Cas12a, thus their detection time was shorter than the RPA-CRISPR/Cas12a platform (Hou et al., 2022; Li et al., 2022; Poirier et al., 2022). The use of LFTS readout was conducive to point-of-care testing, but it was less stable than fluorescent testing when LFTS strips were exposed to air for a long time, which might more easily produce false-positive results (Hou et al., 2022; Li et al., 2022; Park et al., 2022). Furthermore, it was not suitable to use the LFTS strips to identify the results with the naked eye when the intensity of the detection band was very low (Xiong et al., 2022). As for LAMP, the colorimetric methods such as SYBR Green were usually used to achieve results readout by the naked eye, which was not easy to discern color changes when low-concentration samples were detected. Compared to LFTS and the colorimetric methods, the instrumental readout of the fluorescence signal was more stable and accurate. In our study, the FD used a real-time fluorescent quantitative PCR instrument to detect fluorescence signal and was suitable to be applied in the laboratories of resourceful areas. In contrast, the LFTS method used LFTS to detect fluorescence signal and was suitable to be applied for point-of-care testing and in the laboratories of resource-poor areas.

There were some limits in the present study. Firstly, although the RPA and CRISPR/Cas12a reactions could be completed within 50 min, the commonly used DNA extraction from clinical samples by commercial kits was always time-consuming. A fast, convenient, and efficient DNA extraction method for sputum samples needed to be developed in the future. Secondly, although the RPA-CRISPR/Cas12a platform had demonstrated 100% sensitivity and 100% specificity in clinical sputum samples testing, the sample size in our study was still relatively small, and a larger number of clinical samples were still required to further confirm the performance of this platform in the future.

In summary, we developed a novel detection platform for Kp based on the RPA-CRISPR/Cas12a. The platform was capable of detecting Kp quickly, easily, and accurately at room temperature. The platform had detection capabilities comparable to culture and qPCR methods. And it was easy to operate without relying on precise and complex temperature control equipment. It could be used for rapid and accurate clinical diagnosis of Kp in the future, especially in resource-poor areas.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving humans were approved by Medical Ethics Committee of the Affiliated Hospital of Youjiang Medical University for Nationalities. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin because Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements. Written informed consent was not obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article because Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

GW: Writing – review & editing. MT: Data curation, Methodology, Validation, Writing – original draft, Writing – review & editing. LL: Writing – original draft. CL: Writing – review & editing. ZZ: Writing – review & editing. SL: Writing – review & editing. XY: Writing – review & editing. CWa: Writing – review & editing. CWe: Writing – review & editing. JC: Writing – review & editing. XL: Writing – review & editing.