Rapid detection of Mycoplasma mycoides subsp. capri and Mycoplasma capricolum subsp. capripneumoniae using high-resolution melting curve analysis

Mycoplasma capricolum subsp.subsp. capripneumonia (Mccp) and Mycoplasma mycoides subsp.sbusp. capri (Mmc) cause caprine pleuropneumonia (CCPP) and mycoplasmal pneumonia in goats and sheep (MPGS), respectively. These diseases cannot be identified on clinical symptoms alone and it is laborious to distinguish them using biochemical methods. It is therefore important to establish a simple, rapid identification method for Mccp and Mmc. Here, we report a high-resolution melting (HRM) curve analysis using specific primers based on the Mmc 95010 strain MLC_0560 and Mccp F38 strain MCCPF38_00984 gene sequences. The method was highly specific with intra- and inter-batch coefficients of variation < 1%. The lower limit of detection for Mccp and Mmc was 55 copies/μL and 58 copies/μL, respectively. HRM and fluorescence qPCR results were compared using 106 nasal swabs and 47 lung tissue samples from goats (HRM-qPCR coincidence rate 94.8%; 145/153). Mycoplasma isolation and identification was performed on 30 lung tissue samples and 16 nasal swabs (HRM-culturing coincidence rate 87.0%; 40/46). HRM analysis was more sensitive than fluorescence qPCR and Mycoplasma isolation, indicating the practicality of HRM for accurate and rapid identification of Mccp and Mmc, and diagnosis and epidemiology of CCPP and MPGS.

Construction of a positive plasmid standard. Mmc  Sensitivity. The positive plasmid standard and Mmc and Mccp genomic DNA were diluted in tenfold serial dilutions with ddH 2 O, and PCR-HRM analysis was conducted under the optimal reaction conditions described in "Optimization of the HRM detection method" section to establish the minimum detection limit of the HRM method. Each dilution series, along with a blank control, was repeated three times. A standard curve diagram was drawn according to the Ct values and concentration. Ct values and melting curves were used to define the limit of detection (LOD). Data analysis and standard curve was performed with Rotor-Gene Q Series Software. For comparison with the conventional PCR assay, primers and reaction conditions used were the same as for the PCR step of the HRM method.
Reproducibility. The positive control plasmid was diluted to three gradients to determine the reproducibility of the HRM method. All measurements were repeated three times for each sample at each concentration in the batch; the inter-batch test was carried out three times at an interval of 3 days. The HRM reaction system and cycle parameters are as described in "Optimization of the HRM detection method" section. The intra-and inter-batch coefficients of variation were calculated according to the Tm values of the two pathogens. Statistical analysis was performed with SPSS Statistics 25. Tm values between the two groups were compared with the independent t-test. We considered two-sided P values of 0.05 to be significant.

DNA sequencing. To benchmark Mmc and
Mccp detection with HRM analysis, DNA sequencing was performed. After the HRM reaction was completed, the HRM products were subjected to agarose gel electrophoresis, recovered, and purified. The purified target fragment was ligated into pMD 18-T and used to transform DH5α competent cells as described in "Construction of a positive plasmid standard" section. MF-107 and MR-107 were used for PCR to determine the positive clones, which sequenced at Takara Biotechnology (Dalian, China).

Mycoplasma isolation and identification.
Lung tissues were homogenized with modified Hayfick's culture medium. Ten-fold serial dilutions were prepared by combining 0.3 mL of the homogenized supernatant and subsequent dilutions with 2.7 mL of modified Hayfick's culture medium resulting in a dilution series from 10 -3 to 10 -1 . All dilutions were incubated at 37 °C in 5% CO 2 , and three blind passages were performed every 7 days. Afterwards, 0.1 mL of the cultures that turned yellow was diluted tenfold to 10 -5 to 10 -3 , and 0.2 mL of each dilution were streaked onto FRIIS Agar plates and cultured at 37 °C in 5% CO 2 for 5-10 days. A single characteristic  (Fig. 2) showed that only two specific melting peak curves of Mmc and Mccp were formed. The difference in Tm value of two dissolution curves was approximately 1.3, indicating that the two strains could easily be distinguished based on the melting curves, while no specific melting curves were generated from common goat pathogens as Movi and ORFV, indicating that this HRM method was specific for Mmc and Mccp.
HRM reproducibility. As shown in Table 1, the intra-and inter-batch coefficients of variation for melting peak Tm1 were both within 0.08-0.11%, while for Tm2, they were 0.08-0.14% and 0.11-0.14%, respectively. As these values were all < 1%, the reproducibility of this method was verified. Tm values between the two groups were verified as statistically significant by the independent t-test (P = 0.000; P <0.01) ( Table 1).
HRM sensitivity. The positive control plasmids were diluted tenfold and detected using the HRM method described here as well as conventional PCR. The results (Fig. 3) indicated that specific melting curves and Tm values could be determined at Mccp and Mmc plasmid concentrations as low as 55 copies/μL and 58 copies/ μL, respectively. The Ct values of Mccp and Mmc were both less than 40, there was a good correlation between copy number (10 5 to 10 1 copies/μL) and the Ct values, no reaction was detected at concentrations of 10 0 copies/   (Fig. 3). The minimum detectable Ct values between the two groups were compared with independent t-test, the statistical analyses show the average Ct differences between the two pathogens were significant (P = 0.012, P < 0.05) (  (Table 3). Thus, the HRM method established in this study was more sensitive than fluorescence qPCR and pathogen isolation.  Table 2. Comparison of LOD in the detection of Mmc and Mccp. Different lowercase letters in the same row mean significant difference (0.01 < P < 0.05). The average Ct values differences between the two pathogens were significant (P = 0.012 < 0.05). No.

Mccp Mmc
Ct mean χ ± SD CV/% Ct mean χ ± SD CV/%  26 . Fluorescence qPCR is more rapid with greater sensitivity and specificity than conventional PCR, and has been widely used in the clinical detection of various pathogens [27][28][29]  HRM has been widely used for mutation scanning, methylation detection, mononucleotide polymorphism analysis, genotyping, sequence matching, and other applications because of its high sensitivity, good specificity, low cost, high-throughput detection. Douarre et al. 31 described in 2012 a HRM-PCR assay for clearly differentiating the two main types of Mycobacterium avium subsp. paratuberculosis (cattle and sheep) and Gurtler et al. 32 developed in the same year a HRM-PCR method for identification and van genotyping different Enterococcus species. In 2018, Liu et al. 33 developed a highly specific multiplex high resolution melt-curve real-time PCR assay for the reliable detection of Salmonella serotypes. This approach is also suitable for establishing a method for rapidly identifying Mccp and Mmc.
In this study, the Mccp and Mmc sequences were compared to those published in NCBI to identify conserved gene fragments. A primer pair was designed using the Primer Premier 6.0 software to establish a HRM detection method for Mccp and Mmc. This method demonstrated specificity towards Mccp and Mmc without amplifying corresponding sequences from other common goat pathogens. Additionally, five positive samples detected using the HRM method were randomly selected for sequencing. The results showed that the sequence homology between the determined sequence and corresponding Mmc fragment was 100%, which demonstrated the high specificity of the method. The method showed strong reproducibility as melting curves of Mccp and Mmc were essentially the same as those of the positive control samples, and the intra-and inter-batch coefficients of variation for the two melting peaks were < 1%. Results showed that the minimum detection limits of Mccp-and Mmc-positive plasmids using this method were 55 copies/µL, and 58 copies/µL, respectively, whereas the minimum detection limit using SYBR GreenI qRT-PCR was 226 copies/µL and the detection limit with conventional PCR was 2.26 × 10 4 copies/L 19 . Furthermore, the minimum detection limits of the multiplex PCR method were 4.03 × 10 5 copies/µL and 6.78 × 10 5 copies/µL for Mmc and Mccp, respectively 15 , indicating that the HRM method was more sensitive than the conventional and fluorescence qPCR methods for Mmc and Mccp detection.
To verify the practicability of this method in clinical testing of Mmc and Mccp, a total of 153 clinical samples were detected using the HRM method established in this study and compared to the results of fluorescence qPCR and Mycoplasma isolation and identification. The results show that the detection rate of Mmc using HRM was 20.3%, whereas that of SYBR GreenI qRT-PCR Mmc analysis and Mycoplasma isolation was 15.0% and 4.3%, respectively. Additionally, eight samples that were detected to be positive for Mmc with HRM method but negative by SYBR Green I qRT-PCR method were sequenced. The sequencing results confirmed the consistency with the corresponding fragment sequence of Mmc to be 100%, supporting the accuracy of Mmc detection using the HRM method, and indicating that the sensitivity of the HRM method was also higher than that of fluorescence qPCR and Mycoplasma isolation in detecting clinical samples. Furthermore, Thiaucourt 9 established a real-time PCR method for detecting the M. mycoides cluster, which can detect and differentiate Mmc and Mccp, but it requires two separate fluorescence qPCR methods. The HRM method established in this study can differentiate Mmc and Mccp in only one reaction, which is faster and less labor-intensive. Because Mccp has not been found in Fujian Province, it was not detected in any of the clinical samples, which is consistent with the results of Jiang et al. 34 .
In summary, detection of Mmc and Mccp based on the HRM analysis established in this study can quickly and accurately distinguish the two pathogens with high specificity, sensitivity, and reproducibility. Compared to conventional and fluorescence qPCR methods, this method showed higher sensitivity and convenient operation. It can be used for rapid detection and identification of Mmc and Mccp in clinical nasal swab samples, www.nature.com/scientificreports/ rapid diagnosis, and epidemiological investigation of goat mycoplasmal pneumonia and contagious caprine pleuropneumonia.