Development and Evaluation of a New qPCR Assay for the Detection of Mycoplasma in Cell Cultures

In recent years, cell culture has become an important tool not only in research laboratories, but also in diagnostic and biotechnological development laboratories. Mycoplasma contamination is present in up to 35% of cell cultures used in research and in cell therapies. This fact represents a significant problem since such contamination can cause disastrous effects on eukaryotic cells by altering their cellular parameters, which, in turn, can lead to unreliable experimental results. For this reason, it is mandatory to carry out continuous testing for the presence of Mycoplasma in cell culture and the development of appropriate methodologies for this purpose. An ideal detection methodology should be fast, sensitive, and reliable. In this study, we propose an alternative detection method based on real-time PCR in conjunction with a novel combination of primers and probes that have been improved to increase their efficiency. The new PCR method demonstrates 100% sensitivity and specificity results in the detection of common Mycoplasma species that contaminate cell cultures. Whilst 11 of 45 tested supernatants were positive for Mycoplasma (24.4%) using the new PCR method (corresponding to 5 of the 14 lines tested (35.71%)), only 10 of 45 supernatants showed positive results with the commercial Venor®GeM qEP and Plasmotest® kit. In addition, the new PCR method exhibits a high capacity to detect less-frequent Mycoplasma species, such as those related to the M. mycoides cluster. The use of an alternative Mycoplasma-detection method in cell culture labs can guarantee the detection of Mycoplasma contamination, especially in cases when dubious results are recorded.


Introduction
Animal cell cultures have become a basic, frequent, and fundamental tool not only in research laboratories, but also in diagnostic and biotechnological development laboratories [1]. In this sense, animal cell lines are used in in vitro models to investigate new therapeutic approaches against different diseases [2,3]. For this reason, it is essential that cell cultures are kept aseptic. Although different microorganisms (bacterial, fungi, yeast, or viruses) are capable of contaminating cell cultures [4], Mollicutes (or interchangeably Mycoplasma) are the fundamental cause of contamination. Despite the availability of sophisticated technology and specialisations used in manipulating cell cultures, biological contamination primarily due to Mycoplasma contamination remains a significant challenge.
Mycoplasma contamination of cell cultures was first described 50 years ago [5], and it is claimed that approximately 5-35% of current cell cultures are contaminated [6]. The Mycoplasma genus comprises microorganisms with distinctive characteristics, including high resistance to antibiotics [7], reduced metabolism, the lack of a rigid cell wall, a small size (0.3-0.8 µm), and high flexibility, so they are able to pass through common antibacterial filters (0.45 µm) [8]. These microorganisms grow slowly, even if they are in that could compromise future experiments. In our experience, we have observed the detection of sometimes uncertain or confusing results. The importance of successfully detecting Mycoplasma contamination in cell cultures, in conjunction with the current lack of an infallible commercial methodology that can be used to detect Mycoplasma contamination, led us to develop an alternative, highly sensitive, real-time PCR assay detection method that can be used in our laboratory and in the scientific community to assure the correct analysis of Mycoplasma contamination. The developed method is based on the selective amplification of a DNA fragment using a new and original set of universal and degenerate primers targeting the 16S rDNA region from a large selection of the Mycoplasma species genomes.

Supernatant Samples
A total of 14 cell lines were received in the BBSSPA between February 2018 and December 2019. These cell lines were cultivated at different passages, and a total of 45 supernatants from 14 cell lines were tested for Mycoplasma detection (Table 1). Each supernatant sample (1 mL) was recovered and stored at −80 • C until the time of testing using NAT assays.

Samples for Cell Culture Screening
The samples were processed according to Venor ® GeM qEP guidelines for analysis using NAT assays. The Mycoplasma-detection process involved a simplified DNA isolation protocol that did not require an additional DNA isolation kit. For each sample, 500 µL of supernatant was transferred to a 1.5 mL tube and was incubated at 95 • C for 10 min. After a brief period of centrifugation, 2 µL of each sample was used for qPCR detection.
A collection of Mycoplasma Sensitivity Standards (Minerva Biolabs, Heidelberg, Germany, Cat. No. 102-0002) containing 10 CFU/mL was also used to assess the sensitivity and specificity of the qPCR assay. The standards contain inactive Mycoplasma DNA of 9 species: A. laidlawii, M. arginini, M. fermentans, M. gallisepticum, M. hyorhinis, M. orale, M. pneumoniae, M. synoviae, and S. citri, as well as a negative control sample. The lyophilised material was reconstituted using 1 mL of Essential 8 TM DMEM/F12(Ham) (1:1) culture medium (ThermoFisher Scientific, Waltham, MA, USA, Cat. No. A1516901). After 5 min of incubation at room temperature, the samples were vortexed for 10 s. Finally, a Venor ® GeM Sample Preparation Kit (Minerva Biolabs, Heidelberg, Germany, Cat. No. 56-1200) was used to purify the Mycoplasma Standards. Then, 200 µL of the standards was treated and isolated as a sample following the manufacturer's recommendations. The sample was subsequently resuspended in 60 µL of resuspension buffer.

The Design of Primers and Probes
The study and alignment of the 16S rDNA regions of different Mycoplasma species were performed, including species known to commonly contaminate cell cultures and unusual species [7,8,10] Due to the high number of Mycoplasma species that must be detected and the significant degree of diversity in the sequences across the different species examined for this study, only limited areas of the 16S rDNA gene can be used in the design of primers and probes that are common to most Mycoplasma species. Therefore, as proposed by Uphoff and Drexler [8], an equimolecular mixture of the six proposed forward primers was used as a forward primer, except where two of them were simplified into a single primer in anticipation of a degenerate position; specifically, we used the primer CGCCTGAGTAGTACGTWCGC instead of CGCCTGAGTAGTACGTTCGC and CGCCTGAGTAGTACGTACGC, as well as the primer TGCCTGRGTAGTACATTCGC instead of TGCCTGAGTAGTACATTCGC and TGCCTGGGTAGTACATTCGC, as proposed by Uphoff and Drexler [8]. Furthermore, we designed a new reverse primer to amplify a fragment of approximately 256 bp in length. For qPCR detection, we selected a new 'FAM'-labelled probe ( Table 2) designed based on the small region conserved in all selected Mycoplasma species and included it between the forward and reverse oligos used. All of the primers and probes were supplied by Sigma-Aldrich. A synthetic DNA fragment (169 bp) was designed as an internal amplification control. This synthetic amplification sequence was designed to be amplified using the same forward and reverse primers employed in Mycoplasma amplification but presenting a different internal sequence. A specific HEX label probe was used for the internal control, as depicted in Figure 1. This fragment was synthesised by chemical synthesis and cloned into a pMA-T vector by the GeneArt gene synthesis service (ThermoFisher Scientific, Waltham, MA, USA), providing us with a 2543 bp plasmid. Due to the high number of Mycoplasma species that must be detected and the significant degree of diversity in the sequences across the different species examined for this study, only limited areas of the 16S rDNA gene can be used in the design of primers and probes that are common to most Mycoplasma species. Therefore, as proposed by Uphoff and Drexler [8], an equimolecular mixture of the six proposed forward primers was used as a forward primer, except where two of them were simplified into a single primer in anticipation of a degenerate position; specifically, we used the primer CGCCTGAGTAG-TACGTWCGC instead of CGCCTGAGTAGTACGTTCGC and CGCCTGAGTAGTAC-GTACGC, as well as the primer TGCCTGRGTAGTACATTCGC instead of TGCCTGAG-TAGTACATTCGC and TGCCTGGGTAGTACATTCGC, as proposed by Uphoff and Drexler [8]. Furthermore, we designed a new reverse primer to amplify a fragment of approximately 256 bp in length. For qPCR detection, we selected a new 'FAM'-labelled probe ( Table 2) designed based on the small region conserved in all selected Mycoplasma species and included it between the forward and reverse oligos used. All of the primers and probes were supplied by Sigma-Aldrich. A synthetic DNA fragment (169 bp) was designed as an internal amplification control. This synthetic amplification sequence was designed to be amplified using the same forward and reverse primers employed in Mycoplasma amplification but presenting a different internal sequence. A specific HEX label probe was used for the internal control, as depicted in Figure 1. This fragment was synthesised by chemical synthesis and cloned into a pMA-T vector by the GeneArt gene synthesis service (ThermoFisher Scientific, Waltham, MA, USA), providing us with a 2543 bp plasmid.

qPCR Conditions
The optimal qPCR conditions were established to guarantee the maximum sensitivity using serial dilutions of an M. fermentans DNA standard (1 × 10 5 copies/µL to 1 copy/µL). Primers and probes described in the previous section were tested using a TaqMan Fast

qPCR Conditions
The optimal qPCR conditions were established to guarantee the maximum sensitivity using serial dilutions of an M. fermentans DNA standard (1 × 10 5 copies/µL to 1 copy/µL). Primers and probes described in the previous section were tested using a TaqMan Fast Advanced Master Mix (ThermoFisher Scientific, Waltham, MA, USA, Ref: A44260) in a QuanStudio-6 thermocycler (ThermoFisher Scientific, Waltham, MA, USA). Different final primer concentrations for Mycoplasma (250, 500, 750, and 1000 nM) were tested for the sep-up of the assay. In the same way, five probe concentrations for Mycoplasma and for the internal control (Myco-S1 Probe and Myco-IC-S1 Probe, respectively) (62.5, 93.75, 125, 187.5, and 250 nM) were tested. In addition, different dilutions of the internal control, between 1 × 10 −4 and 1 × 10 −9 copies/µL, were tested to guarantee correct detection without compromising the efficiency of the test.
Finally, after qPCR optimisation, each PCR reaction was prepared in a final volume of 20 µL, containing 10 µL TaqMan

Purification of Amplified Bands
The amplified fragments were separated in a 2% agarose gel at 60 V. Specific

DNA Sequencing
The purified 16S rDNA fragments were used in sequencing reactions with a BigDye Terminator v3.

Sequence Assembly and Clustering
The homology of the 16S rDNA sequences was checked using the Nucleotide BLASTn database (from the Nucleotide BLAST, National Center for Biotechnology Information) to compare the sequenced 16S rDNA with other 16S ribosomal RNA sequences available in the rRNA/ITS databases.

Mycoplasma Detection with Plasmotest ®
The Plasmotest ® kit (InvivoGen, San Diego, CA, USA, Rep-pt1) allows for the specific detection of Mycoplasma lipopeptides, which can interact with TLR2 receptors on HEK-Blue TM -2 cells. The activation of the signalling cascade leads to the secretion of alkaline phosphatase, which is detected via colorimetry thanks to the HEK-BlueTM detection medium that presents a purple/blue hue in the presence of alkaline phosphatase. Cells should be used two days after passage and before reaching 80% confluency. Culture flasks for each cell line were used for Plasmotest testing. The testing protocol was performed according to the manufacturer's recommendations, and the test results were interpreted after a 24 h incubation period at 37 • C. The kit used in the test included negative and positive controls.

Controls Used in the Study
The following controls were used for study validation: -Venor ® GeM qEP includes a positive control in the kit that was used as an additional sample. Ultrapure water included in the kit was added as a negative control. -A M. fermentans DNA standard containing 1 × 10 2 copies of strain NCTC 10117 was used as the new qPCR assay positive control. Ultrapure water was added as a negative control. -1/10 dilutions of the negative and positive controls included in the Plasmotest ® kit were used as controls and treated as any other sample.

Mycoplasma qPCR Development
In this study, we conducted a validation test of a novel qPCR technique for the specific detection of a large number of Mycoplasma species. This was achieved by using 16S ribosomal RNA sequences available in rRNA/ITS databases. The PCR conditions were finetuned to ensure optimal detection and included the use of quantified M. fermentans DNA standard dilutions in triplicate. The inclusion and optimisation of an internal amplification control were thus achieved without compromising the sensitivity of the PCR reaction, thereby ensuring the efficiency of the PCR reaction.
Once the PCR conditions were established, the sensitivity of the qPCR was assessed by means of 10 CFU of lyophilised standards resuspended in 1 mL of Essential 8 TM DMEM/F12(Ham) (1:1) culture medium. Following isolation, 2 µL was used for qPCR, and the assays were performed in triplicate. All of the samples were detected via qPCR, and negative and positive controls validated the results that we obtained, as detailed in Table 3.

Comparative Analysis of Detection Methods
To evaluate the new PCR technique, we performed a comparative analysis of Mycoplasma detection methods based on 14 different cell line supernatants of different origins at different passages (see Table 1). A total of 45 culture supernatants were used for Mycoplasma detection, including the new qPCR technique (being tested), a commercial Venor ® GeM qEP kit, and a culture flask of each cell line at the same passage for Plasmotest testing. Positive results were obtained for 11 supernatants from five different cell lines (cell lines 5, 6, 7, 8, and 13). Eleven samples tested positive for Mycoplasma with the qPCR optimised in this study, while only 10 positive samples were detected with Venor ® GeM qEP and Plasmotest kits. Neither the Venor ® GeM qEP kit nor the Plasmotest kit detected one of the samples tested for cell line 5, although at different passages, the Venor ® GeM qEP kit did not detect Sample 42, and the Plasmotest kit did not detect Sample 43 (Table 1). It is important to note that Samples 14, 17, and 18 from cell lines 5 and 6 showed 3 to 10 positive results with the Venor ® GeM qEP kit, despite not exhibiting clear and exponential amplification, as specified in the kit's guidelines (Figure 2

Comparative Analysis of Detection Methods
To evaluate the new PCR technique, we performed a comparative analysis of Mycoplasma detection methods based on 14 different cell line supernatants of different origins at different passages (see Table 1). A total of 45 culture supernatants were used for Mycoplasma detection, including the new qPCR technique (being tested), a commercial Venor ® GeM qEP kit, and a culture flask of each cell line at the same passage for Plasmotest testing. Positive results were obtained for 11 supernatants from five different cell lines (cell lines 5, 6, 7, 8, and 13). Eleven samples tested positive for Mycoplasma with the qPCR optimised in this study, while only 10 positive samples were detected with Venor ® GeM qEP and Plasmotest kits. Neither the Venor ® GeM qEP kit nor the Plasmotest kit detected one of the samples tested for cell line 5, although at different passages, the Venor ® GeM qEP kit did not detect Sample 42, and the Plasmotest kit did not detect Sample 43 (Table 1). It is important to note that Samples 14, 17, and 18 from cell lines 5 and 6 showed 3 to 10 positive results with the Venor ® GeM qEP kit, despite not exhibiting clear and exponential amplification, as specified in the kit's guidelines (Figure 2). The 11 samples that tested positive were sequenced. Their accession numbers are indicated in Table 1. Using the BLASTn database, we detected that one of the samples (Sample 15) presented 100% homology with M. hyorhinis. The other 10 samples showed identical sequences, and the most significant homologies (>99%) were detected with Mycoplasma sp. related to the M. mycoides cluster (M. cottewii and M. yeatsii). The complete 16S rDNA sequence was also sequenced to authenticate Mycoplasma identification (Accession number OP364020) [27]. The 11 samples that tested positive were sequenced. Their accession numbers are indicated in Table 1. Using the BLASTn database, we detected that one of the samples (Sample 15) presented 100% homology with M. hyorhinis. The other 10 samples showed identical sequences, and the most significant homologies (>99%) were detected with Mycoplasma sp. related to the M. mycoides cluster (M. cottewii and M. yeatsii). The complete 16S rDNA sequence was also sequenced to authenticate Mycoplasma identification (Accession number OP364020) [27].

Discussion
Mycoplasma contamination is a significant issue in cell cultures, and therefore, it is essential to have a precise, efficient, sensitive, and rapid diagnostic method for the detection of Mycoplasma contamination in laboratories that work with cell cultures [1]. The presence of Mycoplasma in cell culture alters cell growth and metabolism, producing alterations in DNA, the altered expression of surface antigens and cytokines, and causes changes in cell morphology, among many other consequences. These alterations mean that the results derived from the use of these cells are not credible or reproducible, which entails serious consequences in research and in the pharmaceutical industry [13]. As a result, most academic journals require that the research they publish certify that the study was conducted using Mycoplasma-free cell lines. Furthermore, the sterility of a cell line culture must be demonstrated to be deposited in a cell line bank [2,15,28].
Real-time PCR assays have gained importance in recent years for their role in detecting conserved sequences of Mycoplasma [1,8,10]. Due to the large number of existing Mycoplasma species and their genetic variability, developing a unique and infallible system for detecting Mycoplasma in cell culture is challenging. Although several commercial kits are available, most of them cannot be used for real-time PCR, and many of the available qPCR kits do not enjoy worldwide distribution. It is also important to consider that specific recommended protocols, such as PCR amplification programs and stated amplification sizes, are identical across different kits. These peculiarities lead us to infer that the primers used in many of these testing kits (and (potentially) the manufacturer of these kits) are the same, thus resulting in their shared strengths and weaknesses. It is also important to mention that manufacturers do not disclose what primers are used in their protocols, thereby rendering it impossible for researchers to elucidate the underlying reason(s) for discrepant test results. The documentation that accompanies the different kits includes (i) diverse sensitivity and specificity specifications and (ii) claims regarding the detection of the presence of a highly variable number of Mycoplasma species. Given this scenario, in combination with our experience in the detection of unclear positives with existing commercial methodologies, we have developed a new qPCR technique as an alternative Mycoplasma contamination detection system. The internal control that is integrated into the reaction in our detection system ensures the integrity of the qPCR and the detection of inhibitory agents. We claim that this technique is a valuable option for the scientific community and quality control laboratories for testing cell culture supernatants and resolving ambiguous or discrepant cases of Mycoplasma contamination.
Similar to other PCR methodologies, the new PCR methodology evaluated in this study is based on the 16S ribosomal gene of a cohort of Mycoplasma and Ureoplasma species, owing to this gene's high degree of conservation in specific regions. Our findings regarding this new methodology corroborate its optimal sensitivity regarding the detection of the nine species that were tested, which are responsible for 98% of Mycoplasma contaminations, but also include atypical species related to the M. mycoides cluster, as previously validated through sequencing [7,27,29]. It is important to note that claims concerning the sensitivity of Mycoplasma detection assays are regulated by the European Pharmacopoeia, Section 2.6.7 [26]. As stated in this document, a sensitivity of 10 CFU/mL is required if a PCR assay is to be used for the detection of Mycoplasma. Our experiment detected a minimum of 10 CFU/mL for the nine species tested, with Cq values between 29.4 for S. citri and 36.74 for M. pneumoniae and M. arginini. Other PCR tests have been previously evaluated and proven to be sensitive to at least 10 CFU/mL for detecting Mycoplasma in culture supernatants [30], while the sensitivity of other published assays has been comparatively lower [22,31,32].
In this study, we analysed 45 supernatants from 14 cell lines at different passages. The new PCR developed in this study revealed that 11 out of these 45 supernatants tested positive for Mycoplasma (24.4%), although only 5 of the 14 cell lines tested positive (35.71%) positive. These positivity values are consistent with values reported in the literature [6]. Two Mycoplasma species were detected in the cell cultures through sequencing, namely M. hyorninis in one sample and a Mycoplasma sp. related to the M. mycoides cluster in 10 different samples. While M. hyorninis is a typical contaminant in cell cultures [7,21], the Mycoplasma sp. related to the M. mycoides cluster is usually found in infected cattle and small ruminants [33]. Previous studies have identified these microorganisms in the ear canal of goats and have reported on their presence in bovine milk and bovine lung tissue [34,35]. The contamination of cell cultures with Mycoplasma sp. related to the M. mycoides cluster has only been described twice previously [36,37]. Sequencing revealed that the species were identical in the 10 supernatants. Since they correspond to four cell lines from the same laboratory, we infer a common source of contamination, probably associated with the use of contaminated foetal bovine serum during cell line culture [36]. The complete 16S ribosomal DNA sequence cited in this study has been fully sequenced and is described in Carrillo-Avila et al. [27].
For the evaluation of the results, we have considered as true positives, those samples with a positive result based on at least two of the methodologies used. The novel PCR assay showed a congruence of 97.7% with the Venor ® GeM qEP PCR test and the Plasmotest colorimetric test. Comparing the results of the novel PCR test with the colorimetric assay, we observed a false negative outcome for Sample 43 (cell line 5) using the Plasmotest kit. This false negative could be due to the inadequate growth of this cell culture since the confluence of this cell culture was less than optimal (i.e., <80%). The growth difficulty experienced with this cell line has been reported on by other authors [32]. Thus, determining the optimal growth conditions and confluence may pose a limitation for Mycoplasma detection using the colorimetric assay in the Plasmotest kit. A lack of specificity has also been described as a limitation of this colorimetric test since it can produce results for other bacteria and yeasts [38]. Nonetheless, colorimetric kits are widely accepted as an appropriate alternative tool for verifying test results [39], particularly in laboratories where routine PCR assays cannot be performed.
However, we note that 22.2% (10 out of 45) of the supernatants tested in this study indicated positivity for Mycoplasma using the Venor ® GeM qEP PCR assay. This corresponds to 35.71% (5 out of 17) of the tested cell lines. The findings of this study are consistent with the results obtained using the newly developed PCR assay, except for Sample 42 (cell line 5), which tested negative with the Venor ® GeM qEP PCR assay. Nonetheless, this sample tested positive using the new PCR test, and sequencing analysis revealed the presence of a Mycoplasma sp. related to the M. mycoides cluster that is similar to the other supernatants from cell line 5. Sample 42 also tested positive using the colorimetric assay, indicating that the negative result obtained with the Venor ® GeM qEP PCR test was, in all probability, a false negative. This result may be due to the lower sensitivity of the Venor ® GeM qEP PCR test with regards to Mycoplasma sp. related to the M. mycoides cluster. Notably, the amplification of Mycoplasma sp. in Sample 42 using the new PCR technique occurred at a higher cycle of Ct (31.87), thereby indicating a lower concentration of Mycoplasma in this particular sample. This decrease in sensitivity is also consistent with the non-exponential amplification observed for Samples 14, 17, and 18 from cell lines 5 and 6 using the Venor ® GeM qEP kit, despite being considered 'positive' according to the kit's specifications. Results compatible with those exposed have been previously observed by other authors, since different commercial NAT assays are frequently not capable of detecting some species of Mycoplasma at the <10 CFU/mL required in the European and Japanese pharmacopeias. Even some methods are not able to detect frequent Mycoplasmas, like M. orale and M. synoviae, and other Mycoplasmas, like M. ixodetis. These facts have led to the suggestion of the need to develop species-specific assays to supplement the deficiencies observed with commercial platforms [22].
If we take into account the premise of considering as true positives, those samples with a positive result based on at least two of the methodologies used, we can conclude a 100% sensitivity and specificity for the new PCR assay. The test is even able to detect a minimum of 10 CFU/mL for the nine species regulated by the European Pharmacopoeia, Section 2.6.7 [26]. However, an important limitation of our study is that broad-range sensitive detection cannot be guaranteed because only a limited number of known Mycoplasma species have been tested. It is even necessary to assay a large number of cell culture samples to have more robust sensitivity and specificity data.

Conclusions
The peculiarities of the Mycoplasma genus pose a significant challenge to the development of an infallible method for its detection. The occasional manifestation of inconclusive NAT outcomes and/or the requirement to perform a confirmation test (to confirm one's test results) in conjunction with the enormous degree of heterogeneity in existing Mycoplasma species, the emergence of novel mutations as a consequence of natural evolution, and the ability of these species (including unusual species) to infect cell cultures indicate that the continuous evaluation and adjustment of diagnostic systems is necessary. The proposed new PCR assay for Mycoplasma detection was able to detect the minimum of 10 CFU/mL for the nine species regulated by the European Pharmacopoeia, Section 2.6.7 [40], in addition to showing 100% sensitivity with the 45 supernatant samples from 14 different cell lines. According to the good results obtained, the clinical significance of the new PCR assay is highest with the species and samples analysed. Despite the routine use of a commercial method for Mycoplasma detection, implementing an alternative Mycoplasma detection technique in cell culture laboratories is a step towards guaranteeing the detection of Mycoplasma contamination, particularly contamination by unusual species or in difficult samples that contain a low bacterium titre. Funding: This research was funded by Consejería de Salud y Consumo, Junta de Andalucía, by the ISCIII Platform Biobans and Biomodels (PT20/00065).