A protocol for the detection of genetic markers in saliva by polymerase chain reaction without a nucleic acid puri ﬁ cation step: examples of SARS-CoV-2 and GAPDH markers

mild detergents allow e ﬃ cient detection of external reference and intracellular endogenous markers, while strong detergent, e.g. sodium dodecyl sulfate, inhibited the PCR reaction. Under these conditions, saliva samples can be stored for 24 h at +4°C or –18°C with the preservation of markers. Storage at room temperature led to the deterioration of marker detection. Snap heating of saliva samples at the time of collection, followed by storage at room temperature, provided partial protection. Conclusion. The protocol presented in this report describes the collection and storage of saliva for direct detection of genetic markers and is compatible with PCR and LAMP tests.


Introduction
Polymerase chain reaction is a frequently used highly sensitive diagnostic method for detecting nucleic acids. Nucleic acids (NAs) are embedded in various cellular structures, e.g., nuclei, vacuoles or protein-nucleic acids complexes [1]. Accessibility of NAs for amplifi cation is crucial for the performance and sensitivity of a PCR test. Clinical samples contain many diff erent components that may aff ect PCR reaction, e.g., nucleases and inhibitors [2]. The considerations of DNA accessibility and the complex nature of clinical samples prompted the introduction of a nucleic acid purifi cation step in PCR diagnostic tests. However, this purifi cation step increases the time and cost of each assay and requires dedicated laboratory instrumentation [3; 4]. Failure in the purifi cation procedure may also jeopardize an assay [5].
There have been a number of eff orts to develop protocols that would not require the purifi cation of nucleic acids. Direct detection of genetic markers without any additives to a clinical sample, the addition of organic solvents, buff ers, detergents, and absorbing materials have all been explored to omit or simplify nucleic acid purifi cation [6][7][8]. The rationale of these techniques is the release and collection of targeted genetic material in a form that can be amplifi ed in a PCR reaction. The success of reported methods varies. For example, direct detection of SARS-CoV-2 in nasopharyngeal swabs sample media has been reported, e.g., in Virocult, Transwab [8]. This direct detection, however, required lysis and heat-inactivation of swab samples. The addition of detergents to the sample collection media has also been reported. Detergent-mediated lysis releases NAs from various complexes and structures. However, detergents may have a detrimental eff ect on the stability of the reverse transcriptase and/or DNA polymerase used in PCR tests [6; 8]. An approach of snap-heating swab samples immediately upon collection has been reported [8]. The rationale for this technique is the denaturation of proteins, including nucleases, upon heating the sample to between 60°C and 120°C for few seconds. The drawback of this approach is that the heating of swab samples may induce RNA degradation.
Clinical samples used for PCR diagnostic analysis diff er in their composition, based on the origin of the sample. Blood, plasma, serum, buccal swabs, nasopharyngeal swabs and saliva are the most frequently used clinical materials. Among these, saliva is most suitable for self-collection. The collection of blood or swabs requires trained personnel. Therefore, saliva has been extensively explored as a source of clinical samples. In the ongoing COVID-19 pandemic, saliva is used as a clinical source for testing [9][10][11][12][13][14][15]. The meta-analysis by Butler-Laporte and colleagues showed high sensitivity (74-91 %) and specifi city (98-99 %) of SARS-CoV-2 detection in saliva [16]. The variability of saliva composition was of concern for the reliability of tests, e.g., presence and quantity of marker-containing material, chemical and enzymatic impact on the intactness of markers, protocols for extraction and stabilization of the markers. However, recent reports show promising results and confi rm that saliva has to be considered as a material for testing [9][10][11][12][13][14][15][16].
Here we report that genetic markers can be detected in saliva by collecting a sample in a solution containing mild detergents Triton X-100 and Tween 20. Tests with markers for SARS-CoV-2 and endogenous intracellular human GAPDH confi rmed the effi cacy of using detergent-containing solutions for the collection of saliva samples. The protocol described in this report signifi cantly simplifi es PCR-and LAMP-based tests by direct detection of genetic markers in saliva.  Current news using 2 % agarose (UltraPure Agarose, Invitrogen) in 1x TBE. The gel was stained with SYBR Safe DNA stain (Invitrogen), and the separated PCR products were visualized using the iBright CL1000 imaging system (ThermoFisher Scientifi c). DNA size markers were Trackit 1kb Plus DNA ladder (Invitrogen). Quantifi cations were performed using ImageJ [17].

Real-time PCR
Real-time PCR was performed with QuantStudio 6 Flex Real-Time PCR System tool (Applied Biosystems, ThermoFisher Scientifi c). We performed real-time PCR tests using the SYBR Green and with a fl uorescent dye. Reactions were set as for regular PCR reactions described above. For detection using SYBR Green, the stain was added to the reaction mixture (fi nal concentration 1 μM). The TaqMan Reagents protocol was set to 40 cycles at 92 0 C for 5 sec and 55 0 C for 30 sec. The Ct of amplifi cation were calculated using QuantStudio 6 Flex System. Melting curves were also collected. For the real-time PCR with the fl uorescence dye, a middle primer with a FAM reporting dye and IowaBlack quencher was added to reaction mixtures. The SYBR Green Reagent protocol was set to 40 cycles at 92°C for 5 sec and 55°C for 30 sec. Data were analyzed in Quant-Studio 6 Flex System. Following the real-time PCR analysis, all reactions were subjected to gel electrophoresis to monitor generated products, as described above in the «PCR reactions» section.

Loop-mediated isothermal amplifi cation (LAMP)
LAMP test was used to detect 2 diff erent regions of SARS-CoV-2. LAMP amplifi cation was performed with Bst 3.0 polymerase (New England Biolabs) for 30-60 min at 65°C. Amplifi cation was performed with 4 primers targeting 6 regions in the SARS-CoV-2 genome, and with a synthesized target representing the SARS-CoV-2 region. Primers and targets were obtained from Integrated DNA Technologies (www.idtdna.com), Twist Bioscience (www. twistbioscience.com), and SynBio Technologies (https://www.synbio-tech.com/). Detection was with 50 μM cresol red in the reaction mixture, by monitoring change of color from violet to yellow. The generation of DNA products was monitored by the agarose gel electrophoresis as described above.

Results
Direct PCR on saliva collected with a sample solution containing detergents and ethanol.
To test diff erent extraction components, we used the following sampling solutions: a) 1.0 % SDS, b) 1.0 % Triton X-100, c) 1.0 % Tween 20, d) 40 % ethanol and e) water. Saliva was collected in these solutions at a 1:1 ratio. DNA templates for SARS-CoV-2 (TS22) or GAPDH (TP1) were added to the saliva samples and/or tested as annotated in Figure 1.
The GAPDH TP1 and SARS-Cov-2 TS22 templates were added to sampling solutions in 2 concentrations, 1.5x10 -10 M and 1.5x10 -11 M, respectively. This provides a robust detection with 10 marker molecules per test reaction, Template titration tests showed that the copy detection threshold was at 1.0x10^2 copies per milliliter. Samples were handled at room temperature. One microL of the sample: solution mixture was used for each PCR reaction. On average, the time to prepare reactions was 20-35 min. The PCR cycling protocol is described in Table 2. PCR products were separated by agarose electrophoresis and gels were stained with SYBR Safe (Figure 1A). Specifi c PCR products were quantifi ed using ImageJ ( Figure 1). Quantifi cation showed that SDS strongly inhibited the PCR reaction, while Triton X-100, Tween 20 and ethanol did not aff ect the PCR. Similar results were obtained when testing the SARS-CoV-2 marker; the SARS-CoV-2 template (TS22) and specifi c primers were used ( Figure 1B). SDS inhibited the PCR reaction, while the SARS-CoV-2 marker was detected in samples collected in Triton X-100 and Tween 20 in 40 % ethanol. Therefore, we proceeded with the sampling solution containing 1.0 % Tween 20 with 40 % ethanol.
This solution was found to be compatible with real-time PCR protocols and with the LAMP assay ( Figure 1C, D). For the LAMP assay, the template was a SARS-CoV-2 sequence with 4 primers targeting 6 sites in the sequence. DNA amplifi cation was also monitored by the change of the reaction mixture color from violet to yellow; cresol red was used as a pH sensing dye. Analysis of generated DNA products by gel electrophoresis showed similar quantities of DNA generated from the template mixed with the saliva in the sample solution and the template in water ( Figure 1C).
Real-time PCR is frequently used in diagnostic.
We observed that saliva collected in the sample solution did not interfere with real-time PCR tests ( Figure 1D). We used two protocols of qRT-PCR. The fi rst was a Taqman protocol detecting generation of the fl uorescent dye FAM (with a quencher IowaBlack) and the second was based on detecting generated PCR products with SYBR Green. Ct for the samples with and without saliva was similar, i.e., 26 ( Figure 1D). Titration of the template detection showed that both assays could detect as little as 10 molecules in 1 ml of the sample used in the 25 ml reaction. The titration of detection limit by the qRT-PCR assay showed the range similar to RT-PCR and LAMP tests, i.e. 1x10^2 molecules per milliliter.
Thus, RT-PCR, two qRT-PCR detection methods (SYBR Green and FAM/IowaBlack) and LAMP assays showed that saliva can be used for the direct detection and that 1 % Tween 20 and 40% ethanol solution is suitable for collecting saliva for testing. The copy number threshold of detection for tests was determined as 1.0x10^2 copies per milliliter, with a robust detection of 10 marker molecules per a test-reaction.

Storage conditions: +4°C or freezing are recommended
To evaluate the impact of storage on the detection, saliva samples with added DNA templates were stored for diff erent periods at different temperatures. Samples were stored for 24 h, 5 h, or 0.5 h before using PCR amplification (Figure 2A, B). Testing is recommended within 24 h of sample collection, and therefore storage for more than 24 h was not tested. We observed a decrease of the signal following 5 h storage at room temperature (20-22°C).
After 24 h storage at room temperature, the signal decreased by more than 90 %. Storage of samples at +4°C or -18°C did not aff ect the detection of markers. These temperature conditions are recommended for clinical use. Freezing samples may be complicated at the sites of collection. Therefore, storage at +4°C is a good alternative for the storage and transportation of samples.
Therefore, decreased effi cacy of detection after storage at room temperature for 5 and 24 h was expected. Our data suggest that storage at room temperature for longer than 1 h should be avoided.
Snap-heating of clinical samples has been used to preserve degradation [6; 8]. Short bursts of heating to 100-120 0 C denatures proteins and protects the sample from degradation, as degrading enzymes are proteins. We observed that 5 minutes of heating to 80°C followed by storage at room temperature, prevented sample degradation to a signifi cant extent (Figure 2A, B). Thus, snap-heating can be used if there is no possibility to store samples at +4°C or below. Therefore, the recommended storage and transportation conditions are +4°C or below.
Control experiments with saliva samples included sterility tests and separation of saliva samples on SDS-polyacrylamide gels to monitor the protein pattern in samples (Supplementary Figure S1). These experiments show that the recommended sample solution (1 % Tween 20, 40 % ethanol in water) prevented microbial growth. The electrophoresis profi le of saliva was similar to reported saliva profi les [18]. See supplementary fi gure S1 for examples of these control experiments.

Detection of endogenous intracellular target
Detection of genetic markers requires that they are accessible to primers. Most genetic markers are found in complexes with other molecules, e.g., proteins. Cellular DNA and RNA form complexes with proteins, and viral DNA/RNA is contained within capsids [1]. For detection by PCR, nucleic acids have to be released from these complexes.
To explore how incubation with the sample solution may aff ect access to endogenous targets, we decided to test if we could detect endogenous human GAPDH ( Figure 2C).    mixed with saliva and sample solution, or the cell suspension was mixed with water, as annotated in Figure 2C. The ratio was 1:1:2 for cells:saliva:sampling solution, respectively. Under these conditions, we were able to detect endogenous GAPDH in cell extracts with or without saliva in the sample solution, with the same sensitivity that was obtained with the synthetic DNA template of GAPDH (TP1). Figure 2C shows an example with MCF7 cells; similar results were obtained with ACHN cells. For the PCR reaction, reverse transcriptase was used to generate cDNA from cellular GAPDH mRNA. Two concentrations of the cell extract with and without saliva in the sample solution were tested, i.e., 1x and 100x diluted cell extract annotated as 1.0 and 0.01 respectively ( Figure 2C). Detection of endogenous GAPDH in the presence of saliva shows that the sample solution can be used to detect intracellular genetic markers.

Discussion
Omitting nucleic acid purifi cation may significantly facilitate PCR-based testing. However, the complexity of saliva clinical samples, and the presence of nucleases, as well as the complexing of nucleic acids with proteins complicate effi cient direct detection. The protocol described here overcomes problems associated with nucleic acid purifi cation, preservation and accessibility of genetic markers for testing. Direct detection of genetic markers removes a costly and time-consuming purification step from the testing protocol [6; 8-9; 16]. The composition of the sample solution described here promotes the preservation of genetic markers and also allows for the storage and transportation of clinical samples. This is of great importance since many testing sites do not have access to advanced instrumentation. Saliva is also easier to collect as compared to other types of samples. Saliva can be self-collected and has been extensively explored as a source of testing [9].
The collection solution described in this report contains Tween 20 and ethanol. Mixing saliva with the sampling solution at a 1:1 ratio results in a solution containing 0.5 % of detergent and 20% ethanol. The concentration of Tween 20 was suffi cient to relax protein complexes without aff ecting enzymes in the PCR reaction (Figure 1). SDS, contrary to Tween 20, is a more potent denaturing detergent (https://pubchem.ncbi.nlm.nih.gov/compound/3423265), which was refl ected in the inhibition of the PCR reaction when the sample solution contained SDS (Figure 1). Tween 20 is used in the extraction of proteins and is known as a mild denaturing agent (https://pubchem.ncbi.nlm.nih. gov/compound/Polysorbate-20). This feature of Tween 20 benefi ts the extraction and stabilization of genetic markers (Figures 1 and 2). Saliva contains microorganisms that are present in the oral cavity. The addition of ethanol blocked microbial growth (Supplementary Figure S1). Therefore, the presence of a mild denaturant and ethanol protects from microbial growth and facilitates accessibility of targeted markers.

Control
The storage of samples between collection and analysis is of importance for successful testing. Saliva contains enzymes and chemical entities that may aff ect the stability of markers. Approaches to preserving sample integrity include freezing, keeping at cold, and chemical or thermal stabilization [6; 8-9]. Storage at room temperature, i.e., 20°C and higher, is not recommended. We observed that the storage of samples at +4°C or -18°C for 24 h preserved markers ( Figure 2). Storage at room temperature resulted in the degradation of markers already after 5 h. Snap-heating at +80°C for 5 min immediately following sample collection is aimed at denaturing enzymes in saliva [6,8]. Stabilization of the genetic marker after snap-heating was observed (Figure 2A, B), although the effi ciency of detection was lower as compared to storage at +4°C or -18°C.
We observed that the direct use of saliva without nucleic acid purifi cation was compatible with standard protocols of real-time PCR and LAMP amplifi cation ( Figure 1C, D). We observed that the protocol described here provides for detecting copy numbers in the range described for protocols using the nucleic acid purifi cation step. The detection range of the described protocol is in the range of 10 molecules per test reaction for all 3 tests described herein. It indicates that the performance of primers is of high importance for detection. This is similar to ranges reported for the Genmark ePlex and the Abbott RealTime SARS-CoV-2 tests, e.g., 10^2-10^3 copies per mil-liliter [19]. This protocol can be used in clinical trials to detect diff erent genetic markers, including markers of SARS-CoV-2 and intracellular endogenous markers.
To sum up: This report describes a protocol for the successful use of saliva for direct detection of genetic markers and omitting the nucleic acid purifi cation step. The protocol reports optimization conditions for saliva collection, storage and testing.
The protocol describes the collection of saliva in a solution containing Tween 20 and ethanol, storage conditions (+4°C or frozen), and shows compatibility with PCR and LAMP methods. Our report also describes crucial practical moments in saliva collection and storage that can aff ect results, e.g., troubleshooting by comparison with other solutions and detergents, deviations from the optimal storage conditions and how that may aff ect results. Such troubleshooting would be of help for the implementation of the protocol reported herein. The robustness and simplicity of this protocol are of advantage for its clinical use.
Ethical approval: This work was performed under an IBC permit from Qatar University (QU-IBC-2019/023).
Self-collected saliva was provided by healthy volunteers upon signing an informed consent form of the Qatar University Institutional Review Board.
Availability of data and materials: All data are available upon request. All data generated and analyzed during the study are included in this published article.