Background

Saliva is one of the biofluids suitable for monitoring the systemic conditions. The non-invasive availability of saliva samples enables frequent, timely, and cost-effective diagnostics, which would contribute to realizing personalized medicine. Therefore biomarker discoveries for various diseases have been reported (Wang et al., 2017). Saliva contains microorganisms and also a wide variety of components, such as genome, coding and non-coding RNA, proteins, and metabolites (Bonne and Wong, 2012; Yoshizawa et al., 2013). Traditionally, the profiling of these components has been utilized for the diagnostic of oral diseases (Dawes and Wong, 2019; Martina et al., 2020). Recently, accumulated evidence has revealed the potential to monitor systematic health conditions (Sugimoto et al., 2013; Kaczor-Urbanowicz et al., 2017).

Among various omics technologies, metabolomics has been utilized to discover the biomarkers for metabolic diseases, including cancers (Trezzi et al., 2015). The metabolic aberrance in cancer cells, such as the Warburg effect (Warburg, 1956), potentially reflects metabolite profiles in biofluids. Metabolomics-based liquid biopsy is therefore intensively developed (Armitage and Ciborowski, 2017). In comparison to the blood and urine-based diagnostics, the use of saliva is an emerging approach while salivary metabolic profiles showed the potential to detect various cancers (Sugimoto et al., 2010a).

The salivary diagnostics of oral cancer are the most reported among various cancers (Washio and Takahashi, 2016; Chattopadhyay and Panda, 2019). We previously confirmed the consistently elevated metabolites in saliva and oral cancer tissues (Ishikawa et al., 2016). The effect of saliva collection after various fasting duration on these markers was also evaluated (Ishikawa et al., 2017). The evaluation of the specificity of oral cancer-specific markers against various diseases in the oral cavity, e.g., periodontal diseases, is also conducted (Mikkonen et al., 2016). Discrimination of oral leukoplakia, oral lichen planus, and oral cancer is practically useful in clinical settings (Sridharan et al., 2019; Ishikawa et al., 2020). The storage condition of the saliva samples affected the discrimination abilities of metabolite markers (Wang et al., 2014), which requires the standard of the protocol to deal with the saliva samples for realizing reproducible diagnostics.

Recently, salivary metabolite biomarkers to diagnose the cancers far from the oral cavity have been reported. The quantified polyamines in saliva were utilized for the detection of breast cancers (Takayama et al., 2016; Murata et al., 2019) using liquid chromatography-mass spectrometry (LC-MS). These analyses were conducted using triple-quadrupole-MS (MS) for targeted analyses. We utilized capillary electrophoresis time-of-flight-MS (CE-TOFMS) to conduct untargeted analyses to quantify hundreds of salivary metabolites for pancreatic cancer detections (Asai et al., 2018). MS requires the ionization of metabolites to be detected and therefore various separation system, detection, and their interface are available (Monton and Soga, 2007). Among these systems, CE-MS is one of a powerful tool that can analyze the charged metabolites using only two modes, cation and anion, i.e., positively and negatively charged metabolites (Soga et al., 2003; Soga, 2007). Here, the metabolite profiling protocol for saliva samples using CE-TOFMS is described.