Salivary fatty acids in humans: a comprehensive literature review

Collection

Although there is no standardized technique to collect saliva for FA composition analysis, the collection procedure remains a critical step to maintain the reliability of the analysis, especially for specific FA classes such as SCFAs. Different strategies have been employed so far (Table 1). The collection of whole saliva is mostly performed in the morning before breakfast [14, 22], [23], [24]. In one study, samples were gathered at 6 pm [21] whereas in another trial, saliva samples were collected four times a day [22].

A variable period of fasting was requested: it varied between 30 min [23, 31] and 1 [18, 24], 2 [15, 20, 22, 32, 34], 6 [25], 8 [13], and up to 14 h [35], with the indication not to consume any drinks during that period [25].

In some studies, the participants were asked to avoid routine behaviors potentially impacting FAs content such as toothbrush [13, 25, 31, 34], use of hygiene products [20, 21, 27] and/or oral rinse, consumption of chewing-gum [26, 35] and physical activity practice [13, 25]. In one study, it was requested to avoid smoking within 1 h before collection. In one case, participants were asked to use the same non-anti-microbial fluoridated toothpaste [30].

Most studies analyzed unstimulated saliva, which was obtained using a Pasteur pipette placed under the tongue [22], by aspiration [34], by spitting [34] or letting the subject drool into a collection tube [31], a sterile plastic container [13, 23, 27] or glass tube/cup [25]. Subjects were instructed to avoid swallowing for 10 [21] or 5 min [34] during the sampling procedures. The volume of saliva collected ranged between 0.5 [27, 30] and 100 mL [35]. The salivary flow rate was recently evaluated as a parameter to consider during the collection phase [14]. In other studies, collection time was fixed at 5 [29, 33, 34] or 10 min [21, 23].

Stimulated saliva sampling was performed using a Salivette^® [14, 24] chewing gum base with a Curby cup [26], or chewing on a piece of parafilm [29]. However, as recently observed in the analysis of 90 metabolites [33], stimulation of salivary production may introduce bias in the analysis of fatty acids. Notably, the concentrations among spitting, aspiration and Salivette^® samples were equivalent only for palmitoleic acid, while the levels of palmitic, linoleic, oleic, and arachidonic acid were significantly different with Salivette^® collection [33]. Neyraud et al. showed that the FAs concentration in stimulated saliva is higher than in resting collection [29]. The authors suggest that lipids mainly originate from the parotid glands, which have a higher impact on the secretions of whole saliva [29]. However, results from other studies do not confirm that secretions from parotid glands are superior in lipids than sub-mandibular salivary product [36]. Rinsing of the mouth may also influence salivary FA components, stearic acid in particular, suggesting how collection methods without rinsing are preferable [33].

Sample preparation (centrifugation and storage)

A few studies have centrifuged the saliva samples before the analysis or storage [14, 23, 24, 26, 31, 34], eventually immediately after the collection [20, 22, 30, 37]. Storage was performed at −80 °C immediately after the sampling [21, 30], after [14, 23, 24, 29, 32] or before [21] the centrifugation step. In a few cases, storage was performed at −20 or −14 °C [31] and then the sample was centrifugated to analyze the supernatant [13]. In one study, the sample was spiked with butylated hydroxytoluene to prevent oxidation before storage [32]. Eventually, samples were placed on ice and then stored at −80 °C [27, 32]. In some studies, the collection was performed in ice or the tube was pre-chilled [20]. In one trial, the frosting was not executed [22]. In some cases, details on storage or centrifugation were not reported [18, 35]. Apparently, refrigeration and freezing at −80 °C are required to effectively remove cellular or bacterial membranes that would alter the result. This represents a consistent limitation of the analysis of fatty acids from saliva. If this is impractical, the sample can be collected on ice and quickly proceed with the analysis.

As for SCFAs analysis, Lenzi and colleagues [38] conducted a stability study and confirmed a variation of lactic acid only, after storage at room temperature for up to 6 h. An increase in lactic acid was recorded, probably due to the presence of lactate dehydrogenase in saliva.

Extraction and derivatization

The choice of the lipid extraction technique must be made according to the sample matrix [1]. In the salivary samples considered for this review, total lipids were extracted with Folch’s [20, 21, 25, 26, 28] or Moilanen and Nikkari methods [29, 34], which both consist of the employment of a mixture of chloroform:methanol that allows the successful recovery of lipids. Following the extraction, derivatization methods are chosen according to the analysis technique, the lipids of interest, and the objective of the study [1, 8]. For instance, fatty acids in some sphingolipids are linked also with amidic bonds and require for transmethylation with an acidic reagent, i.e. HCl. Instead, for FFAs a mild reagent is sufficient. Huang and colleagues [23] applied the same protocol of plasma NEFA profiling to salivary NEFA by converting them into fatty acid methyl esters (FAMEs), which best allow the analysis with GC [1]. The sample was mixed with isopropanol, hexane and methylation was performed with methanol and sulphuric acid. Afterward, hexane was used to extract FAMEs. A simultaneous quantification of sterols and FAs was developed using one-step tert-butyldimethylsilyl (TBDMS) derivatization of lipids extracted with tert-butyl methyl ether [15]. Additional details are available in Table 2.

Analysis

FAs in saliva have been measured alone or in combination with other lipidic components. For both purposes, in some cases lipids were previously extracted and then separated by thin-layer chromatography [22, 26]. However, gas or liquid chromatography followed by mass spectrometry is nowadays the most commonly employed tool to assess salivary FAs. Moreover, gas chromatographic methods represent the best practice as they allow to obtaining a satisfactory separation of all the FAs types, including isomers – positional and geometrical – for unsaturated FAs [15, 20, 24, 30]. Variants include high-performance liquid chromatography coupled to time-of-flight mass spectrometry [13, 18], inductively coupled plasma mass spectrometry [39], ultrahigh performance liquid chromatography–tandem mass spectrometry [14, 27], liquid chromatography/electrospray ionization-tandem mass spectrometry [31] and reversed-phased liquid chromatography and hydrophilic interaction liquid chromatography, coupled to high-resolution mass spectrometry [34], gas chromatography-flame ionization detection technique [21, 29, 32], eventually after sample lyophilization [21]. Spectrophotometric techniques were also used [35]. Nuclear magnetic resonance was adopted by Neyraud and colleagues [29]. Additional details are available in Table 2.

Short-chain fatty acids (SCFAs)

They consist of a class of branched-chain saturated FAs with two to seven carbon atoms that are mainly produced by colonic bacteria [38, 40]. The sample preparation and analysis of SCFAs require precautions as they may be lost in the process due to their volatility and water solubility [38]. To prevent sample deterioration, storage at −20 °C or lower temperature is required. Meanwhile, the adoption of an acidifying agent to mitigate the partial hydrophilic nature of SCFAs can improve the extraction [41]. Few studies assessing SCFAs in saliva were conducted with different techniques, including nuclear magnetic resonance (NMR) [42], high-performance liquid chromatography (HPLC) [43], LC-MS [44], isotope dilution method using GC–MS [45] and enzyme-linked immunosorbent assay (ELISA) [46]. Considering their properties, derivatization is often performed in case GC-MS is adopted, although a pilot study quantified SCFAs with a solid-phase microextraction coupled to GC-MS, without the adoption of a derivative agent successfully [47]. Due to the methodological differences, the assessment of SCFAs will not be deepened.

Lingual lipase

Lipids composing the human diet are mainly represented by triglycerides. Their absorption in the small intestine is preceded by their digestion and hydrolyzation in FFAs thanks primarily to the pancreatic lipases action.

The existence of salivary lipase has been debated for decades. Some observations point out that approximately half of the healthy individuals present lipase in the oral cavity [48, 49]. Such lipase might be particularly abundant in breastfed infants [50]. However, the secretion of lipase by human salivary glands has never been demonstrated [48]. It is currently assumed that the gastric or pancreatic lipase may reach the oral cavity by gastroesophageal reflux or duodenogastro-oesophageal reflux, respectively [48]. The lipase activity within the oral cavity is reduced due to the pH of the mouth, which is usually higher than the optimum of its action on lipolysis (pH 5.4) [51]. Studies conducted in periodontal disease-affected subjects suggest that lipolytic activity may be increased in this population, probably due to microbially produced enzymes and to the inflammatory environment [27]. It is currently speculated that lipase in the oral cavity is involved in fat taste sensitivity as the liberation of FAs from dietary lipids increases their binding to receptors responsible for fat perception [52]. The same inference has been proposed by Neyraud et al. who tested the role of salivary lipolytic activity in regulating the basal level of salivary fatty acids in humans [21]. The researchers found a positive correlation between this activity and FFAs concentration, which influenced, in turn, fat taste perception in the mouth [21] with repercussions on fat intake and body mass index [52], [53], [54].

Smoking

Smoking is associated with inflammatory status and several pathological processes in the oral cavity. These effects might involve salivary FAs. Palmerini et al. analyzed the salivary lipid levels of 7 smoking with those of 7 non-smoking subjects aged between 20 and 25 years [22]. Although free FAs were not evaluated, an overall decrease in unsaturated FAs (in fatty acyl chains of phospholipids) was observed in smoking as compared to non-smoking subjects [22].

Dietary intake

Unhealthy dietary habits are modifiable risk factors for chronic non-communicable diseases [55].

In fact, the health status may be affected by both the amounts and quality of some components of the diet, included lipids. Traditional tools such as food frequency questionnaires (FFQs) and 24 h recall are the most commonly used techniques to evaluate nutritional habits. However, they are fraught with inaccuracies (e.g., recall bias). Diet has a plenitude of effects on our organism and may even alter the lipid composition of saliva, especially for lipidic compounds that cannot be synthesized by the organism [25, 37], the analysis of the salivary FAs profile is a promising strategy to extrapolate reliable data on the dietary intake and nutritional status.

A few studies have compared the results from FFQs and salivary FAs analysis. The salivary FAs composition expressed as a percentage of total fatty acids was compared with the dietary FAs, calculated as a percentage of total FAs from each food item of an FFQ of 20 adults following a mixed or vegetarian diet [25]. A significant correlation was found only for myristic acid (14:0) for the mixed diet group (p=0.03) and arachidonic acid (AA, 20:4 n−6) for vegetarians (p=0.05) [25]. A possible explanation provided by the authors is that the different FA profiles between the two groups may reflect the lipid nutritional status, which is different among omnivores and vegetarians. Moreover, significantly higher concentrations of α-linolenic acid (ALA, 18:3 n−3) were found in vegetarian subjects compared to the mixed dietary group [25]. In another study [13], salivary concentrations of eighteen FAs of the omega-3 and omega-6 classes were compared to the estimated intake through an FFQ in seventy-nine healthy subjects. Higher salivary concentrations of linoleic acid (LA, 18:2 n−6) and arachidonic acid (AA, 20:4 n−6) were correlated with the dietary intake of omega-6 vegetable oils (β=0.80, 95 % CI 0.06–0.09 p=0.03) and red meat, cold meat, and viscera (β=0.40, 95 % CI 0.30–0.50, p=0.02) respectively. Salivary concentrations of ALA were significantly correlated with the intakes of nuts [13].

Women across the lifespan

The fatty acid profile in saliva varies according to the reproductive phase, as an evolutive mechanism of sexual recognition. Sixteen different fatty acids were retrieved by analyzing the saliva of 60 young women [28]. Even though the levels of multiple free fatty acids were comparable across the follicular, ovulatory and luteal phases, some differences in salivary FAs composition were observed. Oleic, acetic, palmitic, butyric and lauric acid are influenced by the levels of circulating steroidal hormones and their levels in saliva are high during the ovulatory and luteal phase. Stearic, myristic, butyric, caproic and linoleic acid levels were minimal in the ovulatory phase [28]. The magnitude of the secretory tissue of salivary glands decreases with age, with effects on the amount and composition of saliva. Recently, the metabolic components of saliva were compared between young and older women [34]. Among the lipid metabolites, isovaleric acid and linoleic acid were over-expressed in elderly [34].

Diseases and inflammatory conditions

Sjögren’s syndrome is one of the first pathological conditions that was studied for its alterations in the lipidic content of saliva [56]. Sjögren’s syndrome is an autoimmune, inflammatory disease characterized by the infiltration of lymphocytes in exocrine glands, including salivary glands with implications for dryness in the mouth. Specific FAs, especially palmitic acid, have been studied to unravel and modulate the inflammatory mechanisms of this disease. Palmitic acid was found to enhance the expression of IL-6 and, consequently, the inflammatory response in salivary glands by promoting CD4+ cell differentiation [57]. It was also found to interfere with the activity of antigen-presenting cells and adhesion mechanisms [57]. However, a recent study did not reveal the presence of the FFAs component, neither in patients with Sjögren’s syndrome nor in healthy subjects [18].

The relationship between cystic fibrosis and FFAs has also been evaluated. As chronic inflammation and increased lipolytic activity are key metabolic aspects of cystic fibrosis, FFAs are potentially relevant markers of disease activity [58]. The level of FAs in the saliva of young cystic fibrosis patients was first studied by Slomiany and colleagues in 1980 when a significant increase in FFAs in stimulated saliva was reported in 10 young patients compared to controls [36]. The authors identified a significant rise in salivary total FFAs and cholesterol compared with healthy subjects [36] and a recent study confirmed these findings [32]. In cystic fibrosis adult patients, the total salivary FAs are inversely correlated with serum triglycerides, possibly due to the increase in lipolysis. The ratio of unsaturated to saturated non-esterified FAs was found to be higher in cystic fibrosis patients, especially in those with severe lung disease, compared to healthy controls [32]. However, an exhaustive quantification of FAs was not performed, as only palmitic acid, stearic acid, oleic and linoleic acids were measured [32]. A comprehensive analysis of the FAs, including the complete PUFA profile, is therefore necessary for patients affected by cystic fibrosis, where the role of unsaturated fatty acids may be a key factor in the course of the disease [59].

In pediatric patients, the analysis of salivary metabolic parameters was performed to evaluate the differences between obese with and without hepatic steatosis, and normal-weight subjects, [24]. In obese patients, the level of palmitic acid was significantly higher than controls; obese patients with steatosis had higher levels of propanoic acid and 2-hydroxyisocaproic acid; obese patients without steatosis showed higher levels of butanoic acid and 2-hydroxybutyric acid [24].

Chronic oral inflammation and infection occur in periodontal disease, which is associated with dysbiosis, immunological alteration, and eventually connective tissue damage and bone loss. Dysbiotical alterations of the biofilm on the tooth surface, together with genetic and salivary factors, and exposition to sugar may lead to caries [60]. In both phenomena, bacteria are known to reshape the lipid composition of parotid saliva, and a possible role of the FAs composition in caries susceptibility was investigated [23, 26]. Huang et al. retrieved higher concentrations of lauric acid and myristic acid, while no significant change in the levels of polyunsaturated fatty acids and arachidonic acid was observed [23]. However, multiple metabolites derived from arachidonic acid were significantly increased in subjects with periodontitis [23]. A greater concentration of total FFAs, stearic, linoleic and docosahexaenoic acid were observed in subjects susceptible to caries compared to controls. In this context, SCFAs have started to be explored in saliva due to their possible role in periodontitis [42, 43, 45], gastroesophageal reflux [44] and specific cancer types [46]. In fact, alterations in the production of SCFAs by the oral microorganisms may be implied in the proinflammatory response, alterations with the immune system and carcinogenesis [46].

Patients with diabetes mellitus (DM) frequently develop oral inflammatory conditions such as gingivitis and periodontitis. Barnes et al. [30] compared the metabolomic profile of 80 subjects affected by DM with the one from 81 controls. Among the 161 subjects, 53 suffered from gingivitis and 55 periodontal disease.

An increase in inflammation-associated FAs was observed in subjects with periodontal disease without DM, particularly 12-hydroxyeicosatetraenoic acid, linoleic acid, and arachidonic acid. A similar FAs profile was detected in DM patients with gingivitis as compared to DM patients without gingivitis or periodontal disease [30]. A significant increase in docosapentaenoic acid (DPA 22:5 n−3) was also retrieved recently in acute heart failure subjects compared to chronic heart failure, partly due to the lipid rearrangement that occurs in relapses of heart failure [14].