Salivary peptidome profiling for diagnosis of severe early childhood caries

Dental caries is one common chronic disease among children [1], remaining a major problem in many countries [2‐5]. Severe early childhood caries (s-ECC) is its aggressive form, defined by one or more of the following criteria: (1) any sign of smooth surface caries in children aged <3; (2) one or more decayed, missing (due to caries), or filled tooth surfaces in primary maxillary anterior teeth in those aged 3–5; (3) having decayed, missing, or filled surface score (dmfs) more than four at age 3, or more than five at 4, or more than six at 5 [6]. s-ECC can reduce the children’s ability of eating and drinking, resulting in tooth decay or malocclusion of permanent dentition [7, 8]. These further cause malnutrition, speaking and sleeping problems, as well as impair the school performance, social behaviour, and self-esteem of the patients [4]. Therefore, s-ECC has made many negative impacts on the quality of life of preschool children and their parents [8, 9].

Paediatric dental caries is also a significant problem in China. According to the national epidemiological report, the prevalence of dental caries among 5-year-old children reached 66 %, with mean dmfs 3.5 [10]. Recent studies demonstrated that polarization of caries has been more pronounced [11‐13].

The known factors influencing dental caries in children includes: (1) immature immune systems; (2) characteristics of saliva; (3) cariogenic microorganisms; (4) feeding type; (5) oral hygiene care in early childhood [14]. Recently, the characteristics and components of saliva have also been suggested to forewarn the development of s-ECC [15].

Saliva is one important body fluid containing complex components which has great diagnostic value [16, 17]. As the host-associated factor, saliva plays an essential role in dynamic equilibrium between demineralization and remineralization, changing the sensitivity and resistance to dental caries [18]. Many proteins in saliva function during adhesing and aggregating of cariogenic bacteria [18‐20]. However, some other proteins in saliva may contribute to the defence against microorganisms [21‐23].

Recently, new proteomics techniques, including high-throughput analysis of the salivary proteome, make it possible to characterize its protein composition and investigate its impact on development of dental caries. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) is a useful method for identification of microbes and proteins, which has high efficiency and automation but lower costs [24‐26], making great benefits to clinical and laboratorial experiments. It is one of the most promising techniques to reduce the huge gap between high throughput proteomics and clinical studies with large samples [27, 28]. MALDI-TOF MS can be used in combination with weak cation-exchanger (WCX) magnetic bead to select peptides from saliva in the range of 1000–10,000 Da prior to further identification. The WCX magnetic beads can separate the proteins and/or peptides of different isoelectric points from complex biological fluids with specific anionic ligands, which is confirmed to be suitable for the fractionation of low-molecular-mass range peptides (1–10,000 Da) and can exhibit optimal performance which can be confirmed in many studies, such as some cancer research [29, 30]. The techniques of MALDI-TOF MS combined with WCX magnetic beads incorporate both of their advantages such as low cost and simple purification, which could capture more peptides than some other methods especially in the low-molecular-mass range [31, 32].

In the previous study from our research group, we have designed a cross-sectional study to introduce weak cation-exchanger magnetic bead-based matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (WCX MB-based MALDI-TOF MS) to compare the components of salivary peptides in children with and without s-ECC, and successively found certain differentially expressed peptides between caries and non-caries children [33]. Based on the former research, in the present study we also use this method, WCX MB-based MALDI-TOF MS, to evaluate the profiles of salivary peptides in those children with s-ECC at different time points of before, 1 and 4 weeks after treatment, so as to verify the hypothesis that certain peptides in saliva are associated with different periods of before and after treatment of dental caries in children, and could be used as candidate biomarkers for diagnosis and identification of status of s-ECC.

Paediatric dental caries is a major public health problem, becoming the most common chronic infectious disease in childhood of great concern to parents and dentists [4, 6]. Saliva and its components play important roles in protecting oral structures. Although infection of Streptococcus mutans, inappropriate feeding and diet, and poor oral hygiene care are all causal factors, but endogenous factors, such as characteristics of saliva, may determine why some children develop caries while others do not [15, 35, 36].

In this study, we used proteomic technique to detect and discriminate peptide profiles in stimulated whole saliva from children aged 3–5 years with s-ECC compared with after 1 and 4 weeks of treatment. Conventional salivary proteomics studies mainly used 2-DE and SDS-PAGE, in which proteins are separated in terms of their molecular weight, followed by MS for identification. However, these strategies have several significant disadvantages, including complex procedures, the large sample size required, and that the informative low molecular weight peptides are likely outside their detection range [37].

Currently, the contribution of the different sources to whole saliva, and the evaluation of individual variability and physiological modifications have been investigated by top–down proteomic approaches, disclosing the faceted and complex profile of the human salivary proteome [38]. In a study, the whole saliva of 88 children aged between 0 and 48 months were analysed by a top–down platform based on RP-HPLC–ESI–MS, the result showed how the salivary proteome of human children evolves during infant development and concentrations of some salivary proteins (acidic proline-rich proteins, histatin-1, histatin-3, statherin, P-B peptide, and salivary cystatins) slowly increased with age, with different trends [39]. Top-down proteomic approaches were also used to detect biomarkers in some systemic diseases, such as schizophrenic and bipolar disorder [40], lung cancer [41], and chronic graft-versus-host disease [42].

In our study, we are using MB-based MALDI-TOF MS, which is an approach that overcomes some limitations of conventional salivary proteomics. MBs are micro-particles available with a range of different surface chemistries providing a choice of functionalities [43]. They are used for sample preparation (e.g., saliva, serum, plasma, urine, and cerebrospinal fluid) to enrich target low molecular weight and low-concentration peptides by their specific affinities [31]. In comparison with other time-consuming techniques in proteomics, the MB-based MALDI-TOF MS technique is rapid, high-throughput, and precise, with adequate resolution and sensitivity for the investigation of complex salivary samples [32]. It has also been used to detect and identify specific candidate biomarkers, not only in oral pathologies, such as periodontitis [44] and oral cancer [45], but also in systemic diseases, such as Sjögren’s syndrome [46], rheumatoid arthritis [47], and systemic lupus erythematosus [30].

In our results, the intensities of seven peaks (Figs. 2, 3a) differed significantly between the pre- and post-treatment s-ECC groups, indicating that this method can be used to analyse the peptide profiles in children with dental caries. We also found that several peptides were associated with having experienced caries, and might influence the onset and development of early childhood caries. We verified the diagnostic model established using four peptides (experimental m/z values: 1723.7, 1851.4, 2331.0, and 2995.7 Da) in other six cases and the results showed sensitivity and specificity of this model were both 83.3 %, indicating these salivary peptides should be potential candidate biomarkers of s-ECC. In the previous study by Si et al. [33], the findings revealed different composition of significantly expressed peptides between children with and without dental caries under a cross-sectional study design. The samples were from children in a different kindergarten, and dental caries in that study were not treated before analysis. Our present study was a longitudinal study with caries treatment, which was an entirely different design and could better obtain the potential biomarkers of caries treatment and experience. However, further studies are still needed if we would like to draw a full picture of diagnosis of dental caries, especially to explore the difference between untreated non-caries, untreated caries, well-treated caries and secondary caries.

Some previous studies have identified salivary peptides, including histatins, defensins, and proline-rich proteins (PRPs), may contribute to paediatric caries [18, 48, 49]. In this study, two peptides (experimental m/z values: 1723.7 and 1886.5 Da) were predicted to be segments of histatin-1 (Table 1), and their levels of expression were lower in the pre-treatment than post-treatment s-ECC group. However, the difference between their sequences were only one amino acid (tyrosine), indicating these two peptides might be homogenous. But on the other hand, this will provide much stronger evidence for histatin-1 as a candidate biomarker for dental caries.

Histatins are salivary proteins that have a high affinity for hydroxyapatite and contribute to the acquired enamel pellicle [50]. When histatins are adsorbed onto the enamel surface, they provide protection against acid aggression [51]. In this study, we collected 16 stimulated whole saliva samples from 8 children with s-ECC before and after 4 weeks of treatment (equal amounts of the total protein were used for each sample) who were randomly chosen from the whole sample, to perform western blotting assay in order to learn the difference of protein expression of histatin-1 between different time points before and after treatment of dental caries. Western blotting was performed in parallel with constant conditions, and we found that the levels of salivary histatin-1 were higher in children with s-ECC after treatment for 4 weeks compared with before treatment (P < 0.01), which was consistent with the tendency of expression of the two peptides with experimental m/z values of 1723.7 and 1886.5 during different period of treatment of dental caries in MALDI-TOF MS. In a study performed in 2005, high-performance liquid chromatography-mass spectrometry (HPLC–MS) was used to investigate the correlation between salivary peptides and dental caries in 20 samples, who were divided by dmfs index into two groups [18]. The results indicated strong correlations between high levels of phosphopeptides (PRP 1-3, histatin-1, and statherin) and the absence of dental caries. Results of the present study were consistent with it despite the different method applied. These above suggested that histatin-1 played a significant role in maintaining tooth integrity and protection against cariogenic bacteria, and might be associated with having experienced caries.

In addition, we found prolactin inducible protein (PIP) and myoglobin might also be candidate biomarkers of dental caries. PIP is a 17 kDa single polypeptide chain, known by various names due to its versatile nature and function in human reproductive and immune systems, which expressed in lacrimal, salivary, and sweat glands [52]. It is considered as a prognostic biomarker because of its over-expression in metastatic breast and prostate cancer [53, 54]. Moreover, its aspartyl-proteinase nature suggests its role in tumour progression [55, 56]. In saliva, PIP is a lipopolysaccharide-binding (LPS-binding) protein capable of binding to many bacteria living in oral cavity, including A. actinomycetemcomitans, P. gingivalis [57], Gemella haemolysans, Gemella morbillorium, Streptococcus acidominimus, Streptococcus oralis, Streptococcus salivarius and Streptococcus parasanguinis [58]. PIP may play an important role in the blockage of bacterial adhesion, contributing to pathogenesis and host immunity. We could only find very few related studies on the relationship between PIP and dental caries in recent years. In one study in 2014, magnetic beads and LTQ-Orbitrap hybrid Fourier transform mass spectrometry were used to identify the Streptococcus mutans lipoteichoic acid (Sm.LTA) binding proteins in the saliva of caries-free and caries human subjects. However, the results indicated that the expression of prolactin-inducible protein had no significant difference in both groups [59]. Meanwhile, myoglobin, one protein that was known to bind oxygen in myocardium and skeletal muscles, was not reported relevant to saliva and dental caries in previous studies. Further studies will be needed to explore the probable function of PIP and myoglobin in dental caries.

Although proteome profiling using MB-based MALDI-TOF MS together with LTQ-Orbitrap-MS is effective for the identification of saliva-specific peptides, some limitations are still remaining. First, identifying all the peaks was challenging as some peptides in the saliva may be secreted by epithelial cells of salivary glands and oral microorganisms, or may be produced by proteolysis or inflammatory factors associated with the body’s response to infection [60, 61], which might become confounding factors when searching for candidate diagnostic biomarkers of dental caries. Second, among the seven differentially expressed peptides, only four of them were identified successfully, while the other three had no matching results from LTQ-Orbitrap-MS. We speculate these peptides might have varying stability when using different MS methods, and hopefully these peptides can be identified if we could introduce other precise methods of peptidomics with less impact on peptide stability in future studies. Third, the bio-informative identification of these identified peptides were performed by searching human peptide and protein databases, indicating the authentic origination of these peptides may not fully match the identified sources present in the databases. There should also be some possibility that these peptides might be secreted by some microorganisms or other sources in saliva. This issue will not be fully resolved until the database is improved to become much more complete in the future, and we may have a look back on these peptides periodically as the database is ever-growing all the time. Fourth, as for histatin-1, due to the limitations of current methodology and dynamic transformation of histatin-1 in a variety of physiological, biochemical and metabolic processes, we were not able to detect the entire sequence of histatin-1 by mass spectrometry, though we had confirmed levels of salivary histatin-1 are higher in children with s-ECC after treatment for 4 weeks compared with before treatment by western blotting. In the future, with the development of research technologies and improvement of peptidomic methodology, we are looking forward to performing more in-depth studies on the intrinsic mechanisms between histatin-1 and the occurrence and development of dental caries. Last but not the least, the number of samples analysed was not quite enough, which may limit the validity of the results. Further studies in larger patient cohorts with blinded samples are required to confirm the usefulness of these diagnosis-oriented peptides, together with the function and structures of these novel candidate biomarkers. This will facilitate saliva proteomic profiling of s-ECC at the molecular level and yield a better understanding of s-ECC and its prevention.

In summary, these results indicated that proteomics approaches can help discovery of new candidate biomarkers and identification of s-ECC-related proteins. The combination of 4 peptides here (experimental m/z values: 1723.7, 1851.4, 2331.0, and 2995.7 Da) may be a diagnostic indicator of the status of s-ECC. Besides, 2 differentially expressed peptides (experimental m/z values: 1723.7 and 1886.5), which might be probably homogenous, were predicted to originate from histatin-1, and histatin-1 was also identified as one candidate biomarker at protein level. In the future, caries risk of children is hopeful to be assessed by testing the specific peptides in saliva, which can be useful for drawing up individualised preventive strategies.