Page 48 - COS-FOD2015

Shoji et al.

Discussion

In the present study, the use of human saliva samples recruited were from age range from 19
samples for NMR-based metabolomics was to 64. Studies enclosing a larger number of
1
evaluated. High-resolution H NMR spectra was subjects and specific age group are needed to
obtained, and the major signals were clearly minimize the potential confounders.
assigned consistent with earlier studies. Tech- The samples collected in this study were only
nical advances in NMR spectroscopy, it is now
from around the Klang valley area. Samples
possible to identify and quantify hundreds of from other states should be covered in order to
metabolites from many different types of biolog- establish a database that can represent the
ical samples in relatively short orders [9]. How-
healthy adults in Malaysia. There are few un-
ever, the procedure from sample collection to known peaks needed to be further investigated.
analysis is sensitive. Cellular debris from both Some metabolites in the saliva that were previ-
human and bacterial sources may contribute
ously identified in other studies such as Butyr-
various metabolites, and products of bacterial ate, Choline, Ethanol, Glycine, Histidine and
metabolism may also be present. Steroids, Pyruvate were not found in this study. This
some other hormones, and many drugs and
might be due to the small sample size and thus
antibodies can diffuse from serum into saliva
inadequate of some metabolite concentration to
[13]. Filtration of saliva samples using centri- be quantified. In cases where quantitation was
fuge provides an effective method of removing
not reliable for a particular compound, concen-
both proteins and particulate matter from saliva
trations were either not used (for generation of
samples. Analysis of the sample will be easier a model using multivariate statistics), or report-
when the protein is removed from the sample
ed as being less than 1 µM. It is known that
[13]. It also depends on the method of collec-
gender has an influence over urinary metabolite
tion, food components may be directly observed concentrations [11, 12]. To determine differ-
during targeted profiling of saliva [14]. For ex-
ences in salivary composition due to gender, 20
ample, caffeine can be solutes already dis- females and 5 males was obtained and ana-
solved in the food, and rapidly dissolve into the lysed using H NMR spectroscopy coupled with
1
saliva. All of these factors provide the important
targeted profiling. Figure 3 revealed no signifi-
information for consideration when collecting
cant variation between male and female saliva
saliva sample. The result only showed of 25 metabolites. As compare to the study of
samples out of the total of 50 samples collect-
Takeda et al. [9], the male subjects has higher
ed, 20 females and 5 males. This smaller sam-
in concentration of nearly all to metabolites
ple set might be potentially affecting the result. compared to female subjects, included acetate,
The reason for reduced sample set was mainly
formate, glycine, lactate, methanol, propionate,
due to the inhomogeneities in its field of the
propylene, glycol, pyruvate, succinate, and tau-
samples, outliers with weird peak that may be rine. The sample size was obviously imbalance,
due to sugar, food or underlying systemic dis-
as 20 female samples versus 5 males samples
ease. Some studies indicated that there are
only. Thus, a larger sample set is needed for
age-associated alterations in certain aspects of further validation and future clinical application.
salivary gland function [15]. In this study, the

44