Metabolomics and human breast milk
A unique and inimitable food for infants
Flamina Cesare Marincola, Sara Corbu, Roberta Pintus, Angelica Dessì, and Vassilios Fanos
Human breast milk (HBM) is recognized to be the perfect food for every newborn to promote optimal health and development (1). Besides being rich in nutrients (carbohydrates, lipids, proteins, vitamins, and minerals), HBM provides biologically active compounds (hormones, growth factors, cytokines and chemokines, antimicrobial substances, etc.) and, according to the most recent research, bacteria that promote intestinal development and multipotent stem cells (2). Breastfeeding has many health benefits for both the mother and for infant health (3). In the short term, this practice reduces infant mortality and the incidence of intestinal, respiratory, or urinary infectious diseases. In the long term, it is also associated with less chance of allergies and diabetes, in addition to promoting cognitive development. Additionally, breastfeeding lowers the risk of breast and ovarian cancer, diabetes, and postpartum depression for the mother. According to the indications of the World Health Organization (4), infants should be exclusively breastfed under 6 months of life and, possibly, breastfeeding should continue for 2 years or more, according to the desire of mother and child.
The composition of breast milk varies from woman to woman, between the different stages of breastfeeding (colostrum, transitional milk, mature milk), during the day, and during each feeding. Other factors that influence the composition of breast milk are mother genotype, diet, and gestational age. This variability makes HBM unique for each mother-infant pair. When HBM is contraindicated or women are unable to breastfeed, dedicated infant formulas are the only suitable substitute for breast milk. Indeed, most of these formulated products have as a basic component the cow’s milk, whose composition differs widely from human milk. For this reason, the dietary industries make use of different technological processes to change the composition of the formula milk and to make it closer to that of the mother’s milk (5). However, despite the considerable progress made in the improvement of formula milk, HBM remains an inimitable food not only in terms of its nutrient composition but, in particular, for its inter- and intraindividual variability and its “immune function.” Our belief is that the future of research on human milk will be based on the 3 M’s: metabolomics (presented in this chapter), microbiomics, and multipotent stem cells (6).
The “omics” technologies (genomics, transcriptomics, proteomics, metabolomics, etc.) represent analytical approaches aimed at investigating the complex function of an organism through an holistic study of a large spectrum of molecules (genes, proteins, metabolites, etc.). The integration of the multiple levels of information arising from “omics” studies represents the specific objective of “systems biology,” an interdisciplinary science focused on a comprehensive understanding of the regulation of cell behavior (7).
Metabolomics lies at the end of the “omics cascade” (8). This discipline deals with the study of the whole complement of primary metabolites (metabolome) present in a biological sample such as biofluid, cell, or tissue. Since these molecules are considered as the final products of the genome and its interaction with the environment, metabolomics reports on the actual functional status of an organism, and thus it provides a snapshot of the metabolic phenotype reflecting the physiological, evolutionary, or pathological status of a given biological system.
Metabolomics strategies are divided into two distinct approaches: targeted and untargeted. Targeted metabolomics is used in case of an hypothesis-driven experiment and analyzes a restricted number of known metabolites or compound classes, providing quantitative or semiquantitative results (9). In the untargeted approach, hundreds to thousands of metabolites can typically be measured, thus offering the potential to determine novel biomarkers (10). The two most accepted platforms used in targeted and untargeted metabolomics are nuclear magnetic resonance (NMR) spectroscopy and high-performance chromatography (GC or HPLC) coupled to mass spectrometry (MS) (11). Although these analytical platforms are capable of detecting a huge number of metabolites in a single analysis, no one of these techniques can completely detect, identify, and quantify all metabolites in one analysis, due to the wide chemical diversity and broad range of concentration of the metabolites.
The complexity of the high-throughput data set arising from a metabolomics study imposes the use of appropriate statistical methods for modeling data. To this aim, chemometric techniques lend a powerful hand in analyzing and interpreting the interactions between the hundreds or thousands of measured analytes (12). The most widely used techniques are based on the application of projection methods. They are mainly distinguished as unsupervised or supervised, depending on the type of analytical strategy. In unsupervised methods, the classification of groups is not known a priori. These methods reduce the complexity of the data available so that they can be represented by means of visually interpretable plots. In this way, it is possible to identify particular structures within the data set including clusters, anomalies, or relationships between the measured variables. In the supervised approach, the classification is known a priori. The objective of the analysis is the mathematical formulation of descriptive and/or predictive models and the determination of metabolites that distinguish between groups, i.e., for biomarker discovery.
Many metabolomics applications have been initially focused on medical issues to acquire knowledge on the mechanisms of the disease, the drug action, and to explore biomarkers (13). This approach has been shown to be a valuable tool to advance research in a huge diversity of fields such as environmental sciences (14), sport and exercise science (15), and nutritional and food science (16).
Metabolomics for human breast milk
Metabolomics applications on HBM aim to improve the understanding of the biochemical composition, dynamics, and functions of the metabolites present in this biofluid. Despite only a few investigations performed so far, the findings of these studies have pointed out the potential of metabolomics to study different aspects related to the nutrition and health of the infant. The most relevant results are as follows.
The first metabolomics investigation on HBM was performed in 2012 (17). Water-soluble and lipid fractions were extracted from milk samples obtained from mothers prematurely giving birth (26–36 gestation weeks) and some formula milk samples and analyzed by NMR spectroscopy and GC/MS, respectively. The preliminary results of this study pointed out the presence of biochemical variability between preterm HBM and commercial milk and within the group of HBM samples (Figure 28.1).
Figure 28.1 Score (a) and loading (b) plots (PC1 versus PC2) generated from principal component analysis (PCA) of the 1H-NMR spectra of water-soluble extracts of human breast milk (⦁) and formula milk (◻). (Source: Ref. 17, with permission.)