Magnesium is one of the body’s major electrolytes, fundamentally important for the vital activity of all cell types [1]. Magnesium ions are necessary for stabilizing the structure of the DNA double helix and the spatial structures of RNA.

The human proteome databases (NCBI PROTEIN, UNIPROT, Human Proteome Map, etc.) include more than 20,000 proteins [2]. The total number of proteins in the proteome whose activity is associated with the magnesium ion is unknown. Identifying magnesium-dependent proteins (MDPs) is important for determining the impact of this essential micronutrient on various body functions, including the functioning of the nervous system, and, consequently, for describing the spectrum of clinical applications of magnesium preparations [3].

The aim of this study is to establish the functional categories of magnesium-dependent proteins.

The stages of this study included original algorithms for the recognition, classification, and identification of magnesium-binding proteins in the human proteome [4]. Based on the concepts of elementary amino acid motifs, positional independence of motifs, heuristic assessment of informativeness, and solvability on sets of elementary motifs [5], algorithms were developed to compute sets of the most informative amino acid motifs, which are used for annotating protein functions [6]. The obtained lists of magnesium-dependent proteins were analyzed using the functional linkage method, which includes the analysis of data on the cellular roles of proteins [7]. To assess the functional categories of proteins, the information array of the international Gene Ontology (GO) nomenclature was used [8].

For statistical processing of the research results, the following were used: calculation of numerical characteristics of random variables, testing of statistical hypotheses using parametric and nonparametric criteria, correlation and dispersion analysis. Comparison of predicted and observed frequencies of the studied traits was performed using the chi-square test, the Wilcoxon–Mann–Whitney test, and Student's t-test. For statistical processing of the material, the STATISTICA 6.0 software package and Microsoft Excel spreadsheets were used.

Analysis of the functional categories of magnesium-dependent proteins according to the international Gene Ontology (GO) nomenclature (Fig. 1A) showed that, overall, 1,503 GO categories were associated with the biological functions of magnesium. The largest numbers of proteins were found in the GO categories "cytosol", "ATP binding", "metal ion binding", "cytoplasm", "plasma membrane", and "cell nucleus": each of these categories contains more than 300 "magnesium" proteins. From 100 to 250 magnesium-dependent proteins were represented in the GO functional categories "nucleoplasm", "cell membrane", "extracellular exosome", "mitochondrion", "protein serine/threonine kinase activity", "GTP binding", "intracellular signal transduction", and "Golgi apparatus".

In the case of proteins associated with nervous system functioning (Fig. 1B), the most represented categories were those describing structural components of neurons and their connections:

From 10 to 24 proteins were represented in categories related to CNS development and neuroprotection:

Fig. 1A. Gene Ontology (GO) categories of magnesium-dependent proteins. A) Main categories of magnesium-dependent proteins

Fig. 1B. Gene Ontology (GO) categories of magnesium-dependent proteins. B) Protein categories related to nervous system function

Fewer than 10 magnesium-dependent proteins were included in each of the GO functional categories related to neuronal morphogenesis and migration (axon growth cone, neuron differentiation, axon terminus, myelination, axoneme, neuromuscular junction development, neuronal action potential, axon guidance, retina development, regulation of axonogenesis, regulation of neuron projection development, positive regulation of axon extension) and to synaptic signal transmission (glutamatergic synaptic transmission, synaptic membrane, long-term synaptic plasticity, activation of excitatory postsynaptic signal, regulation of synaptic plasticity, dopamine metabolism, regulation of myocardial contraction by calcium).

Overall, the analysis of protein functional categories showed that at least 172 magnesium-dependent proteins of the human proteome are involved in the implementation of neuroprotective, neurotrophic, and other "neurotropic" effects of the magnesium ion.

Such a wide representation of the element magnesium in the human proteome confirms its ubiquitous role in supporting the normal physiological state under conditions of normal physiological supply of this essential element [9]. The varying representation of magnesium-dependent proteins in the proteome indicates the priority of magnesium in supporting the functions of electrolyte metabolism [10], nerve signal conduction [11], regulation of myocardial contractility [12], and energy metabolism [13]. Due to the depletion of magnesium stores not only in bones and muscles [14] but also in the brain [15], a pathophysiological drift is formed, leading to mitochondrial dysfunction [16], accelerated aging, and multi-organ pathology [17].

In this study, a systems biology analysis of magnesium-dependent proteins was carried out in the context of their functional categories using modern mathematical methods of topological recognition theory. The study took into account the array of information available for magnesium-dependent proteins according to the international GO nomenclature.

Analysis of the functional categories of proteins showed that 1,503 GO categories were associated with the biological functions of magnesium. A total of 172 magnesium-dependent proteins of the human proteome are involved in the functional responses of the nervous system. These proteins are involved in neurotransmitter homeostasis, neuroplasticity, and neuronal survival. Some proteins may simultaneously participate in all of these neurophysiological processes.