Carbohydrate Structure Database: tools for statistical analysis of bacterial, plant and fungal glycomes

Abstract

Carbohydrates are biological blocks participating in diverse and crucial processes both at cellular and organism levels. They protect individual cells, establish intracellular interactions, take part in the immune reaction and participate in many other processes. Glycosylation is considered as one of the most important modifications of proteins and other biologically active molecules. Still, the data on the enzymatic machinery involved in the carbohydrate synthesis and processing are scattered, and the advance on its study is hindered by the vast bulk of accumulated genetic information not supported by any experimental evidences for functions of proteins that are encoded by these genes. In this article, we present novel instruments for statistical analysis of glycomes in taxa. These tools may be helpful for investigating carbohydrate-related enzymatic activities in various groups of organisms and for comparison of their carbohydrate content. The instruments are developed on the Carbohydrate Structure Database (CSDB) platform and are available freely on the CSDB web-site at http://csdb.glycoscience.ru. Database URL: http://csdb.glycoscience.ru.

Figure 1.

Fragment abundance form.

Figure 2.

Monomeric composition for S. japonicus and S. pombe.

Figure 3.

Fragment abundance form. Only fragments unique for the genus Eleutherococcus in its phylum will be processed.

Figure 4.

Dimeric fragments unique for the Eleutherococcus genus in its phylum.

Figure 5.

Coverage statistics form.

Figure 6.

Coverage statistics for the Actinobacteria and Firmicutes phyla.

Figure 7.

Parameter input for clustering of taxa by glycan structural features.

Figure 8.

Clustering result window.

Figure 9.

CSDB coverage by structures and organisms. Absolute numbers of structures/organisms are provided for every domain/phylum/class. The inner diagram shows distribution of structures among domains split into lower ranks; the outer ring shows distribution of organisms among domains. The following color code is used: blue shades, bacteria; red shades, fungi; green shades, plants; orange, archaea, algae and protista.

Figure 10.

Thirty most widespread monomeric residues in carbohydrate structures from major taxonomic groups. Bubble area corresponds to averaged monomer frequency in the domain (see text), varying from 9.7 (aDManp, α-D-mannopyranose) to 0.12 (bDFruf, β-D-fructofuranose). Amino sugar residues include both acetylated and non-acetylated forms and are highlighted in lilac; undetermined residues with unknown anomeric, absolute or ring size configuration are shaded.

Figure 11.

Thirty most widespread dimers in carbohydrate structures from major taxonomic groups. Bubble area corresponds to disaccharide frequency in the domain (see text), varying from 3.15 (aDManp(1-2)aDManp) to 0.09 (aDGlcp(1-3)aDGlcp). Dimers containing amino sugars, which include both acetylated and non-acetylated forms, are highlighted in lilac; dimers containing undetermined residues with unknown anomeric or ring size configuration are shaded.

Figure 12.

Seventeen most widespread dimeric fragments containing non-sugar moieties in plant saccharides. Numbers are absolute abundance in plant structures.

Figure 13.

Glycome-based clustering of most studied genera. Shades of blue represent various bacterial taxonomic groups, shades of green represent plant groups, red and orange represent fungi and protista domains, respectively. The outer arc for bacteria is colored according to the Gram-reaction. Size of the circles reflects the normalized popularity of a given genus in CSDB in terms of assigned organisms (green) and assigned structures (blue). When a blue circle is not visible, it is the same size as a green one. Color of labels denotes the database from which the data came.

Figure 14.

Flow chart of data processing in the reported online tool.

Figure 15.

Circular phenetic trees based on dimers (containing monosaccharides, aliases, aglycons, monovalent residues) present in compounds from bacterial species most populated in BCSDB. Firmicutes are shown in red, Actinobacteria in green, Enterobacteria in cyan and other Proteobacteria in violet. Three dendrograms correspond to different clustering methods: NJ (A), Ward’s minimal variance (B) and balanced minimum evolution (C). Blue and green circles in (A) depict the normalized number of structures and organisms assigned to each species, correspondingly. When a blue circle is invisible, it is the same size as a green one. Gray color code in the outer rim (A) reflects the Gram-reaction. The underlying numbers and trees are available in Supplementary materials.

Figure 16.

Phylogenetic tree based on small ribosomal subunit rRNA sequences. Firmicutes are shown in red, Actinobacteria in green, Enterobacteria in cyan and other Proteobacteria in violet. The underlying data are available in the Supplementary materials.