A polysaccharide called rhamnogalacturonan II is a major component of some fruits, but humans rely on their gut microbiota to digest it. The microbes and processes responsible for this digestion have now been revealed. See Article p.65
When you sip a glass of wine, you might be feeding some of your gut bacteria their favourite food: a polysaccharide named rhamnogalacturonan II (RGII). This is because humans do not produce the enzymes needed to break down most of the polysaccharides in plant-based food1 into single sugar molecules, whereas some bacteria in the human gut flora (the microbiota) do. But precisely how the complex components of fruit cell walls, such as RGII, are decomposed in the human gut, and which microbes are responsible, was not known. Ndeh et al.2 report on page 65 that, surprisingly, bacterial consortia are not required — single bacterial strains common in most human gut microbiota possess the entire enzymatic system needed to break down and metabolize RGII.
It has become clear that human nutrition affects the diversity of the gut microbiota and directly influences health3. But many questions remain about the interplay of different microbial strains in polysaccharide digestion and the underlying molecular mechanisms. RGII, for example, constitutes up to 15% of the cell-wall components of some fruit, particularly grapes. Do different types of mutually beneficial microbes work together to cleave the 21 types of glycosidic bond that connect the individual sugars in RGII (ref. 4) from grapes? Or have highly specialized bacterial phyla developed an enzyme system that breaks down all the bonds?
Ndeh and colleagues used a combination of biochemical, crystallographic and microbial studies to work out the key mechanistic and functional aspects of RGII decomposition. Their detailed analyses led to the discovery of seven families of glycoside hydrolase enzymes (GHs) that cleave glycosidic bonds for substrates that were not known to be cleaved by GHs. They also report several previously unknown substrate specificities for existing GH families. Strikingly, the authors' detailed study of the specificity and mode of action of the newly discovered enzymes reveals that some of the chemical structures thought to be present in RGII were incorrect (Fig. 1).
The tremendous complexity of naturally occurring polysaccharides is one of the main reasons why the biochemical characterization of these compounds still lags far behind that of other naturally occurring polymers. The polysaccharide 'alphabet' is much more complicated than those of DNA or proteins, consisting of about 120 naturally occurring monosaccharides that can be linked and branched in many ways. We do not yet have the methods and techniques to easily access the sequences and detailed structures of polysaccharides, so such analyses remain difficult and time-consuming.
One approach that helps to simplify sequencing and structure determinations, and that was used by Ndeh et al. in their study, is to expose polysaccharides to GHs that break a known type of glycosidic bond. Characterization of the products of several different GH reactions provides clues about the structure of the original polysaccharide. This strategy nevertheless still requires a lot of effort, because many different enzymes and characterizations are needed to identify the structures of complex polysaccharides, as indeed the authors found for RGII.
A particularly impressive aspect of Ndeh and colleagues' work is that they produced and isolated oligosaccharides (short polysaccharides that contain two to ten sugar units) that formed as intermediates in the depolymerization of RGII. They did this either by feeding RGII to mutant bacterial strains that lacked one of the key enzymes in the proposed depolymerization pathway, or by incubating RGII (or oligosaccharides produced from RGII by the mutants) with purified enzymes from the pathway. Once purified and characterized, the oligosaccharides could themselves be used as substrates to probe the specificity of other enzymes along the reaction pathway. This approach allowed Ndeh et al. not only to decipher the entire depolymerization pathway, but also to propose a model for the bacterial deconstruction of RGII.
The huge and rapidly growing carbohydrate-enzyme database5 CAZy groups enzymes that cleave glycosidic bonds (GHs and others) into families on the basis of their amino-acid sequences, and provides valuable information about the key amino-acid residues in active sites, catalytic mechanisms and overall 3D structures of these enzymes. Nevertheless, assigning which glycosidic bonds are cleaved by which family remains challenging, because individual members of most families often have different substrate specificities. Moreover, even though the database contains several hundreds of thousands of sequences, a full biochemical characterization — knowledge of an enzyme's substrate, of the details of the catalytic reaction mechanism, of the sugar units on either side of the glycosidic linkage that is cleaved and of where the linkage is found in the substrate — is known for only about 2% of them2. Ndeh and colleagues' characterization of the enzymes involved in the breakdown of RGII thus provides useful tools for future research: the enzymes can now be used to cleave specific linkages in other complex polysaccharides, and the potential biological roles of the oligosaccharides produced during RGII breakdown can be investigated.
Notably, Ndeh and co-workers identified and described the sophisticated machinery that metabolizes RGII in the genome of the human-gut bacterium Bacteroides thetaiotaomicron, a member of the phylum Bacteroidetes. Members of this phylum are major components of animal microbiota, especially in the gastrointestinal tract. These physiologically diverse bacteria express many enzymes that use carbohydrates as substrates, and are generally considered to be the main players in polysaccharide degradation. They live in numerous ecological niches where carbohydrate use is inextricably linked to the ability of microbes to survive.
Bacteroidetes genomes contain polysaccharide utilization loci6 (PULs) — strictly regulated, co-localized gene clusters that encode all the enzymes needed to break down a given polysaccharide. Since the discovery7 of PULs in 1989, considerable progress has been made in working out their functions and regulation to gain insight into their impact on human health and how their expression is affected by human nutritional habits (see ref. 8 for a review). For example, pioneering work9 on the differential use of polysaccharides called fructans by several symbiotic Bacteroides species in the human gut showed that the bacteria rely on enzymes that cleave specific glycosidic linkages, and that these enzymes are instrumental in defining the nutritional preferences of the bacteria for polysaccharides.
The preference of different bacteria for degrading certain polysaccharides might have a central role in shaping the relationships between organisms of the microbiota. Ndeh and co-workers show that three different PULs induced by RGII in B. thetaiotaomicron are adapted to cope with structural variations in the polysaccharide that are associated with RGII's diverse plant origins. One PUL allows the microbes to grow on apple RGII, for example, whereas another is needed to break down the RGII in wine. Microbes that possess all of the PULs will have a net advantage in the battle for these coveted substrates.
Several questions arising from the new work remain unanswered. For example, do the fragments of RGII released by its depolymerization regulate or influence the presence of other bacterial phyla in the microbiota? And are some of the rare monosaccharides produced from RGII degradation beneficial to humans? Further work is needed to address these issues. Footnote 1
El Kaoutari, A., Armougom, F., Gordon, J. I., Raoult, D. & Henrissat, B. Nature Rev. Microbiol. 11, 497–504 (2013).
Ndeh, D. et al. Nature 544, 65–70 (2017).
The Human Microbiome Project Consortium. Nature 486, 207–214 (2012).
Pellerin, P. et al. Carbohydr. Res. 290, 183–197 (1996).
Lombard, V., Ramulu, H. G., Drula, E., Coutinho, P. M. & Henrissat, B. Nucleic Acids Res. 42, D490–D495 (2014).
Bjursell, M. K., Martens, E. C. & Gordon, J. I. J. Biol. Chem. 281, 36269–36279 (2006).
Anderson, K. L. & Salyers, A. A. J. Bacteriol. 171, 3199–3204 (1989).
Grondin, J. M., Tamura, K., Déjean, G., Abbott, D. W. & Brumer, H. J. Bacteriol. http://dx.doi.org/10.1128/JB.00860-16 (2017).
Sonnenburg, E. D. et al. Cell 141, 1241–1252 (2010).
Related links in Nature Research
About this article
Analytical and Bioanalytical Chemistry (2020)
Journal of Analysis and Testing (2019)
Pathways from epigenomics and glycobiology towards novel biomarkers of addiction and its radical cure
Medical Hypotheses (2018)
Nature Communications (2018)