TY - JOUR
T1 - CAZyme fold architecture is conserved between disparate environments despite extreme sequence divergence
AU - Jeilu, Oliyad
AU - Simachew, Addis
AU - Hartmann, Erica M.
AU - Alexandersson, Erik
AU - Johansson, Eva
N1 - Publisher Copyright:
© 2026 Jeilu et al.
PY - 2026/6
Y1 - 2026/6
N2 - Microbial carbohydrate-active enzymes (CAZymes) underpin carbon cycling across Earth’s ecosystems; however, how contrasting environments shape CAZyme diversity and structural conservation remains poorly understood. Here, we applied shotgun metagenomics to compare the carbohydrate-degradation potential of two functionally prolific but physicochemically opposed ecosystems: the alkaline-saline soda lakes of the East African Rift Valley and the anaerobic ruminant gut. From 34 metagenomes (12 soda lake and 22 rumen), we recovered 371 quality-filtered metagenome-assembled genomes, of which 84% of soda lake and 52% of rumen MAGs represented novel species. Rumen communities, dominated by Bacteroidota, Fibrobacterota, and Bacillota, exhibited significantly higher taxonomic diversity and were enriched in carbohydrate catabolism and fermentation pathways. Soda lake communities, dominated by Pseudomonadota, displayed greater evolutionary divergence (lower RED scores) and were enriched in pH homeostasis, oxidative and osmotic stress, sulfur cycling, and carbon fixation pathways. To assess whether structural conservation persists despite extreme sequence divergence, we predicted three-dimensional structures for 12 representative enzymes from six glycoside hydrolase families (GH1, GH3, GH5_11, GH9, GH10, and GH28) using AlphaFold 3. All 12 structures adopted canonical GH family folds with high confidence (pTM 0.75–0.97). These results demonstrate that environmental selection drives distinct taxonomic and functional strategies for carbon processing while preserving three-dimensional CAZyme architecture, positioning soda lake and rumen metagenomes as complementary reservoirs for bioprospecting industrially relevant enzymes.
AB - Microbial carbohydrate-active enzymes (CAZymes) underpin carbon cycling across Earth’s ecosystems; however, how contrasting environments shape CAZyme diversity and structural conservation remains poorly understood. Here, we applied shotgun metagenomics to compare the carbohydrate-degradation potential of two functionally prolific but physicochemically opposed ecosystems: the alkaline-saline soda lakes of the East African Rift Valley and the anaerobic ruminant gut. From 34 metagenomes (12 soda lake and 22 rumen), we recovered 371 quality-filtered metagenome-assembled genomes, of which 84% of soda lake and 52% of rumen MAGs represented novel species. Rumen communities, dominated by Bacteroidota, Fibrobacterota, and Bacillota, exhibited significantly higher taxonomic diversity and were enriched in carbohydrate catabolism and fermentation pathways. Soda lake communities, dominated by Pseudomonadota, displayed greater evolutionary divergence (lower RED scores) and were enriched in pH homeostasis, oxidative and osmotic stress, sulfur cycling, and carbon fixation pathways. To assess whether structural conservation persists despite extreme sequence divergence, we predicted three-dimensional structures for 12 representative enzymes from six glycoside hydrolase families (GH1, GH3, GH5_11, GH9, GH10, and GH28) using AlphaFold 3. All 12 structures adopted canonical GH family folds with high confidence (pTM 0.75–0.97). These results demonstrate that environmental selection drives distinct taxonomic and functional strategies for carbon processing while preserving three-dimensional CAZyme architecture, positioning soda lake and rumen metagenomes as complementary reservoirs for bioprospecting industrially relevant enzymes.
KW - biogeochemical cycles
KW - carbon cycling
KW - CAZymes
KW - MAGs
KW - microbial diversity
KW - rumen microbiome
KW - shotgun metagenomics
KW - soda lakes
UR - https://www.scopus.com/pages/publications/105044484782
UR - https://res.slu.se/id/publ/094c07e9-03ed-453f-ab59-0e92d161b1f5
U2 - 10.1128/msystems.00485-26
DO - 10.1128/msystems.00485-26
M3 - Journal article
C2 - 42153643
AN - SCOPUS:105044484782
SN - 2379-5077
VL - 11
JO - mSystems
JF - mSystems
IS - 6
M1 - e00485-26
ER -