Page 14 - Read Online

Page 8 of 18 Mueller et al. Microbiome Res Rep 2024;3:33 https://dx.doi.org/10.20517/mrr.2024.09

Figure 2. Akkermansia species are distinguishable by their estimated metabolic capabilities and fatty acid composition. (A) Unique core
gene clusters were defined as those present in all genomes of one species, but not present in every genome of another species. Shared
core gene clusters (n = 1,192) were defined as those present in all genomes of multiple species. All core gene clusters were categorized
by these definitions and tallied to define the sizes of the A. muciniphila, A. massiliensis, and A. biwaensis core genomes; (B) Estimated
metabolic enrichment analysis indicates the gain or loss of some metabolic pathways in Akkermansia phylogroups. Blue indicates
pathway presence, calculated by pathway completeness of 50% or greater; (C) A principal components analysis was performed to
compare fatty acid composition across species. The principal component scores distinguish three clusters comprised primarily of either
A. muciniphila and AmV, A. massiliensis, or A. biwaensis and AmVI, as indicated by dotted ellipses. Isolates are colored by species; (D)
Average fatty acid composition by C14:0 iso and C15:0 iso distinguishes A. massiliensis from other Akkermansia species.

Next, we used the API 20A system to perform biochemical phenotyping on six A. muciniphila (including
Muc , isolated from humans), two AmV (isolated from mice), three A. massiliensis (isolated from humans),
T
four A. biwaensis (isolated from humans), and one AmVI (isolated from humans) isolate [Supplementary
Table 5]. Out of the 21 assays performed, 15 yielded negative results for all isolates. All isolates were positive
for glucose and lactose utilization, and the remaining four assays yielded varied results. Mannitol utilization
was absent in all AmV, A. massiliensis, and AmVI isolates, but varied between isolates of A. muciniphila and
A. biwaensis. Maltose utilization occurred in both AmV isolates, not in the AmVI isolate, and varied
between isolates of A. muciniphila, A. massiliensis, and A. biwaensis. All isolates of AmV and A. biwaensis
could use mannose but not the AmVI isolate, and mannose utilization varied between isolates of
A. muciniphila and A. massiliensis. Catalase activity was present in all isolates of AmV and A. massiliensis,
absent in the AmVI isolate, and varied between isolates of A. muciniphila and A. biwaensis. Interestingly,
this variability in catalase activity correlates to the variable, low, and high sensitivity to oxygen
demonstrated by A. muciniphila, A. massiliensis, and A. biwaensis in a prior study .
[20]

Akkermansia species and subspecies-level assignments can be made from metagenomic
sequences and 16S rRNA V3-V4 regions
Akkermansia species and phylogroups have distinct metabolic, in vitro phenotypic, and in vivo competitive
characteristics [20-23] that could be linked to health or disease risks. We next asked if we could use existing 16S
rRNA and metagenomic sequencing data generated from various patient cohorts to clarify the relationships
between Akkermansia species and human health. We leveraged our pangenomic analysis results to enhance