Page 9 - Read Online

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

Given in vitro phenotypic differences between phylogroups that could be linked to interactions with its host,
[20]
such as TLR stimulation and adherence to epithelial cells , we postulate that which Akkermansia species is
present in the GI tract may influence the ultimate impact on its human host. Indeed, one study found that
A. muciniphila could be distinguished from Akkermansia candidate species based on marker genes from
metagenome-assembled genomes . Based on these markers, the authors suggested that the association
[25]
observed between A. muciniphila and body mass index (BMI) did not apply to the additional candidate
species. However, current analysis tools and databases [32,33] do not, by default, distinguish between these
Akkermansia phylogroups. In this study, we combined all the previously, but independently, described
genomes from isolated strains into a single, large pangenome. Our pangenomic analysis defined a novel
clade of Akkermansia and supported a species re-assignment for phylogroups AmII, IV, and V, which we
corroborate with phenotypic characterizations of representative strains for each species and subspecies.
Using these new classifications, we developed methods to identify Akkermansia species and subspecies from
16S rRNA and metagenomic sequences of stool samples. Finally, we provide three case studies to further
support the premise that the relationship between the relative abundance of Akkermansia and health
outcomes can correlate with specific Akkermansia phylogroups.

METHODS
Isolation of Akkermansia from fecal samples
Bacterial isolation and growth were performed in a Coy Laboratory anaerobic chamber under 5% hydrogen,
5% carbon dioxide, and 90% nitrogen. Eight Akkermansia strains were isolated from the stool of renal cell
carcinoma (RCC) patients at Duke University Hospital (IRB protocol Pro00076768). Isolation of strains was
[20]
performed as described previously . Four Akkermansia strains were isolated from the stool of patients with
amyotrophic lateral sclerosis (people with ALS, PALS) at Duke University Hospital (IRB protocol
Pro00108282). To isolate new Akkermansia strains, approximately 100 mg of frozen stool was used to
inoculate 1 mL of mucin medium as previously described [1,20] , and supplemented with cysteine (0.5 mM),
vancomycin (6 μg/mL), gentamicin (10 μg/mL), and kanamycin (12 μg/mL) and incubated at 37 °C for 48 h.
Three sequential passages at 1:10 dilutions were performed, and a sample was streaked on 1% agar mucin
medium plates, supplemented with cysteine (0.5 mM), to facilitate the isolation of single colonies. After
7 days at 37 °C, individual colonies were picked and cultured in synthetic media [5,20] supplemented with
cysteine (0.5 mM).

Genome sequencing and annotation
Sequencing and annotation of genomes labeled as “RCC” were performed as described previously . For
[20]
“PALS” genomes, genomic DNA extractions were carried out using the DNeasy Blood & Tissue kit (Qiagen
catalog 69504) following the manufacturer’s protocol. DNA concentrations were determined using the
Qubit double-stranded DNA (dsDNA) broad-range kit (Thermo Scientific). Library preparation and
sequencing were performed by SeqCenter (https://www.seqcenter.com/). The Oxford Nanopore
Technologies (ONT) Ligation Sequencing Kit (SQK-NBD114.24) with NEBNext® Companion Module
(E7180L) was used to prepare DNA sequencing libraries, adhering to the manufacturer’s specifications.
Genome sequencing was carried out on Nanopore R10.4.1 flow cells using the MinION Mk1B device. The
Oxford Nanopore data processing toolkit, Guppy (v6.3.8), was used for base calling and demultiplexing.
The super accurate model was used for base calling, as it offers the highest raw read accuracy . The
[34]
resulting fastq files were assembled into a single contig using Flye . In addition, we used an Illumina DNA
[35]
Prep kit and IDT 10bp UDI indices to prepare additional libraries, which were then sequenced on an
Illumina NextSeq 2000, generating 2 × 151 bp reads. The Illumina software app bcl-convert (v3.9.3) is
provided to convert files produced by Illumina systems into FASTQ format. We used this software to
[36]
perform demultiplexing, quality control, and adapter trimming of reads . To refine the Flye assembled
genome, we used Pilon in combination with the Illumina reads for autocorrection . The assembly
[37]
underwent annotation and quality evaluation using the PATRIC genome annotation service .
[38]