Based at MD Anderson Cancer Center, we focus on the genetics/epigenetics of bacteria associated with human cancers, in particular those that infiltrate tumors.
New research from
@BullmanLab
/
@CJohnstonLab1
, with superstar 1st author
@mzepedar
, out now in
@Nature
pinpoints “high-risk” subtype of Fusobacterium nucleatum that dominates the colorectal
cancer (CRC) niche in human patient's
Link :
Tweet🧵below:
Finally, two papers out within a day of each other from the
@BullmanLab
and
@CJohnstonLab1
relating to the intratumoral microbiota, today in
@Nature
() and yesterday in
@CellReports
(). 🧵below to those interested!
Enjoyed working on this with
@BullmanLab
, and Laurence Zitvogel group
@GustaveRoussy
. Here in
@NatureCancer
we discuss the potential of Fecal Microbiota Transplantation to modulate microbiota for improved immunotherapy response in cancer patients.
Here, in
@NatRevGastroHep
, we and
@BullmanLab
@fredhutch
discuss current concepts relating to the intratumoral microbiota and what is needed to gain mechanistic insights on the role of bacteria in human tumors:
Wonderful and thought provoking morning seminar
@fredhutch
@HutchMRI
by Prof. John Cryan
@UCC
on the microbiota as a key regulator of physiology and behaviour. Thanks for joining us (remotely) in Seattle
@jfcryan
First, our
@Nature
manuscript, sterling work from first author Jorge Galeano Ninõ, and coauthors
@kaidenlac
,
@HanruiWu233
, Sam Minot
@fredhutch
and more. Here we show that the intratumoral microbiota impacts spatial, cellular, and transcriptional heterogeneity in human cancers.
Excited to share preprint of our recent collaboration with
@BullmanLab
,work led by
@kaidenlac
'The cancer chemotherapeutic 5-fluorouracil is a potent inhibitor of Fusobacterium nucleatum and its activity is modified by the intratumoral microbiota'
The Johnston lab at
@fredhutch
is seeking two outstanding Staff Scientists (one wetlab/one drylab) to advance our epigenetics/synthetic microbiology/metagenomic projects.
If you could RT to your network would greatly appreciate it! Links below
Collectively, this data reveals that the distribution of the microbiota within a tumor is not random; instead, it is highly organized in microniches with immune and epithelial cell functions that promote cancer progression.
We then asked, what is happening in these microbial microniches? We coupled targeted RNAscope
@ACDbio
and
@nanostringtech
protein digital spatial profiling and see that these regions are highly immunosuppressive, with myeloid cell infiltration and T cell exclusion.
Key finding is that the microbiota is not evenly distributed across the tumor tissue, but are spatially localized in distinct microniches or ‘hotspots’ (this has implications for bulk tissue microbiome analysis in human tumor specimens)
In 3D cancer spheroids (vid above), Fuso infected cells invade surrounding environment as single cells & recruit myeloid cells. Without Fuso, the cancer spheroid grows en masse, but with Fuso they migrate as single cells, bringing intracellular Fuso (implications for metastasis)
We adapted 10x Visium spatial transcriptomics
@10xGenomics
to determine the identity and in situ location of intratumoral microbial communities within patient oral squamous cell carcinoma (OSCC) and colorectal cancer (CRC) tissues.
We then adapted
@10xGenomics
5’ singlecell RNAseq for simultaneous detection of host and bacterial transcripts (which we call INVADEseq). This reveals the identity of intracellular bacteria, the host cells they infect, and their impact on host gene expression!
We know that there are microbes in human tumors (intratumoral microbiota), but where are they physically located? What human cells do they interact with? And what is the consequence of these interactions? This is what we set out (3 years ago) to answer.
Our work with
@KLemonLab
led by
@Steph_BloomingT
on pangenomics/epigenomics of Dolosigranulum pigrum (candidate beneficial bacterium from the human microbiota) is now published in mSystems. Included my personal favorite image of a pangenome via PPanGGOLiN:
We're hiring at multiple levels (Tech/Postdoc/Staff Scientist). Interested in pan-epigenomics (bacterial DNA methylation across human microbiota)?Or want to build new dCas9-effector tools to examine human-microbe epigenetic crosstalk? Consider applying:
With high-resolution complete
@PacBio
genomes in hand, we could go deeper. To our surprise we saw something
interesting, Fna seemed to bifurcate into two distinct lineages or clades (Fna C1 and Fna C2). One
seemingly limited to the oral cavity and the other enriched in CRC.
Applying INVADEseq to OSCC tumors, Fusobacterium & Treponema are dominant intracellular bacteria predominantly within cancer epithelial cells (aneuploid) and macrophage. Bacteria+ve single cells vs Bacteria-ve had upreg of inflammation,
metastasis, dormancy & DNA repair pathways
Last summer my lab and
@BullmanLab
hosted (remotely) two outstanding
@tcddublin
@GeneticsTCD
undergrads Adam McGlinchey and Marija Stepanovica, mentored by
@martha_a_zepeda
. Incredible productivity by all with 2x manuscripts:
And
The
@BullmanLab
, cultured 65 unique Fusobacterium strains directly from 59 patients CRC tumors,
@ForsythResearch
/ KCOM grew 81 Fusobacterium strains from oral cavity of healthy ppl, and we used
@PacBio
SMRTseq to generate complete closed genomes and methylomes for every one!
Our collaborative work with
@KLemonLab
led by
@Steph_BloomingT
'Genomic Stability and Genetic Defense systems in Dolosigranulum pigrim a Candidate Beneficial bacterium from the Human Microbiome' is submitted and available
@biorxiv_micrbio
. Some highlights:
With outstanding assistance from
@nsegata
,
@cibiocm
, we used metagenomic analysis of stool samples from more than 1,200 people,
(n=616 healthy, and n=596 patients with CRC). Again, it was only Fna C2 significantly enriched!
Nice description and application of our 2019 SyngenicDNA method in Bifidobacterium breve, (although we are not cited? Science-burn from my own home city Cork!
@VanSinderenLab
@Pharmabiotic
)
In the paper, we show that these clades are genetically, epigenetically, and phenotypically distinct. The
perfect comparative group to pull apart the genetics of tumor homing Fn capabilities! 192 factors enriched in Fna C2.
Pangenome network visualization plot shown
@axbazin
Oh no, I forgot to acknowledge the beautiful work by scientific illustrator
@markabelan
for this image. It was a pleasure talking science and perspective with him, and highly recommend working with
@ArtsciStudios
!
We identified factors that may influence this stability, including phage, the role of MGEs in strain-level variation & defenses against MGEs. My favourite is a conserved integrative hotspot across strains containing 13 different RM systems (plus 3
@SorekLab
deity systems)
Further, in mouse models for CRC, we show that only Fna C2, and not Fna C1, can increase the number
of large intestinal tumors and alter intestinal metabolism.
We then generated ex-vivo CRC communities and exposed to 5-FU. There was considerable loss of species diversity in a subset of CRC patient communities; suggesting that in addition to Fn, other members of intrumoral microbiota may be sensitive to the drug.
Other members of intratumoral microbiota from CRC specimens diminish 5-FU activity and can rescue Fn and human CRC epithelial cells(!) from 5-FU toxicity. RNAscope visuals of intratumoral communities show they are tissue invasive and surround cancer cells within patient tumors.
So why does this matter? We've found the exact lineage of Fn that is driving the enrichment of this bacterium with human CRC. The Fna C2 lineage is therefore the highly virulent subgroup of Fn that should be the primary focus for mechanistic studies and therapeutic drug design!!
Next we questioned if we saw the same with culture independent approaches, and in matched CRC
tumor and adjacent normal patient tissues, we found that of all Fusobacterium assessed, only Fna C2
was significantly enriched in patient tumors. How about stool?
We confirmed that 5-FU is a potent inhibitor of multiple clinical isolates representing all subspecies of Fn at physiologically relevant concentrations. But, 5-FU does not have pan-antimicrobial activity.
Fn in a healthy mouth or in oral disease are genetically
diverse, with multiple subspecies present; Fn ssp. animalis (Fna), Fn ssp. polymorphum
(Fnp), Fn ssp. vincentii (Fnv) and Fn ssp. nucleatum (Fnn). It was unknown whether all of these ssp. colonize CRC tumors or not!
Through mass-spec analysis, we show that 50% of the patient tumors analyzed harbored an intratumoral microbiota capable of diminishing 5-FU levels within 48 hrs.
Fusobacterium nucleatum (Fn), is normally found in the mouth (rarely found in the lower gastrointestinal tract of healthy ppl, is known to be enriched in human colorectal cancer (CRC) tumors. High Fn
tumor load is consistently linked with treatment resistance + poorer outcomes.
The genomic stability, antibiotic susceptibility, capacity for defense against HGT and lack of known virulence factors across D.pigrum genomes combined with inhibition of nasal pathobionts support it as a candidate for future therapeutic/prophylactic use.
@KLemonLab
@ettwiller
@biorxiv
@illumina
Ah, I really meant 'commercially available at a core with back end pipeline ready to go'. Given the double whammy of m5c detection and hybrid/polishing with illumina reads, I'd certainly supplement all our WG Pacbio runs with this.
This propelled a large scale pangenomic analyses of Fn in cancer and in health, and we found only one of the mentioned subspecies as significantly enriched in CRC tumor tissues – Fn subspecies animalis (Fna)
F. nucleatum is among the most prevalent and dominant bacterial species found in colorectal cancer (CRC) tumor tissues, and this is true across multiple patient cohorts worldwide. Cell culture and animal models support its role in CRC progression.
Explored D.pigrum's pangenome bc nasal microbiota dominated by Dpig is less likely to be colonized by nasal pathobionts, lowering risk of invasive infection. It's inversely associated with S.aureus in adult nostrils and infants with lots of D.pig+Coryne get fewer RT infections.
To support such efforts by the collective Fuso field, we (
@seajaneCchange
,
@mzepedar
) created an interactive website allowing the exploration of Fna pangenome datasets named Fusopangea.
Both spent the summer characterizing the genetics of bacterial isolates from human tumors. Helps when you have an in-lab
@PacBio
Sequel instrument for closed genomes/plasmids/methylomes and great collaborators pulling new species and strains from the intratumoral microbiota.
Very excited to share the news that the W.M. Keck Foundation has funded our super high-risk application that focuses on the potential of host-microbe epigenetic crosstalk. A collaboration two years in the making with
@BullmanLab
and
@Angela_Ting_PhD
.
Considering the intimate association of Fn and CRC, these results suggest that at least part of chemotherapeutic effects of 5-FUcould be via anti-microbial activity against Fn. This also reveals that tumor micriota members, notably E.coli can deplete 5fu to limit its efficacy.
Previously, in 2022, with
@BullmanLab
, we showed that intratumoral Fn could modulate the tumor microenvironment at the spatial and cellular level, supporting cancer progression in oral squamous cell carcinoma (OSCC) and colorectal cancer (CRC) ()
"In development of phages for bioengineering and therapeutics it will be important to consider that nucleic acids of introduced phages may spread into local phage populations through recombination, and the likelihood of transfer is not predictable based on killing host range"
Analysis of 28 genomes (19
@PacBio
completed, plus methylomes) revealed remarkable chromosomal stability among nasal D.pigrum strains over 20-year span. The beautiful image above is a partitioned pangenome graph, generated using PPanGGOLIN (
@vallenet
)
The Johnston Lab at
#Fredhutch
in Seattle is hiring! We're recruiting a Bioinformatics Analyst (level commensurate with experience) for Computational Analyses of Microbial Genomic/Metagenomic/Epigenomic Data.
#job
We used a small-molecule screen of 1,846 bioactive compounds to ID inhibitors for a tumor isolate of this oncomicrobe. Surprisingly, 15% of inhibitors identified were antineoplastic agents incl 5-fluorouracil (5-FU), the first-line chemotherapeutic used to treat CRC worldwide.
Is that gene essential?..kinda sorta..
@RosconiA
and
@tvanop
investigate how a bacterial pan-genome might influence gene essentiality, and whether essential genes that are initially critical for the survival of an organism can evolve to become non-essential.
Our PPanGGOLiN figure describing the synteny of D. pigrum from our Flores Ramos et al. 2021 paper is the featured image for the latest issue of
@ASMicrobiology
! Thanks to
@KLemonLab
@CJohnstonLab1
& our many other collaborators for getting this story out
@pentamorfico
@NAR_Open
Awesome work, I just stumbled across 148 of these in a pangenome project yesterday and had no idea what to make of them. Then this pops up. Great timing.