Scott Bultman, Department of Genetics, Lineberger Comprehensive Cancer Center, UNC-Chapel Hill

Bridging nutrition, metabolism, the microbiome, and cancer susceptibility

Nutrition

Immune/Inflammation

Metabolism Genetic background

Cancer susceptibility

Microbiome

Lumen Intestinal Epithelium Lamina Propria

Gut microbiota

B cell T cell

Macrophage Dendritic cell Immune cells

Fiber

+ Dietary/digestive components Metabolites that influence cancer =

Polyphenols

Meat Bile acids Oncometabolites: 2° bile acids (DCA), hydrogen sulfide Tumor-suppressive metabolites: SCFAs (butyrate), equol, urolithins

Diet & Microbiota Influence Cancer Risk and Progression

Fiber-Microbiota-Butyrate Axis

Fiber

Bacterial fermentation

Butyrate Mitochondria: Energetics

Nucleus: Epigenetics

SCFAs FAO

HDACi Ligand for GPCRs

Evidence for Butyrate & Cancer Prevention Tumor-derived cell lines: Microbiome studies:

Butyrate Proliferation Apoptosis Differentiation

Butyrate producers: CRC cases < controls

A Limitation of Metagenomic Studies

Microbiota composition: Cases vs controls

Association study: ∆microbiome = cause or consequence?

Gnotobiotic Facility

GEMMs

Metagenomic Sequencing

Next Phase of Microbiome Research Bacterial Culture

from cataloging to function

Does Dietary Fiber Protect Against Colorectal Cancer (CRC)?

Complex polysaccharides

Butyrate

Maintains digestive movement

(insoluble)

(soluble)

Two proposed general mechanisms:

But …

#2

#1

…Lack of Universal Consensus: Conflicting Results from Human Epidemiology Studies

Not Protective Protective

Fuchs et. al. 1999

Lanza et. al. 2002

Park et. al. 2005

Lanza et. al. 2007

Ferguson et. al. 2003

Bingham et. al. 2003

Magalhães et. al. 2011

Janike et. al. 2011

grain/fiber

Ecologic and migration supportive; prospective cohort mixed:

Possible Reasons for Conflicting Results

Various fiber sources

Differences in microbiota

Other dietary factors

Compliance

Hypothesis: Dietary fiber protects against CRC in a microbiota- and butyrate-dependent manner.

Genetic heterogeneity

Experimental Approach- Part 1

5 months

AOM Tumors Resemble Human CRC

mouse

human CRC T S M A ED LD adult

normal human

4.00

0.25

1.00

Kaiser et al. (2007) Genome Biol.

Similar mutation spectrum: Similar transcriptome profiles:

Apc Kras Braf β Catenin p53

“Vogelgram”

Experimental Approach- Part 2

Control Control Control Experimental

AOM AOM

Low Fiber Diet Low Fiber Diet High Fiber Diet High Fiber Diet

Butyrate

Hypothesis: tumor burden

Combination of High Fiber & B. fib Confer Tumor Suppressive Effect

High Dose (5 AOM)

Experimental: Diminished incidence & multiplicity

Tum

ors/

mou

se

*

Combination of High Fiber & B. fib Also Inhibits Tumor Growth & Progression

Experimental: Diminished size

*

Combination of High Fiber & B. fib Also Inhibits Tumor Growth & Progression

Experimental: Less progressed

High Grade Low Grade

A Mutant Strain of B. fibrisolvens to Confirm Butyrate Mechanism

Asanuma et al. (2005) Current Microbiology

Butyryl-CoA synthesis operon:

Mutant

Wild type

0.8-kb deletion

Butyrate is a Causal Factor

High Dose (5 AOM)

Tum

ors/

mou

se

Attenuated protective effect

Stronger protective effect

* *

+

# # # #

What is Butyrate Doing Mechanistically?

#1 It accumulates in tumors…

LC-MS:

Tumor

Control Groups

HFD WT B. fib

TB (-) B. fib

Buty

rate

(pm

ols/

µg o

f pro

tein

)

**

*

* *

Tumor Normal

What is Butyrate Doing Mechanistically?

#2 …which increases histone acetylation due to HDAC inhibition and…

Tumor

H3

H3ac

HFD LFD TB HFD LFD

(-) B. fib WT B. fib

IHC: Western Blot:

What is Butyrate Doing Mechanistically? #3 … cell proliferation and apoptosis

Accumulation HDAC inhibitor Epigenetic gene regulation Cell proliferation Apoptosis (Ki-67) (cleaved caspase 3)

Q: Why does butyrate accumulate in tumors in first place?

( p21)

Fewer/smaller tumors

( Fas)

A: Warburg Effect- Part 1 of 2

LC-MS:

*

Normal Tumor

*

LDHA IHC:

Tumor

Tumors undergo Warburg effect …

A: Warburg Effect- Part 2 of 2

Butyrate Proliferation +Warburg

-Warburg

… butyrate dependent on Warburg effect in cell culture model

(siLDHA, glutamine>glucose)

Current Working Model

Fiber

Cell proliferation Apoptosis

Butyrate

Treg cells & anti-inflammatory effects, barrier function, GPCR signaling

Caveats: • Approach is reductionistic for diet and microbiome. One genetic bkg. • Model is simplistic. Focused on cancer cell autonomous effects.

Nevertheless …

… Human Cancer Relevance

LC-MS: H3ac WBA:

*

H3ac

H3

Normal Colon Adenocarcinoma 1 2 1 2

Normal Colon Adenocarcinoma (n = 11) (n = 11)

Advantages of Chemoprevention Strategy

Endogenous HDAC inhibitor: Synthetic HDAC inhibitors:

Probiotics & prebiotics Anticancer chemotherapeutics Minimal off-target/adverse effects Clinical trials & FDA approval

Tumor Proliferation

Apoptosis

Future: Combinatorial Human Studies

Prospective-cohort fiber study

+ GWAS or exome sequencing + Gut microbiome

Hypothesis: Responders and non-responders should have microbiota and/or genetic differences. Resolve previous conflicting results.

Precision medicine

Integrating Precision Medicine into Cancer Prevention

Population-based studies

Biomarkers: Dietary compliance or disease state

Mechanism: Etiology and disease progression

Translational potential

Genetic background/GWAS

Microbiome

Metabolome

Emerging Evidence for Precision Medicine & Interactome

Nutrition (fiber)

Immune/Inflammation (Treg cells)

Metabolism (butyrate)

Genetic background

Cancer susceptibility

Microbiome (Clostridium)

Milk Bifidobacteria SNPs

Diet Choline production/ 1C metabolism

SNPs Tumor Tumor Normal Normal

Bultman lab Dallas Donohoe Darcy Holley Sarah Renner Kaitlin Curry Alicia Greenwalt

Funding:

Acknowledgments Texas A&M David Threadgill Colorado State University Elizabeth Ryan

UNC Maureen Bower Leonard Collins Jim Swenberg Virginia Godfrey Stephanie Montgomery

Questions?

Ongoing Experiments

Control Diet High Fiber Diet

Butyrate

Hypothesis: Butyrate mediates tumor suppression, in part, through a cancer cell non-autonomous manner via diminished inflammation

Wild type RAG2-deficient IL10-deficient

Tumor burden Inflammation

ChIP seq: Colon, tumor, immune cells H3K9ac, H3K27ac, HDACs

Other dietary factors? Omega-3 fatty acids (PUFAs) Synergize with butyrate in vitro

±

Lumen Intestinal Epithelium Lamina Propria

Commensal bacteria

B cell T cell

Macrophage

Dendritic cell

Immune cells

Fiber

+ Dietary/digestive components Metabolites (oncometabolites, tumour-suppressive metabolites)

=

Polyphenols

Meat Bile acids 2° bile acids (DCA), hydrogen sulfide SCFAs (butyrate), equol, urolithins

Cancer incidence

Cancer treatment

Beyond Butyrate: Gut Microbiota & Cancer

CpG oligo, platinum chemotherapy Anti-CTLA4, anti-PD-L1 immunotherapy β-glucuronidase & chemotherapy (CPT)

Systemic effects via circulation

Irinotecan

SN-38-glucuronide

HUMAN ESTERASE

SN-38

HUMAN UGT

HUMAN TOPO I

efficacy

SN-38-glucuronide

SN-38

BACTERIAL GUS

epithelial cell death Dose-Limiting

Diarrhea

Large GI

Prodrug

Active

Active

Inactive

Inactive

Liver

Drugs Targeting Bacterial Enzymes to Prevent Adverse Effects of Chemotherapeutics in GI Tract

Matt Redinbo, UNC Aadra Bhatt, UNC

Gnotobiotic Facility

GEMMs

Metagenomic Sequencing

Next Phase of Microbiome Research Bacterial Culture

from cataloging to function

Organoid co-culture system -Higher throughput -Less $ -Not Caco-2 cells -Not exisiting miniguts

3D Crypts-on-a-Chip

-Lgr5+ stem cells & each differentiated cell type on ECM scaffold in media with Wnt gradient -Anatomically & physiologic normal crypt -Customize genetic background & disease state -Assay barrier function, transport, metabolism, etc. -Amenable to microbial co-culture -Can be interconverted to 2D culture system for high-throughput screening

Nancy Allbritton, UNC Scott Magness, UNC

Human crypts in vivo 3D crypts-on-a-chip

Intestinal Simulacra

The Warburg Effect Cancer cells shift from oxidative metabolism to aerobic glycolysis:

Glycolysis

Pyruvate

2-4 ATP TCA cycle

36 ATP

Glucose

cytosol

mitochondria

+ O2

Lactate

Otto Warburg 1931 Nobel Prize Recently “rediscovered”

OXPHOS

Cancer Cells Take Up Glucose at High Levels

Warburg effect is basis of tumor imaging in the clinic:

FDG

Warburg Effect: Conduit for Raw Materials to Double Cellular Biomass

Butyrate Accumulation & Histone Acetylation

IC50 = 300

13C-Butyrate (mM)

Reported IC50

Warburg Effect No Warburg Effect

LC-MS/MS: WBA:

H3ac

H3

Butyrate (mM)

Hypothesis: Butyrate can also increase histone acetylation by a mechanism other than HDAC inhibition.

Butyrate Regulates Histone Acetylation via ACL-Dependent Mechanism

#1

#2

Butyrate

Butyrate

Butyrate

ACL

TCA cycle

Acetyl CoA

Citrate

β ox

idat

ion

mitochondria

Citrate Acetyl CoA

nucleus

Ac

nucleus HDACs

HATs

Lipid biosynthesis

AKT

Glucose

X

C13

C13

TSA Control

- + - + siACL

Vehicle TSA

H3ac

β actin

ACL

#1

#2

Butyrate

Butyrate

Butyrate

ACL

TCA cycle

Acetyl CoA

Citrate

β ox

idat

ion

mitochondria

Citrate Acetyl CoA

nucleus

Ac

nucleus HDACs

HATs

Lipid biosynthesis

AKT

Glucose

X

ACL-Dependent Mechanism Predominates in Normal Colonocytes

#1

#2

Butyrate

Butyrate

Butyrate

ACL

TCA cycle

Acetyl CoA

Citrate

β ox

idat

ion

mitochondria

Citrate Acetyl CoA

nucleus

Ac

nucleus HDACs

HATs

Lipid biosynthesis

AKT

Glucose

X

Cancerous Colonocyte

Warburg Effect No Warburg Effect

Reg

ulat

ion

of H

isto

ne

Acet

ylat

ion

and

Tran

scrip

tion

0.5 1.0 2.0

Acetyl-CoA Donor/HAT Mechanism

HDAC Inhibitor Mechanism

Butyrate (mM)

A

Butyrate

Lys

Mitochondria

Histone

Ac

HDACs HATs

Butyrate Butyrate Glucose B C

D

Butyrate Glucose

Nucleus

Histone

Ac

HATs HDACs

Histone

Ac

HATs HDACs

CO2

Normal Colonocyte

Proliferation

5.0

Butyrate Increases Histone Acetylation by Stimulating HATs & Inhibiting HDACs

Histone Histone Lys Lys Ac

Ac Coenzyme A

HATs

Coenzyme A

Ac

HDACs

Normal cells; lower doses; proliferation genes

Cancer cells; higher doses; pro-apoptotic genes (Fas)

Biological material (buccal swabs, stool samples, tissue biopsies, etc.): -Can be obtained from disease cases and controls

Principal Component Analysis to identify differences between samples such as disease cases and controls Identification of microbes significantly enriched in different body sites or at one site in cases vs controls

Prepare genomic DNA

Computational assembly and analysis of microbial sequence reads to quantify abundance of microbiota

Next-generation DNA sequencing: -16S rRNA -whole-genome shotgun (discard human sequence reads)

Microbiome Studies

Scott Bultman, Department of Genetics, Lineberger Comprehensive Cancer Center, UNC-Chapel Hill

Bridging nutrition, metabolism, the microbiome, and cancer

Microbiome

Cancer cell metabolism (Warburg effect)

Epigenetics (HDAC inhibitor)

Butyrate