Epigenomics and Microbiome
Michael Snyder
October 19, 2015
Slides from: Joe Ecker, Anne Brunet, Joanna Wysocka, Jason Reuter Disclosures: Personalis, GenapSys, AxioMx
A Symphonic Example
DNA Phenotype
Image: mooogmonster
Epigenetic changes
It is likely that factors outside of DNA sequence may also contribute to human diseases and possibly evolution
Definition: Epigenomics The study of the complete set of epigenetic modifications on the genome that can modify DNA instructions, turning on and off genes, or control the production of proteins. These modifications occur when chemical compounds or proteins bind to DNA and “mark” the genome. They do not alter the sequence itself. Methods for marking the genome: • DNA methylation • Histone modification
The Epigenome – DNA Methylome
Often occurs at CpG >28 million CpG sites in the human genome. Human genome consists of vast oceans of DNA sequence containing few but heavily methylated CpG dinucleotides. Punctuated by short regions with unmethylated CpGs occurring at higher density, forming “CpG islands” in the genome, often at gene promoters. HydroxyC methylation (not discussed)
The Human Epigenome
Maleszewska & Kaminska, Cancers, 2013
Methylation in promoters: Gene expression turned off
Inactive A
ctive Chrom
atin
Ac: Acetylation
The Human Epigenome
Maleszewska & Kaminska, Cancers, 2013
Methylation in promoters: Gene expression turned off
Inactive A
ctive Chrom
atin Ac: Acetylation
Dnmt: DNA methyltransferase TET: “Demethylation” HMT: Histone MethylTransferase KDM: Histone Demethylase HAT: Histone AcetylTransferase HDAC: Histone DeACetylase
Modifying enzymes
DNA Methylation Associated with Many Diseases and Traits
Asthma and Allergies (Nadeau)
Aging Nutrition
Cancer Exercise
Horvath Genome Biology 2013 14:R115 doi:10.1186/gb-2013-14-10-r115
DNA Methylation Correlates With Age
Measure Methylation at 353 CpG Sites “Clock CpGs” Correlates Well with Age Across Many Tissues
Nutrition and DNA Methylation Honey Bees Extensive differential methylation workers and queen “Royal Jelly”silences Dnmt3 DNA Methyl transferase
Nutrition and DNA Methylation
Agouti Mouse • Paracrine signalling molecule
that affect coat color • Yellow, obese, diabetic
unmethylated • Brown healthy methylated • Transposon
Nutrition and DNA Methylation
BPA exposure
Agouti Mouse • Affected by Nutrition and Environment • Methyl rich diet by Mom Suppresses yellow • BPA – (bisphenol A) polycarbonate plastic precursor
Human DNA methylation at putative MEs is influenced by season of conception in the Gambia
Waterland et al., PLoS Genet 2010
Gambia – rainy season: hungry, dry season: lots of food
DNA from peripheral blood leukocytes
Exercise and DNA Methylation
Lindholm et al. Epigenetics. 2014 Dec 2;9(12):1557-69. doi: 10.4161/15592294.2014.982445.
• Many studies have linked exercise and DNA methylation differences are specific genes
Recent study Lindholm et al 23 people. Bike use one leg 4X/Week 45’ for 3 months; other leg is control Measure gene expression and DNA methylation (arrays) beginning and end of training on each leg (muscle biopsy) 4919 sites differed in trained leg (5% FDR)
Father
Mother
T C Inactivated by
mutation
Inactivated by DNA
Methylation M M M
Methylated CpGs
Few RNAs
Lots of RNA
PDE4 DIP Gene
Gene Inactivation by Mutation and Methylation: PDE4 involved in eosinophilia
electron micrograph of chromatin
nucleosome
DNA
histones K. Luger
Epigenetics and Chromatin
Chromatin undergoes reversible chemical modifications, which have regulatory properties
Chromatin modification plays a regulatory role
Covalent modifications of histone proteins
acetylation lysine methylation
arginine methylation
phosphor- ylation
Me
K Me Me K Ac
R Me Me
S/T P Ub
K Sumo
K
ubiquitin- ylation
sumo- ylation
Non-covalent mechanisms: ATP-dependent chromatin remodeling
Mind-boggling array of histone modifications
Writers, readers and erasers of histone modifications
writer eraser reader
acetylation
methylation
acetyltransferases deacetylases
methyltransferases demethylases
kinases phosphatases phosphoryl.
bromodomains
chromodomains, PHD fingers, MBT
14-3-3, BRCT
MODIFICATION:
From J. Wysoka
Identification of proteins recognizing histone lysine acetylation and methylation
SPECIFICITY DOMAIN PROTEIN YEAR NOTES REFERENCES
Acetyl lysine Bromodomain Gcn5 1999 Not lysine specific
Dhalluin et al., 1999
Methyl lysine Chromodomain HP1 2001 H3K9me specific
Bannister et al., 2001; Jacobs, Khorasanizadeh, 2002; Lachner et al., 2001
PHD finger BPTF, MLL etc etc
2006 Specific for histone, lysine & me’n state
Wysocka et al., 2006; Li et al., 2006; Pena et al., 2006;
Tudor, MBT etc 2007 Reviewed by (Taverna et al., 2007)
Multiple histone-binding modules often co-exist within the same polypeptide
Ruthenburg et al., Nat. Rev. MCB, 2007
24
Histone Marks Representing a Variety of Functional Elements
Active Genes
Repressed Genes
Chromatin variation across individuals
SAN
CEU
YRI
Asian
Enhancer related marks are more variable than RNA
Kasowski et al 2013 Science
Epigenetics and Cancer
Cancer is a genetic disease Cancer is also an epigenetic disease
The methylome and major changes that occur in cancer
The cancer methylome is characterized by widespread Hypo-methylation & focal CpG-island Hyper-methylation, often in promoters of tumor suppressor genes. Stirzaker et al., Trends in Genetics, 2014
Epigenetic modifiers are frequent targets for mutation in cancer
Recent sequencing studies show that mutations in the three classes of epigenetic modifiers are frequently observed in various types of cancers, highlighting the crosstalk between genetics and epigenetics: - DNMT3A mutations in AML
(next slide) - IDH1/2 mutations in
glioblastoma - EZH2 mutations in breast,
colon, pancreas, melanoma,…
You & Jones, Cancer Cell, 2012
DNMT3A mutations in leukemia Mutations in DNMT3A, an enzyme concerned with de novo methylation of CpG dinucleotides, are among the most common somatic mutations, occurring in 25% of acute myeloid leukemia (AML). DNMT3A mutations have also been found in myelodysplastic syndromes (MDS). Indicates that aberrant epigenetic modulation of the genome has a pathological role in leukemogenesis.
Location of mutations in DNMT3A in AML
Methylation-based classification of colorectal cancer correlates with cancer
gene mutation status
Hinoue, T et al. (2012). Genome research, 22(2), 271–82
BRCA DNA Methylation and Mutation in Ovarian Cancer
Genes can be inactivated by either Mutation or DNA Methylation
Cancer Genome-Epigenome Interaction: Gene body methylation contributes to cancer-causing mutations
• Methylation of cytosine strongly increases the rate of C→T transition mutations.
• Due to “spontaneous deamination” of cytosine, which forms thymine.
• In somatic cells, gene body methylation is a major
cause of cancer gene mutations in tumor suppressor genes, such as TP53.
The epigenome as a therapeutic target
Abnormal hyper-methylation often silences the expression of tumor suppressor genes. Drugs, such as Dacogen and Vidaza (5-azacytidine), may reverse tumor-associated gene silencing through their ability to disrupt DNA methylation.
Hamm and Costa, Pharmacology & Therapeutics, 2015
The histone deacetylase inhibitors Zolinza and Istodax may reverse tumor-associated gene silencing through their ability to disrupt aberrant posttranslational histone alterations. (Subcutanteous T-cell lymphoma; CTCL)
The reversible nature of epigenetic modifications provides an exciting opportunity for the development of therapeutics.
Methylation of certain genes can influence sensitivity to chemotherapeutic drugs
Toyota et al.,
Cancer Science, 2009
Best known example for promoter methylation influencing treatment: MGMT
• The O(6)-methylguanine-DNA methyltransferase (MGMT) encodes a DNA repair enzyme that can abrogate the effects of alkylating chemotherapy such as temozolamide.
• Alkylating chemotherapy damages DNA and kills tumor cells.
• However, if the MGMT gene is active (unmethylated promoter), the damage is rapidly repaired.
• Malignant gliomas may have the MGMT gene inactivated due to methylation of its promoter region.
• Thus, in gliomas, MGMT promoter methylation is a favorable prognostic marker in the setting of either radiation or chemotherapy.
DNA methylation of specific marker genes can be used as a biomarker for cancer detection and diagnosis
Summers et al., Journal of Cancer, 2013 http://www.epiprocolon.com
History of cancer epigenetics
Hattori et al., Biochem. Biophys. Res. Commun., 2014
Conclusions
1. Epigenomic regulation is widespread 2. Occurs by multiple mechanisms 3. Is involved in many important biological processes Cancer, Aging, Nutrition
Microbiome
What is the microbiome? The collection of microbiomes that exist in a given area, tissue or organism. It refers to all microbes including bacteria, fungi and viruses.
The Human Microbiome The human microbiome is the collection of microbes that live on and in us
30 trillion human cells
300 trillion microbial cells
>10 million genes (compared to 20K for humans)
Enormous biochemical diversity What do the microbes do??
The Role of the Microbiome It breaks down our food into constituents we absorb.
It synthesizes essential vitamins. E.g. Vitamin B12, Biotin, Folic Acid
Body ecology Different Kingdoms Bacteria Eukaryota (e.g fungi) Archaea Viruses Different Phyla/Classes Typically 5-10 major types 100s – 1000s of species Macro-ecology: Plants, Animals, Fungi, Protists
Different niches Aerobic, anaerobic Dry, moist Oily, nutrient-rich/poor Exposed, protected from environment
Where did our microbiome come from?
Microbes have been on earth for 3 billion years We evolved in the midst of microbes We evolved organs to house our microbes
Each body site has its own microbial community Adaptive immunity allowed richer microbiome?
Image: Darryl Leja, NHGRI
When does a person’s microbiome form?
At birth a baby’s gut microbiome is similar to the mother’s vaginal microbiome It forms at lot during the first year as baby ingest food Microbiome from breast fed microbiome has similarities to its Mother’s breast/milk
Non breast fed babies have a different microbiome.
Dominguez-Bello et al. (2010) PNAS 107:119
Sampling of the microbiota 5-60 minutes after birth
• Acinotobactor " Baclllales • Conobactotineae • Corynebacterineae • Haomopnnus • Lactobacd us
• Mlc:roooconeae • NGI$$8t14
Pasteur..aceae • Prevotella • Propion1bacteMeae Sneathia
Staphylococcus Slroptococcus Other
Dominguez-Belloeta!. (2010) PNAS 107:11971x
Vaginal (Baby l ike Vaginal) Caesarian (baby looks l ike skin)
Approaches to metagenomic study of the microbiome
Microbial community is very diverse Bacteria, viruses, eukaryotes 100s-1000s of taxa A few are >80% of total MANY minor taxa 99% of microbes cannot be cultured! Two approaches Targeted sequencing Shotgun sequencing
Targeted sequencing
16S rRNA sequences/sample Average 5000 sequences
Must have >1000 sequences Sample 100s of taxa
Describe communities and genus abundance in many samples –
“Average” community and variations
Cost < $50/sample 16S rRNA = bar code identifier of species
Liu et al. Nucleic Acid. Res. 2007 35. e120
Primers to variable parts of 16S rRNA
Shotgun sequencing Sample every gene in the community by high throughput DNA sequencing
Analyze for: Bacterial community structure, Viruses, Fungi Metabolic pathways, other genes of interest (antibiotic resistance; virulence)
Slide courtesy of Ami Bhatt
Studying communities – shotgun sequencing
Assemble
Illumina shotgun of sample
community DNA
10 Gb/sample 100 million reads 100s of samples
Align
Catalog of 2000 reference genomes
Blast vs DBs (GenBank eg)
Analyze each genome for breadth and depth of read coverage,
SNVs
Gene predictions
Growth of Archae/Bacteria genome sequences
Nikos Kyrpides, Genomes Online Database
1 genome/day added to GenBank
Bacterial sequencing dominates
Major subgroups of the stool biome type
women men
St. Louis Houston
BMI>=30
BMI<25
NA
not hispanic/latino/spanish Hispanic/latino/spanish
25 <= BMI <30
Lots of diversity of individual level: People Have Their Own “Personal Microbiome”
Each row a different sample
Histograms of genera in each sample
Bacteroides
Prevotella
Ruminococcus
Yanjiao Zhou & George Weinstock
Diversity at the organismal level but wide array of shared microbial genes comprising a “core microbiome” at the gene level
Turnbaugh PJ, et al. Nature. 2009
The microbiome differs with different organ sites
Dirk Gevers (Broad Institute) + HMP Consortium
Other cool facts
Niels O. Verhulst PLoS One 2011
HA = Highly Attractive PA = Poorly attractive
Does microbiome affect attraction between humans?
The skin microbiome affects our attractiveness to mosquitos!
The Role of the Microbiome in Health and Disease
Ecological diseases from our microbiome
Not infectious diseases: causative agents are part of the normal flora - e.g. Candida albicans - Disease results from disturbance to the environment, altering microbial balance Antibiotic treatment - Clostridum difficile no longer held in check by excess beneficial organisms - Reduced protection against deleterious organisms Dental caries - Streptococcus mutans overgrowth due to high carbohydrate utilization - Increased amounts of deleterious substances, eg lactic acid Toxic Shock Syndrome - Staphylococcus aureus emergence due to abnormally dry environment - Change in the environment Many more - to
be discovered
Disease metagenomics projects Illness of unknown origin
Necrotizing enterocolitis
IBD/Crohn’s disease
Urethritis
Nasal microbiome after vaccination and flu infection
Acne
Antibiotic effects on physiology
Cystic fibrosis
Periodontitis
Lung microbiome and HIV
Kawasaki disease
Bacterial vaginitis
Obesity
Diabetes
Sepsis
Trachoma and sequelae
Autism
Myocardialinfarction (heart attack)
http://guardianlv.com/2013/10/flu-vaccine-may-cut-heart-attack-risk/
Carnitine from meat is converted to TMAO using gut microbiome - Antibiotics suppress TMAO levels in humans - Germ free mice, no TMAO CUT genes responsible Possible solutions: 1) Change diet to reduce relevant bacteria 2) Produce inhibitors of the CUT gene product. 3) Genetic engineering
High levels of TMAO (trimethylamine-N-oxide) are associated with MI
Some examples: Obesity and the microbiome
Our food affects our microbiome High fat vs low fat diet.
Typically American diet Soy
Vervet 2009
Less gut bacterial diversity in obese vs lean people
Turnbaugh PJ, et al. Nature. 2009
Obesity and the microbiome Differences between obese mouse and lean mouse
Ley et al PNAS 2005
Leptin mutant (ob/ob)
Normal mouse (+/+)
TLR5 (toll like-receptor deficient) mice fight bacteria differently from normal mice.
Vijay-Kumar et al Science 2010
Their transmissible microbiome makes them eat more
Vijay-Kumar et al Science 2010
What about humans?
Ley et al Host Cell & Microbe 2006
Pilot study in humanized mice
Smith et al Science 2013
Gut microbes studies in discordant twins Implant microbiome into Mice
Kawshiokar = malnurished
Diabetes and the microbiome
Diabetics have a different microbiome than nondiabetics in humans and mouse models. Inject bacteria from healthy mouse into mouse prior to inducing diabetes. Reduces diabetes formation.
Wang et al Cell 2012
Inflammatory bowel disease (IBD)
Different microbiome in Chrohn’s disease and IBD In some cases fecal transplants seemed to help (In case you were wondering – Administered through Ingestion or the rectum)
Infections: Organ transplantation Microbial infections often appear after organ transplantation Example: Hematopoetic cell transplantation with umbilical cord cells at Harvard 11 people infected after tranplant e.g. 30 yr old woman with ALL. Pursued cord tranplant. Developed coiltis
WBC 135K/uL, 53% blasts
Acute Leukemia (ALL)
Bhatt et al, NEJM 2012
Organ transplantation
Extract DNA from fixed tissue Shotgun sequencing
Many reads did not map to known bacterial genomes!
Lots!
Bhatt et al, NEJM 2012
Organ transplantation Assemble unmapped reads from two patients into new sequence: Bradyrhizobium enterica
Slide courtesy of Ami Bhatt
Classification of the CCS-associated bacterium
Phylogenetic analysis using the draft genome to classify the organism
Slide courtesy of Ami Bhatt
Febrile children study
Plasma samples positive for HHV6 or Enterovirus 6 8 1
PCR Seq
Greg Storch
Conclusions 1) The microbiome is vast and essential 2) It is associated with many, and possibly all, human diseases. 3) It is not clear what is cause and what is effect. 4) It may provide novel avenues for treating human disease.