Sequence variants in SLITRK1 are associated with Tourette's syndrome

1

Sequence Variants in SLITRK1Are Associated with Tourette’sSyndrome

Jesse F. Abelson, Kenneth Y. Kwan, J. O’Roak, Danielle Y. Baek, Althea A. Stillman, ThomasM. Morgan, Carol A. Mathews, David L. Pauls, Mladen-Roko Rašin, Murat Gunel, Nicole R.Davis, A. Gulhan Ercan-Sencicek, Danielle H. Guez, John A. Spertus, James F. Leckman, LeonS. Dure IV, Roger Kurlan, Harvey S. Singer, Donald L. Gilbert, Anita Farhi, Angeliki Louvi,Richard P. Lifton, Nenad Šestan, Matthew W. State.

Science 310, 317-20. (2005)

Presented by Joanne Brathwaite

Tourette’s Syndrome Define Tourette’s syndrome Early genetic research and TS Current Study

Chromosome 13 inversion Frameshift mutation Noncoding sequence variant (var321)

Expression patterns in mouse and human brain Summary Conclusions Questions

Tourette’s Syndrome (TS) or (GTS)

Definition:A genetically influenced developmentalneuropsychiatric disorder, characterized bychronic motor and vocal tics.

Tourette’s Syndrome Association, Inc.; Diagnostic and statistical manual ofmental disorders : DSM-IV-TR., 2000

Early genetic research Segregation Analysis

Suggested that TS is familial Suggested disorder inheritance is rare, autosomal

dominant Suggested poly- or oligogenic inheritance

Genome-wide Linkage Analysis Disorder implicated on chromosomes 4,5,8,11 and 17

But no disease related mutations identified!Why?

Problems identifying TS genetics

Phenotype decreases in severity with age High population prevalence of transient tics Symptom overlap with common disorders (OCD,

ADHD) Marked locus heterogeneity Gene x Environment interactions Assortative mating

Pauls, 2003; Merette et al., 2000; The Tourette Syndrome AssociationInternational Consortium for Genetics, 1999; Singer, 2005; Hanna et al.,1999

Current Study – Family 1

Recruited a rare subset of TS patients with chromosomalanomalies

Identified a patient with TS and ADHD

Family 1 History negative for:

TS OCD TTM ADHD

2

Chromosome 13 Inversion De novo inv(13)(q31.1:q33.1) Breakpoints spanned by 2 BACs

RP11-375K12 (spans q31.1) andRP11-255P5 (spans q33.1)

FISH used to fine map the chromosomerearrangement

Chromosome 13 Inversion 3 genes mapped within 500kb of both breakpoints ERCC5

Xeroderma pigmentosum group G

SLC10A2 Primary bile acid malabsorption

SLITRK1 Integral membrane protein expressed in neural tissues

Candidate Gene SLITRK1

Slit and TRK-like family member 1 (SLITRK1) Encodes a single-pass transmembrane protein with

2 leucine-rich repeat (LRR) motifs in extracellulardomain

High relative expression in brain regions previouslyimplicated in TS

A suggested role in neurite outgrowth

Aruga J et al., 2003

Possible Role of SLITRK1 in TS

“...if altered SLITRK1 function contributed to the risk forTS in the patient carrying the inversion, we would expecta subset of TS patients to have mutations in this gene.”

Screened SLITRK1 in 174 affected individuals

Family 2 Identified a proband with TS and ADHD Identified a single-base deletion in coding region leading

to a frameshift Mutation present in patient and patient’s mother

SLITRK1 truncated peptide•Lacking LRRs, LRR C-terminal domain and transmembrane domain (TM)

3

SLITRK1 var321 Identified in 2 unrelated TS

patients with obsessivecompulsive (OC) symptoms;absent in 4296 controlchromosomes. Significant association with TS

(P = 0.000056) Both families had a history of

chronic tics and OC symptoms

var321: Non-coding sequence variant (G/A) single base change in 3’ UTR of transcript

Corresponds to a highly conserved nucleotide withinthe predicted binding site for the human miRNA hsa-miR-189

G:U wobble replaced with A:U Watson-Crick basepair

Testing var321 effects Luciferase - pRL-SLITRK1-3’UTR construct was

transfected into Neuro2a (N2a) cells

var321 construct shows dose-dependent further repression ofluciferase expression versuswildtype

Slitrk1 mRNA and miRNA-189 expression

Postnatal day 14 mouse brain Fetal human brain 20 weeks gestation

Dendrite growth•SLITRK1 has high cortical expression levels

•Cortical pyramidal neurons exposed to wild-type SLITRK1produced dendrites significantly longer than those exposed tocontrol and mutant SLITRK1

SLITRK1 and Tourette’s syndromeSummary

SLITRK1 identified as a candidate gene in TS var321 (G/A) in 3’UTR of SLITRK1 mRNA

Altered interaction between SLITRK1 and miR-189

Frameshift mutation of SLITRK1 Overlapping expression patterns for SLITRK1 mRNA and

miR-189 SLITRK1 gene product promotes dendritic growth, and

mutation results in a loss of function

4

SLITRK1 and Tourette’s syndrome

Conclusions Discovery of rare mutations helps to understand disease

pathogenesis Further study of SLITRK1 will serve a role in

understanding TS at a molecular and cellular level

References Tourette Syndrome Association, Inc. http://www.tsa-usa.org/ Diagnostic and statistical manual of mental disorders : DSM-IV-TR.,

2000 D. L. Pauls, J. Psychosom. Res. 55, 7 (2003). C. Merette et al., Am. J. Hum. Genet. 67, 1008 (2000). The Tourette Syndrome Association International Consortium for

Genetics, Am. J. Hum. Genet. 65, 1428 (1999). H. S. Singer, Lancet Neurol. 4, 149 (2005). P. A. Hanna, F. N. Janjua, C. F. Contant, J. Jankovic, Neurology 53, 813 (1999). Materials and methods are available as supporting material on

Science Online. J. Aruga, N. Yokota, K. Mikoshiba, Gene 315, 87 (2003).

Tourette’s Syndrome

Questions?

1

Global Position and Recruitmentof HATs and HDACs in the Yeast

Genome

Presented by Lu Bilyk

Outline

Histone structure Histone modifications HATs HDACs

Core Histone Octamer

Richard Wheeler (Zephyris) 2005, http://en.wikipedia.org/wiki/Histone http://fr.wikipedia.org/wiki/Histone

Chromatin RemodellingThere are 3 general ways in which chromatin

structure can be altered:

1. Remodelling by specific complexesex: SWI/SNF, ISW1

2. Histone variants may replace core histonesex: H2AZ, H2AX, H3.3

3. Histone modificationex: Acetylation, Methylation, Phosphorylation, Ubiquitination

Chromatin RemodellingThere are 3 general ways in which chromatin

structure can be altered:

1. Remodelling by specific complexesex: SWI/SNF, ISW1

2. Histone variants may replace core histonesex: H2AZ, H2AX, H3.3

3. Histone modificationex: Acetylation, Methylation, Phosphorylation, Ubiquitination

2

Histone Modification

•Acetylation

•Methylation

•Phosphorylation

•Ubiquitination

www.upstate.com and also www.histone.com

Histone Acetylation

Acetylated → ActiveDeacetylated → Repressed

HATs, HDACs, and all that jazz..

HATs: Histone Acetyltransferases Gcn5 Esa1

HDACs: Histone Deacetylases Hst1 Rpd3

Global Position and Recruitment

Knowledge of the genome-wide location of a chromatinregulator has the potential to:

1. Determine whether the regulator is associated withall or a subset of genes transcribed by a specificRNA polymerase

2. Reveal whether a regulator is associated with thepromoter or transcribed region of genes

3. Lay the foundation for studies that reveal thefactors responsible for recruiting the regulator tospecific regions of the genome

4. Extend to many genes a model for regulatorfunction based on previous studies of one or a fewgenes

Protocol Overview

HATs and HDACs were epitope taggedat their chromosomal locus

Genome-wide location scans wereperformed in triplicate using DNA arraysrepresenting the yeast genome

Additional information from ChIPexperiments

ChIP

http://www.komabiotech.co.kr/product/immunology/use/chip.htm

1. Cross-link proteins toDNA

2. ImmunoprecipitateDNA-protein complex

3. Identify DNA by PCR

3

Gcn5 and Esa1 are Recruited toActive Protein-Coding GenesGcn5 → catalytic subunit of SAGAEsa1 → catalytic subunit of NuA4

Genome-wide expression studies show that only asubset of protein-coding genes depend on function ofSAGA and NuA4

Is it possible that they are recruited to and function atall active protein-coding genes, but loss-of-functionexperiments do not fully reveal this due to the abilityof other chromatin regulators to compensate?

Gcn5 and Esa1 are Recruited toActive Protein-Coding Genes

Genome-wide occupancy of protein coding genes byboth Gcn5 and Esa1 correlates with transcription rate

Gcn5 and Esa1 are locatedPredominately at UAS When hybridized on ORF arrays,

smaller enrichment was observed

Gcn5 and Esa1 are Recruited toInactive Genes Upon Activation

Uninduced - black

Induced - Gray

Hst1 Regulates Midsporulationand Kynureine Pathway Genes Unline HATs, HDACs are not recruited to active

genes Chip confirmed that Hst1 occupies the promoters of

all midsporulation genes tested and BNA1 and BNA5from kynureine pathway

Hst1 is Recruited by a SingleTranscription Factor Sum1

The set of genes occupied by Sum1and Hst1 are nearly identical

4

Hst1 has HDAC Activity in Vivo

ChIP using antibodies against acetylatedH3/H4

Similar results with sum1Δ strain

Rpd3 is Associated With CellCycle Genes Rpd3 is part of a large protein complex

composed of many proteins, includingSin3, Sap30, and Sds3

Rpd3 complex negatively regulatesearly meiosis genes during vegetativegrowth

Rpd3 and Sin3 Associate withEssentially the Same Genes

Consistent with the data that Rpd3 and Sin3can be purified as a complex

Genome-wide occupancy of Rpd3 and Sin3do not correlate with transcription rate

Target genes associated with cell cycleregulation

Rpd3 Occupies Promoters of CellCycle Regulators

Consistent with previous reports that Rpd3might play a role in cell cycle regulation

INO1 → myo-inositol-1-phosphate synthasePCL1, CDC20, CLB6 → cell cycle regulators

Association of Rpd3 with CycleRegulators Requires Swi4/Swi6

Association of Rpd3 with PCL1, CDC20, andCLB6 requires Swi4/Swi6

Rpd3 continues to occupy genes notregulated by Swi4; Rpd3 may be recruited bymultiple transcription factors

Study by Kurdistani et Al. A similar study by Kurdistani et al. found that

Rpd3 is preferentially associated withpromoters that direct high transcriptionalactivity

5

Rpd3 is not Generally Associatedwith Highly Transcribed ProteinGenes in Rich Media

What is the cause of the discrepancy?

Different Protocols GiveDifferent Enrichment

Significant difference: wash step in ice-cold buffer

Standard protocol → blackKurdistani protocol → gray

Cold Shock Triggers Bindingwith Ribosomal Protein Genes

Summary

Histone acetyltransferases Gcn5 and Esa1are both generally recruited to promoters ofactive protein-coding genes

Histone deacetylases Hst1 and Rpd3 arerecruited to specific sets of genesHst1 → recruited to sporulation and kynureine pathways genesby Sum1Rpd3 → recruited to cell cycle regulators and to ribosomalprotein genes under stress

References

ArticlesFelsenfield, G. and Groudine, M., 2003. Controlling the double helix. Nature 421, pp. 448–453Robert, F., Pokholok D.K., Hannett N.M., Rinaldi N.J., Chandy M., Rolfe A., Workman J.L.,

Gifford D.K., and. Young R.A., 2004. Global Position and Recruitment of HATs and HDACsin the Yeast Genome. Mol. Cell 16, pp.199-209

Kurdistani, S.K., Robyr, D., Tavazoie, S. and Grunstein, M., 2002. Genome-wide binding map ofthe histone deacetylase Rpd3 in yeast. Nat. Genet. 31, pp. 248–254.

BooksWeaver, R.F. 2002, Molecular Biology 3rd ed., McGraw-Hill, NY USA pp. 405-419

Websiteswww.histone.comwww.wikipedia.com

1

Parallel Patterns of Evolution in theGenomes and Transcriptomes ofHumans and Chimpanzees

Philipp Khaitovich et al.

BIO478Presentation by: Jerry ChaoMarch 17, 2006

Outline

Evolution Theories and Terms The Experiments

Defining sample & gene profile Parallel pattern between expression & sequence

evolution Evolution of core promoter sequence vs.

expression Incidence of positive selection Expression & sequence diversity between species

Summary…final thoughts

Evolution

Evolution: Changes in allele frequencies over time

Darwinian Natural Selection Difference in fitness between individual with one phenotype

and individual with another phenotype Negative selection: purifying selection, elimination

because not favorable/deleterious Positive selection: trait is favorable, frequency gradually

increases and may become fixed

Neutral Theory of Molecular Evolution

“Vast majority of base substitutions are neutral withrespect to fitness and genetic drift dominatesevolution at the level of DNA sequence”

Formally published by Motoo Kimura in 1983 Based on observation of redundant genetic code for

amino acids Early 2000s, neutral theory became widely used as

“null model”

Functional Constraints Rates of molecular evolution vary widely among loci Genes that are responsible for most vital functions

have lowest rate of replacement substitution Histone protein have high constraints

Less vital genes have less stringent constraints andlarge proportion follows neutral theory

From Chimpanzee to Human

Human have changed drastically since evolutionarydivergence from common ancestor

Source of evolutionary differences: Structural/Trait (Protein) evolution Regulatory (Gene expression) evolution

Do Protein evolution and Expression evolutionprogress independently?

What drives expression evolution?

2

Source of Evolutionary Differences

Evolutionary variation at the gene sequence levelusually NOT proportional to phenotypic variation Human & chimpanzee have much greater phenotypic

variation than Mus musculus & Mus spretus, but sequencevariations are relatively equal

Dogs have large phenotypic variation while having littleoverall sequence difference

Just for Clarification…

Divergence: Comparison between species;i.e. Humans to Chimpanzees

Diversity: Comparison the lineages withinspecies; i.e. individual humans

Sample Source

6 humans & 5 chimpanzees RNA isolates from various tissues

Heart Liver Kidney Testis Brain (prefrontal cortex)

Choice of Genes to Study

Affymetrix Human Genome U113Plus2 array probedwith RNA isolates

http://www.affymetrix.com/products/arrays/specific/hgu133plus.affx

Search for sequence identical/near identicalbetween human and chimpanzee

51,460 probe sets (~21,000 transcripts) from alltissues were used

Transcriptome Expression Diversity &Divergence

Sum of square differences between expression intensity Brain show lowest divergence Greatest divergence to diversity ratio in testis

Schematic illustration ofexpression intensity

Categorized into tissue-specific and ubiquitouslyexpressed genes

Length proportional todifference in expressionlevel

Observations Similar expression trend Ubiquitous expressed

genes have lowerdiversity and divergence

3

Protein Coding Sequence Divergence

= amount of non-synonymous substitution = amount of synonymous substitution

Estimate of protein divergence =

Low – change in sequence has no obvious effect, possibly due tostrong negative selection that eliminated non-synonymoussubstitution, ultimately confer to neutral selection

High – sequence change with phenotypic effects, amino acidsubstitution enriched by positive selection

Protein Coding Sequence Divergence

Brain had lower protein divergence Ubiquitously expressed proteins had lower protein

divergence as well

Parallel Pattern

Divergence trend similar between expression and genesequence

Perhaps expression evolution follow neutral theory, justas evolution of gene sequence

However, other important factors were not considered“…besides evolutionary mechanism, protein and gene

expression evolution are associated with mRNA abundance,protein length and protein-protein interactions.” (Lemos etal. U of Massachusetts

Evolution of Promoter Sequence vs.ExpressionTheoretically: Changes in promoter sequence have stronger effect on

expression level

The Test: Hypothetical core promoter regions

-1500bp to +500bp of transcriptional start Rate of non-synonymous substitution in hypothetical

promoter region (Kp)

Evolution of Promoter Sequence vs.ExpressionResults: Kp/Ki is weakly correlated with expression divergence

Possible Explanation: ‘Hypothetical’ promoter sequence

Most of the sequence not relevant to transcriptional activityat all

Actual promoter sequence partially included or completelymissed

More Than Just Neutral Evolution?

Consider: Under neutral theory, extent of expression divergence

between species will be determined by Time past since divergence from common ancestor Selection pressure imposed on the tissue

IF: Time is the sole influence

Ratio of expression divergence/diversity in human shouldbe low consistently across all tissues, partly due to therelatively short human evolution period

4

More Than Just Neutral Evolution?

But recall:

More Than Just Neutral Evolution?

The Test: Using ubiquitously expressed gene groups, compare influence of

expression in one tissue on the diversity of expression in another Compare diversity between gene groups that were expressed in

tissue A but not in tissue B, while other three tissues have nearidentical expression profile

12 comparisons for each pair of tissues Reduction in expression profile should indicate tissue specific

constraint on expression

Tissue Selective Constraint

Results: Strong selective constraint

on genes expressed intestis

Implications: Account for low expression

diversity in human Extreme value in

divergence/diversity doesindicate positive selection

Evidence for Positive Selection

Theoretically: Gene expressed in testis will be sex related Positive selected recessive variants accumulate

on X-chromosome will be exposed in male

The Test: Compare…

A) expression divergenceB) sequence divergence (Ka/Ki)

…for genes expressed on X-chromosometo autosomes

Evidence of Positive Selection

Results More expression divergence between human and

chimpanzee for genes on the X-chromosome

Evidence of Positive Selection

Higher sequence divergence for X-chromosome as well

5

Evidence of Positive Selection

Consistent with previous observations that genesinvolved in reproduction tend to evolve underpositive selection

Most rapidly evolving gene families are thoseinvolved in reproduction & host defense

Expression Diversity between Humans &Chimpanzees Test whether expression evolution proceeded at the

same rate between humans and chimpanzees Although less frequent to actually have evolutionary

change taking place, evolution for up-regulated geneexpression have higher amplitude if it is to occur

Unequal rate of evolution between two species can beobserved as a skewed distribution

Expression Diversity between Humans &Chimpanzees Majority of tissues have positively skewed distribution

Humans had more changes in gene expression sincedeparture from common ancestor with chimpanzees,especially the brain

Sequence Diversity between Humans &Chimpanzees Alignment of genes orthologous to human, chimpanzee,

mouse and rats Compare amount of sequence alteration that cause amino

acid changes in humans to chimpanzees Human lineage show higher rate of evolution for genes

involved in brain function and development

In Summary…

Expression evolution show parallel pattern withsequence evolution, suggesting similar evolutionarymechanism that confers to neutral theory

Parallel pattern further indicate the effect of geneexpression in evolution

Two examples did not follow neutral theory Expression evolution in human testis Expression in human brain lineages

Further Implications

Evolutionary study between humans andchimpanzees allow unique insight into humanbiology

Chimpanzee genetic sequence still imperfect andincomplete, quality sequence will be necessary toprovide conclusive studies of genomic evolution

1

A genome-wide comparison of recentchimpanzee and human segmental

duplicationsPresented by: Adrienne

Cheng, Z., Ventura, M., She, X., Khaitovich, P., Graves, T.,Osoegawa, K., Church, D., DeJong, P., Wilson, R.K., Paabo, S.,Rocchi, M. and Eichler, E.E. (2005) A genome-wide comparison ofrecent chimpanzee and human segmental duplications. Nature437(1):88-93.

Outline

Purpose Segmental duplications Methods to study segmental duplications Mechanism of segmental duplications

maintenance Hyperextension Summary References

Purpose

To understand the origin and impact ofsegmental duplications. By comparing thehuman and chimpanzee (Pan troglogytes)genomes that show evidence of shared andlineage-specific duplications

Segmental duplications

Are duplicated blocks of genomic DNAranges in size from 1–200 kb.

Often contain sequence features such as– high-copy repeats– gene sequences with intron–exon structure. Thus,

being composed of apparently normal genomicDNA

(Bailey et al. 2001)

Segmental duplications

Recent segmental duplication influenced theevolution of:

– The architecture of the human genome– The emergence of new genes– The adaptation of humans to the environment– And a host of genetic diseases

Evolutionary maintenance of duplicates could be dueto

– Slow rates of deletion,– High rates of duplications– Gene conversion

Methods to study segmentalduplications

Whole genome assembly comparison method (WGAC)– reduction of more divergent chimpanzee interchromosomal

pair wise alignments compared to humans– Recent duplications were more likely to misassemble or be

fragmented compared to the unique chimpanzee sequence

The depth of coverage of chimpanzee Whole genomeshotgun sequence detection (WSSD)

– This method will only work when the WGS is randomlydistributed and as large as possible

2

Verification of WSSD method

Chimpanzee and Human onlyduplications

Chimpanzee only and human only duplication were 10.3 times more likelyto be located in close proximity to a shared duplication

– This indicates that lineage specific deletion or duplication is occurring inproximity to regions of shared duplications.

– This is termed “duplication shadowing”

U133 Affymetrix gene expression comparison 177 complete and partial genes show evidence of duplication in

humans not in chimps– 56% of human only duplications showed significant differences in gene

expression (83% of these were upregulated) 94 genes were duplicated in chimpanzee not in humans

– 49% of chimpanzee only duplications showed 57% upregulation. This indicates that a significant proportion of the lineage specific

duplications resulted in gene expression differences

Mechanism of segmental duplicationsmaintenance

Comparing chimpanzee only duplicates withother great apes

11 out of 17 are restricted to chimpanzees– Emerged after human-chimpanzee speciation

6/17 are duplicated in gorillas 1/17 in orangutan

– These arose before divergence– And was deleted in human lineage

94-98.2%- pre-speciation, 98.2-99.2% -speciation, 99.2%-post-speciation

Chimpanzee specific hyperexpansion

Hyperexpansion- greater than 100 copies of segmental duplications

Localizations within humans: 296 regions had significant increase in copy number compared to chimps 33% of the human increase mapped within 5Mb of the centromere 21 out of 29 pericentromeric duplications in human genome

Localizations within Chimpanzees: Chimpanzees showed little increase in pericentromeric duplications (13/92)

Array comparative genomic hybridization (CGH) between humans and chimpconfirms results and suggests:

– A genome wide global expansion of pericentromeric duplications in human lineage– Or deletion of such duplications in chimp lineage

3

Chimpanzee specific hyperexpansionHomo sapiens

Old World monkey

Pan troglodytes

Pan paniscus

Gorilla gorilla

Pongo pygmaeus

Macaca fuscata

Papio anubis

Conclusions

1.5% (46Mb) is duplicated in one lineage and not in other Lineage specific differences are due to de novo duplications

(remainder is due to deletions) 26Mb Net gain of segmental duplications within chimpanzee

genome Genomic duplication rate is 4-5Mb per million years since

divergence (assuming divergence 6 million years ago) Single base pair differences – 1.2% Large segmental duplication events have a 2.7% impact

References

Bailey, J.A., Church, D.M., Ventura, M., Rocchi, M. and Eichler, E.E. (2004) Analysis of segmentalduplications and genome assembly in the mouse. Genome Research 14(5):789-801.

Cheng, Z., Ventura, M., She, X., Khaitovich, P., Graves, T., Osoegawa, K., Church, D., DeJong, P.,Wilson, R.K., Paabo, S., Rocchi, M. and Eichler, E.E. (2005) A genome-wide comparison of recentchimpanzee and human segmental duplications. Nature 437(1):88-93.

Eichler, E.E. (2001) Segmental duplications: What’s missing, misassigned, and misassembled – andshould we care? Genomic Research 11(5):653-656.

Mikkelsen TS, Hillier LW, Eichler EE, Zody MC, Jaffe DB, Yang SP, Enard W, Hellmann I,Lindblad-Toh K, Altheide TK, Archidiacono N, Bork P, Butler J, Chang JL, Cheng Z, ChinwallaAT, deJong P, Delehaunty KD, Fronick CC, Fulton LL, Gilad Y, Glusman G, Gnerre S, Graves TA,Hayakawa T, Hayden KE, Huang XQ, Ji HK, Kent WJ, King MC, Kulbokas EJ, Lee MK, Liu G,Lopez-Otin C, Makova KD, Man O, Mardis ER, Mauceli E, Miner TL, Nash WE, Nelson JO, PaaboS, Patterson NJ, Pohl CS, Pollard KS, Prufer K, Puente XS, Reich D, Rocchi M, Rosenbloom K,Ruvolo M, Richter DJ, Schaffner SF, Smit AFA, Smith SM, Suyama M, Taylor J, Torrents D,Tuzun E, Varki A, Velasco G, Ventura M, Wallis JW, Wendl MC, Wilson RK, Lander ES,Waterston RH (2005) Initial sequence of the chimpanzee genome and comparison with the humangenome. Nature 437(7055):69-87.

Yohn, C.T., Jiang, Z.S., McGrath, S.D., Hayden, K.E., Khaitovich, P., Johnson, M.E., Eichler,M.Y., McPherson, J.D., Zhao, S.Y., Paabo, S. and Eichler, E.E. (2005) Lineage-specific expansionsof retrovial insertions within the genomes of African great apes but not humans and orangutans.PLOS Biology 3(4): 577-587.

1

A Haplotype Map of theHuman Genome

International HapMap Consortium, The. (2005) A Haplotype Map of the Human Genome. Nature. 437: 1299-1320

Haplotype

• Genotype: complete

• Haplotype: incomplete; results at specificloci

Note: most organisms have more than four genes

Genes and Disease

• Two Methods currently in use:– Linkage Analysis

• Family based – link genes with phenotype• Low power if >1 gene affects phenotype

– Candidate Gene Analysis• Find a gene and then do association studies• Lower power because it ignores “universe”

– Ideally we’d like to resequence wholegenomes – Not feasible

Common Genetic Variants

• More practical to search for common variants(SNP’s)

• Individuals who carry a particular allele at onesite often carry another at adjacent sites:Linkage Disequilibrium– Higher than expected probability of finding two alleles

together considering recombination and mutation• Allele of interest came from one individual and

mutation / recombination slowly erode thatassociation

Linkage Disequilibrium

LD provides SNP’sthrough whichevolution can betraced reasonablyaccurately

International HapMap ConsortiumLaunched October 2002

1. Availability of whole human genome2. Databases of Common SNP’s3. Insights into human LD4. Accurate technology for HT screening5. Ease of Data Sharing (Internet)6. Ethical and Cultural issues addressed

2

Phase I

• Goal: genotype one common (> 5% MAF)SNP every 5kb using 269 DNA samples

• Compared with ENCODE (ENCyclopediaOf DNA Elements) using 10 representative500kb regions – SNP’s discovered or indbSNP were genotyped in 269 phase Iindividuals

• Phase II to genotype additional 4.6 millionSNP’s

dbSNP – Public SNP Database

Cumulative non-redundant

Validated by genotyping

Double hits

Samples

• Individuals from– YRI (Yoruba in Ibadan, Nigeria) 90– CEU (Utah, USA) 90– CHB (Han Chinese, Beijing) 45– JPT (Japanese, Tokyo, Japan) 44

• Three “Panels”

Results

• SNP density across ENCODE region washigher (10X) than whole genome

• 1,007,329 SNP’s discovered that arepolymorphic across all three panels (1 per279 bp)

• Relatively few of the possible haplotypesare actually found

Figure 2Greater proportion of SNP’s occur at smallerinter-SNP distances – Recombination ratebetween two loci varies inversely with distance

3

Figure 5 – inherent bias to previously characterized and common alleles“levelled” out the graphs

Expected

ENCODE

Phase I

Results• Given 1,000,000 SNP’s, between 75%

and 85% were polymorphic within panels• Also, there are few fixed differences in

which the alternate alleles are only seen inother panels

• Usually the closest relative haplotype isfrom the same panel but 10% of the time itis in a different panel – common and rarehaplotypes are shared acrosspopulations

Minor Allele frequencies(MAF’s) are usually commonacross panels; diagonalpattern indicates similarityacross panels

There are no red areas orpurple areas in strangeplaces

Linkage Disequilibrium• ‘Hotspots’ and ‘coldspots’

– 10-fold variation in recombination frequenciesacross ENCODE region

– Usually a haplotype will “break” at arecombination site• Between hotspots knowing one SNP may allow

you to infer the others• Genome is actually inherited in blocks

hotspot coldspot

DNA

Linkage Disequilibrium

• Centromeres usually have extendedhaplotypes due to lack of recombination

• This study reveals that recombinationrates are very local

Skewness of the lines indicates that recombination happens locally – youonly need half the sequence to see all the recombination

Theoretical co

nstant re

combination rate

4

Association Studies

• Find SNP (allele) that is associated withdisease– Sometimes a causal SNP can be inferred by

the presence of other SNP’s (proxies)– R2 of 1 would indicate perfect correlation– 80% of SNP’s tested had ≥1 proxy– Relaxing correlation coefficient increases

proxy number

Tags• Set of SNP’s used in a given study

– Careful selection can reduce genotypingburden without loss of resolution

• Law of diminishing returns applies whenadding more SNP’s to a tag

• Can show where tumor supressor genesare:– Long runs of homozygous SNP haplotypes can

indicate loss of heterozygosity

DNA Structure

• Deletions show detectible SNP patterns– Strong LD with neighboring regions still exists

allowing LD-based approaches to be usefuleven for discerning DNA structure

• Recombination Hotspots (22,000)– Little known about molecular nature

• Excess of THE1A/B retrotransposon-like elements• Six-fold increase in CCTCCCT within elements

www.tqnyc.org/NYC040844/ image/Chromosome.gif

More LD

Less LDInteresting Findings

Linkage Disequilibrium SND

Highest Gene Densityand highest percentageof bases within codons

5

Interesting Findings

• LD is related to function– Immune system shows less LD

• This allows it to be more diverse– DNA/RNA metabolism shows more LD

• This conserves sequence information – thesetypes of gene show incredible lack of diversityacross eukaryotes

– Where advantageous, there is more LD andwhere it would be a disadvantage there is lessLD

Conclusions

• Can extract extensive information fromgenome without resequencing

• Efficient selection of tags can be used forassociation studies

• Genetic information is inherited in blocksthat are not as random as originallythought (governed by LD)