High-throughput%20Genotyping

1
High-throughput Genotyping to Map Genes for
Complex Disease
Shelley A. Cole, Ph.D. Director, Genetics Core
Laboratory
2
Gene Mapping

• What is gene mapping?
• What is high-throughput genotyping?
• What are some of the logistic issues of running a
  high-throughput genotyping operation?
• High-throughput phenotyping (gene expression)
  analyses

3
Gene Mapping Studies

• - Designed to identify genes that confer risk
• for complex disease
• Uses a genome scan
• Identifies regions of the genome that have
• genes that affect disease risk factors
• Definitions
• genotype - what your gene(s) encode
• phenotype - what results from the expression of
• your genes and environment

4
Complex Diseases

• Not inherited in a Mendelian manner
• Characterized by both genetic and environmental
  risk factors
• Can have multiple genetic and environmental causes

5
Genotyping Projects in the Genetics Core
Laboratory

• The San Antonio Family Study
• The Strong Heart Family Study
• Viva La Familia
• A Neurobehavioral Family Study of Schizophrenia

• Genetic Epidemiology of Childhood Skeletal
  Maturation (Fels)
• Genetics of Coronary Artery Disease in Alaskan
  Natives
• Genetic Modifiers of Severity in Sickle Cell
  Anemia

6
What is needed for genotyping for a genome scan?

• families, with accurate information on family
  structure
• good DNA
• markers

7

• We receive frozen
• cells from blood.
• The blood samples are anonymous.
• We isolate DNA.

Drawing by Darryl Leja, NHGRI
8
Human Chromosomes
9
We build a genetic road map of the
chromosomes using genetic markers as landmarks.
10
What are the genetic markers we use?
11
Genetic Markers for Genome Scans

• Need to be common throughout the genome
• Need to be easily assayed (low cost, high
  throughput)
• Need to be highly informative (have many alleles)
• Need to be reliably inherited

12
chromosome
DNA
A - T C - G G - C T - A T - A G - C C - G A - T
13
Markers are variable repeat units in DNA sequence.
chromosome from one person
?
CACACA
chromosome from another person
?
CACACACACA
chromosome from yet another person
?
CACACACACACACA
14
We all have 2 chromosomes, one from our mother
and one from our father.
chromosome from our mother
?
CACACACA
chromosome from our father
?
CACACACACACA
15
Genetic Markers for Genome Scans

• Microsatellites, short tandem repeats (STRs)
• characterized by JL Weber (1989)
• Have repeat units of 2, 3, or 4 nucleotides
• Can be assayed using PCR in groups of 10 to 20
  (sort by size color on a gel)
• Have many alleles, are highly informative

16
How do we type 400 markers in 1200 people?
480,000 marker typings !

• We try to combine as many steps as possible.
• We use specialized lab equipment.
• We use PCR.
• We label the markers with fluorescent dyes.
• We can sort the markers by size and dye.

17
Genotyping STRs
allele 1, 208bp allele 2, 212bp
(N)100CACACACA (N)100
(N)100 CACACACACACA (N)100
denature and amplify by PCR
run on a gel
allele 1
(CA)4

(GT)4
212
208
allele 2
(CA)6

(GT)6
fluorescently labeled primer
18
Hydra Micro dispenser

• for dispensing DNA
• from 96 wells to 384

19
(No Transcript)
20
(No Transcript)
21
ABI 377 Automated DNA Sequencer
22
(No Transcript)
23
Segregating Genotypes
father
mother
child 1
child 2
child 3
24
Pedigree with some family members
having high cholesterol
High cholesterol
Normal cholesterol
25
Same pedigree with typings at marker A
A1/A2
A3/A4
A3/A2
A1/A3
A1/A4
A5/A3
A1/A5
A1/A3
26
Where is the Gene for high cholesterol?
27
Improvements due to workflow

• Decreasing tedium/ increasing efficiency
• Improvement in data quality (typos,
  inconsistencies, immediate feedback on mendelian
  errors)
• Improvement in data tracking (what was done when
  and by whom, what wasnt done and why, what was
  changed and when, what is and will be done,
  status of data and samples)

28
Table with genotypes additional, pertinent
information
29
Checking Sending data to PopGen ScanABI
30
Pedsys merged-marker file generation
31
Data generated in a genome scan

• Genotype data (400 markers, 1000-3600 subjects)
• phenotype data (recall)
• pedigree information
• gene map (versions)
• analyses
• results (versions)

32
Common Diseases in Mexican Americansin San
Antonio

• Diabetes
• 2-3 times higher prevalence than in non-Hispanic
  whites
• prevalence of 20-25 over age 40
• Obesity
• mean BMI of 29 kg/m2
• 25-35 severely overweight
• Heart disease
• more adverse risk factors than non-Hispanic
  whites
• (higher blood pressure, triglycerides, BMI)

33
San Antonio Family Heart Study Study Design

• Families ascertained through a randomly chosen
  40-60 year old
• Probands were Mexican American
• Probands had a spouse and at least 6 offspring
  and/or brothers and sisters who were at least 16
  years old and living in San Antonio.
• Each family consists of
• The proband and his or her spouse
• All-available first-, second-, and third-degree
  relatives of the proband and the probands spouse
• Individuals who have married into the family
• 1,431 participants from 42 multigenerational
  Mexican American families
• gt850 recalled at 5 and 10 years after initial
  clinic visit

34
San Antonio Family Heart StudyQuestionnaire Data
Age, sex, residence Diet Identities of
relatives Smoking Ancestry Alcohol
use Acculturation Physical activity Socioeconom
ic indicators Prescription drug use History of
MI, stroke, angina pectoris, other disorders Rep
roductive history
35
San Antonio Family Heart StudyPhysical
Examination
Glucose tolerance test Blood pressure EKG Anthropo
metrics Bioimpedance Blood draw for genotype and
phenotype determinations Carotid artery ultrasound
36
San Antonio Family Heart Study
37
San Antonio Family Heart Study Relationships
among 1431 Family Members
Degree of Relationship Relationship
No. of
Pairs First Sibs 1,546
Parent-offspring 1,306 Second
Avuncular 2,996 Grandparent-grandchild
589 Half-sibs
172 Double first cousins
19 Third First cousins
3,313 Grand avuncular
910 Half-avuncular
405 Great
grandparent-grandchild 158
Other third degree 75 Fourth First
cousins once removed 3,205
Half-first cousins
511 Great grand avuncular
99 Other fourth degree
152 Fifth Second cousins
1,007 First cousins, twice
removed 450 Other fifth degree
55 Sixth

329 Total 17,297
38
San Antonio Family Heart Study Data Cleaning and
Analysis
Sequential Oligogenic Linkage Analysis Routines
39
What do we do when we find a chromosomal region
linked to a phenotype?

• Try to narrow down the chromosomal region
• Identify candidate genes in the region
• Sequence the gene regions to identify
  polymorphisms, usually SNPs (single nucleotide
  polymorphisms)
• Genotype the SNPs, and see if any are the
  mutation responsible for the linkage

40
San Antonio Family Heart Study Program Director
Jean W. MacCluer, Ph.D.
SOUTHWEST FOUNDATION LABORATORIES Genetics Core
Lab Population Genetics Genetic
Epidemiology Lipoprotein Genetics Biochemical
Genetics
UT HEALTH SCIENCE CENTER AT SAN
ANTONIO Division of Clinical Epidemiology
CONSULTANTS AND COLLABORATORS AT Univ. of
Maryland Georgia State Univ. Mayo
Clinic NHLBI SUNY Albany Tufts Univ. Boston
University UT Austin UTHSC Houston Univ. of
Vermont Wright State Univ.
41
Gene expression as a novel phenotype in genome
scans
42
Baboon as a model for genetic studies

• Genetically similar to human
• Extended pedigrees
• Chromosome map is available

43
Genome scan to identify QTLs for obesity-related
phenotypes in baboons

• Ongoing program project at SFBR using baboons
  from well-documented pedigrees.
• A 10 cM microsatellite genetic linkage map has
  been constructed to permit genomewide searches
  for novel genes that affect quantitative risk
  factors for disease.
• Omental adipose tissue biopsies from 500 baboons
  are being used to generate quantitative
  expression data from obesity candidate genes.

44
Baboon as a model for human obesity and related
disorders

• Spontaneous development of obesity
• Animals share same environment and diet
• tissue biopsies are obtainable

45
The baboon as a non-human primate model of obesity

• Like humans, fat cell volume, not number,
  increases with increasing fat mass, and fat cell
  volume is heritable (D. Lewis).
• Genes account for 51 of the variance in mean
  adult body weight, and 12 of the variance in
  body weight stability (Jaquish et al., 1997).
• Obesity occurs spontaneously in some captive
  baboons fed ad libitum low fat monkey chow.

46
Why are we collecting omental adipose tissue
samples from 500 pedigreed baboons?
Ultimately To identify genes involved in
obesity. More practically To identify novel
genes that cause differences in gene expression
in adipose tissue among baboons.
47
Candidate genes for obesity expressed in adipose
tissue

• lipoprotein lipase
• leptin
• uncoupling protein
• glucose transporter
• agouti signalling protein
• tumor necrosis factor
• adipsin

• 3-adrenergic receptor
• peroxisome proliferator activated receptor
• leptin receptor
• retinoid receptors
• acid phosphatase (ACP1)

48
Leptin as a candidate gene for obesity

• A hormone expressed in adipose tissue.
• Leptin levels rise after feeding, and fall after
  fasting.
• Leptin signals the brain (and other tissues) as
  to the bodys energy state through the leptin
  receptor.
• Serum leptin levels are highly associated with
  amount of body fat.

49
Is leptin gene expression in baboon adipose
tissue measurable (and variable)?
50
Baboon leptin cDNA (348 bp)
51
Ribonuclease Protection Assay
radiolabeled antisense probe
RNase digestion
gel electrophoresis
hybridize
RNase
hybridized product
total RNA
52
Quantitative variation in baboon omental
adipocyte LEP mRNA
LEP
28S
53
Baboon adipose tissue LEP mRNA expression versus
serum leptin levels

280


140


LEP mRNA (pg)



70





35





2 4 6 8 10 12
14
Serum Leptin Levels (ng/ml)
r 0.397, p lt 0.05
54
Weight versus leptin mRNA expression
55
Mean adipocyte cell volume versus leptin mRNA
expression
56
Is leptin gene expression in baboon adipose
tissue heritable?
57
Heritability of baboon omental adipose tissue LEP
mRNA expression levels
h2
0.30

p 0.04
58
Leptin levels in males and females

• Females have higher leptin levels than males
• Genotype-by-sex interaction in leptin
  levels Martin et al., 2002

59
Genotype by sex interaction
20
15
Phenotypic Level
10
5
Genotype 1
Genotype 2
60
Objective
Confirm and extend our human studies of G x S
interaction on leptin levels in a non-human
primate model.
61
The Method
Sequential Oligogenic Linkage Analysis Routines
http//www.sfbr.org/public/software/solar/index.ht
ml
62
Leptin measurements

• Serum leptin measured by RIA (Linco)
• leptin mRNA measured using ribonuclease
  protection assay

63
Summary

• There is a significant genotype-by-sex
  interaction in the expression of serum leptin
  levels in baboons.
• We were unable to detect a significant
  genotype-by-sex interaction for omental adipose
  tissue leptin mRNA expression.

64
Conclusion
Our data implies that the sexual dimorphism in
serum leptin levels results from genes that are
differentially expressed in males and females,
and that these genes may act on
post-transcriptional processes.
65
Collaborators
Molecular Genetics Laura Cox Jim Hixson Phil
Morin (Axys Pharm.) Jeff Rogers Biochemical
Genetics John VandeBerg Lipoprotein
Genetics David Rainwater
Statistical Genetics Laura Almasy John
Blangero Gerry Brush Tony Comuzzie Bennett
Dyke Candy Kammerer Jean MacCluer Michael
Mahaney Lisa Martin Brackie Mitchell Jeff
Williams Sarah Williams-Blangero
SAFHS Clinic Mike Stern Baboon Care Dee
Carey Karen Rice