Transcript ForwardGeneticsMapping2012

Forward Genetics
Letting the Genome Tell You What
Genes Are Required for the Biological
Process You are Studying
-Random screening as an unbiased appoach for gene discovery
-Key entry point for determining molecular mechanisms
There is no substitute for loss-of function
phenotype for finding out what your gene is doing!
A Little Genetics
Genotype: what alleles you have (heterozygous, homozygous, etc.)
Phenotype: what you look like (wild type or mutant for some trait)
Recessive allele: genotype needs to be homozygous mutant in order to see
mutant phenotype
Dominant allele: have mutant phenotype even when heterozygous for mutant
allele
Zygotic gene: genotype of embryo determines phenotype of embryo
Maternal effect gene: genotype of MOTHER determines phenotype of embryo
A-P polarity set up in egg chamber
bicoid mRNA
microtubule-based
-
oskar mRNA
protein + nanos RNA
+
bicoid, oskar and nanos are examples of genes acting in the OVARY that influence patterning of the EMBRYO
Therefore, the GENOTYPE of the MOTHER determines the PHENOTYPE of the EMBRYO
(Maternal Effect)
Examples:
Zygotic recessive
(phenotype)
m/+ X m/+:
m/+ (wild type)
+/+ (wild type)
m/m (mutant)
Zygotic dominant
m/+ X m/+:
m/+ (mutant)
+/+ (wild type)
m/m (mutant)
Maternal effect recessive
Mom
Dad
Embryo (phenotype)
m/+ X m/+:
m/+ (wild type)
+/+ (wild type)
m/m (wild type)
Mom
Dad
m/m X m/+:
Embryo
m/+ (mutant)
m/m (mutant)
Mom
m/+ X
Embryo
m/+ (wild type)
m/m (wild type)
Dad
m/m:
Types of Mutants
Hypomorph: Loss of function
Often recessive - genotype needs to be
homozygous for mutation (m/m) to see phenotype
Amorph or Null: Complete loss of function
(behaves like a deletion of gene)
Hypermorph: Gain of function
Often dominant - can see a phenotype even if
genotype is only heterozygous (m/+)
Antimorph: Behaves stronger than null
(e.g. dominant negative)
Neomorph: New function (e.g. gene now expressed in
ectopic location)
Chemical Mutagenesis
EMS (Ethyl methane sulfonate): e.g. flies
ENU (n-Ethyl-n-nitrosourea): e.g fish and mice
Chemically modify DNA bases to induce replication errors
Preferentially induce point mutations (but some small deletions)
Advantages:
Most random of mutagens
Alleles of different strengths
Disadvantages:
Harder to identify lesion (clone gene)
Radiation Mutagenesis
X-rays and gamma rays most common
Induce double strand DNA breaks
-Deletions
-Inversions
-Translocations
Advantages:
Often produce null mutations
Easy to identify lesion
(often just by looking at chromosomes)
Disadvantages:
Usually take out multiple genes
Insertional Mutagenesis (Transposon or Retroviral)
Insertion of transposon or viral sequences affects gene function
Can control transposon jumping by separating transposase
from transposon
Advantages:
Easy to identify lesion, clone gene
(gene is “transposon tagged”)
Disadvantages:
Non-random integration/mutagenesis
-affects target distribution
-affects allele strength
Currently being done in flies, fish and mouse
Traditional Screens for Recessive Mutations
EMS
Make mutant sperm
Make heterozygous
mutant individuals
Make mutant
brothers and sisters
Cross heterozygous
brothers and sisters
to make homozygous
mutant offsprint
Nusslein-Volhard and Weischaus Nobel Lectures
http://www.nobel.se/medicine/laureates/1995/
Summary of X, 2 and 3
Nusslein-Volhard and Weischaus Nobel Lectures
http://www.nobel.se/medicine/laureates/1995/
paired
WT
knirps
Large Scale Forward Mutagenesis in Zebrafish
Nusslein-Volhard
and Dreiver Labs
1996
Problems with traditional homozygous mutant screens:
-Genes are not all equally mutable
-some are small targets or tough to induce loss of function
(e.g. microRNAs)
-Genes can be redundant
-knocking out one copy doesn’t always give phenotype
-Genes are pleiotropic
-if embryo dies before your process “happens”, can’t tell if
that gene is required
Sensitized Genetic Screens: the sevenlessTS screen
Simon, Bowtell, Dodson, Laverty and Rubin, 1991
sevenless: receptor tyrosine kinase required for R7 specification
Problem: sev is specific to eye, but downstream RTK pathway common to all
RTKs and therefore embryonic lethal
How do you identify sev pathway components (and therefore components of all
RTK signaling)?
Sensitized Genetic Screens: the sevenlessTS screen
Simon, Bowtell, Dodson, Laverty and Rubin, 1991
sevTS: Making flies “on the edge”
22.7°C R7 mostly present--screen for dominant enhancers of sev
-R7 now lost
24.3°C R7 mostly absent--screen for dominant suppressors of sev
-R7 now restored
e.g. ras
ras-/ras+ = wild type
ras-/ras- = uninformative dead embryo
sevTS 24.3°C with ras-/+ = no R7 cell
Normally ras is recessive but now behaves as a dominant enhancer of sevTS
Sensitized Genetic Screens: generic eye screens
Express programmed cell
death gene in eye
Look for
suppressors/enhancers
of programmed cell
death pathway
Mosaic Screens
Genetic mosaics can be created by mitotic recombination
induced by X-rays or a site-specific recombinase
FLP/FRT in Drosophila: Golic and Lindquist, 1989
Essential reagents: Xu and Rubin, Chou and Perrimon
Treisman Lab
Forward/Reverse Genetics: Whole genome RNAi Screens
Advantages:
-Save gene identification step
Disadvantages:
-Depends on genome sequence/annotation
-Efficiency of RNAi is variable
-Issues with delivery of dsRNA trigger
Cell Culture: Transfect or bathe cells with dsRNA
Cell Culture: Lentiviral vectors expressing dsRNA
C. elegans: Feed worms bacteria expressing dsRNA
Flies: Inject embryos with dsRNA
Flies: UAS-shmiRNAs for every gene in genome
Positional Cloning
Finding a needle in a haystack
Finding a single bp change in 3.6 x 108 (Fly)
3.4 x 109 (Fish)
6.0 x 109 (Human/Mouse)
First fly gene: Ultrabithorax 1979
First human disease genes:
chronic granulomatous disease 1986
Duchenne muscular dystrophy 1987
Cystic fibrosis 1989
(approx. 1200 disease genes now cloned)
First fish gene: one-eyed pinhead 1998
How do you find your gene?
Identify a transposon-induced allele of your gene:
transposon then “tags” genomic region of interest
Find a genomic lesion (deletion/inversion/transposition)
allele of your gene:
breakpoints in genome identify region of interest
For point mutants: meiotic mapping and positional cloning
Genetics 101
Mendelian Inheritance
Linkage
a ;b
+ ;+
a b X + +
a b
a ;b
;
X
+ +
a b
a ; b + ;+
+ ; b a ;+
a b a b
a b a b
1 : 1 : 1 : 1
Parentals = Recombinants
a b
a b
X
+ +
+ +
a b
+ +
X
a b
a b
+ b a +
+ +
a b a b
a b
Parentals > Recombinants
160
40
% Recomb
Recombination = # Recombinants
X 100 = Map Units
Frequency
Total
CentiMorgans (cM)
40/200 x 100 = 20 cM
a b
a b
Three genes: a, b, c
a c
a c
a c
+ +
Meiotic Mapping
X
+ +
+ +
c b
c b
X
+ +
+ +
X
a c
a c
c b
+ +
X
c b
c b
+ c a +
+ +
a c a c
a c
Recombinants
Parentals
190
10
10/200 x 100 = 5 cM
a c
a c
a
+ +
c b
Parentals
c b
c b
c +
c b
Recombinants
170
30
30/200 x 100 = 15 cM
c
5
+ b
c b
b
15
To Clone Gene C
a
1)
2)
3)
4)
c
b
Link meiotic map to physical map (DNA)
Identify markers and map crossovers to define limits of C
Identify genes within this region
Determine which gene is C
Markers for Meiotic/Physical Mapping
“Classically” done using visible dominant and recessive mutations
-Low density of useful markers
-Less rooted in physical map
Can improve the density of visible markers using transgenes
e.g. w+ transposons in flies
Modern methods directly assess DNA polymorphisms
Random markers
e.g. Randomly Amplified Polymorphic DNA (RAPDs)
PCR w/ primers of random sequence, get few random products
Presence or absence of product can depend on as little as single bp change
Don’t require prior knowledge of genome sequence
Allows “entry” into physical map (identifies STS near gene of interest)
Simple sequence length polymorphisms (microsatellite DNA, e.g. CA repeats)
PCR shows small polymorphic changes in repeat number
Advantage: easy to analyze
Disadvantage: Not enough (low density)
Single nucleotide polymorphisms (SNPs)
Advantage: Maximum possible density
Disadvantage: Can be difficult to assay
Syvanen, Nat Rev Gen 2001
SNPs alter oligo annealing
Suitable for microarrays
SNPs alter oligo annealing
+/- PCR product
Single nucleotide
“mini-sequencing”
Suitable for microarrays
Afymetrix offers SNP Chips that can genotype 10-50,000 SNPs
Also,
-Single strand conformation polymorphisms (detected in gels)
-Denaturation HPLC
-Mass-spec DNA sequencing
Sounds easy but…
-Compare mutagenized chromosome with interesting phenotype to
control, parental chromosome that was isogenized before screen
-Found 165 sequence changes on third chromosome.
-Could only verify 103 (some false positives). Others likely not found (false
negatives) since not all regions have good sequence.
-Of these, 11 made changes to ORFs. Therefore, still some work to figure
out correct gene.
The First Association Between
the Meiotic and Physical Maps
RAPDs
Allele 1: products A and B
A
B
Allele 2: Change in site 2
No product A
http://avery.rutgers.edu/WSSP/StudentScholars/project/archives/onions/rapd.html
Bulk Segregant Analysis: Look for Linkage
Start with mutation heterozygous in strain 1
Use strain 2 as polymorphic mapping strain
Strain 1: m/+
a1
m
b1
a1
+
(lack of recombination)
Strain 2: +/+
X
b1
a2
+
a2
+
a1
m
a2
+
X
b2
b2
b1
b2
Mutant = m
Polymorphic Markers = a1, a2, b1, b2
Mutant ‘bros
X
Backcross to
Parental Strain 1
a1
a1
m
+
wt ‘bros
a1
m
a1 a1
m m
a1
m
a2
+
a1
m
a2
+
a1
m
b1
b1 b2
b1
b2
b1
b1
b1
b1
b1
Sort mutant vs. wt ‘bros
Make DNA from pools
a1
+
b1
Any combo
b is mixed 1 and 2, therefore unlinked to m
a is always (or mostly) 1, therefore linked to m (few recombinations)
1) Identify closely linked polymorphic markers
oep
-Bulk segregant analysis identifies 15AH and 20K as “close” to oep
Now look for recombinants between closely
linked marker and gene
Backcross to
Strain 1: m/+
a1
m
a1
+
b1
b1
Strain 2: +/+
X
a2
+
a2
+
a1
a2
m X +
b2
b2
b1
Mutant = m
Markers (alleles) = a1, a2, b1, b2
a2
m
b2
a1
m
b1
b2
X
Parental Strain 1
a1
a1
m
+
b1
b1
Examine INDIVIDUAL
offspring
= individual meioses
(gametes) of parents
-Look for recombination between close marker and gene
-Screen 3100 “meioses” to find rare recombinants
-Save Recombinants: can go back and analyze later with new
markers to further define WHERE recombination took
place and therefore limits of WHERE oep can be!
1) Identify closely linked polymorphic markers
oep
-Bulk segregant analysis identifies 15AH and 20K as “close” to oep
-Analyze DNA from 3122 INDIVIDUAL mutant ‘bros
-find 1 recombinant b/w 15AH and oep
0.03 cM = 18 kb (IF 600 kb/cM)
-find 5 recombinants b/w 20K and oep
0.16 cM = 96 kb
Now close enough to go after DNA in region
(link meiotic map to physical map)
2) Create Physical Map of Region (Genomic “walk”)
oep
Use 15AH and 20K markers to gain “entry” into genomic region
-Make probes using RAPD PCR bands from each
-Probe genomic library to isolate clones 134 and 32
-Use ends of these to isolate contiguous clones (“walk”)
-Stop when two directions of walk meet
If genome sequence is available, don’t have to walk since you know sequence
of interval between markers
2) Create Physical Map of Region (Genomic “walk”)
oep
Clone 134F10 failed the “deletion test”
-see if all of clone is missing in deletion of oep genomic region
-if not missing, then some of clone’s DNA is from other region
-suggests clone is chimeric (contains different parts of genome)
-would be disaster to continue “walking” from chimeric clone
could jump to entire new (irrelevant) region or new chromosome
3) Recombinant Fine Mapping
oep
-Subclone 14 and 240 into cosmids
-Use ends to make STS’s (need spaced sequence info)
-Use sequence to identify additional polymorphic markers
SSCPs and CAPS
-Go back to previously identified recombinants (1 on left, 5 on right)
-Use new markers to map recombination events
e.g. 46T7 has allele of “strain 2”, so recombination is b/w 46T7 and oep
-Therefore oep is between 46T7 and 32T7
4) Identify the Gene
oep
-Use genomic DNA to probe cDNA library from stage of interest (223 cDNAs!)
-Find out that these represent 13 “classes” of cDNA
-Use in situ hybridization to see if any expression patterns fit predictions
-Sequence mutant genomic DNA to identify potential bp changes that are
responsible for mutant phenotype
-Rescue with in vitro synthesized RNA
-Gold Standard for Gene ID: sequence point mutations and
rescue mutant defect with transcript
How would things be changed today?
-High density polymorphism map produced so don’t need to
search for polymorphic markers
-Genome sequence being completed so don’t need to walk
-Large scale EST (cDNA) sequencing so know transcript
distribution and candidate genes (at least those that are
correctly annotated!)
-Can use morpholinos (RNAi in other species) to test
candidate transcripts
-Whole genome sequencing becoming helpful for identifying
mutations
http://www.hapmap.org/
Goal: Determine the existing human haplotypes with a defined set of SNPs
Long Term Goal: Associate haplotypes with phenotypes for
-cloning disease genes
-understanding genetics of complex traits
-pharmacogenomics