1. BioSci D145 Lecture #8
Bruce Blumberg
4103 Nat Sci 2 - office hours Tu, Th 3:30-5:00 (or by appointment)
phone
TA – Riann Egusquiza
4351 Nat Sci 2– office hours M 1-3
Phone
Updated lectures will be posted on web pages after lecture
Term papers due Friday, March 10 by 12 midnight (23:59.59) (-1 point for each day late)
BioSci D145 lecture 8 page 1 ©copyright Bruce Blumberg All rights reserved

2. Term paper requirements and scoring
Actual paper – 5 pages single spaced 1” margins (references not included).
PLEASE DO NOT WRITE AN ABSTRACT
Specific aims – 2 points (this should be about 3/4 to one page)
Write a paragraph introducing the topic, state why it is important and what are the gaps in knowledge that you will address.
State a hypothesis to be tested
Enumerate 2-3 specific aims in the form of questions that test your hypothesis. Use 1-2 sentences to state how you will answer the question
Finish with a paragraph describing what is innovative about your work and what impact it will have. It is VERY IMPORTANT to state
the human health relevance of your research (if you are doing something biomedical) or
the broader impacts on advancing the frontiers of knowledge (for something that is not relevant to human health).
This is a critically important part of a grant application. You have to convince the reviewer that your work is important and worth funding.
BioSci D145 lecture 1 page 2 ©copyright Bruce Blumberg All rights reserved

3. Term paper requirements and scoring
Background and Significance – 3 points (about pages)
Briefly summarize what is known about the problem.
Not a comprehensive review, just a summary of the important points.
Succinctly state what is not known and why it is important that this research be done
Address knowledge gaps
Are you addressing something controversial?
talk about the controversy and why your work will address it directly.
In about one paragraph, state what is important about your proposed research and why will accomplishing it benefit the research community and world at large.
i.e., what is the potential impact if you are successful
Don’t repeat what was said in specific aims exactly but obviously they should be related.
BioSci D145 lecture 1 page 3 ©copyright Bruce Blumberg All rights reserved

4. Term paper requirements and scoring
Research plan – 4 points (about pages)
In a short paragraph, state what you will do and why it is important. (I know it seems repetitive by now, but reviewers are busy and will be skimming your grant. You need to hit them over the head a few times before they will get your point).
Restate each specific aim from the Specific aims section (one by one)
describe what you will do to address the aim
Break into subaims as appropriate
State the hypothesis to be tested in each
Explain the rationale
Describe briefly what approach you will take
Discuss what you expect to find
Point out any possible problems and alternative approaches
I am mostly concerned with your hypothesis and rationale here.
Not an all-encompassing proposal – 4-5 years by a small team (e.g., your PhD thesis research)
BioSci D145 lecture 1 page 4 ©copyright Bruce Blumberg All rights reserved

5. Where/when are they expressed? Known genes (e.g. from genome projects)
Functional Genomics - The challenge: Many new genes of unknown function
Where/when are they expressed?
Known genes (e.g. from genome projects)
Gene chips (Affymetrix)
Microarrays (Oligo, cDNA, protein)
Novel genes
Differential display
Expression profiling (3 papers this week)
SAGE and related approaches
Massively parallel sequencing
Which genes regulate what other genes?
ChIP (chromatin immunoprecipitation)
What is the phenotype of loss-of-function?
Gene targeting by homologous recombination
Genome wide siRNA
CRISPR-Cas9
BioSci D145 lecture 8 page 5 ©copyright Bruce Blumberg All rights reserved

6. Genome wide analysis of gene function
Loss-of-function analysis is the most powerful way to identify gene function
Direct link between genotype and phenotype
Forward vs reverse genetics
Forward genetics->
Reverse genetics ->
Reverse genetics is much more important than forward genetics in post genomic era – why?
random mutagenesis, then phenotypic analysis
Identity of gene involved not known at the start
associating functions with known genes
Directed mutagenesis of individual genes, phenotypic analysis
Because we have identified many genes by sequencing with no known functions, or even hints about function
What are some useful approaches ?
Mutagenesis (forward genetics)
Systematically mutating each gene (required genome sequence)
Random targeting with viruses or transposons, match genes later
Can id new genes as well as known genes
Generate phenocopies of mutant alleles
RNAi (siRNA), morpholinos, virus induced gene silencing
BioSci D145 lecture 8 page 6 ©copyright Bruce Blumberg All rights reserved

7. Construction of transgenic animals
standard transgenesis
Microinject DNA into a fertilized egg (mouse) or embryo (Drosophila)
some embryos undergo integration of DNA into genome
transgene goes to offspring in a few
Why do these mice have stubby tails?
Applications
rescue of a mutation
promoter analysis
ID elements required for expression
verify function of putative elements
model for dominant forms of human diseases
identify effects of misexpression
Large scale mutagenesis
Gene trap
Enhancer trap
RNAi
BioSci D145 lecture 8 page 7 ©copyright Bruce Blumberg All rights reserved

8. Transgenesis is mostly a gain-of-function technique
Gene targeting
Transgenesis is mostly a gain-of-function technique
Loss-of-function preferred for identifying gene function
Targeted gene disruption is very desirable
to understand function of newly identified genes
e.g., from genome projects
Or gene by gene
produce a mutation and evaluate the requirements for your gene of interest
good to create mouse models for human diseases
knockout the same gene disrupted in a human and may be able to understand disease better and develop efficacious treatments
excellent review is Müller (1999) Mechanisms of Development 82, 3-21.
BioSci D145 lecture 8 page 8 ©copyright Bruce Blumberg All rights reserved

9. Gene targeting (contd)
enabling technology is embryonic stem (ES) cells (or iPS cells)
these can be cultured but retain the ability to colonize the germ line
essential for transmission of engineered mutations
derived from inner cell mass of blastula stage embryos
grown on lethally irradiated “feeder” cells which help to mimic the in vivo condition
essential for maintaining stem cell phenotype
Also problematic for human ES lines
ES cells are very touchy in culture
lose ability to colonize germ line with time
easily infected by “mysterious microorganisms” that inhibit ability to colonize germ line
ko labs maintain separate hoods and incubators for ES cell work
ES cells depend critically on the culture conditions maintain an uncommitted, undifferentiated state that allows germ line transmission.
BioSci D145 lecture 8 page 9 ©copyright Bruce Blumberg All rights reserved

10. Gene targeting (contd) stopped here 2015
isolate genomic clones from ES cell library
Restriction map
Especially exons/introns
Make targeting construct
Want ~5kb genomic regions flanking targeted region
Must disrupt essential exon
Want no functional protein
Verify in cell culture
often useful to fuse reporter gene to the coding region of the protein
gene expression can be readily monitored
Insert dominant selectable marker within replacement region
negative selection marker is located outside the region targeted to be replaced
Electroporate DNA into ES cells, select colonies resistant to positive selection
Integration positive cells then subjected to negative selection
homologous recombinants lose this marker
BioSci D145 lecture 8 page 10 ©copyright Bruce Blumberg All rights reserved

11. Gene Targeting (contd)
Targeting vector
Electroporate into ES cells
Recombination
Selection
identification
BioSci D145 lecture 8 page 11 ©copyright Bruce Blumberg All rights reserved

12. Gene targeting (contd)
Technique (contd)
homologous recombination is verified by Southern blotting
factors affecting targeting frequency (success)
length of homologous regions, more is better.
0.5 kb is minimum length for shortest arm
isogenic DNA (ie, from the ES cells) used for targeting construct is best
locus targeted. This may result from differences in chromatin structure and accessibility
Expand ES cell colonies
BioSci D145 lecture 8 page 12 ©copyright Bruce Blumberg All rights reserved

13. Gene targeting (contd)
Transfer into blastocyst of recipient
Implant into foster mothers (white or black)
Progeny will be mixed color
Breed mixed color F1 mice with homozygous white/black mice
Black progeny derive from germ cells harboring the knockout
Heterozygous for knockout
Breed these to establish lines and determine effects of homozygous mutations
BioSci D145 lecture 8 page 13 ©copyright Bruce Blumberg All rights reserved

14. Gene targeting (contd)
problems and pitfalls
incomplete knockouts, ie, protein function is not lost
but such weak alleles may be informative
alteration of expression of adjacent genes
region removed may contain regulatory elements
may remove unintended genes (e.g. on opposite strand)
interference from selection cassette
strong promoters driving these may cause phenotypes
BioSci D145 lecture 8 page 14 ©copyright Bruce Blumberg All rights reserved

15. Gene targeting (contd)
Applications
creating loss-of-function alleles
introducing subtle mutations
chromosome engineering
marking gene with reporter, enabling whole mount detection of expression pattern (knock-in)
BioSci D145 lecture 8 page 15 ©copyright Bruce Blumberg All rights reserved

16. Example of a “knock-in” model
BioSci D145 lecture 8 page 16 ©copyright Bruce Blumberg All rights reserved

17. Gene targeting (contd)
Applications
creating loss-of-function alleles
introducing subtle mutations
chromosome engineering
marking gene with reporter, enabling whole mount detection of expression pattern (knock-in)
advantages
can generate a true loss-of-function alleles
precise control over integration sites
prescreening of ES cells for phenotypes possible
can also “knock in” genes
disadvantages
not trivial to set up
may not be possible to study dominant lethal phenotypes
non-specific embryonic lethality is common (~30%)
difficulties related to selection cassette
BioSci 145B lecture 6 page 17 ©copyright Bruce Blumberg All rights reserved

18. Conditional gene targeting
Many gene knockouts are embryonic lethal
some of these are appropriate and expected
gene activity is required early
others result from failure to form and/or maintain the placenta
~30% of all knockouts
Clearly a big obstacle for gene analysis
How can this be overcome?
Generate conditional knockouts either in particular tissues or after critical developmental windows pass
Sauer (1998) Methods 14,
BioSci D145 lecture 8 page 18 ©copyright Bruce Blumberg All rights reserved

19. Conditional gene targeting - contd
Approach
recombinases perform site-specific excision between recognition sites
FLP system from yeast
doesn’t work well
Cre/lox system from bacteriophage P1
P1 is a temperate phage that hops into and out of the bacterial genome
recombination requires
34 bp recognition sites locus of crossover x in P1 (loxP)
Cre recombinase
if loxP sites are directly repeated then deletions
if inverted repeats then inversions result
Chromosome deleted for target gene
Target reversed
e.g., Rosa26
BioSci D145 lecture 8 page 19 ©copyright Bruce Blumberg All rights reserved

20. Conditional gene targeting (contd)
Strategy
Make targeting construct (minimum needed for grant)
homologous recombination, select for loss of DT-A
transfect CRE, select for loss of tk
Southern to select correct event
Result called “floxed allele” possible outcomes
inject into blastocysts, select chimeras
establish lines
cross with Cre expressing line and analyze function X X
BioSci D145 lecture 8 page 20 ©copyright Bruce Blumberg All rights reserved

21. Conditional gene targeting (contd)
Tissue- or stage-specific knockouts from crossing floxed mouse with specific Cre-expressing line
requirement for Cre lines
must be well characterized
promoters can’t be leaky
Andras Nagy’s database of Cre lines and other knockout resources
BioSci D145 lecture 8 page 21 ©copyright Bruce Blumberg All rights reserved

22. Figure Floxing mice
pgs3e-fig jpg

23. Conditional gene targeting (contd)
advantages
can target recombination to specific tissues and times
can study genes that are embryonic lethal when disrupted
can use for marker eviction
can study the role of a single gene in many different tissues with a single mouse line
can use for engineering translocations and inversions on chromosomes
disadvantages
not trivial to set up, more difficult than std ko but more information possible
requirement for Cre lines
must be well characterized regarding site and time of expression
promoters can’t be leaky (expressed when/where not intended)
BioSci D145 lecture 8 page 23 ©copyright Bruce Blumberg All rights reserved

24. Other approaches to genome manipulation
Transgenic and knockout technology is species dependent (doesn’t work in all species – need ES cell equivalent). How else can we accomplish gene disruption in a targeted way ?
RNAi approaches (Luo paper this week)
Nuclease based methods - introduce double-stranded breaks
ZFN – zinc finger nucleases
TALEN nucleases
Meganuclease
CRISPR/Cas9 – RNA guided nuclease (Gilbert, Liu papers this week)
Meganucleases
Based on naturally occurring restriction enzymes with extended DNA binding specificity (relatively limited application)
TALEN and ZFN nucleases
Artificial fusion proteins combining an engineered DNA binding domain fused to a nonspecific nuclease domain from FokI restriction enzyme
ZNF – zinc finger repeats
TALE – repeats
Main limitation is the sequence specificity that can be engineered in.
BioSci D145 lecture 8 page 24 ©copyright Bruce Blumberg All rights reserved

25. CRISPR-Cas gene editing
CRISPR = clustered, regularly interspaced, short palindromic repeat
Sander & Joung (2014) CRISPR-Cas systems for editing, regulating and targeting genomes, Nat Biotech 32:
Lander (2016) The Heroes of CRISPR Cell 164, 18-28
Cas9 – RNA-guided nuclease
Functions as a bacterial immune system. Foreign DNA is incorporated between CRISPR repeat sequences, transcription generates crRNA.
Hybridizes to tracrRNA (also encoded by CRISPR system), guides Cas9 nuclease to the target
Cas9 nuclease introduces double stranded breaks into target
Repair of this break can introduce deletions frameshifts, etc.
BioSci D145 lecture 8 page 25 ©copyright Bruce Blumberg All rights reserved

26. CRISPR-Cas gene editing
CRISPR/Cas9 (contd)
For gene targeting, we would fuse the target RNA to the tracrRNA and introduce this into the target cell, embryo, etc together with Cas9 nuclease
Introduces ds breaks at target sequence which may introduce desirable mutations.
Potential issues
Can’t specify what happens after ds break (deletion, frameshift, insertion, etc
Some question about specificity of the crRNA introduced – may have off-target effects
Some sequence preferences of Cas9 nuclease may limit utility
BioSci D145 lecture 8 page 26 ©copyright Bruce Blumberg All rights reserved

27. CRISPR-Cas gene editing - applications
BioSci D145 lecture 8 page 27 ©copyright Bruce Blumberg All rights reserved

28. Generating phenocopies of mutant alleles
How to inactivate endogenous genes in a targeted but general way?
RNAi = RNA interference
Observation is that introduction of double-stranded RNAs into cells lead to destruction of corresponding mRNA (if there is one)
Principle is siRNA – small interfering RNAs
These generate small single stranded RNAs that target mRNAs for destruction by
RISC – RNA interference silencing complex
First applied in C. elegans where it works extremely well
Can introduce siRNA into cells even by feeding to the worms!
Works very well in Drosophila
variably in mammalian cells
Poorly in Xenopus – Why?
Xenopus has an endogenous dsRNA helicase
BioSci D145 lecture 8 page 28 ©copyright Bruce Blumberg All rights reserved

29. Dicer complex generates short duplexes from dsRNA in the cell
RNAi (contd)
Dicer complex generates short duplexes from dsRNA in the cell
Important to have 2-nt overhangs
siRNAs are generated from these fragments
Antisense strand binds to mRNA and this recruits the RISC - RNAi silencing complex
Complex leads to mRNA cleavage and destruction
Two important reviews to read
McManus and Sharp (2002) Nature Reviews Genetics 3,
Dykxhoorn et al. (2003) Nature Reviews, Molecular Cellular Biology 4,
BioSci D145 lecture 9 page 29 ©copyright Bruce Blumberg All rights reserved

30. Always form a hairpin structure with mismatches in stem
RNAi (contd)
Micro RNAs are small cellular RNAs that previously lacked any known function
Always form a hairpin structure with mismatches in stem
Micro RNAs direct gene silencing via translational repression
(miRNAs) are mismatched duplexes that dicer processes into stRNAs (small temporal RNAs)
Use same cellular complex as siRNAs
Perfect matches -> target cleavage
Imperfect matches -> translational repression of target
Two important papers
Giraldez et al (2005) Science 308, (microRNAs regulate brain morphogenesis)
Lecellier et al (2005) Science 308, (microRNA mediates antiviral defenses in human cells)
BioSci D145 lecture 9 page 30 ©copyright Bruce Blumberg All rights reserved

31. Parallels between siRNA and miRNA-directed RNAi
RNAi (contd)
Parallels between siRNA and miRNA-directed RNAi
BioSci D145 lecture 9 page 31 ©copyright Bruce Blumberg All rights reserved

32. Ways to generate short RNAs that silence gene expression in vitro
RNAi (contd)
Ways to generate short RNAs that silence gene expression in vitro
a) chemical synthesis of siRNA, introduce into cell
b) synthesize long dsRNA, use dicer to chop into siRNA
c) introduce perfect duplex hairpin, dicer generates siRNA
d) make miRNA based hairpin, dicer generates silencing RNA
Introduce into cells or organism by microinjection, transfection, etc.
Expression is transient, loss of function ALWAYS partial
can only generate phenotypes for a short time after introduction
BioSci D145 lecture 9 page 32 ©copyright Bruce Blumberg All rights reserved

33. a) produce long hairpin from pol II promoter, let dicer make siRNA
RNAi (contd)
Ways to generate short silencing RNAs in vivo – need continuing expression to generate stable phenotype
a) produce long hairpin from pol II promoter, let dicer make siRNA
b) produce two transcripts from pol III promoter, let anneal in cells
c) produce a short hairpin from pol III promoter (or viral vector), let dicer generate siRNAs
d) produce imperfect hairpin from pol II promoter, let dicer generate miRNAs that direct gene silencing
BioSci D145 lecture 8 page 33 ©copyright Bruce Blumberg All rights reserved

34. RNAi for whole genome functional analysis
RNAi (contd)
RNAi for whole genome functional analysis
First generate library of constructs that generates siRNA or stRNA
Or synthesize a complete library of siRNAs
Introduce these into cells, embyos (fly, frog, mouse) or animals (C. elegans, plants)
For C. elegans, make the library in E. coli and simply feed bacteria to worms
Must microinject or transfect with other animals
Evaluate phenotypes
IMPORTANT!
RNAi is inherently transient unless you are expressing the siRNAs from a stably integrated plasmid
RNAi is a knock-down (not knock-out) method.
BioSci D145 lecture 9 page 34 ©copyright Bruce Blumberg All rights reserved

35. Phenotypic rescue by RNAi – synthetic lethal and related approaches **
How can we find other members of pathways we already know something about?
Or, how can we find drugs that act on a pathway to kill cells (e.g., cancer cells?)
Synthetic lethal is one relatively new and promising approach
2 mutations are synthetic lethal if either single mutation is viable but the double mutant is lethal
BioSci D145 lecture 8 page 35 ©copyright Bruce Blumberg All rights reserved

36. Phenotypic rescue by RNAi – synthetic lethal and related approaches
How can we find other members of pathways we already know something about?
In cancer screening, what if we combine a mutation and a drug to find a combination that increases the kill rate, or reveals a phenotype that is similar to the total loss of function?
Could find novel drug targets (pathways that kill cells in presence of siRNA+drug (usually sublethal amount of drug)
Can find genes that are targeted by drugs – pathway analysis
Could find new biomarkers that are required for cell viability following drug treatment
BioSci D145 lecture 8 page 36 ©copyright Bruce Blumberg All rights reserved

37. Mohr et al., 2010, Genomic Screening with RNAi: Results and Challenges. Annual Review of Biochemistry, 29: 37-64
BioSci D145 lecture 8 page 37 ©copyright Bruce Blumberg All rights reserved