1. RNA interference (RNAi)
white RNAi hairpin
Wild-type

2. A Brief History of RNAi Hint First description
1995, Guo and Kemphues: Injection of par-1 gene sense RNA into the gonad of
C. elegans induced par-1 null phenocopies at the same high frequency as injection
of anti-sense RNA.
First description
1998, Fire et al.: Injection of dsRNA for specific genes into C. elegans caused
a specific disappearance of the gene products from somatic cells and F1 progeny
effect was on stability of mRNA
crossed cellular barriers
only a few molecules of dsRNA per cell required
dsRNA from exons but not introns had effect

3. 21-23 nt RNAs = short interfering RNAs (“siRNAs”)
Brief History contd.
Mechanism
1999, Hamilton and Baulcombe: Arabidopsis plants undergoing post-transcriptional gene silencing (PTGS) contained nt long RNAs that were complementary to both strands of the silenced gene and that were processed from a long dsRNA precursor
2000, Zamore et al. : Used Drosophila embryo extracts to show that long dsRNAs are processed to nt RNAs that direct targeted mRNA cleavage
2001, Bernstein et al.: Using Drosophila S2 cell extracts, these authors described the enzyme for producing the 21-23nt RNAs: an Rnase III enzyme, Dicer.
21-23 nt RNAs = short interfering RNAs (“siRNAs”)

4. siRNA Structure and Formation
5’ phosphate
siRNA
Dicer cleaves dsRNA
siRNAs are incorporated into RISC
(RNA Induced Silencing Complex)
siRNAs unwind and guide RISC
to a substrate
mRNA substrate cleavage
(From McManus and Sharp, 2002
& Hannon, 2002)

5. Small temporal RNAs (stRNAs)
C. elegans let-7 and lin-4:
isolated as heterochronic mutants
negative regulators of specific protein-coding genes
(target the 3’ UTR)
encode small RNAs, synthesized as 70 nt precursors
post-transcriptionally processed to a 21 nt mature form
by Dicer
regulate expression at the level of translation
archetypes of a large class of endogenously encoded
small RNAs now called microRNAs (miRNAs)

6. siRNAs vs stRNAs

7. RNAi in worms heritable systemic Methods:
injection of dsRNA into gonad
soak worms in dsRNA
feed worms E. coli transformed with a plasmid expressing
S and AS RNAs

8. Large scale RNAi screen in worms
Gonczy et al., 2000
targeted 96% of ORFs on chromosome III
used PCR primers tailed with T3 and T7 promoter sequences
PCR product size: 500+ bp
product encompassed > 90% coding sequence

9. RNAi in Flies non-inheritable cell autonomous Methods:
injection of dsRNA into syncytial blastoderm embryos
(Kennerdell and Carthew, 1998)
observed variability in interference activities of
different dsRNAs (null phenotypes and mosaics)
phenotype localized to site of injection
Inducible RNAi transgenes
dsRNAs
Snap-back RNAi
extended hairpin loop RNAs
Genomic cDNA hybrids

10. Examples and Results
2000, Lam and Thummel: established P-element transformants
that use the heat-inducible hsp70 promoter to drive expression
of a snapback dsRNA corresponding to the coding region of EcR
and BFTZ-F1
established 8 hs-EcRi and 3 hs-FF1i lines; had variable effects
NotI
BamH1
XbaI
EcoR1
hsp70
Act 5C termination and polyA signals
pCaSpeR-hs-act P-element vector

11. Examples and Results contd.
2000, Kennerdell and Carthew: expressed an extended hairpin-loop
loop RNA from a transgene
UAS-geneRNAi X different Gal4 strains
Kirby et al. 2002: Sod2 RNAi; made inverted-repeat structure of Sod2
cDNA and cloned into EcoR1 site of pPUAST
generated 18 lines and analyzed 2 with “robust” expression
Leulier et al. 2002: dFADD RNAi; 500 bp long cDNA fragment
(nt positions of coding seq) was amplified by PCR and inserted
as an inverted repeat (IR) into a modified Bluescript vector, pSC1, which
possesses an IR formation site consisting of paired CpoI and SfiI
RE sites; IRs in a head-to-head orientation; IR fragment cut out and
cloned into pUAST ; used at least 2 independent lines for each expt.

12. Genomic-cDNA fusions
Kalidas and Smith, 2002: used genomic-cDNA fusion construct
regulated by its own promoter; pUAST system; 3 genes
Genomic fragment contains 5’UTR and intron + exon sequences
cDNA fragment contains corresponding exon sequences
two fragments are fused, head-to-head; apparently more stable
and easier to clone; splices to form mature
mRNA which then forms hairpin dsRNA
analyzed two independent lines for each construct
claim that suppression is greater and more uniform but no direct
comparison between methods

13. RNAi in Drosophila cell cultures
(Perrimon lab/RNAi Genome Project)
25 – 75 nM (0.2 ug) of
500 nt dsRNAs

14. Functional genomic analysis of phagocytosis using Drosophila S2 cells and RNAi
(Ramet et al. 2002)
generated random templates from an S2 cell-derived
cDNA library cloned into pcDNAI plasmids
Pooled two plasmids and generated S and AS RNA using
T7 and SP6
5 x 105 S2 cells were treated with 40ug dsRNA (20ug
per gene) for 60h and then analyzed
examined 1,000 random dsRNAs; found 34 genes with a
detectable effect on phagocytosis

15. RNAi in mammals RNAi used in early mouse embryos
BUT mammalian somatic cells exhibit a nonspecific response
to dsRNA
long dsRNA activates the RNA-dependent protein kinase
(PKR) pathway which phosphorylates EIF-2A and arrests
translation
synthetic siRNAs do not activate PKR (likely too small) and
can induce gene knockdowns

16. RNAi in mammals contd.
siRNA and a lipophilic agent used to transfect cells
Limiting factor is the transfectability of cells;
HeLa cells are the cell line of choice
effects of siRNAs are transient since mammalian cells
lack amplification mechanisms
most recent experimental approach is modelled on miRNAs
short hairpin RNAs (shRNAs) are expressed in vivo from
DNA vectors containing RNA pol III promoters (H1, U6)
shown to induce stable suppression in mammalian cells

17. Designing DNA silencing constructs
Hairpin
siRNA-in-trans
(From McManus and Sharp, 2002)

18. Use of DNA silencing constructs
(From McManus and Sharp, 2002)

19. Practical Considerations of siRNA design
select base-pairing region carefully to avoid chance
complementarity to an unrelated mRNA i.e. BLAST
N.B.: RNAi can tolerate siRNA:mRNA mismatches
of 1 – 2 bp
mRNA region optimal for siRNA targeting is not yet known;
suggested region is the first 50 – 100 nt of a cDNA
sequence, downstream of the translation start site
(want to avoid regulatory protein binding sites)
optimal design for shRNAs not yet known