Non-coding RNA

and RNA interference

F. Guesdon, 2006
 Objectives for this lecture
You will know which recent data indicates the existence
of previously unknown classes of non-coding RNA
molecules

You will know the characteristics and functions of 4

types of small non-coding RNAs :
• snRNAs: mRNA splicing
• siRNAs, shRNAs and miRNAs: gene silencing

1. You will understand how these new findings suggest

new models of gene regulation (control archgitecture)
 A short history of RNA

• 1890’s - second kind of nucleic acid, cytoplasmic

• 1920’s - different sugars for DNA and RNA
• 1958 - tRNA (Hoagland)
• 1960’s - mRNA
• 1982 - ribozyme, snRNAs?
• 1990s - New non-coding RNAs, RNA interference
• 2004 - Widespread transcription
Involvement of snRNAs in mRNA
splicing and surveillance
 Example of “classic” non-
coding RNA: snRNA
• Small nuclear ribonucleoprotein particles -
snRNPs, pronounced "snurps" - are involved in
splicing

• A snRNP consists of
– about 10 different proteins
– a small nuclear RNA, or snRNA (100-200 bases
long)
Types of snRNPs:
U1: binds 5’ splice site
U2: binds Branch point
U3: binds 3’ splice site
U4 & U6: function less well defined
Exon recognition

Spliceosome
 The snRNAs recognise
Splicing sites and branch point
by base-pairing
 snRNAs
• snRNAs are very important in splicing
• They mediate the sequence
specificity and the accuracy of the
splicing reaction
• They fold into secondary structures
(stem-loops) that are important for
catalysis
• In snRNPs it is the RNA part that is
catalytic, not the protein part
 Evidence of unsuspected transcriptional
activity and new classes of RNA
• 1990s: Expressed sequence tags (EST) data indicates presence of excess
number of poly-A+ RNA (about 100,000 unique transcripts)

• New interest in traditionally poorly studied RNA species:

– PolyA- RNA
– Nuclear polyA+ RNA

• Genome tiling arrays indicate widespread transcription of non-coding DNA

• ChIp assays detects transcription factor binding sites for new RNAs

• Bioinformatic analysis of genomic data based on RNA secondary structure

prediction finds RNA-coding genes

• Discovery of small non-coding RNAs involved in RNA interference

 Expressed sequence tags
UTR Coding Region UTR
DNA 5’ 3’

Transcription, splicing

mRNA 5’ AAAAA 3’

Oligo-dT purification, Reverse transcription, Cloning

cDNA 5’ AAAAA 3’
Sequencing

>IMAGE:275615 3', mRNA sequence

ESTs NNTCAAGTTTTATGATTTATTTAACTTGTGGAACAAAAATAA
5’EST ACCAGATTAACCACAACCATGCCTTACTTTATCAAATGTATA
3’EST
AGANGTAAATATGAATCTTATATGACAAAATGTTTCATTCAT
TATAACAAATTTCCAATAATCCTGTCAATNATATTTCTAAAT
TTTCCCCCAAATTCTAAGCAGAGTATGTAAATTGGAAGTTAA
CTTATGCACGCTTAACTATCTTAACAAGCTTTGAGTGCAAGA
GATTGANGAGTTCAAATCTGACCAAGATGTTGATGTTGGATA
AGAGAATTCTCTGCTCCCCACCTCTANGTTGCCAGCCCTC
 Too many ESTs !!!!
• Short sequences (~650bs)
• Contains only exonic sequences; UTRs or
coding regions
• Total number of unique human ESTs: >100,000

• Problem: Analysis of the human genome

sequence detected only about 25,000 genes
 Gene prediction

• Easy for prokaryotes – one gene, one protein

• More difficult for eukaryotes – one gene,
many proteins
• Very difficult for Human – short exons
separated by non-coding long introns
 Genome tiling arrays provide
unbiased transcription mapping
 In some cases, tiling array results confirmed
and extended software-based gene predictions
• Thalamus

(b) testes

(c) uterus
Kapranov et al.
Determination of transcriptional Activity of Chromosomes 21
and 22 using tiling arrays. Science, 2002
49 % of transcripts do not match known exons:
Transcripts of Unknown Function (TUFs)

Kampa D. Novel RNAs identified from an in-depth analysis of the transcriptome of human
chromosomes 21 and 22. Genome Res 2004
 Overall results: detection of unpredicted
transcripts - “Dark Matter in the Genome”

What could they be?

Tiling arrays for human
chromosomes 20 and 22: •Novel protein-coding genes

47% of positive probes outside exons •Novel non-coding genes

• 22% in introns •Antisense transcription

• 25% in intergenic regions •Alternative isoforms

•Biological ‘artifacts’
•False positives
Johnson et al. 2005 TRENDS in Genetics 21:93-102
Detection of new transcription factor
binding sites by Chromatin Immuno-
Precipitation (ChIP)

Cawley et al.

Unbiased mapping
of transcription
factor binding sites
along human
chromosomes
21 and 22 points to
widespread
regulation of
noncoding RNAs.

Cell 2004
Which proportion of the genome
gets transcribed?
• At least 60 % of the genome is transcribed

• Most of the transcripts are of low abundance in the

cytoplasm, so escaped traditional methods of detection

• What are the functions of the new non-coding RNA species?

• Two points of view:

– It is mostly useless transcriptional noise generated by junk DNA

– The new non-coding transcripts have important regulatory roles

• Who is right?
 Why were the new classes of
RNA not discovered earlier?
RNA
RNA
(prep
Markers (prep 2) Markers
1)

Relative Amount
of RNA in E. coli
rRNA 80%
tRNA 15%
mRNA 5%
Total RNA extracted from human cells (Dr C. Nile)
 Bulk RNA content is a poor
indicator of the transcriptional
potential of the genome
• RNA content at any one time reflects the
balance of transcription and RNA degradation
• RNAs species that are either transcribed very
actively or degraded very slowly dominate
siRNAs and RNA interference
 RNA interference (RNAi)
What is it?
• An RNA-mediated process by which specific genes
can be silenced
• Identified 10 years ago (1994)
• Understood recently (last 5 years)

Why is it important?
• Widely used experimental tool (investigation of
function)
• Fast and relatively inexpensive
• Therapeutic potential.
 1970s-1990s: Antisense RNA
technology
• Principle
• Transfect an expression vector encoding an antisense RNA.
• Annealing of antisense RNA with endogenous mRNA produces
double-stranded RNA (dsRNA).
• RNAse H destroys dsRNA (i.e. both mRNA and antisense).

• Pitfalls
• Limited efficiency: Only ~50% downregulation of target mRNA.
• Side-effects: Activates dsRNA-dependent kinase, leading to:
• Expression of antiviral proteins (e.g. interferons)
• General shut-down of protein synthesis.
 Co-suppression in Petunia

van der Krol AR, Mur LA, Beld M, Mol JN, Stuitje AR. Flavonoid genes in petunia: addition of a limited number of gene copies may lead to a
suppression of gene expression. Plant Cell. 1990 Apr;2(4):291-9. PMID: 2152117
Napoli C, Lemieux C, Jorgensen R. Introduction of a Chimeric Chalcone Synthase Gene into Petunia Results in Reversible Co-Suppression of
Homologous Genes in trans. Plant Cell. 1990 Apr;2(4):279-289. PMID: 12354959
 Discovery of RNAi
• Chance discovery in plants (1990)
• In experiments based on antisense RNA technique,
a vector encoding a sense construct was used as
control
• When the sense RNA was co-transfected together
with the antisense, the inhibition of protein
expression was much more efficient than when
antisense alone was used.
• Infecting a few cells generated a systemic effect.
 Mechanisms of RNAi
Core mechanism: Autonomous RNAi
– Found in all Eukaryotes

Extensions to the core mechanism:

• Maintenance and amplification
• Systemic RNAi
• Germline transmission
• These extended functions are restricted to certain organisms

Mechanisms identified mostly through work by Andrew Fire

and Craig Mello in C. elegans - Nobel Prize for Physiology an
Medicine 2006
 Core mechanism of RNAi: Gene
Silencing directed by ~22nt RNAs
dsRNA
DICER processing
~22nt siRNAs

recognition
RISC
target (Argonaute)

mRNA
degradation

RISC: RNA-Induced Silencing Complex (contains effector nuclease)

 Core mechanism:
Autonomous RNAi and RISC
• Autonomous RNAi is triggered by double
stranded RNA (dsRNA)
• The Rnase, DICER (member of the
Rnase III family), cleaves the dsRNA
into small fragments of 21 to 23 bp.
• Cleavage is sequence-specific,
produces dsRNAs with U overhangs
called short interfering RNAs (siRNAs).
• siRNAs bind to proteins of the
Argonaute family to form RNA-induced
silencing comlplexes (RISC)
• RISCs bind to single-stranded RNA
molecules homologous to the siRNA
(recognition by complementary base
pairing)
• RISC cleaves the target RNA
 Extensions to the core
mechanism
• Maintenance and amplification
• Systemic RNAi
• Germline transmission
 Maintenance and amplification:
RdRP
– In some organisms (plants, yeast, worms, insects), siRNAs
that anneals to mRNA do more than activating RISCs
– They also acts as primers for synthesis of antisense strand by
an RNA-dependent RNA polymerase (RdRP), thereby
generating long dsRNA structures
– The resulting dsRNA is then cleaved by Dicer to form a second
generation of siRNAs
– By ensuring continuous production of siRNAs, this amplification
process ensures the RNAi response is maintained until the
target mRNA has been completely eliminated
– Mammalian cells do not have RdRP activity and cannot
amplify the RNAi response
• Refs: Nature 409:363-366; Cell 107:297-307; Cell 107:465-476
Genomic immune system
hypothesis 2
 Systemic RNAi
• In some multicellular organisms (plants, worms), the
siRNAs are propagated from cell to cell
• This allows the RNAi response to spread to the entire
organism faster than the original dsRNA
• Systemic RNAi requires amplification
• Systemic RNAi requires a protein encoded by the Sid-1
gene, which imports siRNA from the extracellular medium
into cells
• There are no homologues of Sid-1in the genome of insects
and vertebrates - these animals cannot mount a systemic
RNAi response
• Recently (2003), transgenic Drosophila expressing Sid-1
have been made: they exhibit systemic RNAi responses
 Germline transmission
• In some experimental systems (plants), the RNAi
response can transmit from one generation to
the next
• The mechanism is still unclear. Two possible
mechanisms (speculative) are:
– Transmission of the siRNA response to gametes
(systemic RNAi)
– Transmission of the epigenetic modification
responsible for silencing of the target gene (DNA
methylation pattern caused by siRNA)
 siRNA has an antiviral function
• Autonomous RNAi can be thought of as an intracellular
adaptive immune system that fights back viral infectious
agents using complementary RNA sequences instead of
complementary protein surfaces
• Like the protein antibodies that circulate in the blood and
lymphatic systems of vertebrates, systemic RNAi provides
a cellular defence in advance of the viral infection front in
plants, and probably also in worms

Adapted from Weiner, Molecular Cell, 2004

shRNAs and miRNAs
 microRNAs (miRNAs)

• Small endogenous RNAs (19 to 23 nt, typically

22 nt)

• Gene-encoded (at least 220 miRNA genes = 0.8

% of total genes in H. sapiens.

• They regulate the expression of other genes

• Most abundant class of regulatory genes

 miRNA expression
In the nucleus
• miRNA genes encode long primary transcripts (pre-
miRNA) that may contain multiple hairpin miRNA
precursors
• Processing of pre-miRNAs by the RNAse III, Drosha,
yields hairpin precursors of 60-80 nucleotides (shRNAs)

In the cytoplasm
• The miRNAs are excised from the shRNAs by DICER
 miRNA / mRNA duplexes
• miRNAs are usually not perfect matches to their targets
• Nucleotides 2-8 of miRNAs (core heptamer or seed) are the
most important for target recognition.
• Outside the nt2-8 region, non-canonical base-pairing (G:U)
or mismatches are allowed
• Annealing to target can lead to formation of loops, where the
5’ and 3’ ends of the miRNA anneal to different regions of
the mRNA
• Mismatch tolerance allows miRNAs to act on several
mRNA targets
miRNAs are produced and recognise
target mRNAs by by similar
mechanisms as siRNAs
In both cases, the mechanisms
involve:

 Formation of small RNAs through

cleavage of precursor dsRNA
structures by DICER
 The small RNAs then guide
protein complexes to the target
RNA by means of specific base-
pairing
 siRNA and miRNA: two arms
of the same system
• Both systems involve processing of dsRNA precursors
by specific RNAse III enzymes
• The excised dsRNAs associate with an Argonaute family
member to form an RNA-induced silencing complex
(RISC)
• The RISC then targets almost perfect complementary
RNAs for either degradation or control of translation.
• Hypothesis: Mismatches in the duplex structure prevent
the RISC complex to function as a cleavage enzyme, but
allows it to prevent translation.
 Functions of miRNAs

• miRNAs hybridise with homologous sequences in

3’ UTRs of target mRNAs
• The resulting dsRNA structures specify fate of
mRNA:
– Repression
– Cleavage
– Translation
 Interactions of miRNAs with
mRNAs
• Class 1 targets : perfect match to miRNA “seed”

Myotrophin 3’ UTR
5’ - - - U U-U A ------- A U - - - 3’
U-G-U-G G-C G-A-A-C-A-A-A
            

G-C-G-C C-G C - U - U - G - U - U - U - 5’
U U G G
G miR-375
C-U-U
A

From Rajewsky,
Nature Genetics
Seed = first 7 nucleotides 38:S8-S13, 2006
 Interactions of miRNAs with
mRNAs (2)
• Class 2 targets : Imperfect match to “seed”, high
complementarity to remainder of miRNA sequence

Lin-41 3’ UTR
5’ - - - U G-U-U A A - - - 3’

U-U-A-G-G-U-U-G-U-A G-U-A-G U-G-A-G

                 

G-A-U-C-C-A-A-C-A-U C - A - U - C - - A - C - U - C - 5’
U A-U Let-7
U

From Rajewsky,
Nature Genetics
38:S8-S13, 2006
 Identification of miRNA targets
• Computer predictions: still uncertain, especially
for class 2 interactions

• Experimental:
– Searches for anti-correlations between miRNA and
putative target in microarray data
– Micro-array-based screens for miRNA-dependent
changes in gene expression
– Overexpression introduces bias
– Mutation / deletion of endogenous miRNA genes:
only done in a few cases
Database of non-coding RNAs

Rfam: an RNA family database.

Nucleic Acids Res. 2003 Jan 1; 31(1): 439-41
http://www.sanger.ac.uk/Software/Rfam/

The microRNA Registry.

Nucleic Acids Res. 2004 Jan 1; 32(1): D109-11
http://www.sanger.ac.uk/Software/Rfam/mirna/
 miRNAs and siRNAs use different
mechanims of post-transcriptional
silencing

- Imperfect match to template (miRNA) 2 - Perfect match (siRNA)

 Blocks translation machinery  Degradation of mRNA by DICER

 NO mRNA degradation  Amplification
 The same miRNA can silence several
mRNA with slightly different sequences
 siRNA-mediated
transcriptional silencing
 siRNA directed to promoters
induce transcriptional silencing
• Mechanism well documented in plants and yeast

• Involves histone methylation and de-acetylation

• In yeast and mammalian cells: H3K9
• In plants: methylation of H3K27 (Histone 3 Lys27)

• In mammalian cells, transcriptional silencing has

been studied on 3 genes:
• EF1A (Science 305:1289, 2004)
• E-cadherin (Nature 431:211, 2004 - Retracted !!)
• RASSF1A Tumor suppressor (Mol. Ther. 12:179, 2005)
 The RITS Complex
Experiments carried out to determine the role of Chp1 in
siRNA mediated heterochromatin led to the identification
of the RITS complex (RNA induced transcriptional
silencing)

• Ago1 (Argonaute protein)

• Chp1 (required for Swi6
recruiment)
• Tas3 (required for H3mK9)

Nature 431, 364 - 370

Both H3mK9 and siRNA association are required for centromeric localization
 RNA interference uses two
types of effector complexes
1. RITS: RNA-Induced Transcriptional Silencing
complex (nuclear)
• Ago1 (Argonaute protein)
• Chp1 (required for Swi6 recruiment)
• Swi6 (Histone-binding)
• Tas3 (required for H3mK9)

2. RISC: RNA-Induced Silencing Complex

(Cytoplasmic)
 RNA interference can act at
three levels
• Transcriptional silencing (chromatin
condensation) - Through RITS complexes
• RISC- mediated Post-transcriptional
silencing (RNA degradation)
• RISC-mediated Translational silencing
(translation block)
RNA interference and the
control architecture of the
genome
Proven physiological functions of
miRNAs (so far)
Involved in regulation of:

• Timing of larval development (C. elegans)

• Cell proliferation
• Cell death
• Fat metabolism (insects)
• Leaf and flower development (plants)
• Antiviral defense
• Control of chromatin modification (epigenetic silencing)
 Defense against transposons?
• In C.elegans and Drosophila mutation of RNAi
components  activation of transposons
 Is RNAi a genomic immune system?

– The vertebrate adaptive immune system:

• Distinquish self from non-self
• Amplify a response
• Kill the intruder
RNAi is involved in the control of
selfish genetic elements
• Selfish elements replicate through either DNA or
RNA intermediates (retro-transposons)
• They generate dsRNA intermediates during
either replication or transcription
• The RNAi system silences centromeric DNA
(rich in retro-elements) in flies and yeast
 Genomic immune system
hypothesis 3
• Self/non-self discrimination (generation of
dsRNA)
– Multicopy transposons:
• read through from flanking promotors create
complementary strands to form dsRNA
– Some transposons have terminal inverted
repeats  hairpin
 Gene expression:
Control Architecture (before 1999)
Genome

Regulation by
proteins
Transcriptome

Proteome
Graphical representation of the transcriptome

From Mattick and

Makunin, Human Mol.
Genet. 15:R17-R29,
2006
 Gene expression:
Control Architecture (after 2002)
Imprinting – methylation
Genome Splicing

RNAi
Regulation
Regulation by
by proteins Transcriptome RNA

Proteome Ribozymes
 Complexity of the transcriptional
lansdscape in mammals

Non-coding exonic A: Antisense transcript (overlapping exons)

Coding exonic B: Nested transcripts on both strands

snoRNA C: Antisense transcripts (interlacing exons)

miRNA D: Retained intron

From Mattick and Makunin, Human Mol. Genet. 15:R17-R29, 2006

 Non-coding DNA correlates with
developmental complexity

From Mattick et al., Nature Rev. Genet. 5:316 (2004)

 Prediction of miRNA targets

Ref: Lewis et al., Cell 115:787-798 (2003)

• miRNA target lists are enriched in genes

encoding transcription factors (in mammals) or
development regulators (in plants)

• This strongly supports the view that miRNAs

have evolved to sustain complex regulatory
networks needed in higher organisms
Why use RNA in regulation?

• Cost
– Smaller genomic space of miRNA vs protein
genes
– Bypasses metabolic cost of making mRNA and
protein
• Speed of response
– RNA produced and active very quickly
• Evolutionary agility
– Smaller ”mutation space” to explore on a 22 nt
RNA sequence than on a protein-coding gene.
 So why are they regulatory
proteins?
• Regulation by proteins widespread in simple
organisms (bacteria)
• RNAi system perahps evolved initially as
defense against viruses, not regulation system
• Proteins may allow more specific or stable
interactions
• Proteins may be better at sensing environment
– responding to signal transduction etc...
• Perhaps also better at modulating output?
 Types of RNA
RNA

ncRNA
mRNA Non-coding RNA. Transcribed RNA with a structural,
functional or catalytic role

tRNA snRNA snoRNA Other

rRNA Transfer RNA Small nuclear RNA Small nucleolar RNA
miRNA Including large RNA
Ribosomal RNA Micro RNA with roles in
Interface between -Incl. RNA that Found in nucleolus,
Participate in Small RNA involved chromotin structure and
mRNA & form part of the involved in modification
protein synthesis regulation of expression imprinting
amino acids spliceosome of rRNA

stRNA siRNA
Small temporal RNA. Small interfering RNA
RNA with a role in Active molecules in
developmental timing RNA interference

Any questions?
 Main take-home points
You must know the following:
What are the meanings of the abbreviations
ncRNA, snRNA, shRNA, siRNA, miRNA and
RNAi
What are spliceosomes, DICER, RICS, RdRP
and RITS
In what processes they are involved
What role do they play in these processes
What distinguish autonomous RNAi from
systemic RNAi
Applications of RNAi:
tomorrow
For Wednesday