Transcript Introduction to quantitative real

Introduction to quantitative realtime PCR
Veryan Codd
[email protected]
Overview
•
•
•
•
•
What is real-time PCR and what is it used for?
Real time platforms available in CVS
Basic Principles
Experimental design, Controls and QC
Quantification methods
What is real time PCR
• Ability to monitor PCR reaction in real time
• Accurate quantification of the amount of starting
material (template) added to a reaction
• Increase dynamic range of detection over end point
methods
Why do we use real-time PCR
• Quantification of gene expression levels
• Quantification of DNA (Telomere assays,
mtDNA content, copy number)
• Virus titre
• Can be used as a quantitative step as part of
another technique e.g. ChIP and DNA
Methylation
Our facilities – ViiA7
•
•
•
•
Life Tech (Applied Biosystems)
96, 384 well plates and Taqman array cards
Upgraded 7900
Advanced optics – 6 excitation and 6 emission
wavelengths – 21 dye combinations
• Value ~75K
Our facilities – Rotorgene-Q
• Qiagen
• Unique rotary system – no well-to-well thermal
variation – improved accuracy
• 0.2ml tubes (36), 0.1ml strip tube (72), rotordisc
(72 or 100)
• Robotic set-up required but available for 100 well
disc
• Additional quant methods in software
• 2-plex
– Green (SYBR green, FAM) and Yellow (VIC,
JOE)channels
Our facilities – Rotorgene-Q
• Value ~20K
Basic Principles
• PCR with a fluorescent dye that can be
detected
• PCR is theoretically exponential – product
doubles with each cycle
• The more starting material the faster the
reaction will occur (lower cycle number)
Experimental design - reagents
• Primers
–
–
–
–
Specific to target
18 – 24bp
Produce product 50 – 150 bp in length (200bp does work)
Product ~50% GC and avoid GC stretches and repetitive
sequences
– Check for primer dimers especially for SYBR green assays
– F and R primers annealing temp within 5˚C
– In separate exons or spanning exon boundaries
• Pre-designed assay available (taqman probes, sybr based
assays)
• Array cards / expression sets
Experimental design – probe or SYBR?
SYBR green
• Dye binds minor groove of all
DNA
• Fluorescence much stronger
when bound
• More product more dye bound
• Cheap
• Not specific
• At early stage check product
on gel and perform
dissociation analysis (melt
curve)
Probe
• Dye and quencher either
end of probe
• Nuclease activity of
polymerase removes dye
and quencher during PCR
• More product produced,
more fluorescence
• No melt curve
• Allows multiplexing
Experimental design- singleplex or
multiplex
Singleplex (simplex)
• One gene per PCR
• SYBR green or Probe – more
flexible
• More variation
• Usually simpler to set up
• Uses more enzyme mix but
can SYBR green with cheap
primers instead of expensive
probes
• More template?
Multiplex
• Multiple genes per PCR
• Probe based – 2 dyes (VIC
FAM)
• Minimises pipetting error
• Competition between
reactions can be problematic
• Cheaper?
• Higher throughput – 2
measurements in one
• Less template?
Experimental design - reagents
• Enzyme or enzyme master mix governed by
primer probe choice and by choice of platform
(e.g. ViiA7 or rotorgene)
• 2x mastermixes widely available
– Consistent
– Pre-optimized (MgCl2 etc)
Experimental design - Template
• For gene expression analysis need to reverse
transcribe mRNA into cDNA
• Good quality RNA
• Accurately quantified
• Variety of RT enzymes commercially available
Reverse transcription - Priming
• Three (or four) basic choices
Oligo-dT priming
• Favourite choice of many as specific to mRNA
• Total RNA
– 80% rRNA
– 15% tRNA
– 5% mRNA
• Sensitive to secondary structure – incomplete
reverse transcription
• Design primers at 3’ end of transcript
• Don’t use 18s as a control (it’s rRNA!)
Random Priming
• Uses random hexamer primers
• Bind throughout RNA molecule
– Secondary structure less of a concern
– Less sensitive to RNA degradation
– Increase yield of cDNA
• Amplify ALL RNA – mRNA, tRNA and rRNA
Sequence specific priming
• Only gene of interest is reverse transcribed
• Limits use of cDNA, used up RNA stocks
• Only used in one-step reverse transcriptionqPCR (one step qRT-PCR)
Option no 4
• Mix of oligo-dT and random primers
• Best of both worlds
Good laboratory practice
• qPCR can be a sensitive technique if performed correctly and
with care
• Keep everything clean
• Aliquot reagents to prevent freeze/thaw
• Accurate pipetting
–
–
–
–
–
–
Make master mix of everything except template then aliquot
Mix everything gently but thoroughly
2x master mixes are viscous
Small volume pipetting less accurate
22ul + 3ul or 23ul + 2ul works well (MM + Template)
Don’t use second stop on pipette to avoid “spraying”
• Perform triplicate PCRs for sample
Contamination
• Big problem for qPCR
• Most common cause of contamination comes from product of
previous PCRs
• Don’t leave PCRs in machine indefinitely after completion
• If running product on get use separate pipettes to those you
use for PCR set-up
• Use PCR dedicated Hood (genomics lab) to set up PCR MM’s
– Keep primers and other reagents free from contamination
– DNA/cDNA MUST NOT BE ADDED IN THE HOOD – pipette
contamination
Getting started
• For pre-designed assays have a test run to
check for good amplification
• If designing own assay
– Test to check product of correct size is formed
– Optimise annealing temp (gradient PCR)
– Test in real-time PCR that amplification looks
reasonable
– Titrate primer concs if/as necessary
Normalisation
• Housekeeping gene(s)
– Accounts for pipetting error and differences in RT efficiencies between
samples
– Constantly expressed at a stable level
– Put the same starting RNA in all RT-reactions when samples are to be
compared
– IT IS NOT THERE TO ACCOUNT FOR MASSIVE DIFFERENCES IN
STARTING MATERIAL!
• Calibrator sample
– Sample run on every assay
• T0 / Untreated
– Compare everything to this sample
Controls
• NTC (No template control)
– PCR contamination detection
– Primer dimers
• no-RT control
– Detection of genomic contamination and/or nonspecific product
– Especially important in transient over-expression
experiments to check for carry over of plasmid DNA
Dissociation or Melt curve analysis
• To check for single specific product in SYBR green based
assays
• Compare to product on gel in early stages of set-up
• Detection of primer dimers and non-specific products
Melt curve Example 1
Melt curve example 2
NTC
Primer
dimers
Specific
product
Melt curve example 3
Specific
product
Non-specific
product
In this case NTC’s showed no product, so no primer dimers, large nonspecific product observed on gel
Standard curves
• Do I need a standard curve???
– YES!
• “My assay is predesigned and guaranteed to
be efficient therefore I don’t need to waste my
time doing a Standard curve”
– Not true! But very common!!
What is a standard curve and why is it
necessary?
• Real time PCR conducted across a series of
serially diluted template/samples
• Tell you what the dynamic (linear) range of the
assay is
• Allows calculation of efficiency
Cycle number Ct or take-off
Log10 template concentration
Cycle number Ct or take-off
Log10 template concentration
Cycle number Ct or take-off
25
20
y = -3.2241x + 25.595
R² = 0.9813
15
10
5
0
0
0.5
1
1.5
2
Log10 template concentration
2.5
14
12
10
8
6
4
2
0
0
0.5
1
1.5
2
2.5
3
Linear range 7 -10 cycles
Therefore experimental samples with Ct’s outside of this should not be
used for quantification
Set the concentration of template in middle of linear range
If input concentrations are highly variable then more likely to get samples
falling outside of LR that cannot be accurately quantified
Efficiency
•
•
•
•
Efficiency = 10(-1/gradient)-1
Gradient of -3.32 = 100% efficient
90-110% considered ok (-3.1 to -3.6)
For some, most commonly used, quantification
methods it is important to have similar efficiencies
between GOI and housekeeping genes
Quantification methods- Absolute
Quantification
• Samples calculated against a standard curve of known
concentration
• Allows quantification as copy number
– Can be used when one condition contains no expression
• Takes some variability into account
• Requires more reagents
• Needs a known standard
– Plasmid DNA containing product – no RT step so efficiencies
may be different
– In-vitro transcription to generate specific mRNA – RT, costly
Quantification methods- ΔCt, ΔΔCt
•
•
•
•
•
•
•
•
•
•
Comparative quantification
Threshold cycle number – Ct
Slightly subjective
Calibrator – i.e T0 or untreated
Normalized to housekeeper (HK)
ΔCt = CtGOI-CtHK
Δ ΔCt = ΔCtsample- Δctcalibrator
Fold change = 2- Δ ΔCt
Assumes efficiency 100% for both GOI and HK
Can substitute 2 for your calc efficiency but must be the
same for GOI and HK
Example
Efficiency is 88% GOI (red)
Efficiency is 1.07% Housekeeper
Calculate average DeltaCt for each dilution:
200
-4.8
100
-4.65
50
-4.5
25
-4.35
12.5
-4.2
6.25
-4.05
3.125
-3.9
1.56
-3.75
One end of range 1.05 cycles different to
other end – equivalent to a two fold
difference when it should be the same
26
24
y = -3.1552x + 25.061
R² = 1
22
20
18
16
14
y = -3.6534x + 21.407
R² = 1
12
10
0
0.5
1
1.5
2
2.5
Comparative quantification - Qiagen
• Method on rotorgene software
• Calculates efficiency for each individual sample based on rate
of fluorescence change and mean efficiency across run
• Calculates a “take off” value – start of exponential phase, less
subjective than setting threshold for Ct
• Set calibrator sample, T0 or untreated etc
• Relative amount calculated against calibrator
• Relative conc = MAE(calibrator takeoff-sample takeoff)
• Effectively a ΔCt but with real efficiency
• Then need to manually normalise to HK or a Δ ΔCt and fold
change calculation
Example
GOI efficiency = 96%
HK efficiency = 104%
Comparative quant each dilution
for GOI and HK separately using
one concentration as calibrator
24
y = -3.2241x + 25.595
R² = 0.9813
22
20
18
16
14
y = -3.4327x + 21.765
R² = 0.9906
12
10
0
200
100
50
25
12.5
GOI/HK
0.999
1.002
1.006
1.021
0.968
Δ ΔRc
0.996
1.000
1.001
1.003
0.997
0.5
1
1.5
Δ Δ Ct (using Ct rather than TO)
1.000
0.749
0.620
0.488
0.357
2
2.5
Accuracy
• Singleplex reaction more prone to variation in pipetting
accuracy
• Replica error
– How consistent is Ct/take off value between triplicate measurements?
– A 1 cycle difference across the triplicate is a 2-fold difference within
the same sample!
– Rotorgene 0.2 or 0.3 cycles easily achievable
– ViiA7 and other block based ~0.5
• Rotorgenes better at accurate quantification of small
differences
– No thermal variation
– Take off calculation less prone to subjectivity
Further resources
• Real-time PCR handbook
• LifeTech
• Molecular Cloning, Green and Sambrook
• MIQE guidlines