Real-Time PCR:
Practical Issues and Troubleshooting
Mehmet Tevfik DORAK, MD PhD
Dept of Environmental & Occupational Health
Robert Stempel College of Public Health and Social Work
Florida International University
Miami, Florida
USA
MOBGAM, Istanbul, Turkey
June 3, 2011
(www)
(www)
(www)
(www)
linear view log view
Linear vs Log View
log view
linear view
Linear vs Log View
6.6 picogram template & CT > 40
Baseline Setting
Baseline Setting
Baseline Setting
Baseline Setting
Baseline Setting
Baseline Setting
Easier to do in log-view
Threshold Setting
Multiple Thresholds May be Necessary
Easier to do in log-view
Threshold Setting
Baseline and Threshold Setting
High Background
Standard Curve
Low DRn/RFU
• Imperfect assay design
• Incorrect quencher
• Bad template quality
(long probe, G at 5’ end, incorrect choice of reporter and quencher pair,
wrong primer-pair concentrations, probe has mismatches)
High efficiency?
Efficiency > 110% (slope <-3.1) is due to:
• Pipetting errors
• Wrong threshold setting (not in the log-linear phase)
• Primer-dimers in SYBR green assays
• Probe degradation in TaqMan assays
• High variability at low concentrations
• PCR inhibition by reverse transcriptase
High efficiency?
• Five CT difference rule
• SYBR green vs TaqMan
• Melting curve analysis
RNA quality
(www)
RNA quality: integrity
(www)
RNA quality: purity
RNA quantity
As long as you are staying within the
dynamic range of the assay established
by initial validation, use any amount, but
no less than 10 picogram and no more
than 1 microgram. Usually 1 to 100
nanogram is sufficient.
High template amount may increase
background.
(www)
Optimal template amount
(www)
Optimal template amount
Reaction volume
• Manufacturers recommend 50 microL
• 25 microL is fine
• 10 microL is generally fine
• 5 microL may be fine (for allelic discrimination) but not recommended
for qPCR
• Aim for 3-5 microL template volume in the reaction
• Aim for duplicates unless using so little template (around 10 picogram
or Ct > 35); then you need to use triplicates
cDNA synthesis
mRNA in formalin fixed samples may have lost their poly-A tails.
cDNA synthesis
Primer design
Primer design
Assay design
(www)
“ Primer concentrations may
result in severe changes in CT
values and should remain
constant in all experiments for
the same assay “
Pipetting
Albumin (ALB) gene dosage by real-time PCR
Laurendeau et al. Clin Chem 1999 (www)
Good efficiency,
good sensitivity and
good predictive
power.
Good Assay
ABI Understanding CT (www)
Good Assay
Good Assay
Optimized for:
• High signal intensity: High RFU / DRn
• Low background (noise)
• Low Ct values
• Maximum Ct = 40 for lowest template amount in dilution series
Controls
Controls
Controls
DDCT Assay
DDCT Assay: Validation
DDCT Assay: Validation
DDCT Assay
Dye selection
Normalizer selection
Normalizer selection
Stable endogenous controls do not
yield DCt values greater than that of
IPC and do not show much variation.
Normalizer selection
Normalizer selection
Sabek et al. Transplantation 2002 (www)
Normalizer selection
• Most abundant RNA: may need singleplex runs using
diluted samples or competimers (Ambion); not suitable
for rare target transcripts
• Forces separate baseline settings in some instruments
• Not mRNA
• Does not have 3’ poly-A tail
• Ct value should be smaller than 22 for valid results
18S as a Normalizer
GAPDH as a Normalizer
• A case of majority not always being
right!
• The most unstable and inconsistent
normalizer!
• Just don’t use it!
(www)
Weak Correlation Between mRNA
and Protein Levels in Eukaryotes
A total of 150 signature genes showed
significant changes at either the protein
and/or the mRNA level in two bovine bone
marrow derived cell lines. 113 signature genes
(76%) exhibited changes for mRNAs and their
cognate proteins in the same direction (1st
and 3rd quadrants), only 29 of them changed
significantly at both mRNA and protein levels
and were thus dubbed correlated genes (red).
In contrast, 67 genes showed significant
changes at the mRNA but not the protein level
(green), whereas 52 genes showed significant
changes at the protein but not the mRNA level
(blue). Another two genes showed opposite
expression patterns of mRNA and protein
(brown). The correlation coefficient between
mRNA and protein is 0.64 for the signature
genes and 0.59 for all the genes examined.
Tian, 2004 (www)
“ Any assessment of the biological consequences of variable mRNA
levels must include additional information regarding regulatory RNAs,
protein levels and protein activity ”
Interpretation
(www)
www.dorak.info
www.dorak.info/mobgam2011.pdf